diff options
| author | Georg Brandl <georg@python.org> | 2020-09-08 20:20:19 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2020-09-08 20:20:19 +0200 |
| commit | 9f5672672bd61f7149d2a165b49f0617a1a9fe8e (patch) | |
| tree | ad3d62c5c167c6a75edf67a88c20341c77566c7e | |
| parent | d9a9e9ee40eb9815ecc3d9ec9d6f5e57499009d2 (diff) | |
| download | pygments-git-9f5672672bd61f7149d2a165b49f0617a1a9fe8e.tar.gz | |
all: remove "u" string prefix (#1536)
* all: remove "u" string prefix
* util: remove unirange
Since Python 3.3, all builds are wide unicode compatible.
* unistring: remove support for narrow-unicode builds
which stopped being relevant with Python 3.3
76 files changed, 3632 insertions, 3745 deletions
diff --git a/doc/conf.py b/doc/conf.py index 9e896890..1a08e0d6 100644 --- a/doc/conf.py +++ b/doc/conf.py @@ -34,8 +34,8 @@ source_suffix = '.rst' master_doc = 'index' # General information about the project. -project = u'Pygments' -copyright = u'2006-2020, Georg Brandl and Pygments contributors' +project = 'Pygments' +copyright = '2006-2020, Georg Brandl and Pygments contributors' # The version info for the project you're documenting, acts as replacement for # |version| and |release|, also used in various other places throughout the @@ -180,8 +180,8 @@ latex_elements = { # Grouping the document tree into LaTeX files. List of tuples # (source start file, target name, title, author, documentclass [howto/manual]). latex_documents = [ - ('docs/index', 'Pygments.tex', u'Pygments Documentation', - u'Pygments authors', 'manual'), + ('docs/index', 'Pygments.tex', 'Pygments Documentation', + 'Pygments authors', 'manual'), ] # The name of an image file (relative to this directory) to place at the top of @@ -210,8 +210,8 @@ latex_documents = [ # One entry per manual page. List of tuples # (source start file, name, description, authors, manual section). man_pages = [ - ('docs/index', 'pygments', u'Pygments Documentation', - [u'Pygments authors'], 1) + ('docs/index', 'pygments', 'Pygments Documentation', + ['Pygments authors'], 1) ] # If true, show URL addresses after external links. diff --git a/external/rst-directive.py b/external/rst-directive.py index a381ce43..937c3765 100644 --- a/external/rst-directive.py +++ b/external/rst-directive.py @@ -76,7 +76,7 @@ class Pygments(Directive): lexer = TextLexer() # take an arbitrary option if more than one is given formatter = self.options and VARIANTS[list(self.options)[0]] or DEFAULT - parsed = highlight(u'\n'.join(self.content), lexer, formatter) + parsed = highlight('\n'.join(self.content), lexer, formatter) return [nodes.raw('', parsed, format='html')] directives.register_directive('sourcecode', Pygments) diff --git a/pygments/filters/__init__.py b/pygments/filters/__init__.py index b57c64d3..37f87722 100644 --- a/pygments/filters/__init__.py +++ b/pygments/filters/__init__.py @@ -107,563 +107,563 @@ class SymbolFilter(Filter): """ latex_symbols = { - '\\alpha' : u'\U000003b1', - '\\beta' : u'\U000003b2', - '\\gamma' : u'\U000003b3', - '\\delta' : u'\U000003b4', - '\\varepsilon' : u'\U000003b5', - '\\zeta' : u'\U000003b6', - '\\eta' : u'\U000003b7', - '\\vartheta' : u'\U000003b8', - '\\iota' : u'\U000003b9', - '\\kappa' : u'\U000003ba', - '\\lambda' : u'\U000003bb', - '\\mu' : u'\U000003bc', - '\\nu' : u'\U000003bd', - '\\xi' : u'\U000003be', - '\\pi' : u'\U000003c0', - '\\varrho' : u'\U000003c1', - '\\sigma' : u'\U000003c3', - '\\tau' : u'\U000003c4', - '\\upsilon' : u'\U000003c5', - '\\varphi' : u'\U000003c6', - '\\chi' : u'\U000003c7', - '\\psi' : u'\U000003c8', - '\\omega' : u'\U000003c9', - '\\Gamma' : u'\U00000393', - '\\Delta' : u'\U00000394', - '\\Theta' : u'\U00000398', - '\\Lambda' : u'\U0000039b', - '\\Xi' : u'\U0000039e', - '\\Pi' : u'\U000003a0', - '\\Sigma' : u'\U000003a3', - '\\Upsilon' : u'\U000003a5', - '\\Phi' : u'\U000003a6', - '\\Psi' : u'\U000003a8', - '\\Omega' : u'\U000003a9', - '\\leftarrow' : u'\U00002190', - '\\longleftarrow' : u'\U000027f5', - '\\rightarrow' : u'\U00002192', - '\\longrightarrow' : u'\U000027f6', - '\\Leftarrow' : u'\U000021d0', - '\\Longleftarrow' : u'\U000027f8', - '\\Rightarrow' : u'\U000021d2', - '\\Longrightarrow' : u'\U000027f9', - '\\leftrightarrow' : u'\U00002194', - '\\longleftrightarrow' : u'\U000027f7', - '\\Leftrightarrow' : u'\U000021d4', - '\\Longleftrightarrow' : u'\U000027fa', - '\\mapsto' : u'\U000021a6', - '\\longmapsto' : u'\U000027fc', - '\\relbar' : u'\U00002500', - '\\Relbar' : u'\U00002550', - '\\hookleftarrow' : u'\U000021a9', - '\\hookrightarrow' : u'\U000021aa', - '\\leftharpoondown' : u'\U000021bd', - '\\rightharpoondown' : u'\U000021c1', - '\\leftharpoonup' : u'\U000021bc', - '\\rightharpoonup' : u'\U000021c0', - '\\rightleftharpoons' : u'\U000021cc', - '\\leadsto' : u'\U0000219d', - '\\downharpoonleft' : u'\U000021c3', - '\\downharpoonright' : u'\U000021c2', - '\\upharpoonleft' : u'\U000021bf', - '\\upharpoonright' : u'\U000021be', - '\\restriction' : u'\U000021be', - '\\uparrow' : u'\U00002191', - '\\Uparrow' : u'\U000021d1', - '\\downarrow' : u'\U00002193', - '\\Downarrow' : u'\U000021d3', - '\\updownarrow' : u'\U00002195', - '\\Updownarrow' : u'\U000021d5', - '\\langle' : u'\U000027e8', - '\\rangle' : u'\U000027e9', - '\\lceil' : u'\U00002308', - '\\rceil' : u'\U00002309', - '\\lfloor' : u'\U0000230a', - '\\rfloor' : u'\U0000230b', - '\\flqq' : u'\U000000ab', - '\\frqq' : u'\U000000bb', - '\\bot' : u'\U000022a5', - '\\top' : u'\U000022a4', - '\\wedge' : u'\U00002227', - '\\bigwedge' : u'\U000022c0', - '\\vee' : u'\U00002228', - '\\bigvee' : u'\U000022c1', - '\\forall' : u'\U00002200', - '\\exists' : u'\U00002203', - '\\nexists' : u'\U00002204', - '\\neg' : u'\U000000ac', - '\\Box' : u'\U000025a1', - '\\Diamond' : u'\U000025c7', - '\\vdash' : u'\U000022a2', - '\\models' : u'\U000022a8', - '\\dashv' : u'\U000022a3', - '\\surd' : u'\U0000221a', - '\\le' : u'\U00002264', - '\\ge' : u'\U00002265', - '\\ll' : u'\U0000226a', - '\\gg' : u'\U0000226b', - '\\lesssim' : u'\U00002272', - '\\gtrsim' : u'\U00002273', - '\\lessapprox' : u'\U00002a85', - '\\gtrapprox' : u'\U00002a86', - '\\in' : u'\U00002208', - '\\notin' : u'\U00002209', - '\\subset' : u'\U00002282', - '\\supset' : u'\U00002283', - '\\subseteq' : u'\U00002286', - '\\supseteq' : u'\U00002287', - '\\sqsubset' : u'\U0000228f', - '\\sqsupset' : u'\U00002290', - '\\sqsubseteq' : u'\U00002291', - '\\sqsupseteq' : u'\U00002292', - '\\cap' : u'\U00002229', - '\\bigcap' : u'\U000022c2', - '\\cup' : u'\U0000222a', - '\\bigcup' : u'\U000022c3', - '\\sqcup' : u'\U00002294', - '\\bigsqcup' : u'\U00002a06', - '\\sqcap' : u'\U00002293', - '\\Bigsqcap' : u'\U00002a05', - '\\setminus' : u'\U00002216', - '\\propto' : u'\U0000221d', - '\\uplus' : u'\U0000228e', - '\\bigplus' : u'\U00002a04', - '\\sim' : u'\U0000223c', - '\\doteq' : u'\U00002250', - '\\simeq' : u'\U00002243', - '\\approx' : u'\U00002248', - '\\asymp' : u'\U0000224d', - '\\cong' : u'\U00002245', - '\\equiv' : u'\U00002261', - '\\Join' : u'\U000022c8', - '\\bowtie' : u'\U00002a1d', - '\\prec' : u'\U0000227a', - '\\succ' : u'\U0000227b', - '\\preceq' : u'\U0000227c', - '\\succeq' : u'\U0000227d', - '\\parallel' : u'\U00002225', - '\\mid' : u'\U000000a6', - '\\pm' : u'\U000000b1', - '\\mp' : u'\U00002213', - '\\times' : u'\U000000d7', - '\\div' : u'\U000000f7', - '\\cdot' : u'\U000022c5', - '\\star' : u'\U000022c6', - '\\circ' : u'\U00002218', - '\\dagger' : u'\U00002020', - '\\ddagger' : u'\U00002021', - '\\lhd' : u'\U000022b2', - '\\rhd' : u'\U000022b3', - '\\unlhd' : u'\U000022b4', - '\\unrhd' : u'\U000022b5', - '\\triangleleft' : u'\U000025c3', - '\\triangleright' : u'\U000025b9', - '\\triangle' : u'\U000025b3', - '\\triangleq' : u'\U0000225c', - '\\oplus' : u'\U00002295', - '\\bigoplus' : u'\U00002a01', - '\\otimes' : u'\U00002297', - '\\bigotimes' : u'\U00002a02', - '\\odot' : u'\U00002299', - '\\bigodot' : u'\U00002a00', - '\\ominus' : u'\U00002296', - '\\oslash' : u'\U00002298', - '\\dots' : u'\U00002026', - '\\cdots' : u'\U000022ef', - '\\sum' : u'\U00002211', - '\\prod' : u'\U0000220f', - '\\coprod' : u'\U00002210', - '\\infty' : u'\U0000221e', - '\\int' : u'\U0000222b', - '\\oint' : u'\U0000222e', - '\\clubsuit' : u'\U00002663', - '\\diamondsuit' : u'\U00002662', - '\\heartsuit' : u'\U00002661', - '\\spadesuit' : u'\U00002660', - '\\aleph' : u'\U00002135', - '\\emptyset' : u'\U00002205', - '\\nabla' : u'\U00002207', - '\\partial' : u'\U00002202', - '\\flat' : u'\U0000266d', - '\\natural' : u'\U0000266e', - '\\sharp' : u'\U0000266f', - '\\angle' : u'\U00002220', - '\\copyright' : u'\U000000a9', - '\\textregistered' : u'\U000000ae', - '\\textonequarter' : u'\U000000bc', - '\\textonehalf' : u'\U000000bd', - '\\textthreequarters' : u'\U000000be', - '\\textordfeminine' : u'\U000000aa', - '\\textordmasculine' : u'\U000000ba', - '\\euro' : u'\U000020ac', - '\\pounds' : u'\U000000a3', - '\\yen' : u'\U000000a5', - '\\textcent' : u'\U000000a2', - '\\textcurrency' : u'\U000000a4', - '\\textdegree' : u'\U000000b0', + '\\alpha' : '\U000003b1', + '\\beta' : '\U000003b2', + '\\gamma' : '\U000003b3', + '\\delta' : '\U000003b4', + '\\varepsilon' : '\U000003b5', + '\\zeta' : '\U000003b6', + '\\eta' : '\U000003b7', + '\\vartheta' : '\U000003b8', + '\\iota' : '\U000003b9', + '\\kappa' : '\U000003ba', + '\\lambda' : '\U000003bb', + '\\mu' : '\U000003bc', + '\\nu' : '\U000003bd', + '\\xi' : '\U000003be', + '\\pi' : '\U000003c0', + '\\varrho' : '\U000003c1', + '\\sigma' : '\U000003c3', + '\\tau' : '\U000003c4', + '\\upsilon' : '\U000003c5', + '\\varphi' : '\U000003c6', + '\\chi' : '\U000003c7', + '\\psi' : '\U000003c8', + '\\omega' : '\U000003c9', + '\\Gamma' : '\U00000393', + '\\Delta' : '\U00000394', + '\\Theta' : '\U00000398', + '\\Lambda' : '\U0000039b', + '\\Xi' : '\U0000039e', + '\\Pi' : '\U000003a0', + '\\Sigma' : '\U000003a3', + '\\Upsilon' : '\U000003a5', + '\\Phi' : '\U000003a6', + '\\Psi' : '\U000003a8', + '\\Omega' : '\U000003a9', + '\\leftarrow' : '\U00002190', + '\\longleftarrow' : '\U000027f5', + '\\rightarrow' : '\U00002192', + '\\longrightarrow' : '\U000027f6', + '\\Leftarrow' : '\U000021d0', + '\\Longleftarrow' : '\U000027f8', + '\\Rightarrow' : '\U000021d2', + '\\Longrightarrow' : '\U000027f9', + '\\leftrightarrow' : '\U00002194', + '\\longleftrightarrow' : '\U000027f7', + '\\Leftrightarrow' : '\U000021d4', + '\\Longleftrightarrow' : '\U000027fa', + '\\mapsto' : '\U000021a6', + '\\longmapsto' : '\U000027fc', + '\\relbar' : '\U00002500', + '\\Relbar' : '\U00002550', + '\\hookleftarrow' : '\U000021a9', + '\\hookrightarrow' : '\U000021aa', + '\\leftharpoondown' : '\U000021bd', + '\\rightharpoondown' : '\U000021c1', + '\\leftharpoonup' : '\U000021bc', + '\\rightharpoonup' : '\U000021c0', + '\\rightleftharpoons' : '\U000021cc', + '\\leadsto' : '\U0000219d', + '\\downharpoonleft' : '\U000021c3', + '\\downharpoonright' : '\U000021c2', + '\\upharpoonleft' : '\U000021bf', + '\\upharpoonright' : '\U000021be', + '\\restriction' : '\U000021be', + '\\uparrow' : '\U00002191', + '\\Uparrow' : '\U000021d1', + '\\downarrow' : '\U00002193', + '\\Downarrow' : '\U000021d3', + '\\updownarrow' : '\U00002195', + '\\Updownarrow' : '\U000021d5', + '\\langle' : '\U000027e8', + '\\rangle' : '\U000027e9', + '\\lceil' : '\U00002308', + '\\rceil' : '\U00002309', + '\\lfloor' : '\U0000230a', + '\\rfloor' : '\U0000230b', + '\\flqq' : '\U000000ab', + '\\frqq' : '\U000000bb', + '\\bot' : '\U000022a5', + '\\top' : '\U000022a4', + '\\wedge' : '\U00002227', + '\\bigwedge' : '\U000022c0', + '\\vee' : '\U00002228', + '\\bigvee' : '\U000022c1', + '\\forall' : '\U00002200', + '\\exists' : '\U00002203', + '\\nexists' : '\U00002204', + '\\neg' : '\U000000ac', + '\\Box' : '\U000025a1', + '\\Diamond' : '\U000025c7', + '\\vdash' : '\U000022a2', + '\\models' : '\U000022a8', + '\\dashv' : '\U000022a3', + '\\surd' : '\U0000221a', + '\\le' : '\U00002264', + '\\ge' : '\U00002265', + '\\ll' : '\U0000226a', + '\\gg' : '\U0000226b', + '\\lesssim' : '\U00002272', + '\\gtrsim' : '\U00002273', + '\\lessapprox' : '\U00002a85', + '\\gtrapprox' : '\U00002a86', + '\\in' : '\U00002208', + '\\notin' : '\U00002209', + '\\subset' : '\U00002282', + '\\supset' : '\U00002283', + '\\subseteq' : '\U00002286', + '\\supseteq' : '\U00002287', + '\\sqsubset' : '\U0000228f', + '\\sqsupset' : '\U00002290', + '\\sqsubseteq' : '\U00002291', + '\\sqsupseteq' : '\U00002292', + '\\cap' : '\U00002229', + '\\bigcap' : '\U000022c2', + '\\cup' : '\U0000222a', + '\\bigcup' : '\U000022c3', + '\\sqcup' : '\U00002294', + '\\bigsqcup' : '\U00002a06', + '\\sqcap' : '\U00002293', + '\\Bigsqcap' : '\U00002a05', + '\\setminus' : '\U00002216', + '\\propto' : '\U0000221d', + '\\uplus' : '\U0000228e', + '\\bigplus' : '\U00002a04', + '\\sim' : '\U0000223c', + '\\doteq' : '\U00002250', + '\\simeq' : '\U00002243', + '\\approx' : '\U00002248', + '\\asymp' : '\U0000224d', + '\\cong' : '\U00002245', + '\\equiv' : '\U00002261', + '\\Join' : '\U000022c8', + '\\bowtie' : '\U00002a1d', + '\\prec' : '\U0000227a', + '\\succ' : '\U0000227b', + '\\preceq' : '\U0000227c', + '\\succeq' : '\U0000227d', + '\\parallel' : '\U00002225', + '\\mid' : '\U000000a6', + '\\pm' : '\U000000b1', + '\\mp' : '\U00002213', + '\\times' : '\U000000d7', + '\\div' : '\U000000f7', + '\\cdot' : '\U000022c5', + '\\star' : '\U000022c6', + '\\circ' : '\U00002218', + '\\dagger' : '\U00002020', + '\\ddagger' : '\U00002021', + '\\lhd' : '\U000022b2', + '\\rhd' : '\U000022b3', + '\\unlhd' : '\U000022b4', + '\\unrhd' : '\U000022b5', + '\\triangleleft' : '\U000025c3', + '\\triangleright' : '\U000025b9', + '\\triangle' : '\U000025b3', + '\\triangleq' : '\U0000225c', + '\\oplus' : '\U00002295', + '\\bigoplus' : '\U00002a01', + '\\otimes' : '\U00002297', + '\\bigotimes' : '\U00002a02', + '\\odot' : '\U00002299', + '\\bigodot' : '\U00002a00', + '\\ominus' : '\U00002296', + '\\oslash' : '\U00002298', + '\\dots' : '\U00002026', + '\\cdots' : '\U000022ef', + '\\sum' : '\U00002211', + '\\prod' : '\U0000220f', + '\\coprod' : '\U00002210', + '\\infty' : '\U0000221e', + '\\int' : '\U0000222b', + '\\oint' : '\U0000222e', + '\\clubsuit' : '\U00002663', + '\\diamondsuit' : '\U00002662', + '\\heartsuit' : '\U00002661', + '\\spadesuit' : '\U00002660', + '\\aleph' : '\U00002135', + '\\emptyset' : '\U00002205', + '\\nabla' : '\U00002207', + '\\partial' : '\U00002202', + '\\flat' : '\U0000266d', + '\\natural' : '\U0000266e', + '\\sharp' : '\U0000266f', + '\\angle' : '\U00002220', + '\\copyright' : '\U000000a9', + '\\textregistered' : '\U000000ae', + '\\textonequarter' : '\U000000bc', + '\\textonehalf' : '\U000000bd', + '\\textthreequarters' : '\U000000be', + '\\textordfeminine' : '\U000000aa', + '\\textordmasculine' : '\U000000ba', + '\\euro' : '\U000020ac', + '\\pounds' : '\U000000a3', + '\\yen' : '\U000000a5', + '\\textcent' : '\U000000a2', + '\\textcurrency' : '\U000000a4', + '\\textdegree' : '\U000000b0', } isabelle_symbols = { - '\\<zero>' : u'\U0001d7ec', - '\\<one>' : u'\U0001d7ed', - '\\<two>' : u'\U0001d7ee', - '\\<three>' : u'\U0001d7ef', - '\\<four>' : u'\U0001d7f0', - '\\<five>' : u'\U0001d7f1', - '\\<six>' : u'\U0001d7f2', - '\\<seven>' : u'\U0001d7f3', - '\\<eight>' : u'\U0001d7f4', - '\\<nine>' : u'\U0001d7f5', - '\\<A>' : u'\U0001d49c', - '\\<B>' : u'\U0000212c', - '\\<C>' : u'\U0001d49e', - '\\<D>' : u'\U0001d49f', - '\\<E>' : u'\U00002130', - '\\<F>' : u'\U00002131', - '\\<G>' : u'\U0001d4a2', - '\\<H>' : u'\U0000210b', - '\\<I>' : u'\U00002110', - '\\<J>' : u'\U0001d4a5', - '\\<K>' : u'\U0001d4a6', - '\\<L>' : u'\U00002112', - '\\<M>' : u'\U00002133', - '\\<N>' : u'\U0001d4a9', - '\\<O>' : u'\U0001d4aa', - '\\<P>' : u'\U0001d4ab', - '\\<Q>' : u'\U0001d4ac', - '\\<R>' : u'\U0000211b', - '\\<S>' : u'\U0001d4ae', - '\\<T>' : u'\U0001d4af', - '\\<U>' : u'\U0001d4b0', - '\\<V>' : u'\U0001d4b1', - '\\<W>' : u'\U0001d4b2', - '\\<X>' : u'\U0001d4b3', - '\\<Y>' : u'\U0001d4b4', - '\\<Z>' : u'\U0001d4b5', - '\\<a>' : u'\U0001d5ba', - '\\<b>' : u'\U0001d5bb', - '\\<c>' : u'\U0001d5bc', - '\\<d>' : u'\U0001d5bd', - '\\<e>' : u'\U0001d5be', - '\\<f>' : u'\U0001d5bf', - '\\<g>' : u'\U0001d5c0', - '\\<h>' : u'\U0001d5c1', - '\\<i>' : u'\U0001d5c2', - '\\<j>' : u'\U0001d5c3', - '\\<k>' : u'\U0001d5c4', - '\\<l>' : u'\U0001d5c5', - '\\<m>' : u'\U0001d5c6', - '\\<n>' : u'\U0001d5c7', - '\\<o>' : u'\U0001d5c8', - '\\<p>' : u'\U0001d5c9', - '\\<q>' : u'\U0001d5ca', - '\\<r>' : u'\U0001d5cb', - '\\<s>' : u'\U0001d5cc', - '\\<t>' : u'\U0001d5cd', - '\\<u>' : u'\U0001d5ce', - '\\<v>' : u'\U0001d5cf', - '\\<w>' : u'\U0001d5d0', - '\\<x>' : u'\U0001d5d1', - '\\<y>' : u'\U0001d5d2', - '\\<z>' : u'\U0001d5d3', - '\\<AA>' : u'\U0001d504', - '\\<BB>' : u'\U0001d505', - '\\<CC>' : u'\U0000212d', - '\\<DD>' : u'\U0001d507', - '\\<EE>' : u'\U0001d508', - '\\<FF>' : u'\U0001d509', - '\\<GG>' : u'\U0001d50a', - '\\<HH>' : u'\U0000210c', - '\\<II>' : u'\U00002111', - '\\<JJ>' : u'\U0001d50d', - '\\<KK>' : u'\U0001d50e', - '\\<LL>' : u'\U0001d50f', - '\\<MM>' : u'\U0001d510', - '\\<NN>' : u'\U0001d511', - '\\<OO>' : u'\U0001d512', - '\\<PP>' : u'\U0001d513', - '\\<QQ>' : u'\U0001d514', - '\\<RR>' : u'\U0000211c', - '\\<SS>' : u'\U0001d516', - '\\<TT>' : u'\U0001d517', - '\\<UU>' : u'\U0001d518', - '\\<VV>' : u'\U0001d519', - '\\<WW>' : u'\U0001d51a', - '\\<XX>' : u'\U0001d51b', - '\\<YY>' : u'\U0001d51c', - '\\<ZZ>' : u'\U00002128', - '\\<aa>' : u'\U0001d51e', - '\\<bb>' : u'\U0001d51f', - '\\<cc>' : u'\U0001d520', - '\\<dd>' : u'\U0001d521', - '\\<ee>' : u'\U0001d522', - '\\<ff>' : u'\U0001d523', - '\\<gg>' : u'\U0001d524', - '\\<hh>' : u'\U0001d525', - '\\<ii>' : u'\U0001d526', - '\\<jj>' : u'\U0001d527', - '\\<kk>' : u'\U0001d528', - '\\<ll>' : u'\U0001d529', - '\\<mm>' : u'\U0001d52a', - '\\<nn>' : u'\U0001d52b', - '\\<oo>' : u'\U0001d52c', - '\\<pp>' : u'\U0001d52d', - '\\<qq>' : u'\U0001d52e', - '\\<rr>' : u'\U0001d52f', - '\\<ss>' : u'\U0001d530', - '\\<tt>' : u'\U0001d531', - '\\<uu>' : u'\U0001d532', - '\\<vv>' : u'\U0001d533', - '\\<ww>' : u'\U0001d534', - '\\<xx>' : u'\U0001d535', - '\\<yy>' : u'\U0001d536', - '\\<zz>' : u'\U0001d537', - '\\<alpha>' : u'\U000003b1', - '\\<beta>' : u'\U000003b2', - '\\<gamma>' : u'\U000003b3', - '\\<delta>' : u'\U000003b4', - '\\<epsilon>' : u'\U000003b5', - '\\<zeta>' : u'\U000003b6', - '\\<eta>' : u'\U000003b7', - '\\<theta>' : u'\U000003b8', - '\\<iota>' : u'\U000003b9', - '\\<kappa>' : u'\U000003ba', - '\\<lambda>' : u'\U000003bb', - '\\<mu>' : u'\U000003bc', - '\\<nu>' : u'\U000003bd', - '\\<xi>' : u'\U000003be', - '\\<pi>' : u'\U000003c0', - '\\<rho>' : u'\U000003c1', - '\\<sigma>' : u'\U000003c3', - '\\<tau>' : u'\U000003c4', - '\\<upsilon>' : u'\U000003c5', - '\\<phi>' : u'\U000003c6', - '\\<chi>' : u'\U000003c7', - '\\<psi>' : u'\U000003c8', - '\\<omega>' : u'\U000003c9', - '\\<Gamma>' : u'\U00000393', - '\\<Delta>' : u'\U00000394', - '\\<Theta>' : u'\U00000398', - '\\<Lambda>' : u'\U0000039b', - '\\<Xi>' : u'\U0000039e', - '\\<Pi>' : u'\U000003a0', - '\\<Sigma>' : u'\U000003a3', - '\\<Upsilon>' : u'\U000003a5', - '\\<Phi>' : u'\U000003a6', - '\\<Psi>' : u'\U000003a8', - '\\<Omega>' : u'\U000003a9', - '\\<bool>' : u'\U0001d539', - '\\<complex>' : u'\U00002102', - '\\<nat>' : u'\U00002115', - '\\<rat>' : u'\U0000211a', - '\\<real>' : u'\U0000211d', - '\\<int>' : u'\U00002124', - '\\<leftarrow>' : u'\U00002190', - '\\<longleftarrow>' : u'\U000027f5', - '\\<rightarrow>' : u'\U00002192', - '\\<longrightarrow>' : u'\U000027f6', - '\\<Leftarrow>' : u'\U000021d0', - '\\<Longleftarrow>' : u'\U000027f8', - '\\<Rightarrow>' : u'\U000021d2', - '\\<Longrightarrow>' : u'\U000027f9', - '\\<leftrightarrow>' : u'\U00002194', - '\\<longleftrightarrow>' : u'\U000027f7', - '\\<Leftrightarrow>' : u'\U000021d4', - '\\<Longleftrightarrow>' : u'\U000027fa', - '\\<mapsto>' : u'\U000021a6', - '\\<longmapsto>' : u'\U000027fc', - '\\<midarrow>' : u'\U00002500', - '\\<Midarrow>' : u'\U00002550', - '\\<hookleftarrow>' : u'\U000021a9', - '\\<hookrightarrow>' : u'\U000021aa', - '\\<leftharpoondown>' : u'\U000021bd', - '\\<rightharpoondown>' : u'\U000021c1', - '\\<leftharpoonup>' : u'\U000021bc', - '\\<rightharpoonup>' : u'\U000021c0', - '\\<rightleftharpoons>' : u'\U000021cc', - '\\<leadsto>' : u'\U0000219d', - '\\<downharpoonleft>' : u'\U000021c3', - '\\<downharpoonright>' : u'\U000021c2', - '\\<upharpoonleft>' : u'\U000021bf', - '\\<upharpoonright>' : u'\U000021be', - '\\<restriction>' : u'\U000021be', - '\\<Colon>' : u'\U00002237', - '\\<up>' : u'\U00002191', - '\\<Up>' : u'\U000021d1', - '\\<down>' : u'\U00002193', - '\\<Down>' : u'\U000021d3', - '\\<updown>' : u'\U00002195', - '\\<Updown>' : u'\U000021d5', - '\\<langle>' : u'\U000027e8', - '\\<rangle>' : u'\U000027e9', - '\\<lceil>' : u'\U00002308', - '\\<rceil>' : u'\U00002309', - '\\<lfloor>' : u'\U0000230a', - '\\<rfloor>' : u'\U0000230b', - '\\<lparr>' : u'\U00002987', - '\\<rparr>' : u'\U00002988', - '\\<lbrakk>' : u'\U000027e6', - '\\<rbrakk>' : u'\U000027e7', - '\\<lbrace>' : u'\U00002983', - '\\<rbrace>' : u'\U00002984', - '\\<guillemotleft>' : u'\U000000ab', - '\\<guillemotright>' : u'\U000000bb', - '\\<bottom>' : u'\U000022a5', - '\\<top>' : u'\U000022a4', - '\\<and>' : u'\U00002227', - '\\<And>' : u'\U000022c0', - '\\<or>' : u'\U00002228', - '\\<Or>' : u'\U000022c1', - '\\<forall>' : u'\U00002200', - '\\<exists>' : u'\U00002203', - '\\<nexists>' : u'\U00002204', - '\\<not>' : u'\U000000ac', - '\\<box>' : u'\U000025a1', - '\\<diamond>' : u'\U000025c7', - '\\<turnstile>' : u'\U000022a2', - '\\<Turnstile>' : u'\U000022a8', - '\\<tturnstile>' : u'\U000022a9', - '\\<TTurnstile>' : u'\U000022ab', - '\\<stileturn>' : u'\U000022a3', - '\\<surd>' : u'\U0000221a', - '\\<le>' : u'\U00002264', - '\\<ge>' : u'\U00002265', - '\\<lless>' : u'\U0000226a', - '\\<ggreater>' : u'\U0000226b', - '\\<lesssim>' : u'\U00002272', - '\\<greatersim>' : u'\U00002273', - '\\<lessapprox>' : u'\U00002a85', - '\\<greaterapprox>' : u'\U00002a86', - '\\<in>' : u'\U00002208', - '\\<notin>' : u'\U00002209', - '\\<subset>' : u'\U00002282', - '\\<supset>' : u'\U00002283', - '\\<subseteq>' : u'\U00002286', - '\\<supseteq>' : u'\U00002287', - '\\<sqsubset>' : u'\U0000228f', - '\\<sqsupset>' : u'\U00002290', - '\\<sqsubseteq>' : u'\U00002291', - '\\<sqsupseteq>' : u'\U00002292', - '\\<inter>' : u'\U00002229', - '\\<Inter>' : u'\U000022c2', - '\\<union>' : u'\U0000222a', - '\\<Union>' : u'\U000022c3', - '\\<squnion>' : u'\U00002294', - '\\<Squnion>' : u'\U00002a06', - '\\<sqinter>' : u'\U00002293', - '\\<Sqinter>' : u'\U00002a05', - '\\<setminus>' : u'\U00002216', - '\\<propto>' : u'\U0000221d', - '\\<uplus>' : u'\U0000228e', - '\\<Uplus>' : u'\U00002a04', - '\\<noteq>' : u'\U00002260', - '\\<sim>' : u'\U0000223c', - '\\<doteq>' : u'\U00002250', - '\\<simeq>' : u'\U00002243', - '\\<approx>' : u'\U00002248', - '\\<asymp>' : u'\U0000224d', - '\\<cong>' : u'\U00002245', - '\\<smile>' : u'\U00002323', - '\\<equiv>' : u'\U00002261', - '\\<frown>' : u'\U00002322', - '\\<Join>' : u'\U000022c8', - '\\<bowtie>' : u'\U00002a1d', - '\\<prec>' : u'\U0000227a', - '\\<succ>' : u'\U0000227b', - '\\<preceq>' : u'\U0000227c', - '\\<succeq>' : u'\U0000227d', - '\\<parallel>' : u'\U00002225', - '\\<bar>' : u'\U000000a6', - '\\<plusminus>' : u'\U000000b1', - '\\<minusplus>' : u'\U00002213', - '\\<times>' : u'\U000000d7', - '\\<div>' : u'\U000000f7', - '\\<cdot>' : u'\U000022c5', - '\\<star>' : u'\U000022c6', - '\\<bullet>' : u'\U00002219', - '\\<circ>' : u'\U00002218', - '\\<dagger>' : u'\U00002020', - '\\<ddagger>' : u'\U00002021', - '\\<lhd>' : u'\U000022b2', - '\\<rhd>' : u'\U000022b3', - '\\<unlhd>' : u'\U000022b4', - '\\<unrhd>' : u'\U000022b5', - '\\<triangleleft>' : u'\U000025c3', - '\\<triangleright>' : u'\U000025b9', - '\\<triangle>' : u'\U000025b3', - '\\<triangleq>' : u'\U0000225c', - '\\<oplus>' : u'\U00002295', - '\\<Oplus>' : u'\U00002a01', - '\\<otimes>' : u'\U00002297', - '\\<Otimes>' : u'\U00002a02', - '\\<odot>' : u'\U00002299', - '\\<Odot>' : u'\U00002a00', - '\\<ominus>' : u'\U00002296', - '\\<oslash>' : u'\U00002298', - '\\<dots>' : u'\U00002026', - '\\<cdots>' : u'\U000022ef', - '\\<Sum>' : u'\U00002211', - '\\<Prod>' : u'\U0000220f', - '\\<Coprod>' : u'\U00002210', - '\\<infinity>' : u'\U0000221e', - '\\<integral>' : u'\U0000222b', - '\\<ointegral>' : u'\U0000222e', - '\\<clubsuit>' : u'\U00002663', - '\\<diamondsuit>' : u'\U00002662', - '\\<heartsuit>' : u'\U00002661', - '\\<spadesuit>' : u'\U00002660', - '\\<aleph>' : u'\U00002135', - '\\<emptyset>' : u'\U00002205', - '\\<nabla>' : u'\U00002207', - '\\<partial>' : u'\U00002202', - '\\<flat>' : u'\U0000266d', - '\\<natural>' : u'\U0000266e', - '\\<sharp>' : u'\U0000266f', - '\\<angle>' : u'\U00002220', - '\\<copyright>' : u'\U000000a9', - '\\<registered>' : u'\U000000ae', - '\\<hyphen>' : u'\U000000ad', - '\\<inverse>' : u'\U000000af', - '\\<onequarter>' : u'\U000000bc', - '\\<onehalf>' : u'\U000000bd', - '\\<threequarters>' : u'\U000000be', - '\\<ordfeminine>' : u'\U000000aa', - '\\<ordmasculine>' : u'\U000000ba', - '\\<section>' : u'\U000000a7', - '\\<paragraph>' : u'\U000000b6', - '\\<exclamdown>' : u'\U000000a1', - '\\<questiondown>' : u'\U000000bf', - '\\<euro>' : u'\U000020ac', - '\\<pounds>' : u'\U000000a3', - '\\<yen>' : u'\U000000a5', - '\\<cent>' : u'\U000000a2', - '\\<currency>' : u'\U000000a4', - '\\<degree>' : u'\U000000b0', - '\\<amalg>' : u'\U00002a3f', - '\\<mho>' : u'\U00002127', - '\\<lozenge>' : u'\U000025ca', - '\\<wp>' : u'\U00002118', - '\\<wrong>' : u'\U00002240', - '\\<struct>' : u'\U000022c4', - '\\<acute>' : u'\U000000b4', - '\\<index>' : u'\U00000131', - '\\<dieresis>' : u'\U000000a8', - '\\<cedilla>' : u'\U000000b8', - '\\<hungarumlaut>' : u'\U000002dd', - '\\<some>' : u'\U000003f5', - '\\<newline>' : u'\U000023ce', - '\\<open>' : u'\U00002039', - '\\<close>' : u'\U0000203a', - '\\<here>' : u'\U00002302', - '\\<^sub>' : u'\U000021e9', - '\\<^sup>' : u'\U000021e7', - '\\<^bold>' : u'\U00002759', - '\\<^bsub>' : u'\U000021d8', - '\\<^esub>' : u'\U000021d9', - '\\<^bsup>' : u'\U000021d7', - '\\<^esup>' : u'\U000021d6', + '\\<zero>' : '\U0001d7ec', + '\\<one>' : '\U0001d7ed', + '\\<two>' : '\U0001d7ee', + '\\<three>' : '\U0001d7ef', + '\\<four>' : '\U0001d7f0', + '\\<five>' : '\U0001d7f1', + '\\<six>' : '\U0001d7f2', + '\\<seven>' : '\U0001d7f3', + '\\<eight>' : '\U0001d7f4', + '\\<nine>' : '\U0001d7f5', + '\\<A>' : '\U0001d49c', + '\\<B>' : '\U0000212c', + '\\<C>' : '\U0001d49e', + '\\<D>' : '\U0001d49f', + '\\<E>' : '\U00002130', + '\\<F>' : '\U00002131', + '\\<G>' : '\U0001d4a2', + '\\<H>' : '\U0000210b', + '\\<I>' : '\U00002110', + '\\<J>' : '\U0001d4a5', + '\\<K>' : '\U0001d4a6', + '\\<L>' : '\U00002112', + '\\<M>' : '\U00002133', + '\\<N>' : '\U0001d4a9', + '\\<O>' : '\U0001d4aa', + '\\<P>' : '\U0001d4ab', + '\\<Q>' : '\U0001d4ac', + '\\<R>' : '\U0000211b', + '\\<S>' : '\U0001d4ae', + '\\<T>' : '\U0001d4af', + '\\<U>' : '\U0001d4b0', + '\\<V>' : '\U0001d4b1', + '\\<W>' : '\U0001d4b2', + '\\<X>' : '\U0001d4b3', + '\\<Y>' : '\U0001d4b4', + '\\<Z>' : '\U0001d4b5', + '\\<a>' : '\U0001d5ba', + '\\<b>' : '\U0001d5bb', + '\\<c>' : '\U0001d5bc', + '\\<d>' : '\U0001d5bd', + '\\<e>' : '\U0001d5be', + '\\<f>' : '\U0001d5bf', + '\\<g>' : '\U0001d5c0', + '\\<h>' : '\U0001d5c1', + '\\<i>' : '\U0001d5c2', + '\\<j>' : '\U0001d5c3', + '\\<k>' : '\U0001d5c4', + '\\<l>' : '\U0001d5c5', + '\\<m>' : '\U0001d5c6', + '\\<n>' : '\U0001d5c7', + '\\<o>' : '\U0001d5c8', + '\\<p>' : '\U0001d5c9', + '\\<q>' : '\U0001d5ca', + '\\<r>' : '\U0001d5cb', + '\\<s>' : '\U0001d5cc', + '\\<t>' : '\U0001d5cd', + '\\<u>' : '\U0001d5ce', + '\\<v>' : '\U0001d5cf', + '\\<w>' : '\U0001d5d0', + '\\<x>' : '\U0001d5d1', + '\\<y>' : '\U0001d5d2', + '\\<z>' : '\U0001d5d3', + '\\<AA>' : '\U0001d504', + '\\<BB>' : '\U0001d505', + '\\<CC>' : '\U0000212d', + '\\<DD>' : '\U0001d507', + '\\<EE>' : '\U0001d508', + '\\<FF>' : '\U0001d509', + '\\<GG>' : '\U0001d50a', + '\\<HH>' : '\U0000210c', + '\\<II>' : '\U00002111', + '\\<JJ>' : '\U0001d50d', + '\\<KK>' : '\U0001d50e', + '\\<LL>' : '\U0001d50f', + '\\<MM>' : '\U0001d510', + '\\<NN>' : '\U0001d511', + '\\<OO>' : '\U0001d512', + '\\<PP>' : '\U0001d513', + '\\<QQ>' : '\U0001d514', + '\\<RR>' : '\U0000211c', + '\\<SS>' : '\U0001d516', + '\\<TT>' : '\U0001d517', + '\\<UU>' : '\U0001d518', + '\\<VV>' : '\U0001d519', + '\\<WW>' : '\U0001d51a', + '\\<XX>' : '\U0001d51b', + '\\<YY>' : '\U0001d51c', + '\\<ZZ>' : '\U00002128', + '\\<aa>' : '\U0001d51e', + '\\<bb>' : '\U0001d51f', + '\\<cc>' : '\U0001d520', + '\\<dd>' : '\U0001d521', + '\\<ee>' : '\U0001d522', + '\\<ff>' : '\U0001d523', + '\\<gg>' : '\U0001d524', + '\\<hh>' : '\U0001d525', + '\\<ii>' : '\U0001d526', + '\\<jj>' : '\U0001d527', + '\\<kk>' : '\U0001d528', + '\\<ll>' : '\U0001d529', + '\\<mm>' : '\U0001d52a', + '\\<nn>' : '\U0001d52b', + '\\<oo>' : '\U0001d52c', + '\\<pp>' : '\U0001d52d', + '\\<qq>' : '\U0001d52e', + '\\<rr>' : '\U0001d52f', + '\\<ss>' : '\U0001d530', + '\\<tt>' : '\U0001d531', + '\\<uu>' : '\U0001d532', + '\\<vv>' : '\U0001d533', + '\\<ww>' : '\U0001d534', + '\\<xx>' : '\U0001d535', + '\\<yy>' : '\U0001d536', + '\\<zz>' : '\U0001d537', + '\\<alpha>' : '\U000003b1', + '\\<beta>' : '\U000003b2', + '\\<gamma>' : '\U000003b3', + '\\<delta>' : '\U000003b4', + '\\<epsilon>' : '\U000003b5', + '\\<zeta>' : '\U000003b6', + '\\<eta>' : '\U000003b7', + '\\<theta>' : '\U000003b8', + '\\<iota>' : '\U000003b9', + '\\<kappa>' : '\U000003ba', + '\\<lambda>' : '\U000003bb', + '\\<mu>' : '\U000003bc', + '\\<nu>' : '\U000003bd', + '\\<xi>' : '\U000003be', + '\\<pi>' : '\U000003c0', + '\\<rho>' : '\U000003c1', + '\\<sigma>' : '\U000003c3', + '\\<tau>' : '\U000003c4', + '\\<upsilon>' : '\U000003c5', + '\\<phi>' : '\U000003c6', + '\\<chi>' : '\U000003c7', + '\\<psi>' : '\U000003c8', + '\\<omega>' : '\U000003c9', + '\\<Gamma>' : '\U00000393', + '\\<Delta>' : '\U00000394', + '\\<Theta>' : '\U00000398', + '\\<Lambda>' : '\U0000039b', + '\\<Xi>' : '\U0000039e', + '\\<Pi>' : '\U000003a0', + '\\<Sigma>' : '\U000003a3', + '\\<Upsilon>' : '\U000003a5', + '\\<Phi>' : '\U000003a6', + '\\<Psi>' : '\U000003a8', + '\\<Omega>' : '\U000003a9', + '\\<bool>' : '\U0001d539', + '\\<complex>' : '\U00002102', + '\\<nat>' : '\U00002115', + '\\<rat>' : '\U0000211a', + '\\<real>' : '\U0000211d', + '\\<int>' : '\U00002124', + '\\<leftarrow>' : '\U00002190', + '\\<longleftarrow>' : '\U000027f5', + '\\<rightarrow>' : '\U00002192', + '\\<longrightarrow>' : '\U000027f6', + '\\<Leftarrow>' : '\U000021d0', + '\\<Longleftarrow>' : '\U000027f8', + '\\<Rightarrow>' : '\U000021d2', + '\\<Longrightarrow>' : '\U000027f9', + '\\<leftrightarrow>' : '\U00002194', + '\\<longleftrightarrow>' : '\U000027f7', + '\\<Leftrightarrow>' : '\U000021d4', + '\\<Longleftrightarrow>' : '\U000027fa', + '\\<mapsto>' : '\U000021a6', + '\\<longmapsto>' : '\U000027fc', + '\\<midarrow>' : '\U00002500', + '\\<Midarrow>' : '\U00002550', + '\\<hookleftarrow>' : '\U000021a9', + '\\<hookrightarrow>' : '\U000021aa', + '\\<leftharpoondown>' : '\U000021bd', + '\\<rightharpoondown>' : '\U000021c1', + '\\<leftharpoonup>' : '\U000021bc', + '\\<rightharpoonup>' : '\U000021c0', + '\\<rightleftharpoons>' : '\U000021cc', + '\\<leadsto>' : '\U0000219d', + '\\<downharpoonleft>' : '\U000021c3', + '\\<downharpoonright>' : '\U000021c2', + '\\<upharpoonleft>' : '\U000021bf', + '\\<upharpoonright>' : '\U000021be', + '\\<restriction>' : '\U000021be', + '\\<Colon>' : '\U00002237', + '\\<up>' : '\U00002191', + '\\<Up>' : '\U000021d1', + '\\<down>' : '\U00002193', + '\\<Down>' : '\U000021d3', + '\\<updown>' : '\U00002195', + '\\<Updown>' : '\U000021d5', + '\\<langle>' : '\U000027e8', + '\\<rangle>' : '\U000027e9', + '\\<lceil>' : '\U00002308', + '\\<rceil>' : '\U00002309', + '\\<lfloor>' : '\U0000230a', + '\\<rfloor>' : '\U0000230b', + '\\<lparr>' : '\U00002987', + '\\<rparr>' : '\U00002988', + '\\<lbrakk>' : '\U000027e6', + '\\<rbrakk>' : '\U000027e7', + '\\<lbrace>' : '\U00002983', + '\\<rbrace>' : '\U00002984', + '\\<guillemotleft>' : '\U000000ab', + '\\<guillemotright>' : '\U000000bb', + '\\<bottom>' : '\U000022a5', + '\\<top>' : '\U000022a4', + '\\<and>' : '\U00002227', + '\\<And>' : '\U000022c0', + '\\<or>' : '\U00002228', + '\\<Or>' : '\U000022c1', + '\\<forall>' : '\U00002200', + '\\<exists>' : '\U00002203', + '\\<nexists>' : '\U00002204', + '\\<not>' : '\U000000ac', + '\\<box>' : '\U000025a1', + '\\<diamond>' : '\U000025c7', + '\\<turnstile>' : '\U000022a2', + '\\<Turnstile>' : '\U000022a8', + '\\<tturnstile>' : '\U000022a9', + '\\<TTurnstile>' : '\U000022ab', + '\\<stileturn>' : '\U000022a3', + '\\<surd>' : '\U0000221a', + '\\<le>' : '\U00002264', + '\\<ge>' : '\U00002265', + '\\<lless>' : '\U0000226a', + '\\<ggreater>' : '\U0000226b', + '\\<lesssim>' : '\U00002272', + '\\<greatersim>' : '\U00002273', + '\\<lessapprox>' : '\U00002a85', + '\\<greaterapprox>' : '\U00002a86', + '\\<in>' : '\U00002208', + '\\<notin>' : '\U00002209', + '\\<subset>' : '\U00002282', + '\\<supset>' : '\U00002283', + '\\<subseteq>' : '\U00002286', + '\\<supseteq>' : '\U00002287', + '\\<sqsubset>' : '\U0000228f', + '\\<sqsupset>' : '\U00002290', + '\\<sqsubseteq>' : '\U00002291', + '\\<sqsupseteq>' : '\U00002292', + '\\<inter>' : '\U00002229', + '\\<Inter>' : '\U000022c2', + '\\<union>' : '\U0000222a', + '\\<Union>' : '\U000022c3', + '\\<squnion>' : '\U00002294', + '\\<Squnion>' : '\U00002a06', + '\\<sqinter>' : '\U00002293', + '\\<Sqinter>' : '\U00002a05', + '\\<setminus>' : '\U00002216', + '\\<propto>' : '\U0000221d', + '\\<uplus>' : '\U0000228e', + '\\<Uplus>' : '\U00002a04', + '\\<noteq>' : '\U00002260', + '\\<sim>' : '\U0000223c', + '\\<doteq>' : '\U00002250', + '\\<simeq>' : '\U00002243', + '\\<approx>' : '\U00002248', + '\\<asymp>' : '\U0000224d', + '\\<cong>' : '\U00002245', + '\\<smile>' : '\U00002323', + '\\<equiv>' : '\U00002261', + '\\<frown>' : '\U00002322', + '\\<Join>' : '\U000022c8', + '\\<bowtie>' : '\U00002a1d', + '\\<prec>' : '\U0000227a', + '\\<succ>' : '\U0000227b', + '\\<preceq>' : '\U0000227c', + '\\<succeq>' : '\U0000227d', + '\\<parallel>' : '\U00002225', + '\\<bar>' : '\U000000a6', + '\\<plusminus>' : '\U000000b1', + '\\<minusplus>' : '\U00002213', + '\\<times>' : '\U000000d7', + '\\<div>' : '\U000000f7', + '\\<cdot>' : '\U000022c5', + '\\<star>' : '\U000022c6', + '\\<bullet>' : '\U00002219', + '\\<circ>' : '\U00002218', + '\\<dagger>' : '\U00002020', + '\\<ddagger>' : '\U00002021', + '\\<lhd>' : '\U000022b2', + '\\<rhd>' : '\U000022b3', + '\\<unlhd>' : '\U000022b4', + '\\<unrhd>' : '\U000022b5', + '\\<triangleleft>' : '\U000025c3', + '\\<triangleright>' : '\U000025b9', + '\\<triangle>' : '\U000025b3', + '\\<triangleq>' : '\U0000225c', + '\\<oplus>' : '\U00002295', + '\\<Oplus>' : '\U00002a01', + '\\<otimes>' : '\U00002297', + '\\<Otimes>' : '\U00002a02', + '\\<odot>' : '\U00002299', + '\\<Odot>' : '\U00002a00', + '\\<ominus>' : '\U00002296', + '\\<oslash>' : '\U00002298', + '\\<dots>' : '\U00002026', + '\\<cdots>' : '\U000022ef', + '\\<Sum>' : '\U00002211', + '\\<Prod>' : '\U0000220f', + '\\<Coprod>' : '\U00002210', + '\\<infinity>' : '\U0000221e', + '\\<integral>' : '\U0000222b', + '\\<ointegral>' : '\U0000222e', + '\\<clubsuit>' : '\U00002663', + '\\<diamondsuit>' : '\U00002662', + '\\<heartsuit>' : '\U00002661', + '\\<spadesuit>' : '\U00002660', + '\\<aleph>' : '\U00002135', + '\\<emptyset>' : '\U00002205', + '\\<nabla>' : '\U00002207', + '\\<partial>' : '\U00002202', + '\\<flat>' : '\U0000266d', + '\\<natural>' : '\U0000266e', + '\\<sharp>' : '\U0000266f', + '\\<angle>' : '\U00002220', + '\\<copyright>' : '\U000000a9', + '\\<registered>' : '\U000000ae', + '\\<hyphen>' : '\U000000ad', + '\\<inverse>' : '\U000000af', + '\\<onequarter>' : '\U000000bc', + '\\<onehalf>' : '\U000000bd', + '\\<threequarters>' : '\U000000be', + '\\<ordfeminine>' : '\U000000aa', + '\\<ordmasculine>' : '\U000000ba', + '\\<section>' : '\U000000a7', + '\\<paragraph>' : '\U000000b6', + '\\<exclamdown>' : '\U000000a1', + '\\<questiondown>' : '\U000000bf', + '\\<euro>' : '\U000020ac', + '\\<pounds>' : '\U000000a3', + '\\<yen>' : '\U000000a5', + '\\<cent>' : '\U000000a2', + '\\<currency>' : '\U000000a4', + '\\<degree>' : '\U000000b0', + '\\<amalg>' : '\U00002a3f', + '\\<mho>' : '\U00002127', + '\\<lozenge>' : '\U000025ca', + '\\<wp>' : '\U00002118', + '\\<wrong>' : '\U00002240', + '\\<struct>' : '\U000022c4', + '\\<acute>' : '\U000000b4', + '\\<index>' : '\U00000131', + '\\<dieresis>' : '\U000000a8', + '\\<cedilla>' : '\U000000b8', + '\\<hungarumlaut>' : '\U000002dd', + '\\<some>' : '\U000003f5', + '\\<newline>' : '\U000023ce', + '\\<open>' : '\U00002039', + '\\<close>' : '\U0000203a', + '\\<here>' : '\U00002302', + '\\<^sub>' : '\U000021e9', + '\\<^sup>' : '\U000021e7', + '\\<^bold>' : '\U00002759', + '\\<^bsub>' : '\U000021d8', + '\\<^esub>' : '\U000021d9', + '\\<^bsup>' : '\U000021d7', + '\\<^esup>' : '\U000021d6', } lang_map = {'isabelle' : isabelle_symbols, 'latex' : latex_symbols} @@ -816,9 +816,9 @@ class VisibleWhitespaceFilter(Filter): def __init__(self, **options): Filter.__init__(self, **options) - for name, default in [('spaces', u'·'), - ('tabs', u'»'), - ('newlines', u'¶')]: + for name, default in [('spaces', '·'), + ('tabs', '»'), + ('newlines', '¶')]: opt = options.get(name, False) if isinstance(opt, str) and len(opt) == 1: setattr(self, name, opt) @@ -833,9 +833,9 @@ class VisibleWhitespaceFilter(Filter): def filter(self, lexer, stream): if self.wstt: - spaces = self.spaces or u' ' - tabs = self.tabs or u'\t' - newlines = self.newlines or u'\n' + spaces = self.spaces or ' ' + tabs = self.tabs or '\t' + newlines = self.newlines or '\n' regex = re.compile(r'\s') def replacefunc(wschar): @@ -885,7 +885,7 @@ class GobbleFilter(Filter): if left < len(value): return value[left:], 0 else: - return u'', left - len(value) + return '', left - len(value) def filter(self, lexer, stream): n = self.n @@ -896,7 +896,7 @@ class GobbleFilter(Filter): (parts[0], left) = self.gobble(parts[0], left) for i in range(1, len(parts)): (parts[i], left) = self.gobble(parts[i], n) - value = u'\n'.join(parts) + value = '\n'.join(parts) if value != '': yield ttype, value diff --git a/pygments/formatters/html.py b/pygments/formatters/html.py index 421710f7..29cc9fb8 100644 --- a/pygments/formatters/html.py +++ b/pygments/formatters/html.py @@ -27,11 +27,11 @@ __all__ = ['HtmlFormatter'] _escape_html_table = { - ord('&'): u'&', - ord('<'): u'<', - ord('>'): u'>', - ord('"'): u'"', - ord("'"): u''', + ord('&'): '&', + ord('<'): '<', + ord('>'): '>', + ord('"'): '"', + ord("'"): ''', } @@ -433,7 +433,7 @@ class HtmlFormatter(Formatter): self.linenostep = abs(get_int_opt(options, 'linenostep', 1)) self.linenospecial = abs(get_int_opt(options, 'linenospecial', 0)) self.nobackground = get_bool_opt(options, 'nobackground', False) - self.lineseparator = options.get('lineseparator', u'\n') + self.lineseparator = options.get('lineseparator', '\n') self.lineanchors = options.get('lineanchors', '') self.linespans = options.get('linespans', '') self.anchorlinenos = get_bool_opt(options, 'anchorlinenos', False) diff --git a/pygments/formatters/latex.py b/pygments/formatters/latex.py index 9e23740d..a0e9421b 100644 --- a/pygments/formatters/latex.py +++ b/pygments/formatters/latex.py @@ -265,7 +265,7 @@ class LatexFormatter(Formatter): self.right = self.escapeinside[1] else: self.escapeinside = '' - self.envname = options.get('envname', u'Verbatim') + self.envname = options.get('envname', 'Verbatim') self._create_stylesheet() @@ -336,17 +336,17 @@ class LatexFormatter(Formatter): realoutfile = outfile outfile = StringIO() - outfile.write(u'\\begin{' + self.envname + u'}[commandchars=\\\\\\{\\}') + outfile.write('\\begin{' + self.envname + '}[commandchars=\\\\\\{\\}') if self.linenos: start, step = self.linenostart, self.linenostep - outfile.write(u',numbers=left' + - (start and u',firstnumber=%d' % start or u'') + - (step and u',stepnumber=%d' % step or u'')) + outfile.write(',numbers=left' + + (start and ',firstnumber=%d' % start or '') + + (step and ',stepnumber=%d' % step or '')) if self.mathescape or self.texcomments or self.escapeinside: - outfile.write(u',codes={\\catcode`\\$=3\\catcode`\\^=7\\catcode`\\_=8}') + outfile.write(',codes={\\catcode`\\$=3\\catcode`\\^=7\\catcode`\\_=8}') if self.verboptions: - outfile.write(u',' + self.verboptions) - outfile.write(u']\n') + outfile.write(',' + self.verboptions) + outfile.write(']\n') for ttype, value in tokensource: if ttype in Token.Comment: @@ -409,7 +409,7 @@ class LatexFormatter(Formatter): else: outfile.write(value) - outfile.write(u'\\end{' + self.envname + u'}\n') + outfile.write('\\end{' + self.envname + '}\n') if self.full: encoding = self.encoding or 'utf8' diff --git a/pygments/formatters/other.py b/pygments/formatters/other.py index 9f8bab4b..a86e81e0 100644 --- a/pygments/formatters/other.py +++ b/pygments/formatters/other.py @@ -119,12 +119,12 @@ class RawTokenFormatter(Formatter): flush() -TESTCASE_BEFORE = u'''\ +TESTCASE_BEFORE = '''\ def testNeedsName(lexer): fragment = %r tokens = [ ''' -TESTCASE_AFTER = u'''\ +TESTCASE_AFTER = '''\ ] assert list(lexer.get_tokens(fragment)) == tokens ''' @@ -152,8 +152,8 @@ class TestcaseFormatter(Formatter): rawbuf.append(value) outbuf.append('%s(%s, %r),\n' % (indentation, ttype, value)) - before = TESTCASE_BEFORE % (u''.join(rawbuf),) - during = u''.join(outbuf) + before = TESTCASE_BEFORE % (''.join(rawbuf),) + during = ''.join(outbuf) after = TESTCASE_AFTER if self.encoding is None: outfile.write(before + during + after) diff --git a/pygments/formatters/rtf.py b/pygments/formatters/rtf.py index 7f51bacd..34393d47 100644 --- a/pygments/formatters/rtf.py +++ b/pygments/formatters/rtf.py @@ -10,7 +10,7 @@ """ from pygments.formatter import Formatter -from pygments.util import get_int_opt, _surrogatepair +from pygments.util import get_int_opt, surrogatepair __all__ = ['RtfFormatter'] @@ -65,14 +65,14 @@ class RtfFormatter(Formatter): self.fontsize = get_int_opt(options, 'fontsize', 0) def _escape(self, text): - return text.replace(u'\\', u'\\\\') \ - .replace(u'{', u'\\{') \ - .replace(u'}', u'\\}') + return text.replace('\\', '\\\\') \ + .replace('{', '\\{') \ + .replace('}', '\\}') def _escape_text(self, text): # empty strings, should give a small performance improvement if not text: - return u'' + return '' # escape text text = self._escape(text) @@ -85,21 +85,21 @@ class RtfFormatter(Formatter): buf.append(str(c)) elif (2**7) <= cn < (2**16): # single unicode escape sequence - buf.append(u'{\\u%d}' % cn) + buf.append('{\\u%d}' % cn) elif (2**16) <= cn: # RTF limits unicode to 16 bits. # Force surrogate pairs - buf.append(u'{\\u%d}{\\u%d}' % _surrogatepair(cn)) + buf.append('{\\u%d}{\\u%d}' % surrogatepair(cn)) - return u''.join(buf).replace(u'\n', u'\\par\n') + return ''.join(buf).replace('\n', '\\par\n') def format_unencoded(self, tokensource, outfile): # rtf 1.8 header - outfile.write(u'{\\rtf1\\ansi\\uc0\\deff0' - u'{\\fonttbl{\\f0\\fmodern\\fprq1\\fcharset0%s;}}' - u'{\\colortbl;' % (self.fontface and - u' ' + self._escape(self.fontface) or - u'')) + outfile.write('{\\rtf1\\ansi\\uc0\\deff0' + '{\\fonttbl{\\f0\\fmodern\\fprq1\\fcharset0%s;}}' + '{\\colortbl;' % (self.fontface and + ' ' + self._escape(self.fontface) or + '')) # convert colors and save them in a mapping to access them later. color_mapping = {} @@ -108,15 +108,15 @@ class RtfFormatter(Formatter): for color in style['color'], style['bgcolor'], style['border']: if color and color not in color_mapping: color_mapping[color] = offset - outfile.write(u'\\red%d\\green%d\\blue%d;' % ( + outfile.write('\\red%d\\green%d\\blue%d;' % ( int(color[0:2], 16), int(color[2:4], 16), int(color[4:6], 16) )) offset += 1 - outfile.write(u'}\\f0 ') + outfile.write('}\\f0 ') if self.fontsize: - outfile.write(u'\\fs%d' % (self.fontsize)) + outfile.write('\\fs%d' % self.fontsize) # highlight stream for ttype, value in tokensource: @@ -125,23 +125,23 @@ class RtfFormatter(Formatter): style = self.style.style_for_token(ttype) buf = [] if style['bgcolor']: - buf.append(u'\\cb%d' % color_mapping[style['bgcolor']]) + buf.append('\\cb%d' % color_mapping[style['bgcolor']]) if style['color']: - buf.append(u'\\cf%d' % color_mapping[style['color']]) + buf.append('\\cf%d' % color_mapping[style['color']]) if style['bold']: - buf.append(u'\\b') + buf.append('\\b') if style['italic']: - buf.append(u'\\i') + buf.append('\\i') if style['underline']: - buf.append(u'\\ul') + buf.append('\\ul') if style['border']: - buf.append(u'\\chbrdr\\chcfpat%d' % + buf.append('\\chbrdr\\chcfpat%d' % color_mapping[style['border']]) - start = u''.join(buf) + start = ''.join(buf) if start: - outfile.write(u'{%s ' % start) + outfile.write('{%s ' % start) outfile.write(self._escape_text(value)) if start: - outfile.write(u'}') + outfile.write('}') - outfile.write(u'}') + outfile.write('}') diff --git a/pygments/lexer.py b/pygments/lexer.py index 3f0df88e..804aafe0 100644 --- a/pygments/lexer.py +++ b/pygments/lexer.py @@ -166,11 +166,11 @@ class Lexer(metaclass=LexerMeta): text = decoded else: text = text.decode(self.encoding) - if text.startswith(u'\ufeff'): - text = text[len(u'\ufeff'):] + if text.startswith('\ufeff'): + text = text[len('\ufeff'):] else: - if text.startswith(u'\ufeff'): - text = text[len(u'\ufeff'):] + if text.startswith('\ufeff'): + text = text[len('\ufeff'):] # text now *is* a unicode string text = text.replace('\r\n', '\n') @@ -663,7 +663,7 @@ class RegexLexer(Lexer, metaclass=RegexLexerMeta): # at EOL, reset state to "root" statestack = ['root'] statetokens = tokendefs['root'] - yield pos, Text, u'\n' + yield pos, Text, '\n' pos += 1 continue yield pos, Error, text[pos] @@ -751,7 +751,7 @@ class ExtendedRegexLexer(RegexLexer): # at EOL, reset state to "root" ctx.stack = ['root'] statetokens = tokendefs['root'] - yield ctx.pos, Text, u'\n' + yield ctx.pos, Text, '\n' ctx.pos += 1 continue yield ctx.pos, Error, text[ctx.pos] diff --git a/pygments/lexers/apl.py b/pygments/lexers/apl.py index 68c4ffe1..4e2de92c 100644 --- a/pygments/lexers/apl.py +++ b/pygments/lexers/apl.py @@ -35,7 +35,7 @@ class APLLexer(RegexLexer): # Comment # ======= # '⍝' is traditional; '#' is supported by GNU APL and NGN (but not Dyalog) - (u'[⍝#].*$', Comment.Single), + (r'[⍝#].*$', Comment.Single), # # Strings # ======= @@ -46,7 +46,7 @@ class APLLexer(RegexLexer): # =========== # This token type is used for diamond and parenthesis # but not for bracket and ; (see below) - (u'[⋄◇()]', Punctuation), + (r'[⋄◇()]', Punctuation), # # Array indexing # ============== @@ -57,45 +57,45 @@ class APLLexer(RegexLexer): # Distinguished names # =================== # following IBM APL2 standard - (u'⎕[A-Za-zΔ∆⍙][A-Za-zΔ∆⍙_¯0-9]*', Name.Function), + (r'⎕[A-Za-zΔ∆⍙][A-Za-zΔ∆⍙_¯0-9]*', Name.Function), # # Labels # ====== # following IBM APL2 standard - # (u'[A-Za-zΔ∆⍙][A-Za-zΔ∆⍙_¯0-9]*:', Name.Label), + # (r'[A-Za-zΔ∆⍙][A-Za-zΔ∆⍙_¯0-9]*:', Name.Label), # # Variables # ========= # following IBM APL2 standard - (u'[A-Za-zΔ∆⍙][A-Za-zΔ∆⍙_¯0-9]*', Name.Variable), + (r'[A-Za-zΔ∆⍙][A-Za-zΔ∆⍙_¯0-9]*', Name.Variable), # # Numbers # ======= - (u'¯?(0[Xx][0-9A-Fa-f]+|[0-9]*\\.?[0-9]+([Ee][+¯]?[0-9]+)?|¯|∞)' - u'([Jj]¯?(0[Xx][0-9A-Fa-f]+|[0-9]*\\.?[0-9]+([Ee][+¯]?[0-9]+)?|¯|∞))?', + (r'¯?(0[Xx][0-9A-Fa-f]+|[0-9]*\.?[0-9]+([Ee][+¯]?[0-9]+)?|¯|∞)' + r'([Jj]¯?(0[Xx][0-9A-Fa-f]+|[0-9]*\.?[0-9]+([Ee][+¯]?[0-9]+)?|¯|∞))?', Number), # # Operators # ========== - (u'[\\.\\\\\\/⌿⍀¨⍣⍨⍠⍤∘⌸&⌶@⌺⍥⍛⍢]', Name.Attribute), # closest token type - (u'[+\\-×÷⌈⌊∣|⍳?*⍟○!⌹<≤=>≥≠≡≢∊⍷∪∩~∨∧⍱⍲⍴,⍪⌽⊖⍉↑↓⊂⊃⌷⍋⍒⊤⊥⍕⍎⊣⊢⍁⍂≈⌸⍯↗⊆⊇⍸√⌾…⍮]', + (r'[\.\\\/⌿⍀¨⍣⍨⍠⍤∘⌸&⌶@⌺⍥⍛⍢]', Name.Attribute), # closest token type + (r'[+\-×÷⌈⌊∣|⍳?*⍟○!⌹<≤=>≥≠≡≢∊⍷∪∩~∨∧⍱⍲⍴,⍪⌽⊖⍉↑↓⊂⊃⌷⍋⍒⊤⊥⍕⍎⊣⊢⍁⍂≈⌸⍯↗⊆⊇⍸√⌾…⍮]', Operator), # # Constant # ======== - (u'⍬', Name.Constant), + (r'⍬', Name.Constant), # # Quad symbol # =========== - (u'[⎕⍞]', Name.Variable.Global), + (r'[⎕⍞]', Name.Variable.Global), # # Arrows left/right # ================= - (u'[←→]', Keyword.Declaration), + (r'[←→]', Keyword.Declaration), # # D-Fn # ==== - (u'[⍺⍵⍶⍹∇:]', Name.Builtin.Pseudo), + (r'[⍺⍵⍶⍹∇:]', Name.Builtin.Pseudo), (r'[{}]', Keyword.Type), ], } diff --git a/pygments/lexers/archetype.py b/pygments/lexers/archetype.py index bca9cbbb..713970ec 100644 --- a/pygments/lexers/archetype.py +++ b/pygments/lexers/archetype.py @@ -212,9 +212,9 @@ class CadlLexer(AtomsLexer): (r'(not)\W', Operator), (r'(matches|is_in)\W', Operator), # is_in / not is_in char - (u'(\u2208|\u2209)', Operator), + ('(\u2208|\u2209)', Operator), # there_exists / not there_exists / for_all / and / or - (u'(\u2203|\u2204|\u2200|\u2227|\u2228|\u22BB|\223C)', + ('(\u2203|\u2204|\u2200|\u2227|\u2228|\u22BB|\223C)', Operator), # regex in slot or as string constraint (r'(\{)(\s*/[^}]+/\s*)(\})', diff --git a/pygments/lexers/erlang.py b/pygments/lexers/erlang.py index fdc83451..829a1c23 100644 --- a/pygments/lexers/erlang.py +++ b/pygments/lexers/erlang.py @@ -504,7 +504,7 @@ class ElixirConsoleLexer(Lexer): insertions = [] for match in line_re.finditer(text): line = match.group() - if line.startswith(u'** '): + if line.startswith('** '): in_error = True insertions.append((len(curcode), [(0, Generic.Error, line[:-1])])) diff --git a/pygments/lexers/esoteric.py b/pygments/lexers/esoteric.py index 6ac3dd9d..e1e02a42 100644 --- a/pygments/lexers/esoteric.py +++ b/pygments/lexers/esoteric.py @@ -255,23 +255,23 @@ class AheuiLexer(RegexLexer): tokens = { 'root': [ - (u'[' - u'나-낳냐-냫너-넣녀-녛노-놓뇨-눟뉴-닇' - u'다-닿댜-댷더-덯뎌-뎧도-돟됴-둫듀-딓' - u'따-땋땨-떃떠-떻뗘-뗳또-똫뚀-뚷뜌-띟' - u'라-랗랴-럏러-렇려-렿로-롷료-뤃류-릫' - u'마-맣먀-먛머-멓며-몋모-뫃묘-뭏뮤-믷' - u'바-밯뱌-뱧버-벟벼-볗보-봏뵤-붛뷰-빃' - u'빠-빻뺘-뺳뻐-뻫뼈-뼣뽀-뽛뾰-뿧쀼-삏' - u'사-샇샤-샿서-섷셔-셯소-솧쇼-숳슈-싛' - u'싸-쌓쌰-썋써-쎃쎠-쎻쏘-쏳쑈-쑿쓔-씧' - u'자-잫쟈-쟣저-젛져-졓조-좋죠-줗쥬-즿' - u'차-챃챠-챻처-첳쳐-쳫초-촣쵸-춯츄-칗' - u'카-캏캬-컇커-컿켜-켷코-콯쿄-쿻큐-킣' - u'타-탛탸-턓터-텋텨-톃토-톻툐-퉇튜-틯' - u'파-팧퍄-퍟퍼-펗펴-폏포-퐇표-풓퓨-픻' - u'하-핳햐-햫허-헣혀-혛호-홓효-훟휴-힇' - u']', Operator), + ('[' + '나-낳냐-냫너-넣녀-녛노-놓뇨-눟뉴-닇' + '다-닿댜-댷더-덯뎌-뎧도-돟됴-둫듀-딓' + '따-땋땨-떃떠-떻뗘-뗳또-똫뚀-뚷뜌-띟' + '라-랗랴-럏러-렇려-렿로-롷료-뤃류-릫' + '마-맣먀-먛머-멓며-몋모-뫃묘-뭏뮤-믷' + '바-밯뱌-뱧버-벟벼-볗보-봏뵤-붛뷰-빃' + '빠-빻뺘-뺳뻐-뻫뼈-뼣뽀-뽛뾰-뿧쀼-삏' + '사-샇샤-샿서-섷셔-셯소-솧쇼-숳슈-싛' + '싸-쌓쌰-썋써-쎃쎠-쎻쏘-쏳쑈-쑿쓔-씧' + '자-잫쟈-쟣저-젛져-졓조-좋죠-줗쥬-즿' + '차-챃챠-챻처-첳쳐-쳫초-촣쵸-춯츄-칗' + '카-캏캬-컇커-컿켜-켷코-콯쿄-쿻큐-킣' + '타-탛탸-턓터-텋텨-톃토-톻툐-퉇튜-틯' + '파-팧퍄-퍟퍼-펗펴-폏포-퐇표-풓퓨-픻' + '하-핳햐-햫허-헣혀-혛호-홓효-훟휴-힇' + ']', Operator), ('.', Comment), ], } diff --git a/pygments/lexers/ezhil.py b/pygments/lexers/ezhil.py index 37d793dd..ea5bba3c 100644 --- a/pygments/lexers/ezhil.py +++ b/pygments/lexers/ezhil.py @@ -30,20 +30,20 @@ class EzhilLexer(RegexLexer): flags = re.MULTILINE | re.UNICODE # Refer to tamil.utf8.tamil_letters from open-tamil for a stricter version of this. # This much simpler version is close enough, and includes combining marks. - _TALETTERS = u'[a-zA-Z_]|[\u0b80-\u0bff]' + _TALETTERS = '[a-zA-Z_]|[\u0b80-\u0bff]' tokens = { 'root': [ include('keywords'), (r'#.*\n', Comment.Single), (r'[@+/*,^\-%]|[!<>=]=?|&&?|\|\|?', Operator), - (u'இல்', Operator.Word), - (words((u'assert', u'max', u'min', - u'நீளம்', u'சரம்_இடமாற்று', u'சரம்_கண்டுபிடி', - u'பட்டியல்', u'பின்இணை', u'வரிசைப்படுத்து', - u'எடு', u'தலைகீழ்', u'நீட்டிக்க', u'நுழைக்க', u'வை', - u'கோப்பை_திற', u'கோப்பை_எழுது', u'கோப்பை_மூடு', - u'pi', u'sin', u'cos', u'tan', u'sqrt', u'hypot', u'pow', - u'exp', u'log', u'log10', u'exit', + ('இல்', Operator.Word), + (words(('assert', 'max', 'min', + 'நீளம்', 'சரம்_இடமாற்று', 'சரம்_கண்டுபிடி', + 'பட்டியல்', 'பின்இணை', 'வரிசைப்படுத்து', + 'எடு', 'தலைகீழ்', 'நீட்டிக்க', 'நுழைக்க', 'வை', + 'கோப்பை_திற', 'கோப்பை_எழுது', 'கோப்பை_மூடு', + 'pi', 'sin', 'cos', 'tan', 'sqrt', 'hypot', 'pow', + 'exp', 'log', 'log10', 'exit', ), suffix=r'\b'), Name.Builtin), (r'(True|False)\b', Keyword.Constant), (r'[^\S\n]+', Text), @@ -52,10 +52,10 @@ class EzhilLexer(RegexLexer): (r'[(){}\[\]:;.]', Punctuation), ], 'keywords': [ - (u'பதிப்பி|தேர்ந்தெடு|தேர்வு|ஏதேனில்|ஆனால்|இல்லைஆனால்|இல்லை|ஆக|ஒவ்வொன்றாக|இல்|வரை|செய்|முடியேனில்|பின்கொடு|முடி|நிரல்பாகம்|தொடர்|நிறுத்து|நிரல்பாகம்', Keyword), + ('பதிப்பி|தேர்ந்தெடு|தேர்வு|ஏதேனில்|ஆனால்|இல்லைஆனால்|இல்லை|ஆக|ஒவ்வொன்றாக|இல்|வரை|செய்|முடியேனில்|பின்கொடு|முடி|நிரல்பாகம்|தொடர்|நிறுத்து|நிரல்பாகம்', Keyword), ], 'identifier': [ - (u'(?:'+_TALETTERS+u')(?:[0-9]|'+_TALETTERS+u')*', Name), + ('(?:'+_TALETTERS+')(?:[0-9]|'+_TALETTERS+')*', Name), ], 'literal': [ (r'".*?"', String), diff --git a/pygments/lexers/haskell.py b/pygments/lexers/haskell.py index da86a507..497d4cf3 100644 --- a/pygments/lexers/haskell.py +++ b/pygments/lexers/haskell.py @@ -327,10 +327,10 @@ class AgdaLexer(RegexLexer): # Identifiers (r'\b(%s)(?!\')\b' % '|'.join(reserved), Keyword.Reserved), (r'(import|module)(\s+)', bygroups(Keyword.Reserved, Text), 'module'), - (u'\\b(Set|Prop)[\u2080-\u2089]*\\b', Keyword.Type), + (r'\b(Set|Prop)[\u2080-\u2089]*\b', Keyword.Type), # Special Symbols (r'(\(|\)|\{|\})', Operator), - (u'(\\.{1,3}|\\||\u03BB|\u2200|\u2192|:|=|->)', Operator.Word), + (r'(\.{1,3}|\||\u03BB|\u2200|\u2192|:|=|->)', Operator.Word), # Numbers (r'\d+[eE][+-]?\d+', Number.Float), (r'\d+\.\d+([eE][+-]?\d+)?', Number.Float), diff --git a/pygments/lexers/haxe.py b/pygments/lexers/haxe.py index 0a58aefc..f95e4556 100644 --- a/pygments/lexers/haxe.py +++ b/pygments/lexers/haxe.py @@ -79,7 +79,7 @@ class HaxeLexer(ExtendedRegexLexer): if proc in ['error']: ctx.stack.append('preproc-error') - yield match.start(), Comment.Preproc, u'#' + proc + yield match.start(), Comment.Preproc, '#' + proc ctx.pos = match.end() tokens = { diff --git a/pygments/lexers/int_fiction.py b/pygments/lexers/int_fiction.py index e3d35088..a9194415 100644 --- a/pygments/lexers/int_fiction.py +++ b/pygments/lexers/int_fiction.py @@ -38,10 +38,10 @@ class Inform6Lexer(RegexLexer): # Inform 7 maps these four character classes to their ASCII # equivalents. To support Inform 6 inclusions within Inform 7, # Inform6Lexer maps them too. - _dash = u'\\-\u2010-\u2014' - _dquote = u'"\u201c\u201d' - _squote = u"'\u2018\u2019" - _newline = u'\\n\u0085\u2028\u2029' + _dash = '\\-\u2010-\u2014' + _dquote = '"\u201c\u201d' + _squote = "'\u2018\u2019" + _newline = '\\n\u0085\u2028\u2029' tokens = { 'root': [ @@ -858,7 +858,7 @@ class Tads3Lexer(RegexLexer): tokens = { 'root': [ - (u'\ufeff', Text), + ('\ufeff', Text), (r'\{', Punctuation, 'object-body'), (r';+', Punctuation), (r'(?=(argcount|break|case|catch|continue|default|definingobj|' diff --git a/pygments/lexers/javascript.py b/pygments/lexers/javascript.py index f36863dc..0a689eab 100644 --- a/pygments/lexers/javascript.py +++ b/pygments/lexers/javascript.py @@ -26,7 +26,7 @@ JS_IDENT_START = ('(?:[$_' + uni.combine('Lu', 'Ll', 'Lt', 'Lm', 'Lo', 'Nl') + ']|\\\\u[a-fA-F0-9]{4})') JS_IDENT_PART = ('(?:[$' + uni.combine('Lu', 'Ll', 'Lt', 'Lm', 'Lo', 'Nl', 'Mn', 'Mc', 'Nd', 'Pc') + - u'\u200c\u200d]|\\\\u[a-fA-F0-9]{4})') + '\u200c\u200d]|\\\\u[a-fA-F0-9]{4})') JS_IDENT = JS_IDENT_START + '(?:' + JS_IDENT_PART + ')*' diff --git a/pygments/lexers/julia.py b/pygments/lexers/julia.py index 7978074a..b512e248 100644 --- a/pygments/lexers/julia.py +++ b/pygments/lexers/julia.py @@ -15,13 +15,12 @@ from pygments.lexer import Lexer, RegexLexer, bygroups, do_insertions, \ words, include from pygments.token import Text, Comment, Operator, Keyword, Name, String, \ Number, Punctuation, Generic -from pygments.util import shebang_matches, unirange +from pygments.util import shebang_matches __all__ = ['JuliaLexer', 'JuliaConsoleLexer'] -allowed_variable = ( - u'(?:[a-zA-Z_\u00A1-\uffff]|%s)(?:[a-zA-Z_0-9\u00A1-\uffff]|%s)*!*' % - ((unirange(0x10000, 0x10ffff),) * 2)) +allowed_variable = \ + '(?:[a-zA-Z_\u00A1-\U0010ffff]|%s)(?:[a-zA-Z_0-9\u00A1-\U0010ffff])*!*' class JuliaLexer(RegexLexer): @@ -131,59 +130,58 @@ class JuliaLexer(RegexLexer): # builtins (words([ - u'ARGS', u'CPU_CORES', u'C_NULL', u'DevNull', u'ENDIAN_BOM', - u'ENV', u'I', u'Inf', u'Inf16', u'Inf32', u'Inf64', - u'InsertionSort', u'JULIA_HOME', u'LOAD_PATH', u'MergeSort', - u'NaN', u'NaN16', u'NaN32', u'NaN64', u'OS_NAME', - u'QuickSort', u'RoundDown', u'RoundFromZero', u'RoundNearest', - u'RoundNearestTiesAway', u'RoundNearestTiesUp', - u'RoundToZero', u'RoundUp', u'STDERR', u'STDIN', u'STDOUT', - u'VERSION', u'WORD_SIZE', u'catalan', u'e', u'eu', - u'eulergamma', u'golden', u'im', u'nothing', u'pi', u'γ', - u'π', u'φ'], + 'ARGS', 'CPU_CORES', 'C_NULL', 'DevNull', 'ENDIAN_BOM', + 'ENV', 'I', 'Inf', 'Inf16', 'Inf32', 'Inf64', + 'InsertionSort', 'JULIA_HOME', 'LOAD_PATH', 'MergeSort', + 'NaN', 'NaN16', 'NaN32', 'NaN64', 'OS_NAME', + 'QuickSort', 'RoundDown', 'RoundFromZero', 'RoundNearest', + 'RoundNearestTiesAway', 'RoundNearestTiesUp', + 'RoundToZero', 'RoundUp', 'STDERR', 'STDIN', 'STDOUT', + 'VERSION', 'WORD_SIZE', 'catalan', 'e', 'eu', + 'eulergamma', 'golden', 'im', 'nothing', 'pi', 'γ', 'π', 'φ'], suffix=r'\b'), Name.Builtin), # operators # see: https://github.com/JuliaLang/julia/blob/master/src/julia-parser.scm (words(( # prec-assignment - u'=', u':=', u'+=', u'-=', u'*=', u'/=', u'//=', u'.//=', u'.*=', u'./=', - u'\\=', u'.\\=', u'^=', u'.^=', u'÷=', u'.÷=', u'%=', u'.%=', u'|=', u'&=', - u'$=', u'=>', u'<<=', u'>>=', u'>>>=', u'~', u'.+=', u'.-=', + '=', ':=', '+=', '-=', '*=', '/=', '//=', './/=', '.*=', './=', + '\\=', '.\\=', '^=', '.^=', '÷=', '.÷=', '%=', '.%=', '|=', '&=', + '$=', '=>', '<<=', '>>=', '>>>=', '~', '.+=', '.-=', # prec-conditional - u'?', + '?', # prec-arrow - u'--', u'-->', + '--', '-->', # prec-lazy-or - u'||', + '||', # prec-lazy-and - u'&&', + '&&', # prec-comparison - u'>', u'<', u'>=', u'≥', u'<=', u'≤', u'==', u'===', u'≡', u'!=', u'≠', - u'!==', u'≢', u'.>', u'.<', u'.>=', u'.≥', u'.<=', u'.≤', u'.==', u'.!=', - u'.≠', u'.=', u'.!', u'<:', u'>:', u'∈', u'∉', u'∋', u'∌', u'⊆', - u'⊈', u'⊂', - u'⊄', u'⊊', + '>', '<', '>=', '≥', '<=', '≤', '==', '===', '≡', '!=', '≠', + '!==', '≢', '.>', '.<', '.>=', '.≥', '.<=', '.≤', '.==', '.!=', + '.≠', '.=', '.!', '<:', '>:', '∈', '∉', '∋', '∌', '⊆', + '⊈', '⊂', + '⊄', '⊊', # prec-pipe - u'|>', u'<|', + '|>', '<|', # prec-colon - u':', + ':', # prec-plus - u'.+', u'.-', u'|', u'∪', u'$', + '.+', '.-', '|', '∪', '$', # prec-bitshift - u'<<', u'>>', u'>>>', u'.<<', u'.>>', u'.>>>', + '<<', '>>', '>>>', '.<<', '.>>', '.>>>', # prec-times - u'*', u'/', u'./', u'÷', u'.÷', u'%', u'⋅', u'.%', u'.*', u'\\', u'.\\', u'&', u'∩', + '*', '/', './', '÷', '.÷', '%', '⋅', '.%', '.*', '\\', '.\\', '&', '∩', # prec-rational - u'//', u'.//', + '//', './/', # prec-power - u'^', u'.^', + '^', '.^', # prec-decl - u'::', + '::', # prec-dot - u'.', + '.', # unary op - u'+', u'-', u'!', u'√', u'∛', u'∜' + '+', '-', '!', '√', '∛', '∜', )), Operator), # chars diff --git a/pygments/lexers/jvm.py b/pygments/lexers/jvm.py index 86af6d9f..ee0bc7af 100644 --- a/pygments/lexers/jvm.py +++ b/pygments/lexers/jvm.py @@ -151,120 +151,120 @@ class ScalaLexer(RegexLexer): flags = re.MULTILINE | re.DOTALL # don't use raw unicode strings! - op = (u'[-~\\^\\*!%&\\\\<>\\|+=:/?@\u00a6-\u00a7\u00a9\u00ac\u00ae\u00b0-\u00b1' - u'\u00b6\u00d7\u00f7\u03f6\u0482\u0606-\u0608\u060e-\u060f\u06e9' - u'\u06fd-\u06fe\u07f6\u09fa\u0b70\u0bf3-\u0bf8\u0bfa\u0c7f\u0cf1-\u0cf2' - u'\u0d79\u0f01-\u0f03\u0f13-\u0f17\u0f1a-\u0f1f\u0f34\u0f36\u0f38' - u'\u0fbe-\u0fc5\u0fc7-\u0fcf\u109e-\u109f\u1360\u1390-\u1399\u1940' - u'\u19e0-\u19ff\u1b61-\u1b6a\u1b74-\u1b7c\u2044\u2052\u207a-\u207c' - u'\u208a-\u208c\u2100-\u2101\u2103-\u2106\u2108-\u2109\u2114\u2116-\u2118' - u'\u211e-\u2123\u2125\u2127\u2129\u212e\u213a-\u213b\u2140-\u2144' - u'\u214a-\u214d\u214f\u2190-\u2328\u232b-\u244a\u249c-\u24e9\u2500-\u2767' - u'\u2794-\u27c4\u27c7-\u27e5\u27f0-\u2982\u2999-\u29d7\u29dc-\u29fb' - u'\u29fe-\u2b54\u2ce5-\u2cea\u2e80-\u2ffb\u3004\u3012-\u3013\u3020' - u'\u3036-\u3037\u303e-\u303f\u3190-\u3191\u3196-\u319f\u31c0-\u31e3' - u'\u3200-\u321e\u322a-\u3250\u3260-\u327f\u328a-\u32b0\u32c0-\u33ff' - u'\u4dc0-\u4dff\ua490-\ua4c6\ua828-\ua82b\ufb29\ufdfd\ufe62\ufe64-\ufe66' - u'\uff0b\uff1c-\uff1e\uff5c\uff5e\uffe2\uffe4\uffe8-\uffee\ufffc-\ufffd]+') - - letter = (u'[a-zA-Z\\$_\u00aa\u00b5\u00ba\u00c0-\u00d6\u00d8-\u00f6' - u'\u00f8-\u02af\u0370-\u0373\u0376-\u0377\u037b-\u037d\u0386' - u'\u0388-\u03f5\u03f7-\u0481\u048a-\u0556\u0561-\u0587\u05d0-\u05f2' - u'\u0621-\u063f\u0641-\u064a\u066e-\u066f\u0671-\u06d3\u06d5' - u'\u06ee-\u06ef\u06fa-\u06fc\u06ff\u0710\u0712-\u072f\u074d-\u07a5' - u'\u07b1\u07ca-\u07ea\u0904-\u0939\u093d\u0950\u0958-\u0961' - u'\u0972-\u097f\u0985-\u09b9\u09bd\u09ce\u09dc-\u09e1\u09f0-\u09f1' - u'\u0a05-\u0a39\u0a59-\u0a5e\u0a72-\u0a74\u0a85-\u0ab9\u0abd' - u'\u0ad0-\u0ae1\u0b05-\u0b39\u0b3d\u0b5c-\u0b61\u0b71\u0b83-\u0bb9' - u'\u0bd0\u0c05-\u0c3d\u0c58-\u0c61\u0c85-\u0cb9\u0cbd\u0cde-\u0ce1' - u'\u0d05-\u0d3d\u0d60-\u0d61\u0d7a-\u0d7f\u0d85-\u0dc6\u0e01-\u0e30' - u'\u0e32-\u0e33\u0e40-\u0e45\u0e81-\u0eb0\u0eb2-\u0eb3\u0ebd-\u0ec4' - u'\u0edc-\u0f00\u0f40-\u0f6c\u0f88-\u0f8b\u1000-\u102a\u103f' - u'\u1050-\u1055\u105a-\u105d\u1061\u1065-\u1066\u106e-\u1070' - u'\u1075-\u1081\u108e\u10a0-\u10fa\u1100-\u135a\u1380-\u138f' - u'\u13a0-\u166c\u166f-\u1676\u1681-\u169a\u16a0-\u16ea\u16ee-\u1711' - u'\u1720-\u1731\u1740-\u1751\u1760-\u1770\u1780-\u17b3\u17dc' - u'\u1820-\u1842\u1844-\u18a8\u18aa-\u191c\u1950-\u19a9\u19c1-\u19c7' - u'\u1a00-\u1a16\u1b05-\u1b33\u1b45-\u1b4b\u1b83-\u1ba0\u1bae-\u1baf' - u'\u1c00-\u1c23\u1c4d-\u1c4f\u1c5a-\u1c77\u1d00-\u1d2b\u1d62-\u1d77' - u'\u1d79-\u1d9a\u1e00-\u1fbc\u1fbe\u1fc2-\u1fcc\u1fd0-\u1fdb' - u'\u1fe0-\u1fec\u1ff2-\u1ffc\u2071\u207f\u2102\u2107\u210a-\u2113' - u'\u2115\u2119-\u211d\u2124\u2126\u2128\u212a-\u212d\u212f-\u2139' - u'\u213c-\u213f\u2145-\u2149\u214e\u2160-\u2188\u2c00-\u2c7c' - u'\u2c80-\u2ce4\u2d00-\u2d65\u2d80-\u2dde\u3006-\u3007\u3021-\u3029' - u'\u3038-\u303a\u303c\u3041-\u3096\u309f\u30a1-\u30fa\u30ff-\u318e' - u'\u31a0-\u31b7\u31f0-\u31ff\u3400-\u4db5\u4e00-\ua014\ua016-\ua48c' - u'\ua500-\ua60b\ua610-\ua61f\ua62a-\ua66e\ua680-\ua697\ua722-\ua76f' - u'\ua771-\ua787\ua78b-\ua801\ua803-\ua805\ua807-\ua80a\ua80c-\ua822' - u'\ua840-\ua873\ua882-\ua8b3\ua90a-\ua925\ua930-\ua946\uaa00-\uaa28' - u'\uaa40-\uaa42\uaa44-\uaa4b\uac00-\ud7a3\uf900-\ufb1d\ufb1f-\ufb28' - u'\ufb2a-\ufd3d\ufd50-\ufdfb\ufe70-\ufefc\uff21-\uff3a\uff41-\uff5a' - u'\uff66-\uff6f\uff71-\uff9d\uffa0-\uffdc]') - - upper = (u'[A-Z\\$_\u00c0-\u00d6\u00d8-\u00de\u0100\u0102\u0104\u0106\u0108' - u'\u010a\u010c\u010e\u0110\u0112\u0114\u0116\u0118\u011a\u011c' - u'\u011e\u0120\u0122\u0124\u0126\u0128\u012a\u012c\u012e\u0130' - u'\u0132\u0134\u0136\u0139\u013b\u013d\u013f\u0141\u0143\u0145' - u'\u0147\u014a\u014c\u014e\u0150\u0152\u0154\u0156\u0158\u015a' - u'\u015c\u015e\u0160\u0162\u0164\u0166\u0168\u016a\u016c\u016e' - u'\u0170\u0172\u0174\u0176\u0178-\u0179\u017b\u017d\u0181-\u0182' - u'\u0184\u0186-\u0187\u0189-\u018b\u018e-\u0191\u0193-\u0194' - u'\u0196-\u0198\u019c-\u019d\u019f-\u01a0\u01a2\u01a4\u01a6-\u01a7' - u'\u01a9\u01ac\u01ae-\u01af\u01b1-\u01b3\u01b5\u01b7-\u01b8\u01bc' - u'\u01c4\u01c7\u01ca\u01cd\u01cf\u01d1\u01d3\u01d5\u01d7\u01d9' - u'\u01db\u01de\u01e0\u01e2\u01e4\u01e6\u01e8\u01ea\u01ec\u01ee' - u'\u01f1\u01f4\u01f6-\u01f8\u01fa\u01fc\u01fe\u0200\u0202\u0204' - u'\u0206\u0208\u020a\u020c\u020e\u0210\u0212\u0214\u0216\u0218' - u'\u021a\u021c\u021e\u0220\u0222\u0224\u0226\u0228\u022a\u022c' - u'\u022e\u0230\u0232\u023a-\u023b\u023d-\u023e\u0241\u0243-\u0246' - u'\u0248\u024a\u024c\u024e\u0370\u0372\u0376\u0386\u0388-\u038f' - u'\u0391-\u03ab\u03cf\u03d2-\u03d4\u03d8\u03da\u03dc\u03de\u03e0' - u'\u03e2\u03e4\u03e6\u03e8\u03ea\u03ec\u03ee\u03f4\u03f7' - u'\u03f9-\u03fa\u03fd-\u042f\u0460\u0462\u0464\u0466\u0468\u046a' - u'\u046c\u046e\u0470\u0472\u0474\u0476\u0478\u047a\u047c\u047e' - u'\u0480\u048a\u048c\u048e\u0490\u0492\u0494\u0496\u0498\u049a' - u'\u049c\u049e\u04a0\u04a2\u04a4\u04a6\u04a8\u04aa\u04ac\u04ae' - u'\u04b0\u04b2\u04b4\u04b6\u04b8\u04ba\u04bc\u04be\u04c0-\u04c1' - u'\u04c3\u04c5\u04c7\u04c9\u04cb\u04cd\u04d0\u04d2\u04d4\u04d6' - u'\u04d8\u04da\u04dc\u04de\u04e0\u04e2\u04e4\u04e6\u04e8\u04ea' - u'\u04ec\u04ee\u04f0\u04f2\u04f4\u04f6\u04f8\u04fa\u04fc\u04fe' - u'\u0500\u0502\u0504\u0506\u0508\u050a\u050c\u050e\u0510\u0512' - u'\u0514\u0516\u0518\u051a\u051c\u051e\u0520\u0522\u0531-\u0556' - u'\u10a0-\u10c5\u1e00\u1e02\u1e04\u1e06\u1e08\u1e0a\u1e0c\u1e0e' - u'\u1e10\u1e12\u1e14\u1e16\u1e18\u1e1a\u1e1c\u1e1e\u1e20\u1e22' - u'\u1e24\u1e26\u1e28\u1e2a\u1e2c\u1e2e\u1e30\u1e32\u1e34\u1e36' - u'\u1e38\u1e3a\u1e3c\u1e3e\u1e40\u1e42\u1e44\u1e46\u1e48\u1e4a' - u'\u1e4c\u1e4e\u1e50\u1e52\u1e54\u1e56\u1e58\u1e5a\u1e5c\u1e5e' - u'\u1e60\u1e62\u1e64\u1e66\u1e68\u1e6a\u1e6c\u1e6e\u1e70\u1e72' - u'\u1e74\u1e76\u1e78\u1e7a\u1e7c\u1e7e\u1e80\u1e82\u1e84\u1e86' - u'\u1e88\u1e8a\u1e8c\u1e8e\u1e90\u1e92\u1e94\u1e9e\u1ea0\u1ea2' - u'\u1ea4\u1ea6\u1ea8\u1eaa\u1eac\u1eae\u1eb0\u1eb2\u1eb4\u1eb6' - u'\u1eb8\u1eba\u1ebc\u1ebe\u1ec0\u1ec2\u1ec4\u1ec6\u1ec8\u1eca' - u'\u1ecc\u1ece\u1ed0\u1ed2\u1ed4\u1ed6\u1ed8\u1eda\u1edc\u1ede' - u'\u1ee0\u1ee2\u1ee4\u1ee6\u1ee8\u1eea\u1eec\u1eee\u1ef0\u1ef2' - u'\u1ef4\u1ef6\u1ef8\u1efa\u1efc\u1efe\u1f08-\u1f0f\u1f18-\u1f1d' - u'\u1f28-\u1f2f\u1f38-\u1f3f\u1f48-\u1f4d\u1f59-\u1f5f' - u'\u1f68-\u1f6f\u1fb8-\u1fbb\u1fc8-\u1fcb\u1fd8-\u1fdb' - u'\u1fe8-\u1fec\u1ff8-\u1ffb\u2102\u2107\u210b-\u210d\u2110-\u2112' - u'\u2115\u2119-\u211d\u2124\u2126\u2128\u212a-\u212d\u2130-\u2133' - u'\u213e-\u213f\u2145\u2183\u2c00-\u2c2e\u2c60\u2c62-\u2c64\u2c67' - u'\u2c69\u2c6b\u2c6d-\u2c6f\u2c72\u2c75\u2c80\u2c82\u2c84\u2c86' - u'\u2c88\u2c8a\u2c8c\u2c8e\u2c90\u2c92\u2c94\u2c96\u2c98\u2c9a' - u'\u2c9c\u2c9e\u2ca0\u2ca2\u2ca4\u2ca6\u2ca8\u2caa\u2cac\u2cae' - u'\u2cb0\u2cb2\u2cb4\u2cb6\u2cb8\u2cba\u2cbc\u2cbe\u2cc0\u2cc2' - u'\u2cc4\u2cc6\u2cc8\u2cca\u2ccc\u2cce\u2cd0\u2cd2\u2cd4\u2cd6' - u'\u2cd8\u2cda\u2cdc\u2cde\u2ce0\u2ce2\ua640\ua642\ua644\ua646' - u'\ua648\ua64a\ua64c\ua64e\ua650\ua652\ua654\ua656\ua658\ua65a' - u'\ua65c\ua65e\ua662\ua664\ua666\ua668\ua66a\ua66c\ua680\ua682' - u'\ua684\ua686\ua688\ua68a\ua68c\ua68e\ua690\ua692\ua694\ua696' - u'\ua722\ua724\ua726\ua728\ua72a\ua72c\ua72e\ua732\ua734\ua736' - u'\ua738\ua73a\ua73c\ua73e\ua740\ua742\ua744\ua746\ua748\ua74a' - u'\ua74c\ua74e\ua750\ua752\ua754\ua756\ua758\ua75a\ua75c\ua75e' - u'\ua760\ua762\ua764\ua766\ua768\ua76a\ua76c\ua76e\ua779\ua77b' - u'\ua77d-\ua77e\ua780\ua782\ua784\ua786\ua78b\uff21-\uff3a]') - - idrest = u'%s(?:%s|[0-9])*(?:(?<=_)%s)?' % (letter, letter, op) - letter_letter_digit = u'%s(?:%s|\\d)*' % (letter, letter) + op = ('[-~\\^\\*!%&\\\\<>\\|+=:/?@\u00a6-\u00a7\u00a9\u00ac\u00ae\u00b0-\u00b1' + '\u00b6\u00d7\u00f7\u03f6\u0482\u0606-\u0608\u060e-\u060f\u06e9' + '\u06fd-\u06fe\u07f6\u09fa\u0b70\u0bf3-\u0bf8\u0bfa\u0c7f\u0cf1-\u0cf2' + '\u0d79\u0f01-\u0f03\u0f13-\u0f17\u0f1a-\u0f1f\u0f34\u0f36\u0f38' + '\u0fbe-\u0fc5\u0fc7-\u0fcf\u109e-\u109f\u1360\u1390-\u1399\u1940' + '\u19e0-\u19ff\u1b61-\u1b6a\u1b74-\u1b7c\u2044\u2052\u207a-\u207c' + '\u208a-\u208c\u2100-\u2101\u2103-\u2106\u2108-\u2109\u2114\u2116-\u2118' + '\u211e-\u2123\u2125\u2127\u2129\u212e\u213a-\u213b\u2140-\u2144' + '\u214a-\u214d\u214f\u2190-\u2328\u232b-\u244a\u249c-\u24e9\u2500-\u2767' + '\u2794-\u27c4\u27c7-\u27e5\u27f0-\u2982\u2999-\u29d7\u29dc-\u29fb' + '\u29fe-\u2b54\u2ce5-\u2cea\u2e80-\u2ffb\u3004\u3012-\u3013\u3020' + '\u3036-\u3037\u303e-\u303f\u3190-\u3191\u3196-\u319f\u31c0-\u31e3' + '\u3200-\u321e\u322a-\u3250\u3260-\u327f\u328a-\u32b0\u32c0-\u33ff' + '\u4dc0-\u4dff\ua490-\ua4c6\ua828-\ua82b\ufb29\ufdfd\ufe62\ufe64-\ufe66' + '\uff0b\uff1c-\uff1e\uff5c\uff5e\uffe2\uffe4\uffe8-\uffee\ufffc-\ufffd]+') + + letter = ('[a-zA-Z\\$_\u00aa\u00b5\u00ba\u00c0-\u00d6\u00d8-\u00f6' + '\u00f8-\u02af\u0370-\u0373\u0376-\u0377\u037b-\u037d\u0386' + '\u0388-\u03f5\u03f7-\u0481\u048a-\u0556\u0561-\u0587\u05d0-\u05f2' + '\u0621-\u063f\u0641-\u064a\u066e-\u066f\u0671-\u06d3\u06d5' + '\u06ee-\u06ef\u06fa-\u06fc\u06ff\u0710\u0712-\u072f\u074d-\u07a5' + '\u07b1\u07ca-\u07ea\u0904-\u0939\u093d\u0950\u0958-\u0961' + '\u0972-\u097f\u0985-\u09b9\u09bd\u09ce\u09dc-\u09e1\u09f0-\u09f1' + '\u0a05-\u0a39\u0a59-\u0a5e\u0a72-\u0a74\u0a85-\u0ab9\u0abd' + '\u0ad0-\u0ae1\u0b05-\u0b39\u0b3d\u0b5c-\u0b61\u0b71\u0b83-\u0bb9' + '\u0bd0\u0c05-\u0c3d\u0c58-\u0c61\u0c85-\u0cb9\u0cbd\u0cde-\u0ce1' + '\u0d05-\u0d3d\u0d60-\u0d61\u0d7a-\u0d7f\u0d85-\u0dc6\u0e01-\u0e30' + '\u0e32-\u0e33\u0e40-\u0e45\u0e81-\u0eb0\u0eb2-\u0eb3\u0ebd-\u0ec4' + '\u0edc-\u0f00\u0f40-\u0f6c\u0f88-\u0f8b\u1000-\u102a\u103f' + '\u1050-\u1055\u105a-\u105d\u1061\u1065-\u1066\u106e-\u1070' + '\u1075-\u1081\u108e\u10a0-\u10fa\u1100-\u135a\u1380-\u138f' + '\u13a0-\u166c\u166f-\u1676\u1681-\u169a\u16a0-\u16ea\u16ee-\u1711' + '\u1720-\u1731\u1740-\u1751\u1760-\u1770\u1780-\u17b3\u17dc' + '\u1820-\u1842\u1844-\u18a8\u18aa-\u191c\u1950-\u19a9\u19c1-\u19c7' + '\u1a00-\u1a16\u1b05-\u1b33\u1b45-\u1b4b\u1b83-\u1ba0\u1bae-\u1baf' + '\u1c00-\u1c23\u1c4d-\u1c4f\u1c5a-\u1c77\u1d00-\u1d2b\u1d62-\u1d77' + '\u1d79-\u1d9a\u1e00-\u1fbc\u1fbe\u1fc2-\u1fcc\u1fd0-\u1fdb' + '\u1fe0-\u1fec\u1ff2-\u1ffc\u2071\u207f\u2102\u2107\u210a-\u2113' + '\u2115\u2119-\u211d\u2124\u2126\u2128\u212a-\u212d\u212f-\u2139' + '\u213c-\u213f\u2145-\u2149\u214e\u2160-\u2188\u2c00-\u2c7c' + '\u2c80-\u2ce4\u2d00-\u2d65\u2d80-\u2dde\u3006-\u3007\u3021-\u3029' + '\u3038-\u303a\u303c\u3041-\u3096\u309f\u30a1-\u30fa\u30ff-\u318e' + '\u31a0-\u31b7\u31f0-\u31ff\u3400-\u4db5\u4e00-\ua014\ua016-\ua48c' + '\ua500-\ua60b\ua610-\ua61f\ua62a-\ua66e\ua680-\ua697\ua722-\ua76f' + '\ua771-\ua787\ua78b-\ua801\ua803-\ua805\ua807-\ua80a\ua80c-\ua822' + '\ua840-\ua873\ua882-\ua8b3\ua90a-\ua925\ua930-\ua946\uaa00-\uaa28' + '\uaa40-\uaa42\uaa44-\uaa4b\uac00-\ud7a3\uf900-\ufb1d\ufb1f-\ufb28' + '\ufb2a-\ufd3d\ufd50-\ufdfb\ufe70-\ufefc\uff21-\uff3a\uff41-\uff5a' + '\uff66-\uff6f\uff71-\uff9d\uffa0-\uffdc]') + + upper = ('[A-Z\\$_\u00c0-\u00d6\u00d8-\u00de\u0100\u0102\u0104\u0106\u0108' + '\u010a\u010c\u010e\u0110\u0112\u0114\u0116\u0118\u011a\u011c' + '\u011e\u0120\u0122\u0124\u0126\u0128\u012a\u012c\u012e\u0130' + '\u0132\u0134\u0136\u0139\u013b\u013d\u013f\u0141\u0143\u0145' + '\u0147\u014a\u014c\u014e\u0150\u0152\u0154\u0156\u0158\u015a' + '\u015c\u015e\u0160\u0162\u0164\u0166\u0168\u016a\u016c\u016e' + '\u0170\u0172\u0174\u0176\u0178-\u0179\u017b\u017d\u0181-\u0182' + '\u0184\u0186-\u0187\u0189-\u018b\u018e-\u0191\u0193-\u0194' + '\u0196-\u0198\u019c-\u019d\u019f-\u01a0\u01a2\u01a4\u01a6-\u01a7' + '\u01a9\u01ac\u01ae-\u01af\u01b1-\u01b3\u01b5\u01b7-\u01b8\u01bc' + '\u01c4\u01c7\u01ca\u01cd\u01cf\u01d1\u01d3\u01d5\u01d7\u01d9' + '\u01db\u01de\u01e0\u01e2\u01e4\u01e6\u01e8\u01ea\u01ec\u01ee' + '\u01f1\u01f4\u01f6-\u01f8\u01fa\u01fc\u01fe\u0200\u0202\u0204' + '\u0206\u0208\u020a\u020c\u020e\u0210\u0212\u0214\u0216\u0218' + '\u021a\u021c\u021e\u0220\u0222\u0224\u0226\u0228\u022a\u022c' + '\u022e\u0230\u0232\u023a-\u023b\u023d-\u023e\u0241\u0243-\u0246' + '\u0248\u024a\u024c\u024e\u0370\u0372\u0376\u0386\u0388-\u038f' + '\u0391-\u03ab\u03cf\u03d2-\u03d4\u03d8\u03da\u03dc\u03de\u03e0' + '\u03e2\u03e4\u03e6\u03e8\u03ea\u03ec\u03ee\u03f4\u03f7' + '\u03f9-\u03fa\u03fd-\u042f\u0460\u0462\u0464\u0466\u0468\u046a' + '\u046c\u046e\u0470\u0472\u0474\u0476\u0478\u047a\u047c\u047e' + '\u0480\u048a\u048c\u048e\u0490\u0492\u0494\u0496\u0498\u049a' + '\u049c\u049e\u04a0\u04a2\u04a4\u04a6\u04a8\u04aa\u04ac\u04ae' + '\u04b0\u04b2\u04b4\u04b6\u04b8\u04ba\u04bc\u04be\u04c0-\u04c1' + '\u04c3\u04c5\u04c7\u04c9\u04cb\u04cd\u04d0\u04d2\u04d4\u04d6' + '\u04d8\u04da\u04dc\u04de\u04e0\u04e2\u04e4\u04e6\u04e8\u04ea' + '\u04ec\u04ee\u04f0\u04f2\u04f4\u04f6\u04f8\u04fa\u04fc\u04fe' + '\u0500\u0502\u0504\u0506\u0508\u050a\u050c\u050e\u0510\u0512' + '\u0514\u0516\u0518\u051a\u051c\u051e\u0520\u0522\u0531-\u0556' + '\u10a0-\u10c5\u1e00\u1e02\u1e04\u1e06\u1e08\u1e0a\u1e0c\u1e0e' + '\u1e10\u1e12\u1e14\u1e16\u1e18\u1e1a\u1e1c\u1e1e\u1e20\u1e22' + '\u1e24\u1e26\u1e28\u1e2a\u1e2c\u1e2e\u1e30\u1e32\u1e34\u1e36' + '\u1e38\u1e3a\u1e3c\u1e3e\u1e40\u1e42\u1e44\u1e46\u1e48\u1e4a' + '\u1e4c\u1e4e\u1e50\u1e52\u1e54\u1e56\u1e58\u1e5a\u1e5c\u1e5e' + '\u1e60\u1e62\u1e64\u1e66\u1e68\u1e6a\u1e6c\u1e6e\u1e70\u1e72' + '\u1e74\u1e76\u1e78\u1e7a\u1e7c\u1e7e\u1e80\u1e82\u1e84\u1e86' + '\u1e88\u1e8a\u1e8c\u1e8e\u1e90\u1e92\u1e94\u1e9e\u1ea0\u1ea2' + '\u1ea4\u1ea6\u1ea8\u1eaa\u1eac\u1eae\u1eb0\u1eb2\u1eb4\u1eb6' + '\u1eb8\u1eba\u1ebc\u1ebe\u1ec0\u1ec2\u1ec4\u1ec6\u1ec8\u1eca' + '\u1ecc\u1ece\u1ed0\u1ed2\u1ed4\u1ed6\u1ed8\u1eda\u1edc\u1ede' + '\u1ee0\u1ee2\u1ee4\u1ee6\u1ee8\u1eea\u1eec\u1eee\u1ef0\u1ef2' + '\u1ef4\u1ef6\u1ef8\u1efa\u1efc\u1efe\u1f08-\u1f0f\u1f18-\u1f1d' + '\u1f28-\u1f2f\u1f38-\u1f3f\u1f48-\u1f4d\u1f59-\u1f5f' + '\u1f68-\u1f6f\u1fb8-\u1fbb\u1fc8-\u1fcb\u1fd8-\u1fdb' + '\u1fe8-\u1fec\u1ff8-\u1ffb\u2102\u2107\u210b-\u210d\u2110-\u2112' + '\u2115\u2119-\u211d\u2124\u2126\u2128\u212a-\u212d\u2130-\u2133' + '\u213e-\u213f\u2145\u2183\u2c00-\u2c2e\u2c60\u2c62-\u2c64\u2c67' + '\u2c69\u2c6b\u2c6d-\u2c6f\u2c72\u2c75\u2c80\u2c82\u2c84\u2c86' + '\u2c88\u2c8a\u2c8c\u2c8e\u2c90\u2c92\u2c94\u2c96\u2c98\u2c9a' + '\u2c9c\u2c9e\u2ca0\u2ca2\u2ca4\u2ca6\u2ca8\u2caa\u2cac\u2cae' + '\u2cb0\u2cb2\u2cb4\u2cb6\u2cb8\u2cba\u2cbc\u2cbe\u2cc0\u2cc2' + '\u2cc4\u2cc6\u2cc8\u2cca\u2ccc\u2cce\u2cd0\u2cd2\u2cd4\u2cd6' + '\u2cd8\u2cda\u2cdc\u2cde\u2ce0\u2ce2\ua640\ua642\ua644\ua646' + '\ua648\ua64a\ua64c\ua64e\ua650\ua652\ua654\ua656\ua658\ua65a' + '\ua65c\ua65e\ua662\ua664\ua666\ua668\ua66a\ua66c\ua680\ua682' + '\ua684\ua686\ua688\ua68a\ua68c\ua68e\ua690\ua692\ua694\ua696' + '\ua722\ua724\ua726\ua728\ua72a\ua72c\ua72e\ua732\ua734\ua736' + '\ua738\ua73a\ua73c\ua73e\ua740\ua742\ua744\ua746\ua748\ua74a' + '\ua74c\ua74e\ua750\ua752\ua754\ua756\ua758\ua75a\ua75c\ua75e' + '\ua760\ua762\ua764\ua766\ua768\ua76a\ua76c\ua76e\ua779\ua77b' + '\ua77d-\ua77e\ua780\ua782\ua784\ua786\ua78b\uff21-\uff3a]') + + idrest = '%s(?:%s|[0-9])*(?:(?<=_)%s)?' % (letter, letter, op) + letter_letter_digit = '%s(?:%s|\\d)*' % (letter, letter) tokens = { 'root': [ @@ -272,26 +272,26 @@ class ScalaLexer(RegexLexer): (r'(class|trait|object)(\s+)', bygroups(Keyword, Text), 'class'), (r'[^\S\n]+', Text), include('comments'), - (u'@%s' % idrest, Name.Decorator), - (u'(abstract|ca(?:se|tch)|d(?:ef|o)|e(?:lse|xtends)|' - u'f(?:inal(?:ly)?|or(?:Some)?)|i(?:f|mplicit)|' - u'lazy|match|new|override|pr(?:ivate|otected)' - u'|re(?:quires|turn)|s(?:ealed|uper)|' - u't(?:h(?:is|row)|ry)|va[lr]|w(?:hile|ith)|yield)\\b|' - u'(<[%:-]|=>|>:|[#=@_\u21D2\u2190])\\b', Keyword), - (u':(?!%s)' % op, Keyword, 'type'), - (u'%s%s\\b' % (upper, idrest), Name.Class), + (r'@%s' % idrest, Name.Decorator), + (r'(abstract|ca(?:se|tch)|d(?:ef|o)|e(?:lse|xtends)|' + r'f(?:inal(?:ly)?|or(?:Some)?)|i(?:f|mplicit)|' + r'lazy|match|new|override|pr(?:ivate|otected)' + r'|re(?:quires|turn)|s(?:ealed|uper)|' + r't(?:h(?:is|row)|ry)|va[lr]|w(?:hile|ith)|yield)\b|' + r'(<[%:-]|=>|>:|[#=@_\u21D2\u2190])\b', Keyword), + (r':(?!%s)' % op, Keyword, 'type'), + (r'%s%s\b' % (upper, idrest), Name.Class), (r'(true|false|null)\b', Keyword.Constant), (r'(import|package)(\s+)', bygroups(Keyword, Text), 'import'), (r'(type)(\s+)', bygroups(Keyword, Text), 'type'), (r'""".*?"""(?!")', String), (r'"(\\\\|\\"|[^"])*"', String), (r"'\\.'|'[^\\]'|'\\u[0-9a-fA-F]{4}'", String.Char), - (u"'%s" % idrest, Text.Symbol), + (r"'%s" % idrest, Text.Symbol), (r'[fs]"""', String, 'interptriplestring'), # interpolated strings (r'[fs]"', String, 'interpstring'), # interpolated strings (r'raw"(\\\\|\\"|[^"])*"', String), # raw strings - # (ur'(\.)(%s|%s|`[^`]+`)' % (idrest, op), bygroups(Operator, + # (r'(\.)(%s|%s|`[^`]+`)' % (idrest, op), bygroups(Operator, # Name.Attribute)), (idrest, Name), (r'`[^`]+`', Name), @@ -305,27 +305,27 @@ class ScalaLexer(RegexLexer): (r'\n', Text) ], 'class': [ - (u'(%s|%s|`[^`]+`)(\\s*)(\\[)' % (idrest, op), + (r'(%s|%s|`[^`]+`)(\s*)(\[)' % (idrest, op), bygroups(Name.Class, Text, Operator), ('#pop', 'typeparam')), (r'\s+', Text), include('comments'), (r'\{', Operator, '#pop'), (r'\(', Operator, '#pop'), - (u'%s|%s|`[^`]+`' % (idrest, op), Name.Class, '#pop'), + (r'%s|%s|`[^`]+`' % (idrest, op), Name.Class, '#pop'), ], 'type': [ (r'\s+', Text), include('comments'), (r'<[%:]|>:|[#_]|\bforSome\b|\btype\b', Keyword), - (u'([,);}]|=>|=|\u21d2)(\\s*)', bygroups(Operator, Text), '#pop'), + (r'([,);}]|=>|=|\u21d2)(\s*)', bygroups(Operator, Text), '#pop'), (r'[({]', Operator, '#push'), - (u'((?:%s|%s|`[^`]+`)(?:\\.(?:%s|%s|`[^`]+`))*)(\\s*)(\\[)' % + (r'((?:%s|%s|`[^`]+`)(?:\.(?:%s|%s|`[^`]+`))*)(\s*)(\[)' % (idrest, op, idrest, op), bygroups(Keyword.Type, Text, Operator), ('#pop', 'typeparam')), - (u'((?:%s|%s|`[^`]+`)(?:\\.(?:%s|%s|`[^`]+`))*)(\\s*)$' % + (r'((?:%s|%s|`[^`]+`)(?:\.(?:%s|%s|`[^`]+`))*)(\s*)$' % (idrest, op, idrest, op), bygroups(Keyword.Type, Text), '#pop'), - (u'\\.|%s|%s|`[^`]+`' % (idrest, op), Keyword.Type) + (r'\.|%s|%s|`[^`]+`' % (idrest, op), Keyword.Type) ], 'typeparam': [ (r'\s+', Text), @@ -334,7 +334,7 @@ class ScalaLexer(RegexLexer): (r'<[%:]|=>|>:|[#_\u21D2]|\bforSome\b|\btype\b', Keyword), (r'([\])}])', Operator, '#pop'), (r'[(\[{]', Operator, '#push'), - (u'\\.|%s|%s|`[^`]+`' % (idrest, op), Keyword.Type) + (r'\.|%s|%s|`[^`]+`' % (idrest, op), Keyword.Type) ], 'comments': [ (r'//.*?\n', Comment.Single), @@ -347,7 +347,7 @@ class ScalaLexer(RegexLexer): (r'[*/]', Comment.Multiline) ], 'import': [ - (u'(%s|\\.)+' % idrest, Name.Namespace, '#pop') + (r'(%s|\.)+' % idrest, Name.Namespace, '#pop') ], 'interpstringcommon': [ (r'[^"$\\]+', String), @@ -698,9 +698,9 @@ class IokeLexer(RegexLexer): r'System|Text|Tuple)(?![\w!:?])', Name.Builtin), # functions - (u'(generateMatchMethod|aliasMethod|\u03bb|\u028E|fnx|fn|method|' - u'dmacro|dlecro|syntax|macro|dlecrox|lecrox|lecro|syntax)' - u'(?![\\w!:?])', Name.Function), + ('(generateMatchMethod|aliasMethod|\u03bb|\u028E|fnx|fn|method|' + 'dmacro|dlecro|syntax|macro|dlecrox|lecrox|lecro|syntax)' + '(?![\\w!:?])', Name.Function), # Numbers (r'-?0[xX][0-9a-fA-F]+', Number.Hex), @@ -716,7 +716,7 @@ class IokeLexer(RegexLexer): r'\-\-|<=|>=|==|!=|&&|\.\.|\+=|\-=|\*=|\/=|%=|&=|\^=|\|=|<\-|' r'\+>|!>|<>|&>|%>|#>|\@>|\/>|\*>|\?>|\|>|\^>|~>|\$>|<\->|\->|' r'<<|>>|\*\*|\?\||\?&|\|\||>|<|\*|\/|%|\+|\-|&|\^|\||=|\$|!|~|' - u'\\?|#|\u2260|\u2218|\u2208|\u2209)', Operator), + r'\?|#|\u2260|\u2218|\u2208|\u2209)', Operator), (r'(and|nand|or|xor|nor|return|import)(?![\w!?])', Operator), @@ -1017,7 +1017,7 @@ class KotlinLexer(RegexLexer): .. versionadded:: 1.5 """ - + name = 'Kotlin' aliases = ['kotlin'] filenames = ['*.kt'] @@ -1028,7 +1028,7 @@ class KotlinLexer(RegexLexer): kt_name = ('@?[_' + uni.combine('Lu', 'Ll', 'Lt', 'Lm', 'Nl') + ']' + '[' + uni.combine('Lu', 'Ll', 'Lt', 'Lm', 'Nl', 'Nd', 'Pc', 'Cf', 'Mn', 'Mc') + ']*') - + kt_space_name = ('@?[_' + uni.combine('Lu', 'Ll', 'Lt', 'Lm', 'Nl') + ']' + '[' + uni.combine('Lu', 'Ll', 'Lt', 'Lm', 'Nl', 'Nd', 'Pc', 'Cf', 'Mn', 'Mc', 'Zs') + ',-]*') @@ -1142,7 +1142,7 @@ class XtendLexer(RegexLexer): 'class'), (r'(import)(\s+)', bygroups(Keyword.Namespace, Text), 'import'), (r"(''')", String, 'template'), - (u'(\u00BB)', String, 'template'), + (r'(\u00BB)', String, 'template'), (r'"(\\\\|\\"|[^"])*"', String), (r"'(\\\\|\\'|[^'])*'", String), (r'[a-zA-Z_]\w*:', Name.Label), @@ -1161,7 +1161,7 @@ class XtendLexer(RegexLexer): ], 'template': [ (r"'''", String, '#pop'), - (u'\u00AB', String, '#pop'), + (r'\u00AB', String, '#pop'), (r'.', String) ], } diff --git a/pygments/lexers/lisp.py b/pygments/lexers/lisp.py index 5dee6c6f..47efb4fb 100644 --- a/pygments/lexers/lisp.py +++ b/pygments/lexers/lisp.py @@ -471,779 +471,779 @@ class RacketLexer(RegexLexer): # Generated by example.rkt _keywords = ( - u'#%app', u'#%datum', u'#%declare', u'#%expression', u'#%module-begin', - u'#%plain-app', u'#%plain-lambda', u'#%plain-module-begin', - u'#%printing-module-begin', u'#%provide', u'#%require', - u'#%stratified-body', u'#%top', u'#%top-interaction', - u'#%variable-reference', u'->', u'->*', u'->*m', u'->d', u'->dm', u'->i', - u'->m', u'...', u':do-in', u'==', u'=>', u'_', u'absent', u'abstract', - u'all-defined-out', u'all-from-out', u'and', u'any', u'augment', u'augment*', - u'augment-final', u'augment-final*', u'augride', u'augride*', u'begin', - u'begin-for-syntax', u'begin0', u'case', u'case->', u'case->m', - u'case-lambda', u'class', u'class*', u'class-field-accessor', - u'class-field-mutator', u'class/c', u'class/derived', u'combine-in', - u'combine-out', u'command-line', u'compound-unit', u'compound-unit/infer', - u'cond', u'cons/dc', u'contract', u'contract-out', u'contract-struct', - u'contracted', u'define', u'define-compound-unit', - u'define-compound-unit/infer', u'define-contract-struct', - u'define-custom-hash-types', u'define-custom-set-types', - u'define-for-syntax', u'define-local-member-name', u'define-logger', - u'define-match-expander', u'define-member-name', - u'define-module-boundary-contract', u'define-namespace-anchor', - u'define-opt/c', u'define-sequence-syntax', u'define-serializable-class', - u'define-serializable-class*', u'define-signature', - u'define-signature-form', u'define-struct', u'define-struct/contract', - u'define-struct/derived', u'define-syntax', u'define-syntax-rule', - u'define-syntaxes', u'define-unit', u'define-unit-binding', - u'define-unit-from-context', u'define-unit/contract', - u'define-unit/new-import-export', u'define-unit/s', u'define-values', - u'define-values-for-export', u'define-values-for-syntax', - u'define-values/invoke-unit', u'define-values/invoke-unit/infer', - u'define/augment', u'define/augment-final', u'define/augride', - u'define/contract', u'define/final-prop', u'define/match', - u'define/overment', u'define/override', u'define/override-final', - u'define/private', u'define/public', u'define/public-final', - u'define/pubment', u'define/subexpression-pos-prop', - u'define/subexpression-pos-prop/name', u'delay', u'delay/idle', - u'delay/name', u'delay/strict', u'delay/sync', u'delay/thread', u'do', - u'else', u'except', u'except-in', u'except-out', u'export', u'extends', - u'failure-cont', u'false', u'false/c', u'field', u'field-bound?', u'file', - u'flat-murec-contract', u'flat-rec-contract', u'for', u'for*', u'for*/and', - u'for*/async', u'for*/first', u'for*/fold', u'for*/fold/derived', - u'for*/hash', u'for*/hasheq', u'for*/hasheqv', u'for*/last', u'for*/list', - u'for*/lists', u'for*/mutable-set', u'for*/mutable-seteq', - u'for*/mutable-seteqv', u'for*/or', u'for*/product', u'for*/set', - u'for*/seteq', u'for*/seteqv', u'for*/stream', u'for*/sum', u'for*/vector', - u'for*/weak-set', u'for*/weak-seteq', u'for*/weak-seteqv', u'for-label', - u'for-meta', u'for-syntax', u'for-template', u'for/and', u'for/async', - u'for/first', u'for/fold', u'for/fold/derived', u'for/hash', u'for/hasheq', - u'for/hasheqv', u'for/last', u'for/list', u'for/lists', u'for/mutable-set', - u'for/mutable-seteq', u'for/mutable-seteqv', u'for/or', u'for/product', - u'for/set', u'for/seteq', u'for/seteqv', u'for/stream', u'for/sum', - u'for/vector', u'for/weak-set', u'for/weak-seteq', u'for/weak-seteqv', - u'gen:custom-write', u'gen:dict', u'gen:equal+hash', u'gen:set', - u'gen:stream', u'generic', u'get-field', u'hash/dc', u'if', u'implies', - u'import', u'include', u'include-at/relative-to', - u'include-at/relative-to/reader', u'include/reader', u'inherit', - u'inherit-field', u'inherit/inner', u'inherit/super', u'init', - u'init-depend', u'init-field', u'init-rest', u'inner', u'inspect', - u'instantiate', u'interface', u'interface*', u'invariant-assertion', - u'invoke-unit', u'invoke-unit/infer', u'lambda', u'lazy', u'let', u'let*', - u'let*-values', u'let-syntax', u'let-syntaxes', u'let-values', u'let/cc', - u'let/ec', u'letrec', u'letrec-syntax', u'letrec-syntaxes', - u'letrec-syntaxes+values', u'letrec-values', u'lib', u'link', u'local', - u'local-require', u'log-debug', u'log-error', u'log-fatal', u'log-info', - u'log-warning', u'match', u'match*', u'match*/derived', u'match-define', - u'match-define-values', u'match-lambda', u'match-lambda*', - u'match-lambda**', u'match-let', u'match-let*', u'match-let*-values', - u'match-let-values', u'match-letrec', u'match-letrec-values', - u'match/derived', u'match/values', u'member-name-key', u'mixin', u'module', - u'module*', u'module+', u'nand', u'new', u'nor', u'object-contract', - u'object/c', u'only', u'only-in', u'only-meta-in', u'open', u'opt/c', u'or', - u'overment', u'overment*', u'override', u'override*', u'override-final', - u'override-final*', u'parameterize', u'parameterize*', - u'parameterize-break', u'parametric->/c', u'place', u'place*', - u'place/context', u'planet', u'prefix', u'prefix-in', u'prefix-out', - u'private', u'private*', u'prompt-tag/c', u'protect-out', u'provide', - u'provide-signature-elements', u'provide/contract', u'public', u'public*', - u'public-final', u'public-final*', u'pubment', u'pubment*', u'quasiquote', - u'quasisyntax', u'quasisyntax/loc', u'quote', u'quote-syntax', - u'quote-syntax/prune', u'recontract-out', u'recursive-contract', - u'relative-in', u'rename', u'rename-in', u'rename-inner', u'rename-out', - u'rename-super', u'require', u'send', u'send*', u'send+', u'send-generic', - u'send/apply', u'send/keyword-apply', u'set!', u'set!-values', - u'set-field!', u'shared', u'stream', u'stream*', u'stream-cons', u'struct', - u'struct*', u'struct-copy', u'struct-field-index', u'struct-out', - u'struct/c', u'struct/ctc', u'struct/dc', u'submod', u'super', - u'super-instantiate', u'super-make-object', u'super-new', u'syntax', - u'syntax-case', u'syntax-case*', u'syntax-id-rules', u'syntax-rules', - u'syntax/loc', u'tag', u'this', u'this%', u'thunk', u'thunk*', u'time', - u'unconstrained-domain->', u'unit', u'unit-from-context', u'unit/c', - u'unit/new-import-export', u'unit/s', u'unless', u'unquote', - u'unquote-splicing', u'unsyntax', u'unsyntax-splicing', u'values/drop', - u'when', u'with-continuation-mark', u'with-contract', - u'with-contract-continuation-mark', u'with-handlers', u'with-handlers*', - u'with-method', u'with-syntax', u'λ' + '#%app', '#%datum', '#%declare', '#%expression', '#%module-begin', + '#%plain-app', '#%plain-lambda', '#%plain-module-begin', + '#%printing-module-begin', '#%provide', '#%require', + '#%stratified-body', '#%top', '#%top-interaction', + '#%variable-reference', '->', '->*', '->*m', '->d', '->dm', '->i', + '->m', '...', ':do-in', '==', '=>', '_', 'absent', 'abstract', + 'all-defined-out', 'all-from-out', 'and', 'any', 'augment', 'augment*', + 'augment-final', 'augment-final*', 'augride', 'augride*', 'begin', + 'begin-for-syntax', 'begin0', 'case', 'case->', 'case->m', + 'case-lambda', 'class', 'class*', 'class-field-accessor', + 'class-field-mutator', 'class/c', 'class/derived', 'combine-in', + 'combine-out', 'command-line', 'compound-unit', 'compound-unit/infer', + 'cond', 'cons/dc', 'contract', 'contract-out', 'contract-struct', + 'contracted', 'define', 'define-compound-unit', + 'define-compound-unit/infer', 'define-contract-struct', + 'define-custom-hash-types', 'define-custom-set-types', + 'define-for-syntax', 'define-local-member-name', 'define-logger', + 'define-match-expander', 'define-member-name', + 'define-module-boundary-contract', 'define-namespace-anchor', + 'define-opt/c', 'define-sequence-syntax', 'define-serializable-class', + 'define-serializable-class*', 'define-signature', + 'define-signature-form', 'define-struct', 'define-struct/contract', + 'define-struct/derived', 'define-syntax', 'define-syntax-rule', + 'define-syntaxes', 'define-unit', 'define-unit-binding', + 'define-unit-from-context', 'define-unit/contract', + 'define-unit/new-import-export', 'define-unit/s', 'define-values', + 'define-values-for-export', 'define-values-for-syntax', + 'define-values/invoke-unit', 'define-values/invoke-unit/infer', + 'define/augment', 'define/augment-final', 'define/augride', + 'define/contract', 'define/final-prop', 'define/match', + 'define/overment', 'define/override', 'define/override-final', + 'define/private', 'define/public', 'define/public-final', + 'define/pubment', 'define/subexpression-pos-prop', + 'define/subexpression-pos-prop/name', 'delay', 'delay/idle', + 'delay/name', 'delay/strict', 'delay/sync', 'delay/thread', 'do', + 'else', 'except', 'except-in', 'except-out', 'export', 'extends', + 'failure-cont', 'false', 'false/c', 'field', 'field-bound?', 'file', + 'flat-murec-contract', 'flat-rec-contract', 'for', 'for*', 'for*/and', + 'for*/async', 'for*/first', 'for*/fold', 'for*/fold/derived', + 'for*/hash', 'for*/hasheq', 'for*/hasheqv', 'for*/last', 'for*/list', + 'for*/lists', 'for*/mutable-set', 'for*/mutable-seteq', + 'for*/mutable-seteqv', 'for*/or', 'for*/product', 'for*/set', + 'for*/seteq', 'for*/seteqv', 'for*/stream', 'for*/sum', 'for*/vector', + 'for*/weak-set', 'for*/weak-seteq', 'for*/weak-seteqv', 'for-label', + 'for-meta', 'for-syntax', 'for-template', 'for/and', 'for/async', + 'for/first', 'for/fold', 'for/fold/derived', 'for/hash', 'for/hasheq', + 'for/hasheqv', 'for/last', 'for/list', 'for/lists', 'for/mutable-set', + 'for/mutable-seteq', 'for/mutable-seteqv', 'for/or', 'for/product', + 'for/set', 'for/seteq', 'for/seteqv', 'for/stream', 'for/sum', + 'for/vector', 'for/weak-set', 'for/weak-seteq', 'for/weak-seteqv', + 'gen:custom-write', 'gen:dict', 'gen:equal+hash', 'gen:set', + 'gen:stream', 'generic', 'get-field', 'hash/dc', 'if', 'implies', + 'import', 'include', 'include-at/relative-to', + 'include-at/relative-to/reader', 'include/reader', 'inherit', + 'inherit-field', 'inherit/inner', 'inherit/super', 'init', + 'init-depend', 'init-field', 'init-rest', 'inner', 'inspect', + 'instantiate', 'interface', 'interface*', 'invariant-assertion', + 'invoke-unit', 'invoke-unit/infer', 'lambda', 'lazy', 'let', 'let*', + 'let*-values', 'let-syntax', 'let-syntaxes', 'let-values', 'let/cc', + 'let/ec', 'letrec', 'letrec-syntax', 'letrec-syntaxes', + 'letrec-syntaxes+values', 'letrec-values', 'lib', 'link', 'local', + 'local-require', 'log-debug', 'log-error', 'log-fatal', 'log-info', + 'log-warning', 'match', 'match*', 'match*/derived', 'match-define', + 'match-define-values', 'match-lambda', 'match-lambda*', + 'match-lambda**', 'match-let', 'match-let*', 'match-let*-values', + 'match-let-values', 'match-letrec', 'match-letrec-values', + 'match/derived', 'match/values', 'member-name-key', 'mixin', 'module', + 'module*', 'module+', 'nand', 'new', 'nor', 'object-contract', + 'object/c', 'only', 'only-in', 'only-meta-in', 'open', 'opt/c', 'or', + 'overment', 'overment*', 'override', 'override*', 'override-final', + 'override-final*', 'parameterize', 'parameterize*', + 'parameterize-break', 'parametric->/c', 'place', 'place*', + 'place/context', 'planet', 'prefix', 'prefix-in', 'prefix-out', + 'private', 'private*', 'prompt-tag/c', 'protect-out', 'provide', + 'provide-signature-elements', 'provide/contract', 'public', 'public*', + 'public-final', 'public-final*', 'pubment', 'pubment*', 'quasiquote', + 'quasisyntax', 'quasisyntax/loc', 'quote', 'quote-syntax', + 'quote-syntax/prune', 'recontract-out', 'recursive-contract', + 'relative-in', 'rename', 'rename-in', 'rename-inner', 'rename-out', + 'rename-super', 'require', 'send', 'send*', 'send+', 'send-generic', + 'send/apply', 'send/keyword-apply', 'set!', 'set!-values', + 'set-field!', 'shared', 'stream', 'stream*', 'stream-cons', 'struct', + 'struct*', 'struct-copy', 'struct-field-index', 'struct-out', + 'struct/c', 'struct/ctc', 'struct/dc', 'submod', 'super', + 'super-instantiate', 'super-make-object', 'super-new', 'syntax', + 'syntax-case', 'syntax-case*', 'syntax-id-rules', 'syntax-rules', + 'syntax/loc', 'tag', 'this', 'this%', 'thunk', 'thunk*', 'time', + 'unconstrained-domain->', 'unit', 'unit-from-context', 'unit/c', + 'unit/new-import-export', 'unit/s', 'unless', 'unquote', + 'unquote-splicing', 'unsyntax', 'unsyntax-splicing', 'values/drop', + 'when', 'with-continuation-mark', 'with-contract', + 'with-contract-continuation-mark', 'with-handlers', 'with-handlers*', + 'with-method', 'with-syntax', 'λ' ) # Generated by example.rkt _builtins = ( - u'*', u'*list/c', u'+', u'-', u'/', u'<', u'</c', u'<=', u'<=/c', u'=', u'=/c', - u'>', u'>/c', u'>=', u'>=/c', u'abort-current-continuation', u'abs', - u'absolute-path?', u'acos', u'add-between', u'add1', u'alarm-evt', - u'always-evt', u'and/c', u'andmap', u'angle', u'any/c', u'append', u'append*', - u'append-map', u'apply', u'argmax', u'argmin', u'arithmetic-shift', - u'arity-at-least', u'arity-at-least-value', u'arity-at-least?', - u'arity-checking-wrapper', u'arity-includes?', u'arity=?', - u'arrow-contract-info', u'arrow-contract-info-accepts-arglist', - u'arrow-contract-info-chaperone-procedure', - u'arrow-contract-info-check-first-order', u'arrow-contract-info?', - u'asin', u'assf', u'assoc', u'assq', u'assv', u'atan', - u'bad-number-of-results', u'banner', u'base->-doms/c', u'base->-rngs/c', - u'base->?', u'between/c', u'bitwise-and', u'bitwise-bit-field', - u'bitwise-bit-set?', u'bitwise-ior', u'bitwise-not', u'bitwise-xor', - u'blame-add-car-context', u'blame-add-cdr-context', u'blame-add-context', - u'blame-add-missing-party', u'blame-add-nth-arg-context', - u'blame-add-range-context', u'blame-add-unknown-context', - u'blame-context', u'blame-contract', u'blame-fmt->-string', - u'blame-missing-party?', u'blame-negative', u'blame-original?', - u'blame-positive', u'blame-replace-negative', u'blame-source', - u'blame-swap', u'blame-swapped?', u'blame-update', u'blame-value', - u'blame?', u'boolean=?', u'boolean?', u'bound-identifier=?', u'box', - u'box-cas!', u'box-immutable', u'box-immutable/c', u'box/c', u'box?', - u'break-enabled', u'break-parameterization?', u'break-thread', - u'build-chaperone-contract-property', u'build-compound-type-name', - u'build-contract-property', u'build-flat-contract-property', - u'build-list', u'build-path', u'build-path/convention-type', - u'build-string', u'build-vector', u'byte-pregexp', u'byte-pregexp?', - u'byte-ready?', u'byte-regexp', u'byte-regexp?', u'byte?', u'bytes', - u'bytes->immutable-bytes', u'bytes->list', u'bytes->path', - u'bytes->path-element', u'bytes->string/latin-1', u'bytes->string/locale', - u'bytes->string/utf-8', u'bytes-append', u'bytes-append*', - u'bytes-close-converter', u'bytes-convert', u'bytes-convert-end', - u'bytes-converter?', u'bytes-copy', u'bytes-copy!', - u'bytes-environment-variable-name?', u'bytes-fill!', u'bytes-join', - u'bytes-length', u'bytes-no-nuls?', u'bytes-open-converter', u'bytes-ref', - u'bytes-set!', u'bytes-utf-8-index', u'bytes-utf-8-length', - u'bytes-utf-8-ref', u'bytes<?', u'bytes=?', u'bytes>?', u'bytes?', u'caaaar', - u'caaadr', u'caaar', u'caadar', u'caaddr', u'caadr', u'caar', u'cadaar', - u'cadadr', u'cadar', u'caddar', u'cadddr', u'caddr', u'cadr', - u'call-in-nested-thread', u'call-with-atomic-output-file', - u'call-with-break-parameterization', - u'call-with-composable-continuation', u'call-with-continuation-barrier', - u'call-with-continuation-prompt', u'call-with-current-continuation', - u'call-with-default-reading-parameterization', - u'call-with-escape-continuation', u'call-with-exception-handler', - u'call-with-file-lock/timeout', u'call-with-immediate-continuation-mark', - u'call-with-input-bytes', u'call-with-input-file', - u'call-with-input-file*', u'call-with-input-string', - u'call-with-output-bytes', u'call-with-output-file', - u'call-with-output-file*', u'call-with-output-string', - u'call-with-parameterization', u'call-with-semaphore', - u'call-with-semaphore/enable-break', u'call-with-values', u'call/cc', - u'call/ec', u'car', u'cartesian-product', u'cdaaar', u'cdaadr', u'cdaar', - u'cdadar', u'cdaddr', u'cdadr', u'cdar', u'cddaar', u'cddadr', u'cddar', - u'cdddar', u'cddddr', u'cdddr', u'cddr', u'cdr', u'ceiling', u'channel-get', - u'channel-put', u'channel-put-evt', u'channel-put-evt?', - u'channel-try-get', u'channel/c', u'channel?', u'chaperone-box', - u'chaperone-channel', u'chaperone-continuation-mark-key', - u'chaperone-contract-property?', u'chaperone-contract?', u'chaperone-evt', - u'chaperone-hash', u'chaperone-hash-set', u'chaperone-of?', - u'chaperone-procedure', u'chaperone-procedure*', u'chaperone-prompt-tag', - u'chaperone-struct', u'chaperone-struct-type', u'chaperone-vector', - u'chaperone?', u'char->integer', u'char-alphabetic?', u'char-blank?', - u'char-ci<=?', u'char-ci<?', u'char-ci=?', u'char-ci>=?', u'char-ci>?', - u'char-downcase', u'char-foldcase', u'char-general-category', - u'char-graphic?', u'char-in', u'char-in/c', u'char-iso-control?', - u'char-lower-case?', u'char-numeric?', u'char-punctuation?', - u'char-ready?', u'char-symbolic?', u'char-title-case?', u'char-titlecase', - u'char-upcase', u'char-upper-case?', u'char-utf-8-length', - u'char-whitespace?', u'char<=?', u'char<?', u'char=?', u'char>=?', u'char>?', - u'char?', u'check-duplicate-identifier', u'check-duplicates', - u'checked-procedure-check-and-extract', u'choice-evt', - u'class->interface', u'class-info', u'class-seal', u'class-unseal', - u'class?', u'cleanse-path', u'close-input-port', u'close-output-port', - u'coerce-chaperone-contract', u'coerce-chaperone-contracts', - u'coerce-contract', u'coerce-contract/f', u'coerce-contracts', - u'coerce-flat-contract', u'coerce-flat-contracts', u'collect-garbage', - u'collection-file-path', u'collection-path', u'combinations', u'compile', - u'compile-allow-set!-undefined', u'compile-context-preservation-enabled', - u'compile-enforce-module-constants', u'compile-syntax', - u'compiled-expression-recompile', u'compiled-expression?', - u'compiled-module-expression?', u'complete-path?', u'complex?', u'compose', - u'compose1', u'conjoin', u'conjugate', u'cons', u'cons/c', u'cons?', u'const', - u'continuation-mark-key/c', u'continuation-mark-key?', - u'continuation-mark-set->context', u'continuation-mark-set->list', - u'continuation-mark-set->list*', u'continuation-mark-set-first', - u'continuation-mark-set?', u'continuation-marks', - u'continuation-prompt-available?', u'continuation-prompt-tag?', - u'continuation?', u'contract-continuation-mark-key', - u'contract-custom-write-property-proc', u'contract-exercise', - u'contract-first-order', u'contract-first-order-passes?', - u'contract-late-neg-projection', u'contract-name', u'contract-proc', - u'contract-projection', u'contract-property?', - u'contract-random-generate', u'contract-random-generate-fail', - u'contract-random-generate-fail?', - u'contract-random-generate-get-current-environment', - u'contract-random-generate-stash', u'contract-random-generate/choose', - u'contract-stronger?', u'contract-struct-exercise', - u'contract-struct-generate', u'contract-struct-late-neg-projection', - u'contract-struct-list-contract?', u'contract-val-first-projection', - u'contract?', u'convert-stream', u'copy-directory/files', u'copy-file', - u'copy-port', u'cos', u'cosh', u'count', u'current-blame-format', - u'current-break-parameterization', u'current-code-inspector', - u'current-command-line-arguments', u'current-compile', - u'current-compiled-file-roots', u'current-continuation-marks', - u'current-contract-region', u'current-custodian', u'current-directory', - u'current-directory-for-user', u'current-drive', - u'current-environment-variables', u'current-error-port', u'current-eval', - u'current-evt-pseudo-random-generator', - u'current-force-delete-permissions', u'current-future', - u'current-gc-milliseconds', u'current-get-interaction-input-port', - u'current-inexact-milliseconds', u'current-input-port', - u'current-inspector', u'current-library-collection-links', - u'current-library-collection-paths', u'current-load', - u'current-load-extension', u'current-load-relative-directory', - u'current-load/use-compiled', u'current-locale', u'current-logger', - u'current-memory-use', u'current-milliseconds', - u'current-module-declare-name', u'current-module-declare-source', - u'current-module-name-resolver', u'current-module-path-for-load', - u'current-namespace', u'current-output-port', u'current-parameterization', - u'current-plumber', u'current-preserved-thread-cell-values', - u'current-print', u'current-process-milliseconds', u'current-prompt-read', - u'current-pseudo-random-generator', u'current-read-interaction', - u'current-reader-guard', u'current-readtable', u'current-seconds', - u'current-security-guard', u'current-subprocess-custodian-mode', - u'current-thread', u'current-thread-group', - u'current-thread-initial-stack-size', - u'current-write-relative-directory', u'curry', u'curryr', - u'custodian-box-value', u'custodian-box?', u'custodian-limit-memory', - u'custodian-managed-list', u'custodian-memory-accounting-available?', - u'custodian-require-memory', u'custodian-shutdown-all', u'custodian?', - u'custom-print-quotable-accessor', u'custom-print-quotable?', - u'custom-write-accessor', u'custom-write-property-proc', u'custom-write?', - u'date', u'date*', u'date*-nanosecond', u'date*-time-zone-name', u'date*?', - u'date-day', u'date-dst?', u'date-hour', u'date-minute', u'date-month', - u'date-second', u'date-time-zone-offset', u'date-week-day', u'date-year', - u'date-year-day', u'date?', u'datum->syntax', u'datum-intern-literal', - u'default-continuation-prompt-tag', u'degrees->radians', - u'delete-directory', u'delete-directory/files', u'delete-file', - u'denominator', u'dict->list', u'dict-can-functional-set?', - u'dict-can-remove-keys?', u'dict-clear', u'dict-clear!', u'dict-copy', - u'dict-count', u'dict-empty?', u'dict-for-each', u'dict-has-key?', - u'dict-implements/c', u'dict-implements?', u'dict-iter-contract', - u'dict-iterate-first', u'dict-iterate-key', u'dict-iterate-next', - u'dict-iterate-value', u'dict-key-contract', u'dict-keys', u'dict-map', - u'dict-mutable?', u'dict-ref', u'dict-ref!', u'dict-remove', - u'dict-remove!', u'dict-set', u'dict-set!', u'dict-set*', u'dict-set*!', - u'dict-update', u'dict-update!', u'dict-value-contract', u'dict-values', - u'dict?', u'directory-exists?', u'directory-list', u'disjoin', u'display', - u'display-lines', u'display-lines-to-file', u'display-to-file', - u'displayln', u'double-flonum?', u'drop', u'drop-common-prefix', - u'drop-right', u'dropf', u'dropf-right', u'dump-memory-stats', - u'dup-input-port', u'dup-output-port', u'dynamic->*', u'dynamic-get-field', - u'dynamic-object/c', u'dynamic-place', u'dynamic-place*', - u'dynamic-require', u'dynamic-require-for-syntax', u'dynamic-send', - u'dynamic-set-field!', u'dynamic-wind', u'eighth', u'empty', - u'empty-sequence', u'empty-stream', u'empty?', - u'environment-variables-copy', u'environment-variables-names', - u'environment-variables-ref', u'environment-variables-set!', - u'environment-variables?', u'eof', u'eof-evt', u'eof-object?', - u'ephemeron-value', u'ephemeron?', u'eprintf', u'eq-contract-val', - u'eq-contract?', u'eq-hash-code', u'eq?', u'equal-contract-val', - u'equal-contract?', u'equal-hash-code', u'equal-secondary-hash-code', - u'equal<%>', u'equal?', u'equal?/recur', u'eqv-hash-code', u'eqv?', u'error', - u'error-display-handler', u'error-escape-handler', - u'error-print-context-length', u'error-print-source-location', - u'error-print-width', u'error-value->string-handler', u'eval', - u'eval-jit-enabled', u'eval-syntax', u'even?', u'evt/c', u'evt?', - u'exact->inexact', u'exact-ceiling', u'exact-floor', u'exact-integer?', - u'exact-nonnegative-integer?', u'exact-positive-integer?', u'exact-round', - u'exact-truncate', u'exact?', u'executable-yield-handler', u'exit', - u'exit-handler', u'exn', u'exn-continuation-marks', u'exn-message', - u'exn:break', u'exn:break-continuation', u'exn:break:hang-up', - u'exn:break:hang-up?', u'exn:break:terminate', u'exn:break:terminate?', - u'exn:break?', u'exn:fail', u'exn:fail:contract', - u'exn:fail:contract:arity', u'exn:fail:contract:arity?', - u'exn:fail:contract:blame', u'exn:fail:contract:blame-object', - u'exn:fail:contract:blame?', u'exn:fail:contract:continuation', - u'exn:fail:contract:continuation?', u'exn:fail:contract:divide-by-zero', - u'exn:fail:contract:divide-by-zero?', - u'exn:fail:contract:non-fixnum-result', - u'exn:fail:contract:non-fixnum-result?', u'exn:fail:contract:variable', - u'exn:fail:contract:variable-id', u'exn:fail:contract:variable?', - u'exn:fail:contract?', u'exn:fail:filesystem', - u'exn:fail:filesystem:errno', u'exn:fail:filesystem:errno-errno', - u'exn:fail:filesystem:errno?', u'exn:fail:filesystem:exists', - u'exn:fail:filesystem:exists?', u'exn:fail:filesystem:missing-module', - u'exn:fail:filesystem:missing-module-path', - u'exn:fail:filesystem:missing-module?', u'exn:fail:filesystem:version', - u'exn:fail:filesystem:version?', u'exn:fail:filesystem?', - u'exn:fail:network', u'exn:fail:network:errno', - u'exn:fail:network:errno-errno', u'exn:fail:network:errno?', - u'exn:fail:network?', u'exn:fail:object', u'exn:fail:object?', - u'exn:fail:out-of-memory', u'exn:fail:out-of-memory?', u'exn:fail:read', - u'exn:fail:read-srclocs', u'exn:fail:read:eof', u'exn:fail:read:eof?', - u'exn:fail:read:non-char', u'exn:fail:read:non-char?', u'exn:fail:read?', - u'exn:fail:syntax', u'exn:fail:syntax-exprs', - u'exn:fail:syntax:missing-module', - u'exn:fail:syntax:missing-module-path', - u'exn:fail:syntax:missing-module?', u'exn:fail:syntax:unbound', - u'exn:fail:syntax:unbound?', u'exn:fail:syntax?', u'exn:fail:unsupported', - u'exn:fail:unsupported?', u'exn:fail:user', u'exn:fail:user?', - u'exn:fail?', u'exn:misc:match?', u'exn:missing-module-accessor', - u'exn:missing-module?', u'exn:srclocs-accessor', u'exn:srclocs?', u'exn?', - u'exp', u'expand', u'expand-once', u'expand-syntax', u'expand-syntax-once', - u'expand-syntax-to-top-form', u'expand-to-top-form', u'expand-user-path', - u'explode-path', u'expt', u'externalizable<%>', u'failure-result/c', - u'false?', u'field-names', u'fifth', u'file->bytes', u'file->bytes-lines', - u'file->lines', u'file->list', u'file->string', u'file->value', - u'file-exists?', u'file-name-from-path', u'file-or-directory-identity', - u'file-or-directory-modify-seconds', u'file-or-directory-permissions', - u'file-position', u'file-position*', u'file-size', - u'file-stream-buffer-mode', u'file-stream-port?', u'file-truncate', - u'filename-extension', u'filesystem-change-evt', - u'filesystem-change-evt-cancel', u'filesystem-change-evt?', - u'filesystem-root-list', u'filter', u'filter-map', u'filter-not', - u'filter-read-input-port', u'find-executable-path', u'find-files', - u'find-library-collection-links', u'find-library-collection-paths', - u'find-relative-path', u'find-system-path', u'findf', u'first', - u'first-or/c', u'fixnum?', u'flat-contract', u'flat-contract-predicate', - u'flat-contract-property?', u'flat-contract?', u'flat-named-contract', - u'flatten', u'floating-point-bytes->real', u'flonum?', u'floor', - u'flush-output', u'fold-files', u'foldl', u'foldr', u'for-each', u'force', - u'format', u'fourth', u'fprintf', u'free-identifier=?', - u'free-label-identifier=?', u'free-template-identifier=?', - u'free-transformer-identifier=?', u'fsemaphore-count', u'fsemaphore-post', - u'fsemaphore-try-wait?', u'fsemaphore-wait', u'fsemaphore?', u'future', - u'future?', u'futures-enabled?', u'gcd', u'generate-member-key', - u'generate-temporaries', u'generic-set?', u'generic?', u'gensym', - u'get-output-bytes', u'get-output-string', u'get-preference', - u'get/build-late-neg-projection', u'get/build-val-first-projection', - u'getenv', u'global-port-print-handler', u'group-by', u'group-execute-bit', - u'group-read-bit', u'group-write-bit', u'guard-evt', u'handle-evt', - u'handle-evt?', u'has-blame?', u'has-contract?', u'hash', u'hash->list', - u'hash-clear', u'hash-clear!', u'hash-copy', u'hash-copy-clear', - u'hash-count', u'hash-empty?', u'hash-eq?', u'hash-equal?', u'hash-eqv?', - u'hash-for-each', u'hash-has-key?', u'hash-iterate-first', - u'hash-iterate-key', u'hash-iterate-key+value', u'hash-iterate-next', - u'hash-iterate-pair', u'hash-iterate-value', u'hash-keys', u'hash-map', - u'hash-placeholder?', u'hash-ref', u'hash-ref!', u'hash-remove', - u'hash-remove!', u'hash-set', u'hash-set!', u'hash-set*', u'hash-set*!', - u'hash-update', u'hash-update!', u'hash-values', u'hash-weak?', u'hash/c', - u'hash?', u'hasheq', u'hasheqv', u'identifier-binding', - u'identifier-binding-symbol', u'identifier-label-binding', - u'identifier-prune-lexical-context', - u'identifier-prune-to-source-module', - u'identifier-remove-from-definition-context', - u'identifier-template-binding', u'identifier-transformer-binding', - u'identifier?', u'identity', u'if/c', u'imag-part', u'immutable?', - u'impersonate-box', u'impersonate-channel', - u'impersonate-continuation-mark-key', u'impersonate-hash', - u'impersonate-hash-set', u'impersonate-procedure', - u'impersonate-procedure*', u'impersonate-prompt-tag', - u'impersonate-struct', u'impersonate-vector', u'impersonator-contract?', - u'impersonator-ephemeron', u'impersonator-of?', - u'impersonator-prop:application-mark', u'impersonator-prop:blame', - u'impersonator-prop:contracted', - u'impersonator-property-accessor-procedure?', u'impersonator-property?', - u'impersonator?', u'implementation?', u'implementation?/c', u'in-bytes', - u'in-bytes-lines', u'in-combinations', u'in-cycle', u'in-dict', - u'in-dict-keys', u'in-dict-pairs', u'in-dict-values', u'in-directory', - u'in-hash', u'in-hash-keys', u'in-hash-pairs', u'in-hash-values', - u'in-immutable-hash', u'in-immutable-hash-keys', - u'in-immutable-hash-pairs', u'in-immutable-hash-values', - u'in-immutable-set', u'in-indexed', u'in-input-port-bytes', - u'in-input-port-chars', u'in-lines', u'in-list', u'in-mlist', - u'in-mutable-hash', u'in-mutable-hash-keys', u'in-mutable-hash-pairs', - u'in-mutable-hash-values', u'in-mutable-set', u'in-naturals', - u'in-parallel', u'in-permutations', u'in-port', u'in-producer', u'in-range', - u'in-sequences', u'in-set', u'in-slice', u'in-stream', u'in-string', - u'in-syntax', u'in-value', u'in-values*-sequence', u'in-values-sequence', - u'in-vector', u'in-weak-hash', u'in-weak-hash-keys', u'in-weak-hash-pairs', - u'in-weak-hash-values', u'in-weak-set', u'inexact->exact', - u'inexact-real?', u'inexact?', u'infinite?', u'input-port-append', - u'input-port?', u'inspector?', u'instanceof/c', u'integer->char', - u'integer->integer-bytes', u'integer-bytes->integer', u'integer-in', - u'integer-length', u'integer-sqrt', u'integer-sqrt/remainder', u'integer?', - u'interface->method-names', u'interface-extension?', u'interface?', - u'internal-definition-context-binding-identifiers', - u'internal-definition-context-introduce', - u'internal-definition-context-seal', u'internal-definition-context?', - u'is-a?', u'is-a?/c', u'keyword->string', u'keyword-apply', u'keyword<?', - u'keyword?', u'keywords-match', u'kill-thread', u'last', u'last-pair', - u'lcm', u'length', u'liberal-define-context?', u'link-exists?', u'list', - u'list*', u'list*of', u'list->bytes', u'list->mutable-set', - u'list->mutable-seteq', u'list->mutable-seteqv', u'list->set', - u'list->seteq', u'list->seteqv', u'list->string', u'list->vector', - u'list->weak-set', u'list->weak-seteq', u'list->weak-seteqv', - u'list-contract?', u'list-prefix?', u'list-ref', u'list-set', u'list-tail', - u'list-update', u'list/c', u'list?', u'listen-port-number?', u'listof', - u'load', u'load-extension', u'load-on-demand-enabled', u'load-relative', - u'load-relative-extension', u'load/cd', u'load/use-compiled', - u'local-expand', u'local-expand/capture-lifts', - u'local-transformer-expand', u'local-transformer-expand/capture-lifts', - u'locale-string-encoding', u'log', u'log-all-levels', u'log-level-evt', - u'log-level?', u'log-max-level', u'log-message', u'log-receiver?', - u'logger-name', u'logger?', u'magnitude', u'make-arity-at-least', - u'make-base-empty-namespace', u'make-base-namespace', u'make-bytes', - u'make-channel', u'make-chaperone-contract', - u'make-continuation-mark-key', u'make-continuation-prompt-tag', - u'make-contract', u'make-custodian', u'make-custodian-box', - u'make-custom-hash', u'make-custom-hash-types', u'make-custom-set', - u'make-custom-set-types', u'make-date', u'make-date*', - u'make-derived-parameter', u'make-directory', u'make-directory*', - u'make-do-sequence', u'make-empty-namespace', - u'make-environment-variables', u'make-ephemeron', u'make-exn', - u'make-exn:break', u'make-exn:break:hang-up', u'make-exn:break:terminate', - u'make-exn:fail', u'make-exn:fail:contract', - u'make-exn:fail:contract:arity', u'make-exn:fail:contract:blame', - u'make-exn:fail:contract:continuation', - u'make-exn:fail:contract:divide-by-zero', - u'make-exn:fail:contract:non-fixnum-result', - u'make-exn:fail:contract:variable', u'make-exn:fail:filesystem', - u'make-exn:fail:filesystem:errno', u'make-exn:fail:filesystem:exists', - u'make-exn:fail:filesystem:missing-module', - u'make-exn:fail:filesystem:version', u'make-exn:fail:network', - u'make-exn:fail:network:errno', u'make-exn:fail:object', - u'make-exn:fail:out-of-memory', u'make-exn:fail:read', - u'make-exn:fail:read:eof', u'make-exn:fail:read:non-char', - u'make-exn:fail:syntax', u'make-exn:fail:syntax:missing-module', - u'make-exn:fail:syntax:unbound', u'make-exn:fail:unsupported', - u'make-exn:fail:user', u'make-file-or-directory-link', - u'make-flat-contract', u'make-fsemaphore', u'make-generic', - u'make-handle-get-preference-locked', u'make-hash', - u'make-hash-placeholder', u'make-hasheq', u'make-hasheq-placeholder', - u'make-hasheqv', u'make-hasheqv-placeholder', - u'make-immutable-custom-hash', u'make-immutable-hash', - u'make-immutable-hasheq', u'make-immutable-hasheqv', - u'make-impersonator-property', u'make-input-port', - u'make-input-port/read-to-peek', u'make-inspector', - u'make-keyword-procedure', u'make-known-char-range-list', - u'make-limited-input-port', u'make-list', u'make-lock-file-name', - u'make-log-receiver', u'make-logger', u'make-mixin-contract', - u'make-mutable-custom-set', u'make-none/c', u'make-object', - u'make-output-port', u'make-parameter', u'make-parent-directory*', - u'make-phantom-bytes', u'make-pipe', u'make-pipe-with-specials', - u'make-placeholder', u'make-plumber', u'make-polar', u'make-prefab-struct', - u'make-primitive-class', u'make-proj-contract', - u'make-pseudo-random-generator', u'make-reader-graph', u'make-readtable', - u'make-rectangular', u'make-rename-transformer', - u'make-resolved-module-path', u'make-security-guard', u'make-semaphore', - u'make-set!-transformer', u'make-shared-bytes', u'make-sibling-inspector', - u'make-special-comment', u'make-srcloc', u'make-string', - u'make-struct-field-accessor', u'make-struct-field-mutator', - u'make-struct-type', u'make-struct-type-property', - u'make-syntax-delta-introducer', u'make-syntax-introducer', - u'make-temporary-file', u'make-tentative-pretty-print-output-port', - u'make-thread-cell', u'make-thread-group', u'make-vector', - u'make-weak-box', u'make-weak-custom-hash', u'make-weak-custom-set', - u'make-weak-hash', u'make-weak-hasheq', u'make-weak-hasheqv', - u'make-will-executor', u'map', u'match-equality-test', - u'matches-arity-exactly?', u'max', u'mcar', u'mcdr', u'mcons', u'member', - u'member-name-key-hash-code', u'member-name-key=?', u'member-name-key?', - u'memf', u'memq', u'memv', u'merge-input', u'method-in-interface?', u'min', - u'mixin-contract', u'module->exports', u'module->imports', - u'module->language-info', u'module->namespace', - u'module-compiled-cross-phase-persistent?', u'module-compiled-exports', - u'module-compiled-imports', u'module-compiled-language-info', - u'module-compiled-name', u'module-compiled-submodules', - u'module-declared?', u'module-path-index-join', - u'module-path-index-resolve', u'module-path-index-split', - u'module-path-index-submodule', u'module-path-index?', u'module-path?', - u'module-predefined?', u'module-provide-protected?', u'modulo', u'mpair?', - u'mutable-set', u'mutable-seteq', u'mutable-seteqv', u'n->th', - u'nack-guard-evt', u'namespace-anchor->empty-namespace', - u'namespace-anchor->namespace', u'namespace-anchor?', - u'namespace-attach-module', u'namespace-attach-module-declaration', - u'namespace-base-phase', u'namespace-mapped-symbols', - u'namespace-module-identifier', u'namespace-module-registry', - u'namespace-require', u'namespace-require/constant', - u'namespace-require/copy', u'namespace-require/expansion-time', - u'namespace-set-variable-value!', u'namespace-symbol->identifier', - u'namespace-syntax-introduce', u'namespace-undefine-variable!', - u'namespace-unprotect-module', u'namespace-variable-value', u'namespace?', - u'nan?', u'natural-number/c', u'negate', u'negative?', u'never-evt', - u'new-∀/c', u'new-∃/c', u'newline', u'ninth', u'non-empty-listof', - u'non-empty-string?', u'none/c', u'normal-case-path', u'normalize-arity', - u'normalize-path', u'normalized-arity?', u'not', u'not/c', u'null', u'null?', - u'number->string', u'number?', u'numerator', u'object%', u'object->vector', - u'object-info', u'object-interface', u'object-method-arity-includes?', - u'object-name', u'object-or-false=?', u'object=?', u'object?', u'odd?', - u'one-of/c', u'open-input-bytes', u'open-input-file', - u'open-input-output-file', u'open-input-string', u'open-output-bytes', - u'open-output-file', u'open-output-nowhere', u'open-output-string', - u'or/c', u'order-of-magnitude', u'ormap', u'other-execute-bit', - u'other-read-bit', u'other-write-bit', u'output-port?', u'pair?', - u'parameter-procedure=?', u'parameter/c', u'parameter?', - u'parameterization?', u'parse-command-line', u'partition', u'path->bytes', - u'path->complete-path', u'path->directory-path', u'path->string', - u'path-add-suffix', u'path-convention-type', u'path-element->bytes', - u'path-element->string', u'path-element?', u'path-for-some-system?', - u'path-list-string->path-list', u'path-only', u'path-replace-suffix', - u'path-string?', u'path<?', u'path?', u'pathlist-closure', u'peek-byte', - u'peek-byte-or-special', u'peek-bytes', u'peek-bytes!', u'peek-bytes!-evt', - u'peek-bytes-avail!', u'peek-bytes-avail!*', u'peek-bytes-avail!-evt', - u'peek-bytes-avail!/enable-break', u'peek-bytes-evt', u'peek-char', - u'peek-char-or-special', u'peek-string', u'peek-string!', - u'peek-string!-evt', u'peek-string-evt', u'peeking-input-port', - u'permutations', u'phantom-bytes?', u'pi', u'pi.f', u'pipe-content-length', - u'place-break', u'place-channel', u'place-channel-get', - u'place-channel-put', u'place-channel-put/get', u'place-channel?', - u'place-dead-evt', u'place-enabled?', u'place-kill', u'place-location?', - u'place-message-allowed?', u'place-sleep', u'place-wait', u'place?', - u'placeholder-get', u'placeholder-set!', u'placeholder?', - u'plumber-add-flush!', u'plumber-flush-all', - u'plumber-flush-handle-remove!', u'plumber-flush-handle?', u'plumber?', - u'poll-guard-evt', u'port->bytes', u'port->bytes-lines', u'port->lines', - u'port->list', u'port->string', u'port-closed-evt', u'port-closed?', - u'port-commit-peeked', u'port-count-lines!', u'port-count-lines-enabled', - u'port-counts-lines?', u'port-display-handler', u'port-file-identity', - u'port-file-unlock', u'port-next-location', u'port-number?', - u'port-print-handler', u'port-progress-evt', - u'port-provides-progress-evts?', u'port-read-handler', - u'port-try-file-lock?', u'port-write-handler', u'port-writes-atomic?', - u'port-writes-special?', u'port?', u'positive?', u'predicate/c', - u'prefab-key->struct-type', u'prefab-key?', u'prefab-struct-key', - u'preferences-lock-file-mode', u'pregexp', u'pregexp?', u'pretty-display', - u'pretty-format', u'pretty-print', u'pretty-print-.-symbol-without-bars', - u'pretty-print-abbreviate-read-macros', u'pretty-print-columns', - u'pretty-print-current-style-table', u'pretty-print-depth', - u'pretty-print-exact-as-decimal', u'pretty-print-extend-style-table', - u'pretty-print-handler', u'pretty-print-newline', - u'pretty-print-post-print-hook', u'pretty-print-pre-print-hook', - u'pretty-print-print-hook', u'pretty-print-print-line', - u'pretty-print-remap-stylable', u'pretty-print-show-inexactness', - u'pretty-print-size-hook', u'pretty-print-style-table?', - u'pretty-printing', u'pretty-write', u'primitive-closure?', - u'primitive-result-arity', u'primitive?', u'print', u'print-as-expression', - u'print-boolean-long-form', u'print-box', u'print-graph', - u'print-hash-table', u'print-mpair-curly-braces', - u'print-pair-curly-braces', u'print-reader-abbreviations', - u'print-struct', u'print-syntax-width', u'print-unreadable', - u'print-vector-length', u'printable/c', u'printable<%>', u'printf', - u'println', u'procedure->method', u'procedure-arity', - u'procedure-arity-includes/c', u'procedure-arity-includes?', - u'procedure-arity?', u'procedure-closure-contents-eq?', - u'procedure-extract-target', u'procedure-keywords', - u'procedure-reduce-arity', u'procedure-reduce-keyword-arity', - u'procedure-rename', u'procedure-result-arity', u'procedure-specialize', - u'procedure-struct-type?', u'procedure?', u'process', u'process*', - u'process*/ports', u'process/ports', u'processor-count', u'progress-evt?', - u'promise-forced?', u'promise-running?', u'promise/c', u'promise/name?', - u'promise?', u'prop:arity-string', u'prop:arrow-contract', - u'prop:arrow-contract-get-info', u'prop:arrow-contract?', u'prop:blame', - u'prop:chaperone-contract', u'prop:checked-procedure', u'prop:contract', - u'prop:contracted', u'prop:custom-print-quotable', u'prop:custom-write', - u'prop:dict', u'prop:dict/contract', u'prop:equal+hash', u'prop:evt', - u'prop:exn:missing-module', u'prop:exn:srclocs', - u'prop:expansion-contexts', u'prop:flat-contract', - u'prop:impersonator-of', u'prop:input-port', - u'prop:liberal-define-context', u'prop:object-name', - u'prop:opt-chaperone-contract', u'prop:opt-chaperone-contract-get-test', - u'prop:opt-chaperone-contract?', u'prop:orc-contract', - u'prop:orc-contract-get-subcontracts', u'prop:orc-contract?', - u'prop:output-port', u'prop:place-location', u'prop:procedure', - u'prop:recursive-contract', u'prop:recursive-contract-unroll', - u'prop:recursive-contract?', u'prop:rename-transformer', u'prop:sequence', - u'prop:set!-transformer', u'prop:stream', u'proper-subset?', - u'pseudo-random-generator->vector', u'pseudo-random-generator-vector?', - u'pseudo-random-generator?', u'put-preferences', u'putenv', u'quotient', - u'quotient/remainder', u'radians->degrees', u'raise', - u'raise-argument-error', u'raise-arguments-error', u'raise-arity-error', - u'raise-blame-error', u'raise-contract-error', u'raise-mismatch-error', - u'raise-not-cons-blame-error', u'raise-range-error', - u'raise-result-error', u'raise-syntax-error', u'raise-type-error', - u'raise-user-error', u'random', u'random-seed', u'range', u'rational?', - u'rationalize', u'read', u'read-accept-bar-quote', u'read-accept-box', - u'read-accept-compiled', u'read-accept-dot', u'read-accept-graph', - u'read-accept-infix-dot', u'read-accept-lang', u'read-accept-quasiquote', - u'read-accept-reader', u'read-byte', u'read-byte-or-special', - u'read-bytes', u'read-bytes!', u'read-bytes!-evt', u'read-bytes-avail!', - u'read-bytes-avail!*', u'read-bytes-avail!-evt', - u'read-bytes-avail!/enable-break', u'read-bytes-evt', u'read-bytes-line', - u'read-bytes-line-evt', u'read-case-sensitive', u'read-cdot', u'read-char', - u'read-char-or-special', u'read-curly-brace-as-paren', - u'read-curly-brace-with-tag', u'read-decimal-as-inexact', - u'read-eval-print-loop', u'read-language', u'read-line', u'read-line-evt', - u'read-on-demand-source', u'read-square-bracket-as-paren', - u'read-square-bracket-with-tag', u'read-string', u'read-string!', - u'read-string!-evt', u'read-string-evt', u'read-syntax', - u'read-syntax/recursive', u'read/recursive', u'readtable-mapping', - u'readtable?', u'real->decimal-string', u'real->double-flonum', - u'real->floating-point-bytes', u'real->single-flonum', u'real-in', - u'real-part', u'real?', u'reencode-input-port', u'reencode-output-port', - u'regexp', u'regexp-match', u'regexp-match*', u'regexp-match-evt', - u'regexp-match-exact?', u'regexp-match-peek', - u'regexp-match-peek-immediate', u'regexp-match-peek-positions', - u'regexp-match-peek-positions*', - u'regexp-match-peek-positions-immediate', - u'regexp-match-peek-positions-immediate/end', - u'regexp-match-peek-positions/end', u'regexp-match-positions', - u'regexp-match-positions*', u'regexp-match-positions/end', - u'regexp-match/end', u'regexp-match?', u'regexp-max-lookbehind', - u'regexp-quote', u'regexp-replace', u'regexp-replace*', - u'regexp-replace-quote', u'regexp-replaces', u'regexp-split', - u'regexp-try-match', u'regexp?', u'relative-path?', u'relocate-input-port', - u'relocate-output-port', u'remainder', u'remf', u'remf*', u'remove', - u'remove*', u'remove-duplicates', u'remq', u'remq*', u'remv', u'remv*', - u'rename-contract', u'rename-file-or-directory', - u'rename-transformer-target', u'rename-transformer?', u'replace-evt', - u'reroot-path', u'resolve-path', u'resolved-module-path-name', - u'resolved-module-path?', u'rest', u'reverse', u'round', u'second', - u'seconds->date', u'security-guard?', u'semaphore-peek-evt', - u'semaphore-peek-evt?', u'semaphore-post', u'semaphore-try-wait?', - u'semaphore-wait', u'semaphore-wait/enable-break', u'semaphore?', - u'sequence->list', u'sequence->stream', u'sequence-add-between', - u'sequence-andmap', u'sequence-append', u'sequence-count', - u'sequence-filter', u'sequence-fold', u'sequence-for-each', - u'sequence-generate', u'sequence-generate*', u'sequence-length', - u'sequence-map', u'sequence-ormap', u'sequence-ref', u'sequence-tail', - u'sequence/c', u'sequence?', u'set', u'set!-transformer-procedure', - u'set!-transformer?', u'set->list', u'set->stream', u'set-add', u'set-add!', - u'set-box!', u'set-clear', u'set-clear!', u'set-copy', u'set-copy-clear', - u'set-count', u'set-empty?', u'set-eq?', u'set-equal?', u'set-eqv?', - u'set-first', u'set-for-each', u'set-implements/c', u'set-implements?', - u'set-intersect', u'set-intersect!', u'set-map', u'set-mcar!', u'set-mcdr!', - u'set-member?', u'set-mutable?', u'set-phantom-bytes!', - u'set-port-next-location!', u'set-remove', u'set-remove!', u'set-rest', - u'set-some-basic-contracts!', u'set-subtract', u'set-subtract!', - u'set-symmetric-difference', u'set-symmetric-difference!', u'set-union', - u'set-union!', u'set-weak?', u'set/c', u'set=?', u'set?', u'seteq', u'seteqv', - u'seventh', u'sgn', u'shared-bytes', u'shell-execute', u'shrink-path-wrt', - u'shuffle', u'simple-form-path', u'simplify-path', u'sin', - u'single-flonum?', u'sinh', u'sixth', u'skip-projection-wrapper?', u'sleep', - u'some-system-path->string', u'sort', u'special-comment-value', - u'special-comment?', u'special-filter-input-port', u'split-at', - u'split-at-right', u'split-common-prefix', u'split-path', u'splitf-at', - u'splitf-at-right', u'sqr', u'sqrt', u'srcloc', u'srcloc->string', - u'srcloc-column', u'srcloc-line', u'srcloc-position', u'srcloc-source', - u'srcloc-span', u'srcloc?', u'stop-after', u'stop-before', u'stream->list', - u'stream-add-between', u'stream-andmap', u'stream-append', u'stream-count', - u'stream-empty?', u'stream-filter', u'stream-first', u'stream-fold', - u'stream-for-each', u'stream-length', u'stream-map', u'stream-ormap', - u'stream-ref', u'stream-rest', u'stream-tail', u'stream/c', u'stream?', - u'string', u'string->bytes/latin-1', u'string->bytes/locale', - u'string->bytes/utf-8', u'string->immutable-string', u'string->keyword', - u'string->list', u'string->number', u'string->path', - u'string->path-element', u'string->some-system-path', u'string->symbol', - u'string->uninterned-symbol', u'string->unreadable-symbol', - u'string-append', u'string-append*', u'string-ci<=?', u'string-ci<?', - u'string-ci=?', u'string-ci>=?', u'string-ci>?', u'string-contains?', - u'string-copy', u'string-copy!', u'string-downcase', - u'string-environment-variable-name?', u'string-fill!', u'string-foldcase', - u'string-join', u'string-len/c', u'string-length', u'string-locale-ci<?', - u'string-locale-ci=?', u'string-locale-ci>?', u'string-locale-downcase', - u'string-locale-upcase', u'string-locale<?', u'string-locale=?', - u'string-locale>?', u'string-no-nuls?', u'string-normalize-nfc', - u'string-normalize-nfd', u'string-normalize-nfkc', - u'string-normalize-nfkd', u'string-normalize-spaces', u'string-port?', - u'string-prefix?', u'string-ref', u'string-replace', u'string-set!', - u'string-split', u'string-suffix?', u'string-titlecase', u'string-trim', - u'string-upcase', u'string-utf-8-length', u'string<=?', u'string<?', - u'string=?', u'string>=?', u'string>?', u'string?', u'struct->vector', - u'struct-accessor-procedure?', u'struct-constructor-procedure?', - u'struct-info', u'struct-mutator-procedure?', - u'struct-predicate-procedure?', u'struct-type-info', - u'struct-type-make-constructor', u'struct-type-make-predicate', - u'struct-type-property-accessor-procedure?', u'struct-type-property/c', - u'struct-type-property?', u'struct-type?', u'struct:arity-at-least', - u'struct:arrow-contract-info', u'struct:date', u'struct:date*', - u'struct:exn', u'struct:exn:break', u'struct:exn:break:hang-up', - u'struct:exn:break:terminate', u'struct:exn:fail', - u'struct:exn:fail:contract', u'struct:exn:fail:contract:arity', - u'struct:exn:fail:contract:blame', - u'struct:exn:fail:contract:continuation', - u'struct:exn:fail:contract:divide-by-zero', - u'struct:exn:fail:contract:non-fixnum-result', - u'struct:exn:fail:contract:variable', u'struct:exn:fail:filesystem', - u'struct:exn:fail:filesystem:errno', - u'struct:exn:fail:filesystem:exists', - u'struct:exn:fail:filesystem:missing-module', - u'struct:exn:fail:filesystem:version', u'struct:exn:fail:network', - u'struct:exn:fail:network:errno', u'struct:exn:fail:object', - u'struct:exn:fail:out-of-memory', u'struct:exn:fail:read', - u'struct:exn:fail:read:eof', u'struct:exn:fail:read:non-char', - u'struct:exn:fail:syntax', u'struct:exn:fail:syntax:missing-module', - u'struct:exn:fail:syntax:unbound', u'struct:exn:fail:unsupported', - u'struct:exn:fail:user', u'struct:srcloc', - u'struct:wrapped-extra-arg-arrow', u'struct?', u'sub1', u'subbytes', - u'subclass?', u'subclass?/c', u'subprocess', u'subprocess-group-enabled', - u'subprocess-kill', u'subprocess-pid', u'subprocess-status', - u'subprocess-wait', u'subprocess?', u'subset?', u'substring', u'suggest/c', - u'symbol->string', u'symbol-interned?', u'symbol-unreadable?', u'symbol<?', - u'symbol=?', u'symbol?', u'symbols', u'sync', u'sync/enable-break', - u'sync/timeout', u'sync/timeout/enable-break', u'syntax->datum', - u'syntax->list', u'syntax-arm', u'syntax-column', u'syntax-debug-info', - u'syntax-disarm', u'syntax-e', u'syntax-line', - u'syntax-local-bind-syntaxes', u'syntax-local-certifier', - u'syntax-local-context', u'syntax-local-expand-expression', - u'syntax-local-get-shadower', u'syntax-local-identifier-as-binding', - u'syntax-local-introduce', u'syntax-local-lift-context', - u'syntax-local-lift-expression', u'syntax-local-lift-module', - u'syntax-local-lift-module-end-declaration', - u'syntax-local-lift-provide', u'syntax-local-lift-require', - u'syntax-local-lift-values-expression', - u'syntax-local-make-definition-context', - u'syntax-local-make-delta-introducer', - u'syntax-local-module-defined-identifiers', - u'syntax-local-module-exports', - u'syntax-local-module-required-identifiers', u'syntax-local-name', - u'syntax-local-phase-level', u'syntax-local-submodules', - u'syntax-local-transforming-module-provides?', u'syntax-local-value', - u'syntax-local-value/immediate', u'syntax-original?', u'syntax-position', - u'syntax-property', u'syntax-property-preserved?', - u'syntax-property-symbol-keys', u'syntax-protect', u'syntax-rearm', - u'syntax-recertify', u'syntax-shift-phase-level', u'syntax-source', - u'syntax-source-module', u'syntax-span', u'syntax-taint', - u'syntax-tainted?', u'syntax-track-origin', - u'syntax-transforming-module-expression?', - u'syntax-transforming-with-lifts?', u'syntax-transforming?', u'syntax/c', - u'syntax?', u'system', u'system*', u'system*/exit-code', - u'system-big-endian?', u'system-idle-evt', u'system-language+country', - u'system-library-subpath', u'system-path-convention-type', u'system-type', - u'system/exit-code', u'tail-marks-match?', u'take', u'take-common-prefix', - u'take-right', u'takef', u'takef-right', u'tan', u'tanh', - u'tcp-abandon-port', u'tcp-accept', u'tcp-accept-evt', - u'tcp-accept-ready?', u'tcp-accept/enable-break', u'tcp-addresses', - u'tcp-close', u'tcp-connect', u'tcp-connect/enable-break', u'tcp-listen', - u'tcp-listener?', u'tcp-port?', u'tentative-pretty-print-port-cancel', - u'tentative-pretty-print-port-transfer', u'tenth', u'terminal-port?', - u'the-unsupplied-arg', u'third', u'thread', u'thread-cell-ref', - u'thread-cell-set!', u'thread-cell-values?', u'thread-cell?', - u'thread-dead-evt', u'thread-dead?', u'thread-group?', u'thread-receive', - u'thread-receive-evt', u'thread-resume', u'thread-resume-evt', - u'thread-rewind-receive', u'thread-running?', u'thread-send', - u'thread-suspend', u'thread-suspend-evt', u'thread-try-receive', - u'thread-wait', u'thread/suspend-to-kill', u'thread?', u'time-apply', - u'touch', u'transplant-input-port', u'transplant-output-port', u'true', - u'truncate', u'udp-addresses', u'udp-bind!', u'udp-bound?', u'udp-close', - u'udp-connect!', u'udp-connected?', u'udp-multicast-interface', - u'udp-multicast-join-group!', u'udp-multicast-leave-group!', - u'udp-multicast-loopback?', u'udp-multicast-set-interface!', - u'udp-multicast-set-loopback!', u'udp-multicast-set-ttl!', - u'udp-multicast-ttl', u'udp-open-socket', u'udp-receive!', - u'udp-receive!*', u'udp-receive!-evt', u'udp-receive!/enable-break', - u'udp-receive-ready-evt', u'udp-send', u'udp-send*', u'udp-send-evt', - u'udp-send-ready-evt', u'udp-send-to', u'udp-send-to*', u'udp-send-to-evt', - u'udp-send-to/enable-break', u'udp-send/enable-break', u'udp?', u'unbox', - u'uncaught-exception-handler', u'unit?', u'unspecified-dom', - u'unsupplied-arg?', u'use-collection-link-paths', - u'use-compiled-file-paths', u'use-user-specific-search-paths', - u'user-execute-bit', u'user-read-bit', u'user-write-bit', u'value-blame', - u'value-contract', u'values', u'variable-reference->empty-namespace', - u'variable-reference->module-base-phase', - u'variable-reference->module-declaration-inspector', - u'variable-reference->module-path-index', - u'variable-reference->module-source', u'variable-reference->namespace', - u'variable-reference->phase', - u'variable-reference->resolved-module-path', - u'variable-reference-constant?', u'variable-reference?', u'vector', - u'vector->immutable-vector', u'vector->list', - u'vector->pseudo-random-generator', u'vector->pseudo-random-generator!', - u'vector->values', u'vector-append', u'vector-argmax', u'vector-argmin', - u'vector-copy', u'vector-copy!', u'vector-count', u'vector-drop', - u'vector-drop-right', u'vector-fill!', u'vector-filter', - u'vector-filter-not', u'vector-immutable', u'vector-immutable/c', - u'vector-immutableof', u'vector-length', u'vector-map', u'vector-map!', - u'vector-member', u'vector-memq', u'vector-memv', u'vector-ref', - u'vector-set!', u'vector-set*!', u'vector-set-performance-stats!', - u'vector-split-at', u'vector-split-at-right', u'vector-take', - u'vector-take-right', u'vector/c', u'vector?', u'vectorof', u'version', - u'void', u'void?', u'weak-box-value', u'weak-box?', u'weak-set', - u'weak-seteq', u'weak-seteqv', u'will-execute', u'will-executor?', - u'will-register', u'will-try-execute', u'with-input-from-bytes', - u'with-input-from-file', u'with-input-from-string', - u'with-output-to-bytes', u'with-output-to-file', u'with-output-to-string', - u'would-be-future', u'wrap-evt', u'wrapped-extra-arg-arrow', - u'wrapped-extra-arg-arrow-extra-neg-party-argument', - u'wrapped-extra-arg-arrow-real-func', u'wrapped-extra-arg-arrow?', - u'writable<%>', u'write', u'write-byte', u'write-bytes', - u'write-bytes-avail', u'write-bytes-avail*', u'write-bytes-avail-evt', - u'write-bytes-avail/enable-break', u'write-char', u'write-special', - u'write-special-avail*', u'write-special-evt', u'write-string', - u'write-to-file', u'writeln', u'xor', u'zero?', u'~.a', u'~.s', u'~.v', u'~a', - u'~e', u'~r', u'~s', u'~v' + '*', '*list/c', '+', '-', '/', '<', '</c', '<=', '<=/c', '=', '=/c', + '>', '>/c', '>=', '>=/c', 'abort-current-continuation', 'abs', + 'absolute-path?', 'acos', 'add-between', 'add1', 'alarm-evt', + 'always-evt', 'and/c', 'andmap', 'angle', 'any/c', 'append', 'append*', + 'append-map', 'apply', 'argmax', 'argmin', 'arithmetic-shift', + 'arity-at-least', 'arity-at-least-value', 'arity-at-least?', + 'arity-checking-wrapper', 'arity-includes?', 'arity=?', + 'arrow-contract-info', 'arrow-contract-info-accepts-arglist', + 'arrow-contract-info-chaperone-procedure', + 'arrow-contract-info-check-first-order', 'arrow-contract-info?', + 'asin', 'assf', 'assoc', 'assq', 'assv', 'atan', + 'bad-number-of-results', 'banner', 'base->-doms/c', 'base->-rngs/c', + 'base->?', 'between/c', 'bitwise-and', 'bitwise-bit-field', + 'bitwise-bit-set?', 'bitwise-ior', 'bitwise-not', 'bitwise-xor', + 'blame-add-car-context', 'blame-add-cdr-context', 'blame-add-context', + 'blame-add-missing-party', 'blame-add-nth-arg-context', + 'blame-add-range-context', 'blame-add-unknown-context', + 'blame-context', 'blame-contract', 'blame-fmt->-string', + 'blame-missing-party?', 'blame-negative', 'blame-original?', + 'blame-positive', 'blame-replace-negative', 'blame-source', + 'blame-swap', 'blame-swapped?', 'blame-update', 'blame-value', + 'blame?', 'boolean=?', 'boolean?', 'bound-identifier=?', 'box', + 'box-cas!', 'box-immutable', 'box-immutable/c', 'box/c', 'box?', + 'break-enabled', 'break-parameterization?', 'break-thread', + 'build-chaperone-contract-property', 'build-compound-type-name', + 'build-contract-property', 'build-flat-contract-property', + 'build-list', 'build-path', 'build-path/convention-type', + 'build-string', 'build-vector', 'byte-pregexp', 'byte-pregexp?', + 'byte-ready?', 'byte-regexp', 'byte-regexp?', 'byte?', 'bytes', + 'bytes->immutable-bytes', 'bytes->list', 'bytes->path', + 'bytes->path-element', 'bytes->string/latin-1', 'bytes->string/locale', + 'bytes->string/utf-8', 'bytes-append', 'bytes-append*', + 'bytes-close-converter', 'bytes-convert', 'bytes-convert-end', + 'bytes-converter?', 'bytes-copy', 'bytes-copy!', + 'bytes-environment-variable-name?', 'bytes-fill!', 'bytes-join', + 'bytes-length', 'bytes-no-nuls?', 'bytes-open-converter', 'bytes-ref', + 'bytes-set!', 'bytes-utf-8-index', 'bytes-utf-8-length', + 'bytes-utf-8-ref', 'bytes<?', 'bytes=?', 'bytes>?', 'bytes?', 'caaaar', + 'caaadr', 'caaar', 'caadar', 'caaddr', 'caadr', 'caar', 'cadaar', + 'cadadr', 'cadar', 'caddar', 'cadddr', 'caddr', 'cadr', + 'call-in-nested-thread', 'call-with-atomic-output-file', + 'call-with-break-parameterization', + 'call-with-composable-continuation', 'call-with-continuation-barrier', + 'call-with-continuation-prompt', 'call-with-current-continuation', + 'call-with-default-reading-parameterization', + 'call-with-escape-continuation', 'call-with-exception-handler', + 'call-with-file-lock/timeout', 'call-with-immediate-continuation-mark', + 'call-with-input-bytes', 'call-with-input-file', + 'call-with-input-file*', 'call-with-input-string', + 'call-with-output-bytes', 'call-with-output-file', + 'call-with-output-file*', 'call-with-output-string', + 'call-with-parameterization', 'call-with-semaphore', + 'call-with-semaphore/enable-break', 'call-with-values', 'call/cc', + 'call/ec', 'car', 'cartesian-product', 'cdaaar', 'cdaadr', 'cdaar', + 'cdadar', 'cdaddr', 'cdadr', 'cdar', 'cddaar', 'cddadr', 'cddar', + 'cdddar', 'cddddr', 'cdddr', 'cddr', 'cdr', 'ceiling', 'channel-get', + 'channel-put', 'channel-put-evt', 'channel-put-evt?', + 'channel-try-get', 'channel/c', 'channel?', 'chaperone-box', + 'chaperone-channel', 'chaperone-continuation-mark-key', + 'chaperone-contract-property?', 'chaperone-contract?', 'chaperone-evt', + 'chaperone-hash', 'chaperone-hash-set', 'chaperone-of?', + 'chaperone-procedure', 'chaperone-procedure*', 'chaperone-prompt-tag', + 'chaperone-struct', 'chaperone-struct-type', 'chaperone-vector', + 'chaperone?', 'char->integer', 'char-alphabetic?', 'char-blank?', + 'char-ci<=?', 'char-ci<?', 'char-ci=?', 'char-ci>=?', 'char-ci>?', + 'char-downcase', 'char-foldcase', 'char-general-category', + 'char-graphic?', 'char-in', 'char-in/c', 'char-iso-control?', + 'char-lower-case?', 'char-numeric?', 'char-punctuation?', + 'char-ready?', 'char-symbolic?', 'char-title-case?', 'char-titlecase', + 'char-upcase', 'char-upper-case?', 'char-utf-8-length', + 'char-whitespace?', 'char<=?', 'char<?', 'char=?', 'char>=?', 'char>?', + 'char?', 'check-duplicate-identifier', 'check-duplicates', + 'checked-procedure-check-and-extract', 'choice-evt', + 'class->interface', 'class-info', 'class-seal', 'class-unseal', + 'class?', 'cleanse-path', 'close-input-port', 'close-output-port', + 'coerce-chaperone-contract', 'coerce-chaperone-contracts', + 'coerce-contract', 'coerce-contract/f', 'coerce-contracts', + 'coerce-flat-contract', 'coerce-flat-contracts', 'collect-garbage', + 'collection-file-path', 'collection-path', 'combinations', 'compile', + 'compile-allow-set!-undefined', 'compile-context-preservation-enabled', + 'compile-enforce-module-constants', 'compile-syntax', + 'compiled-expression-recompile', 'compiled-expression?', + 'compiled-module-expression?', 'complete-path?', 'complex?', 'compose', + 'compose1', 'conjoin', 'conjugate', 'cons', 'cons/c', 'cons?', 'const', + 'continuation-mark-key/c', 'continuation-mark-key?', + 'continuation-mark-set->context', 'continuation-mark-set->list', + 'continuation-mark-set->list*', 'continuation-mark-set-first', + 'continuation-mark-set?', 'continuation-marks', + 'continuation-prompt-available?', 'continuation-prompt-tag?', + 'continuation?', 'contract-continuation-mark-key', + 'contract-custom-write-property-proc', 'contract-exercise', + 'contract-first-order', 'contract-first-order-passes?', + 'contract-late-neg-projection', 'contract-name', 'contract-proc', + 'contract-projection', 'contract-property?', + 'contract-random-generate', 'contract-random-generate-fail', + 'contract-random-generate-fail?', + 'contract-random-generate-get-current-environment', + 'contract-random-generate-stash', 'contract-random-generate/choose', + 'contract-stronger?', 'contract-struct-exercise', + 'contract-struct-generate', 'contract-struct-late-neg-projection', + 'contract-struct-list-contract?', 'contract-val-first-projection', + 'contract?', 'convert-stream', 'copy-directory/files', 'copy-file', + 'copy-port', 'cos', 'cosh', 'count', 'current-blame-format', + 'current-break-parameterization', 'current-code-inspector', + 'current-command-line-arguments', 'current-compile', + 'current-compiled-file-roots', 'current-continuation-marks', + 'current-contract-region', 'current-custodian', 'current-directory', + 'current-directory-for-user', 'current-drive', + 'current-environment-variables', 'current-error-port', 'current-eval', + 'current-evt-pseudo-random-generator', + 'current-force-delete-permissions', 'current-future', + 'current-gc-milliseconds', 'current-get-interaction-input-port', + 'current-inexact-milliseconds', 'current-input-port', + 'current-inspector', 'current-library-collection-links', + 'current-library-collection-paths', 'current-load', + 'current-load-extension', 'current-load-relative-directory', + 'current-load/use-compiled', 'current-locale', 'current-logger', + 'current-memory-use', 'current-milliseconds', + 'current-module-declare-name', 'current-module-declare-source', + 'current-module-name-resolver', 'current-module-path-for-load', + 'current-namespace', 'current-output-port', 'current-parameterization', + 'current-plumber', 'current-preserved-thread-cell-values', + 'current-print', 'current-process-milliseconds', 'current-prompt-read', + 'current-pseudo-random-generator', 'current-read-interaction', + 'current-reader-guard', 'current-readtable', 'current-seconds', + 'current-security-guard', 'current-subprocess-custodian-mode', + 'current-thread', 'current-thread-group', + 'current-thread-initial-stack-size', + 'current-write-relative-directory', 'curry', 'curryr', + 'custodian-box-value', 'custodian-box?', 'custodian-limit-memory', + 'custodian-managed-list', 'custodian-memory-accounting-available?', + 'custodian-require-memory', 'custodian-shutdown-all', 'custodian?', + 'custom-print-quotable-accessor', 'custom-print-quotable?', + 'custom-write-accessor', 'custom-write-property-proc', 'custom-write?', + 'date', 'date*', 'date*-nanosecond', 'date*-time-zone-name', 'date*?', + 'date-day', 'date-dst?', 'date-hour', 'date-minute', 'date-month', + 'date-second', 'date-time-zone-offset', 'date-week-day', 'date-year', + 'date-year-day', 'date?', 'datum->syntax', 'datum-intern-literal', + 'default-continuation-prompt-tag', 'degrees->radians', + 'delete-directory', 'delete-directory/files', 'delete-file', + 'denominator', 'dict->list', 'dict-can-functional-set?', + 'dict-can-remove-keys?', 'dict-clear', 'dict-clear!', 'dict-copy', + 'dict-count', 'dict-empty?', 'dict-for-each', 'dict-has-key?', + 'dict-implements/c', 'dict-implements?', 'dict-iter-contract', + 'dict-iterate-first', 'dict-iterate-key', 'dict-iterate-next', + 'dict-iterate-value', 'dict-key-contract', 'dict-keys', 'dict-map', + 'dict-mutable?', 'dict-ref', 'dict-ref!', 'dict-remove', + 'dict-remove!', 'dict-set', 'dict-set!', 'dict-set*', 'dict-set*!', + 'dict-update', 'dict-update!', 'dict-value-contract', 'dict-values', + 'dict?', 'directory-exists?', 'directory-list', 'disjoin', 'display', + 'display-lines', 'display-lines-to-file', 'display-to-file', + 'displayln', 'double-flonum?', 'drop', 'drop-common-prefix', + 'drop-right', 'dropf', 'dropf-right', 'dump-memory-stats', + 'dup-input-port', 'dup-output-port', 'dynamic->*', 'dynamic-get-field', + 'dynamic-object/c', 'dynamic-place', 'dynamic-place*', + 'dynamic-require', 'dynamic-require-for-syntax', 'dynamic-send', + 'dynamic-set-field!', 'dynamic-wind', 'eighth', 'empty', + 'empty-sequence', 'empty-stream', 'empty?', + 'environment-variables-copy', 'environment-variables-names', + 'environment-variables-ref', 'environment-variables-set!', + 'environment-variables?', 'eof', 'eof-evt', 'eof-object?', + 'ephemeron-value', 'ephemeron?', 'eprintf', 'eq-contract-val', + 'eq-contract?', 'eq-hash-code', 'eq?', 'equal-contract-val', + 'equal-contract?', 'equal-hash-code', 'equal-secondary-hash-code', + 'equal<%>', 'equal?', 'equal?/recur', 'eqv-hash-code', 'eqv?', 'error', + 'error-display-handler', 'error-escape-handler', + 'error-print-context-length', 'error-print-source-location', + 'error-print-width', 'error-value->string-handler', 'eval', + 'eval-jit-enabled', 'eval-syntax', 'even?', 'evt/c', 'evt?', + 'exact->inexact', 'exact-ceiling', 'exact-floor', 'exact-integer?', + 'exact-nonnegative-integer?', 'exact-positive-integer?', 'exact-round', + 'exact-truncate', 'exact?', 'executable-yield-handler', 'exit', + 'exit-handler', 'exn', 'exn-continuation-marks', 'exn-message', + 'exn:break', 'exn:break-continuation', 'exn:break:hang-up', + 'exn:break:hang-up?', 'exn:break:terminate', 'exn:break:terminate?', + 'exn:break?', 'exn:fail', 'exn:fail:contract', + 'exn:fail:contract:arity', 'exn:fail:contract:arity?', + 'exn:fail:contract:blame', 'exn:fail:contract:blame-object', + 'exn:fail:contract:blame?', 'exn:fail:contract:continuation', + 'exn:fail:contract:continuation?', 'exn:fail:contract:divide-by-zero', + 'exn:fail:contract:divide-by-zero?', + 'exn:fail:contract:non-fixnum-result', + 'exn:fail:contract:non-fixnum-result?', 'exn:fail:contract:variable', + 'exn:fail:contract:variable-id', 'exn:fail:contract:variable?', + 'exn:fail:contract?', 'exn:fail:filesystem', + 'exn:fail:filesystem:errno', 'exn:fail:filesystem:errno-errno', + 'exn:fail:filesystem:errno?', 'exn:fail:filesystem:exists', + 'exn:fail:filesystem:exists?', 'exn:fail:filesystem:missing-module', + 'exn:fail:filesystem:missing-module-path', + 'exn:fail:filesystem:missing-module?', 'exn:fail:filesystem:version', + 'exn:fail:filesystem:version?', 'exn:fail:filesystem?', + 'exn:fail:network', 'exn:fail:network:errno', + 'exn:fail:network:errno-errno', 'exn:fail:network:errno?', + 'exn:fail:network?', 'exn:fail:object', 'exn:fail:object?', + 'exn:fail:out-of-memory', 'exn:fail:out-of-memory?', 'exn:fail:read', + 'exn:fail:read-srclocs', 'exn:fail:read:eof', 'exn:fail:read:eof?', + 'exn:fail:read:non-char', 'exn:fail:read:non-char?', 'exn:fail:read?', + 'exn:fail:syntax', 'exn:fail:syntax-exprs', + 'exn:fail:syntax:missing-module', + 'exn:fail:syntax:missing-module-path', + 'exn:fail:syntax:missing-module?', 'exn:fail:syntax:unbound', + 'exn:fail:syntax:unbound?', 'exn:fail:syntax?', 'exn:fail:unsupported', + 'exn:fail:unsupported?', 'exn:fail:user', 'exn:fail:user?', + 'exn:fail?', 'exn:misc:match?', 'exn:missing-module-accessor', + 'exn:missing-module?', 'exn:srclocs-accessor', 'exn:srclocs?', 'exn?', + 'exp', 'expand', 'expand-once', 'expand-syntax', 'expand-syntax-once', + 'expand-syntax-to-top-form', 'expand-to-top-form', 'expand-user-path', + 'explode-path', 'expt', 'externalizable<%>', 'failure-result/c', + 'false?', 'field-names', 'fifth', 'file->bytes', 'file->bytes-lines', + 'file->lines', 'file->list', 'file->string', 'file->value', + 'file-exists?', 'file-name-from-path', 'file-or-directory-identity', + 'file-or-directory-modify-seconds', 'file-or-directory-permissions', + 'file-position', 'file-position*', 'file-size', + 'file-stream-buffer-mode', 'file-stream-port?', 'file-truncate', + 'filename-extension', 'filesystem-change-evt', + 'filesystem-change-evt-cancel', 'filesystem-change-evt?', + 'filesystem-root-list', 'filter', 'filter-map', 'filter-not', + 'filter-read-input-port', 'find-executable-path', 'find-files', + 'find-library-collection-links', 'find-library-collection-paths', + 'find-relative-path', 'find-system-path', 'findf', 'first', + 'first-or/c', 'fixnum?', 'flat-contract', 'flat-contract-predicate', + 'flat-contract-property?', 'flat-contract?', 'flat-named-contract', + 'flatten', 'floating-point-bytes->real', 'flonum?', 'floor', + 'flush-output', 'fold-files', 'foldl', 'foldr', 'for-each', 'force', + 'format', 'fourth', 'fprintf', 'free-identifier=?', + 'free-label-identifier=?', 'free-template-identifier=?', + 'free-transformer-identifier=?', 'fsemaphore-count', 'fsemaphore-post', + 'fsemaphore-try-wait?', 'fsemaphore-wait', 'fsemaphore?', 'future', + 'future?', 'futures-enabled?', 'gcd', 'generate-member-key', + 'generate-temporaries', 'generic-set?', 'generic?', 'gensym', + 'get-output-bytes', 'get-output-string', 'get-preference', + 'get/build-late-neg-projection', 'get/build-val-first-projection', + 'getenv', 'global-port-print-handler', 'group-by', 'group-execute-bit', + 'group-read-bit', 'group-write-bit', 'guard-evt', 'handle-evt', + 'handle-evt?', 'has-blame?', 'has-contract?', 'hash', 'hash->list', + 'hash-clear', 'hash-clear!', 'hash-copy', 'hash-copy-clear', + 'hash-count', 'hash-empty?', 'hash-eq?', 'hash-equal?', 'hash-eqv?', + 'hash-for-each', 'hash-has-key?', 'hash-iterate-first', + 'hash-iterate-key', 'hash-iterate-key+value', 'hash-iterate-next', + 'hash-iterate-pair', 'hash-iterate-value', 'hash-keys', 'hash-map', + 'hash-placeholder?', 'hash-ref', 'hash-ref!', 'hash-remove', + 'hash-remove!', 'hash-set', 'hash-set!', 'hash-set*', 'hash-set*!', + 'hash-update', 'hash-update!', 'hash-values', 'hash-weak?', 'hash/c', + 'hash?', 'hasheq', 'hasheqv', 'identifier-binding', + 'identifier-binding-symbol', 'identifier-label-binding', + 'identifier-prune-lexical-context', + 'identifier-prune-to-source-module', + 'identifier-remove-from-definition-context', + 'identifier-template-binding', 'identifier-transformer-binding', + 'identifier?', 'identity', 'if/c', 'imag-part', 'immutable?', + 'impersonate-box', 'impersonate-channel', + 'impersonate-continuation-mark-key', 'impersonate-hash', + 'impersonate-hash-set', 'impersonate-procedure', + 'impersonate-procedure*', 'impersonate-prompt-tag', + 'impersonate-struct', 'impersonate-vector', 'impersonator-contract?', + 'impersonator-ephemeron', 'impersonator-of?', + 'impersonator-prop:application-mark', 'impersonator-prop:blame', + 'impersonator-prop:contracted', + 'impersonator-property-accessor-procedure?', 'impersonator-property?', + 'impersonator?', 'implementation?', 'implementation?/c', 'in-bytes', + 'in-bytes-lines', 'in-combinations', 'in-cycle', 'in-dict', + 'in-dict-keys', 'in-dict-pairs', 'in-dict-values', 'in-directory', + 'in-hash', 'in-hash-keys', 'in-hash-pairs', 'in-hash-values', + 'in-immutable-hash', 'in-immutable-hash-keys', + 'in-immutable-hash-pairs', 'in-immutable-hash-values', + 'in-immutable-set', 'in-indexed', 'in-input-port-bytes', + 'in-input-port-chars', 'in-lines', 'in-list', 'in-mlist', + 'in-mutable-hash', 'in-mutable-hash-keys', 'in-mutable-hash-pairs', + 'in-mutable-hash-values', 'in-mutable-set', 'in-naturals', + 'in-parallel', 'in-permutations', 'in-port', 'in-producer', 'in-range', + 'in-sequences', 'in-set', 'in-slice', 'in-stream', 'in-string', + 'in-syntax', 'in-value', 'in-values*-sequence', 'in-values-sequence', + 'in-vector', 'in-weak-hash', 'in-weak-hash-keys', 'in-weak-hash-pairs', + 'in-weak-hash-values', 'in-weak-set', 'inexact->exact', + 'inexact-real?', 'inexact?', 'infinite?', 'input-port-append', + 'input-port?', 'inspector?', 'instanceof/c', 'integer->char', + 'integer->integer-bytes', 'integer-bytes->integer', 'integer-in', + 'integer-length', 'integer-sqrt', 'integer-sqrt/remainder', 'integer?', + 'interface->method-names', 'interface-extension?', 'interface?', + 'internal-definition-context-binding-identifiers', + 'internal-definition-context-introduce', + 'internal-definition-context-seal', 'internal-definition-context?', + 'is-a?', 'is-a?/c', 'keyword->string', 'keyword-apply', 'keyword<?', + 'keyword?', 'keywords-match', 'kill-thread', 'last', 'last-pair', + 'lcm', 'length', 'liberal-define-context?', 'link-exists?', 'list', + 'list*', 'list*of', 'list->bytes', 'list->mutable-set', + 'list->mutable-seteq', 'list->mutable-seteqv', 'list->set', + 'list->seteq', 'list->seteqv', 'list->string', 'list->vector', + 'list->weak-set', 'list->weak-seteq', 'list->weak-seteqv', + 'list-contract?', 'list-prefix?', 'list-ref', 'list-set', 'list-tail', + 'list-update', 'list/c', 'list?', 'listen-port-number?', 'listof', + 'load', 'load-extension', 'load-on-demand-enabled', 'load-relative', + 'load-relative-extension', 'load/cd', 'load/use-compiled', + 'local-expand', 'local-expand/capture-lifts', + 'local-transformer-expand', 'local-transformer-expand/capture-lifts', + 'locale-string-encoding', 'log', 'log-all-levels', 'log-level-evt', + 'log-level?', 'log-max-level', 'log-message', 'log-receiver?', + 'logger-name', 'logger?', 'magnitude', 'make-arity-at-least', + 'make-base-empty-namespace', 'make-base-namespace', 'make-bytes', + 'make-channel', 'make-chaperone-contract', + 'make-continuation-mark-key', 'make-continuation-prompt-tag', + 'make-contract', 'make-custodian', 'make-custodian-box', + 'make-custom-hash', 'make-custom-hash-types', 'make-custom-set', + 'make-custom-set-types', 'make-date', 'make-date*', + 'make-derived-parameter', 'make-directory', 'make-directory*', + 'make-do-sequence', 'make-empty-namespace', + 'make-environment-variables', 'make-ephemeron', 'make-exn', + 'make-exn:break', 'make-exn:break:hang-up', 'make-exn:break:terminate', + 'make-exn:fail', 'make-exn:fail:contract', + 'make-exn:fail:contract:arity', 'make-exn:fail:contract:blame', + 'make-exn:fail:contract:continuation', + 'make-exn:fail:contract:divide-by-zero', + 'make-exn:fail:contract:non-fixnum-result', + 'make-exn:fail:contract:variable', 'make-exn:fail:filesystem', + 'make-exn:fail:filesystem:errno', 'make-exn:fail:filesystem:exists', + 'make-exn:fail:filesystem:missing-module', + 'make-exn:fail:filesystem:version', 'make-exn:fail:network', + 'make-exn:fail:network:errno', 'make-exn:fail:object', + 'make-exn:fail:out-of-memory', 'make-exn:fail:read', + 'make-exn:fail:read:eof', 'make-exn:fail:read:non-char', + 'make-exn:fail:syntax', 'make-exn:fail:syntax:missing-module', + 'make-exn:fail:syntax:unbound', 'make-exn:fail:unsupported', + 'make-exn:fail:user', 'make-file-or-directory-link', + 'make-flat-contract', 'make-fsemaphore', 'make-generic', + 'make-handle-get-preference-locked', 'make-hash', + 'make-hash-placeholder', 'make-hasheq', 'make-hasheq-placeholder', + 'make-hasheqv', 'make-hasheqv-placeholder', + 'make-immutable-custom-hash', 'make-immutable-hash', + 'make-immutable-hasheq', 'make-immutable-hasheqv', + 'make-impersonator-property', 'make-input-port', + 'make-input-port/read-to-peek', 'make-inspector', + 'make-keyword-procedure', 'make-known-char-range-list', + 'make-limited-input-port', 'make-list', 'make-lock-file-name', + 'make-log-receiver', 'make-logger', 'make-mixin-contract', + 'make-mutable-custom-set', 'make-none/c', 'make-object', + 'make-output-port', 'make-parameter', 'make-parent-directory*', + 'make-phantom-bytes', 'make-pipe', 'make-pipe-with-specials', + 'make-placeholder', 'make-plumber', 'make-polar', 'make-prefab-struct', + 'make-primitive-class', 'make-proj-contract', + 'make-pseudo-random-generator', 'make-reader-graph', 'make-readtable', + 'make-rectangular', 'make-rename-transformer', + 'make-resolved-module-path', 'make-security-guard', 'make-semaphore', + 'make-set!-transformer', 'make-shared-bytes', 'make-sibling-inspector', + 'make-special-comment', 'make-srcloc', 'make-string', + 'make-struct-field-accessor', 'make-struct-field-mutator', + 'make-struct-type', 'make-struct-type-property', + 'make-syntax-delta-introducer', 'make-syntax-introducer', + 'make-temporary-file', 'make-tentative-pretty-print-output-port', + 'make-thread-cell', 'make-thread-group', 'make-vector', + 'make-weak-box', 'make-weak-custom-hash', 'make-weak-custom-set', + 'make-weak-hash', 'make-weak-hasheq', 'make-weak-hasheqv', + 'make-will-executor', 'map', 'match-equality-test', + 'matches-arity-exactly?', 'max', 'mcar', 'mcdr', 'mcons', 'member', + 'member-name-key-hash-code', 'member-name-key=?', 'member-name-key?', + 'memf', 'memq', 'memv', 'merge-input', 'method-in-interface?', 'min', + 'mixin-contract', 'module->exports', 'module->imports', + 'module->language-info', 'module->namespace', + 'module-compiled-cross-phase-persistent?', 'module-compiled-exports', + 'module-compiled-imports', 'module-compiled-language-info', + 'module-compiled-name', 'module-compiled-submodules', + 'module-declared?', 'module-path-index-join', + 'module-path-index-resolve', 'module-path-index-split', + 'module-path-index-submodule', 'module-path-index?', 'module-path?', + 'module-predefined?', 'module-provide-protected?', 'modulo', 'mpair?', + 'mutable-set', 'mutable-seteq', 'mutable-seteqv', 'n->th', + 'nack-guard-evt', 'namespace-anchor->empty-namespace', + 'namespace-anchor->namespace', 'namespace-anchor?', + 'namespace-attach-module', 'namespace-attach-module-declaration', + 'namespace-base-phase', 'namespace-mapped-symbols', + 'namespace-module-identifier', 'namespace-module-registry', + 'namespace-require', 'namespace-require/constant', + 'namespace-require/copy', 'namespace-require/expansion-time', + 'namespace-set-variable-value!', 'namespace-symbol->identifier', + 'namespace-syntax-introduce', 'namespace-undefine-variable!', + 'namespace-unprotect-module', 'namespace-variable-value', 'namespace?', + 'nan?', 'natural-number/c', 'negate', 'negative?', 'never-evt', + 'new-∀/c', 'new-∃/c', 'newline', 'ninth', 'non-empty-listof', + 'non-empty-string?', 'none/c', 'normal-case-path', 'normalize-arity', + 'normalize-path', 'normalized-arity?', 'not', 'not/c', 'null', 'null?', + 'number->string', 'number?', 'numerator', 'object%', 'object->vector', + 'object-info', 'object-interface', 'object-method-arity-includes?', + 'object-name', 'object-or-false=?', 'object=?', 'object?', 'odd?', + 'one-of/c', 'open-input-bytes', 'open-input-file', + 'open-input-output-file', 'open-input-string', 'open-output-bytes', + 'open-output-file', 'open-output-nowhere', 'open-output-string', + 'or/c', 'order-of-magnitude', 'ormap', 'other-execute-bit', + 'other-read-bit', 'other-write-bit', 'output-port?', 'pair?', + 'parameter-procedure=?', 'parameter/c', 'parameter?', + 'parameterization?', 'parse-command-line', 'partition', 'path->bytes', + 'path->complete-path', 'path->directory-path', 'path->string', + 'path-add-suffix', 'path-convention-type', 'path-element->bytes', + 'path-element->string', 'path-element?', 'path-for-some-system?', + 'path-list-string->path-list', 'path-only', 'path-replace-suffix', + 'path-string?', 'path<?', 'path?', 'pathlist-closure', 'peek-byte', + 'peek-byte-or-special', 'peek-bytes', 'peek-bytes!', 'peek-bytes!-evt', + 'peek-bytes-avail!', 'peek-bytes-avail!*', 'peek-bytes-avail!-evt', + 'peek-bytes-avail!/enable-break', 'peek-bytes-evt', 'peek-char', + 'peek-char-or-special', 'peek-string', 'peek-string!', + 'peek-string!-evt', 'peek-string-evt', 'peeking-input-port', + 'permutations', 'phantom-bytes?', 'pi', 'pi.f', 'pipe-content-length', + 'place-break', 'place-channel', 'place-channel-get', + 'place-channel-put', 'place-channel-put/get', 'place-channel?', + 'place-dead-evt', 'place-enabled?', 'place-kill', 'place-location?', + 'place-message-allowed?', 'place-sleep', 'place-wait', 'place?', + 'placeholder-get', 'placeholder-set!', 'placeholder?', + 'plumber-add-flush!', 'plumber-flush-all', + 'plumber-flush-handle-remove!', 'plumber-flush-handle?', 'plumber?', + 'poll-guard-evt', 'port->bytes', 'port->bytes-lines', 'port->lines', + 'port->list', 'port->string', 'port-closed-evt', 'port-closed?', + 'port-commit-peeked', 'port-count-lines!', 'port-count-lines-enabled', + 'port-counts-lines?', 'port-display-handler', 'port-file-identity', + 'port-file-unlock', 'port-next-location', 'port-number?', + 'port-print-handler', 'port-progress-evt', + 'port-provides-progress-evts?', 'port-read-handler', + 'port-try-file-lock?', 'port-write-handler', 'port-writes-atomic?', + 'port-writes-special?', 'port?', 'positive?', 'predicate/c', + 'prefab-key->struct-type', 'prefab-key?', 'prefab-struct-key', + 'preferences-lock-file-mode', 'pregexp', 'pregexp?', 'pretty-display', + 'pretty-format', 'pretty-print', 'pretty-print-.-symbol-without-bars', + 'pretty-print-abbreviate-read-macros', 'pretty-print-columns', + 'pretty-print-current-style-table', 'pretty-print-depth', + 'pretty-print-exact-as-decimal', 'pretty-print-extend-style-table', + 'pretty-print-handler', 'pretty-print-newline', + 'pretty-print-post-print-hook', 'pretty-print-pre-print-hook', + 'pretty-print-print-hook', 'pretty-print-print-line', + 'pretty-print-remap-stylable', 'pretty-print-show-inexactness', + 'pretty-print-size-hook', 'pretty-print-style-table?', + 'pretty-printing', 'pretty-write', 'primitive-closure?', + 'primitive-result-arity', 'primitive?', 'print', 'print-as-expression', + 'print-boolean-long-form', 'print-box', 'print-graph', + 'print-hash-table', 'print-mpair-curly-braces', + 'print-pair-curly-braces', 'print-reader-abbreviations', + 'print-struct', 'print-syntax-width', 'print-unreadable', + 'print-vector-length', 'printable/c', 'printable<%>', 'printf', + 'println', 'procedure->method', 'procedure-arity', + 'procedure-arity-includes/c', 'procedure-arity-includes?', + 'procedure-arity?', 'procedure-closure-contents-eq?', + 'procedure-extract-target', 'procedure-keywords', + 'procedure-reduce-arity', 'procedure-reduce-keyword-arity', + 'procedure-rename', 'procedure-result-arity', 'procedure-specialize', + 'procedure-struct-type?', 'procedure?', 'process', 'process*', + 'process*/ports', 'process/ports', 'processor-count', 'progress-evt?', + 'promise-forced?', 'promise-running?', 'promise/c', 'promise/name?', + 'promise?', 'prop:arity-string', 'prop:arrow-contract', + 'prop:arrow-contract-get-info', 'prop:arrow-contract?', 'prop:blame', + 'prop:chaperone-contract', 'prop:checked-procedure', 'prop:contract', + 'prop:contracted', 'prop:custom-print-quotable', 'prop:custom-write', + 'prop:dict', 'prop:dict/contract', 'prop:equal+hash', 'prop:evt', + 'prop:exn:missing-module', 'prop:exn:srclocs', + 'prop:expansion-contexts', 'prop:flat-contract', + 'prop:impersonator-of', 'prop:input-port', + 'prop:liberal-define-context', 'prop:object-name', + 'prop:opt-chaperone-contract', 'prop:opt-chaperone-contract-get-test', + 'prop:opt-chaperone-contract?', 'prop:orc-contract', + 'prop:orc-contract-get-subcontracts', 'prop:orc-contract?', + 'prop:output-port', 'prop:place-location', 'prop:procedure', + 'prop:recursive-contract', 'prop:recursive-contract-unroll', + 'prop:recursive-contract?', 'prop:rename-transformer', 'prop:sequence', + 'prop:set!-transformer', 'prop:stream', 'proper-subset?', + 'pseudo-random-generator->vector', 'pseudo-random-generator-vector?', + 'pseudo-random-generator?', 'put-preferences', 'putenv', 'quotient', + 'quotient/remainder', 'radians->degrees', 'raise', + 'raise-argument-error', 'raise-arguments-error', 'raise-arity-error', + 'raise-blame-error', 'raise-contract-error', 'raise-mismatch-error', + 'raise-not-cons-blame-error', 'raise-range-error', + 'raise-result-error', 'raise-syntax-error', 'raise-type-error', + 'raise-user-error', 'random', 'random-seed', 'range', 'rational?', + 'rationalize', 'read', 'read-accept-bar-quote', 'read-accept-box', + 'read-accept-compiled', 'read-accept-dot', 'read-accept-graph', + 'read-accept-infix-dot', 'read-accept-lang', 'read-accept-quasiquote', + 'read-accept-reader', 'read-byte', 'read-byte-or-special', + 'read-bytes', 'read-bytes!', 'read-bytes!-evt', 'read-bytes-avail!', + 'read-bytes-avail!*', 'read-bytes-avail!-evt', + 'read-bytes-avail!/enable-break', 'read-bytes-evt', 'read-bytes-line', + 'read-bytes-line-evt', 'read-case-sensitive', 'read-cdot', 'read-char', + 'read-char-or-special', 'read-curly-brace-as-paren', + 'read-curly-brace-with-tag', 'read-decimal-as-inexact', + 'read-eval-print-loop', 'read-language', 'read-line', 'read-line-evt', + 'read-on-demand-source', 'read-square-bracket-as-paren', + 'read-square-bracket-with-tag', 'read-string', 'read-string!', + 'read-string!-evt', 'read-string-evt', 'read-syntax', + 'read-syntax/recursive', 'read/recursive', 'readtable-mapping', + 'readtable?', 'real->decimal-string', 'real->double-flonum', + 'real->floating-point-bytes', 'real->single-flonum', 'real-in', + 'real-part', 'real?', 'reencode-input-port', 'reencode-output-port', + 'regexp', 'regexp-match', 'regexp-match*', 'regexp-match-evt', + 'regexp-match-exact?', 'regexp-match-peek', + 'regexp-match-peek-immediate', 'regexp-match-peek-positions', + 'regexp-match-peek-positions*', + 'regexp-match-peek-positions-immediate', + 'regexp-match-peek-positions-immediate/end', + 'regexp-match-peek-positions/end', 'regexp-match-positions', + 'regexp-match-positions*', 'regexp-match-positions/end', + 'regexp-match/end', 'regexp-match?', 'regexp-max-lookbehind', + 'regexp-quote', 'regexp-replace', 'regexp-replace*', + 'regexp-replace-quote', 'regexp-replaces', 'regexp-split', + 'regexp-try-match', 'regexp?', 'relative-path?', 'relocate-input-port', + 'relocate-output-port', 'remainder', 'remf', 'remf*', 'remove', + 'remove*', 'remove-duplicates', 'remq', 'remq*', 'remv', 'remv*', + 'rename-contract', 'rename-file-or-directory', + 'rename-transformer-target', 'rename-transformer?', 'replace-evt', + 'reroot-path', 'resolve-path', 'resolved-module-path-name', + 'resolved-module-path?', 'rest', 'reverse', 'round', 'second', + 'seconds->date', 'security-guard?', 'semaphore-peek-evt', + 'semaphore-peek-evt?', 'semaphore-post', 'semaphore-try-wait?', + 'semaphore-wait', 'semaphore-wait/enable-break', 'semaphore?', + 'sequence->list', 'sequence->stream', 'sequence-add-between', + 'sequence-andmap', 'sequence-append', 'sequence-count', + 'sequence-filter', 'sequence-fold', 'sequence-for-each', + 'sequence-generate', 'sequence-generate*', 'sequence-length', + 'sequence-map', 'sequence-ormap', 'sequence-ref', 'sequence-tail', + 'sequence/c', 'sequence?', 'set', 'set!-transformer-procedure', + 'set!-transformer?', 'set->list', 'set->stream', 'set-add', 'set-add!', + 'set-box!', 'set-clear', 'set-clear!', 'set-copy', 'set-copy-clear', + 'set-count', 'set-empty?', 'set-eq?', 'set-equal?', 'set-eqv?', + 'set-first', 'set-for-each', 'set-implements/c', 'set-implements?', + 'set-intersect', 'set-intersect!', 'set-map', 'set-mcar!', 'set-mcdr!', + 'set-member?', 'set-mutable?', 'set-phantom-bytes!', + 'set-port-next-location!', 'set-remove', 'set-remove!', 'set-rest', + 'set-some-basic-contracts!', 'set-subtract', 'set-subtract!', + 'set-symmetric-difference', 'set-symmetric-difference!', 'set-union', + 'set-union!', 'set-weak?', 'set/c', 'set=?', 'set?', 'seteq', 'seteqv', + 'seventh', 'sgn', 'shared-bytes', 'shell-execute', 'shrink-path-wrt', + 'shuffle', 'simple-form-path', 'simplify-path', 'sin', + 'single-flonum?', 'sinh', 'sixth', 'skip-projection-wrapper?', 'sleep', + 'some-system-path->string', 'sort', 'special-comment-value', + 'special-comment?', 'special-filter-input-port', 'split-at', + 'split-at-right', 'split-common-prefix', 'split-path', 'splitf-at', + 'splitf-at-right', 'sqr', 'sqrt', 'srcloc', 'srcloc->string', + 'srcloc-column', 'srcloc-line', 'srcloc-position', 'srcloc-source', + 'srcloc-span', 'srcloc?', 'stop-after', 'stop-before', 'stream->list', + 'stream-add-between', 'stream-andmap', 'stream-append', 'stream-count', + 'stream-empty?', 'stream-filter', 'stream-first', 'stream-fold', + 'stream-for-each', 'stream-length', 'stream-map', 'stream-ormap', + 'stream-ref', 'stream-rest', 'stream-tail', 'stream/c', 'stream?', + 'string', 'string->bytes/latin-1', 'string->bytes/locale', + 'string->bytes/utf-8', 'string->immutable-string', 'string->keyword', + 'string->list', 'string->number', 'string->path', + 'string->path-element', 'string->some-system-path', 'string->symbol', + 'string->uninterned-symbol', 'string->unreadable-symbol', + 'string-append', 'string-append*', 'string-ci<=?', 'string-ci<?', + 'string-ci=?', 'string-ci>=?', 'string-ci>?', 'string-contains?', + 'string-copy', 'string-copy!', 'string-downcase', + 'string-environment-variable-name?', 'string-fill!', 'string-foldcase', + 'string-join', 'string-len/c', 'string-length', 'string-locale-ci<?', + 'string-locale-ci=?', 'string-locale-ci>?', 'string-locale-downcase', + 'string-locale-upcase', 'string-locale<?', 'string-locale=?', + 'string-locale>?', 'string-no-nuls?', 'string-normalize-nfc', + 'string-normalize-nfd', 'string-normalize-nfkc', + 'string-normalize-nfkd', 'string-normalize-spaces', 'string-port?', + 'string-prefix?', 'string-ref', 'string-replace', 'string-set!', + 'string-split', 'string-suffix?', 'string-titlecase', 'string-trim', + 'string-upcase', 'string-utf-8-length', 'string<=?', 'string<?', + 'string=?', 'string>=?', 'string>?', 'string?', 'struct->vector', + 'struct-accessor-procedure?', 'struct-constructor-procedure?', + 'struct-info', 'struct-mutator-procedure?', + 'struct-predicate-procedure?', 'struct-type-info', + 'struct-type-make-constructor', 'struct-type-make-predicate', + 'struct-type-property-accessor-procedure?', 'struct-type-property/c', + 'struct-type-property?', 'struct-type?', 'struct:arity-at-least', + 'struct:arrow-contract-info', 'struct:date', 'struct:date*', + 'struct:exn', 'struct:exn:break', 'struct:exn:break:hang-up', + 'struct:exn:break:terminate', 'struct:exn:fail', + 'struct:exn:fail:contract', 'struct:exn:fail:contract:arity', + 'struct:exn:fail:contract:blame', + 'struct:exn:fail:contract:continuation', + 'struct:exn:fail:contract:divide-by-zero', + 'struct:exn:fail:contract:non-fixnum-result', + 'struct:exn:fail:contract:variable', 'struct:exn:fail:filesystem', + 'struct:exn:fail:filesystem:errno', + 'struct:exn:fail:filesystem:exists', + 'struct:exn:fail:filesystem:missing-module', + 'struct:exn:fail:filesystem:version', 'struct:exn:fail:network', + 'struct:exn:fail:network:errno', 'struct:exn:fail:object', + 'struct:exn:fail:out-of-memory', 'struct:exn:fail:read', + 'struct:exn:fail:read:eof', 'struct:exn:fail:read:non-char', + 'struct:exn:fail:syntax', 'struct:exn:fail:syntax:missing-module', + 'struct:exn:fail:syntax:unbound', 'struct:exn:fail:unsupported', + 'struct:exn:fail:user', 'struct:srcloc', + 'struct:wrapped-extra-arg-arrow', 'struct?', 'sub1', 'subbytes', + 'subclass?', 'subclass?/c', 'subprocess', 'subprocess-group-enabled', + 'subprocess-kill', 'subprocess-pid', 'subprocess-status', + 'subprocess-wait', 'subprocess?', 'subset?', 'substring', 'suggest/c', + 'symbol->string', 'symbol-interned?', 'symbol-unreadable?', 'symbol<?', + 'symbol=?', 'symbol?', 'symbols', 'sync', 'sync/enable-break', + 'sync/timeout', 'sync/timeout/enable-break', 'syntax->datum', + 'syntax->list', 'syntax-arm', 'syntax-column', 'syntax-debug-info', + 'syntax-disarm', 'syntax-e', 'syntax-line', + 'syntax-local-bind-syntaxes', 'syntax-local-certifier', + 'syntax-local-context', 'syntax-local-expand-expression', + 'syntax-local-get-shadower', 'syntax-local-identifier-as-binding', + 'syntax-local-introduce', 'syntax-local-lift-context', + 'syntax-local-lift-expression', 'syntax-local-lift-module', + 'syntax-local-lift-module-end-declaration', + 'syntax-local-lift-provide', 'syntax-local-lift-require', + 'syntax-local-lift-values-expression', + 'syntax-local-make-definition-context', + 'syntax-local-make-delta-introducer', + 'syntax-local-module-defined-identifiers', + 'syntax-local-module-exports', + 'syntax-local-module-required-identifiers', 'syntax-local-name', + 'syntax-local-phase-level', 'syntax-local-submodules', + 'syntax-local-transforming-module-provides?', 'syntax-local-value', + 'syntax-local-value/immediate', 'syntax-original?', 'syntax-position', + 'syntax-property', 'syntax-property-preserved?', + 'syntax-property-symbol-keys', 'syntax-protect', 'syntax-rearm', + 'syntax-recertify', 'syntax-shift-phase-level', 'syntax-source', + 'syntax-source-module', 'syntax-span', 'syntax-taint', + 'syntax-tainted?', 'syntax-track-origin', + 'syntax-transforming-module-expression?', + 'syntax-transforming-with-lifts?', 'syntax-transforming?', 'syntax/c', + 'syntax?', 'system', 'system*', 'system*/exit-code', + 'system-big-endian?', 'system-idle-evt', 'system-language+country', + 'system-library-subpath', 'system-path-convention-type', 'system-type', + 'system/exit-code', 'tail-marks-match?', 'take', 'take-common-prefix', + 'take-right', 'takef', 'takef-right', 'tan', 'tanh', + 'tcp-abandon-port', 'tcp-accept', 'tcp-accept-evt', + 'tcp-accept-ready?', 'tcp-accept/enable-break', 'tcp-addresses', + 'tcp-close', 'tcp-connect', 'tcp-connect/enable-break', 'tcp-listen', + 'tcp-listener?', 'tcp-port?', 'tentative-pretty-print-port-cancel', + 'tentative-pretty-print-port-transfer', 'tenth', 'terminal-port?', + 'the-unsupplied-arg', 'third', 'thread', 'thread-cell-ref', + 'thread-cell-set!', 'thread-cell-values?', 'thread-cell?', + 'thread-dead-evt', 'thread-dead?', 'thread-group?', 'thread-receive', + 'thread-receive-evt', 'thread-resume', 'thread-resume-evt', + 'thread-rewind-receive', 'thread-running?', 'thread-send', + 'thread-suspend', 'thread-suspend-evt', 'thread-try-receive', + 'thread-wait', 'thread/suspend-to-kill', 'thread?', 'time-apply', + 'touch', 'transplant-input-port', 'transplant-output-port', 'true', + 'truncate', 'udp-addresses', 'udp-bind!', 'udp-bound?', 'udp-close', + 'udp-connect!', 'udp-connected?', 'udp-multicast-interface', + 'udp-multicast-join-group!', 'udp-multicast-leave-group!', + 'udp-multicast-loopback?', 'udp-multicast-set-interface!', + 'udp-multicast-set-loopback!', 'udp-multicast-set-ttl!', + 'udp-multicast-ttl', 'udp-open-socket', 'udp-receive!', + 'udp-receive!*', 'udp-receive!-evt', 'udp-receive!/enable-break', + 'udp-receive-ready-evt', 'udp-send', 'udp-send*', 'udp-send-evt', + 'udp-send-ready-evt', 'udp-send-to', 'udp-send-to*', 'udp-send-to-evt', + 'udp-send-to/enable-break', 'udp-send/enable-break', 'udp?', 'unbox', + 'uncaught-exception-handler', 'unit?', 'unspecified-dom', + 'unsupplied-arg?', 'use-collection-link-paths', + 'use-compiled-file-paths', 'use-user-specific-search-paths', + 'user-execute-bit', 'user-read-bit', 'user-write-bit', 'value-blame', + 'value-contract', 'values', 'variable-reference->empty-namespace', + 'variable-reference->module-base-phase', + 'variable-reference->module-declaration-inspector', + 'variable-reference->module-path-index', + 'variable-reference->module-source', 'variable-reference->namespace', + 'variable-reference->phase', + 'variable-reference->resolved-module-path', + 'variable-reference-constant?', 'variable-reference?', 'vector', + 'vector->immutable-vector', 'vector->list', + 'vector->pseudo-random-generator', 'vector->pseudo-random-generator!', + 'vector->values', 'vector-append', 'vector-argmax', 'vector-argmin', + 'vector-copy', 'vector-copy!', 'vector-count', 'vector-drop', + 'vector-drop-right', 'vector-fill!', 'vector-filter', + 'vector-filter-not', 'vector-immutable', 'vector-immutable/c', + 'vector-immutableof', 'vector-length', 'vector-map', 'vector-map!', + 'vector-member', 'vector-memq', 'vector-memv', 'vector-ref', + 'vector-set!', 'vector-set*!', 'vector-set-performance-stats!', + 'vector-split-at', 'vector-split-at-right', 'vector-take', + 'vector-take-right', 'vector/c', 'vector?', 'vectorof', 'version', + 'void', 'void?', 'weak-box-value', 'weak-box?', 'weak-set', + 'weak-seteq', 'weak-seteqv', 'will-execute', 'will-executor?', + 'will-register', 'will-try-execute', 'with-input-from-bytes', + 'with-input-from-file', 'with-input-from-string', + 'with-output-to-bytes', 'with-output-to-file', 'with-output-to-string', + 'would-be-future', 'wrap-evt', 'wrapped-extra-arg-arrow', + 'wrapped-extra-arg-arrow-extra-neg-party-argument', + 'wrapped-extra-arg-arrow-real-func', 'wrapped-extra-arg-arrow?', + 'writable<%>', 'write', 'write-byte', 'write-bytes', + 'write-bytes-avail', 'write-bytes-avail*', 'write-bytes-avail-evt', + 'write-bytes-avail/enable-break', 'write-char', 'write-special', + 'write-special-avail*', 'write-special-evt', 'write-string', + 'write-to-file', 'writeln', 'xor', 'zero?', '~.a', '~.s', '~.v', '~a', + '~e', '~r', '~s', '~v' ) _opening_parenthesis = r'[([{]' @@ -1270,7 +1270,7 @@ class RacketLexer(RegexLexer): ], 'datum': [ (r'(?s)#;|#*', Comment), - (u';[^\\n\\r\x85\u2028\u2029]*', Comment.Single), + (r';[^\n\r\x85\u2028\u2029]*', Comment.Single), (r'#\|', Comment.Multiline, 'block-comment'), # Whitespaces diff --git a/pygments/lexers/markup.py b/pygments/lexers/markup.py index bd814a54..7ee161c1 100644 --- a/pygments/lexers/markup.py +++ b/pygments/lexers/markup.py @@ -168,8 +168,8 @@ class RstLexer(RegexLexer): yield from do_insertions(ins, lexer.get_tokens_unprocessed(code)) # from docutils.parsers.rst.states - closers = u'\'")]}>\u2019\u201d\xbb!?' - unicode_delimiters = u'\u2010\u2011\u2012\u2013\u2014\u00a0' + closers = '\'")]}>\u2019\u201d\xbb!?' + unicode_delimiters = '\u2010\u2011\u2012\u2013\u2014\u00a0' end_string_suffix = (r'((?=$)|(?=[-/:.,; \n\x00%s%s]))' % (re.escape(unicode_delimiters), re.escape(closers))) diff --git a/pygments/lexers/mime.py b/pygments/lexers/mime.py index f5bae8bd..4c6bbd5f 100644 --- a/pygments/lexers/mime.py +++ b/pygments/lexers/mime.py @@ -100,7 +100,7 @@ class MIMELexer(RegexLexer): # skip first newline if entire_body[0] == '\n': - yield pos_body_start, Text.Whitespace, u'\n' + yield pos_body_start, Text.Whitespace, '\n' pos_body_start = pos_body_start + 1 entire_body = entire_body[1:] @@ -176,7 +176,7 @@ class MIMELexer(RegexLexer): prefix_len = match.start(1) - match.start(0) yield match.start(0), Text.Whitespace, match.group(0)[:prefix_len] yield match.start(1), Name.Label, match.group(2) - yield match.end(2), String.Delimiter, u"/" + yield match.end(2), String.Delimiter, '/' yield match.start(3), Name.Label, match.group(3) def get_content_type_subtokens(self, match): diff --git a/pygments/lexers/modeling.py b/pygments/lexers/modeling.py index 167ec86a..5a9071d6 100644 --- a/pygments/lexers/modeling.py +++ b/pygments/lexers/modeling.py @@ -38,7 +38,7 @@ class ModelicaLexer(RegexLexer): tokens = { 'whitespace': [ - (u'[\\s\ufeff]+', Text), + (r'[\s\ufeff]+', Text), (r'//[^\n]*\n?', Comment.Single), (r'/\*.*?\*/', Comment.Multiline) ], diff --git a/pygments/lexers/modula2.py b/pygments/lexers/modula2.py index b95bfaec..4794ebce 100644 --- a/pygments/lexers/modula2.py +++ b/pygments/lexers/modula2.py @@ -1547,15 +1547,15 @@ class Modula2Lexer(RegexLexer): # substitute lexemes when in Algol mode if self.algol_publication_mode: if value == '#': - value = u'≠' + value = '≠' elif value == '<=': - value = u'≤' + value = '≤' elif value == '>=': - value = u'≥' + value = '≥' elif value == '==': - value = u'≡' + value = '≡' elif value == '*.': - value = u'•' + value = '•' # return result yield index, token, value diff --git a/pygments/lexers/perl.py b/pygments/lexers/perl.py index 4a3ca300..741de3fd 100644 --- a/pygments/lexers/perl.py +++ b/pygments/lexers/perl.py @@ -415,70 +415,70 @@ class Perl6Lexer(ExtendedRegexLexer): # Perl 6 has a *lot* of possible bracketing characters # this list was lifted from STD.pm6 (https://github.com/perl6/std) PERL6_BRACKETS = { - u'\u0028': u'\u0029', u'\u003c': u'\u003e', u'\u005b': u'\u005d', - u'\u007b': u'\u007d', u'\u00ab': u'\u00bb', u'\u0f3a': u'\u0f3b', - u'\u0f3c': u'\u0f3d', u'\u169b': u'\u169c', u'\u2018': u'\u2019', - u'\u201a': u'\u2019', u'\u201b': u'\u2019', u'\u201c': u'\u201d', - u'\u201e': u'\u201d', u'\u201f': u'\u201d', u'\u2039': u'\u203a', - u'\u2045': u'\u2046', u'\u207d': u'\u207e', u'\u208d': u'\u208e', - u'\u2208': u'\u220b', u'\u2209': u'\u220c', u'\u220a': u'\u220d', - u'\u2215': u'\u29f5', u'\u223c': u'\u223d', u'\u2243': u'\u22cd', - u'\u2252': u'\u2253', u'\u2254': u'\u2255', u'\u2264': u'\u2265', - u'\u2266': u'\u2267', u'\u2268': u'\u2269', u'\u226a': u'\u226b', - u'\u226e': u'\u226f', u'\u2270': u'\u2271', u'\u2272': u'\u2273', - u'\u2274': u'\u2275', u'\u2276': u'\u2277', u'\u2278': u'\u2279', - u'\u227a': u'\u227b', u'\u227c': u'\u227d', u'\u227e': u'\u227f', - u'\u2280': u'\u2281', u'\u2282': u'\u2283', u'\u2284': u'\u2285', - u'\u2286': u'\u2287', u'\u2288': u'\u2289', u'\u228a': u'\u228b', - u'\u228f': u'\u2290', u'\u2291': u'\u2292', u'\u2298': u'\u29b8', - u'\u22a2': u'\u22a3', u'\u22a6': u'\u2ade', u'\u22a8': u'\u2ae4', - u'\u22a9': u'\u2ae3', u'\u22ab': u'\u2ae5', u'\u22b0': u'\u22b1', - u'\u22b2': u'\u22b3', u'\u22b4': u'\u22b5', u'\u22b6': u'\u22b7', - u'\u22c9': u'\u22ca', u'\u22cb': u'\u22cc', u'\u22d0': u'\u22d1', - u'\u22d6': u'\u22d7', u'\u22d8': u'\u22d9', u'\u22da': u'\u22db', - u'\u22dc': u'\u22dd', u'\u22de': u'\u22df', u'\u22e0': u'\u22e1', - u'\u22e2': u'\u22e3', u'\u22e4': u'\u22e5', u'\u22e6': u'\u22e7', - u'\u22e8': u'\u22e9', u'\u22ea': u'\u22eb', u'\u22ec': u'\u22ed', - u'\u22f0': u'\u22f1', u'\u22f2': u'\u22fa', u'\u22f3': u'\u22fb', - u'\u22f4': u'\u22fc', u'\u22f6': u'\u22fd', u'\u22f7': u'\u22fe', - u'\u2308': u'\u2309', u'\u230a': u'\u230b', u'\u2329': u'\u232a', - u'\u23b4': u'\u23b5', u'\u2768': u'\u2769', u'\u276a': u'\u276b', - u'\u276c': u'\u276d', u'\u276e': u'\u276f', u'\u2770': u'\u2771', - u'\u2772': u'\u2773', u'\u2774': u'\u2775', u'\u27c3': u'\u27c4', - u'\u27c5': u'\u27c6', u'\u27d5': u'\u27d6', u'\u27dd': u'\u27de', - u'\u27e2': u'\u27e3', u'\u27e4': u'\u27e5', u'\u27e6': u'\u27e7', - u'\u27e8': u'\u27e9', u'\u27ea': u'\u27eb', u'\u2983': u'\u2984', - u'\u2985': u'\u2986', u'\u2987': u'\u2988', u'\u2989': u'\u298a', - u'\u298b': u'\u298c', u'\u298d': u'\u298e', u'\u298f': u'\u2990', - u'\u2991': u'\u2992', u'\u2993': u'\u2994', u'\u2995': u'\u2996', - u'\u2997': u'\u2998', u'\u29c0': u'\u29c1', u'\u29c4': u'\u29c5', - u'\u29cf': u'\u29d0', u'\u29d1': u'\u29d2', u'\u29d4': u'\u29d5', - u'\u29d8': u'\u29d9', u'\u29da': u'\u29db', u'\u29f8': u'\u29f9', - u'\u29fc': u'\u29fd', u'\u2a2b': u'\u2a2c', u'\u2a2d': u'\u2a2e', - u'\u2a34': u'\u2a35', u'\u2a3c': u'\u2a3d', u'\u2a64': u'\u2a65', - u'\u2a79': u'\u2a7a', u'\u2a7d': u'\u2a7e', u'\u2a7f': u'\u2a80', - u'\u2a81': u'\u2a82', u'\u2a83': u'\u2a84', u'\u2a8b': u'\u2a8c', - u'\u2a91': u'\u2a92', u'\u2a93': u'\u2a94', u'\u2a95': u'\u2a96', - u'\u2a97': u'\u2a98', u'\u2a99': u'\u2a9a', u'\u2a9b': u'\u2a9c', - u'\u2aa1': u'\u2aa2', u'\u2aa6': u'\u2aa7', u'\u2aa8': u'\u2aa9', - u'\u2aaa': u'\u2aab', u'\u2aac': u'\u2aad', u'\u2aaf': u'\u2ab0', - u'\u2ab3': u'\u2ab4', u'\u2abb': u'\u2abc', u'\u2abd': u'\u2abe', - u'\u2abf': u'\u2ac0', u'\u2ac1': u'\u2ac2', u'\u2ac3': u'\u2ac4', - u'\u2ac5': u'\u2ac6', u'\u2acd': u'\u2ace', u'\u2acf': u'\u2ad0', - u'\u2ad1': u'\u2ad2', u'\u2ad3': u'\u2ad4', u'\u2ad5': u'\u2ad6', - u'\u2aec': u'\u2aed', u'\u2af7': u'\u2af8', u'\u2af9': u'\u2afa', - u'\u2e02': u'\u2e03', u'\u2e04': u'\u2e05', u'\u2e09': u'\u2e0a', - u'\u2e0c': u'\u2e0d', u'\u2e1c': u'\u2e1d', u'\u2e20': u'\u2e21', - u'\u3008': u'\u3009', u'\u300a': u'\u300b', u'\u300c': u'\u300d', - u'\u300e': u'\u300f', u'\u3010': u'\u3011', u'\u3014': u'\u3015', - u'\u3016': u'\u3017', u'\u3018': u'\u3019', u'\u301a': u'\u301b', - u'\u301d': u'\u301e', u'\ufd3e': u'\ufd3f', u'\ufe17': u'\ufe18', - u'\ufe35': u'\ufe36', u'\ufe37': u'\ufe38', u'\ufe39': u'\ufe3a', - u'\ufe3b': u'\ufe3c', u'\ufe3d': u'\ufe3e', u'\ufe3f': u'\ufe40', - u'\ufe41': u'\ufe42', u'\ufe43': u'\ufe44', u'\ufe47': u'\ufe48', - u'\ufe59': u'\ufe5a', u'\ufe5b': u'\ufe5c', u'\ufe5d': u'\ufe5e', - u'\uff08': u'\uff09', u'\uff1c': u'\uff1e', u'\uff3b': u'\uff3d', - u'\uff5b': u'\uff5d', u'\uff5f': u'\uff60', u'\uff62': u'\uff63', + '\u0028': '\u0029', '\u003c': '\u003e', '\u005b': '\u005d', + '\u007b': '\u007d', '\u00ab': '\u00bb', '\u0f3a': '\u0f3b', + '\u0f3c': '\u0f3d', '\u169b': '\u169c', '\u2018': '\u2019', + '\u201a': '\u2019', '\u201b': '\u2019', '\u201c': '\u201d', + '\u201e': '\u201d', '\u201f': '\u201d', '\u2039': '\u203a', + '\u2045': '\u2046', '\u207d': '\u207e', '\u208d': '\u208e', + '\u2208': '\u220b', '\u2209': '\u220c', '\u220a': '\u220d', + '\u2215': '\u29f5', '\u223c': '\u223d', '\u2243': '\u22cd', + '\u2252': '\u2253', '\u2254': '\u2255', '\u2264': '\u2265', + '\u2266': '\u2267', '\u2268': '\u2269', '\u226a': '\u226b', + '\u226e': '\u226f', '\u2270': '\u2271', '\u2272': '\u2273', + '\u2274': '\u2275', '\u2276': '\u2277', '\u2278': '\u2279', + '\u227a': '\u227b', '\u227c': '\u227d', '\u227e': '\u227f', + '\u2280': '\u2281', '\u2282': '\u2283', '\u2284': '\u2285', + '\u2286': '\u2287', '\u2288': '\u2289', '\u228a': '\u228b', + '\u228f': '\u2290', '\u2291': '\u2292', '\u2298': '\u29b8', + '\u22a2': '\u22a3', '\u22a6': '\u2ade', '\u22a8': '\u2ae4', + '\u22a9': '\u2ae3', '\u22ab': '\u2ae5', '\u22b0': '\u22b1', + '\u22b2': '\u22b3', '\u22b4': '\u22b5', '\u22b6': '\u22b7', + '\u22c9': '\u22ca', '\u22cb': '\u22cc', '\u22d0': '\u22d1', + '\u22d6': '\u22d7', '\u22d8': '\u22d9', '\u22da': '\u22db', + '\u22dc': '\u22dd', '\u22de': '\u22df', '\u22e0': '\u22e1', + '\u22e2': '\u22e3', '\u22e4': '\u22e5', '\u22e6': '\u22e7', + '\u22e8': '\u22e9', '\u22ea': '\u22eb', '\u22ec': '\u22ed', + '\u22f0': '\u22f1', '\u22f2': '\u22fa', '\u22f3': '\u22fb', + '\u22f4': '\u22fc', '\u22f6': '\u22fd', '\u22f7': '\u22fe', + '\u2308': '\u2309', '\u230a': '\u230b', '\u2329': '\u232a', + '\u23b4': '\u23b5', '\u2768': '\u2769', '\u276a': '\u276b', + '\u276c': '\u276d', '\u276e': '\u276f', '\u2770': '\u2771', + '\u2772': '\u2773', '\u2774': '\u2775', '\u27c3': '\u27c4', + '\u27c5': '\u27c6', '\u27d5': '\u27d6', '\u27dd': '\u27de', + '\u27e2': '\u27e3', '\u27e4': '\u27e5', '\u27e6': '\u27e7', + '\u27e8': '\u27e9', '\u27ea': '\u27eb', '\u2983': '\u2984', + '\u2985': '\u2986', '\u2987': '\u2988', '\u2989': '\u298a', + '\u298b': '\u298c', '\u298d': '\u298e', '\u298f': '\u2990', + '\u2991': '\u2992', '\u2993': '\u2994', '\u2995': '\u2996', + '\u2997': '\u2998', '\u29c0': '\u29c1', '\u29c4': '\u29c5', + '\u29cf': '\u29d0', '\u29d1': '\u29d2', '\u29d4': '\u29d5', + '\u29d8': '\u29d9', '\u29da': '\u29db', '\u29f8': '\u29f9', + '\u29fc': '\u29fd', '\u2a2b': '\u2a2c', '\u2a2d': '\u2a2e', + '\u2a34': '\u2a35', '\u2a3c': '\u2a3d', '\u2a64': '\u2a65', + '\u2a79': '\u2a7a', '\u2a7d': '\u2a7e', '\u2a7f': '\u2a80', + '\u2a81': '\u2a82', '\u2a83': '\u2a84', '\u2a8b': '\u2a8c', + '\u2a91': '\u2a92', '\u2a93': '\u2a94', '\u2a95': '\u2a96', + '\u2a97': '\u2a98', '\u2a99': '\u2a9a', '\u2a9b': '\u2a9c', + '\u2aa1': '\u2aa2', '\u2aa6': '\u2aa7', '\u2aa8': '\u2aa9', + '\u2aaa': '\u2aab', '\u2aac': '\u2aad', '\u2aaf': '\u2ab0', + '\u2ab3': '\u2ab4', '\u2abb': '\u2abc', '\u2abd': '\u2abe', + '\u2abf': '\u2ac0', '\u2ac1': '\u2ac2', '\u2ac3': '\u2ac4', + '\u2ac5': '\u2ac6', '\u2acd': '\u2ace', '\u2acf': '\u2ad0', + '\u2ad1': '\u2ad2', '\u2ad3': '\u2ad4', '\u2ad5': '\u2ad6', + '\u2aec': '\u2aed', '\u2af7': '\u2af8', '\u2af9': '\u2afa', + '\u2e02': '\u2e03', '\u2e04': '\u2e05', '\u2e09': '\u2e0a', + '\u2e0c': '\u2e0d', '\u2e1c': '\u2e1d', '\u2e20': '\u2e21', + '\u3008': '\u3009', '\u300a': '\u300b', '\u300c': '\u300d', + '\u300e': '\u300f', '\u3010': '\u3011', '\u3014': '\u3015', + '\u3016': '\u3017', '\u3018': '\u3019', '\u301a': '\u301b', + '\u301d': '\u301e', '\ufd3e': '\ufd3f', '\ufe17': '\ufe18', + '\ufe35': '\ufe36', '\ufe37': '\ufe38', '\ufe39': '\ufe3a', + '\ufe3b': '\ufe3c', '\ufe3d': '\ufe3e', '\ufe3f': '\ufe40', + '\ufe41': '\ufe42', '\ufe43': '\ufe44', '\ufe47': '\ufe48', + '\ufe59': '\ufe5a', '\ufe5b': '\ufe5c', '\ufe5d': '\ufe5e', + '\uff08': '\uff09', '\uff1c': '\uff1e', '\uff3b': '\uff3d', + '\uff5b': '\uff5d', '\uff5f': '\uff60', '\uff62': '\uff63', } def _build_word_match(words, boundary_regex_fragment=None, prefix='', suffix=''): @@ -602,11 +602,11 @@ class Perl6Lexer(ExtendedRegexLexer): Name.Builtin), (_build_word_match(PERL6_BUILTINS, PERL6_IDENTIFIER_RANGE), Name.Builtin), # copied from PerlLexer - (r'[$@%&][.^:?=!~]?' + PERL6_IDENTIFIER_RANGE + u'+(?:<<.*?>>|<.*?>|«.*?»)*', + (r'[$@%&][.^:?=!~]?' + PERL6_IDENTIFIER_RANGE + '+(?:<<.*?>>|<.*?>|«.*?»)*', Name.Variable), (r'\$[!/](?:<<.*?>>|<.*?>|«.*?»)*', Name.Variable.Global), (r'::\?\w+', Name.Variable.Global), - (r'[$@%&]\*' + PERL6_IDENTIFIER_RANGE + u'+(?:<<.*?>>|<.*?>|«.*?»)*', + (r'[$@%&]\*' + PERL6_IDENTIFIER_RANGE + '+(?:<<.*?>>|<.*?>|«.*?»)*', Name.Variable.Global), (r'\$(?:<.*?>)+', Name.Variable), (r'(?:q|qq|Q)[a-zA-Z]?\s*(?P<adverbs>:[\w\s:]+)?\s*(?P<delimiter>(?P<first_char>[^0-9a-zA-Z:\s])' diff --git a/pygments/lexers/php.py b/pygments/lexers/php.py index 2bad339a..aab502e2 100644 --- a/pygments/lexers/php.py +++ b/pygments/lexers/php.py @@ -116,13 +116,13 @@ class PsyshConsoleLexer(Lexer): insertions = [] for match in line_re.finditer(text): line = match.group() - if line.startswith(u'>>> ') or line.startswith(u'... '): + if line.startswith('>>> ') or line.startswith('... '): insertions.append((len(curcode), [(0, Generic.Prompt, line[:4])])) curcode += line[4:] - elif line.rstrip() == u'...': + elif line.rstrip() == '...': insertions.append((len(curcode), - [(0, Generic.Prompt, u'...')])) + [(0, Generic.Prompt, '...')])) curcode += line[3:] else: if curcode: diff --git a/pygments/lexers/prolog.py b/pygments/lexers/prolog.py index 40ef0df3..94f4024c 100644 --- a/pygments/lexers/prolog.py +++ b/pygments/lexers/prolog.py @@ -55,22 +55,22 @@ class PrologLexer(RegexLexer): (r'(mod|div|not)\b', Operator), (r'_', Keyword), # The don't-care variable (r'([a-z]+)(:)', bygroups(Name.Namespace, Punctuation)), - (u'([a-z\u00c0-\u1fff\u3040-\ud7ff\ue000-\uffef]' - u'[\\w$\u00c0-\u1fff\u3040-\ud7ff\ue000-\uffef]*)' - u'(\\s*)(:-|-->)', + (r'([a-z\u00c0-\u1fff\u3040-\ud7ff\ue000-\uffef]' + r'[\w$\u00c0-\u1fff\u3040-\ud7ff\ue000-\uffef]*)' + r'(\s*)(:-|-->)', bygroups(Name.Function, Text, Operator)), # function defn - (u'([a-z\u00c0-\u1fff\u3040-\ud7ff\ue000-\uffef]' - u'[\\w$\u00c0-\u1fff\u3040-\ud7ff\ue000-\uffef]*)' - u'(\\s*)(\\()', + (r'([a-z\u00c0-\u1fff\u3040-\ud7ff\ue000-\uffef]' + r'[\w$\u00c0-\u1fff\u3040-\ud7ff\ue000-\uffef]*)' + r'(\s*)(\()', bygroups(Name.Function, Text, Punctuation)), - (u'[a-z\u00c0-\u1fff\u3040-\ud7ff\ue000-\uffef]' - u'[\\w$\u00c0-\u1fff\u3040-\ud7ff\ue000-\uffef]*', + (r'[a-z\u00c0-\u1fff\u3040-\ud7ff\ue000-\uffef]' + r'[\w$\u00c0-\u1fff\u3040-\ud7ff\ue000-\uffef]*', String.Atom), # atom, characters # This one includes ! - (u'[#&*+\\-./:<=>?@\\\\^~\u00a1-\u00bf\u2010-\u303f]+', + (r'[#&*+\-./:<=>?@\\^~\u00a1-\u00bf\u2010-\u303f]+', String.Atom), # atom, graphics (r'[A-Z_]\w*', Name.Variable), - (u'\\s+|[\u2000-\u200f\ufff0-\ufffe\uffef]', Text), + (r'\s+|[\u2000-\u200f\ufff0-\ufffe\uffef]', Text), ], 'nested-comment': [ (r'\*/', Comment.Multiline, '#pop'), diff --git a/pygments/lexers/python.py b/pygments/lexers/python.py index 22d21430..10157999 100644 --- a/pygments/lexers/python.py +++ b/pygments/lexers/python.py @@ -646,17 +646,17 @@ class PythonConsoleLexer(Lexer): tb = 0 for match in line_re.finditer(text): line = match.group() - if line.startswith(u'>>> ') or line.startswith(u'... '): + if line.startswith('>>> ') or line.startswith('... '): tb = 0 insertions.append((len(curcode), [(0, Generic.Prompt, line[:4])])) curcode += line[4:] - elif line.rstrip() == u'...' and not tb: + elif line.rstrip() == '...' and not tb: # only a new >>> prompt can end an exception block # otherwise an ellipsis in place of the traceback frames # will be mishandled insertions.append((len(curcode), - [(0, Generic.Prompt, u'...')])) + [(0, Generic.Prompt, '...')])) curcode += line[3:] else: if curcode: @@ -664,8 +664,8 @@ class PythonConsoleLexer(Lexer): insertions, pylexer.get_tokens_unprocessed(curcode)) curcode = '' insertions = [] - if (line.startswith(u'Traceback (most recent call last):') or - re.match(u' File "[^"]+", line \\d+\\n$', line)): + if (line.startswith('Traceback (most recent call last):') or + re.match(' File "[^"]+", line \\d+\\n$', line)): tb = 1 curtb = line tbindex = match.start() @@ -673,7 +673,7 @@ class PythonConsoleLexer(Lexer): yield match.start(), Name.Class, line elif tb: curtb += line - if not (line.startswith(' ') or line.strip() == u'...'): + if not (line.startswith(' ') or line.strip() == '...'): tb = 0 for i, t, v in tblexer.get_tokens_unprocessed(curtb): yield tbindex+i, t, v diff --git a/pygments/lexers/qvt.py b/pygments/lexers/qvt.py index 515d7270..680d3fb8 100644 --- a/pygments/lexers/qvt.py +++ b/pygments/lexers/qvt.py @@ -18,7 +18,7 @@ __all__ = ['QVToLexer'] class QVToLexer(RegexLexer): - u""" + """ For the `QVT Operational Mapping language <http://www.omg.org/spec/QVT/1.1/>`_. Reference for implementing this: «Meta Object Facility (MOF) 2.0 diff --git a/pygments/lexers/rdf.py b/pygments/lexers/rdf.py index 8bbd4b0c..136850bd 100644 --- a/pygments/lexers/rdf.py +++ b/pygments/lexers/rdf.py @@ -31,27 +31,27 @@ class SparqlLexer(RegexLexer): # character group definitions :: - PN_CHARS_BASE_GRP = (u'a-zA-Z' - u'\u00c0-\u00d6' - u'\u00d8-\u00f6' - u'\u00f8-\u02ff' - u'\u0370-\u037d' - u'\u037f-\u1fff' - u'\u200c-\u200d' - u'\u2070-\u218f' - u'\u2c00-\u2fef' - u'\u3001-\ud7ff' - u'\uf900-\ufdcf' - u'\ufdf0-\ufffd') + PN_CHARS_BASE_GRP = ('a-zA-Z' + '\u00c0-\u00d6' + '\u00d8-\u00f6' + '\u00f8-\u02ff' + '\u0370-\u037d' + '\u037f-\u1fff' + '\u200c-\u200d' + '\u2070-\u218f' + '\u2c00-\u2fef' + '\u3001-\ud7ff' + '\uf900-\ufdcf' + '\ufdf0-\ufffd') PN_CHARS_U_GRP = (PN_CHARS_BASE_GRP + '_') PN_CHARS_GRP = (PN_CHARS_U_GRP + r'\-' + r'0-9' + - u'\u00b7' + - u'\u0300-\u036f' + - u'\u203f-\u2040') + '\u00b7' + + '\u0300-\u036f' + + '\u203f-\u2040') HEX_GRP = '0-9A-Fa-f' @@ -76,8 +76,8 @@ class SparqlLexer(RegexLexer): PN_PREFIX = PN_CHARS_BASE + '(?:[' + PN_CHARS_GRP + '.]*' + PN_CHARS + ')?' - VARNAME = u'[0-9' + PN_CHARS_U_GRP + '][' + PN_CHARS_U_GRP + \ - u'0-9\u00b7\u0300-\u036f\u203f-\u2040]*' + VARNAME = '[0-9' + PN_CHARS_U_GRP + '][' + PN_CHARS_U_GRP + \ + '0-9\u00b7\u0300-\u036f\u203f-\u2040]*' PERCENT = '%' + HEX + HEX @@ -288,27 +288,27 @@ class ShExCLexer(RegexLexer): # character group definitions :: - PN_CHARS_BASE_GRP = (u'a-zA-Z' - u'\u00c0-\u00d6' - u'\u00d8-\u00f6' - u'\u00f8-\u02ff' - u'\u0370-\u037d' - u'\u037f-\u1fff' - u'\u200c-\u200d' - u'\u2070-\u218f' - u'\u2c00-\u2fef' - u'\u3001-\ud7ff' - u'\uf900-\ufdcf' - u'\ufdf0-\ufffd') + PN_CHARS_BASE_GRP = ('a-zA-Z' + '\u00c0-\u00d6' + '\u00d8-\u00f6' + '\u00f8-\u02ff' + '\u0370-\u037d' + '\u037f-\u1fff' + '\u200c-\u200d' + '\u2070-\u218f' + '\u2c00-\u2fef' + '\u3001-\ud7ff' + '\uf900-\ufdcf' + '\ufdf0-\ufffd') PN_CHARS_U_GRP = (PN_CHARS_BASE_GRP + '_') PN_CHARS_GRP = (PN_CHARS_U_GRP + r'\-' + r'0-9' + - u'\u00b7' + - u'\u0300-\u036f' + - u'\u203f-\u2040') + '\u00b7' + + '\u0300-\u036f' + + '\u203f-\u2040') HEX_GRP = '0-9A-Fa-f' diff --git a/pygments/lexers/scripting.py b/pygments/lexers/scripting.py index 843cddfc..50f47681 100644 --- a/pygments/lexers/scripting.py +++ b/pygments/lexers/scripting.py @@ -157,7 +157,7 @@ class LuaLexer(RegexLexer): elif '.' in value: a, b = value.split('.') yield index, Name, a - yield index + len(a), Punctuation, u'.' + yield index + len(a), Punctuation, '.' yield index + len(a) + 1, Name, b continue yield index, token, value @@ -659,18 +659,18 @@ class AppleScriptLexer(RegexLexer): tokens = { 'root': [ (r'\s+', Text), - (u'¬\\n', String.Escape), + (r'¬\n', String.Escape), (r"'s\s+", Text), # This is a possessive, consider moving (r'(--|#).*?$', Comment), (r'\(\*', Comment.Multiline, 'comment'), (r'[(){}!,.:]', Punctuation), - (u'(«)([^»]+)(»)', + (r'(«)([^»]+)(»)', bygroups(Text, Name.Builtin, Text)), (r'\b((?:considering|ignoring)\s*)' r'(application responses|case|diacriticals|hyphens|' r'numeric strings|punctuation|white space)', bygroups(Keyword, Name.Builtin)), - (u'(-|\\*|\\+|&|≠|>=?|<=?|=|≥|≤|/|÷|\\^)', Operator), + (r'(-|\*|\+|&|≠|>=?|<=?|=|≥|≤|/|÷|\^)', Operator), (r"\b(%s)\b" % '|'.join(Operators), Operator.Word), (r'^(\s*(?:on|end)\s+)' r'(%s)' % '|'.join(StudioEvents[::-1]), @@ -976,7 +976,7 @@ class EasytrieveLexer(RegexLexer): _DELIMITER_PATTERN = '[' + _DELIMITERS + ']' _DELIMITER_PATTERN_CAPTURE = '(' + _DELIMITER_PATTERN + ')' _NON_DELIMITER_OR_COMMENT_PATTERN = '[^' + _DELIMITERS_OR_COMENT + ']' - _OPERATORS_PATTERN = u'[.+\\-/=\\[\\](){}<>;,&%¬]' + _OPERATORS_PATTERN = '[.+\\-/=\\[\\](){}<>;,&%¬]' _KEYWORDS = [ 'AFTER-BREAK', 'AFTER-LINE', 'AFTER-SCREEN', 'AIM', 'AND', 'ATTR', 'BEFORE', 'BEFORE-BREAK', 'BEFORE-LINE', 'BEFORE-SCREEN', 'BUSHU', diff --git a/pygments/lexers/slash.py b/pygments/lexers/slash.py index 8c9d53d2..84e1d2e6 100644 --- a/pygments/lexers/slash.py +++ b/pygments/lexers/slash.py @@ -26,7 +26,7 @@ class SlashLanguageLexer(ExtendedRegexLexer): def right_angle_bracket(lexer, match, ctx): if len(ctx.stack) > 1 and ctx.stack[-2] == "string": ctx.stack.pop() - yield match.start(), String.Interpol, u"}" + yield match.start(), String.Interpol, '}' ctx.pos = match.end() pass diff --git a/pygments/lexers/testing.py b/pygments/lexers/testing.py index 914bc1ec..919cc774 100644 --- a/pygments/lexers/testing.py +++ b/pygments/lexers/testing.py @@ -26,10 +26,10 @@ class GherkinLexer(RegexLexer): filenames = ['*.feature'] mimetypes = ['text/x-gherkin'] - feature_keywords = u'^(기능|機能|功能|フィーチャ|خاصية|תכונה|Функціонал|Функционалност|Функционал|Фича|Особина|Могућност|Özellik|Właściwość|Tính năng|Trajto|Savybė|Požiadavka|Požadavek|Osobina|Ominaisuus|Omadus|OH HAI|Mogućnost|Mogucnost|Jellemző|Fīča|Funzionalità|Funktionalität|Funkcionalnost|Funkcionalitāte|Funcționalitate|Functionaliteit|Functionalitate|Funcionalitat|Funcionalidade|Fonctionnalité|Fitur|Feature|Egenskap|Egenskab|Crikey|Característica|Arwedd)(:)(.*)$' - feature_element_keywords = u'^(\\s*)(시나리오 개요|시나리오|배경|背景|場景大綱|場景|场景大纲|场景|劇本大綱|劇本|剧本大纲|剧本|テンプレ|シナリオテンプレート|シナリオテンプレ|シナリオアウトライン|シナリオ|سيناريو مخطط|سيناريو|الخلفية|תרחיש|תבנית תרחיש|רקע|Тарих|Сценарій|Сценарио|Сценарий структураси|Сценарий|Структура сценарію|Структура сценарија|Структура сценария|Скица|Рамка на сценарий|Пример|Предыстория|Предистория|Позадина|Передумова|Основа|Концепт|Контекст|Założenia|Wharrimean is|Tình huống|The thing of it is|Tausta|Taust|Tapausaihio|Tapaus|Szenariogrundriss|Szenario|Szablon scenariusza|Stsenaarium|Struktura scenarija|Skica|Skenario konsep|Skenario|Situācija|Senaryo taslağı|Senaryo|Scénář|Scénario|Schema dello scenario|Scenārijs pēc parauga|Scenārijs|Scenár|Scenaro|Scenariusz|Scenariul de şablon|Scenariul de sablon|Scenariu|Scenario Outline|Scenario Amlinellol|Scenario|Scenarijus|Scenarijaus šablonas|Scenarij|Scenarie|Rerefons|Raamstsenaarium|Primer|Pozadí|Pozadina|Pozadie|Plan du scénario|Plan du Scénario|Osnova scénáře|Osnova|Náčrt Scénáře|Náčrt Scenáru|Mate|MISHUN SRSLY|MISHUN|Kịch bản|Konturo de la scenaro|Kontext|Konteksts|Kontekstas|Kontekst|Koncept|Khung tình huống|Khung kịch bản|Háttér|Grundlage|Geçmiş|Forgatókönyv vázlat|Forgatókönyv|Fono|Esquema do Cenário|Esquema do Cenario|Esquema del escenario|Esquema de l\'escenari|Escenario|Escenari|Dis is what went down|Dasar|Contexto|Contexte|Contesto|Condiţii|Conditii|Cenário|Cenario|Cefndir|Bối cảnh|Blokes|Bakgrunn|Bakgrund|Baggrund|Background|B4|Antecedents|Antecedentes|All y\'all|Achtergrond|Abstrakt Scenario|Abstract Scenario)(:)(.*)$' - examples_keywords = u'^(\\s*)(예|例子|例|サンプル|امثلة|דוגמאות|Сценарији|Примери|Приклади|Мисоллар|Значения|Örnekler|Voorbeelden|Variantai|Tapaukset|Scenarios|Scenariji|Scenarijai|Příklady|Példák|Príklady|Przykłady|Primjeri|Primeri|Piemēri|Pavyzdžiai|Paraugs|Juhtumid|Exemplos|Exemples|Exemplele|Exempel|Examples|Esempi|Enghreifftiau|Ekzemploj|Eksempler|Ejemplos|EXAMPLZ|Dữ liệu|Contoh|Cobber|Beispiele)(:)(.*)$' - step_keywords = u'^(\\s*)(하지만|조건|먼저|만일|만약|단|그리고|그러면|那麼|那么|而且|當|当|前提|假設|假设|假如|假定|但是|但し|並且|并且|同時|同时|もし|ならば|ただし|しかし|かつ|و |متى |لكن |عندما |ثم |بفرض |اذاً |כאשר |וגם |בהינתן |אזי |אז |אבל |Якщо |Унда |То |Припустимо, що |Припустимо |Онда |Но |Нехай |Лекин |Когато |Када |Кад |К тому же |И |Задато |Задати |Задате |Если |Допустим |Дадено |Ва |Бирок |Аммо |Али |Але |Агар |А |І |Și |És |Zatati |Zakładając |Zadato |Zadate |Zadano |Zadani |Zadan |Youse know when youse got |Youse know like when |Yna |Ya know how |Ya gotta |Y |Wun |Wtedy |When y\'all |When |Wenn |WEN |Và |Ve |Und |Un |Thì |Then y\'all |Then |Tapi |Tak |Tada |Tad |Så |Stel |Soit |Siis |Si |Sed |Se |Quando |Quand |Quan |Pryd |Pokud |Pokiaľ |Però |Pero |Pak |Oraz |Onda |Ond |Oletetaan |Og |Och |O zaman |Når |När |Niin |Nhưng |N |Mutta |Men |Mas |Maka |Majd |Mais |Maar |Ma |Lorsque |Lorsqu\'|Kun |Kuid |Kui |Khi |Keď |Ketika |Když |Kaj |Kai |Kada |Kad |Jeżeli |Ja |Ir |I CAN HAZ |I |Ha |Givun |Givet |Given y\'all |Given |Gitt |Gegeven |Gegeben sei |Fakat |Eğer ki |Etant donné |Et |Então |Entonces |Entao |En |Eeldades |E |Duota |Dun |Donitaĵo |Donat |Donada |Do |Diyelim ki |Dengan |Den youse gotta |De |Dato |Dar |Dann |Dan |Dado |Dacă |Daca |DEN |Când |Cuando |Cho |Cept |Cand |Cal |But y\'all |But |Buh |Biết |Bet |BUT |Atès |Atunci |Atesa |Anrhegedig a |Angenommen |And y\'all |And |An |Ama |Als |Alors |Allora |Ali |Aleshores |Ale |Akkor |Aber |AN |A také |A |\\* )' + feature_keywords = '^(기능|機能|功能|フィーチャ|خاصية|תכונה|Функціонал|Функционалност|Функционал|Фича|Особина|Могућност|Özellik|Właściwość|Tính năng|Trajto|Savybė|Požiadavka|Požadavek|Osobina|Ominaisuus|Omadus|OH HAI|Mogućnost|Mogucnost|Jellemző|Fīča|Funzionalità|Funktionalität|Funkcionalnost|Funkcionalitāte|Funcționalitate|Functionaliteit|Functionalitate|Funcionalitat|Funcionalidade|Fonctionnalité|Fitur|Feature|Egenskap|Egenskab|Crikey|Característica|Arwedd)(:)(.*)$' + feature_element_keywords = '^(\\s*)(시나리오 개요|시나리오|배경|背景|場景大綱|場景|场景大纲|场景|劇本大綱|劇本|剧本大纲|剧本|テンプレ|シナリオテンプレート|シナリオテンプレ|シナリオアウトライン|シナリオ|سيناريو مخطط|سيناريو|الخلفية|תרחיש|תבנית תרחיש|רקע|Тарих|Сценарій|Сценарио|Сценарий структураси|Сценарий|Структура сценарію|Структура сценарија|Структура сценария|Скица|Рамка на сценарий|Пример|Предыстория|Предистория|Позадина|Передумова|Основа|Концепт|Контекст|Założenia|Wharrimean is|Tình huống|The thing of it is|Tausta|Taust|Tapausaihio|Tapaus|Szenariogrundriss|Szenario|Szablon scenariusza|Stsenaarium|Struktura scenarija|Skica|Skenario konsep|Skenario|Situācija|Senaryo taslağı|Senaryo|Scénář|Scénario|Schema dello scenario|Scenārijs pēc parauga|Scenārijs|Scenár|Scenaro|Scenariusz|Scenariul de şablon|Scenariul de sablon|Scenariu|Scenario Outline|Scenario Amlinellol|Scenario|Scenarijus|Scenarijaus šablonas|Scenarij|Scenarie|Rerefons|Raamstsenaarium|Primer|Pozadí|Pozadina|Pozadie|Plan du scénario|Plan du Scénario|Osnova scénáře|Osnova|Náčrt Scénáře|Náčrt Scenáru|Mate|MISHUN SRSLY|MISHUN|Kịch bản|Konturo de la scenaro|Kontext|Konteksts|Kontekstas|Kontekst|Koncept|Khung tình huống|Khung kịch bản|Háttér|Grundlage|Geçmiş|Forgatókönyv vázlat|Forgatókönyv|Fono|Esquema do Cenário|Esquema do Cenario|Esquema del escenario|Esquema de l\'escenari|Escenario|Escenari|Dis is what went down|Dasar|Contexto|Contexte|Contesto|Condiţii|Conditii|Cenário|Cenario|Cefndir|Bối cảnh|Blokes|Bakgrunn|Bakgrund|Baggrund|Background|B4|Antecedents|Antecedentes|All y\'all|Achtergrond|Abstrakt Scenario|Abstract Scenario)(:)(.*)$' + examples_keywords = '^(\\s*)(예|例子|例|サンプル|امثلة|דוגמאות|Сценарији|Примери|Приклади|Мисоллар|Значения|Örnekler|Voorbeelden|Variantai|Tapaukset|Scenarios|Scenariji|Scenarijai|Příklady|Példák|Príklady|Przykłady|Primjeri|Primeri|Piemēri|Pavyzdžiai|Paraugs|Juhtumid|Exemplos|Exemples|Exemplele|Exempel|Examples|Esempi|Enghreifftiau|Ekzemploj|Eksempler|Ejemplos|EXAMPLZ|Dữ liệu|Contoh|Cobber|Beispiele)(:)(.*)$' + step_keywords = '^(\\s*)(하지만|조건|먼저|만일|만약|단|그리고|그러면|那麼|那么|而且|當|当|前提|假設|假设|假如|假定|但是|但し|並且|并且|同時|同时|もし|ならば|ただし|しかし|かつ|و |متى |لكن |عندما |ثم |بفرض |اذاً |כאשר |וגם |בהינתן |אזי |אז |אבל |Якщо |Унда |То |Припустимо, що |Припустимо |Онда |Но |Нехай |Лекин |Когато |Када |Кад |К тому же |И |Задато |Задати |Задате |Если |Допустим |Дадено |Ва |Бирок |Аммо |Али |Але |Агар |А |І |Și |És |Zatati |Zakładając |Zadato |Zadate |Zadano |Zadani |Zadan |Youse know when youse got |Youse know like when |Yna |Ya know how |Ya gotta |Y |Wun |Wtedy |When y\'all |When |Wenn |WEN |Và |Ve |Und |Un |Thì |Then y\'all |Then |Tapi |Tak |Tada |Tad |Så |Stel |Soit |Siis |Si |Sed |Se |Quando |Quand |Quan |Pryd |Pokud |Pokiaľ |Però |Pero |Pak |Oraz |Onda |Ond |Oletetaan |Og |Och |O zaman |Når |När |Niin |Nhưng |N |Mutta |Men |Mas |Maka |Majd |Mais |Maar |Ma |Lorsque |Lorsqu\'|Kun |Kuid |Kui |Khi |Keď |Ketika |Když |Kaj |Kai |Kada |Kad |Jeżeli |Ja |Ir |I CAN HAZ |I |Ha |Givun |Givet |Given y\'all |Given |Gitt |Gegeven |Gegeben sei |Fakat |Eğer ki |Etant donné |Et |Então |Entonces |Entao |En |Eeldades |E |Duota |Dun |Donitaĵo |Donat |Donada |Do |Diyelim ki |Dengan |Den youse gotta |De |Dato |Dar |Dann |Dan |Dado |Dacă |Daca |DEN |Când |Cuando |Cho |Cept |Cand |Cal |But y\'all |But |Buh |Biết |Bet |BUT |Atès |Atunci |Atesa |Anrhegedig a |Angenommen |And y\'all |And |An |Ama |Als |Alors |Allora |Ali |Aleshores |Ale |Akkor |Aber |AN |A také |A |\\* )' tokens = { 'comments': [ diff --git a/pygments/lexers/theorem.py b/pygments/lexers/theorem.py index fd34de2c..c4c857d4 100644 --- a/pygments/lexers/theorem.py +++ b/pygments/lexers/theorem.py @@ -93,7 +93,7 @@ class CoqLexer(RegexLexer): '<->', '=', '>', '>]', r'>\}', r'\?', r'\?\?', r'\[', r'\[<', r'\[>', r'\[\|', ']', '_', '`', r'\{', r'\{<', r'\|', r'\|]', r'\}', '~', '=>', r'/\\', r'\\/', r'\{\|', r'\|\}', - u'Π', u'λ', + 'Π', 'λ', ) operators = r'[!$%&*+\./:<=>?@^|~-]' prefix_syms = r'[!?~]' @@ -440,11 +440,11 @@ class LeanLexer(RegexLexer): '#print', '#help', ), suffix=r'\b'), Keyword), (words(( - '(', ')', ':', '{', '}', '[', ']', u'⟨', u'⟩', u'‹', u'›', u'⦃', u'⦄', ':=', ',', + '(', ')', ':', '{', '}', '[', ']', '⟨', '⟩', '‹', '›', '⦃', '⦄', ':=', ',', )), Operator), - (u"[A-Za-z_\u03b1-\u03ba\u03bc-\u03fb\u1f00-\u1ffe\u2100-\u214f]" - u"[.A-Za-z_'\u03b1-\u03ba\u03bc-\u03fb\u1f00-\u1ffe\u2070-\u2079" - u"\u207f-\u2089\u2090-\u209c\u2100-\u214f0-9]*", Name), + (r'[A-Za-z_\u03b1-\u03ba\u03bc-\u03fb\u1f00-\u1ffe\u2100-\u214f]' + r'[.A-Za-z_\'\u03b1-\u03ba\u03bc-\u03fb\u1f00-\u1ffe\u2070-\u2079' + r'\u207f-\u2089\u2090-\u209c\u2100-\u214f0-9]*', Name), (r'0x[A-Za-z0-9]+', Number.Integer), (r'0b[01]+', Number.Integer), (r'\d+', Number.Integer), diff --git a/pygments/lexers/webmisc.py b/pygments/lexers/webmisc.py index 84ba490c..dab36aa6 100644 --- a/pygments/lexers/webmisc.py +++ b/pygments/lexers/webmisc.py @@ -15,7 +15,6 @@ from pygments.lexer import RegexLexer, ExtendedRegexLexer, include, bygroups, \ default, using from pygments.token import Text, Comment, Operator, Keyword, Name, String, \ Number, Punctuation, Literal -from pygments.util import unirange from pygments.lexers.css import _indentation, _starts_block from pygments.lexers.html import HtmlLexer @@ -74,15 +73,15 @@ class XQueryLexer(ExtendedRegexLexer): # FIX UNICODE LATER # ncnamestartchar = ( - # ur"[A-Z]|_|[a-z]|[\u00C0-\u00D6]|[\u00D8-\u00F6]|[\u00F8-\u02FF]|" - # ur"[\u0370-\u037D]|[\u037F-\u1FFF]|[\u200C-\u200D]|[\u2070-\u218F]|" - # ur"[\u2C00-\u2FEF]|[\u3001-\uD7FF]|[\uF900-\uFDCF]|[\uFDF0-\uFFFD]|" - # ur"[\u10000-\uEFFFF]" + # r"[A-Z]|_|[a-z]|[\u00C0-\u00D6]|[\u00D8-\u00F6]|[\u00F8-\u02FF]|" + # r"[\u0370-\u037D]|[\u037F-\u1FFF]|[\u200C-\u200D]|[\u2070-\u218F]|" + # r"[\u2C00-\u2FEF]|[\u3001-\uD7FF]|[\uF900-\uFDCF]|[\uFDF0-\uFFFD]|" + # r"[\u10000-\uEFFFF]" # ) ncnamestartchar = r"(?:[A-Z]|_|[a-z])" # FIX UNICODE LATER - # ncnamechar = ncnamestartchar + (ur"|-|\.|[0-9]|\u00B7|[\u0300-\u036F]|" - # ur"[\u203F-\u2040]") + # ncnamechar = ncnamestartchar + (r"|-|\.|[0-9]|\u00B7|[\u0300-\u036F]|" + # r"[\u203F-\u2040]") ncnamechar = r"(?:" + ncnamestartchar + r"|-|\.|[0-9])" ncname = "(?:%s+%s*)" % (ncnamestartchar, ncnamechar) pitarget_namestartchar = r"(?:[A-KN-WYZ]|_|:|[a-kn-wyz])" @@ -99,14 +98,14 @@ class XQueryLexer(ExtendedRegexLexer): stringsingle = r"(?:'(?:" + entityref + r"|" + charref + r"|''|[^&'])*')" # FIX UNICODE LATER - # elementcontentchar = (ur'\t|\r|\n|[\u0020-\u0025]|[\u0028-\u003b]|' - # ur'[\u003d-\u007a]|\u007c|[\u007e-\u007F]') + # elementcontentchar = (r'\t|\r|\n|[\u0020-\u0025]|[\u0028-\u003b]|' + # r'[\u003d-\u007a]|\u007c|[\u007e-\u007F]') elementcontentchar = r'[A-Za-z]|\s|\d|[!"#$%()*+,\-./:;=?@\[\\\]^_\'`|~]' - # quotattrcontentchar = (ur'\t|\r|\n|[\u0020-\u0021]|[\u0023-\u0025]|' - # ur'[\u0027-\u003b]|[\u003d-\u007a]|\u007c|[\u007e-\u007F]') + # quotattrcontentchar = (r'\t|\r|\n|[\u0020-\u0021]|[\u0023-\u0025]|' + # r'[\u0027-\u003b]|[\u003d-\u007a]|\u007c|[\u007e-\u007F]') quotattrcontentchar = r'[A-Za-z]|\s|\d|[!#$%()*+,\-./:;=?@\[\\\]^_\'`|~]' - # aposattrcontentchar = (ur'\t|\r|\n|[\u0020-\u0025]|[\u0028-\u003b]|' - # ur'[\u003d-\u007a]|\u007c|[\u007e-\u007F]') + # aposattrcontentchar = (r'\t|\r|\n|[\u0020-\u0025]|[\u0028-\u003b]|' + # r'[\u003d-\u007a]|\u007c|[\u007e-\u007F]') aposattrcontentchar = r'[A-Za-z]|\s|\d|[!"#$%()*+,\-./:;=?@\[\\\]^_`|~]' # CHAR elements - fix the above elementcontentchar, quotattrcontentchar, @@ -518,8 +517,8 @@ class XQueryLexer(ExtendedRegexLexer): 'xml_comment': [ (r'(-->)', popstate_xmlcomment_callback), (r'[^-]{1,2}', Literal), - (u'\\t|\\r|\\n|[\u0020-\uD7FF]|[\uE000-\uFFFD]|' + - unirange(0x10000, 0x10ffff), Literal), + (r'\t|\r|\n|[\u0020-\uD7FF]|[\uE000-\uFFFD]|[\U00010000-\U0010FFFF]', + Literal), ], 'processing_instruction': [ (r'\s+', Text, 'processing_instruction_content'), @@ -528,13 +527,13 @@ class XQueryLexer(ExtendedRegexLexer): ], 'processing_instruction_content': [ (r'\?>', String.Doc, '#pop'), - (u'\\t|\\r|\\n|[\u0020-\uD7FF]|[\uE000-\uFFFD]|' + - unirange(0x10000, 0x10ffff), Literal), + (r'\t|\r|\n|[\u0020-\uD7FF]|[\uE000-\uFFFD]|[\U00010000-\U0010FFFF]', + Literal), ], 'cdata_section': [ (r']]>', String.Doc, '#pop'), - (u'\\t|\\r|\\n|[\u0020-\uD7FF]|[\uE000-\uFFFD]|' + - unirange(0x10000, 0x10ffff), Literal), + (r'\t|\r|\n|[\u0020-\uD7FF]|[\uE000-\uFFFD]|[\U00010000-\U0010FFFF]', + Literal), ], 'start_tag': [ include('whitespace'), @@ -603,8 +602,8 @@ class XQueryLexer(ExtendedRegexLexer): ], 'pragmacontents': [ (r'#\)', Punctuation, 'operator'), - (u'\\t|\\r|\\n|[\u0020-\uD7FF]|[\uE000-\uFFFD]|' + - unirange(0x10000, 0x10ffff), Literal), + (r'\t|\r|\n|[\u0020-\uD7FF]|[\uE000-\uFFFD]|[\U00010000-\U0010FFFF]', + Literal), (r'(\s+)', Text), ], 'occurrenceindicator': [ diff --git a/pygments/lexers/whiley.py b/pygments/lexers/whiley.py index 255e9d22..fad94e90 100644 --- a/pygments/lexers/whiley.py +++ b/pygments/lexers/whiley.py @@ -97,11 +97,11 @@ class WhileyLexer(RegexLexer): # operators and punctuation (r'[{}()\[\],.;]', Punctuation), - (u'[+\\-*/%&|<>^!~@=:?' + (r'[+\-*/%&|<>^!~@=:?' # unicode operators - u'\u2200\u2203\u2205\u2282\u2286\u2283\u2287' - u'\u222A\u2229\u2264\u2265\u2208\u2227\u2228' - u']', Operator), + r'\u2200\u2203\u2205\u2282\u2286\u2283\u2287' + r'\u222A\u2229\u2264\u2265\u2208\u2227\u2228' + r']', Operator), # identifier (r'[a-zA-Z_]\w*', Name), diff --git a/pygments/styles/arduino.py b/pygments/styles/arduino.py index c734642f..3630ffc2 100644 --- a/pygments/styles/arduino.py +++ b/pygments/styles/arduino.py @@ -15,7 +15,7 @@ from pygments.token import Keyword, Name, Comment, String, Error, \ class ArduinoStyle(Style): - u""" + """ The Arduino® language style. This style is designed to highlight the Arduino source code, so exepect the best results with it. """ diff --git a/pygments/unistring.py b/pygments/unistring.py index 027df53e..6292ad27 100644 --- a/pygments/unistring.py +++ b/pygments/unistring.py @@ -14,126 +14,83 @@ import sys -Cc = u'\x00-\x1f\x7f-\x9f' +Cc = '\x00-\x1f\x7f-\x9f' -Cf = u'\xad\u0600-\u0605\u061c\u06dd\u070f\u08e2\u180e\u200b-\u200f\u202a-\u202e\u2060-\u2064\u2066-\u206f\ufeff\ufff9-\ufffb' +Cf = '\xad\u0600-\u0605\u061c\u06dd\u070f\u08e2\u180e\u200b-\u200f\u202a-\u202e\u2060-\u2064\u2066-\u206f\ufeff\ufff9-\ufffb\U000110bd\U000110cd\U0001bca0-\U0001bca3\U0001d173-\U0001d17a\U000e0001\U000e0020-\U000e007f' -Cn = u'\u0378-\u0379\u0380-\u0383\u038b\u038d\u03a2\u0530\u0557-\u0558\u058b-\u058c\u0590\u05c8-\u05cf\u05eb-\u05ee\u05f5-\u05ff\u061d\u070e\u074b-\u074c\u07b2-\u07bf\u07fb-\u07fc\u082e-\u082f\u083f\u085c-\u085d\u085f\u086b-\u089f\u08b5\u08be-\u08d2\u0984\u098d-\u098e\u0991-\u0992\u09a9\u09b1\u09b3-\u09b5\u09ba-\u09bb\u09c5-\u09c6\u09c9-\u09ca\u09cf-\u09d6\u09d8-\u09db\u09de\u09e4-\u09e5\u09ff-\u0a00\u0a04\u0a0b-\u0a0e\u0a11-\u0a12\u0a29\u0a31\u0a34\u0a37\u0a3a-\u0a3b\u0a3d\u0a43-\u0a46\u0a49-\u0a4a\u0a4e-\u0a50\u0a52-\u0a58\u0a5d\u0a5f-\u0a65\u0a77-\u0a80\u0a84\u0a8e\u0a92\u0aa9\u0ab1\u0ab4\u0aba-\u0abb\u0ac6\u0aca\u0ace-\u0acf\u0ad1-\u0adf\u0ae4-\u0ae5\u0af2-\u0af8\u0b00\u0b04\u0b0d-\u0b0e\u0b11-\u0b12\u0b29\u0b31\u0b34\u0b3a-\u0b3b\u0b45-\u0b46\u0b49-\u0b4a\u0b4e-\u0b55\u0b58-\u0b5b\u0b5e\u0b64-\u0b65\u0b78-\u0b81\u0b84\u0b8b-\u0b8d\u0b91\u0b96-\u0b98\u0b9b\u0b9d\u0ba0-\u0ba2\u0ba5-\u0ba7\u0bab-\u0bad\u0bba-\u0bbd\u0bc3-\u0bc5\u0bc9\u0bce-\u0bcf\u0bd1-\u0bd6\u0bd8-\u0be5\u0bfb-\u0bff\u0c0d\u0c11\u0c29\u0c3a-\u0c3c\u0c45\u0c49\u0c4e-\u0c54\u0c57\u0c5b-\u0c5f\u0c64-\u0c65\u0c70-\u0c77\u0c8d\u0c91\u0ca9\u0cb4\u0cba-\u0cbb\u0cc5\u0cc9\u0cce-\u0cd4\u0cd7-\u0cdd\u0cdf\u0ce4-\u0ce5\u0cf0\u0cf3-\u0cff\u0d04\u0d0d\u0d11\u0d45\u0d49\u0d50-\u0d53\u0d64-\u0d65\u0d80-\u0d81\u0d84\u0d97-\u0d99\u0db2\u0dbc\u0dbe-\u0dbf\u0dc7-\u0dc9\u0dcb-\u0dce\u0dd5\u0dd7\u0de0-\u0de5\u0df0-\u0df1\u0df5-\u0e00\u0e3b-\u0e3e\u0e5c-\u0e80\u0e83\u0e85-\u0e86\u0e89\u0e8b-\u0e8c\u0e8e-\u0e93\u0e98\u0ea0\u0ea4\u0ea6\u0ea8-\u0ea9\u0eac\u0eba\u0ebe-\u0ebf\u0ec5\u0ec7\u0ece-\u0ecf\u0eda-\u0edb\u0ee0-\u0eff\u0f48\u0f6d-\u0f70\u0f98\u0fbd\u0fcd\u0fdb-\u0fff\u10c6\u10c8-\u10cc\u10ce-\u10cf\u1249\u124e-\u124f\u1257\u1259\u125e-\u125f\u1289\u128e-\u128f\u12b1\u12b6-\u12b7\u12bf\u12c1\u12c6-\u12c7\u12d7\u1311\u1316-\u1317\u135b-\u135c\u137d-\u137f\u139a-\u139f\u13f6-\u13f7\u13fe-\u13ff\u169d-\u169f\u16f9-\u16ff\u170d\u1715-\u171f\u1737-\u173f\u1754-\u175f\u176d\u1771\u1774-\u177f\u17de-\u17df\u17ea-\u17ef\u17fa-\u17ff\u180f\u181a-\u181f\u1879-\u187f\u18ab-\u18af\u18f6-\u18ff\u191f\u192c-\u192f\u193c-\u193f\u1941-\u1943\u196e-\u196f\u1975-\u197f\u19ac-\u19af\u19ca-\u19cf\u19db-\u19dd\u1a1c-\u1a1d\u1a5f\u1a7d-\u1a7e\u1a8a-\u1a8f\u1a9a-\u1a9f\u1aae-\u1aaf\u1abf-\u1aff\u1b4c-\u1b4f\u1b7d-\u1b7f\u1bf4-\u1bfb\u1c38-\u1c3a\u1c4a-\u1c4c\u1c89-\u1c8f\u1cbb-\u1cbc\u1cc8-\u1ccf\u1cfa-\u1cff\u1dfa\u1f16-\u1f17\u1f1e-\u1f1f\u1f46-\u1f47\u1f4e-\u1f4f\u1f58\u1f5a\u1f5c\u1f5e\u1f7e-\u1f7f\u1fb5\u1fc5\u1fd4-\u1fd5\u1fdc\u1ff0-\u1ff1\u1ff5\u1fff\u2065\u2072-\u2073\u208f\u209d-\u209f\u20c0-\u20cf\u20f1-\u20ff\u218c-\u218f\u2427-\u243f\u244b-\u245f\u2b74-\u2b75\u2b96-\u2b97\u2bc9\u2bff\u2c2f\u2c5f\u2cf4-\u2cf8\u2d26\u2d28-\u2d2c\u2d2e-\u2d2f\u2d68-\u2d6e\u2d71-\u2d7e\u2d97-\u2d9f\u2da7\u2daf\u2db7\u2dbf\u2dc7\u2dcf\u2dd7\u2ddf\u2e4f-\u2e7f\u2e9a\u2ef4-\u2eff\u2fd6-\u2fef\u2ffc-\u2fff\u3040\u3097-\u3098\u3100-\u3104\u3130\u318f\u31bb-\u31bf\u31e4-\u31ef\u321f\u32ff\u4db6-\u4dbf\u9ff0-\u9fff\ua48d-\ua48f\ua4c7-\ua4cf\ua62c-\ua63f\ua6f8-\ua6ff\ua7ba-\ua7f6\ua82c-\ua82f\ua83a-\ua83f\ua878-\ua87f\ua8c6-\ua8cd\ua8da-\ua8df\ua954-\ua95e\ua97d-\ua97f\ua9ce\ua9da-\ua9dd\ua9ff\uaa37-\uaa3f\uaa4e-\uaa4f\uaa5a-\uaa5b\uaac3-\uaada\uaaf7-\uab00\uab07-\uab08\uab0f-\uab10\uab17-\uab1f\uab27\uab2f\uab66-\uab6f\uabee-\uabef\uabfa-\uabff\ud7a4-\ud7af\ud7c7-\ud7ca\ud7fc-\ud7ff\ufa6e-\ufa6f\ufada-\ufaff\ufb07-\ufb12\ufb18-\ufb1c\ufb37\ufb3d\ufb3f\ufb42\ufb45\ufbc2-\ufbd2\ufd40-\ufd4f\ufd90-\ufd91\ufdc8-\ufdef\ufdfe-\ufdff\ufe1a-\ufe1f\ufe53\ufe67\ufe6c-\ufe6f\ufe75\ufefd-\ufefe\uff00\uffbf-\uffc1\uffc8-\uffc9\uffd0-\uffd1\uffd8-\uffd9\uffdd-\uffdf\uffe7\uffef-\ufff8\ufffe-\uffff' +Cn = '\u0378-\u0379\u0380-\u0383\u038b\u038d\u03a2\u0530\u0557-\u0558\u058b-\u058c\u0590\u05c8-\u05cf\u05eb-\u05ee\u05f5-\u05ff\u061d\u070e\u074b-\u074c\u07b2-\u07bf\u07fb-\u07fc\u082e-\u082f\u083f\u085c-\u085d\u085f\u086b-\u089f\u08b5\u08be-\u08d2\u0984\u098d-\u098e\u0991-\u0992\u09a9\u09b1\u09b3-\u09b5\u09ba-\u09bb\u09c5-\u09c6\u09c9-\u09ca\u09cf-\u09d6\u09d8-\u09db\u09de\u09e4-\u09e5\u09ff-\u0a00\u0a04\u0a0b-\u0a0e\u0a11-\u0a12\u0a29\u0a31\u0a34\u0a37\u0a3a-\u0a3b\u0a3d\u0a43-\u0a46\u0a49-\u0a4a\u0a4e-\u0a50\u0a52-\u0a58\u0a5d\u0a5f-\u0a65\u0a77-\u0a80\u0a84\u0a8e\u0a92\u0aa9\u0ab1\u0ab4\u0aba-\u0abb\u0ac6\u0aca\u0ace-\u0acf\u0ad1-\u0adf\u0ae4-\u0ae5\u0af2-\u0af8\u0b00\u0b04\u0b0d-\u0b0e\u0b11-\u0b12\u0b29\u0b31\u0b34\u0b3a-\u0b3b\u0b45-\u0b46\u0b49-\u0b4a\u0b4e-\u0b55\u0b58-\u0b5b\u0b5e\u0b64-\u0b65\u0b78-\u0b81\u0b84\u0b8b-\u0b8d\u0b91\u0b96-\u0b98\u0b9b\u0b9d\u0ba0-\u0ba2\u0ba5-\u0ba7\u0bab-\u0bad\u0bba-\u0bbd\u0bc3-\u0bc5\u0bc9\u0bce-\u0bcf\u0bd1-\u0bd6\u0bd8-\u0be5\u0bfb-\u0bff\u0c0d\u0c11\u0c29\u0c3a-\u0c3c\u0c45\u0c49\u0c4e-\u0c54\u0c57\u0c5b-\u0c5f\u0c64-\u0c65\u0c70-\u0c77\u0c8d\u0c91\u0ca9\u0cb4\u0cba-\u0cbb\u0cc5\u0cc9\u0cce-\u0cd4\u0cd7-\u0cdd\u0cdf\u0ce4-\u0ce5\u0cf0\u0cf3-\u0cff\u0d04\u0d0d\u0d11\u0d45\u0d49\u0d50-\u0d53\u0d64-\u0d65\u0d80-\u0d81\u0d84\u0d97-\u0d99\u0db2\u0dbc\u0dbe-\u0dbf\u0dc7-\u0dc9\u0dcb-\u0dce\u0dd5\u0dd7\u0de0-\u0de5\u0df0-\u0df1\u0df5-\u0e00\u0e3b-\u0e3e\u0e5c-\u0e80\u0e83\u0e85-\u0e86\u0e89\u0e8b-\u0e8c\u0e8e-\u0e93\u0e98\u0ea0\u0ea4\u0ea6\u0ea8-\u0ea9\u0eac\u0eba\u0ebe-\u0ebf\u0ec5\u0ec7\u0ece-\u0ecf\u0eda-\u0edb\u0ee0-\u0eff\u0f48\u0f6d-\u0f70\u0f98\u0fbd\u0fcd\u0fdb-\u0fff\u10c6\u10c8-\u10cc\u10ce-\u10cf\u1249\u124e-\u124f\u1257\u1259\u125e-\u125f\u1289\u128e-\u128f\u12b1\u12b6-\u12b7\u12bf\u12c1\u12c6-\u12c7\u12d7\u1311\u1316-\u1317\u135b-\u135c\u137d-\u137f\u139a-\u139f\u13f6-\u13f7\u13fe-\u13ff\u169d-\u169f\u16f9-\u16ff\u170d\u1715-\u171f\u1737-\u173f\u1754-\u175f\u176d\u1771\u1774-\u177f\u17de-\u17df\u17ea-\u17ef\u17fa-\u17ff\u180f\u181a-\u181f\u1879-\u187f\u18ab-\u18af\u18f6-\u18ff\u191f\u192c-\u192f\u193c-\u193f\u1941-\u1943\u196e-\u196f\u1975-\u197f\u19ac-\u19af\u19ca-\u19cf\u19db-\u19dd\u1a1c-\u1a1d\u1a5f\u1a7d-\u1a7e\u1a8a-\u1a8f\u1a9a-\u1a9f\u1aae-\u1aaf\u1abf-\u1aff\u1b4c-\u1b4f\u1b7d-\u1b7f\u1bf4-\u1bfb\u1c38-\u1c3a\u1c4a-\u1c4c\u1c89-\u1c8f\u1cbb-\u1cbc\u1cc8-\u1ccf\u1cfa-\u1cff\u1dfa\u1f16-\u1f17\u1f1e-\u1f1f\u1f46-\u1f47\u1f4e-\u1f4f\u1f58\u1f5a\u1f5c\u1f5e\u1f7e-\u1f7f\u1fb5\u1fc5\u1fd4-\u1fd5\u1fdc\u1ff0-\u1ff1\u1ff5\u1fff\u2065\u2072-\u2073\u208f\u209d-\u209f\u20c0-\u20cf\u20f1-\u20ff\u218c-\u218f\u2427-\u243f\u244b-\u245f\u2b74-\u2b75\u2b96-\u2b97\u2bc9\u2bff\u2c2f\u2c5f\u2cf4-\u2cf8\u2d26\u2d28-\u2d2c\u2d2e-\u2d2f\u2d68-\u2d6e\u2d71-\u2d7e\u2d97-\u2d9f\u2da7\u2daf\u2db7\u2dbf\u2dc7\u2dcf\u2dd7\u2ddf\u2e4f-\u2e7f\u2e9a\u2ef4-\u2eff\u2fd6-\u2fef\u2ffc-\u2fff\u3040\u3097-\u3098\u3100-\u3104\u3130\u318f\u31bb-\u31bf\u31e4-\u31ef\u321f\u32ff\u4db6-\u4dbf\u9ff0-\u9fff\ua48d-\ua48f\ua4c7-\ua4cf\ua62c-\ua63f\ua6f8-\ua6ff\ua7ba-\ua7f6\ua82c-\ua82f\ua83a-\ua83f\ua878-\ua87f\ua8c6-\ua8cd\ua8da-\ua8df\ua954-\ua95e\ua97d-\ua97f\ua9ce\ua9da-\ua9dd\ua9ff\uaa37-\uaa3f\uaa4e-\uaa4f\uaa5a-\uaa5b\uaac3-\uaada\uaaf7-\uab00\uab07-\uab08\uab0f-\uab10\uab17-\uab1f\uab27\uab2f\uab66-\uab6f\uabee-\uabef\uabfa-\uabff\ud7a4-\ud7af\ud7c7-\ud7ca\ud7fc-\ud7ff\ufa6e-\ufa6f\ufada-\ufaff\ufb07-\ufb12\ufb18-\ufb1c\ufb37\ufb3d\ufb3f\ufb42\ufb45\ufbc2-\ufbd2\ufd40-\ufd4f\ufd90-\ufd91\ufdc8-\ufdef\ufdfe-\ufdff\ufe1a-\ufe1f\ufe53\ufe67\ufe6c-\ufe6f\ufe75\ufefd-\ufefe\uff00\uffbf-\uffc1\uffc8-\uffc9\uffd0-\uffd1\uffd8-\uffd9\uffdd-\uffdf\uffe7\uffef-\ufff8\ufffe-\uffff\U0001000c\U00010027\U0001003b\U0001003e\U0001004e-\U0001004f\U0001005e-\U0001007f\U000100fb-\U000100ff\U00010103-\U00010106\U00010134-\U00010136\U0001018f\U0001019c-\U0001019f\U000101a1-\U000101cf\U000101fe-\U0001027f\U0001029d-\U0001029f\U000102d1-\U000102df\U000102fc-\U000102ff\U00010324-\U0001032c\U0001034b-\U0001034f\U0001037b-\U0001037f\U0001039e\U000103c4-\U000103c7\U000103d6-\U000103ff\U0001049e-\U0001049f\U000104aa-\U000104af\U000104d4-\U000104d7\U000104fc-\U000104ff\U00010528-\U0001052f\U00010564-\U0001056e\U00010570-\U000105ff\U00010737-\U0001073f\U00010756-\U0001075f\U00010768-\U000107ff\U00010806-\U00010807\U00010809\U00010836\U00010839-\U0001083b\U0001083d-\U0001083e\U00010856\U0001089f-\U000108a6\U000108b0-\U000108df\U000108f3\U000108f6-\U000108fa\U0001091c-\U0001091e\U0001093a-\U0001093e\U00010940-\U0001097f\U000109b8-\U000109bb\U000109d0-\U000109d1\U00010a04\U00010a07-\U00010a0b\U00010a14\U00010a18\U00010a36-\U00010a37\U00010a3b-\U00010a3e\U00010a49-\U00010a4f\U00010a59-\U00010a5f\U00010aa0-\U00010abf\U00010ae7-\U00010aea\U00010af7-\U00010aff\U00010b36-\U00010b38\U00010b56-\U00010b57\U00010b73-\U00010b77\U00010b92-\U00010b98\U00010b9d-\U00010ba8\U00010bb0-\U00010bff\U00010c49-\U00010c7f\U00010cb3-\U00010cbf\U00010cf3-\U00010cf9\U00010d28-\U00010d2f\U00010d3a-\U00010e5f\U00010e7f-\U00010eff\U00010f28-\U00010f2f\U00010f5a-\U00010fff\U0001104e-\U00011051\U00011070-\U0001107e\U000110c2-\U000110cc\U000110ce-\U000110cf\U000110e9-\U000110ef\U000110fa-\U000110ff\U00011135\U00011147-\U0001114f\U00011177-\U0001117f\U000111ce-\U000111cf\U000111e0\U000111f5-\U000111ff\U00011212\U0001123f-\U0001127f\U00011287\U00011289\U0001128e\U0001129e\U000112aa-\U000112af\U000112eb-\U000112ef\U000112fa-\U000112ff\U00011304\U0001130d-\U0001130e\U00011311-\U00011312\U00011329\U00011331\U00011334\U0001133a\U00011345-\U00011346\U00011349-\U0001134a\U0001134e-\U0001134f\U00011351-\U00011356\U00011358-\U0001135c\U00011364-\U00011365\U0001136d-\U0001136f\U00011375-\U000113ff\U0001145a\U0001145c\U0001145f-\U0001147f\U000114c8-\U000114cf\U000114da-\U0001157f\U000115b6-\U000115b7\U000115de-\U000115ff\U00011645-\U0001164f\U0001165a-\U0001165f\U0001166d-\U0001167f\U000116b8-\U000116bf\U000116ca-\U000116ff\U0001171b-\U0001171c\U0001172c-\U0001172f\U00011740-\U000117ff\U0001183c-\U0001189f\U000118f3-\U000118fe\U00011900-\U000119ff\U00011a48-\U00011a4f\U00011a84-\U00011a85\U00011aa3-\U00011abf\U00011af9-\U00011bff\U00011c09\U00011c37\U00011c46-\U00011c4f\U00011c6d-\U00011c6f\U00011c90-\U00011c91\U00011ca8\U00011cb7-\U00011cff\U00011d07\U00011d0a\U00011d37-\U00011d39\U00011d3b\U00011d3e\U00011d48-\U00011d4f\U00011d5a-\U00011d5f\U00011d66\U00011d69\U00011d8f\U00011d92\U00011d99-\U00011d9f\U00011daa-\U00011edf\U00011ef9-\U00011fff\U0001239a-\U000123ff\U0001246f\U00012475-\U0001247f\U00012544-\U00012fff\U0001342f-\U000143ff\U00014647-\U000167ff\U00016a39-\U00016a3f\U00016a5f\U00016a6a-\U00016a6d\U00016a70-\U00016acf\U00016aee-\U00016aef\U00016af6-\U00016aff\U00016b46-\U00016b4f\U00016b5a\U00016b62\U00016b78-\U00016b7c\U00016b90-\U00016e3f\U00016e9b-\U00016eff\U00016f45-\U00016f4f\U00016f7f-\U00016f8e\U00016fa0-\U00016fdf\U00016fe2-\U00016fff\U000187f2-\U000187ff\U00018af3-\U0001afff\U0001b11f-\U0001b16f\U0001b2fc-\U0001bbff\U0001bc6b-\U0001bc6f\U0001bc7d-\U0001bc7f\U0001bc89-\U0001bc8f\U0001bc9a-\U0001bc9b\U0001bca4-\U0001cfff\U0001d0f6-\U0001d0ff\U0001d127-\U0001d128\U0001d1e9-\U0001d1ff\U0001d246-\U0001d2df\U0001d2f4-\U0001d2ff\U0001d357-\U0001d35f\U0001d379-\U0001d3ff\U0001d455\U0001d49d\U0001d4a0-\U0001d4a1\U0001d4a3-\U0001d4a4\U0001d4a7-\U0001d4a8\U0001d4ad\U0001d4ba\U0001d4bc\U0001d4c4\U0001d506\U0001d50b-\U0001d50c\U0001d515\U0001d51d\U0001d53a\U0001d53f\U0001d545\U0001d547-\U0001d549\U0001d551\U0001d6a6-\U0001d6a7\U0001d7cc-\U0001d7cd\U0001da8c-\U0001da9a\U0001daa0\U0001dab0-\U0001dfff\U0001e007\U0001e019-\U0001e01a\U0001e022\U0001e025\U0001e02b-\U0001e7ff\U0001e8c5-\U0001e8c6\U0001e8d7-\U0001e8ff\U0001e94b-\U0001e94f\U0001e95a-\U0001e95d\U0001e960-\U0001ec70\U0001ecb5-\U0001edff\U0001ee04\U0001ee20\U0001ee23\U0001ee25-\U0001ee26\U0001ee28\U0001ee33\U0001ee38\U0001ee3a\U0001ee3c-\U0001ee41\U0001ee43-\U0001ee46\U0001ee48\U0001ee4a\U0001ee4c\U0001ee50\U0001ee53\U0001ee55-\U0001ee56\U0001ee58\U0001ee5a\U0001ee5c\U0001ee5e\U0001ee60\U0001ee63\U0001ee65-\U0001ee66\U0001ee6b\U0001ee73\U0001ee78\U0001ee7d\U0001ee7f\U0001ee8a\U0001ee9c-\U0001eea0\U0001eea4\U0001eeaa\U0001eebc-\U0001eeef\U0001eef2-\U0001efff\U0001f02c-\U0001f02f\U0001f094-\U0001f09f\U0001f0af-\U0001f0b0\U0001f0c0\U0001f0d0\U0001f0f6-\U0001f0ff\U0001f10d-\U0001f10f\U0001f16c-\U0001f16f\U0001f1ad-\U0001f1e5\U0001f203-\U0001f20f\U0001f23c-\U0001f23f\U0001f249-\U0001f24f\U0001f252-\U0001f25f\U0001f266-\U0001f2ff\U0001f6d5-\U0001f6df\U0001f6ed-\U0001f6ef\U0001f6fa-\U0001f6ff\U0001f774-\U0001f77f\U0001f7d9-\U0001f7ff\U0001f80c-\U0001f80f\U0001f848-\U0001f84f\U0001f85a-\U0001f85f\U0001f888-\U0001f88f\U0001f8ae-\U0001f8ff\U0001f90c-\U0001f90f\U0001f93f\U0001f971-\U0001f972\U0001f977-\U0001f979\U0001f97b\U0001f9a3-\U0001f9af\U0001f9ba-\U0001f9bf\U0001f9c3-\U0001f9cf\U0001fa00-\U0001fa5f\U0001fa6e-\U0001ffff\U0002a6d7-\U0002a6ff\U0002b735-\U0002b73f\U0002b81e-\U0002b81f\U0002cea2-\U0002ceaf\U0002ebe1-\U0002f7ff\U0002fa1e-\U000e0000\U000e0002-\U000e001f\U000e0080-\U000e00ff\U000e01f0-\U000effff\U000ffffe-\U000fffff\U0010fffe-\U0010ffff' -Co = u'\ue000-\uf8ff' +Co = '\ue000-\uf8ff\U000f0000-\U000ffffd\U00100000-\U0010fffd' -try: - Cs = eval(r"u'\ud800-\udbff\\\udc00\udc01-\udfff'") -except UnicodeDecodeError: - Cs = '' # Jython can't handle isolated surrogates +Cs = '\ud800-\udbff\\\udc00\udc01-\udfff' -Ll = u'a-z\xb5\xdf-\xf6\xf8-\xff\u0101\u0103\u0105\u0107\u0109\u010b\u010d\u010f\u0111\u0113\u0115\u0117\u0119\u011b\u011d\u011f\u0121\u0123\u0125\u0127\u0129\u012b\u012d\u012f\u0131\u0133\u0135\u0137-\u0138\u013a\u013c\u013e\u0140\u0142\u0144\u0146\u0148-\u0149\u014b\u014d\u014f\u0151\u0153\u0155\u0157\u0159\u015b\u015d\u015f\u0161\u0163\u0165\u0167\u0169\u016b\u016d\u016f\u0171\u0173\u0175\u0177\u017a\u017c\u017e-\u0180\u0183\u0185\u0188\u018c-\u018d\u0192\u0195\u0199-\u019b\u019e\u01a1\u01a3\u01a5\u01a8\u01aa-\u01ab\u01ad\u01b0\u01b4\u01b6\u01b9-\u01ba\u01bd-\u01bf\u01c6\u01c9\u01cc\u01ce\u01d0\u01d2\u01d4\u01d6\u01d8\u01da\u01dc-\u01dd\u01df\u01e1\u01e3\u01e5\u01e7\u01e9\u01eb\u01ed\u01ef-\u01f0\u01f3\u01f5\u01f9\u01fb\u01fd\u01ff\u0201\u0203\u0205\u0207\u0209\u020b\u020d\u020f\u0211\u0213\u0215\u0217\u0219\u021b\u021d\u021f\u0221\u0223\u0225\u0227\u0229\u022b\u022d\u022f\u0231\u0233-\u0239\u023c\u023f-\u0240\u0242\u0247\u0249\u024b\u024d\u024f-\u0293\u0295-\u02af\u0371\u0373\u0377\u037b-\u037d\u0390\u03ac-\u03ce\u03d0-\u03d1\u03d5-\u03d7\u03d9\u03db\u03dd\u03df\u03e1\u03e3\u03e5\u03e7\u03e9\u03eb\u03ed\u03ef-\u03f3\u03f5\u03f8\u03fb-\u03fc\u0430-\u045f\u0461\u0463\u0465\u0467\u0469\u046b\u046d\u046f\u0471\u0473\u0475\u0477\u0479\u047b\u047d\u047f\u0481\u048b\u048d\u048f\u0491\u0493\u0495\u0497\u0499\u049b\u049d\u049f\u04a1\u04a3\u04a5\u04a7\u04a9\u04ab\u04ad\u04af\u04b1\u04b3\u04b5\u04b7\u04b9\u04bb\u04bd\u04bf\u04c2\u04c4\u04c6\u04c8\u04ca\u04cc\u04ce-\u04cf\u04d1\u04d3\u04d5\u04d7\u04d9\u04db\u04dd\u04df\u04e1\u04e3\u04e5\u04e7\u04e9\u04eb\u04ed\u04ef\u04f1\u04f3\u04f5\u04f7\u04f9\u04fb\u04fd\u04ff\u0501\u0503\u0505\u0507\u0509\u050b\u050d\u050f\u0511\u0513\u0515\u0517\u0519\u051b\u051d\u051f\u0521\u0523\u0525\u0527\u0529\u052b\u052d\u052f\u0560-\u0588\u10d0-\u10fa\u10fd-\u10ff\u13f8-\u13fd\u1c80-\u1c88\u1d00-\u1d2b\u1d6b-\u1d77\u1d79-\u1d9a\u1e01\u1e03\u1e05\u1e07\u1e09\u1e0b\u1e0d\u1e0f\u1e11\u1e13\u1e15\u1e17\u1e19\u1e1b\u1e1d\u1e1f\u1e21\u1e23\u1e25\u1e27\u1e29\u1e2b\u1e2d\u1e2f\u1e31\u1e33\u1e35\u1e37\u1e39\u1e3b\u1e3d\u1e3f\u1e41\u1e43\u1e45\u1e47\u1e49\u1e4b\u1e4d\u1e4f\u1e51\u1e53\u1e55\u1e57\u1e59\u1e5b\u1e5d\u1e5f\u1e61\u1e63\u1e65\u1e67\u1e69\u1e6b\u1e6d\u1e6f\u1e71\u1e73\u1e75\u1e77\u1e79\u1e7b\u1e7d\u1e7f\u1e81\u1e83\u1e85\u1e87\u1e89\u1e8b\u1e8d\u1e8f\u1e91\u1e93\u1e95-\u1e9d\u1e9f\u1ea1\u1ea3\u1ea5\u1ea7\u1ea9\u1eab\u1ead\u1eaf\u1eb1\u1eb3\u1eb5\u1eb7\u1eb9\u1ebb\u1ebd\u1ebf\u1ec1\u1ec3\u1ec5\u1ec7\u1ec9\u1ecb\u1ecd\u1ecf\u1ed1\u1ed3\u1ed5\u1ed7\u1ed9\u1edb\u1edd\u1edf\u1ee1\u1ee3\u1ee5\u1ee7\u1ee9\u1eeb\u1eed\u1eef\u1ef1\u1ef3\u1ef5\u1ef7\u1ef9\u1efb\u1efd\u1eff-\u1f07\u1f10-\u1f15\u1f20-\u1f27\u1f30-\u1f37\u1f40-\u1f45\u1f50-\u1f57\u1f60-\u1f67\u1f70-\u1f7d\u1f80-\u1f87\u1f90-\u1f97\u1fa0-\u1fa7\u1fb0-\u1fb4\u1fb6-\u1fb7\u1fbe\u1fc2-\u1fc4\u1fc6-\u1fc7\u1fd0-\u1fd3\u1fd6-\u1fd7\u1fe0-\u1fe7\u1ff2-\u1ff4\u1ff6-\u1ff7\u210a\u210e-\u210f\u2113\u212f\u2134\u2139\u213c-\u213d\u2146-\u2149\u214e\u2184\u2c30-\u2c5e\u2c61\u2c65-\u2c66\u2c68\u2c6a\u2c6c\u2c71\u2c73-\u2c74\u2c76-\u2c7b\u2c81\u2c83\u2c85\u2c87\u2c89\u2c8b\u2c8d\u2c8f\u2c91\u2c93\u2c95\u2c97\u2c99\u2c9b\u2c9d\u2c9f\u2ca1\u2ca3\u2ca5\u2ca7\u2ca9\u2cab\u2cad\u2caf\u2cb1\u2cb3\u2cb5\u2cb7\u2cb9\u2cbb\u2cbd\u2cbf\u2cc1\u2cc3\u2cc5\u2cc7\u2cc9\u2ccb\u2ccd\u2ccf\u2cd1\u2cd3\u2cd5\u2cd7\u2cd9\u2cdb\u2cdd\u2cdf\u2ce1\u2ce3-\u2ce4\u2cec\u2cee\u2cf3\u2d00-\u2d25\u2d27\u2d2d\ua641\ua643\ua645\ua647\ua649\ua64b\ua64d\ua64f\ua651\ua653\ua655\ua657\ua659\ua65b\ua65d\ua65f\ua661\ua663\ua665\ua667\ua669\ua66b\ua66d\ua681\ua683\ua685\ua687\ua689\ua68b\ua68d\ua68f\ua691\ua693\ua695\ua697\ua699\ua69b\ua723\ua725\ua727\ua729\ua72b\ua72d\ua72f-\ua731\ua733\ua735\ua737\ua739\ua73b\ua73d\ua73f\ua741\ua743\ua745\ua747\ua749\ua74b\ua74d\ua74f\ua751\ua753\ua755\ua757\ua759\ua75b\ua75d\ua75f\ua761\ua763\ua765\ua767\ua769\ua76b\ua76d\ua76f\ua771-\ua778\ua77a\ua77c\ua77f\ua781\ua783\ua785\ua787\ua78c\ua78e\ua791\ua793-\ua795\ua797\ua799\ua79b\ua79d\ua79f\ua7a1\ua7a3\ua7a5\ua7a7\ua7a9\ua7af\ua7b5\ua7b7\ua7b9\ua7fa\uab30-\uab5a\uab60-\uab65\uab70-\uabbf\ufb00-\ufb06\ufb13-\ufb17\uff41-\uff5a' +Ll = 'a-z\xb5\xdf-\xf6\xf8-\xff\u0101\u0103\u0105\u0107\u0109\u010b\u010d\u010f\u0111\u0113\u0115\u0117\u0119\u011b\u011d\u011f\u0121\u0123\u0125\u0127\u0129\u012b\u012d\u012f\u0131\u0133\u0135\u0137-\u0138\u013a\u013c\u013e\u0140\u0142\u0144\u0146\u0148-\u0149\u014b\u014d\u014f\u0151\u0153\u0155\u0157\u0159\u015b\u015d\u015f\u0161\u0163\u0165\u0167\u0169\u016b\u016d\u016f\u0171\u0173\u0175\u0177\u017a\u017c\u017e-\u0180\u0183\u0185\u0188\u018c-\u018d\u0192\u0195\u0199-\u019b\u019e\u01a1\u01a3\u01a5\u01a8\u01aa-\u01ab\u01ad\u01b0\u01b4\u01b6\u01b9-\u01ba\u01bd-\u01bf\u01c6\u01c9\u01cc\u01ce\u01d0\u01d2\u01d4\u01d6\u01d8\u01da\u01dc-\u01dd\u01df\u01e1\u01e3\u01e5\u01e7\u01e9\u01eb\u01ed\u01ef-\u01f0\u01f3\u01f5\u01f9\u01fb\u01fd\u01ff\u0201\u0203\u0205\u0207\u0209\u020b\u020d\u020f\u0211\u0213\u0215\u0217\u0219\u021b\u021d\u021f\u0221\u0223\u0225\u0227\u0229\u022b\u022d\u022f\u0231\u0233-\u0239\u023c\u023f-\u0240\u0242\u0247\u0249\u024b\u024d\u024f-\u0293\u0295-\u02af\u0371\u0373\u0377\u037b-\u037d\u0390\u03ac-\u03ce\u03d0-\u03d1\u03d5-\u03d7\u03d9\u03db\u03dd\u03df\u03e1\u03e3\u03e5\u03e7\u03e9\u03eb\u03ed\u03ef-\u03f3\u03f5\u03f8\u03fb-\u03fc\u0430-\u045f\u0461\u0463\u0465\u0467\u0469\u046b\u046d\u046f\u0471\u0473\u0475\u0477\u0479\u047b\u047d\u047f\u0481\u048b\u048d\u048f\u0491\u0493\u0495\u0497\u0499\u049b\u049d\u049f\u04a1\u04a3\u04a5\u04a7\u04a9\u04ab\u04ad\u04af\u04b1\u04b3\u04b5\u04b7\u04b9\u04bb\u04bd\u04bf\u04c2\u04c4\u04c6\u04c8\u04ca\u04cc\u04ce-\u04cf\u04d1\u04d3\u04d5\u04d7\u04d9\u04db\u04dd\u04df\u04e1\u04e3\u04e5\u04e7\u04e9\u04eb\u04ed\u04ef\u04f1\u04f3\u04f5\u04f7\u04f9\u04fb\u04fd\u04ff\u0501\u0503\u0505\u0507\u0509\u050b\u050d\u050f\u0511\u0513\u0515\u0517\u0519\u051b\u051d\u051f\u0521\u0523\u0525\u0527\u0529\u052b\u052d\u052f\u0560-\u0588\u10d0-\u10fa\u10fd-\u10ff\u13f8-\u13fd\u1c80-\u1c88\u1d00-\u1d2b\u1d6b-\u1d77\u1d79-\u1d9a\u1e01\u1e03\u1e05\u1e07\u1e09\u1e0b\u1e0d\u1e0f\u1e11\u1e13\u1e15\u1e17\u1e19\u1e1b\u1e1d\u1e1f\u1e21\u1e23\u1e25\u1e27\u1e29\u1e2b\u1e2d\u1e2f\u1e31\u1e33\u1e35\u1e37\u1e39\u1e3b\u1e3d\u1e3f\u1e41\u1e43\u1e45\u1e47\u1e49\u1e4b\u1e4d\u1e4f\u1e51\u1e53\u1e55\u1e57\u1e59\u1e5b\u1e5d\u1e5f\u1e61\u1e63\u1e65\u1e67\u1e69\u1e6b\u1e6d\u1e6f\u1e71\u1e73\u1e75\u1e77\u1e79\u1e7b\u1e7d\u1e7f\u1e81\u1e83\u1e85\u1e87\u1e89\u1e8b\u1e8d\u1e8f\u1e91\u1e93\u1e95-\u1e9d\u1e9f\u1ea1\u1ea3\u1ea5\u1ea7\u1ea9\u1eab\u1ead\u1eaf\u1eb1\u1eb3\u1eb5\u1eb7\u1eb9\u1ebb\u1ebd\u1ebf\u1ec1\u1ec3\u1ec5\u1ec7\u1ec9\u1ecb\u1ecd\u1ecf\u1ed1\u1ed3\u1ed5\u1ed7\u1ed9\u1edb\u1edd\u1edf\u1ee1\u1ee3\u1ee5\u1ee7\u1ee9\u1eeb\u1eed\u1eef\u1ef1\u1ef3\u1ef5\u1ef7\u1ef9\u1efb\u1efd\u1eff-\u1f07\u1f10-\u1f15\u1f20-\u1f27\u1f30-\u1f37\u1f40-\u1f45\u1f50-\u1f57\u1f60-\u1f67\u1f70-\u1f7d\u1f80-\u1f87\u1f90-\u1f97\u1fa0-\u1fa7\u1fb0-\u1fb4\u1fb6-\u1fb7\u1fbe\u1fc2-\u1fc4\u1fc6-\u1fc7\u1fd0-\u1fd3\u1fd6-\u1fd7\u1fe0-\u1fe7\u1ff2-\u1ff4\u1ff6-\u1ff7\u210a\u210e-\u210f\u2113\u212f\u2134\u2139\u213c-\u213d\u2146-\u2149\u214e\u2184\u2c30-\u2c5e\u2c61\u2c65-\u2c66\u2c68\u2c6a\u2c6c\u2c71\u2c73-\u2c74\u2c76-\u2c7b\u2c81\u2c83\u2c85\u2c87\u2c89\u2c8b\u2c8d\u2c8f\u2c91\u2c93\u2c95\u2c97\u2c99\u2c9b\u2c9d\u2c9f\u2ca1\u2ca3\u2ca5\u2ca7\u2ca9\u2cab\u2cad\u2caf\u2cb1\u2cb3\u2cb5\u2cb7\u2cb9\u2cbb\u2cbd\u2cbf\u2cc1\u2cc3\u2cc5\u2cc7\u2cc9\u2ccb\u2ccd\u2ccf\u2cd1\u2cd3\u2cd5\u2cd7\u2cd9\u2cdb\u2cdd\u2cdf\u2ce1\u2ce3-\u2ce4\u2cec\u2cee\u2cf3\u2d00-\u2d25\u2d27\u2d2d\ua641\ua643\ua645\ua647\ua649\ua64b\ua64d\ua64f\ua651\ua653\ua655\ua657\ua659\ua65b\ua65d\ua65f\ua661\ua663\ua665\ua667\ua669\ua66b\ua66d\ua681\ua683\ua685\ua687\ua689\ua68b\ua68d\ua68f\ua691\ua693\ua695\ua697\ua699\ua69b\ua723\ua725\ua727\ua729\ua72b\ua72d\ua72f-\ua731\ua733\ua735\ua737\ua739\ua73b\ua73d\ua73f\ua741\ua743\ua745\ua747\ua749\ua74b\ua74d\ua74f\ua751\ua753\ua755\ua757\ua759\ua75b\ua75d\ua75f\ua761\ua763\ua765\ua767\ua769\ua76b\ua76d\ua76f\ua771-\ua778\ua77a\ua77c\ua77f\ua781\ua783\ua785\ua787\ua78c\ua78e\ua791\ua793-\ua795\ua797\ua799\ua79b\ua79d\ua79f\ua7a1\ua7a3\ua7a5\ua7a7\ua7a9\ua7af\ua7b5\ua7b7\ua7b9\ua7fa\uab30-\uab5a\uab60-\uab65\uab70-\uabbf\ufb00-\ufb06\ufb13-\ufb17\uff41-\uff5a\U00010428-\U0001044f\U000104d8-\U000104fb\U00010cc0-\U00010cf2\U000118c0-\U000118df\U00016e60-\U00016e7f\U0001d41a-\U0001d433\U0001d44e-\U0001d454\U0001d456-\U0001d467\U0001d482-\U0001d49b\U0001d4b6-\U0001d4b9\U0001d4bb\U0001d4bd-\U0001d4c3\U0001d4c5-\U0001d4cf\U0001d4ea-\U0001d503\U0001d51e-\U0001d537\U0001d552-\U0001d56b\U0001d586-\U0001d59f\U0001d5ba-\U0001d5d3\U0001d5ee-\U0001d607\U0001d622-\U0001d63b\U0001d656-\U0001d66f\U0001d68a-\U0001d6a5\U0001d6c2-\U0001d6da\U0001d6dc-\U0001d6e1\U0001d6fc-\U0001d714\U0001d716-\U0001d71b\U0001d736-\U0001d74e\U0001d750-\U0001d755\U0001d770-\U0001d788\U0001d78a-\U0001d78f\U0001d7aa-\U0001d7c2\U0001d7c4-\U0001d7c9\U0001d7cb\U0001e922-\U0001e943' -Lm = u'\u02b0-\u02c1\u02c6-\u02d1\u02e0-\u02e4\u02ec\u02ee\u0374\u037a\u0559\u0640\u06e5-\u06e6\u07f4-\u07f5\u07fa\u081a\u0824\u0828\u0971\u0e46\u0ec6\u10fc\u17d7\u1843\u1aa7\u1c78-\u1c7d\u1d2c-\u1d6a\u1d78\u1d9b-\u1dbf\u2071\u207f\u2090-\u209c\u2c7c-\u2c7d\u2d6f\u2e2f\u3005\u3031-\u3035\u303b\u309d-\u309e\u30fc-\u30fe\ua015\ua4f8-\ua4fd\ua60c\ua67f\ua69c-\ua69d\ua717-\ua71f\ua770\ua788\ua7f8-\ua7f9\ua9cf\ua9e6\uaa70\uaadd\uaaf3-\uaaf4\uab5c-\uab5f\uff70\uff9e-\uff9f' +Lm = '\u02b0-\u02c1\u02c6-\u02d1\u02e0-\u02e4\u02ec\u02ee\u0374\u037a\u0559\u0640\u06e5-\u06e6\u07f4-\u07f5\u07fa\u081a\u0824\u0828\u0971\u0e46\u0ec6\u10fc\u17d7\u1843\u1aa7\u1c78-\u1c7d\u1d2c-\u1d6a\u1d78\u1d9b-\u1dbf\u2071\u207f\u2090-\u209c\u2c7c-\u2c7d\u2d6f\u2e2f\u3005\u3031-\u3035\u303b\u309d-\u309e\u30fc-\u30fe\ua015\ua4f8-\ua4fd\ua60c\ua67f\ua69c-\ua69d\ua717-\ua71f\ua770\ua788\ua7f8-\ua7f9\ua9cf\ua9e6\uaa70\uaadd\uaaf3-\uaaf4\uab5c-\uab5f\uff70\uff9e-\uff9f\U00016b40-\U00016b43\U00016f93-\U00016f9f\U00016fe0-\U00016fe1' -Lo = u'\xaa\xba\u01bb\u01c0-\u01c3\u0294\u05d0-\u05ea\u05ef-\u05f2\u0620-\u063f\u0641-\u064a\u066e-\u066f\u0671-\u06d3\u06d5\u06ee-\u06ef\u06fa-\u06fc\u06ff\u0710\u0712-\u072f\u074d-\u07a5\u07b1\u07ca-\u07ea\u0800-\u0815\u0840-\u0858\u0860-\u086a\u08a0-\u08b4\u08b6-\u08bd\u0904-\u0939\u093d\u0950\u0958-\u0961\u0972-\u0980\u0985-\u098c\u098f-\u0990\u0993-\u09a8\u09aa-\u09b0\u09b2\u09b6-\u09b9\u09bd\u09ce\u09dc-\u09dd\u09df-\u09e1\u09f0-\u09f1\u09fc\u0a05-\u0a0a\u0a0f-\u0a10\u0a13-\u0a28\u0a2a-\u0a30\u0a32-\u0a33\u0a35-\u0a36\u0a38-\u0a39\u0a59-\u0a5c\u0a5e\u0a72-\u0a74\u0a85-\u0a8d\u0a8f-\u0a91\u0a93-\u0aa8\u0aaa-\u0ab0\u0ab2-\u0ab3\u0ab5-\u0ab9\u0abd\u0ad0\u0ae0-\u0ae1\u0af9\u0b05-\u0b0c\u0b0f-\u0b10\u0b13-\u0b28\u0b2a-\u0b30\u0b32-\u0b33\u0b35-\u0b39\u0b3d\u0b5c-\u0b5d\u0b5f-\u0b61\u0b71\u0b83\u0b85-\u0b8a\u0b8e-\u0b90\u0b92-\u0b95\u0b99-\u0b9a\u0b9c\u0b9e-\u0b9f\u0ba3-\u0ba4\u0ba8-\u0baa\u0bae-\u0bb9\u0bd0\u0c05-\u0c0c\u0c0e-\u0c10\u0c12-\u0c28\u0c2a-\u0c39\u0c3d\u0c58-\u0c5a\u0c60-\u0c61\u0c80\u0c85-\u0c8c\u0c8e-\u0c90\u0c92-\u0ca8\u0caa-\u0cb3\u0cb5-\u0cb9\u0cbd\u0cde\u0ce0-\u0ce1\u0cf1-\u0cf2\u0d05-\u0d0c\u0d0e-\u0d10\u0d12-\u0d3a\u0d3d\u0d4e\u0d54-\u0d56\u0d5f-\u0d61\u0d7a-\u0d7f\u0d85-\u0d96\u0d9a-\u0db1\u0db3-\u0dbb\u0dbd\u0dc0-\u0dc6\u0e01-\u0e30\u0e32-\u0e33\u0e40-\u0e45\u0e81-\u0e82\u0e84\u0e87-\u0e88\u0e8a\u0e8d\u0e94-\u0e97\u0e99-\u0e9f\u0ea1-\u0ea3\u0ea5\u0ea7\u0eaa-\u0eab\u0ead-\u0eb0\u0eb2-\u0eb3\u0ebd\u0ec0-\u0ec4\u0edc-\u0edf\u0f00\u0f40-\u0f47\u0f49-\u0f6c\u0f88-\u0f8c\u1000-\u102a\u103f\u1050-\u1055\u105a-\u105d\u1061\u1065-\u1066\u106e-\u1070\u1075-\u1081\u108e\u1100-\u1248\u124a-\u124d\u1250-\u1256\u1258\u125a-\u125d\u1260-\u1288\u128a-\u128d\u1290-\u12b0\u12b2-\u12b5\u12b8-\u12be\u12c0\u12c2-\u12c5\u12c8-\u12d6\u12d8-\u1310\u1312-\u1315\u1318-\u135a\u1380-\u138f\u1401-\u166c\u166f-\u167f\u1681-\u169a\u16a0-\u16ea\u16f1-\u16f8\u1700-\u170c\u170e-\u1711\u1720-\u1731\u1740-\u1751\u1760-\u176c\u176e-\u1770\u1780-\u17b3\u17dc\u1820-\u1842\u1844-\u1878\u1880-\u1884\u1887-\u18a8\u18aa\u18b0-\u18f5\u1900-\u191e\u1950-\u196d\u1970-\u1974\u1980-\u19ab\u19b0-\u19c9\u1a00-\u1a16\u1a20-\u1a54\u1b05-\u1b33\u1b45-\u1b4b\u1b83-\u1ba0\u1bae-\u1baf\u1bba-\u1be5\u1c00-\u1c23\u1c4d-\u1c4f\u1c5a-\u1c77\u1ce9-\u1cec\u1cee-\u1cf1\u1cf5-\u1cf6\u2135-\u2138\u2d30-\u2d67\u2d80-\u2d96\u2da0-\u2da6\u2da8-\u2dae\u2db0-\u2db6\u2db8-\u2dbe\u2dc0-\u2dc6\u2dc8-\u2dce\u2dd0-\u2dd6\u2dd8-\u2dde\u3006\u303c\u3041-\u3096\u309f\u30a1-\u30fa\u30ff\u3105-\u312f\u3131-\u318e\u31a0-\u31ba\u31f0-\u31ff\u3400-\u4db5\u4e00-\u9fef\ua000-\ua014\ua016-\ua48c\ua4d0-\ua4f7\ua500-\ua60b\ua610-\ua61f\ua62a-\ua62b\ua66e\ua6a0-\ua6e5\ua78f\ua7f7\ua7fb-\ua801\ua803-\ua805\ua807-\ua80a\ua80c-\ua822\ua840-\ua873\ua882-\ua8b3\ua8f2-\ua8f7\ua8fb\ua8fd-\ua8fe\ua90a-\ua925\ua930-\ua946\ua960-\ua97c\ua984-\ua9b2\ua9e0-\ua9e4\ua9e7-\ua9ef\ua9fa-\ua9fe\uaa00-\uaa28\uaa40-\uaa42\uaa44-\uaa4b\uaa60-\uaa6f\uaa71-\uaa76\uaa7a\uaa7e-\uaaaf\uaab1\uaab5-\uaab6\uaab9-\uaabd\uaac0\uaac2\uaadb-\uaadc\uaae0-\uaaea\uaaf2\uab01-\uab06\uab09-\uab0e\uab11-\uab16\uab20-\uab26\uab28-\uab2e\uabc0-\uabe2\uac00-\ud7a3\ud7b0-\ud7c6\ud7cb-\ud7fb\uf900-\ufa6d\ufa70-\ufad9\ufb1d\ufb1f-\ufb28\ufb2a-\ufb36\ufb38-\ufb3c\ufb3e\ufb40-\ufb41\ufb43-\ufb44\ufb46-\ufbb1\ufbd3-\ufd3d\ufd50-\ufd8f\ufd92-\ufdc7\ufdf0-\ufdfb\ufe70-\ufe74\ufe76-\ufefc\uff66-\uff6f\uff71-\uff9d\uffa0-\uffbe\uffc2-\uffc7\uffca-\uffcf\uffd2-\uffd7\uffda-\uffdc' +Lo = '\xaa\xba\u01bb\u01c0-\u01c3\u0294\u05d0-\u05ea\u05ef-\u05f2\u0620-\u063f\u0641-\u064a\u066e-\u066f\u0671-\u06d3\u06d5\u06ee-\u06ef\u06fa-\u06fc\u06ff\u0710\u0712-\u072f\u074d-\u07a5\u07b1\u07ca-\u07ea\u0800-\u0815\u0840-\u0858\u0860-\u086a\u08a0-\u08b4\u08b6-\u08bd\u0904-\u0939\u093d\u0950\u0958-\u0961\u0972-\u0980\u0985-\u098c\u098f-\u0990\u0993-\u09a8\u09aa-\u09b0\u09b2\u09b6-\u09b9\u09bd\u09ce\u09dc-\u09dd\u09df-\u09e1\u09f0-\u09f1\u09fc\u0a05-\u0a0a\u0a0f-\u0a10\u0a13-\u0a28\u0a2a-\u0a30\u0a32-\u0a33\u0a35-\u0a36\u0a38-\u0a39\u0a59-\u0a5c\u0a5e\u0a72-\u0a74\u0a85-\u0a8d\u0a8f-\u0a91\u0a93-\u0aa8\u0aaa-\u0ab0\u0ab2-\u0ab3\u0ab5-\u0ab9\u0abd\u0ad0\u0ae0-\u0ae1\u0af9\u0b05-\u0b0c\u0b0f-\u0b10\u0b13-\u0b28\u0b2a-\u0b30\u0b32-\u0b33\u0b35-\u0b39\u0b3d\u0b5c-\u0b5d\u0b5f-\u0b61\u0b71\u0b83\u0b85-\u0b8a\u0b8e-\u0b90\u0b92-\u0b95\u0b99-\u0b9a\u0b9c\u0b9e-\u0b9f\u0ba3-\u0ba4\u0ba8-\u0baa\u0bae-\u0bb9\u0bd0\u0c05-\u0c0c\u0c0e-\u0c10\u0c12-\u0c28\u0c2a-\u0c39\u0c3d\u0c58-\u0c5a\u0c60-\u0c61\u0c80\u0c85-\u0c8c\u0c8e-\u0c90\u0c92-\u0ca8\u0caa-\u0cb3\u0cb5-\u0cb9\u0cbd\u0cde\u0ce0-\u0ce1\u0cf1-\u0cf2\u0d05-\u0d0c\u0d0e-\u0d10\u0d12-\u0d3a\u0d3d\u0d4e\u0d54-\u0d56\u0d5f-\u0d61\u0d7a-\u0d7f\u0d85-\u0d96\u0d9a-\u0db1\u0db3-\u0dbb\u0dbd\u0dc0-\u0dc6\u0e01-\u0e30\u0e32-\u0e33\u0e40-\u0e45\u0e81-\u0e82\u0e84\u0e87-\u0e88\u0e8a\u0e8d\u0e94-\u0e97\u0e99-\u0e9f\u0ea1-\u0ea3\u0ea5\u0ea7\u0eaa-\u0eab\u0ead-\u0eb0\u0eb2-\u0eb3\u0ebd\u0ec0-\u0ec4\u0edc-\u0edf\u0f00\u0f40-\u0f47\u0f49-\u0f6c\u0f88-\u0f8c\u1000-\u102a\u103f\u1050-\u1055\u105a-\u105d\u1061\u1065-\u1066\u106e-\u1070\u1075-\u1081\u108e\u1100-\u1248\u124a-\u124d\u1250-\u1256\u1258\u125a-\u125d\u1260-\u1288\u128a-\u128d\u1290-\u12b0\u12b2-\u12b5\u12b8-\u12be\u12c0\u12c2-\u12c5\u12c8-\u12d6\u12d8-\u1310\u1312-\u1315\u1318-\u135a\u1380-\u138f\u1401-\u166c\u166f-\u167f\u1681-\u169a\u16a0-\u16ea\u16f1-\u16f8\u1700-\u170c\u170e-\u1711\u1720-\u1731\u1740-\u1751\u1760-\u176c\u176e-\u1770\u1780-\u17b3\u17dc\u1820-\u1842\u1844-\u1878\u1880-\u1884\u1887-\u18a8\u18aa\u18b0-\u18f5\u1900-\u191e\u1950-\u196d\u1970-\u1974\u1980-\u19ab\u19b0-\u19c9\u1a00-\u1a16\u1a20-\u1a54\u1b05-\u1b33\u1b45-\u1b4b\u1b83-\u1ba0\u1bae-\u1baf\u1bba-\u1be5\u1c00-\u1c23\u1c4d-\u1c4f\u1c5a-\u1c77\u1ce9-\u1cec\u1cee-\u1cf1\u1cf5-\u1cf6\u2135-\u2138\u2d30-\u2d67\u2d80-\u2d96\u2da0-\u2da6\u2da8-\u2dae\u2db0-\u2db6\u2db8-\u2dbe\u2dc0-\u2dc6\u2dc8-\u2dce\u2dd0-\u2dd6\u2dd8-\u2dde\u3006\u303c\u3041-\u3096\u309f\u30a1-\u30fa\u30ff\u3105-\u312f\u3131-\u318e\u31a0-\u31ba\u31f0-\u31ff\u3400-\u4db5\u4e00-\u9fef\ua000-\ua014\ua016-\ua48c\ua4d0-\ua4f7\ua500-\ua60b\ua610-\ua61f\ua62a-\ua62b\ua66e\ua6a0-\ua6e5\ua78f\ua7f7\ua7fb-\ua801\ua803-\ua805\ua807-\ua80a\ua80c-\ua822\ua840-\ua873\ua882-\ua8b3\ua8f2-\ua8f7\ua8fb\ua8fd-\ua8fe\ua90a-\ua925\ua930-\ua946\ua960-\ua97c\ua984-\ua9b2\ua9e0-\ua9e4\ua9e7-\ua9ef\ua9fa-\ua9fe\uaa00-\uaa28\uaa40-\uaa42\uaa44-\uaa4b\uaa60-\uaa6f\uaa71-\uaa76\uaa7a\uaa7e-\uaaaf\uaab1\uaab5-\uaab6\uaab9-\uaabd\uaac0\uaac2\uaadb-\uaadc\uaae0-\uaaea\uaaf2\uab01-\uab06\uab09-\uab0e\uab11-\uab16\uab20-\uab26\uab28-\uab2e\uabc0-\uabe2\uac00-\ud7a3\ud7b0-\ud7c6\ud7cb-\ud7fb\uf900-\ufa6d\ufa70-\ufad9\ufb1d\ufb1f-\ufb28\ufb2a-\ufb36\ufb38-\ufb3c\ufb3e\ufb40-\ufb41\ufb43-\ufb44\ufb46-\ufbb1\ufbd3-\ufd3d\ufd50-\ufd8f\ufd92-\ufdc7\ufdf0-\ufdfb\ufe70-\ufe74\ufe76-\ufefc\uff66-\uff6f\uff71-\uff9d\uffa0-\uffbe\uffc2-\uffc7\uffca-\uffcf\uffd2-\uffd7\uffda-\uffdc\U00010000-\U0001000b\U0001000d-\U00010026\U00010028-\U0001003a\U0001003c-\U0001003d\U0001003f-\U0001004d\U00010050-\U0001005d\U00010080-\U000100fa\U00010280-\U0001029c\U000102a0-\U000102d0\U00010300-\U0001031f\U0001032d-\U00010340\U00010342-\U00010349\U00010350-\U00010375\U00010380-\U0001039d\U000103a0-\U000103c3\U000103c8-\U000103cf\U00010450-\U0001049d\U00010500-\U00010527\U00010530-\U00010563\U00010600-\U00010736\U00010740-\U00010755\U00010760-\U00010767\U00010800-\U00010805\U00010808\U0001080a-\U00010835\U00010837-\U00010838\U0001083c\U0001083f-\U00010855\U00010860-\U00010876\U00010880-\U0001089e\U000108e0-\U000108f2\U000108f4-\U000108f5\U00010900-\U00010915\U00010920-\U00010939\U00010980-\U000109b7\U000109be-\U000109bf\U00010a00\U00010a10-\U00010a13\U00010a15-\U00010a17\U00010a19-\U00010a35\U00010a60-\U00010a7c\U00010a80-\U00010a9c\U00010ac0-\U00010ac7\U00010ac9-\U00010ae4\U00010b00-\U00010b35\U00010b40-\U00010b55\U00010b60-\U00010b72\U00010b80-\U00010b91\U00010c00-\U00010c48\U00010d00-\U00010d23\U00010f00-\U00010f1c\U00010f27\U00010f30-\U00010f45\U00011003-\U00011037\U00011083-\U000110af\U000110d0-\U000110e8\U00011103-\U00011126\U00011144\U00011150-\U00011172\U00011176\U00011183-\U000111b2\U000111c1-\U000111c4\U000111da\U000111dc\U00011200-\U00011211\U00011213-\U0001122b\U00011280-\U00011286\U00011288\U0001128a-\U0001128d\U0001128f-\U0001129d\U0001129f-\U000112a8\U000112b0-\U000112de\U00011305-\U0001130c\U0001130f-\U00011310\U00011313-\U00011328\U0001132a-\U00011330\U00011332-\U00011333\U00011335-\U00011339\U0001133d\U00011350\U0001135d-\U00011361\U00011400-\U00011434\U00011447-\U0001144a\U00011480-\U000114af\U000114c4-\U000114c5\U000114c7\U00011580-\U000115ae\U000115d8-\U000115db\U00011600-\U0001162f\U00011644\U00011680-\U000116aa\U00011700-\U0001171a\U00011800-\U0001182b\U000118ff\U00011a00\U00011a0b-\U00011a32\U00011a3a\U00011a50\U00011a5c-\U00011a83\U00011a86-\U00011a89\U00011a9d\U00011ac0-\U00011af8\U00011c00-\U00011c08\U00011c0a-\U00011c2e\U00011c40\U00011c72-\U00011c8f\U00011d00-\U00011d06\U00011d08-\U00011d09\U00011d0b-\U00011d30\U00011d46\U00011d60-\U00011d65\U00011d67-\U00011d68\U00011d6a-\U00011d89\U00011d98\U00011ee0-\U00011ef2\U00012000-\U00012399\U00012480-\U00012543\U00013000-\U0001342e\U00014400-\U00014646\U00016800-\U00016a38\U00016a40-\U00016a5e\U00016ad0-\U00016aed\U00016b00-\U00016b2f\U00016b63-\U00016b77\U00016b7d-\U00016b8f\U00016f00-\U00016f44\U00016f50\U00017000-\U000187f1\U00018800-\U00018af2\U0001b000-\U0001b11e\U0001b170-\U0001b2fb\U0001bc00-\U0001bc6a\U0001bc70-\U0001bc7c\U0001bc80-\U0001bc88\U0001bc90-\U0001bc99\U0001e800-\U0001e8c4\U0001ee00-\U0001ee03\U0001ee05-\U0001ee1f\U0001ee21-\U0001ee22\U0001ee24\U0001ee27\U0001ee29-\U0001ee32\U0001ee34-\U0001ee37\U0001ee39\U0001ee3b\U0001ee42\U0001ee47\U0001ee49\U0001ee4b\U0001ee4d-\U0001ee4f\U0001ee51-\U0001ee52\U0001ee54\U0001ee57\U0001ee59\U0001ee5b\U0001ee5d\U0001ee5f\U0001ee61-\U0001ee62\U0001ee64\U0001ee67-\U0001ee6a\U0001ee6c-\U0001ee72\U0001ee74-\U0001ee77\U0001ee79-\U0001ee7c\U0001ee7e\U0001ee80-\U0001ee89\U0001ee8b-\U0001ee9b\U0001eea1-\U0001eea3\U0001eea5-\U0001eea9\U0001eeab-\U0001eebb\U00020000-\U0002a6d6\U0002a700-\U0002b734\U0002b740-\U0002b81d\U0002b820-\U0002cea1\U0002ceb0-\U0002ebe0\U0002f800-\U0002fa1d' -Lt = u'\u01c5\u01c8\u01cb\u01f2\u1f88-\u1f8f\u1f98-\u1f9f\u1fa8-\u1faf\u1fbc\u1fcc\u1ffc' +Lt = '\u01c5\u01c8\u01cb\u01f2\u1f88-\u1f8f\u1f98-\u1f9f\u1fa8-\u1faf\u1fbc\u1fcc\u1ffc' -Lu = u'A-Z\xc0-\xd6\xd8-\xde\u0100\u0102\u0104\u0106\u0108\u010a\u010c\u010e\u0110\u0112\u0114\u0116\u0118\u011a\u011c\u011e\u0120\u0122\u0124\u0126\u0128\u012a\u012c\u012e\u0130\u0132\u0134\u0136\u0139\u013b\u013d\u013f\u0141\u0143\u0145\u0147\u014a\u014c\u014e\u0150\u0152\u0154\u0156\u0158\u015a\u015c\u015e\u0160\u0162\u0164\u0166\u0168\u016a\u016c\u016e\u0170\u0172\u0174\u0176\u0178-\u0179\u017b\u017d\u0181-\u0182\u0184\u0186-\u0187\u0189-\u018b\u018e-\u0191\u0193-\u0194\u0196-\u0198\u019c-\u019d\u019f-\u01a0\u01a2\u01a4\u01a6-\u01a7\u01a9\u01ac\u01ae-\u01af\u01b1-\u01b3\u01b5\u01b7-\u01b8\u01bc\u01c4\u01c7\u01ca\u01cd\u01cf\u01d1\u01d3\u01d5\u01d7\u01d9\u01db\u01de\u01e0\u01e2\u01e4\u01e6\u01e8\u01ea\u01ec\u01ee\u01f1\u01f4\u01f6-\u01f8\u01fa\u01fc\u01fe\u0200\u0202\u0204\u0206\u0208\u020a\u020c\u020e\u0210\u0212\u0214\u0216\u0218\u021a\u021c\u021e\u0220\u0222\u0224\u0226\u0228\u022a\u022c\u022e\u0230\u0232\u023a-\u023b\u023d-\u023e\u0241\u0243-\u0246\u0248\u024a\u024c\u024e\u0370\u0372\u0376\u037f\u0386\u0388-\u038a\u038c\u038e-\u038f\u0391-\u03a1\u03a3-\u03ab\u03cf\u03d2-\u03d4\u03d8\u03da\u03dc\u03de\u03e0\u03e2\u03e4\u03e6\u03e8\u03ea\u03ec\u03ee\u03f4\u03f7\u03f9-\u03fa\u03fd-\u042f\u0460\u0462\u0464\u0466\u0468\u046a\u046c\u046e\u0470\u0472\u0474\u0476\u0478\u047a\u047c\u047e\u0480\u048a\u048c\u048e\u0490\u0492\u0494\u0496\u0498\u049a\u049c\u049e\u04a0\u04a2\u04a4\u04a6\u04a8\u04aa\u04ac\u04ae\u04b0\u04b2\u04b4\u04b6\u04b8\u04ba\u04bc\u04be\u04c0-\u04c1\u04c3\u04c5\u04c7\u04c9\u04cb\u04cd\u04d0\u04d2\u04d4\u04d6\u04d8\u04da\u04dc\u04de\u04e0\u04e2\u04e4\u04e6\u04e8\u04ea\u04ec\u04ee\u04f0\u04f2\u04f4\u04f6\u04f8\u04fa\u04fc\u04fe\u0500\u0502\u0504\u0506\u0508\u050a\u050c\u050e\u0510\u0512\u0514\u0516\u0518\u051a\u051c\u051e\u0520\u0522\u0524\u0526\u0528\u052a\u052c\u052e\u0531-\u0556\u10a0-\u10c5\u10c7\u10cd\u13a0-\u13f5\u1c90-\u1cba\u1cbd-\u1cbf\u1e00\u1e02\u1e04\u1e06\u1e08\u1e0a\u1e0c\u1e0e\u1e10\u1e12\u1e14\u1e16\u1e18\u1e1a\u1e1c\u1e1e\u1e20\u1e22\u1e24\u1e26\u1e28\u1e2a\u1e2c\u1e2e\u1e30\u1e32\u1e34\u1e36\u1e38\u1e3a\u1e3c\u1e3e\u1e40\u1e42\u1e44\u1e46\u1e48\u1e4a\u1e4c\u1e4e\u1e50\u1e52\u1e54\u1e56\u1e58\u1e5a\u1e5c\u1e5e\u1e60\u1e62\u1e64\u1e66\u1e68\u1e6a\u1e6c\u1e6e\u1e70\u1e72\u1e74\u1e76\u1e78\u1e7a\u1e7c\u1e7e\u1e80\u1e82\u1e84\u1e86\u1e88\u1e8a\u1e8c\u1e8e\u1e90\u1e92\u1e94\u1e9e\u1ea0\u1ea2\u1ea4\u1ea6\u1ea8\u1eaa\u1eac\u1eae\u1eb0\u1eb2\u1eb4\u1eb6\u1eb8\u1eba\u1ebc\u1ebe\u1ec0\u1ec2\u1ec4\u1ec6\u1ec8\u1eca\u1ecc\u1ece\u1ed0\u1ed2\u1ed4\u1ed6\u1ed8\u1eda\u1edc\u1ede\u1ee0\u1ee2\u1ee4\u1ee6\u1ee8\u1eea\u1eec\u1eee\u1ef0\u1ef2\u1ef4\u1ef6\u1ef8\u1efa\u1efc\u1efe\u1f08-\u1f0f\u1f18-\u1f1d\u1f28-\u1f2f\u1f38-\u1f3f\u1f48-\u1f4d\u1f59\u1f5b\u1f5d\u1f5f\u1f68-\u1f6f\u1fb8-\u1fbb\u1fc8-\u1fcb\u1fd8-\u1fdb\u1fe8-\u1fec\u1ff8-\u1ffb\u2102\u2107\u210b-\u210d\u2110-\u2112\u2115\u2119-\u211d\u2124\u2126\u2128\u212a-\u212d\u2130-\u2133\u213e-\u213f\u2145\u2183\u2c00-\u2c2e\u2c60\u2c62-\u2c64\u2c67\u2c69\u2c6b\u2c6d-\u2c70\u2c72\u2c75\u2c7e-\u2c80\u2c82\u2c84\u2c86\u2c88\u2c8a\u2c8c\u2c8e\u2c90\u2c92\u2c94\u2c96\u2c98\u2c9a\u2c9c\u2c9e\u2ca0\u2ca2\u2ca4\u2ca6\u2ca8\u2caa\u2cac\u2cae\u2cb0\u2cb2\u2cb4\u2cb6\u2cb8\u2cba\u2cbc\u2cbe\u2cc0\u2cc2\u2cc4\u2cc6\u2cc8\u2cca\u2ccc\u2cce\u2cd0\u2cd2\u2cd4\u2cd6\u2cd8\u2cda\u2cdc\u2cde\u2ce0\u2ce2\u2ceb\u2ced\u2cf2\ua640\ua642\ua644\ua646\ua648\ua64a\ua64c\ua64e\ua650\ua652\ua654\ua656\ua658\ua65a\ua65c\ua65e\ua660\ua662\ua664\ua666\ua668\ua66a\ua66c\ua680\ua682\ua684\ua686\ua688\ua68a\ua68c\ua68e\ua690\ua692\ua694\ua696\ua698\ua69a\ua722\ua724\ua726\ua728\ua72a\ua72c\ua72e\ua732\ua734\ua736\ua738\ua73a\ua73c\ua73e\ua740\ua742\ua744\ua746\ua748\ua74a\ua74c\ua74e\ua750\ua752\ua754\ua756\ua758\ua75a\ua75c\ua75e\ua760\ua762\ua764\ua766\ua768\ua76a\ua76c\ua76e\ua779\ua77b\ua77d-\ua77e\ua780\ua782\ua784\ua786\ua78b\ua78d\ua790\ua792\ua796\ua798\ua79a\ua79c\ua79e\ua7a0\ua7a2\ua7a4\ua7a6\ua7a8\ua7aa-\ua7ae\ua7b0-\ua7b4\ua7b6\ua7b8\uff21-\uff3a' +Lu = 'A-Z\xc0-\xd6\xd8-\xde\u0100\u0102\u0104\u0106\u0108\u010a\u010c\u010e\u0110\u0112\u0114\u0116\u0118\u011a\u011c\u011e\u0120\u0122\u0124\u0126\u0128\u012a\u012c\u012e\u0130\u0132\u0134\u0136\u0139\u013b\u013d\u013f\u0141\u0143\u0145\u0147\u014a\u014c\u014e\u0150\u0152\u0154\u0156\u0158\u015a\u015c\u015e\u0160\u0162\u0164\u0166\u0168\u016a\u016c\u016e\u0170\u0172\u0174\u0176\u0178-\u0179\u017b\u017d\u0181-\u0182\u0184\u0186-\u0187\u0189-\u018b\u018e-\u0191\u0193-\u0194\u0196-\u0198\u019c-\u019d\u019f-\u01a0\u01a2\u01a4\u01a6-\u01a7\u01a9\u01ac\u01ae-\u01af\u01b1-\u01b3\u01b5\u01b7-\u01b8\u01bc\u01c4\u01c7\u01ca\u01cd\u01cf\u01d1\u01d3\u01d5\u01d7\u01d9\u01db\u01de\u01e0\u01e2\u01e4\u01e6\u01e8\u01ea\u01ec\u01ee\u01f1\u01f4\u01f6-\u01f8\u01fa\u01fc\u01fe\u0200\u0202\u0204\u0206\u0208\u020a\u020c\u020e\u0210\u0212\u0214\u0216\u0218\u021a\u021c\u021e\u0220\u0222\u0224\u0226\u0228\u022a\u022c\u022e\u0230\u0232\u023a-\u023b\u023d-\u023e\u0241\u0243-\u0246\u0248\u024a\u024c\u024e\u0370\u0372\u0376\u037f\u0386\u0388-\u038a\u038c\u038e-\u038f\u0391-\u03a1\u03a3-\u03ab\u03cf\u03d2-\u03d4\u03d8\u03da\u03dc\u03de\u03e0\u03e2\u03e4\u03e6\u03e8\u03ea\u03ec\u03ee\u03f4\u03f7\u03f9-\u03fa\u03fd-\u042f\u0460\u0462\u0464\u0466\u0468\u046a\u046c\u046e\u0470\u0472\u0474\u0476\u0478\u047a\u047c\u047e\u0480\u048a\u048c\u048e\u0490\u0492\u0494\u0496\u0498\u049a\u049c\u049e\u04a0\u04a2\u04a4\u04a6\u04a8\u04aa\u04ac\u04ae\u04b0\u04b2\u04b4\u04b6\u04b8\u04ba\u04bc\u04be\u04c0-\u04c1\u04c3\u04c5\u04c7\u04c9\u04cb\u04cd\u04d0\u04d2\u04d4\u04d6\u04d8\u04da\u04dc\u04de\u04e0\u04e2\u04e4\u04e6\u04e8\u04ea\u04ec\u04ee\u04f0\u04f2\u04f4\u04f6\u04f8\u04fa\u04fc\u04fe\u0500\u0502\u0504\u0506\u0508\u050a\u050c\u050e\u0510\u0512\u0514\u0516\u0518\u051a\u051c\u051e\u0520\u0522\u0524\u0526\u0528\u052a\u052c\u052e\u0531-\u0556\u10a0-\u10c5\u10c7\u10cd\u13a0-\u13f5\u1c90-\u1cba\u1cbd-\u1cbf\u1e00\u1e02\u1e04\u1e06\u1e08\u1e0a\u1e0c\u1e0e\u1e10\u1e12\u1e14\u1e16\u1e18\u1e1a\u1e1c\u1e1e\u1e20\u1e22\u1e24\u1e26\u1e28\u1e2a\u1e2c\u1e2e\u1e30\u1e32\u1e34\u1e36\u1e38\u1e3a\u1e3c\u1e3e\u1e40\u1e42\u1e44\u1e46\u1e48\u1e4a\u1e4c\u1e4e\u1e50\u1e52\u1e54\u1e56\u1e58\u1e5a\u1e5c\u1e5e\u1e60\u1e62\u1e64\u1e66\u1e68\u1e6a\u1e6c\u1e6e\u1e70\u1e72\u1e74\u1e76\u1e78\u1e7a\u1e7c\u1e7e\u1e80\u1e82\u1e84\u1e86\u1e88\u1e8a\u1e8c\u1e8e\u1e90\u1e92\u1e94\u1e9e\u1ea0\u1ea2\u1ea4\u1ea6\u1ea8\u1eaa\u1eac\u1eae\u1eb0\u1eb2\u1eb4\u1eb6\u1eb8\u1eba\u1ebc\u1ebe\u1ec0\u1ec2\u1ec4\u1ec6\u1ec8\u1eca\u1ecc\u1ece\u1ed0\u1ed2\u1ed4\u1ed6\u1ed8\u1eda\u1edc\u1ede\u1ee0\u1ee2\u1ee4\u1ee6\u1ee8\u1eea\u1eec\u1eee\u1ef0\u1ef2\u1ef4\u1ef6\u1ef8\u1efa\u1efc\u1efe\u1f08-\u1f0f\u1f18-\u1f1d\u1f28-\u1f2f\u1f38-\u1f3f\u1f48-\u1f4d\u1f59\u1f5b\u1f5d\u1f5f\u1f68-\u1f6f\u1fb8-\u1fbb\u1fc8-\u1fcb\u1fd8-\u1fdb\u1fe8-\u1fec\u1ff8-\u1ffb\u2102\u2107\u210b-\u210d\u2110-\u2112\u2115\u2119-\u211d\u2124\u2126\u2128\u212a-\u212d\u2130-\u2133\u213e-\u213f\u2145\u2183\u2c00-\u2c2e\u2c60\u2c62-\u2c64\u2c67\u2c69\u2c6b\u2c6d-\u2c70\u2c72\u2c75\u2c7e-\u2c80\u2c82\u2c84\u2c86\u2c88\u2c8a\u2c8c\u2c8e\u2c90\u2c92\u2c94\u2c96\u2c98\u2c9a\u2c9c\u2c9e\u2ca0\u2ca2\u2ca4\u2ca6\u2ca8\u2caa\u2cac\u2cae\u2cb0\u2cb2\u2cb4\u2cb6\u2cb8\u2cba\u2cbc\u2cbe\u2cc0\u2cc2\u2cc4\u2cc6\u2cc8\u2cca\u2ccc\u2cce\u2cd0\u2cd2\u2cd4\u2cd6\u2cd8\u2cda\u2cdc\u2cde\u2ce0\u2ce2\u2ceb\u2ced\u2cf2\ua640\ua642\ua644\ua646\ua648\ua64a\ua64c\ua64e\ua650\ua652\ua654\ua656\ua658\ua65a\ua65c\ua65e\ua660\ua662\ua664\ua666\ua668\ua66a\ua66c\ua680\ua682\ua684\ua686\ua688\ua68a\ua68c\ua68e\ua690\ua692\ua694\ua696\ua698\ua69a\ua722\ua724\ua726\ua728\ua72a\ua72c\ua72e\ua732\ua734\ua736\ua738\ua73a\ua73c\ua73e\ua740\ua742\ua744\ua746\ua748\ua74a\ua74c\ua74e\ua750\ua752\ua754\ua756\ua758\ua75a\ua75c\ua75e\ua760\ua762\ua764\ua766\ua768\ua76a\ua76c\ua76e\ua779\ua77b\ua77d-\ua77e\ua780\ua782\ua784\ua786\ua78b\ua78d\ua790\ua792\ua796\ua798\ua79a\ua79c\ua79e\ua7a0\ua7a2\ua7a4\ua7a6\ua7a8\ua7aa-\ua7ae\ua7b0-\ua7b4\ua7b6\ua7b8\uff21-\uff3a\U00010400-\U00010427\U000104b0-\U000104d3\U00010c80-\U00010cb2\U000118a0-\U000118bf\U00016e40-\U00016e5f\U0001d400-\U0001d419\U0001d434-\U0001d44d\U0001d468-\U0001d481\U0001d49c\U0001d49e-\U0001d49f\U0001d4a2\U0001d4a5-\U0001d4a6\U0001d4a9-\U0001d4ac\U0001d4ae-\U0001d4b5\U0001d4d0-\U0001d4e9\U0001d504-\U0001d505\U0001d507-\U0001d50a\U0001d50d-\U0001d514\U0001d516-\U0001d51c\U0001d538-\U0001d539\U0001d53b-\U0001d53e\U0001d540-\U0001d544\U0001d546\U0001d54a-\U0001d550\U0001d56c-\U0001d585\U0001d5a0-\U0001d5b9\U0001d5d4-\U0001d5ed\U0001d608-\U0001d621\U0001d63c-\U0001d655\U0001d670-\U0001d689\U0001d6a8-\U0001d6c0\U0001d6e2-\U0001d6fa\U0001d71c-\U0001d734\U0001d756-\U0001d76e\U0001d790-\U0001d7a8\U0001d7ca\U0001e900-\U0001e921' -Mc = u'\u0903\u093b\u093e-\u0940\u0949-\u094c\u094e-\u094f\u0982-\u0983\u09be-\u09c0\u09c7-\u09c8\u09cb-\u09cc\u09d7\u0a03\u0a3e-\u0a40\u0a83\u0abe-\u0ac0\u0ac9\u0acb-\u0acc\u0b02-\u0b03\u0b3e\u0b40\u0b47-\u0b48\u0b4b-\u0b4c\u0b57\u0bbe-\u0bbf\u0bc1-\u0bc2\u0bc6-\u0bc8\u0bca-\u0bcc\u0bd7\u0c01-\u0c03\u0c41-\u0c44\u0c82-\u0c83\u0cbe\u0cc0-\u0cc4\u0cc7-\u0cc8\u0cca-\u0ccb\u0cd5-\u0cd6\u0d02-\u0d03\u0d3e-\u0d40\u0d46-\u0d48\u0d4a-\u0d4c\u0d57\u0d82-\u0d83\u0dcf-\u0dd1\u0dd8-\u0ddf\u0df2-\u0df3\u0f3e-\u0f3f\u0f7f\u102b-\u102c\u1031\u1038\u103b-\u103c\u1056-\u1057\u1062-\u1064\u1067-\u106d\u1083-\u1084\u1087-\u108c\u108f\u109a-\u109c\u17b6\u17be-\u17c5\u17c7-\u17c8\u1923-\u1926\u1929-\u192b\u1930-\u1931\u1933-\u1938\u1a19-\u1a1a\u1a55\u1a57\u1a61\u1a63-\u1a64\u1a6d-\u1a72\u1b04\u1b35\u1b3b\u1b3d-\u1b41\u1b43-\u1b44\u1b82\u1ba1\u1ba6-\u1ba7\u1baa\u1be7\u1bea-\u1bec\u1bee\u1bf2-\u1bf3\u1c24-\u1c2b\u1c34-\u1c35\u1ce1\u1cf2-\u1cf3\u1cf7\u302e-\u302f\ua823-\ua824\ua827\ua880-\ua881\ua8b4-\ua8c3\ua952-\ua953\ua983\ua9b4-\ua9b5\ua9ba-\ua9bb\ua9bd-\ua9c0\uaa2f-\uaa30\uaa33-\uaa34\uaa4d\uaa7b\uaa7d\uaaeb\uaaee-\uaaef\uaaf5\uabe3-\uabe4\uabe6-\uabe7\uabe9-\uabea\uabec' +Mc = '\u0903\u093b\u093e-\u0940\u0949-\u094c\u094e-\u094f\u0982-\u0983\u09be-\u09c0\u09c7-\u09c8\u09cb-\u09cc\u09d7\u0a03\u0a3e-\u0a40\u0a83\u0abe-\u0ac0\u0ac9\u0acb-\u0acc\u0b02-\u0b03\u0b3e\u0b40\u0b47-\u0b48\u0b4b-\u0b4c\u0b57\u0bbe-\u0bbf\u0bc1-\u0bc2\u0bc6-\u0bc8\u0bca-\u0bcc\u0bd7\u0c01-\u0c03\u0c41-\u0c44\u0c82-\u0c83\u0cbe\u0cc0-\u0cc4\u0cc7-\u0cc8\u0cca-\u0ccb\u0cd5-\u0cd6\u0d02-\u0d03\u0d3e-\u0d40\u0d46-\u0d48\u0d4a-\u0d4c\u0d57\u0d82-\u0d83\u0dcf-\u0dd1\u0dd8-\u0ddf\u0df2-\u0df3\u0f3e-\u0f3f\u0f7f\u102b-\u102c\u1031\u1038\u103b-\u103c\u1056-\u1057\u1062-\u1064\u1067-\u106d\u1083-\u1084\u1087-\u108c\u108f\u109a-\u109c\u17b6\u17be-\u17c5\u17c7-\u17c8\u1923-\u1926\u1929-\u192b\u1930-\u1931\u1933-\u1938\u1a19-\u1a1a\u1a55\u1a57\u1a61\u1a63-\u1a64\u1a6d-\u1a72\u1b04\u1b35\u1b3b\u1b3d-\u1b41\u1b43-\u1b44\u1b82\u1ba1\u1ba6-\u1ba7\u1baa\u1be7\u1bea-\u1bec\u1bee\u1bf2-\u1bf3\u1c24-\u1c2b\u1c34-\u1c35\u1ce1\u1cf2-\u1cf3\u1cf7\u302e-\u302f\ua823-\ua824\ua827\ua880-\ua881\ua8b4-\ua8c3\ua952-\ua953\ua983\ua9b4-\ua9b5\ua9ba-\ua9bb\ua9bd-\ua9c0\uaa2f-\uaa30\uaa33-\uaa34\uaa4d\uaa7b\uaa7d\uaaeb\uaaee-\uaaef\uaaf5\uabe3-\uabe4\uabe6-\uabe7\uabe9-\uabea\uabec\U00011000\U00011002\U00011082\U000110b0-\U000110b2\U000110b7-\U000110b8\U0001112c\U00011145-\U00011146\U00011182\U000111b3-\U000111b5\U000111bf-\U000111c0\U0001122c-\U0001122e\U00011232-\U00011233\U00011235\U000112e0-\U000112e2\U00011302-\U00011303\U0001133e-\U0001133f\U00011341-\U00011344\U00011347-\U00011348\U0001134b-\U0001134d\U00011357\U00011362-\U00011363\U00011435-\U00011437\U00011440-\U00011441\U00011445\U000114b0-\U000114b2\U000114b9\U000114bb-\U000114be\U000114c1\U000115af-\U000115b1\U000115b8-\U000115bb\U000115be\U00011630-\U00011632\U0001163b-\U0001163c\U0001163e\U000116ac\U000116ae-\U000116af\U000116b6\U00011720-\U00011721\U00011726\U0001182c-\U0001182e\U00011838\U00011a39\U00011a57-\U00011a58\U00011a97\U00011c2f\U00011c3e\U00011ca9\U00011cb1\U00011cb4\U00011d8a-\U00011d8e\U00011d93-\U00011d94\U00011d96\U00011ef5-\U00011ef6\U00016f51-\U00016f7e\U0001d165-\U0001d166\U0001d16d-\U0001d172' -Me = u'\u0488-\u0489\u1abe\u20dd-\u20e0\u20e2-\u20e4\ua670-\ua672' +Me = '\u0488-\u0489\u1abe\u20dd-\u20e0\u20e2-\u20e4\ua670-\ua672' -Mn = u'\u0300-\u036f\u0483-\u0487\u0591-\u05bd\u05bf\u05c1-\u05c2\u05c4-\u05c5\u05c7\u0610-\u061a\u064b-\u065f\u0670\u06d6-\u06dc\u06df-\u06e4\u06e7-\u06e8\u06ea-\u06ed\u0711\u0730-\u074a\u07a6-\u07b0\u07eb-\u07f3\u07fd\u0816-\u0819\u081b-\u0823\u0825-\u0827\u0829-\u082d\u0859-\u085b\u08d3-\u08e1\u08e3-\u0902\u093a\u093c\u0941-\u0948\u094d\u0951-\u0957\u0962-\u0963\u0981\u09bc\u09c1-\u09c4\u09cd\u09e2-\u09e3\u09fe\u0a01-\u0a02\u0a3c\u0a41-\u0a42\u0a47-\u0a48\u0a4b-\u0a4d\u0a51\u0a70-\u0a71\u0a75\u0a81-\u0a82\u0abc\u0ac1-\u0ac5\u0ac7-\u0ac8\u0acd\u0ae2-\u0ae3\u0afa-\u0aff\u0b01\u0b3c\u0b3f\u0b41-\u0b44\u0b4d\u0b56\u0b62-\u0b63\u0b82\u0bc0\u0bcd\u0c00\u0c04\u0c3e-\u0c40\u0c46-\u0c48\u0c4a-\u0c4d\u0c55-\u0c56\u0c62-\u0c63\u0c81\u0cbc\u0cbf\u0cc6\u0ccc-\u0ccd\u0ce2-\u0ce3\u0d00-\u0d01\u0d3b-\u0d3c\u0d41-\u0d44\u0d4d\u0d62-\u0d63\u0dca\u0dd2-\u0dd4\u0dd6\u0e31\u0e34-\u0e3a\u0e47-\u0e4e\u0eb1\u0eb4-\u0eb9\u0ebb-\u0ebc\u0ec8-\u0ecd\u0f18-\u0f19\u0f35\u0f37\u0f39\u0f71-\u0f7e\u0f80-\u0f84\u0f86-\u0f87\u0f8d-\u0f97\u0f99-\u0fbc\u0fc6\u102d-\u1030\u1032-\u1037\u1039-\u103a\u103d-\u103e\u1058-\u1059\u105e-\u1060\u1071-\u1074\u1082\u1085-\u1086\u108d\u109d\u135d-\u135f\u1712-\u1714\u1732-\u1734\u1752-\u1753\u1772-\u1773\u17b4-\u17b5\u17b7-\u17bd\u17c6\u17c9-\u17d3\u17dd\u180b-\u180d\u1885-\u1886\u18a9\u1920-\u1922\u1927-\u1928\u1932\u1939-\u193b\u1a17-\u1a18\u1a1b\u1a56\u1a58-\u1a5e\u1a60\u1a62\u1a65-\u1a6c\u1a73-\u1a7c\u1a7f\u1ab0-\u1abd\u1b00-\u1b03\u1b34\u1b36-\u1b3a\u1b3c\u1b42\u1b6b-\u1b73\u1b80-\u1b81\u1ba2-\u1ba5\u1ba8-\u1ba9\u1bab-\u1bad\u1be6\u1be8-\u1be9\u1bed\u1bef-\u1bf1\u1c2c-\u1c33\u1c36-\u1c37\u1cd0-\u1cd2\u1cd4-\u1ce0\u1ce2-\u1ce8\u1ced\u1cf4\u1cf8-\u1cf9\u1dc0-\u1df9\u1dfb-\u1dff\u20d0-\u20dc\u20e1\u20e5-\u20f0\u2cef-\u2cf1\u2d7f\u2de0-\u2dff\u302a-\u302d\u3099-\u309a\ua66f\ua674-\ua67d\ua69e-\ua69f\ua6f0-\ua6f1\ua802\ua806\ua80b\ua825-\ua826\ua8c4-\ua8c5\ua8e0-\ua8f1\ua8ff\ua926-\ua92d\ua947-\ua951\ua980-\ua982\ua9b3\ua9b6-\ua9b9\ua9bc\ua9e5\uaa29-\uaa2e\uaa31-\uaa32\uaa35-\uaa36\uaa43\uaa4c\uaa7c\uaab0\uaab2-\uaab4\uaab7-\uaab8\uaabe-\uaabf\uaac1\uaaec-\uaaed\uaaf6\uabe5\uabe8\uabed\ufb1e\ufe00-\ufe0f\ufe20-\ufe2f' +Mn = '\u0300-\u036f\u0483-\u0487\u0591-\u05bd\u05bf\u05c1-\u05c2\u05c4-\u05c5\u05c7\u0610-\u061a\u064b-\u065f\u0670\u06d6-\u06dc\u06df-\u06e4\u06e7-\u06e8\u06ea-\u06ed\u0711\u0730-\u074a\u07a6-\u07b0\u07eb-\u07f3\u07fd\u0816-\u0819\u081b-\u0823\u0825-\u0827\u0829-\u082d\u0859-\u085b\u08d3-\u08e1\u08e3-\u0902\u093a\u093c\u0941-\u0948\u094d\u0951-\u0957\u0962-\u0963\u0981\u09bc\u09c1-\u09c4\u09cd\u09e2-\u09e3\u09fe\u0a01-\u0a02\u0a3c\u0a41-\u0a42\u0a47-\u0a48\u0a4b-\u0a4d\u0a51\u0a70-\u0a71\u0a75\u0a81-\u0a82\u0abc\u0ac1-\u0ac5\u0ac7-\u0ac8\u0acd\u0ae2-\u0ae3\u0afa-\u0aff\u0b01\u0b3c\u0b3f\u0b41-\u0b44\u0b4d\u0b56\u0b62-\u0b63\u0b82\u0bc0\u0bcd\u0c00\u0c04\u0c3e-\u0c40\u0c46-\u0c48\u0c4a-\u0c4d\u0c55-\u0c56\u0c62-\u0c63\u0c81\u0cbc\u0cbf\u0cc6\u0ccc-\u0ccd\u0ce2-\u0ce3\u0d00-\u0d01\u0d3b-\u0d3c\u0d41-\u0d44\u0d4d\u0d62-\u0d63\u0dca\u0dd2-\u0dd4\u0dd6\u0e31\u0e34-\u0e3a\u0e47-\u0e4e\u0eb1\u0eb4-\u0eb9\u0ebb-\u0ebc\u0ec8-\u0ecd\u0f18-\u0f19\u0f35\u0f37\u0f39\u0f71-\u0f7e\u0f80-\u0f84\u0f86-\u0f87\u0f8d-\u0f97\u0f99-\u0fbc\u0fc6\u102d-\u1030\u1032-\u1037\u1039-\u103a\u103d-\u103e\u1058-\u1059\u105e-\u1060\u1071-\u1074\u1082\u1085-\u1086\u108d\u109d\u135d-\u135f\u1712-\u1714\u1732-\u1734\u1752-\u1753\u1772-\u1773\u17b4-\u17b5\u17b7-\u17bd\u17c6\u17c9-\u17d3\u17dd\u180b-\u180d\u1885-\u1886\u18a9\u1920-\u1922\u1927-\u1928\u1932\u1939-\u193b\u1a17-\u1a18\u1a1b\u1a56\u1a58-\u1a5e\u1a60\u1a62\u1a65-\u1a6c\u1a73-\u1a7c\u1a7f\u1ab0-\u1abd\u1b00-\u1b03\u1b34\u1b36-\u1b3a\u1b3c\u1b42\u1b6b-\u1b73\u1b80-\u1b81\u1ba2-\u1ba5\u1ba8-\u1ba9\u1bab-\u1bad\u1be6\u1be8-\u1be9\u1bed\u1bef-\u1bf1\u1c2c-\u1c33\u1c36-\u1c37\u1cd0-\u1cd2\u1cd4-\u1ce0\u1ce2-\u1ce8\u1ced\u1cf4\u1cf8-\u1cf9\u1dc0-\u1df9\u1dfb-\u1dff\u20d0-\u20dc\u20e1\u20e5-\u20f0\u2cef-\u2cf1\u2d7f\u2de0-\u2dff\u302a-\u302d\u3099-\u309a\ua66f\ua674-\ua67d\ua69e-\ua69f\ua6f0-\ua6f1\ua802\ua806\ua80b\ua825-\ua826\ua8c4-\ua8c5\ua8e0-\ua8f1\ua8ff\ua926-\ua92d\ua947-\ua951\ua980-\ua982\ua9b3\ua9b6-\ua9b9\ua9bc\ua9e5\uaa29-\uaa2e\uaa31-\uaa32\uaa35-\uaa36\uaa43\uaa4c\uaa7c\uaab0\uaab2-\uaab4\uaab7-\uaab8\uaabe-\uaabf\uaac1\uaaec-\uaaed\uaaf6\uabe5\uabe8\uabed\ufb1e\ufe00-\ufe0f\ufe20-\ufe2f\U000101fd\U000102e0\U00010376-\U0001037a\U00010a01-\U00010a03\U00010a05-\U00010a06\U00010a0c-\U00010a0f\U00010a38-\U00010a3a\U00010a3f\U00010ae5-\U00010ae6\U00010d24-\U00010d27\U00010f46-\U00010f50\U00011001\U00011038-\U00011046\U0001107f-\U00011081\U000110b3-\U000110b6\U000110b9-\U000110ba\U00011100-\U00011102\U00011127-\U0001112b\U0001112d-\U00011134\U00011173\U00011180-\U00011181\U000111b6-\U000111be\U000111c9-\U000111cc\U0001122f-\U00011231\U00011234\U00011236-\U00011237\U0001123e\U000112df\U000112e3-\U000112ea\U00011300-\U00011301\U0001133b-\U0001133c\U00011340\U00011366-\U0001136c\U00011370-\U00011374\U00011438-\U0001143f\U00011442-\U00011444\U00011446\U0001145e\U000114b3-\U000114b8\U000114ba\U000114bf-\U000114c0\U000114c2-\U000114c3\U000115b2-\U000115b5\U000115bc-\U000115bd\U000115bf-\U000115c0\U000115dc-\U000115dd\U00011633-\U0001163a\U0001163d\U0001163f-\U00011640\U000116ab\U000116ad\U000116b0-\U000116b5\U000116b7\U0001171d-\U0001171f\U00011722-\U00011725\U00011727-\U0001172b\U0001182f-\U00011837\U00011839-\U0001183a\U00011a01-\U00011a0a\U00011a33-\U00011a38\U00011a3b-\U00011a3e\U00011a47\U00011a51-\U00011a56\U00011a59-\U00011a5b\U00011a8a-\U00011a96\U00011a98-\U00011a99\U00011c30-\U00011c36\U00011c38-\U00011c3d\U00011c3f\U00011c92-\U00011ca7\U00011caa-\U00011cb0\U00011cb2-\U00011cb3\U00011cb5-\U00011cb6\U00011d31-\U00011d36\U00011d3a\U00011d3c-\U00011d3d\U00011d3f-\U00011d45\U00011d47\U00011d90-\U00011d91\U00011d95\U00011d97\U00011ef3-\U00011ef4\U00016af0-\U00016af4\U00016b30-\U00016b36\U00016f8f-\U00016f92\U0001bc9d-\U0001bc9e\U0001d167-\U0001d169\U0001d17b-\U0001d182\U0001d185-\U0001d18b\U0001d1aa-\U0001d1ad\U0001d242-\U0001d244\U0001da00-\U0001da36\U0001da3b-\U0001da6c\U0001da75\U0001da84\U0001da9b-\U0001da9f\U0001daa1-\U0001daaf\U0001e000-\U0001e006\U0001e008-\U0001e018\U0001e01b-\U0001e021\U0001e023-\U0001e024\U0001e026-\U0001e02a\U0001e8d0-\U0001e8d6\U0001e944-\U0001e94a\U000e0100-\U000e01ef' -Nd = u'0-9\u0660-\u0669\u06f0-\u06f9\u07c0-\u07c9\u0966-\u096f\u09e6-\u09ef\u0a66-\u0a6f\u0ae6-\u0aef\u0b66-\u0b6f\u0be6-\u0bef\u0c66-\u0c6f\u0ce6-\u0cef\u0d66-\u0d6f\u0de6-\u0def\u0e50-\u0e59\u0ed0-\u0ed9\u0f20-\u0f29\u1040-\u1049\u1090-\u1099\u17e0-\u17e9\u1810-\u1819\u1946-\u194f\u19d0-\u19d9\u1a80-\u1a89\u1a90-\u1a99\u1b50-\u1b59\u1bb0-\u1bb9\u1c40-\u1c49\u1c50-\u1c59\ua620-\ua629\ua8d0-\ua8d9\ua900-\ua909\ua9d0-\ua9d9\ua9f0-\ua9f9\uaa50-\uaa59\uabf0-\uabf9\uff10-\uff19' +Nd = '0-9\u0660-\u0669\u06f0-\u06f9\u07c0-\u07c9\u0966-\u096f\u09e6-\u09ef\u0a66-\u0a6f\u0ae6-\u0aef\u0b66-\u0b6f\u0be6-\u0bef\u0c66-\u0c6f\u0ce6-\u0cef\u0d66-\u0d6f\u0de6-\u0def\u0e50-\u0e59\u0ed0-\u0ed9\u0f20-\u0f29\u1040-\u1049\u1090-\u1099\u17e0-\u17e9\u1810-\u1819\u1946-\u194f\u19d0-\u19d9\u1a80-\u1a89\u1a90-\u1a99\u1b50-\u1b59\u1bb0-\u1bb9\u1c40-\u1c49\u1c50-\u1c59\ua620-\ua629\ua8d0-\ua8d9\ua900-\ua909\ua9d0-\ua9d9\ua9f0-\ua9f9\uaa50-\uaa59\uabf0-\uabf9\uff10-\uff19\U000104a0-\U000104a9\U00010d30-\U00010d39\U00011066-\U0001106f\U000110f0-\U000110f9\U00011136-\U0001113f\U000111d0-\U000111d9\U000112f0-\U000112f9\U00011450-\U00011459\U000114d0-\U000114d9\U00011650-\U00011659\U000116c0-\U000116c9\U00011730-\U00011739\U000118e0-\U000118e9\U00011c50-\U00011c59\U00011d50-\U00011d59\U00011da0-\U00011da9\U00016a60-\U00016a69\U00016b50-\U00016b59\U0001d7ce-\U0001d7ff\U0001e950-\U0001e959' -Nl = u'\u16ee-\u16f0\u2160-\u2182\u2185-\u2188\u3007\u3021-\u3029\u3038-\u303a\ua6e6-\ua6ef' +Nl = '\u16ee-\u16f0\u2160-\u2182\u2185-\u2188\u3007\u3021-\u3029\u3038-\u303a\ua6e6-\ua6ef\U00010140-\U00010174\U00010341\U0001034a\U000103d1-\U000103d5\U00012400-\U0001246e' -No = u'\xb2-\xb3\xb9\xbc-\xbe\u09f4-\u09f9\u0b72-\u0b77\u0bf0-\u0bf2\u0c78-\u0c7e\u0d58-\u0d5e\u0d70-\u0d78\u0f2a-\u0f33\u1369-\u137c\u17f0-\u17f9\u19da\u2070\u2074-\u2079\u2080-\u2089\u2150-\u215f\u2189\u2460-\u249b\u24ea-\u24ff\u2776-\u2793\u2cfd\u3192-\u3195\u3220-\u3229\u3248-\u324f\u3251-\u325f\u3280-\u3289\u32b1-\u32bf\ua830-\ua835' +No = '\xb2-\xb3\xb9\xbc-\xbe\u09f4-\u09f9\u0b72-\u0b77\u0bf0-\u0bf2\u0c78-\u0c7e\u0d58-\u0d5e\u0d70-\u0d78\u0f2a-\u0f33\u1369-\u137c\u17f0-\u17f9\u19da\u2070\u2074-\u2079\u2080-\u2089\u2150-\u215f\u2189\u2460-\u249b\u24ea-\u24ff\u2776-\u2793\u2cfd\u3192-\u3195\u3220-\u3229\u3248-\u324f\u3251-\u325f\u3280-\u3289\u32b1-\u32bf\ua830-\ua835\U00010107-\U00010133\U00010175-\U00010178\U0001018a-\U0001018b\U000102e1-\U000102fb\U00010320-\U00010323\U00010858-\U0001085f\U00010879-\U0001087f\U000108a7-\U000108af\U000108fb-\U000108ff\U00010916-\U0001091b\U000109bc-\U000109bd\U000109c0-\U000109cf\U000109d2-\U000109ff\U00010a40-\U00010a48\U00010a7d-\U00010a7e\U00010a9d-\U00010a9f\U00010aeb-\U00010aef\U00010b58-\U00010b5f\U00010b78-\U00010b7f\U00010ba9-\U00010baf\U00010cfa-\U00010cff\U00010e60-\U00010e7e\U00010f1d-\U00010f26\U00010f51-\U00010f54\U00011052-\U00011065\U000111e1-\U000111f4\U0001173a-\U0001173b\U000118ea-\U000118f2\U00011c5a-\U00011c6c\U00016b5b-\U00016b61\U00016e80-\U00016e96\U0001d2e0-\U0001d2f3\U0001d360-\U0001d378\U0001e8c7-\U0001e8cf\U0001ec71-\U0001ecab\U0001ecad-\U0001ecaf\U0001ecb1-\U0001ecb4\U0001f100-\U0001f10c' -Pc = u'_\u203f-\u2040\u2054\ufe33-\ufe34\ufe4d-\ufe4f\uff3f' +Pc = '_\u203f-\u2040\u2054\ufe33-\ufe34\ufe4d-\ufe4f\uff3f' -Pd = u'\\-\u058a\u05be\u1400\u1806\u2010-\u2015\u2e17\u2e1a\u2e3a-\u2e3b\u2e40\u301c\u3030\u30a0\ufe31-\ufe32\ufe58\ufe63\uff0d' +Pd = '\\-\u058a\u05be\u1400\u1806\u2010-\u2015\u2e17\u2e1a\u2e3a-\u2e3b\u2e40\u301c\u3030\u30a0\ufe31-\ufe32\ufe58\ufe63\uff0d' -Pe = u')\\]}\u0f3b\u0f3d\u169c\u2046\u207e\u208e\u2309\u230b\u232a\u2769\u276b\u276d\u276f\u2771\u2773\u2775\u27c6\u27e7\u27e9\u27eb\u27ed\u27ef\u2984\u2986\u2988\u298a\u298c\u298e\u2990\u2992\u2994\u2996\u2998\u29d9\u29db\u29fd\u2e23\u2e25\u2e27\u2e29\u3009\u300b\u300d\u300f\u3011\u3015\u3017\u3019\u301b\u301e-\u301f\ufd3e\ufe18\ufe36\ufe38\ufe3a\ufe3c\ufe3e\ufe40\ufe42\ufe44\ufe48\ufe5a\ufe5c\ufe5e\uff09\uff3d\uff5d\uff60\uff63' +Pe = ')\\]}\u0f3b\u0f3d\u169c\u2046\u207e\u208e\u2309\u230b\u232a\u2769\u276b\u276d\u276f\u2771\u2773\u2775\u27c6\u27e7\u27e9\u27eb\u27ed\u27ef\u2984\u2986\u2988\u298a\u298c\u298e\u2990\u2992\u2994\u2996\u2998\u29d9\u29db\u29fd\u2e23\u2e25\u2e27\u2e29\u3009\u300b\u300d\u300f\u3011\u3015\u3017\u3019\u301b\u301e-\u301f\ufd3e\ufe18\ufe36\ufe38\ufe3a\ufe3c\ufe3e\ufe40\ufe42\ufe44\ufe48\ufe5a\ufe5c\ufe5e\uff09\uff3d\uff5d\uff60\uff63' -Pf = u'\xbb\u2019\u201d\u203a\u2e03\u2e05\u2e0a\u2e0d\u2e1d\u2e21' +Pf = '\xbb\u2019\u201d\u203a\u2e03\u2e05\u2e0a\u2e0d\u2e1d\u2e21' -Pi = u'\xab\u2018\u201b-\u201c\u201f\u2039\u2e02\u2e04\u2e09\u2e0c\u2e1c\u2e20' +Pi = '\xab\u2018\u201b-\u201c\u201f\u2039\u2e02\u2e04\u2e09\u2e0c\u2e1c\u2e20' -Po = u"!-#%-'*,.-/:-;?-@\\\\\xa1\xa7\xb6-\xb7\xbf\u037e\u0387\u055a-\u055f\u0589\u05c0\u05c3\u05c6\u05f3-\u05f4\u0609-\u060a\u060c-\u060d\u061b\u061e-\u061f\u066a-\u066d\u06d4\u0700-\u070d\u07f7-\u07f9\u0830-\u083e\u085e\u0964-\u0965\u0970\u09fd\u0a76\u0af0\u0c84\u0df4\u0e4f\u0e5a-\u0e5b\u0f04-\u0f12\u0f14\u0f85\u0fd0-\u0fd4\u0fd9-\u0fda\u104a-\u104f\u10fb\u1360-\u1368\u166d-\u166e\u16eb-\u16ed\u1735-\u1736\u17d4-\u17d6\u17d8-\u17da\u1800-\u1805\u1807-\u180a\u1944-\u1945\u1a1e-\u1a1f\u1aa0-\u1aa6\u1aa8-\u1aad\u1b5a-\u1b60\u1bfc-\u1bff\u1c3b-\u1c3f\u1c7e-\u1c7f\u1cc0-\u1cc7\u1cd3\u2016-\u2017\u2020-\u2027\u2030-\u2038\u203b-\u203e\u2041-\u2043\u2047-\u2051\u2053\u2055-\u205e\u2cf9-\u2cfc\u2cfe-\u2cff\u2d70\u2e00-\u2e01\u2e06-\u2e08\u2e0b\u2e0e-\u2e16\u2e18-\u2e19\u2e1b\u2e1e-\u2e1f\u2e2a-\u2e2e\u2e30-\u2e39\u2e3c-\u2e3f\u2e41\u2e43-\u2e4e\u3001-\u3003\u303d\u30fb\ua4fe-\ua4ff\ua60d-\ua60f\ua673\ua67e\ua6f2-\ua6f7\ua874-\ua877\ua8ce-\ua8cf\ua8f8-\ua8fa\ua8fc\ua92e-\ua92f\ua95f\ua9c1-\ua9cd\ua9de-\ua9df\uaa5c-\uaa5f\uaade-\uaadf\uaaf0-\uaaf1\uabeb\ufe10-\ufe16\ufe19\ufe30\ufe45-\ufe46\ufe49-\ufe4c\ufe50-\ufe52\ufe54-\ufe57\ufe5f-\ufe61\ufe68\ufe6a-\ufe6b\uff01-\uff03\uff05-\uff07\uff0a\uff0c\uff0e-\uff0f\uff1a-\uff1b\uff1f-\uff20\uff3c\uff61\uff64-\uff65" +Po = "!-#%-'*,.-/:-;?-@\\\\\xa1\xa7\xb6-\xb7\xbf\u037e\u0387\u055a-\u055f\u0589\u05c0\u05c3\u05c6\u05f3-\u05f4\u0609-\u060a\u060c-\u060d\u061b\u061e-\u061f\u066a-\u066d\u06d4\u0700-\u070d\u07f7-\u07f9\u0830-\u083e\u085e\u0964-\u0965\u0970\u09fd\u0a76\u0af0\u0c84\u0df4\u0e4f\u0e5a-\u0e5b\u0f04-\u0f12\u0f14\u0f85\u0fd0-\u0fd4\u0fd9-\u0fda\u104a-\u104f\u10fb\u1360-\u1368\u166d-\u166e\u16eb-\u16ed\u1735-\u1736\u17d4-\u17d6\u17d8-\u17da\u1800-\u1805\u1807-\u180a\u1944-\u1945\u1a1e-\u1a1f\u1aa0-\u1aa6\u1aa8-\u1aad\u1b5a-\u1b60\u1bfc-\u1bff\u1c3b-\u1c3f\u1c7e-\u1c7f\u1cc0-\u1cc7\u1cd3\u2016-\u2017\u2020-\u2027\u2030-\u2038\u203b-\u203e\u2041-\u2043\u2047-\u2051\u2053\u2055-\u205e\u2cf9-\u2cfc\u2cfe-\u2cff\u2d70\u2e00-\u2e01\u2e06-\u2e08\u2e0b\u2e0e-\u2e16\u2e18-\u2e19\u2e1b\u2e1e-\u2e1f\u2e2a-\u2e2e\u2e30-\u2e39\u2e3c-\u2e3f\u2e41\u2e43-\u2e4e\u3001-\u3003\u303d\u30fb\ua4fe-\ua4ff\ua60d-\ua60f\ua673\ua67e\ua6f2-\ua6f7\ua874-\ua877\ua8ce-\ua8cf\ua8f8-\ua8fa\ua8fc\ua92e-\ua92f\ua95f\ua9c1-\ua9cd\ua9de-\ua9df\uaa5c-\uaa5f\uaade-\uaadf\uaaf0-\uaaf1\uabeb\ufe10-\ufe16\ufe19\ufe30\ufe45-\ufe46\ufe49-\ufe4c\ufe50-\ufe52\ufe54-\ufe57\ufe5f-\ufe61\ufe68\ufe6a-\ufe6b\uff01-\uff03\uff05-\uff07\uff0a\uff0c\uff0e-\uff0f\uff1a-\uff1b\uff1f-\uff20\uff3c\uff61\uff64-\uff65\U00010100-\U00010102\U0001039f\U000103d0\U0001056f\U00010857\U0001091f\U0001093f\U00010a50-\U00010a58\U00010a7f\U00010af0-\U00010af6\U00010b39-\U00010b3f\U00010b99-\U00010b9c\U00010f55-\U00010f59\U00011047-\U0001104d\U000110bb-\U000110bc\U000110be-\U000110c1\U00011140-\U00011143\U00011174-\U00011175\U000111c5-\U000111c8\U000111cd\U000111db\U000111dd-\U000111df\U00011238-\U0001123d\U000112a9\U0001144b-\U0001144f\U0001145b\U0001145d\U000114c6\U000115c1-\U000115d7\U00011641-\U00011643\U00011660-\U0001166c\U0001173c-\U0001173e\U0001183b\U00011a3f-\U00011a46\U00011a9a-\U00011a9c\U00011a9e-\U00011aa2\U00011c41-\U00011c45\U00011c70-\U00011c71\U00011ef7-\U00011ef8\U00012470-\U00012474\U00016a6e-\U00016a6f\U00016af5\U00016b37-\U00016b3b\U00016b44\U00016e97-\U00016e9a\U0001bc9f\U0001da87-\U0001da8b\U0001e95e-\U0001e95f" -Ps = u'(\\[{\u0f3a\u0f3c\u169b\u201a\u201e\u2045\u207d\u208d\u2308\u230a\u2329\u2768\u276a\u276c\u276e\u2770\u2772\u2774\u27c5\u27e6\u27e8\u27ea\u27ec\u27ee\u2983\u2985\u2987\u2989\u298b\u298d\u298f\u2991\u2993\u2995\u2997\u29d8\u29da\u29fc\u2e22\u2e24\u2e26\u2e28\u2e42\u3008\u300a\u300c\u300e\u3010\u3014\u3016\u3018\u301a\u301d\ufd3f\ufe17\ufe35\ufe37\ufe39\ufe3b\ufe3d\ufe3f\ufe41\ufe43\ufe47\ufe59\ufe5b\ufe5d\uff08\uff3b\uff5b\uff5f\uff62' +Ps = '(\\[{\u0f3a\u0f3c\u169b\u201a\u201e\u2045\u207d\u208d\u2308\u230a\u2329\u2768\u276a\u276c\u276e\u2770\u2772\u2774\u27c5\u27e6\u27e8\u27ea\u27ec\u27ee\u2983\u2985\u2987\u2989\u298b\u298d\u298f\u2991\u2993\u2995\u2997\u29d8\u29da\u29fc\u2e22\u2e24\u2e26\u2e28\u2e42\u3008\u300a\u300c\u300e\u3010\u3014\u3016\u3018\u301a\u301d\ufd3f\ufe17\ufe35\ufe37\ufe39\ufe3b\ufe3d\ufe3f\ufe41\ufe43\ufe47\ufe59\ufe5b\ufe5d\uff08\uff3b\uff5b\uff5f\uff62' -Sc = u'$\xa2-\xa5\u058f\u060b\u07fe-\u07ff\u09f2-\u09f3\u09fb\u0af1\u0bf9\u0e3f\u17db\u20a0-\u20bf\ua838\ufdfc\ufe69\uff04\uffe0-\uffe1\uffe5-\uffe6' +Sc = '$\xa2-\xa5\u058f\u060b\u07fe-\u07ff\u09f2-\u09f3\u09fb\u0af1\u0bf9\u0e3f\u17db\u20a0-\u20bf\ua838\ufdfc\ufe69\uff04\uffe0-\uffe1\uffe5-\uffe6\U0001ecb0' -Sk = u'\\^`\xa8\xaf\xb4\xb8\u02c2-\u02c5\u02d2-\u02df\u02e5-\u02eb\u02ed\u02ef-\u02ff\u0375\u0384-\u0385\u1fbd\u1fbf-\u1fc1\u1fcd-\u1fcf\u1fdd-\u1fdf\u1fed-\u1fef\u1ffd-\u1ffe\u309b-\u309c\ua700-\ua716\ua720-\ua721\ua789-\ua78a\uab5b\ufbb2-\ufbc1\uff3e\uff40\uffe3' +Sk = '\\^`\xa8\xaf\xb4\xb8\u02c2-\u02c5\u02d2-\u02df\u02e5-\u02eb\u02ed\u02ef-\u02ff\u0375\u0384-\u0385\u1fbd\u1fbf-\u1fc1\u1fcd-\u1fcf\u1fdd-\u1fdf\u1fed-\u1fef\u1ffd-\u1ffe\u309b-\u309c\ua700-\ua716\ua720-\ua721\ua789-\ua78a\uab5b\ufbb2-\ufbc1\uff3e\uff40\uffe3\U0001f3fb-\U0001f3ff' -Sm = u'+<->|~\xac\xb1\xd7\xf7\u03f6\u0606-\u0608\u2044\u2052\u207a-\u207c\u208a-\u208c\u2118\u2140-\u2144\u214b\u2190-\u2194\u219a-\u219b\u21a0\u21a3\u21a6\u21ae\u21ce-\u21cf\u21d2\u21d4\u21f4-\u22ff\u2320-\u2321\u237c\u239b-\u23b3\u23dc-\u23e1\u25b7\u25c1\u25f8-\u25ff\u266f\u27c0-\u27c4\u27c7-\u27e5\u27f0-\u27ff\u2900-\u2982\u2999-\u29d7\u29dc-\u29fb\u29fe-\u2aff\u2b30-\u2b44\u2b47-\u2b4c\ufb29\ufe62\ufe64-\ufe66\uff0b\uff1c-\uff1e\uff5c\uff5e\uffe2\uffe9-\uffec' +Sm = '+<->|~\xac\xb1\xd7\xf7\u03f6\u0606-\u0608\u2044\u2052\u207a-\u207c\u208a-\u208c\u2118\u2140-\u2144\u214b\u2190-\u2194\u219a-\u219b\u21a0\u21a3\u21a6\u21ae\u21ce-\u21cf\u21d2\u21d4\u21f4-\u22ff\u2320-\u2321\u237c\u239b-\u23b3\u23dc-\u23e1\u25b7\u25c1\u25f8-\u25ff\u266f\u27c0-\u27c4\u27c7-\u27e5\u27f0-\u27ff\u2900-\u2982\u2999-\u29d7\u29dc-\u29fb\u29fe-\u2aff\u2b30-\u2b44\u2b47-\u2b4c\ufb29\ufe62\ufe64-\ufe66\uff0b\uff1c-\uff1e\uff5c\uff5e\uffe2\uffe9-\uffec\U0001d6c1\U0001d6db\U0001d6fb\U0001d715\U0001d735\U0001d74f\U0001d76f\U0001d789\U0001d7a9\U0001d7c3\U0001eef0-\U0001eef1' -So = u'\xa6\xa9\xae\xb0\u0482\u058d-\u058e\u060e-\u060f\u06de\u06e9\u06fd-\u06fe\u07f6\u09fa\u0b70\u0bf3-\u0bf8\u0bfa\u0c7f\u0d4f\u0d79\u0f01-\u0f03\u0f13\u0f15-\u0f17\u0f1a-\u0f1f\u0f34\u0f36\u0f38\u0fbe-\u0fc5\u0fc7-\u0fcc\u0fce-\u0fcf\u0fd5-\u0fd8\u109e-\u109f\u1390-\u1399\u1940\u19de-\u19ff\u1b61-\u1b6a\u1b74-\u1b7c\u2100-\u2101\u2103-\u2106\u2108-\u2109\u2114\u2116-\u2117\u211e-\u2123\u2125\u2127\u2129\u212e\u213a-\u213b\u214a\u214c-\u214d\u214f\u218a-\u218b\u2195-\u2199\u219c-\u219f\u21a1-\u21a2\u21a4-\u21a5\u21a7-\u21ad\u21af-\u21cd\u21d0-\u21d1\u21d3\u21d5-\u21f3\u2300-\u2307\u230c-\u231f\u2322-\u2328\u232b-\u237b\u237d-\u239a\u23b4-\u23db\u23e2-\u2426\u2440-\u244a\u249c-\u24e9\u2500-\u25b6\u25b8-\u25c0\u25c2-\u25f7\u2600-\u266e\u2670-\u2767\u2794-\u27bf\u2800-\u28ff\u2b00-\u2b2f\u2b45-\u2b46\u2b4d-\u2b73\u2b76-\u2b95\u2b98-\u2bc8\u2bca-\u2bfe\u2ce5-\u2cea\u2e80-\u2e99\u2e9b-\u2ef3\u2f00-\u2fd5\u2ff0-\u2ffb\u3004\u3012-\u3013\u3020\u3036-\u3037\u303e-\u303f\u3190-\u3191\u3196-\u319f\u31c0-\u31e3\u3200-\u321e\u322a-\u3247\u3250\u3260-\u327f\u328a-\u32b0\u32c0-\u32fe\u3300-\u33ff\u4dc0-\u4dff\ua490-\ua4c6\ua828-\ua82b\ua836-\ua837\ua839\uaa77-\uaa79\ufdfd\uffe4\uffe8\uffed-\uffee\ufffc-\ufffd' +So = '\xa6\xa9\xae\xb0\u0482\u058d-\u058e\u060e-\u060f\u06de\u06e9\u06fd-\u06fe\u07f6\u09fa\u0b70\u0bf3-\u0bf8\u0bfa\u0c7f\u0d4f\u0d79\u0f01-\u0f03\u0f13\u0f15-\u0f17\u0f1a-\u0f1f\u0f34\u0f36\u0f38\u0fbe-\u0fc5\u0fc7-\u0fcc\u0fce-\u0fcf\u0fd5-\u0fd8\u109e-\u109f\u1390-\u1399\u1940\u19de-\u19ff\u1b61-\u1b6a\u1b74-\u1b7c\u2100-\u2101\u2103-\u2106\u2108-\u2109\u2114\u2116-\u2117\u211e-\u2123\u2125\u2127\u2129\u212e\u213a-\u213b\u214a\u214c-\u214d\u214f\u218a-\u218b\u2195-\u2199\u219c-\u219f\u21a1-\u21a2\u21a4-\u21a5\u21a7-\u21ad\u21af-\u21cd\u21d0-\u21d1\u21d3\u21d5-\u21f3\u2300-\u2307\u230c-\u231f\u2322-\u2328\u232b-\u237b\u237d-\u239a\u23b4-\u23db\u23e2-\u2426\u2440-\u244a\u249c-\u24e9\u2500-\u25b6\u25b8-\u25c0\u25c2-\u25f7\u2600-\u266e\u2670-\u2767\u2794-\u27bf\u2800-\u28ff\u2b00-\u2b2f\u2b45-\u2b46\u2b4d-\u2b73\u2b76-\u2b95\u2b98-\u2bc8\u2bca-\u2bfe\u2ce5-\u2cea\u2e80-\u2e99\u2e9b-\u2ef3\u2f00-\u2fd5\u2ff0-\u2ffb\u3004\u3012-\u3013\u3020\u3036-\u3037\u303e-\u303f\u3190-\u3191\u3196-\u319f\u31c0-\u31e3\u3200-\u321e\u322a-\u3247\u3250\u3260-\u327f\u328a-\u32b0\u32c0-\u32fe\u3300-\u33ff\u4dc0-\u4dff\ua490-\ua4c6\ua828-\ua82b\ua836-\ua837\ua839\uaa77-\uaa79\ufdfd\uffe4\uffe8\uffed-\uffee\ufffc-\ufffd\U00010137-\U0001013f\U00010179-\U00010189\U0001018c-\U0001018e\U00010190-\U0001019b\U000101a0\U000101d0-\U000101fc\U00010877-\U00010878\U00010ac8\U0001173f\U00016b3c-\U00016b3f\U00016b45\U0001bc9c\U0001d000-\U0001d0f5\U0001d100-\U0001d126\U0001d129-\U0001d164\U0001d16a-\U0001d16c\U0001d183-\U0001d184\U0001d18c-\U0001d1a9\U0001d1ae-\U0001d1e8\U0001d200-\U0001d241\U0001d245\U0001d300-\U0001d356\U0001d800-\U0001d9ff\U0001da37-\U0001da3a\U0001da6d-\U0001da74\U0001da76-\U0001da83\U0001da85-\U0001da86\U0001ecac\U0001f000-\U0001f02b\U0001f030-\U0001f093\U0001f0a0-\U0001f0ae\U0001f0b1-\U0001f0bf\U0001f0c1-\U0001f0cf\U0001f0d1-\U0001f0f5\U0001f110-\U0001f16b\U0001f170-\U0001f1ac\U0001f1e6-\U0001f202\U0001f210-\U0001f23b\U0001f240-\U0001f248\U0001f250-\U0001f251\U0001f260-\U0001f265\U0001f300-\U0001f3fa\U0001f400-\U0001f6d4\U0001f6e0-\U0001f6ec\U0001f6f0-\U0001f6f9\U0001f700-\U0001f773\U0001f780-\U0001f7d8\U0001f800-\U0001f80b\U0001f810-\U0001f847\U0001f850-\U0001f859\U0001f860-\U0001f887\U0001f890-\U0001f8ad\U0001f900-\U0001f90b\U0001f910-\U0001f93e\U0001f940-\U0001f970\U0001f973-\U0001f976\U0001f97a\U0001f97c-\U0001f9a2\U0001f9b0-\U0001f9b9\U0001f9c0-\U0001f9c2\U0001f9d0-\U0001f9ff\U0001fa60-\U0001fa6d' -Zl = u'\u2028' +Zl = '\u2028' -Zp = u'\u2029' +Zp = '\u2029' -Zs = u' \xa0\u1680\u2000-\u200a\u202f\u205f\u3000' +Zs = ' \xa0\u1680\u2000-\u200a\u202f\u205f\u3000' -xid_continue = u'0-9A-Z_a-z\xaa\xb5\xb7\xba\xc0-\xd6\xd8-\xf6\xf8-\u02c1\u02c6-\u02d1\u02e0-\u02e4\u02ec\u02ee\u0300-\u0374\u0376-\u0377\u037b-\u037d\u037f\u0386-\u038a\u038c\u038e-\u03a1\u03a3-\u03f5\u03f7-\u0481\u0483-\u0487\u048a-\u052f\u0531-\u0556\u0559\u0560-\u0588\u0591-\u05bd\u05bf\u05c1-\u05c2\u05c4-\u05c5\u05c7\u05d0-\u05ea\u05ef-\u05f2\u0610-\u061a\u0620-\u0669\u066e-\u06d3\u06d5-\u06dc\u06df-\u06e8\u06ea-\u06fc\u06ff\u0710-\u074a\u074d-\u07b1\u07c0-\u07f5\u07fa\u07fd\u0800-\u082d\u0840-\u085b\u0860-\u086a\u08a0-\u08b4\u08b6-\u08bd\u08d3-\u08e1\u08e3-\u0963\u0966-\u096f\u0971-\u0983\u0985-\u098c\u098f-\u0990\u0993-\u09a8\u09aa-\u09b0\u09b2\u09b6-\u09b9\u09bc-\u09c4\u09c7-\u09c8\u09cb-\u09ce\u09d7\u09dc-\u09dd\u09df-\u09e3\u09e6-\u09f1\u09fc\u09fe\u0a01-\u0a03\u0a05-\u0a0a\u0a0f-\u0a10\u0a13-\u0a28\u0a2a-\u0a30\u0a32-\u0a33\u0a35-\u0a36\u0a38-\u0a39\u0a3c\u0a3e-\u0a42\u0a47-\u0a48\u0a4b-\u0a4d\u0a51\u0a59-\u0a5c\u0a5e\u0a66-\u0a75\u0a81-\u0a83\u0a85-\u0a8d\u0a8f-\u0a91\u0a93-\u0aa8\u0aaa-\u0ab0\u0ab2-\u0ab3\u0ab5-\u0ab9\u0abc-\u0ac5\u0ac7-\u0ac9\u0acb-\u0acd\u0ad0\u0ae0-\u0ae3\u0ae6-\u0aef\u0af9-\u0aff\u0b01-\u0b03\u0b05-\u0b0c\u0b0f-\u0b10\u0b13-\u0b28\u0b2a-\u0b30\u0b32-\u0b33\u0b35-\u0b39\u0b3c-\u0b44\u0b47-\u0b48\u0b4b-\u0b4d\u0b56-\u0b57\u0b5c-\u0b5d\u0b5f-\u0b63\u0b66-\u0b6f\u0b71\u0b82-\u0b83\u0b85-\u0b8a\u0b8e-\u0b90\u0b92-\u0b95\u0b99-\u0b9a\u0b9c\u0b9e-\u0b9f\u0ba3-\u0ba4\u0ba8-\u0baa\u0bae-\u0bb9\u0bbe-\u0bc2\u0bc6-\u0bc8\u0bca-\u0bcd\u0bd0\u0bd7\u0be6-\u0bef\u0c00-\u0c0c\u0c0e-\u0c10\u0c12-\u0c28\u0c2a-\u0c39\u0c3d-\u0c44\u0c46-\u0c48\u0c4a-\u0c4d\u0c55-\u0c56\u0c58-\u0c5a\u0c60-\u0c63\u0c66-\u0c6f\u0c80-\u0c83\u0c85-\u0c8c\u0c8e-\u0c90\u0c92-\u0ca8\u0caa-\u0cb3\u0cb5-\u0cb9\u0cbc-\u0cc4\u0cc6-\u0cc8\u0cca-\u0ccd\u0cd5-\u0cd6\u0cde\u0ce0-\u0ce3\u0ce6-\u0cef\u0cf1-\u0cf2\u0d00-\u0d03\u0d05-\u0d0c\u0d0e-\u0d10\u0d12-\u0d44\u0d46-\u0d48\u0d4a-\u0d4e\u0d54-\u0d57\u0d5f-\u0d63\u0d66-\u0d6f\u0d7a-\u0d7f\u0d82-\u0d83\u0d85-\u0d96\u0d9a-\u0db1\u0db3-\u0dbb\u0dbd\u0dc0-\u0dc6\u0dca\u0dcf-\u0dd4\u0dd6\u0dd8-\u0ddf\u0de6-\u0def\u0df2-\u0df3\u0e01-\u0e3a\u0e40-\u0e4e\u0e50-\u0e59\u0e81-\u0e82\u0e84\u0e87-\u0e88\u0e8a\u0e8d\u0e94-\u0e97\u0e99-\u0e9f\u0ea1-\u0ea3\u0ea5\u0ea7\u0eaa-\u0eab\u0ead-\u0eb9\u0ebb-\u0ebd\u0ec0-\u0ec4\u0ec6\u0ec8-\u0ecd\u0ed0-\u0ed9\u0edc-\u0edf\u0f00\u0f18-\u0f19\u0f20-\u0f29\u0f35\u0f37\u0f39\u0f3e-\u0f47\u0f49-\u0f6c\u0f71-\u0f84\u0f86-\u0f97\u0f99-\u0fbc\u0fc6\u1000-\u1049\u1050-\u109d\u10a0-\u10c5\u10c7\u10cd\u10d0-\u10fa\u10fc-\u1248\u124a-\u124d\u1250-\u1256\u1258\u125a-\u125d\u1260-\u1288\u128a-\u128d\u1290-\u12b0\u12b2-\u12b5\u12b8-\u12be\u12c0\u12c2-\u12c5\u12c8-\u12d6\u12d8-\u1310\u1312-\u1315\u1318-\u135a\u135d-\u135f\u1369-\u1371\u1380-\u138f\u13a0-\u13f5\u13f8-\u13fd\u1401-\u166c\u166f-\u167f\u1681-\u169a\u16a0-\u16ea\u16ee-\u16f8\u1700-\u170c\u170e-\u1714\u1720-\u1734\u1740-\u1753\u1760-\u176c\u176e-\u1770\u1772-\u1773\u1780-\u17d3\u17d7\u17dc-\u17dd\u17e0-\u17e9\u180b-\u180d\u1810-\u1819\u1820-\u1878\u1880-\u18aa\u18b0-\u18f5\u1900-\u191e\u1920-\u192b\u1930-\u193b\u1946-\u196d\u1970-\u1974\u1980-\u19ab\u19b0-\u19c9\u19d0-\u19da\u1a00-\u1a1b\u1a20-\u1a5e\u1a60-\u1a7c\u1a7f-\u1a89\u1a90-\u1a99\u1aa7\u1ab0-\u1abd\u1b00-\u1b4b\u1b50-\u1b59\u1b6b-\u1b73\u1b80-\u1bf3\u1c00-\u1c37\u1c40-\u1c49\u1c4d-\u1c7d\u1c80-\u1c88\u1c90-\u1cba\u1cbd-\u1cbf\u1cd0-\u1cd2\u1cd4-\u1cf9\u1d00-\u1df9\u1dfb-\u1f15\u1f18-\u1f1d\u1f20-\u1f45\u1f48-\u1f4d\u1f50-\u1f57\u1f59\u1f5b\u1f5d\u1f5f-\u1f7d\u1f80-\u1fb4\u1fb6-\u1fbc\u1fbe\u1fc2-\u1fc4\u1fc6-\u1fcc\u1fd0-\u1fd3\u1fd6-\u1fdb\u1fe0-\u1fec\u1ff2-\u1ff4\u1ff6-\u1ffc\u203f-\u2040\u2054\u2071\u207f\u2090-\u209c\u20d0-\u20dc\u20e1\u20e5-\u20f0\u2102\u2107\u210a-\u2113\u2115\u2118-\u211d\u2124\u2126\u2128\u212a-\u2139\u213c-\u213f\u2145-\u2149\u214e\u2160-\u2188\u2c00-\u2c2e\u2c30-\u2c5e\u2c60-\u2ce4\u2ceb-\u2cf3\u2d00-\u2d25\u2d27\u2d2d\u2d30-\u2d67\u2d6f\u2d7f-\u2d96\u2da0-\u2da6\u2da8-\u2dae\u2db0-\u2db6\u2db8-\u2dbe\u2dc0-\u2dc6\u2dc8-\u2dce\u2dd0-\u2dd6\u2dd8-\u2dde\u2de0-\u2dff\u3005-\u3007\u3021-\u302f\u3031-\u3035\u3038-\u303c\u3041-\u3096\u3099-\u309a\u309d-\u309f\u30a1-\u30fa\u30fc-\u30ff\u3105-\u312f\u3131-\u318e\u31a0-\u31ba\u31f0-\u31ff\u3400-\u4db5\u4e00-\u9fef\ua000-\ua48c\ua4d0-\ua4fd\ua500-\ua60c\ua610-\ua62b\ua640-\ua66f\ua674-\ua67d\ua67f-\ua6f1\ua717-\ua71f\ua722-\ua788\ua78b-\ua7b9\ua7f7-\ua827\ua840-\ua873\ua880-\ua8c5\ua8d0-\ua8d9\ua8e0-\ua8f7\ua8fb\ua8fd-\ua92d\ua930-\ua953\ua960-\ua97c\ua980-\ua9c0\ua9cf-\ua9d9\ua9e0-\ua9fe\uaa00-\uaa36\uaa40-\uaa4d\uaa50-\uaa59\uaa60-\uaa76\uaa7a-\uaac2\uaadb-\uaadd\uaae0-\uaaef\uaaf2-\uaaf6\uab01-\uab06\uab09-\uab0e\uab11-\uab16\uab20-\uab26\uab28-\uab2e\uab30-\uab5a\uab5c-\uab65\uab70-\uabea\uabec-\uabed\uabf0-\uabf9\uac00-\ud7a3\ud7b0-\ud7c6\ud7cb-\ud7fb\uf900-\ufa6d\ufa70-\ufad9\ufb00-\ufb06\ufb13-\ufb17\ufb1d-\ufb28\ufb2a-\ufb36\ufb38-\ufb3c\ufb3e\ufb40-\ufb41\ufb43-\ufb44\ufb46-\ufbb1\ufbd3-\ufc5d\ufc64-\ufd3d\ufd50-\ufd8f\ufd92-\ufdc7\ufdf0-\ufdf9\ufe00-\ufe0f\ufe20-\ufe2f\ufe33-\ufe34\ufe4d-\ufe4f\ufe71\ufe73\ufe77\ufe79\ufe7b\ufe7d\ufe7f-\ufefc\uff10-\uff19\uff21-\uff3a\uff3f\uff41-\uff5a\uff66-\uffbe\uffc2-\uffc7\uffca-\uffcf\uffd2-\uffd7\uffda-\uffdc' +xid_continue = '0-9A-Z_a-z\xaa\xb5\xb7\xba\xc0-\xd6\xd8-\xf6\xf8-\u02c1\u02c6-\u02d1\u02e0-\u02e4\u02ec\u02ee\u0300-\u0374\u0376-\u0377\u037b-\u037d\u037f\u0386-\u038a\u038c\u038e-\u03a1\u03a3-\u03f5\u03f7-\u0481\u0483-\u0487\u048a-\u052f\u0531-\u0556\u0559\u0560-\u0588\u0591-\u05bd\u05bf\u05c1-\u05c2\u05c4-\u05c5\u05c7\u05d0-\u05ea\u05ef-\u05f2\u0610-\u061a\u0620-\u0669\u066e-\u06d3\u06d5-\u06dc\u06df-\u06e8\u06ea-\u06fc\u06ff\u0710-\u074a\u074d-\u07b1\u07c0-\u07f5\u07fa\u07fd\u0800-\u082d\u0840-\u085b\u0860-\u086a\u08a0-\u08b4\u08b6-\u08bd\u08d3-\u08e1\u08e3-\u0963\u0966-\u096f\u0971-\u0983\u0985-\u098c\u098f-\u0990\u0993-\u09a8\u09aa-\u09b0\u09b2\u09b6-\u09b9\u09bc-\u09c4\u09c7-\u09c8\u09cb-\u09ce\u09d7\u09dc-\u09dd\u09df-\u09e3\u09e6-\u09f1\u09fc\u09fe\u0a01-\u0a03\u0a05-\u0a0a\u0a0f-\u0a10\u0a13-\u0a28\u0a2a-\u0a30\u0a32-\u0a33\u0a35-\u0a36\u0a38-\u0a39\u0a3c\u0a3e-\u0a42\u0a47-\u0a48\u0a4b-\u0a4d\u0a51\u0a59-\u0a5c\u0a5e\u0a66-\u0a75\u0a81-\u0a83\u0a85-\u0a8d\u0a8f-\u0a91\u0a93-\u0aa8\u0aaa-\u0ab0\u0ab2-\u0ab3\u0ab5-\u0ab9\u0abc-\u0ac5\u0ac7-\u0ac9\u0acb-\u0acd\u0ad0\u0ae0-\u0ae3\u0ae6-\u0aef\u0af9-\u0aff\u0b01-\u0b03\u0b05-\u0b0c\u0b0f-\u0b10\u0b13-\u0b28\u0b2a-\u0b30\u0b32-\u0b33\u0b35-\u0b39\u0b3c-\u0b44\u0b47-\u0b48\u0b4b-\u0b4d\u0b56-\u0b57\u0b5c-\u0b5d\u0b5f-\u0b63\u0b66-\u0b6f\u0b71\u0b82-\u0b83\u0b85-\u0b8a\u0b8e-\u0b90\u0b92-\u0b95\u0b99-\u0b9a\u0b9c\u0b9e-\u0b9f\u0ba3-\u0ba4\u0ba8-\u0baa\u0bae-\u0bb9\u0bbe-\u0bc2\u0bc6-\u0bc8\u0bca-\u0bcd\u0bd0\u0bd7\u0be6-\u0bef\u0c00-\u0c0c\u0c0e-\u0c10\u0c12-\u0c28\u0c2a-\u0c39\u0c3d-\u0c44\u0c46-\u0c48\u0c4a-\u0c4d\u0c55-\u0c56\u0c58-\u0c5a\u0c60-\u0c63\u0c66-\u0c6f\u0c80-\u0c83\u0c85-\u0c8c\u0c8e-\u0c90\u0c92-\u0ca8\u0caa-\u0cb3\u0cb5-\u0cb9\u0cbc-\u0cc4\u0cc6-\u0cc8\u0cca-\u0ccd\u0cd5-\u0cd6\u0cde\u0ce0-\u0ce3\u0ce6-\u0cef\u0cf1-\u0cf2\u0d00-\u0d03\u0d05-\u0d0c\u0d0e-\u0d10\u0d12-\u0d44\u0d46-\u0d48\u0d4a-\u0d4e\u0d54-\u0d57\u0d5f-\u0d63\u0d66-\u0d6f\u0d7a-\u0d7f\u0d82-\u0d83\u0d85-\u0d96\u0d9a-\u0db1\u0db3-\u0dbb\u0dbd\u0dc0-\u0dc6\u0dca\u0dcf-\u0dd4\u0dd6\u0dd8-\u0ddf\u0de6-\u0def\u0df2-\u0df3\u0e01-\u0e3a\u0e40-\u0e4e\u0e50-\u0e59\u0e81-\u0e82\u0e84\u0e87-\u0e88\u0e8a\u0e8d\u0e94-\u0e97\u0e99-\u0e9f\u0ea1-\u0ea3\u0ea5\u0ea7\u0eaa-\u0eab\u0ead-\u0eb9\u0ebb-\u0ebd\u0ec0-\u0ec4\u0ec6\u0ec8-\u0ecd\u0ed0-\u0ed9\u0edc-\u0edf\u0f00\u0f18-\u0f19\u0f20-\u0f29\u0f35\u0f37\u0f39\u0f3e-\u0f47\u0f49-\u0f6c\u0f71-\u0f84\u0f86-\u0f97\u0f99-\u0fbc\u0fc6\u1000-\u1049\u1050-\u109d\u10a0-\u10c5\u10c7\u10cd\u10d0-\u10fa\u10fc-\u1248\u124a-\u124d\u1250-\u1256\u1258\u125a-\u125d\u1260-\u1288\u128a-\u128d\u1290-\u12b0\u12b2-\u12b5\u12b8-\u12be\u12c0\u12c2-\u12c5\u12c8-\u12d6\u12d8-\u1310\u1312-\u1315\u1318-\u135a\u135d-\u135f\u1369-\u1371\u1380-\u138f\u13a0-\u13f5\u13f8-\u13fd\u1401-\u166c\u166f-\u167f\u1681-\u169a\u16a0-\u16ea\u16ee-\u16f8\u1700-\u170c\u170e-\u1714\u1720-\u1734\u1740-\u1753\u1760-\u176c\u176e-\u1770\u1772-\u1773\u1780-\u17d3\u17d7\u17dc-\u17dd\u17e0-\u17e9\u180b-\u180d\u1810-\u1819\u1820-\u1878\u1880-\u18aa\u18b0-\u18f5\u1900-\u191e\u1920-\u192b\u1930-\u193b\u1946-\u196d\u1970-\u1974\u1980-\u19ab\u19b0-\u19c9\u19d0-\u19da\u1a00-\u1a1b\u1a20-\u1a5e\u1a60-\u1a7c\u1a7f-\u1a89\u1a90-\u1a99\u1aa7\u1ab0-\u1abd\u1b00-\u1b4b\u1b50-\u1b59\u1b6b-\u1b73\u1b80-\u1bf3\u1c00-\u1c37\u1c40-\u1c49\u1c4d-\u1c7d\u1c80-\u1c88\u1c90-\u1cba\u1cbd-\u1cbf\u1cd0-\u1cd2\u1cd4-\u1cf9\u1d00-\u1df9\u1dfb-\u1f15\u1f18-\u1f1d\u1f20-\u1f45\u1f48-\u1f4d\u1f50-\u1f57\u1f59\u1f5b\u1f5d\u1f5f-\u1f7d\u1f80-\u1fb4\u1fb6-\u1fbc\u1fbe\u1fc2-\u1fc4\u1fc6-\u1fcc\u1fd0-\u1fd3\u1fd6-\u1fdb\u1fe0-\u1fec\u1ff2-\u1ff4\u1ff6-\u1ffc\u203f-\u2040\u2054\u2071\u207f\u2090-\u209c\u20d0-\u20dc\u20e1\u20e5-\u20f0\u2102\u2107\u210a-\u2113\u2115\u2118-\u211d\u2124\u2126\u2128\u212a-\u2139\u213c-\u213f\u2145-\u2149\u214e\u2160-\u2188\u2c00-\u2c2e\u2c30-\u2c5e\u2c60-\u2ce4\u2ceb-\u2cf3\u2d00-\u2d25\u2d27\u2d2d\u2d30-\u2d67\u2d6f\u2d7f-\u2d96\u2da0-\u2da6\u2da8-\u2dae\u2db0-\u2db6\u2db8-\u2dbe\u2dc0-\u2dc6\u2dc8-\u2dce\u2dd0-\u2dd6\u2dd8-\u2dde\u2de0-\u2dff\u3005-\u3007\u3021-\u302f\u3031-\u3035\u3038-\u303c\u3041-\u3096\u3099-\u309a\u309d-\u309f\u30a1-\u30fa\u30fc-\u30ff\u3105-\u312f\u3131-\u318e\u31a0-\u31ba\u31f0-\u31ff\u3400-\u4db5\u4e00-\u9fef\ua000-\ua48c\ua4d0-\ua4fd\ua500-\ua60c\ua610-\ua62b\ua640-\ua66f\ua674-\ua67d\ua67f-\ua6f1\ua717-\ua71f\ua722-\ua788\ua78b-\ua7b9\ua7f7-\ua827\ua840-\ua873\ua880-\ua8c5\ua8d0-\ua8d9\ua8e0-\ua8f7\ua8fb\ua8fd-\ua92d\ua930-\ua953\ua960-\ua97c\ua980-\ua9c0\ua9cf-\ua9d9\ua9e0-\ua9fe\uaa00-\uaa36\uaa40-\uaa4d\uaa50-\uaa59\uaa60-\uaa76\uaa7a-\uaac2\uaadb-\uaadd\uaae0-\uaaef\uaaf2-\uaaf6\uab01-\uab06\uab09-\uab0e\uab11-\uab16\uab20-\uab26\uab28-\uab2e\uab30-\uab5a\uab5c-\uab65\uab70-\uabea\uabec-\uabed\uabf0-\uabf9\uac00-\ud7a3\ud7b0-\ud7c6\ud7cb-\ud7fb\uf900-\ufa6d\ufa70-\ufad9\ufb00-\ufb06\ufb13-\ufb17\ufb1d-\ufb28\ufb2a-\ufb36\ufb38-\ufb3c\ufb3e\ufb40-\ufb41\ufb43-\ufb44\ufb46-\ufbb1\ufbd3-\ufc5d\ufc64-\ufd3d\ufd50-\ufd8f\ufd92-\ufdc7\ufdf0-\ufdf9\ufe00-\ufe0f\ufe20-\ufe2f\ufe33-\ufe34\ufe4d-\ufe4f\ufe71\ufe73\ufe77\ufe79\ufe7b\ufe7d\ufe7f-\ufefc\uff10-\uff19\uff21-\uff3a\uff3f\uff41-\uff5a\uff66-\uffbe\uffc2-\uffc7\uffca-\uffcf\uffd2-\uffd7\uffda-\uffdc\U00010000-\U0001000b\U0001000d-\U00010026\U00010028-\U0001003a\U0001003c-\U0001003d\U0001003f-\U0001004d\U00010050-\U0001005d\U00010080-\U000100fa\U00010140-\U00010174\U000101fd\U00010280-\U0001029c\U000102a0-\U000102d0\U000102e0\U00010300-\U0001031f\U0001032d-\U0001034a\U00010350-\U0001037a\U00010380-\U0001039d\U000103a0-\U000103c3\U000103c8-\U000103cf\U000103d1-\U000103d5\U00010400-\U0001049d\U000104a0-\U000104a9\U000104b0-\U000104d3\U000104d8-\U000104fb\U00010500-\U00010527\U00010530-\U00010563\U00010600-\U00010736\U00010740-\U00010755\U00010760-\U00010767\U00010800-\U00010805\U00010808\U0001080a-\U00010835\U00010837-\U00010838\U0001083c\U0001083f-\U00010855\U00010860-\U00010876\U00010880-\U0001089e\U000108e0-\U000108f2\U000108f4-\U000108f5\U00010900-\U00010915\U00010920-\U00010939\U00010980-\U000109b7\U000109be-\U000109bf\U00010a00-\U00010a03\U00010a05-\U00010a06\U00010a0c-\U00010a13\U00010a15-\U00010a17\U00010a19-\U00010a35\U00010a38-\U00010a3a\U00010a3f\U00010a60-\U00010a7c\U00010a80-\U00010a9c\U00010ac0-\U00010ac7\U00010ac9-\U00010ae6\U00010b00-\U00010b35\U00010b40-\U00010b55\U00010b60-\U00010b72\U00010b80-\U00010b91\U00010c00-\U00010c48\U00010c80-\U00010cb2\U00010cc0-\U00010cf2\U00010d00-\U00010d27\U00010d30-\U00010d39\U00010f00-\U00010f1c\U00010f27\U00010f30-\U00010f50\U00011000-\U00011046\U00011066-\U0001106f\U0001107f-\U000110ba\U000110d0-\U000110e8\U000110f0-\U000110f9\U00011100-\U00011134\U00011136-\U0001113f\U00011144-\U00011146\U00011150-\U00011173\U00011176\U00011180-\U000111c4\U000111c9-\U000111cc\U000111d0-\U000111da\U000111dc\U00011200-\U00011211\U00011213-\U00011237\U0001123e\U00011280-\U00011286\U00011288\U0001128a-\U0001128d\U0001128f-\U0001129d\U0001129f-\U000112a8\U000112b0-\U000112ea\U000112f0-\U000112f9\U00011300-\U00011303\U00011305-\U0001130c\U0001130f-\U00011310\U00011313-\U00011328\U0001132a-\U00011330\U00011332-\U00011333\U00011335-\U00011339\U0001133b-\U00011344\U00011347-\U00011348\U0001134b-\U0001134d\U00011350\U00011357\U0001135d-\U00011363\U00011366-\U0001136c\U00011370-\U00011374\U00011400-\U0001144a\U00011450-\U00011459\U0001145e\U00011480-\U000114c5\U000114c7\U000114d0-\U000114d9\U00011580-\U000115b5\U000115b8-\U000115c0\U000115d8-\U000115dd\U00011600-\U00011640\U00011644\U00011650-\U00011659\U00011680-\U000116b7\U000116c0-\U000116c9\U00011700-\U0001171a\U0001171d-\U0001172b\U00011730-\U00011739\U00011800-\U0001183a\U000118a0-\U000118e9\U000118ff\U00011a00-\U00011a3e\U00011a47\U00011a50-\U00011a83\U00011a86-\U00011a99\U00011a9d\U00011ac0-\U00011af8\U00011c00-\U00011c08\U00011c0a-\U00011c36\U00011c38-\U00011c40\U00011c50-\U00011c59\U00011c72-\U00011c8f\U00011c92-\U00011ca7\U00011ca9-\U00011cb6\U00011d00-\U00011d06\U00011d08-\U00011d09\U00011d0b-\U00011d36\U00011d3a\U00011d3c-\U00011d3d\U00011d3f-\U00011d47\U00011d50-\U00011d59\U00011d60-\U00011d65\U00011d67-\U00011d68\U00011d6a-\U00011d8e\U00011d90-\U00011d91\U00011d93-\U00011d98\U00011da0-\U00011da9\U00011ee0-\U00011ef6\U00012000-\U00012399\U00012400-\U0001246e\U00012480-\U00012543\U00013000-\U0001342e\U00014400-\U00014646\U00016800-\U00016a38\U00016a40-\U00016a5e\U00016a60-\U00016a69\U00016ad0-\U00016aed\U00016af0-\U00016af4\U00016b00-\U00016b36\U00016b40-\U00016b43\U00016b50-\U00016b59\U00016b63-\U00016b77\U00016b7d-\U00016b8f\U00016e40-\U00016e7f\U00016f00-\U00016f44\U00016f50-\U00016f7e\U00016f8f-\U00016f9f\U00016fe0-\U00016fe1\U00017000-\U000187f1\U00018800-\U00018af2\U0001b000-\U0001b11e\U0001b170-\U0001b2fb\U0001bc00-\U0001bc6a\U0001bc70-\U0001bc7c\U0001bc80-\U0001bc88\U0001bc90-\U0001bc99\U0001bc9d-\U0001bc9e\U0001d165-\U0001d169\U0001d16d-\U0001d172\U0001d17b-\U0001d182\U0001d185-\U0001d18b\U0001d1aa-\U0001d1ad\U0001d242-\U0001d244\U0001d400-\U0001d454\U0001d456-\U0001d49c\U0001d49e-\U0001d49f\U0001d4a2\U0001d4a5-\U0001d4a6\U0001d4a9-\U0001d4ac\U0001d4ae-\U0001d4b9\U0001d4bb\U0001d4bd-\U0001d4c3\U0001d4c5-\U0001d505\U0001d507-\U0001d50a\U0001d50d-\U0001d514\U0001d516-\U0001d51c\U0001d51e-\U0001d539\U0001d53b-\U0001d53e\U0001d540-\U0001d544\U0001d546\U0001d54a-\U0001d550\U0001d552-\U0001d6a5\U0001d6a8-\U0001d6c0\U0001d6c2-\U0001d6da\U0001d6dc-\U0001d6fa\U0001d6fc-\U0001d714\U0001d716-\U0001d734\U0001d736-\U0001d74e\U0001d750-\U0001d76e\U0001d770-\U0001d788\U0001d78a-\U0001d7a8\U0001d7aa-\U0001d7c2\U0001d7c4-\U0001d7cb\U0001d7ce-\U0001d7ff\U0001da00-\U0001da36\U0001da3b-\U0001da6c\U0001da75\U0001da84\U0001da9b-\U0001da9f\U0001daa1-\U0001daaf\U0001e000-\U0001e006\U0001e008-\U0001e018\U0001e01b-\U0001e021\U0001e023-\U0001e024\U0001e026-\U0001e02a\U0001e800-\U0001e8c4\U0001e8d0-\U0001e8d6\U0001e900-\U0001e94a\U0001e950-\U0001e959\U0001ee00-\U0001ee03\U0001ee05-\U0001ee1f\U0001ee21-\U0001ee22\U0001ee24\U0001ee27\U0001ee29-\U0001ee32\U0001ee34-\U0001ee37\U0001ee39\U0001ee3b\U0001ee42\U0001ee47\U0001ee49\U0001ee4b\U0001ee4d-\U0001ee4f\U0001ee51-\U0001ee52\U0001ee54\U0001ee57\U0001ee59\U0001ee5b\U0001ee5d\U0001ee5f\U0001ee61-\U0001ee62\U0001ee64\U0001ee67-\U0001ee6a\U0001ee6c-\U0001ee72\U0001ee74-\U0001ee77\U0001ee79-\U0001ee7c\U0001ee7e\U0001ee80-\U0001ee89\U0001ee8b-\U0001ee9b\U0001eea1-\U0001eea3\U0001eea5-\U0001eea9\U0001eeab-\U0001eebb\U00020000-\U0002a6d6\U0002a700-\U0002b734\U0002b740-\U0002b81d\U0002b820-\U0002cea1\U0002ceb0-\U0002ebe0\U0002f800-\U0002fa1d\U000e0100-\U000e01ef' -xid_start = u'A-Z_a-z\xaa\xb5\xba\xc0-\xd6\xd8-\xf6\xf8-\u02c1\u02c6-\u02d1\u02e0-\u02e4\u02ec\u02ee\u0370-\u0374\u0376-\u0377\u037b-\u037d\u037f\u0386\u0388-\u038a\u038c\u038e-\u03a1\u03a3-\u03f5\u03f7-\u0481\u048a-\u052f\u0531-\u0556\u0559\u0560-\u0588\u05d0-\u05ea\u05ef-\u05f2\u0620-\u064a\u066e-\u066f\u0671-\u06d3\u06d5\u06e5-\u06e6\u06ee-\u06ef\u06fa-\u06fc\u06ff\u0710\u0712-\u072f\u074d-\u07a5\u07b1\u07ca-\u07ea\u07f4-\u07f5\u07fa\u0800-\u0815\u081a\u0824\u0828\u0840-\u0858\u0860-\u086a\u08a0-\u08b4\u08b6-\u08bd\u0904-\u0939\u093d\u0950\u0958-\u0961\u0971-\u0980\u0985-\u098c\u098f-\u0990\u0993-\u09a8\u09aa-\u09b0\u09b2\u09b6-\u09b9\u09bd\u09ce\u09dc-\u09dd\u09df-\u09e1\u09f0-\u09f1\u09fc\u0a05-\u0a0a\u0a0f-\u0a10\u0a13-\u0a28\u0a2a-\u0a30\u0a32-\u0a33\u0a35-\u0a36\u0a38-\u0a39\u0a59-\u0a5c\u0a5e\u0a72-\u0a74\u0a85-\u0a8d\u0a8f-\u0a91\u0a93-\u0aa8\u0aaa-\u0ab0\u0ab2-\u0ab3\u0ab5-\u0ab9\u0abd\u0ad0\u0ae0-\u0ae1\u0af9\u0b05-\u0b0c\u0b0f-\u0b10\u0b13-\u0b28\u0b2a-\u0b30\u0b32-\u0b33\u0b35-\u0b39\u0b3d\u0b5c-\u0b5d\u0b5f-\u0b61\u0b71\u0b83\u0b85-\u0b8a\u0b8e-\u0b90\u0b92-\u0b95\u0b99-\u0b9a\u0b9c\u0b9e-\u0b9f\u0ba3-\u0ba4\u0ba8-\u0baa\u0bae-\u0bb9\u0bd0\u0c05-\u0c0c\u0c0e-\u0c10\u0c12-\u0c28\u0c2a-\u0c39\u0c3d\u0c58-\u0c5a\u0c60-\u0c61\u0c80\u0c85-\u0c8c\u0c8e-\u0c90\u0c92-\u0ca8\u0caa-\u0cb3\u0cb5-\u0cb9\u0cbd\u0cde\u0ce0-\u0ce1\u0cf1-\u0cf2\u0d05-\u0d0c\u0d0e-\u0d10\u0d12-\u0d3a\u0d3d\u0d4e\u0d54-\u0d56\u0d5f-\u0d61\u0d7a-\u0d7f\u0d85-\u0d96\u0d9a-\u0db1\u0db3-\u0dbb\u0dbd\u0dc0-\u0dc6\u0e01-\u0e30\u0e32\u0e40-\u0e46\u0e81-\u0e82\u0e84\u0e87-\u0e88\u0e8a\u0e8d\u0e94-\u0e97\u0e99-\u0e9f\u0ea1-\u0ea3\u0ea5\u0ea7\u0eaa-\u0eab\u0ead-\u0eb0\u0eb2\u0ebd\u0ec0-\u0ec4\u0ec6\u0edc-\u0edf\u0f00\u0f40-\u0f47\u0f49-\u0f6c\u0f88-\u0f8c\u1000-\u102a\u103f\u1050-\u1055\u105a-\u105d\u1061\u1065-\u1066\u106e-\u1070\u1075-\u1081\u108e\u10a0-\u10c5\u10c7\u10cd\u10d0-\u10fa\u10fc-\u1248\u124a-\u124d\u1250-\u1256\u1258\u125a-\u125d\u1260-\u1288\u128a-\u128d\u1290-\u12b0\u12b2-\u12b5\u12b8-\u12be\u12c0\u12c2-\u12c5\u12c8-\u12d6\u12d8-\u1310\u1312-\u1315\u1318-\u135a\u1380-\u138f\u13a0-\u13f5\u13f8-\u13fd\u1401-\u166c\u166f-\u167f\u1681-\u169a\u16a0-\u16ea\u16ee-\u16f8\u1700-\u170c\u170e-\u1711\u1720-\u1731\u1740-\u1751\u1760-\u176c\u176e-\u1770\u1780-\u17b3\u17d7\u17dc\u1820-\u1878\u1880-\u18a8\u18aa\u18b0-\u18f5\u1900-\u191e\u1950-\u196d\u1970-\u1974\u1980-\u19ab\u19b0-\u19c9\u1a00-\u1a16\u1a20-\u1a54\u1aa7\u1b05-\u1b33\u1b45-\u1b4b\u1b83-\u1ba0\u1bae-\u1baf\u1bba-\u1be5\u1c00-\u1c23\u1c4d-\u1c4f\u1c5a-\u1c7d\u1c80-\u1c88\u1c90-\u1cba\u1cbd-\u1cbf\u1ce9-\u1cec\u1cee-\u1cf1\u1cf5-\u1cf6\u1d00-\u1dbf\u1e00-\u1f15\u1f18-\u1f1d\u1f20-\u1f45\u1f48-\u1f4d\u1f50-\u1f57\u1f59\u1f5b\u1f5d\u1f5f-\u1f7d\u1f80-\u1fb4\u1fb6-\u1fbc\u1fbe\u1fc2-\u1fc4\u1fc6-\u1fcc\u1fd0-\u1fd3\u1fd6-\u1fdb\u1fe0-\u1fec\u1ff2-\u1ff4\u1ff6-\u1ffc\u2071\u207f\u2090-\u209c\u2102\u2107\u210a-\u2113\u2115\u2118-\u211d\u2124\u2126\u2128\u212a-\u2139\u213c-\u213f\u2145-\u2149\u214e\u2160-\u2188\u2c00-\u2c2e\u2c30-\u2c5e\u2c60-\u2ce4\u2ceb-\u2cee\u2cf2-\u2cf3\u2d00-\u2d25\u2d27\u2d2d\u2d30-\u2d67\u2d6f\u2d80-\u2d96\u2da0-\u2da6\u2da8-\u2dae\u2db0-\u2db6\u2db8-\u2dbe\u2dc0-\u2dc6\u2dc8-\u2dce\u2dd0-\u2dd6\u2dd8-\u2dde\u3005-\u3007\u3021-\u3029\u3031-\u3035\u3038-\u303c\u3041-\u3096\u309d-\u309f\u30a1-\u30fa\u30fc-\u30ff\u3105-\u312f\u3131-\u318e\u31a0-\u31ba\u31f0-\u31ff\u3400-\u4db5\u4e00-\u9fef\ua000-\ua48c\ua4d0-\ua4fd\ua500-\ua60c\ua610-\ua61f\ua62a-\ua62b\ua640-\ua66e\ua67f-\ua69d\ua6a0-\ua6ef\ua717-\ua71f\ua722-\ua788\ua78b-\ua7b9\ua7f7-\ua801\ua803-\ua805\ua807-\ua80a\ua80c-\ua822\ua840-\ua873\ua882-\ua8b3\ua8f2-\ua8f7\ua8fb\ua8fd-\ua8fe\ua90a-\ua925\ua930-\ua946\ua960-\ua97c\ua984-\ua9b2\ua9cf\ua9e0-\ua9e4\ua9e6-\ua9ef\ua9fa-\ua9fe\uaa00-\uaa28\uaa40-\uaa42\uaa44-\uaa4b\uaa60-\uaa76\uaa7a\uaa7e-\uaaaf\uaab1\uaab5-\uaab6\uaab9-\uaabd\uaac0\uaac2\uaadb-\uaadd\uaae0-\uaaea\uaaf2-\uaaf4\uab01-\uab06\uab09-\uab0e\uab11-\uab16\uab20-\uab26\uab28-\uab2e\uab30-\uab5a\uab5c-\uab65\uab70-\uabe2\uac00-\ud7a3\ud7b0-\ud7c6\ud7cb-\ud7fb\uf900-\ufa6d\ufa70-\ufad9\ufb00-\ufb06\ufb13-\ufb17\ufb1d\ufb1f-\ufb28\ufb2a-\ufb36\ufb38-\ufb3c\ufb3e\ufb40-\ufb41\ufb43-\ufb44\ufb46-\ufbb1\ufbd3-\ufc5d\ufc64-\ufd3d\ufd50-\ufd8f\ufd92-\ufdc7\ufdf0-\ufdf9\ufe71\ufe73\ufe77\ufe79\ufe7b\ufe7d\ufe7f-\ufefc\uff21-\uff3a\uff41-\uff5a\uff66-\uff9d\uffa0-\uffbe\uffc2-\uffc7\uffca-\uffcf\uffd2-\uffd7\uffda-\uffdc' - -if sys.maxunicode > 0xFFFF: - # non-BMP characters, use only on wide Unicode builds - Cf += u'\U000110bd\U000110cd\U0001bca0-\U0001bca3\U0001d173-\U0001d17a\U000e0001\U000e0020-\U000e007f' - - Cn += u'\U0001000c\U00010027\U0001003b\U0001003e\U0001004e-\U0001004f\U0001005e-\U0001007f\U000100fb-\U000100ff\U00010103-\U00010106\U00010134-\U00010136\U0001018f\U0001019c-\U0001019f\U000101a1-\U000101cf\U000101fe-\U0001027f\U0001029d-\U0001029f\U000102d1-\U000102df\U000102fc-\U000102ff\U00010324-\U0001032c\U0001034b-\U0001034f\U0001037b-\U0001037f\U0001039e\U000103c4-\U000103c7\U000103d6-\U000103ff\U0001049e-\U0001049f\U000104aa-\U000104af\U000104d4-\U000104d7\U000104fc-\U000104ff\U00010528-\U0001052f\U00010564-\U0001056e\U00010570-\U000105ff\U00010737-\U0001073f\U00010756-\U0001075f\U00010768-\U000107ff\U00010806-\U00010807\U00010809\U00010836\U00010839-\U0001083b\U0001083d-\U0001083e\U00010856\U0001089f-\U000108a6\U000108b0-\U000108df\U000108f3\U000108f6-\U000108fa\U0001091c-\U0001091e\U0001093a-\U0001093e\U00010940-\U0001097f\U000109b8-\U000109bb\U000109d0-\U000109d1\U00010a04\U00010a07-\U00010a0b\U00010a14\U00010a18\U00010a36-\U00010a37\U00010a3b-\U00010a3e\U00010a49-\U00010a4f\U00010a59-\U00010a5f\U00010aa0-\U00010abf\U00010ae7-\U00010aea\U00010af7-\U00010aff\U00010b36-\U00010b38\U00010b56-\U00010b57\U00010b73-\U00010b77\U00010b92-\U00010b98\U00010b9d-\U00010ba8\U00010bb0-\U00010bff\U00010c49-\U00010c7f\U00010cb3-\U00010cbf\U00010cf3-\U00010cf9\U00010d28-\U00010d2f\U00010d3a-\U00010e5f\U00010e7f-\U00010eff\U00010f28-\U00010f2f\U00010f5a-\U00010fff\U0001104e-\U00011051\U00011070-\U0001107e\U000110c2-\U000110cc\U000110ce-\U000110cf\U000110e9-\U000110ef\U000110fa-\U000110ff\U00011135\U00011147-\U0001114f\U00011177-\U0001117f\U000111ce-\U000111cf\U000111e0\U000111f5-\U000111ff\U00011212\U0001123f-\U0001127f\U00011287\U00011289\U0001128e\U0001129e\U000112aa-\U000112af\U000112eb-\U000112ef\U000112fa-\U000112ff\U00011304\U0001130d-\U0001130e\U00011311-\U00011312\U00011329\U00011331\U00011334\U0001133a\U00011345-\U00011346\U00011349-\U0001134a\U0001134e-\U0001134f\U00011351-\U00011356\U00011358-\U0001135c\U00011364-\U00011365\U0001136d-\U0001136f\U00011375-\U000113ff\U0001145a\U0001145c\U0001145f-\U0001147f\U000114c8-\U000114cf\U000114da-\U0001157f\U000115b6-\U000115b7\U000115de-\U000115ff\U00011645-\U0001164f\U0001165a-\U0001165f\U0001166d-\U0001167f\U000116b8-\U000116bf\U000116ca-\U000116ff\U0001171b-\U0001171c\U0001172c-\U0001172f\U00011740-\U000117ff\U0001183c-\U0001189f\U000118f3-\U000118fe\U00011900-\U000119ff\U00011a48-\U00011a4f\U00011a84-\U00011a85\U00011aa3-\U00011abf\U00011af9-\U00011bff\U00011c09\U00011c37\U00011c46-\U00011c4f\U00011c6d-\U00011c6f\U00011c90-\U00011c91\U00011ca8\U00011cb7-\U00011cff\U00011d07\U00011d0a\U00011d37-\U00011d39\U00011d3b\U00011d3e\U00011d48-\U00011d4f\U00011d5a-\U00011d5f\U00011d66\U00011d69\U00011d8f\U00011d92\U00011d99-\U00011d9f\U00011daa-\U00011edf\U00011ef9-\U00011fff\U0001239a-\U000123ff\U0001246f\U00012475-\U0001247f\U00012544-\U00012fff\U0001342f-\U000143ff\U00014647-\U000167ff\U00016a39-\U00016a3f\U00016a5f\U00016a6a-\U00016a6d\U00016a70-\U00016acf\U00016aee-\U00016aef\U00016af6-\U00016aff\U00016b46-\U00016b4f\U00016b5a\U00016b62\U00016b78-\U00016b7c\U00016b90-\U00016e3f\U00016e9b-\U00016eff\U00016f45-\U00016f4f\U00016f7f-\U00016f8e\U00016fa0-\U00016fdf\U00016fe2-\U00016fff\U000187f2-\U000187ff\U00018af3-\U0001afff\U0001b11f-\U0001b16f\U0001b2fc-\U0001bbff\U0001bc6b-\U0001bc6f\U0001bc7d-\U0001bc7f\U0001bc89-\U0001bc8f\U0001bc9a-\U0001bc9b\U0001bca4-\U0001cfff\U0001d0f6-\U0001d0ff\U0001d127-\U0001d128\U0001d1e9-\U0001d1ff\U0001d246-\U0001d2df\U0001d2f4-\U0001d2ff\U0001d357-\U0001d35f\U0001d379-\U0001d3ff\U0001d455\U0001d49d\U0001d4a0-\U0001d4a1\U0001d4a3-\U0001d4a4\U0001d4a7-\U0001d4a8\U0001d4ad\U0001d4ba\U0001d4bc\U0001d4c4\U0001d506\U0001d50b-\U0001d50c\U0001d515\U0001d51d\U0001d53a\U0001d53f\U0001d545\U0001d547-\U0001d549\U0001d551\U0001d6a6-\U0001d6a7\U0001d7cc-\U0001d7cd\U0001da8c-\U0001da9a\U0001daa0\U0001dab0-\U0001dfff\U0001e007\U0001e019-\U0001e01a\U0001e022\U0001e025\U0001e02b-\U0001e7ff\U0001e8c5-\U0001e8c6\U0001e8d7-\U0001e8ff\U0001e94b-\U0001e94f\U0001e95a-\U0001e95d\U0001e960-\U0001ec70\U0001ecb5-\U0001edff\U0001ee04\U0001ee20\U0001ee23\U0001ee25-\U0001ee26\U0001ee28\U0001ee33\U0001ee38\U0001ee3a\U0001ee3c-\U0001ee41\U0001ee43-\U0001ee46\U0001ee48\U0001ee4a\U0001ee4c\U0001ee50\U0001ee53\U0001ee55-\U0001ee56\U0001ee58\U0001ee5a\U0001ee5c\U0001ee5e\U0001ee60\U0001ee63\U0001ee65-\U0001ee66\U0001ee6b\U0001ee73\U0001ee78\U0001ee7d\U0001ee7f\U0001ee8a\U0001ee9c-\U0001eea0\U0001eea4\U0001eeaa\U0001eebc-\U0001eeef\U0001eef2-\U0001efff\U0001f02c-\U0001f02f\U0001f094-\U0001f09f\U0001f0af-\U0001f0b0\U0001f0c0\U0001f0d0\U0001f0f6-\U0001f0ff\U0001f10d-\U0001f10f\U0001f16c-\U0001f16f\U0001f1ad-\U0001f1e5\U0001f203-\U0001f20f\U0001f23c-\U0001f23f\U0001f249-\U0001f24f\U0001f252-\U0001f25f\U0001f266-\U0001f2ff\U0001f6d5-\U0001f6df\U0001f6ed-\U0001f6ef\U0001f6fa-\U0001f6ff\U0001f774-\U0001f77f\U0001f7d9-\U0001f7ff\U0001f80c-\U0001f80f\U0001f848-\U0001f84f\U0001f85a-\U0001f85f\U0001f888-\U0001f88f\U0001f8ae-\U0001f8ff\U0001f90c-\U0001f90f\U0001f93f\U0001f971-\U0001f972\U0001f977-\U0001f979\U0001f97b\U0001f9a3-\U0001f9af\U0001f9ba-\U0001f9bf\U0001f9c3-\U0001f9cf\U0001fa00-\U0001fa5f\U0001fa6e-\U0001ffff\U0002a6d7-\U0002a6ff\U0002b735-\U0002b73f\U0002b81e-\U0002b81f\U0002cea2-\U0002ceaf\U0002ebe1-\U0002f7ff\U0002fa1e-\U000e0000\U000e0002-\U000e001f\U000e0080-\U000e00ff\U000e01f0-\U000effff\U000ffffe-\U000fffff\U0010fffe-\U0010ffff' - - Co += u'\U000f0000-\U000ffffd\U00100000-\U0010fffd' - - Ll += u'\U00010428-\U0001044f\U000104d8-\U000104fb\U00010cc0-\U00010cf2\U000118c0-\U000118df\U00016e60-\U00016e7f\U0001d41a-\U0001d433\U0001d44e-\U0001d454\U0001d456-\U0001d467\U0001d482-\U0001d49b\U0001d4b6-\U0001d4b9\U0001d4bb\U0001d4bd-\U0001d4c3\U0001d4c5-\U0001d4cf\U0001d4ea-\U0001d503\U0001d51e-\U0001d537\U0001d552-\U0001d56b\U0001d586-\U0001d59f\U0001d5ba-\U0001d5d3\U0001d5ee-\U0001d607\U0001d622-\U0001d63b\U0001d656-\U0001d66f\U0001d68a-\U0001d6a5\U0001d6c2-\U0001d6da\U0001d6dc-\U0001d6e1\U0001d6fc-\U0001d714\U0001d716-\U0001d71b\U0001d736-\U0001d74e\U0001d750-\U0001d755\U0001d770-\U0001d788\U0001d78a-\U0001d78f\U0001d7aa-\U0001d7c2\U0001d7c4-\U0001d7c9\U0001d7cb\U0001e922-\U0001e943' - - Lm += u'\U00016b40-\U00016b43\U00016f93-\U00016f9f\U00016fe0-\U00016fe1' - - Lo += u'\U00010000-\U0001000b\U0001000d-\U00010026\U00010028-\U0001003a\U0001003c-\U0001003d\U0001003f-\U0001004d\U00010050-\U0001005d\U00010080-\U000100fa\U00010280-\U0001029c\U000102a0-\U000102d0\U00010300-\U0001031f\U0001032d-\U00010340\U00010342-\U00010349\U00010350-\U00010375\U00010380-\U0001039d\U000103a0-\U000103c3\U000103c8-\U000103cf\U00010450-\U0001049d\U00010500-\U00010527\U00010530-\U00010563\U00010600-\U00010736\U00010740-\U00010755\U00010760-\U00010767\U00010800-\U00010805\U00010808\U0001080a-\U00010835\U00010837-\U00010838\U0001083c\U0001083f-\U00010855\U00010860-\U00010876\U00010880-\U0001089e\U000108e0-\U000108f2\U000108f4-\U000108f5\U00010900-\U00010915\U00010920-\U00010939\U00010980-\U000109b7\U000109be-\U000109bf\U00010a00\U00010a10-\U00010a13\U00010a15-\U00010a17\U00010a19-\U00010a35\U00010a60-\U00010a7c\U00010a80-\U00010a9c\U00010ac0-\U00010ac7\U00010ac9-\U00010ae4\U00010b00-\U00010b35\U00010b40-\U00010b55\U00010b60-\U00010b72\U00010b80-\U00010b91\U00010c00-\U00010c48\U00010d00-\U00010d23\U00010f00-\U00010f1c\U00010f27\U00010f30-\U00010f45\U00011003-\U00011037\U00011083-\U000110af\U000110d0-\U000110e8\U00011103-\U00011126\U00011144\U00011150-\U00011172\U00011176\U00011183-\U000111b2\U000111c1-\U000111c4\U000111da\U000111dc\U00011200-\U00011211\U00011213-\U0001122b\U00011280-\U00011286\U00011288\U0001128a-\U0001128d\U0001128f-\U0001129d\U0001129f-\U000112a8\U000112b0-\U000112de\U00011305-\U0001130c\U0001130f-\U00011310\U00011313-\U00011328\U0001132a-\U00011330\U00011332-\U00011333\U00011335-\U00011339\U0001133d\U00011350\U0001135d-\U00011361\U00011400-\U00011434\U00011447-\U0001144a\U00011480-\U000114af\U000114c4-\U000114c5\U000114c7\U00011580-\U000115ae\U000115d8-\U000115db\U00011600-\U0001162f\U00011644\U00011680-\U000116aa\U00011700-\U0001171a\U00011800-\U0001182b\U000118ff\U00011a00\U00011a0b-\U00011a32\U00011a3a\U00011a50\U00011a5c-\U00011a83\U00011a86-\U00011a89\U00011a9d\U00011ac0-\U00011af8\U00011c00-\U00011c08\U00011c0a-\U00011c2e\U00011c40\U00011c72-\U00011c8f\U00011d00-\U00011d06\U00011d08-\U00011d09\U00011d0b-\U00011d30\U00011d46\U00011d60-\U00011d65\U00011d67-\U00011d68\U00011d6a-\U00011d89\U00011d98\U00011ee0-\U00011ef2\U00012000-\U00012399\U00012480-\U00012543\U00013000-\U0001342e\U00014400-\U00014646\U00016800-\U00016a38\U00016a40-\U00016a5e\U00016ad0-\U00016aed\U00016b00-\U00016b2f\U00016b63-\U00016b77\U00016b7d-\U00016b8f\U00016f00-\U00016f44\U00016f50\U00017000-\U000187f1\U00018800-\U00018af2\U0001b000-\U0001b11e\U0001b170-\U0001b2fb\U0001bc00-\U0001bc6a\U0001bc70-\U0001bc7c\U0001bc80-\U0001bc88\U0001bc90-\U0001bc99\U0001e800-\U0001e8c4\U0001ee00-\U0001ee03\U0001ee05-\U0001ee1f\U0001ee21-\U0001ee22\U0001ee24\U0001ee27\U0001ee29-\U0001ee32\U0001ee34-\U0001ee37\U0001ee39\U0001ee3b\U0001ee42\U0001ee47\U0001ee49\U0001ee4b\U0001ee4d-\U0001ee4f\U0001ee51-\U0001ee52\U0001ee54\U0001ee57\U0001ee59\U0001ee5b\U0001ee5d\U0001ee5f\U0001ee61-\U0001ee62\U0001ee64\U0001ee67-\U0001ee6a\U0001ee6c-\U0001ee72\U0001ee74-\U0001ee77\U0001ee79-\U0001ee7c\U0001ee7e\U0001ee80-\U0001ee89\U0001ee8b-\U0001ee9b\U0001eea1-\U0001eea3\U0001eea5-\U0001eea9\U0001eeab-\U0001eebb\U00020000-\U0002a6d6\U0002a700-\U0002b734\U0002b740-\U0002b81d\U0002b820-\U0002cea1\U0002ceb0-\U0002ebe0\U0002f800-\U0002fa1d' - - Lu += u'\U00010400-\U00010427\U000104b0-\U000104d3\U00010c80-\U00010cb2\U000118a0-\U000118bf\U00016e40-\U00016e5f\U0001d400-\U0001d419\U0001d434-\U0001d44d\U0001d468-\U0001d481\U0001d49c\U0001d49e-\U0001d49f\U0001d4a2\U0001d4a5-\U0001d4a6\U0001d4a9-\U0001d4ac\U0001d4ae-\U0001d4b5\U0001d4d0-\U0001d4e9\U0001d504-\U0001d505\U0001d507-\U0001d50a\U0001d50d-\U0001d514\U0001d516-\U0001d51c\U0001d538-\U0001d539\U0001d53b-\U0001d53e\U0001d540-\U0001d544\U0001d546\U0001d54a-\U0001d550\U0001d56c-\U0001d585\U0001d5a0-\U0001d5b9\U0001d5d4-\U0001d5ed\U0001d608-\U0001d621\U0001d63c-\U0001d655\U0001d670-\U0001d689\U0001d6a8-\U0001d6c0\U0001d6e2-\U0001d6fa\U0001d71c-\U0001d734\U0001d756-\U0001d76e\U0001d790-\U0001d7a8\U0001d7ca\U0001e900-\U0001e921' - - Mc += u'\U00011000\U00011002\U00011082\U000110b0-\U000110b2\U000110b7-\U000110b8\U0001112c\U00011145-\U00011146\U00011182\U000111b3-\U000111b5\U000111bf-\U000111c0\U0001122c-\U0001122e\U00011232-\U00011233\U00011235\U000112e0-\U000112e2\U00011302-\U00011303\U0001133e-\U0001133f\U00011341-\U00011344\U00011347-\U00011348\U0001134b-\U0001134d\U00011357\U00011362-\U00011363\U00011435-\U00011437\U00011440-\U00011441\U00011445\U000114b0-\U000114b2\U000114b9\U000114bb-\U000114be\U000114c1\U000115af-\U000115b1\U000115b8-\U000115bb\U000115be\U00011630-\U00011632\U0001163b-\U0001163c\U0001163e\U000116ac\U000116ae-\U000116af\U000116b6\U00011720-\U00011721\U00011726\U0001182c-\U0001182e\U00011838\U00011a39\U00011a57-\U00011a58\U00011a97\U00011c2f\U00011c3e\U00011ca9\U00011cb1\U00011cb4\U00011d8a-\U00011d8e\U00011d93-\U00011d94\U00011d96\U00011ef5-\U00011ef6\U00016f51-\U00016f7e\U0001d165-\U0001d166\U0001d16d-\U0001d172' - - Mn += u'\U000101fd\U000102e0\U00010376-\U0001037a\U00010a01-\U00010a03\U00010a05-\U00010a06\U00010a0c-\U00010a0f\U00010a38-\U00010a3a\U00010a3f\U00010ae5-\U00010ae6\U00010d24-\U00010d27\U00010f46-\U00010f50\U00011001\U00011038-\U00011046\U0001107f-\U00011081\U000110b3-\U000110b6\U000110b9-\U000110ba\U00011100-\U00011102\U00011127-\U0001112b\U0001112d-\U00011134\U00011173\U00011180-\U00011181\U000111b6-\U000111be\U000111c9-\U000111cc\U0001122f-\U00011231\U00011234\U00011236-\U00011237\U0001123e\U000112df\U000112e3-\U000112ea\U00011300-\U00011301\U0001133b-\U0001133c\U00011340\U00011366-\U0001136c\U00011370-\U00011374\U00011438-\U0001143f\U00011442-\U00011444\U00011446\U0001145e\U000114b3-\U000114b8\U000114ba\U000114bf-\U000114c0\U000114c2-\U000114c3\U000115b2-\U000115b5\U000115bc-\U000115bd\U000115bf-\U000115c0\U000115dc-\U000115dd\U00011633-\U0001163a\U0001163d\U0001163f-\U00011640\U000116ab\U000116ad\U000116b0-\U000116b5\U000116b7\U0001171d-\U0001171f\U00011722-\U00011725\U00011727-\U0001172b\U0001182f-\U00011837\U00011839-\U0001183a\U00011a01-\U00011a0a\U00011a33-\U00011a38\U00011a3b-\U00011a3e\U00011a47\U00011a51-\U00011a56\U00011a59-\U00011a5b\U00011a8a-\U00011a96\U00011a98-\U00011a99\U00011c30-\U00011c36\U00011c38-\U00011c3d\U00011c3f\U00011c92-\U00011ca7\U00011caa-\U00011cb0\U00011cb2-\U00011cb3\U00011cb5-\U00011cb6\U00011d31-\U00011d36\U00011d3a\U00011d3c-\U00011d3d\U00011d3f-\U00011d45\U00011d47\U00011d90-\U00011d91\U00011d95\U00011d97\U00011ef3-\U00011ef4\U00016af0-\U00016af4\U00016b30-\U00016b36\U00016f8f-\U00016f92\U0001bc9d-\U0001bc9e\U0001d167-\U0001d169\U0001d17b-\U0001d182\U0001d185-\U0001d18b\U0001d1aa-\U0001d1ad\U0001d242-\U0001d244\U0001da00-\U0001da36\U0001da3b-\U0001da6c\U0001da75\U0001da84\U0001da9b-\U0001da9f\U0001daa1-\U0001daaf\U0001e000-\U0001e006\U0001e008-\U0001e018\U0001e01b-\U0001e021\U0001e023-\U0001e024\U0001e026-\U0001e02a\U0001e8d0-\U0001e8d6\U0001e944-\U0001e94a\U000e0100-\U000e01ef' - - Nd += u'\U000104a0-\U000104a9\U00010d30-\U00010d39\U00011066-\U0001106f\U000110f0-\U000110f9\U00011136-\U0001113f\U000111d0-\U000111d9\U000112f0-\U000112f9\U00011450-\U00011459\U000114d0-\U000114d9\U00011650-\U00011659\U000116c0-\U000116c9\U00011730-\U00011739\U000118e0-\U000118e9\U00011c50-\U00011c59\U00011d50-\U00011d59\U00011da0-\U00011da9\U00016a60-\U00016a69\U00016b50-\U00016b59\U0001d7ce-\U0001d7ff\U0001e950-\U0001e959' - - Nl += u'\U00010140-\U00010174\U00010341\U0001034a\U000103d1-\U000103d5\U00012400-\U0001246e' - - No += u'\U00010107-\U00010133\U00010175-\U00010178\U0001018a-\U0001018b\U000102e1-\U000102fb\U00010320-\U00010323\U00010858-\U0001085f\U00010879-\U0001087f\U000108a7-\U000108af\U000108fb-\U000108ff\U00010916-\U0001091b\U000109bc-\U000109bd\U000109c0-\U000109cf\U000109d2-\U000109ff\U00010a40-\U00010a48\U00010a7d-\U00010a7e\U00010a9d-\U00010a9f\U00010aeb-\U00010aef\U00010b58-\U00010b5f\U00010b78-\U00010b7f\U00010ba9-\U00010baf\U00010cfa-\U00010cff\U00010e60-\U00010e7e\U00010f1d-\U00010f26\U00010f51-\U00010f54\U00011052-\U00011065\U000111e1-\U000111f4\U0001173a-\U0001173b\U000118ea-\U000118f2\U00011c5a-\U00011c6c\U00016b5b-\U00016b61\U00016e80-\U00016e96\U0001d2e0-\U0001d2f3\U0001d360-\U0001d378\U0001e8c7-\U0001e8cf\U0001ec71-\U0001ecab\U0001ecad-\U0001ecaf\U0001ecb1-\U0001ecb4\U0001f100-\U0001f10c' - - Po += u'\U00010100-\U00010102\U0001039f\U000103d0\U0001056f\U00010857\U0001091f\U0001093f\U00010a50-\U00010a58\U00010a7f\U00010af0-\U00010af6\U00010b39-\U00010b3f\U00010b99-\U00010b9c\U00010f55-\U00010f59\U00011047-\U0001104d\U000110bb-\U000110bc\U000110be-\U000110c1\U00011140-\U00011143\U00011174-\U00011175\U000111c5-\U000111c8\U000111cd\U000111db\U000111dd-\U000111df\U00011238-\U0001123d\U000112a9\U0001144b-\U0001144f\U0001145b\U0001145d\U000114c6\U000115c1-\U000115d7\U00011641-\U00011643\U00011660-\U0001166c\U0001173c-\U0001173e\U0001183b\U00011a3f-\U00011a46\U00011a9a-\U00011a9c\U00011a9e-\U00011aa2\U00011c41-\U00011c45\U00011c70-\U00011c71\U00011ef7-\U00011ef8\U00012470-\U00012474\U00016a6e-\U00016a6f\U00016af5\U00016b37-\U00016b3b\U00016b44\U00016e97-\U00016e9a\U0001bc9f\U0001da87-\U0001da8b\U0001e95e-\U0001e95f' - - Sc += u'\U0001ecb0' - - Sk += u'\U0001f3fb-\U0001f3ff' - - Sm += u'\U0001d6c1\U0001d6db\U0001d6fb\U0001d715\U0001d735\U0001d74f\U0001d76f\U0001d789\U0001d7a9\U0001d7c3\U0001eef0-\U0001eef1' - - So += u'\U00010137-\U0001013f\U00010179-\U00010189\U0001018c-\U0001018e\U00010190-\U0001019b\U000101a0\U000101d0-\U000101fc\U00010877-\U00010878\U00010ac8\U0001173f\U00016b3c-\U00016b3f\U00016b45\U0001bc9c\U0001d000-\U0001d0f5\U0001d100-\U0001d126\U0001d129-\U0001d164\U0001d16a-\U0001d16c\U0001d183-\U0001d184\U0001d18c-\U0001d1a9\U0001d1ae-\U0001d1e8\U0001d200-\U0001d241\U0001d245\U0001d300-\U0001d356\U0001d800-\U0001d9ff\U0001da37-\U0001da3a\U0001da6d-\U0001da74\U0001da76-\U0001da83\U0001da85-\U0001da86\U0001ecac\U0001f000-\U0001f02b\U0001f030-\U0001f093\U0001f0a0-\U0001f0ae\U0001f0b1-\U0001f0bf\U0001f0c1-\U0001f0cf\U0001f0d1-\U0001f0f5\U0001f110-\U0001f16b\U0001f170-\U0001f1ac\U0001f1e6-\U0001f202\U0001f210-\U0001f23b\U0001f240-\U0001f248\U0001f250-\U0001f251\U0001f260-\U0001f265\U0001f300-\U0001f3fa\U0001f400-\U0001f6d4\U0001f6e0-\U0001f6ec\U0001f6f0-\U0001f6f9\U0001f700-\U0001f773\U0001f780-\U0001f7d8\U0001f800-\U0001f80b\U0001f810-\U0001f847\U0001f850-\U0001f859\U0001f860-\U0001f887\U0001f890-\U0001f8ad\U0001f900-\U0001f90b\U0001f910-\U0001f93e\U0001f940-\U0001f970\U0001f973-\U0001f976\U0001f97a\U0001f97c-\U0001f9a2\U0001f9b0-\U0001f9b9\U0001f9c0-\U0001f9c2\U0001f9d0-\U0001f9ff\U0001fa60-\U0001fa6d' - - xid_continue += u'\U00010000-\U0001000b\U0001000d-\U00010026\U00010028-\U0001003a\U0001003c-\U0001003d\U0001003f-\U0001004d\U00010050-\U0001005d\U00010080-\U000100fa\U00010140-\U00010174\U000101fd\U00010280-\U0001029c\U000102a0-\U000102d0\U000102e0\U00010300-\U0001031f\U0001032d-\U0001034a\U00010350-\U0001037a\U00010380-\U0001039d\U000103a0-\U000103c3\U000103c8-\U000103cf\U000103d1-\U000103d5\U00010400-\U0001049d\U000104a0-\U000104a9\U000104b0-\U000104d3\U000104d8-\U000104fb\U00010500-\U00010527\U00010530-\U00010563\U00010600-\U00010736\U00010740-\U00010755\U00010760-\U00010767\U00010800-\U00010805\U00010808\U0001080a-\U00010835\U00010837-\U00010838\U0001083c\U0001083f-\U00010855\U00010860-\U00010876\U00010880-\U0001089e\U000108e0-\U000108f2\U000108f4-\U000108f5\U00010900-\U00010915\U00010920-\U00010939\U00010980-\U000109b7\U000109be-\U000109bf\U00010a00-\U00010a03\U00010a05-\U00010a06\U00010a0c-\U00010a13\U00010a15-\U00010a17\U00010a19-\U00010a35\U00010a38-\U00010a3a\U00010a3f\U00010a60-\U00010a7c\U00010a80-\U00010a9c\U00010ac0-\U00010ac7\U00010ac9-\U00010ae6\U00010b00-\U00010b35\U00010b40-\U00010b55\U00010b60-\U00010b72\U00010b80-\U00010b91\U00010c00-\U00010c48\U00010c80-\U00010cb2\U00010cc0-\U00010cf2\U00010d00-\U00010d27\U00010d30-\U00010d39\U00010f00-\U00010f1c\U00010f27\U00010f30-\U00010f50\U00011000-\U00011046\U00011066-\U0001106f\U0001107f-\U000110ba\U000110d0-\U000110e8\U000110f0-\U000110f9\U00011100-\U00011134\U00011136-\U0001113f\U00011144-\U00011146\U00011150-\U00011173\U00011176\U00011180-\U000111c4\U000111c9-\U000111cc\U000111d0-\U000111da\U000111dc\U00011200-\U00011211\U00011213-\U00011237\U0001123e\U00011280-\U00011286\U00011288\U0001128a-\U0001128d\U0001128f-\U0001129d\U0001129f-\U000112a8\U000112b0-\U000112ea\U000112f0-\U000112f9\U00011300-\U00011303\U00011305-\U0001130c\U0001130f-\U00011310\U00011313-\U00011328\U0001132a-\U00011330\U00011332-\U00011333\U00011335-\U00011339\U0001133b-\U00011344\U00011347-\U00011348\U0001134b-\U0001134d\U00011350\U00011357\U0001135d-\U00011363\U00011366-\U0001136c\U00011370-\U00011374\U00011400-\U0001144a\U00011450-\U00011459\U0001145e\U00011480-\U000114c5\U000114c7\U000114d0-\U000114d9\U00011580-\U000115b5\U000115b8-\U000115c0\U000115d8-\U000115dd\U00011600-\U00011640\U00011644\U00011650-\U00011659\U00011680-\U000116b7\U000116c0-\U000116c9\U00011700-\U0001171a\U0001171d-\U0001172b\U00011730-\U00011739\U00011800-\U0001183a\U000118a0-\U000118e9\U000118ff\U00011a00-\U00011a3e\U00011a47\U00011a50-\U00011a83\U00011a86-\U00011a99\U00011a9d\U00011ac0-\U00011af8\U00011c00-\U00011c08\U00011c0a-\U00011c36\U00011c38-\U00011c40\U00011c50-\U00011c59\U00011c72-\U00011c8f\U00011c92-\U00011ca7\U00011ca9-\U00011cb6\U00011d00-\U00011d06\U00011d08-\U00011d09\U00011d0b-\U00011d36\U00011d3a\U00011d3c-\U00011d3d\U00011d3f-\U00011d47\U00011d50-\U00011d59\U00011d60-\U00011d65\U00011d67-\U00011d68\U00011d6a-\U00011d8e\U00011d90-\U00011d91\U00011d93-\U00011d98\U00011da0-\U00011da9\U00011ee0-\U00011ef6\U00012000-\U00012399\U00012400-\U0001246e\U00012480-\U00012543\U00013000-\U0001342e\U00014400-\U00014646\U00016800-\U00016a38\U00016a40-\U00016a5e\U00016a60-\U00016a69\U00016ad0-\U00016aed\U00016af0-\U00016af4\U00016b00-\U00016b36\U00016b40-\U00016b43\U00016b50-\U00016b59\U00016b63-\U00016b77\U00016b7d-\U00016b8f\U00016e40-\U00016e7f\U00016f00-\U00016f44\U00016f50-\U00016f7e\U00016f8f-\U00016f9f\U00016fe0-\U00016fe1\U00017000-\U000187f1\U00018800-\U00018af2\U0001b000-\U0001b11e\U0001b170-\U0001b2fb\U0001bc00-\U0001bc6a\U0001bc70-\U0001bc7c\U0001bc80-\U0001bc88\U0001bc90-\U0001bc99\U0001bc9d-\U0001bc9e\U0001d165-\U0001d169\U0001d16d-\U0001d172\U0001d17b-\U0001d182\U0001d185-\U0001d18b\U0001d1aa-\U0001d1ad\U0001d242-\U0001d244\U0001d400-\U0001d454\U0001d456-\U0001d49c\U0001d49e-\U0001d49f\U0001d4a2\U0001d4a5-\U0001d4a6\U0001d4a9-\U0001d4ac\U0001d4ae-\U0001d4b9\U0001d4bb\U0001d4bd-\U0001d4c3\U0001d4c5-\U0001d505\U0001d507-\U0001d50a\U0001d50d-\U0001d514\U0001d516-\U0001d51c\U0001d51e-\U0001d539\U0001d53b-\U0001d53e\U0001d540-\U0001d544\U0001d546\U0001d54a-\U0001d550\U0001d552-\U0001d6a5\U0001d6a8-\U0001d6c0\U0001d6c2-\U0001d6da\U0001d6dc-\U0001d6fa\U0001d6fc-\U0001d714\U0001d716-\U0001d734\U0001d736-\U0001d74e\U0001d750-\U0001d76e\U0001d770-\U0001d788\U0001d78a-\U0001d7a8\U0001d7aa-\U0001d7c2\U0001d7c4-\U0001d7cb\U0001d7ce-\U0001d7ff\U0001da00-\U0001da36\U0001da3b-\U0001da6c\U0001da75\U0001da84\U0001da9b-\U0001da9f\U0001daa1-\U0001daaf\U0001e000-\U0001e006\U0001e008-\U0001e018\U0001e01b-\U0001e021\U0001e023-\U0001e024\U0001e026-\U0001e02a\U0001e800-\U0001e8c4\U0001e8d0-\U0001e8d6\U0001e900-\U0001e94a\U0001e950-\U0001e959\U0001ee00-\U0001ee03\U0001ee05-\U0001ee1f\U0001ee21-\U0001ee22\U0001ee24\U0001ee27\U0001ee29-\U0001ee32\U0001ee34-\U0001ee37\U0001ee39\U0001ee3b\U0001ee42\U0001ee47\U0001ee49\U0001ee4b\U0001ee4d-\U0001ee4f\U0001ee51-\U0001ee52\U0001ee54\U0001ee57\U0001ee59\U0001ee5b\U0001ee5d\U0001ee5f\U0001ee61-\U0001ee62\U0001ee64\U0001ee67-\U0001ee6a\U0001ee6c-\U0001ee72\U0001ee74-\U0001ee77\U0001ee79-\U0001ee7c\U0001ee7e\U0001ee80-\U0001ee89\U0001ee8b-\U0001ee9b\U0001eea1-\U0001eea3\U0001eea5-\U0001eea9\U0001eeab-\U0001eebb\U00020000-\U0002a6d6\U0002a700-\U0002b734\U0002b740-\U0002b81d\U0002b820-\U0002cea1\U0002ceb0-\U0002ebe0\U0002f800-\U0002fa1d\U000e0100-\U000e01ef' - - xid_start += u'\U00010000-\U0001000b\U0001000d-\U00010026\U00010028-\U0001003a\U0001003c-\U0001003d\U0001003f-\U0001004d\U00010050-\U0001005d\U00010080-\U000100fa\U00010140-\U00010174\U00010280-\U0001029c\U000102a0-\U000102d0\U00010300-\U0001031f\U0001032d-\U0001034a\U00010350-\U00010375\U00010380-\U0001039d\U000103a0-\U000103c3\U000103c8-\U000103cf\U000103d1-\U000103d5\U00010400-\U0001049d\U000104b0-\U000104d3\U000104d8-\U000104fb\U00010500-\U00010527\U00010530-\U00010563\U00010600-\U00010736\U00010740-\U00010755\U00010760-\U00010767\U00010800-\U00010805\U00010808\U0001080a-\U00010835\U00010837-\U00010838\U0001083c\U0001083f-\U00010855\U00010860-\U00010876\U00010880-\U0001089e\U000108e0-\U000108f2\U000108f4-\U000108f5\U00010900-\U00010915\U00010920-\U00010939\U00010980-\U000109b7\U000109be-\U000109bf\U00010a00\U00010a10-\U00010a13\U00010a15-\U00010a17\U00010a19-\U00010a35\U00010a60-\U00010a7c\U00010a80-\U00010a9c\U00010ac0-\U00010ac7\U00010ac9-\U00010ae4\U00010b00-\U00010b35\U00010b40-\U00010b55\U00010b60-\U00010b72\U00010b80-\U00010b91\U00010c00-\U00010c48\U00010c80-\U00010cb2\U00010cc0-\U00010cf2\U00010d00-\U00010d23\U00010f00-\U00010f1c\U00010f27\U00010f30-\U00010f45\U00011003-\U00011037\U00011083-\U000110af\U000110d0-\U000110e8\U00011103-\U00011126\U00011144\U00011150-\U00011172\U00011176\U00011183-\U000111b2\U000111c1-\U000111c4\U000111da\U000111dc\U00011200-\U00011211\U00011213-\U0001122b\U00011280-\U00011286\U00011288\U0001128a-\U0001128d\U0001128f-\U0001129d\U0001129f-\U000112a8\U000112b0-\U000112de\U00011305-\U0001130c\U0001130f-\U00011310\U00011313-\U00011328\U0001132a-\U00011330\U00011332-\U00011333\U00011335-\U00011339\U0001133d\U00011350\U0001135d-\U00011361\U00011400-\U00011434\U00011447-\U0001144a\U00011480-\U000114af\U000114c4-\U000114c5\U000114c7\U00011580-\U000115ae\U000115d8-\U000115db\U00011600-\U0001162f\U00011644\U00011680-\U000116aa\U00011700-\U0001171a\U00011800-\U0001182b\U000118a0-\U000118df\U000118ff\U00011a00\U00011a0b-\U00011a32\U00011a3a\U00011a50\U00011a5c-\U00011a83\U00011a86-\U00011a89\U00011a9d\U00011ac0-\U00011af8\U00011c00-\U00011c08\U00011c0a-\U00011c2e\U00011c40\U00011c72-\U00011c8f\U00011d00-\U00011d06\U00011d08-\U00011d09\U00011d0b-\U00011d30\U00011d46\U00011d60-\U00011d65\U00011d67-\U00011d68\U00011d6a-\U00011d89\U00011d98\U00011ee0-\U00011ef2\U00012000-\U00012399\U00012400-\U0001246e\U00012480-\U00012543\U00013000-\U0001342e\U00014400-\U00014646\U00016800-\U00016a38\U00016a40-\U00016a5e\U00016ad0-\U00016aed\U00016b00-\U00016b2f\U00016b40-\U00016b43\U00016b63-\U00016b77\U00016b7d-\U00016b8f\U00016e40-\U00016e7f\U00016f00-\U00016f44\U00016f50\U00016f93-\U00016f9f\U00016fe0-\U00016fe1\U00017000-\U000187f1\U00018800-\U00018af2\U0001b000-\U0001b11e\U0001b170-\U0001b2fb\U0001bc00-\U0001bc6a\U0001bc70-\U0001bc7c\U0001bc80-\U0001bc88\U0001bc90-\U0001bc99\U0001d400-\U0001d454\U0001d456-\U0001d49c\U0001d49e-\U0001d49f\U0001d4a2\U0001d4a5-\U0001d4a6\U0001d4a9-\U0001d4ac\U0001d4ae-\U0001d4b9\U0001d4bb\U0001d4bd-\U0001d4c3\U0001d4c5-\U0001d505\U0001d507-\U0001d50a\U0001d50d-\U0001d514\U0001d516-\U0001d51c\U0001d51e-\U0001d539\U0001d53b-\U0001d53e\U0001d540-\U0001d544\U0001d546\U0001d54a-\U0001d550\U0001d552-\U0001d6a5\U0001d6a8-\U0001d6c0\U0001d6c2-\U0001d6da\U0001d6dc-\U0001d6fa\U0001d6fc-\U0001d714\U0001d716-\U0001d734\U0001d736-\U0001d74e\U0001d750-\U0001d76e\U0001d770-\U0001d788\U0001d78a-\U0001d7a8\U0001d7aa-\U0001d7c2\U0001d7c4-\U0001d7cb\U0001e800-\U0001e8c4\U0001e900-\U0001e943\U0001ee00-\U0001ee03\U0001ee05-\U0001ee1f\U0001ee21-\U0001ee22\U0001ee24\U0001ee27\U0001ee29-\U0001ee32\U0001ee34-\U0001ee37\U0001ee39\U0001ee3b\U0001ee42\U0001ee47\U0001ee49\U0001ee4b\U0001ee4d-\U0001ee4f\U0001ee51-\U0001ee52\U0001ee54\U0001ee57\U0001ee59\U0001ee5b\U0001ee5d\U0001ee5f\U0001ee61-\U0001ee62\U0001ee64\U0001ee67-\U0001ee6a\U0001ee6c-\U0001ee72\U0001ee74-\U0001ee77\U0001ee79-\U0001ee7c\U0001ee7e\U0001ee80-\U0001ee89\U0001ee8b-\U0001ee9b\U0001eea1-\U0001eea3\U0001eea5-\U0001eea9\U0001eeab-\U0001eebb\U00020000-\U0002a6d6\U0002a700-\U0002b734\U0002b740-\U0002b81d\U0002b820-\U0002cea1\U0002ceb0-\U0002ebe0\U0002f800-\U0002fa1d' +xid_start = 'A-Z_a-z\xaa\xb5\xba\xc0-\xd6\xd8-\xf6\xf8-\u02c1\u02c6-\u02d1\u02e0-\u02e4\u02ec\u02ee\u0370-\u0374\u0376-\u0377\u037b-\u037d\u037f\u0386\u0388-\u038a\u038c\u038e-\u03a1\u03a3-\u03f5\u03f7-\u0481\u048a-\u052f\u0531-\u0556\u0559\u0560-\u0588\u05d0-\u05ea\u05ef-\u05f2\u0620-\u064a\u066e-\u066f\u0671-\u06d3\u06d5\u06e5-\u06e6\u06ee-\u06ef\u06fa-\u06fc\u06ff\u0710\u0712-\u072f\u074d-\u07a5\u07b1\u07ca-\u07ea\u07f4-\u07f5\u07fa\u0800-\u0815\u081a\u0824\u0828\u0840-\u0858\u0860-\u086a\u08a0-\u08b4\u08b6-\u08bd\u0904-\u0939\u093d\u0950\u0958-\u0961\u0971-\u0980\u0985-\u098c\u098f-\u0990\u0993-\u09a8\u09aa-\u09b0\u09b2\u09b6-\u09b9\u09bd\u09ce\u09dc-\u09dd\u09df-\u09e1\u09f0-\u09f1\u09fc\u0a05-\u0a0a\u0a0f-\u0a10\u0a13-\u0a28\u0a2a-\u0a30\u0a32-\u0a33\u0a35-\u0a36\u0a38-\u0a39\u0a59-\u0a5c\u0a5e\u0a72-\u0a74\u0a85-\u0a8d\u0a8f-\u0a91\u0a93-\u0aa8\u0aaa-\u0ab0\u0ab2-\u0ab3\u0ab5-\u0ab9\u0abd\u0ad0\u0ae0-\u0ae1\u0af9\u0b05-\u0b0c\u0b0f-\u0b10\u0b13-\u0b28\u0b2a-\u0b30\u0b32-\u0b33\u0b35-\u0b39\u0b3d\u0b5c-\u0b5d\u0b5f-\u0b61\u0b71\u0b83\u0b85-\u0b8a\u0b8e-\u0b90\u0b92-\u0b95\u0b99-\u0b9a\u0b9c\u0b9e-\u0b9f\u0ba3-\u0ba4\u0ba8-\u0baa\u0bae-\u0bb9\u0bd0\u0c05-\u0c0c\u0c0e-\u0c10\u0c12-\u0c28\u0c2a-\u0c39\u0c3d\u0c58-\u0c5a\u0c60-\u0c61\u0c80\u0c85-\u0c8c\u0c8e-\u0c90\u0c92-\u0ca8\u0caa-\u0cb3\u0cb5-\u0cb9\u0cbd\u0cde\u0ce0-\u0ce1\u0cf1-\u0cf2\u0d05-\u0d0c\u0d0e-\u0d10\u0d12-\u0d3a\u0d3d\u0d4e\u0d54-\u0d56\u0d5f-\u0d61\u0d7a-\u0d7f\u0d85-\u0d96\u0d9a-\u0db1\u0db3-\u0dbb\u0dbd\u0dc0-\u0dc6\u0e01-\u0e30\u0e32\u0e40-\u0e46\u0e81-\u0e82\u0e84\u0e87-\u0e88\u0e8a\u0e8d\u0e94-\u0e97\u0e99-\u0e9f\u0ea1-\u0ea3\u0ea5\u0ea7\u0eaa-\u0eab\u0ead-\u0eb0\u0eb2\u0ebd\u0ec0-\u0ec4\u0ec6\u0edc-\u0edf\u0f00\u0f40-\u0f47\u0f49-\u0f6c\u0f88-\u0f8c\u1000-\u102a\u103f\u1050-\u1055\u105a-\u105d\u1061\u1065-\u1066\u106e-\u1070\u1075-\u1081\u108e\u10a0-\u10c5\u10c7\u10cd\u10d0-\u10fa\u10fc-\u1248\u124a-\u124d\u1250-\u1256\u1258\u125a-\u125d\u1260-\u1288\u128a-\u128d\u1290-\u12b0\u12b2-\u12b5\u12b8-\u12be\u12c0\u12c2-\u12c5\u12c8-\u12d6\u12d8-\u1310\u1312-\u1315\u1318-\u135a\u1380-\u138f\u13a0-\u13f5\u13f8-\u13fd\u1401-\u166c\u166f-\u167f\u1681-\u169a\u16a0-\u16ea\u16ee-\u16f8\u1700-\u170c\u170e-\u1711\u1720-\u1731\u1740-\u1751\u1760-\u176c\u176e-\u1770\u1780-\u17b3\u17d7\u17dc\u1820-\u1878\u1880-\u18a8\u18aa\u18b0-\u18f5\u1900-\u191e\u1950-\u196d\u1970-\u1974\u1980-\u19ab\u19b0-\u19c9\u1a00-\u1a16\u1a20-\u1a54\u1aa7\u1b05-\u1b33\u1b45-\u1b4b\u1b83-\u1ba0\u1bae-\u1baf\u1bba-\u1be5\u1c00-\u1c23\u1c4d-\u1c4f\u1c5a-\u1c7d\u1c80-\u1c88\u1c90-\u1cba\u1cbd-\u1cbf\u1ce9-\u1cec\u1cee-\u1cf1\u1cf5-\u1cf6\u1d00-\u1dbf\u1e00-\u1f15\u1f18-\u1f1d\u1f20-\u1f45\u1f48-\u1f4d\u1f50-\u1f57\u1f59\u1f5b\u1f5d\u1f5f-\u1f7d\u1f80-\u1fb4\u1fb6-\u1fbc\u1fbe\u1fc2-\u1fc4\u1fc6-\u1fcc\u1fd0-\u1fd3\u1fd6-\u1fdb\u1fe0-\u1fec\u1ff2-\u1ff4\u1ff6-\u1ffc\u2071\u207f\u2090-\u209c\u2102\u2107\u210a-\u2113\u2115\u2118-\u211d\u2124\u2126\u2128\u212a-\u2139\u213c-\u213f\u2145-\u2149\u214e\u2160-\u2188\u2c00-\u2c2e\u2c30-\u2c5e\u2c60-\u2ce4\u2ceb-\u2cee\u2cf2-\u2cf3\u2d00-\u2d25\u2d27\u2d2d\u2d30-\u2d67\u2d6f\u2d80-\u2d96\u2da0-\u2da6\u2da8-\u2dae\u2db0-\u2db6\u2db8-\u2dbe\u2dc0-\u2dc6\u2dc8-\u2dce\u2dd0-\u2dd6\u2dd8-\u2dde\u3005-\u3007\u3021-\u3029\u3031-\u3035\u3038-\u303c\u3041-\u3096\u309d-\u309f\u30a1-\u30fa\u30fc-\u30ff\u3105-\u312f\u3131-\u318e\u31a0-\u31ba\u31f0-\u31ff\u3400-\u4db5\u4e00-\u9fef\ua000-\ua48c\ua4d0-\ua4fd\ua500-\ua60c\ua610-\ua61f\ua62a-\ua62b\ua640-\ua66e\ua67f-\ua69d\ua6a0-\ua6ef\ua717-\ua71f\ua722-\ua788\ua78b-\ua7b9\ua7f7-\ua801\ua803-\ua805\ua807-\ua80a\ua80c-\ua822\ua840-\ua873\ua882-\ua8b3\ua8f2-\ua8f7\ua8fb\ua8fd-\ua8fe\ua90a-\ua925\ua930-\ua946\ua960-\ua97c\ua984-\ua9b2\ua9cf\ua9e0-\ua9e4\ua9e6-\ua9ef\ua9fa-\ua9fe\uaa00-\uaa28\uaa40-\uaa42\uaa44-\uaa4b\uaa60-\uaa76\uaa7a\uaa7e-\uaaaf\uaab1\uaab5-\uaab6\uaab9-\uaabd\uaac0\uaac2\uaadb-\uaadd\uaae0-\uaaea\uaaf2-\uaaf4\uab01-\uab06\uab09-\uab0e\uab11-\uab16\uab20-\uab26\uab28-\uab2e\uab30-\uab5a\uab5c-\uab65\uab70-\uabe2\uac00-\ud7a3\ud7b0-\ud7c6\ud7cb-\ud7fb\uf900-\ufa6d\ufa70-\ufad9\ufb00-\ufb06\ufb13-\ufb17\ufb1d\ufb1f-\ufb28\ufb2a-\ufb36\ufb38-\ufb3c\ufb3e\ufb40-\ufb41\ufb43-\ufb44\ufb46-\ufbb1\ufbd3-\ufc5d\ufc64-\ufd3d\ufd50-\ufd8f\ufd92-\ufdc7\ufdf0-\ufdf9\ufe71\ufe73\ufe77\ufe79\ufe7b\ufe7d\ufe7f-\ufefc\uff21-\uff3a\uff41-\uff5a\uff66-\uff9d\uffa0-\uffbe\uffc2-\uffc7\uffca-\uffcf\uffd2-\uffd7\uffda-\uffdc\U00010000-\U0001000b\U0001000d-\U00010026\U00010028-\U0001003a\U0001003c-\U0001003d\U0001003f-\U0001004d\U00010050-\U0001005d\U00010080-\U000100fa\U00010140-\U00010174\U00010280-\U0001029c\U000102a0-\U000102d0\U00010300-\U0001031f\U0001032d-\U0001034a\U00010350-\U00010375\U00010380-\U0001039d\U000103a0-\U000103c3\U000103c8-\U000103cf\U000103d1-\U000103d5\U00010400-\U0001049d\U000104b0-\U000104d3\U000104d8-\U000104fb\U00010500-\U00010527\U00010530-\U00010563\U00010600-\U00010736\U00010740-\U00010755\U00010760-\U00010767\U00010800-\U00010805\U00010808\U0001080a-\U00010835\U00010837-\U00010838\U0001083c\U0001083f-\U00010855\U00010860-\U00010876\U00010880-\U0001089e\U000108e0-\U000108f2\U000108f4-\U000108f5\U00010900-\U00010915\U00010920-\U00010939\U00010980-\U000109b7\U000109be-\U000109bf\U00010a00\U00010a10-\U00010a13\U00010a15-\U00010a17\U00010a19-\U00010a35\U00010a60-\U00010a7c\U00010a80-\U00010a9c\U00010ac0-\U00010ac7\U00010ac9-\U00010ae4\U00010b00-\U00010b35\U00010b40-\U00010b55\U00010b60-\U00010b72\U00010b80-\U00010b91\U00010c00-\U00010c48\U00010c80-\U00010cb2\U00010cc0-\U00010cf2\U00010d00-\U00010d23\U00010f00-\U00010f1c\U00010f27\U00010f30-\U00010f45\U00011003-\U00011037\U00011083-\U000110af\U000110d0-\U000110e8\U00011103-\U00011126\U00011144\U00011150-\U00011172\U00011176\U00011183-\U000111b2\U000111c1-\U000111c4\U000111da\U000111dc\U00011200-\U00011211\U00011213-\U0001122b\U00011280-\U00011286\U00011288\U0001128a-\U0001128d\U0001128f-\U0001129d\U0001129f-\U000112a8\U000112b0-\U000112de\U00011305-\U0001130c\U0001130f-\U00011310\U00011313-\U00011328\U0001132a-\U00011330\U00011332-\U00011333\U00011335-\U00011339\U0001133d\U00011350\U0001135d-\U00011361\U00011400-\U00011434\U00011447-\U0001144a\U00011480-\U000114af\U000114c4-\U000114c5\U000114c7\U00011580-\U000115ae\U000115d8-\U000115db\U00011600-\U0001162f\U00011644\U00011680-\U000116aa\U00011700-\U0001171a\U00011800-\U0001182b\U000118a0-\U000118df\U000118ff\U00011a00\U00011a0b-\U00011a32\U00011a3a\U00011a50\U00011a5c-\U00011a83\U00011a86-\U00011a89\U00011a9d\U00011ac0-\U00011af8\U00011c00-\U00011c08\U00011c0a-\U00011c2e\U00011c40\U00011c72-\U00011c8f\U00011d00-\U00011d06\U00011d08-\U00011d09\U00011d0b-\U00011d30\U00011d46\U00011d60-\U00011d65\U00011d67-\U00011d68\U00011d6a-\U00011d89\U00011d98\U00011ee0-\U00011ef2\U00012000-\U00012399\U00012400-\U0001246e\U00012480-\U00012543\U00013000-\U0001342e\U00014400-\U00014646\U00016800-\U00016a38\U00016a40-\U00016a5e\U00016ad0-\U00016aed\U00016b00-\U00016b2f\U00016b40-\U00016b43\U00016b63-\U00016b77\U00016b7d-\U00016b8f\U00016e40-\U00016e7f\U00016f00-\U00016f44\U00016f50\U00016f93-\U00016f9f\U00016fe0-\U00016fe1\U00017000-\U000187f1\U00018800-\U00018af2\U0001b000-\U0001b11e\U0001b170-\U0001b2fb\U0001bc00-\U0001bc6a\U0001bc70-\U0001bc7c\U0001bc80-\U0001bc88\U0001bc90-\U0001bc99\U0001d400-\U0001d454\U0001d456-\U0001d49c\U0001d49e-\U0001d49f\U0001d4a2\U0001d4a5-\U0001d4a6\U0001d4a9-\U0001d4ac\U0001d4ae-\U0001d4b9\U0001d4bb\U0001d4bd-\U0001d4c3\U0001d4c5-\U0001d505\U0001d507-\U0001d50a\U0001d50d-\U0001d514\U0001d516-\U0001d51c\U0001d51e-\U0001d539\U0001d53b-\U0001d53e\U0001d540-\U0001d544\U0001d546\U0001d54a-\U0001d550\U0001d552-\U0001d6a5\U0001d6a8-\U0001d6c0\U0001d6c2-\U0001d6da\U0001d6dc-\U0001d6fa\U0001d6fc-\U0001d714\U0001d716-\U0001d734\U0001d736-\U0001d74e\U0001d750-\U0001d76e\U0001d770-\U0001d788\U0001d78a-\U0001d7a8\U0001d7aa-\U0001d7c2\U0001d7c4-\U0001d7cb\U0001e800-\U0001e8c4\U0001e900-\U0001e943\U0001ee00-\U0001ee03\U0001ee05-\U0001ee1f\U0001ee21-\U0001ee22\U0001ee24\U0001ee27\U0001ee29-\U0001ee32\U0001ee34-\U0001ee37\U0001ee39\U0001ee3b\U0001ee42\U0001ee47\U0001ee49\U0001ee4b\U0001ee4d-\U0001ee4f\U0001ee51-\U0001ee52\U0001ee54\U0001ee57\U0001ee59\U0001ee5b\U0001ee5d\U0001ee5f\U0001ee61-\U0001ee62\U0001ee64\U0001ee67-\U0001ee6a\U0001ee6c-\U0001ee72\U0001ee74-\U0001ee77\U0001ee79-\U0001ee7c\U0001ee7e\U0001ee80-\U0001ee89\U0001ee8b-\U0001ee9b\U0001eea1-\U0001eea3\U0001eea5-\U0001eea9\U0001eeab-\U0001eebb\U00020000-\U0002a6d6\U0002a700-\U0002b734\U0002b740-\U0002b81d\U0002b820-\U0002cea1\U0002ceb0-\U0002ebe0\U0002f800-\U0002fa1d' cats = ['Cc', 'Cf', 'Cn', 'Co', 'Cs', 'Ll', 'Lm', 'Lo', 'Lt', 'Lu', 'Mc', 'Me', 'Mn', 'Nd', 'Nl', 'No', 'Pc', 'Pd', 'Pe', 'Pf', 'Pi', 'Po', 'Ps', 'Sc', 'Sk', 'Sm', 'So', 'Zl', 'Zp', 'Zs'] # Generated from unidata 11.0.0 def combine(*args): - return u''.join(globals()[cat] for cat in args) + return ''.join(globals()[cat] for cat in args) def allexcept(*args): newcats = cats[:] for arg in args: newcats.remove(arg) - return u''.join(globals()[cat] for cat in newcats) + return ''.join(globals()[cat] for cat in newcats) def _handle_runs(char_list): # pragma: no cover @@ -150,14 +107,13 @@ def _handle_runs(char_list): # pragma: no cover if a == b: yield a else: - yield u'%s-%s' % (a, b) + yield '%s-%s' % (a, b) if __name__ == '__main__': # pragma: no cover import unicodedata - categories_bmp = {'xid_start': [], 'xid_continue': []} - categories_nonbmp = {'xid_start': [], 'xid_continue': []} + categories = {'xid_start': [], 'xid_continue': []} with open(__file__) as fp: content = fp.read() @@ -171,43 +127,26 @@ if __name__ == '__main__': # pragma: no cover if ord(c) == 0xdc00: # Hack to avoid combining this combining with the preceeding high # surrogate, 0xdbff, when doing a repr. - c = u'\\' + c + c = '\\' + c elif ord(c) in (0x2d, 0x5b, 0x5c, 0x5d, 0x5e): # Escape regex metachars. - c = u'\\' + c - cat_dic = categories_bmp if code < 0x10000 else categories_nonbmp - cat_dic.setdefault(cat, []).append(c) + c = '\\' + c + categories.setdefault(cat, []).append(c) # XID_START and XID_CONTINUE are special categories used for matching # identifiers in Python 3. if c.isidentifier(): - cat_dic['xid_start'].append(c) + categories['xid_start'].append(c) if ('a' + c).isidentifier(): - cat_dic['xid_continue'].append(c) + categories['xid_continue'].append(c) with open(__file__, 'w') as fp: fp.write(header) - for cat in sorted(categories_bmp): - val = u''.join(_handle_runs(categories_bmp[cat])) - if cat == 'Cs': - # Jython can't handle isolated surrogates - fp.write("""\ -try: - Cs = eval(r"u%s") -except UnicodeDecodeError: - Cs = '' # Jython can't handle isolated surrogates\n\n""" % ascii(val)) - else: - fp.write('%s = u%a\n\n' % (cat, val)) - - fp.write('if sys.maxunicode > 0xFFFF:\n') - fp.write(' # non-BMP characters, use only on wide Unicode builds\n') - for cat in sorted(categories_nonbmp): - # no special case for Cs needed, since there are no surrogates - # in the higher planes - val = u''.join(_handle_runs(categories_nonbmp[cat])) - fp.write(' %s += u%a\n\n' % (cat, val)) + for cat in sorted(categories): + val = ''.join(_handle_runs(categories[cat])) + fp.write('%s = %a\n\n' % (cat, val)) - cats = sorted(categories_bmp) + cats = sorted(categories) cats.remove('xid_start') cats.remove('xid_continue') fp.write('cats = %r\n\n' % cats) diff --git a/pygments/util.py b/pygments/util.py index febde921..3a79029f 100644 --- a/pygments/util.py +++ b/pygments/util.py @@ -205,55 +205,15 @@ def looks_like_xml(text): return rv -# Python narrow build compatibility - -def _surrogatepair(c): - # Given a unicode character code - # with length greater than 16 bits, - # return the two 16 bit surrogate pair. +def surrogatepair(c): + """Given a unicode character code with length greater than 16 bits, + return the two 16 bit surrogate pair. + """ # From example D28 of: # http://www.unicode.org/book/ch03.pdf return (0xd7c0 + (c >> 10), (0xdc00 + (c & 0x3ff))) -def unirange(a, b): - """Returns a regular expression string to match the given non-BMP range.""" - if b < a: - raise ValueError("Bad character range") - if a < 0x10000 or b < 0x10000: - raise ValueError("unirange is only defined for non-BMP ranges") - - if sys.maxunicode > 0xffff: - # wide build - return u'[%s-%s]' % (chr(a), chr(b)) - else: - # narrow build stores surrogates, and the 're' module handles them - # (incorrectly) as characters. Since there is still ordering among - # these characters, expand the range to one that it understands. Some - # background in http://bugs.python.org/issue3665 and - # http://bugs.python.org/issue12749 - # - # Additionally, the lower constants are using chr rather than - # literals because jython [which uses the wide path] can't load this - # file if they are literals. - ah, al = _surrogatepair(a) - bh, bl = _surrogatepair(b) - if ah == bh: - return u'(?:%s[%s-%s])' % (chr(ah), chr(al), chr(bl)) - else: - buf = [] - buf.append(u'%s[%s-%s]' % (chr(ah), chr(al), - ah == bh and chr(bl) or chr(0xdfff))) - if ah - bh > 1: - buf.append(u'[%s-%s][%s-%s]' % - chr(ah+1), chr(bh-1), chr(0xdc00), chr(0xdfff)) - if ah != bh: - buf.append(u'%s[%s-%s]' % - (chr(bh), chr(0xdc00), chr(bl))) - - return u'(?:' + u'|'.join(buf) + u')' - - def format_lines(var_name, seq, raw=False, indent_level=0): """Formats a sequence of strings for output.""" lines = [] diff --git a/scripts/check_sources.py b/scripts/check_sources.py index 4c0a27b5..0762c967 100755 --- a/scripts/check_sources.py +++ b/scripts/check_sources.py @@ -197,7 +197,7 @@ def main(argv): if not in_pygments_pkg and checker.only_pkg: continue for lno, msg in checker(fn, lines): - print(u"%s:%d: %s" % (fn, lno, msg), file=out) + print('%s:%d: %s' % (fn, lno, msg), file=out) num += 1 if verbose: print() diff --git a/scripts/debug_lexer.py b/scripts/debug_lexer.py index d00b6102..0f1dc0bd 100755 --- a/scripts/debug_lexer.py +++ b/scripts/debug_lexer.py @@ -86,7 +86,7 @@ class DebuggingRegexLexer(ExtendedRegexLexer): # at EOL, reset state to 'root' ctx.stack = ['root'] statetokens = tokendefs['root'] - yield ctx.pos, Text, u'\n' + yield ctx.pos, Text, '\n' ctx.pos += 1 continue yield ctx.pos, Error, text[ctx.pos] diff --git a/tests/test_asm.py b/tests/test_asm.py index d351ce30..23e9f344 100644 --- a/tests/test_asm.py +++ b/tests/test_asm.py @@ -62,7 +62,7 @@ def test_cpuid(lexer_nasm): # cpu id, but as a single token. See bug #1517 fragment = 'cpuid' expected = [ - (Token.Name.Function, u'cpuid'), - (Token.Text, u'\n'), + (Token.Name.Function, 'cpuid'), + (Token.Text, '\n'), ] assert expected == list(lexer_nasm.get_tokens(fragment)) diff --git a/tests/test_basic.py b/tests/test_basic.py index e3826b28..384be6ec 100644 --- a/tests/test_basic.py +++ b/tests/test_basic.py @@ -57,7 +57,7 @@ def test_can_lex_integer(lexer): def test_can_lex_names(lexer): - assert_are_tokens_of_type(lexer, u'thingy thingy123 _thingy _123', Name) + assert_are_tokens_of_type(lexer, 'thingy thingy123 _thingy _123', Name) def test_can_recover_after_unterminated_string(lexer): diff --git a/tests/test_basic_api.py b/tests/test_basic_api.py index 378ea5e4..bbfbb14e 100644 --- a/tests/test_basic_api.py +++ b/tests/test_basic_api.py @@ -173,20 +173,20 @@ def test_formatter_encodings(): # unicode output fmt = HtmlFormatter() - tokens = [(Text, u"ä")] + tokens = [(Text, "ä")] out = format(tokens, fmt) assert type(out) is str - assert u"ä" in out + assert "ä" in out # encoding option fmt = HtmlFormatter(encoding="latin1") - tokens = [(Text, u"ä")] - assert u"ä".encode("latin1") in format(tokens, fmt) + tokens = [(Text, "ä")] + assert "ä".encode("latin1") in format(tokens, fmt) # encoding and outencoding option fmt = HtmlFormatter(encoding="latin1", outencoding="utf8") - tokens = [(Text, u"ä")] - assert u"ä".encode("utf8") in format(tokens, fmt) + tokens = [(Text, "ä")] + assert "ä".encode("utf8") in format(tokens, fmt) @pytest.mark.parametrize('cls', [getattr(formatters, name) @@ -307,7 +307,7 @@ class TestFilters: def test_codetag(self): lx = lexers.PythonLexer() lx.add_filter('codetagify') - text = u'# BUG: text' + text = '# BUG: text' tokens = list(lx.get_tokens(text)) assert '# ' == tokens[0][1] assert 'BUG' == tokens[1][1] @@ -316,15 +316,15 @@ class TestFilters: # ticket #368 lx = lexers.PythonLexer() lx.add_filter('codetagify') - text = u'# DEBUG: text' + text = '# DEBUG: text' tokens = list(lx.get_tokens(text)) assert '# DEBUG: text' == tokens[0][1] def test_symbols(self): lx = lexers.IsabelleLexer() lx.add_filter('symbols') - text = u'lemma "A \\<Longrightarrow> B"' + text = 'lemma "A \\<Longrightarrow> B"' tokens = list(lx.get_tokens(text)) assert 'lemma' == tokens[0][1] assert 'A ' == tokens[3][1] - assert u'\U000027f9' == tokens[4][1] + assert '\U000027f9' == tokens[4][1] diff --git a/tests/test_bibtex.py b/tests/test_bibtex.py index dfa668f2..d7bc02ac 100644 --- a/tests/test_bibtex.py +++ b/tests/test_bibtex.py @@ -21,39 +21,39 @@ def lexer(): def test_preamble(lexer): - data = u'@PREAMBLE{"% some LaTeX code here"}' + data = '@PREAMBLE{"% some LaTeX code here"}' tokens = [ - (Token.Name.Class, u'@PREAMBLE'), - (Token.Punctuation, u'{'), - (Token.String, u'"'), - (Token.String, u'% some LaTeX code here'), - (Token.String, u'"'), - (Token.Punctuation, u'}'), - (Token.Text, u'\n'), + (Token.Name.Class, '@PREAMBLE'), + (Token.Punctuation, '{'), + (Token.String, '"'), + (Token.String, '% some LaTeX code here'), + (Token.String, '"'), + (Token.Punctuation, '}'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(data)) == tokens def test_string(lexer): - data = u'@STRING(SCI = "Science")' + data = '@STRING(SCI = "Science")' tokens = [ - (Token.Name.Class, u'@STRING'), - (Token.Punctuation, u'('), - (Token.Name.Attribute, u'SCI'), - (Token.Text, u' '), - (Token.Punctuation, u'='), - (Token.Text, u' '), - (Token.String, u'"'), - (Token.String, u'Science'), - (Token.String, u'"'), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), + (Token.Name.Class, '@STRING'), + (Token.Punctuation, '('), + (Token.Name.Attribute, 'SCI'), + (Token.Text, ' '), + (Token.Punctuation, '='), + (Token.Text, ' '), + (Token.String, '"'), + (Token.String, 'Science'), + (Token.String, '"'), + (Token.Punctuation, ')'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(data)) == tokens def test_entry(lexer): - data = u""" + data = """ This is a comment. @ARTICLE{ruckenstein-diffusion, @@ -65,56 +65,56 @@ def test_entry(lexer): """ tokens = [ - (Token.Comment, u'This is a comment.'), - (Token.Text, u'\n\n'), - (Token.Name.Class, u'@ARTICLE'), - (Token.Punctuation, u'{'), - (Token.Name.Label, u'ruckenstein-diffusion'), - (Token.Punctuation, u','), - (Token.Text, u'\n '), - (Token.Name.Attribute, u'author'), - (Token.Text, u' '), - (Token.Punctuation, u'='), - (Token.Text, u' '), - (Token.String, u'"'), - (Token.String, u'Liu, Hongquin'), - (Token.String, u'"'), - (Token.Text, u' '), - (Token.Punctuation, u'#'), - (Token.Text, u' '), - (Token.Name.Variable, u'and'), - (Token.Text, u' '), - (Token.Punctuation, u'#'), - (Token.Text, u' '), - (Token.String, u'"'), - (Token.String, u'Ruckenstein, Eli'), - (Token.String, u'"'), - (Token.Punctuation, u','), - (Token.Text, u'\n '), - (Token.Name.Attribute, u'year'), - (Token.Text, u' '), - (Token.Punctuation, u'='), - (Token.Text, u' '), - (Token.Number, u'1997'), - (Token.Punctuation, u','), - (Token.Text, u'\n '), - (Token.Name.Attribute, u'month'), - (Token.Text, u' '), - (Token.Punctuation, u'='), - (Token.Text, u' '), - (Token.Name.Variable, u'JAN'), - (Token.Punctuation, u','), - (Token.Text, u'\n '), - (Token.Name.Attribute, u'pages'), - (Token.Text, u' '), - (Token.Punctuation, u'='), - (Token.Text, u' '), - (Token.String, u'"'), - (Token.String, u'888-895'), - (Token.String, u'"'), - (Token.Text, u'\n'), - (Token.Punctuation, u'}'), - (Token.Text, u'\n'), + (Token.Comment, 'This is a comment.'), + (Token.Text, '\n\n'), + (Token.Name.Class, '@ARTICLE'), + (Token.Punctuation, '{'), + (Token.Name.Label, 'ruckenstein-diffusion'), + (Token.Punctuation, ','), + (Token.Text, '\n '), + (Token.Name.Attribute, 'author'), + (Token.Text, ' '), + (Token.Punctuation, '='), + (Token.Text, ' '), + (Token.String, '"'), + (Token.String, 'Liu, Hongquin'), + (Token.String, '"'), + (Token.Text, ' '), + (Token.Punctuation, '#'), + (Token.Text, ' '), + (Token.Name.Variable, 'and'), + (Token.Text, ' '), + (Token.Punctuation, '#'), + (Token.Text, ' '), + (Token.String, '"'), + (Token.String, 'Ruckenstein, Eli'), + (Token.String, '"'), + (Token.Punctuation, ','), + (Token.Text, '\n '), + (Token.Name.Attribute, 'year'), + (Token.Text, ' '), + (Token.Punctuation, '='), + (Token.Text, ' '), + (Token.Number, '1997'), + (Token.Punctuation, ','), + (Token.Text, '\n '), + (Token.Name.Attribute, 'month'), + (Token.Text, ' '), + (Token.Punctuation, '='), + (Token.Text, ' '), + (Token.Name.Variable, 'JAN'), + (Token.Punctuation, ','), + (Token.Text, '\n '), + (Token.Name.Attribute, 'pages'), + (Token.Text, ' '), + (Token.Punctuation, '='), + (Token.Text, ' '), + (Token.String, '"'), + (Token.String, '888-895'), + (Token.String, '"'), + (Token.Text, '\n'), + (Token.Punctuation, '}'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(textwrap.dedent(data))) == tokens @@ -122,9 +122,9 @@ def test_entry(lexer): def test_comment(lexer): data = '@COMMENT{test}' tokens = [ - (Token.Comment, u'@COMMENT'), - (Token.Comment, u'{test}'), - (Token.Text, u'\n'), + (Token.Comment, '@COMMENT'), + (Token.Comment, '{test}'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(data)) == tokens @@ -132,12 +132,12 @@ def test_comment(lexer): def test_missing_body(lexer): data = '@ARTICLE xxx' tokens = [ - (Token.Name.Class, u'@ARTICLE'), - (Token.Text, u' '), - (Token.Error, u'x'), - (Token.Error, u'x'), - (Token.Error, u'x'), - (Token.Text, u'\n'), + (Token.Name.Class, '@ARTICLE'), + (Token.Text, ' '), + (Token.Error, 'x'), + (Token.Error, 'x'), + (Token.Error, 'x'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(data)) == tokens @@ -145,12 +145,12 @@ def test_missing_body(lexer): def test_mismatched_brace(lexer): data = '@PREAMBLE(""}' tokens = [ - (Token.Name.Class, u'@PREAMBLE'), - (Token.Punctuation, u'('), - (Token.String, u'"'), - (Token.String, u'"'), - (Token.Error, u'}'), - (Token.Text, u'\n'), + (Token.Name.Class, '@PREAMBLE'), + (Token.Punctuation, '('), + (Token.String, '"'), + (Token.String, '"'), + (Token.Error, '}'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(data)) == tokens @@ -177,64 +177,64 @@ def test_basic_bst(): """ tokens = [ (Token.Comment.SingleLine, "% BibTeX standard bibliography style `plain'"), - (Token.Text, u'\n\n'), - (Token.Keyword, u'INTEGERS'), - (Token.Text, u' '), - (Token.Punctuation, u'{'), - (Token.Text, u' '), - (Token.Name.Variable, u'output.state'), - (Token.Text, u' '), - (Token.Name.Variable, u'before.all'), - (Token.Text, u' '), - (Token.Punctuation, u'}'), - (Token.Text, u'\n\n'), - (Token.Keyword, u'FUNCTION'), - (Token.Text, u' '), - (Token.Punctuation, u'{'), - (Token.Name.Variable, u'sort.format.title'), - (Token.Punctuation, u'}'), - (Token.Text, u'\n'), - (Token.Punctuation, u'{'), - (Token.Text, u' '), - (Token.Name.Function, u"'t"), - (Token.Text, u' '), - (Token.Name.Variable, u':='), - (Token.Text, u'\n'), - (Token.Literal.String, u'"A "'), - (Token.Text, u' '), - (Token.Literal.Number, u'#2'), - (Token.Text, u'\n '), - (Token.Literal.String, u'"An "'), - (Token.Text, u' '), - (Token.Literal.Number, u'#3'), - (Token.Text, u'\n '), - (Token.Literal.String, u'"The "'), - (Token.Text, u' '), - (Token.Literal.Number, u'#4'), - (Token.Text, u' '), - (Token.Name.Variable, u't'), - (Token.Text, u' '), - (Token.Name.Variable, u'chop.word'), - (Token.Text, u'\n '), - (Token.Name.Variable, u'chop.word'), - (Token.Text, u'\n'), - (Token.Name.Variable, u'chop.word'), - (Token.Text, u'\n'), - (Token.Name.Variable, u'sortify'), - (Token.Text, u'\n'), - (Token.Literal.Number, u'#1'), - (Token.Text, u' '), - (Token.Name.Builtin, u'global.max$'), - (Token.Text, u' '), - (Token.Name.Builtin, u'substring$'), - (Token.Text, u'\n'), - (Token.Punctuation, u'}'), - (Token.Text, u'\n\n'), - (Token.Keyword, u'ITERATE'), - (Token.Text, u' '), - (Token.Punctuation, u'{'), - (Token.Name.Builtin, u'call.type$'), - (Token.Punctuation, u'}'), - (Token.Text, u'\n'), + (Token.Text, '\n\n'), + (Token.Keyword, 'INTEGERS'), + (Token.Text, ' '), + (Token.Punctuation, '{'), + (Token.Text, ' '), + (Token.Name.Variable, 'output.state'), + (Token.Text, ' '), + (Token.Name.Variable, 'before.all'), + (Token.Text, ' '), + (Token.Punctuation, '}'), + (Token.Text, '\n\n'), + (Token.Keyword, 'FUNCTION'), + (Token.Text, ' '), + (Token.Punctuation, '{'), + (Token.Name.Variable, 'sort.format.title'), + (Token.Punctuation, '}'), + (Token.Text, '\n'), + (Token.Punctuation, '{'), + (Token.Text, ' '), + (Token.Name.Function, "'t"), + (Token.Text, ' '), + (Token.Name.Variable, ':='), + (Token.Text, '\n'), + (Token.Literal.String, '"A "'), + (Token.Text, ' '), + (Token.Literal.Number, '#2'), + (Token.Text, '\n '), + (Token.Literal.String, '"An "'), + (Token.Text, ' '), + (Token.Literal.Number, '#3'), + (Token.Text, '\n '), + (Token.Literal.String, '"The "'), + (Token.Text, ' '), + (Token.Literal.Number, '#4'), + (Token.Text, ' '), + (Token.Name.Variable, 't'), + (Token.Text, ' '), + (Token.Name.Variable, 'chop.word'), + (Token.Text, '\n '), + (Token.Name.Variable, 'chop.word'), + (Token.Text, '\n'), + (Token.Name.Variable, 'chop.word'), + (Token.Text, '\n'), + (Token.Name.Variable, 'sortify'), + (Token.Text, '\n'), + (Token.Literal.Number, '#1'), + (Token.Text, ' '), + (Token.Name.Builtin, 'global.max$'), + (Token.Text, ' '), + (Token.Name.Builtin, 'substring$'), + (Token.Text, '\n'), + (Token.Punctuation, '}'), + (Token.Text, '\n\n'), + (Token.Keyword, 'ITERATE'), + (Token.Text, ' '), + (Token.Punctuation, '{'), + (Token.Name.Builtin, 'call.type$'), + (Token.Punctuation, '}'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(textwrap.dedent(data))) == tokens diff --git a/tests/test_cfm.py b/tests/test_cfm.py index a1600944..a89e160d 100644 --- a/tests/test_cfm.py +++ b/tests/test_cfm.py @@ -19,28 +19,28 @@ def lexer(): def test_basic_comment(lexer): - fragment = u'<!--- cfcomment --->' + fragment = '<!--- cfcomment --->' expected = [ - (Token.Text, u''), - (Token.Comment.Multiline, u'<!---'), - (Token.Comment.Multiline, u' cfcomment '), - (Token.Comment.Multiline, u'--->'), - (Token.Text, u'\n'), + (Token.Text, ''), + (Token.Comment.Multiline, '<!---'), + (Token.Comment.Multiline, ' cfcomment '), + (Token.Comment.Multiline, '--->'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == expected def test_nested_comment(lexer): - fragment = u'<!--- nested <!--- cfcomment ---> --->' + fragment = '<!--- nested <!--- cfcomment ---> --->' expected = [ - (Token.Text, u''), - (Token.Comment.Multiline, u'<!---'), - (Token.Comment.Multiline, u' nested '), - (Token.Comment.Multiline, u'<!---'), - (Token.Comment.Multiline, u' cfcomment '), - (Token.Comment.Multiline, u'--->'), - (Token.Comment.Multiline, u' '), - (Token.Comment.Multiline, u'--->'), - (Token.Text, u'\n'), + (Token.Text, ''), + (Token.Comment.Multiline, '<!---'), + (Token.Comment.Multiline, ' nested '), + (Token.Comment.Multiline, '<!---'), + (Token.Comment.Multiline, ' cfcomment '), + (Token.Comment.Multiline, '--->'), + (Token.Comment.Multiline, ' '), + (Token.Comment.Multiline, '--->'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == expected diff --git a/tests/test_clexer.py b/tests/test_clexer.py index d40ec491..d6561243 100644 --- a/tests/test_clexer.py +++ b/tests/test_clexer.py @@ -33,7 +33,7 @@ def test_numbers(lexer): def test_switch(lexer): - fragment = u'''\ + fragment = '''\ int main() { switch (0) @@ -45,48 +45,48 @@ def test_switch(lexer): } ''' tokens = [ - (Token.Keyword.Type, u'int'), - (Token.Text, u' '), - (Token.Name.Function, u'main'), - (Token.Punctuation, u'('), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), - (Token.Punctuation, u'{'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Keyword, u'switch'), - (Token.Text, u' '), - (Token.Punctuation, u'('), - (Token.Literal.Number.Integer, u'0'), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Punctuation, u'{'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Keyword, u'case'), - (Token.Text, u' '), - (Token.Literal.Number.Integer, u'0'), - (Token.Operator, u':'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Keyword, u'default'), - (Token.Operator, u':'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Punctuation, u';'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Punctuation, u'}'), - (Token.Text, u'\n'), - (Token.Punctuation, u'}'), - (Token.Text, u'\n'), + (Token.Keyword.Type, 'int'), + (Token.Text, ' '), + (Token.Name.Function, 'main'), + (Token.Punctuation, '('), + (Token.Punctuation, ')'), + (Token.Text, '\n'), + (Token.Punctuation, '{'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Keyword, 'switch'), + (Token.Text, ' '), + (Token.Punctuation, '('), + (Token.Literal.Number.Integer, '0'), + (Token.Punctuation, ')'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Punctuation, '{'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Keyword, 'case'), + (Token.Text, ' '), + (Token.Literal.Number.Integer, '0'), + (Token.Operator, ':'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Keyword, 'default'), + (Token.Operator, ':'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Punctuation, ';'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Punctuation, '}'), + (Token.Text, '\n'), + (Token.Punctuation, '}'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(textwrap.dedent(fragment))) == tokens def test_switch_space_before_colon(lexer): - fragment = u'''\ + fragment = '''\ int main() { switch (0) @@ -98,50 +98,50 @@ def test_switch_space_before_colon(lexer): } ''' tokens = [ - (Token.Keyword.Type, u'int'), - (Token.Text, u' '), - (Token.Name.Function, u'main'), - (Token.Punctuation, u'('), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), - (Token.Punctuation, u'{'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Keyword, u'switch'), - (Token.Text, u' '), - (Token.Punctuation, u'('), - (Token.Literal.Number.Integer, u'0'), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Punctuation, u'{'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Keyword, u'case'), - (Token.Text, u' '), - (Token.Literal.Number.Integer, u'0'), - (Token.Text, u' '), - (Token.Operator, u':'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Keyword, u'default'), - (Token.Text, u' '), - (Token.Operator, u':'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Punctuation, u';'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Punctuation, u'}'), - (Token.Text, u'\n'), - (Token.Punctuation, u'}'), - (Token.Text, u'\n'), + (Token.Keyword.Type, 'int'), + (Token.Text, ' '), + (Token.Name.Function, 'main'), + (Token.Punctuation, '('), + (Token.Punctuation, ')'), + (Token.Text, '\n'), + (Token.Punctuation, '{'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Keyword, 'switch'), + (Token.Text, ' '), + (Token.Punctuation, '('), + (Token.Literal.Number.Integer, '0'), + (Token.Punctuation, ')'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Punctuation, '{'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Keyword, 'case'), + (Token.Text, ' '), + (Token.Literal.Number.Integer, '0'), + (Token.Text, ' '), + (Token.Operator, ':'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Keyword, 'default'), + (Token.Text, ' '), + (Token.Operator, ':'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Punctuation, ';'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Punctuation, '}'), + (Token.Text, '\n'), + (Token.Punctuation, '}'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(textwrap.dedent(fragment))) == tokens def test_label(lexer): - fragment = u'''\ + fragment = '''\ int main() { foo: @@ -149,31 +149,31 @@ def test_label(lexer): } ''' tokens = [ - (Token.Keyword.Type, u'int'), - (Token.Text, u' '), - (Token.Name.Function, u'main'), - (Token.Punctuation, u'('), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), - (Token.Punctuation, u'{'), - (Token.Text, u'\n'), - (Token.Name.Label, u'foo'), - (Token.Punctuation, u':'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Keyword, u'goto'), - (Token.Text, u' '), - (Token.Name, u'foo'), - (Token.Punctuation, u';'), - (Token.Text, u'\n'), - (Token.Punctuation, u'}'), - (Token.Text, u'\n'), + (Token.Keyword.Type, 'int'), + (Token.Text, ' '), + (Token.Name.Function, 'main'), + (Token.Punctuation, '('), + (Token.Punctuation, ')'), + (Token.Text, '\n'), + (Token.Punctuation, '{'), + (Token.Text, '\n'), + (Token.Name.Label, 'foo'), + (Token.Punctuation, ':'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Keyword, 'goto'), + (Token.Text, ' '), + (Token.Name, 'foo'), + (Token.Punctuation, ';'), + (Token.Text, '\n'), + (Token.Punctuation, '}'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(textwrap.dedent(fragment))) == tokens def test_label_space_before_colon(lexer): - fragment = u'''\ + fragment = '''\ int main() { foo : @@ -181,32 +181,32 @@ def test_label_space_before_colon(lexer): } ''' tokens = [ - (Token.Keyword.Type, u'int'), - (Token.Text, u' '), - (Token.Name.Function, u'main'), - (Token.Punctuation, u'('), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), - (Token.Punctuation, u'{'), - (Token.Text, u'\n'), - (Token.Name.Label, u'foo'), - (Token.Text, u' '), - (Token.Punctuation, u':'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Keyword, u'goto'), - (Token.Text, u' '), - (Token.Name, u'foo'), - (Token.Punctuation, u';'), - (Token.Text, u'\n'), - (Token.Punctuation, u'}'), - (Token.Text, u'\n'), + (Token.Keyword.Type, 'int'), + (Token.Text, ' '), + (Token.Name.Function, 'main'), + (Token.Punctuation, '('), + (Token.Punctuation, ')'), + (Token.Text, '\n'), + (Token.Punctuation, '{'), + (Token.Text, '\n'), + (Token.Name.Label, 'foo'), + (Token.Text, ' '), + (Token.Punctuation, ':'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Keyword, 'goto'), + (Token.Text, ' '), + (Token.Name, 'foo'), + (Token.Punctuation, ';'), + (Token.Text, '\n'), + (Token.Punctuation, '}'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(textwrap.dedent(fragment))) == tokens def test_label_followed_by_statement(lexer): - fragment = u'''\ + fragment = '''\ int main() { foo:return 0; @@ -214,52 +214,52 @@ def test_label_followed_by_statement(lexer): } ''' tokens = [ - (Token.Keyword.Type, u'int'), - (Token.Text, u' '), - (Token.Name.Function, u'main'), - (Token.Punctuation, u'('), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), - (Token.Punctuation, u'{'), - (Token.Text, u'\n'), - (Token.Name.Label, u'foo'), - (Token.Punctuation, u':'), - (Token.Keyword, u'return'), - (Token.Text, u' '), - (Token.Literal.Number.Integer, u'0'), - (Token.Punctuation, u';'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Keyword, u'goto'), - (Token.Text, u' '), - (Token.Name, u'foo'), - (Token.Punctuation, u';'), - (Token.Text, u'\n'), - (Token.Punctuation, u'}'), - (Token.Text, u'\n'), + (Token.Keyword.Type, 'int'), + (Token.Text, ' '), + (Token.Name.Function, 'main'), + (Token.Punctuation, '('), + (Token.Punctuation, ')'), + (Token.Text, '\n'), + (Token.Punctuation, '{'), + (Token.Text, '\n'), + (Token.Name.Label, 'foo'), + (Token.Punctuation, ':'), + (Token.Keyword, 'return'), + (Token.Text, ' '), + (Token.Literal.Number.Integer, '0'), + (Token.Punctuation, ';'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Keyword, 'goto'), + (Token.Text, ' '), + (Token.Name, 'foo'), + (Token.Punctuation, ';'), + (Token.Text, '\n'), + (Token.Punctuation, '}'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(textwrap.dedent(fragment))) == tokens def test_preproc_file(lexer): - fragment = u'#include <foo>\n' + fragment = '#include <foo>\n' tokens = [ - (Token.Comment.Preproc, u'#'), - (Token.Comment.Preproc, u'include'), - (Token.Text, u' '), - (Token.Comment.PreprocFile, u'<foo>'), - (Token.Comment.Preproc, u'\n'), + (Token.Comment.Preproc, '#'), + (Token.Comment.Preproc, 'include'), + (Token.Text, ' '), + (Token.Comment.PreprocFile, '<foo>'), + (Token.Comment.Preproc, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_preproc_file2(lexer): - fragment = u'#include "foo.h"\n' + fragment = '#include "foo.h"\n' tokens = [ - (Token.Comment.Preproc, u'#'), - (Token.Comment.Preproc, u'include'), - (Token.Text, u' '), - (Token.Comment.PreprocFile, u'"foo.h"'), - (Token.Comment.Preproc, u'\n'), + (Token.Comment.Preproc, '#'), + (Token.Comment.Preproc, 'include'), + (Token.Text, ' '), + (Token.Comment.PreprocFile, '"foo.h"'), + (Token.Comment.Preproc, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens diff --git a/tests/test_cpp.py b/tests/test_cpp.py index a3ef33a3..e847079d 100644 --- a/tests/test_cpp.py +++ b/tests/test_cpp.py @@ -21,18 +21,18 @@ def lexer(): def test_good_comment(lexer): - fragment = u'/* foo */\n' + fragment = '/* foo */\n' tokens = [ - (Token.Comment.Multiline, u'/* foo */'), - (Token.Text, u'\n'), + (Token.Comment.Multiline, '/* foo */'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_open_comment(lexer): - fragment = u'/* foo\n' + fragment = '/* foo\n' tokens = [ - (Token.Comment.Multiline, u'/* foo\n'), + (Token.Comment.Multiline, '/* foo\n'), ] assert list(lexer.get_tokens(fragment)) == tokens @@ -52,4 +52,4 @@ def test_guess_c_lexer(): } ''' lexer = guess_lexer(code) - assert isinstance(lexer, CLexer)
\ No newline at end of file + assert isinstance(lexer, CLexer) diff --git a/tests/test_csound.py b/tests/test_csound.py index 0186da7a..26672043 100644 --- a/tests/test_csound.py +++ b/tests/test_csound.py @@ -30,12 +30,12 @@ def test_comments(lexer): // comment ''') tokens = [ - (Comment.Multiline, u'/*\n * comment\n */'), - (Text, u'\n'), - (Comment.Single, u'; comment'), - (Text, u'\n'), - (Comment.Single, u'// comment'), - (Text, u'\n') + (Comment.Multiline, '/*\n * comment\n */'), + (Text, '\n'), + (Comment.Single, '; comment'), + (Text, '\n'), + (Comment.Single, '// comment'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -48,38 +48,38 @@ def test_instrument_blocks(lexer): endin ''') tokens = [ - (Keyword.Declaration, u'instr'), - (Comment.Multiline, u'/**/'), - (Name.Function, u'1'), - (Punctuation, u','), - (Comment.Multiline, u'/**/'), - (Name.Function, u'N_a_M_e_'), - (Punctuation, u','), - (Comment.Multiline, u'/**/'), - (Punctuation, u'+'), - (Name.Function, u'Name'), - (Comment.Multiline, u'/**/'), - (Comment.Single, u'//'), - (Text, u'\n'), - (Text, u' '), - (Keyword.Type, u'i'), - (Name, u'Duration'), - (Text, u' '), - (Operator, u'='), - (Text, u' '), - (Name.Variable.Instance, u'p3'), - (Text, u'\n'), - (Text, u' '), - (Name.Builtin, u'outc'), - (Punctuation, u':'), - (Keyword.Type, u'a'), - (Punctuation, u'('), - (Keyword.Type, u'a'), - (Name, u'Signal'), - (Punctuation, u')'), - (Text, u'\n'), - (Keyword.Declaration, u'endin'), - (Text, u'\n') + (Keyword.Declaration, 'instr'), + (Comment.Multiline, '/**/'), + (Name.Function, '1'), + (Punctuation, ','), + (Comment.Multiline, '/**/'), + (Name.Function, 'N_a_M_e_'), + (Punctuation, ','), + (Comment.Multiline, '/**/'), + (Punctuation, '+'), + (Name.Function, 'Name'), + (Comment.Multiline, '/**/'), + (Comment.Single, '//'), + (Text, '\n'), + (Text, ' '), + (Keyword.Type, 'i'), + (Name, 'Duration'), + (Text, ' '), + (Operator, '='), + (Text, ' '), + (Name.Variable.Instance, 'p3'), + (Text, '\n'), + (Text, ' '), + (Name.Builtin, 'outc'), + (Punctuation, ':'), + (Keyword.Type, 'a'), + (Punctuation, '('), + (Keyword.Type, 'a'), + (Name, 'Signal'), + (Punctuation, ')'), + (Text, '\n'), + (Keyword.Declaration, 'endin'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -91,22 +91,22 @@ def test_user_defined_opcodes(lexer): endop ''') tokens = [ - (Keyword.Declaration, u'opcode'), - (Comment.Multiline, u'/**/'), - (Name.Function, u'aUDO'), - (Punctuation, u','), - (Comment.Multiline, u'/**/'), - (Keyword.Type, u'i[]'), - (Punctuation, u','), - (Comment.Multiline, u'/**/'), - (Keyword.Type, u'aik'), - (Comment.Single, u'//'), - (Text, u'\n'), - (Text, u' '), - (Name.Function, u'aUDO'), - (Text, u'\n'), - (Keyword.Declaration, u'endop'), - (Text, u'\n') + (Keyword.Declaration, 'opcode'), + (Comment.Multiline, '/**/'), + (Name.Function, 'aUDO'), + (Punctuation, ','), + (Comment.Multiline, '/**/'), + (Keyword.Type, 'i[]'), + (Punctuation, ','), + (Comment.Multiline, '/**/'), + (Keyword.Type, 'aik'), + (Comment.Single, '//'), + (Text, '\n'), + (Text, ' '), + (Name.Function, 'aUDO'), + (Text, '\n'), + (Keyword.Declaration, 'endop'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -114,27 +114,27 @@ def test_user_defined_opcodes(lexer): def test_numbers(lexer): fragment = '123 0123456789' tokens = [ - (Number.Integer, u'123'), - (Text, u' '), - (Number.Integer, u'0123456789'), - (Text, u'\n') + (Number.Integer, '123'), + (Text, ' '), + (Number.Integer, '0123456789'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens fragment = '0xabcdef0123456789 0XABCDEF' tokens = [ - (Keyword.Type, u'0x'), - (Number.Hex, u'abcdef0123456789'), - (Text, u' '), - (Keyword.Type, u'0X'), - (Number.Hex, u'ABCDEF'), - (Text, u'\n') + (Keyword.Type, '0x'), + (Number.Hex, 'abcdef0123456789'), + (Text, ' '), + (Keyword.Type, '0X'), + (Number.Hex, 'ABCDEF'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens fragments = ['1e2', '3e+4', '5e-6', '7E8', '9E+0', '1E-2', '3.', '4.56', '.789'] for fragment in fragments: tokens = [ (Number.Float, fragment), - (Text, u'\n') + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -142,11 +142,11 @@ def test_numbers(lexer): def test_quoted_strings(lexer): fragment = '"characters$MACRO."' tokens = [ - (String, u'"'), - (String, u'characters'), - (Comment.Preproc, u'$MACRO.'), - (String, u'"'), - (Text, u'\n') + (String, '"'), + (String, 'characters'), + (Comment.Preproc, '$MACRO.'), + (String, '"'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -158,10 +158,10 @@ def test_braced_strings(lexer): }} ''') tokens = [ - (String, u'{{'), - (String, u'\ncharacters$MACRO.\n'), - (String, u'}}'), - (Text, u'\n') + (String, '{{'), + (String, '\ncharacters$MACRO.\n'), + (String, '}}'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -171,30 +171,30 @@ def test_escape_sequences(lexer): escapedCharacter = '\\' + character fragment = '"' + escapedCharacter + '"' tokens = [ - (String, u'"'), + (String, '"'), (String.Escape, escapedCharacter), - (String, u'"'), - (Text, u'\n') + (String, '"'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens fragment = '{{' + escapedCharacter + '}}' tokens = [ - (String, u'{{'), + (String, '{{'), (String.Escape, escapedCharacter), - (String, u'}}'), - (Text, u'\n') + (String, '}}'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens def test_operators(lexer): - fragments = ['+', '-', '~', u'¬', '!', '*', '/', '^', '%', '<<', '>>', '<', '>', + fragments = ['+', '-', '~', '¬', '!', '*', '/', '^', '%', '<<', '>>', '<', '>', '<=', '>=', '==', '!=', '&', '#', '|', '&&', '||', '?', ':', '+=', '-=', '*=', '/='] for fragment in fragments: tokens = [ (Operator, fragment), - (Text, u'\n') + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -203,7 +203,7 @@ def test_global_value_identifiers(lexer): for fragment in ['0dbfs', 'A4', 'kr', 'ksmps', 'nchnls', 'nchnls_i', 'sr']: tokens = [ (Name.Variable.Global, fragment), - (Text, u'\n') + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -214,13 +214,13 @@ def test_keywords(lexer): for fragment in fragments: tokens = [ (Keyword, fragment), - (Text, u'\n') + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens for fragment in ['return', 'rireturn']: tokens = [ (Keyword.Pseudo, fragment), - (Text, u'\n') + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -231,13 +231,13 @@ def test_labels(lexer): label2: ''') tokens = [ - (Name.Label, u'aLabel'), - (Punctuation, u':'), - (Text, u'\n'), - (Text, u' '), - (Name.Label, u'label2'), - (Punctuation, u':'), - (Text, u'\n') + (Name.Label, 'aLabel'), + (Punctuation, ':'), + (Text, '\n'), + (Text, ' '), + (Name.Label, 'label2'), + (Punctuation, ':'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -251,11 +251,11 @@ def test_printks_and_prints_escape_sequences(lexer): fragment = opcode + ' "' + escapedCharacter + '"' tokens = [ (Name.Builtin, opcode), - (Text, u' '), - (String, u'"'), + (Text, ' '), + (String, '"'), (String.Escape, escapedCharacter), - (String, u'"'), - (Text, u'\n') + (String, '"'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -265,64 +265,64 @@ def test_goto_statements(lexer): fragment = keyword + ' aLabel' tokens = [ (Keyword, keyword), - (Text, u' '), - (Name.Label, u'aLabel'), - (Text, u'\n') + (Text, ' '), + (Name.Label, 'aLabel'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens for opcode in ['reinit', 'rigoto', 'tigoto']: fragment = opcode + ' aLabel' tokens = [ (Keyword.Pseudo, opcode), - (Text, u' '), - (Name.Label, u'aLabel'), - (Text, u'\n') + (Text, ' '), + (Name.Label, 'aLabel'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens for opcode in ['cggoto', 'cigoto', 'cingoto', 'ckgoto', 'cngoto', 'cnkgoto']: fragment = opcode + ' 1==0, aLabel' tokens = [ (Keyword.Pseudo, opcode), - (Text, u' '), - (Number.Integer, u'1'), - (Operator, u'=='), - (Number.Integer, u'0'), - (Punctuation, u','), - (Text, u' '), - (Name.Label, u'aLabel'), - (Text, u'\n') + (Text, ' '), + (Number.Integer, '1'), + (Operator, '=='), + (Number.Integer, '0'), + (Punctuation, ','), + (Text, ' '), + (Name.Label, 'aLabel'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens fragment = 'timout 0, 0, aLabel' tokens = [ (Keyword.Pseudo, 'timout'), - (Text, u' '), - (Number.Integer, u'0'), - (Punctuation, u','), - (Text, u' '), - (Number.Integer, u'0'), - (Punctuation, u','), - (Text, u' '), - (Name.Label, u'aLabel'), - (Text, u'\n') + (Text, ' '), + (Number.Integer, '0'), + (Punctuation, ','), + (Text, ' '), + (Number.Integer, '0'), + (Punctuation, ','), + (Text, ' '), + (Name.Label, 'aLabel'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens for opcode in ['loop_ge', 'loop_gt', 'loop_le', 'loop_lt']: fragment = opcode + ' 0, 0, 0, aLabel' tokens = [ (Keyword.Pseudo, opcode), - (Text, u' '), - (Number.Integer, u'0'), - (Punctuation, u','), - (Text, u' '), - (Number.Integer, u'0'), - (Punctuation, u','), - (Text, u' '), - (Number.Integer, u'0'), - (Punctuation, u','), - (Text, u' '), - (Name.Label, u'aLabel'), - (Text, u'\n') + (Text, ' '), + (Number.Integer, '0'), + (Punctuation, ','), + (Text, ' '), + (Number.Integer, '0'), + (Punctuation, ','), + (Text, ' '), + (Number.Integer, '0'), + (Punctuation, ','), + (Text, ' '), + (Name.Label, 'aLabel'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -331,10 +331,10 @@ def test_include_directives(lexer): for character in ['"', '|']: fragment = '#include/**/' + character + 'file.udo' + character tokens = [ - (Comment.Preproc, u'#include'), - (Comment.Multiline, u'/**/'), - (String, character + u'file.udo' + character), - (Text, u'\n') + (Comment.Preproc, '#include'), + (Comment.Multiline, '/**/'), + (String, character + 'file.udo' + character), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -342,13 +342,13 @@ def test_include_directives(lexer): def test_includestr_directives(lexer): fragment = '#includestr/**/"$MACRO..udo"' tokens = [ - (Comment.Preproc, u'#includestr'), - (Comment.Multiline, u'/**/'), - (String, u'"'), - (Comment.Preproc, u'$MACRO.'), - (String, u'.udo'), - (String, u'"'), - (Text, u'\n') + (Comment.Preproc, '#includestr'), + (Comment.Multiline, '/**/'), + (String, '"'), + (Comment.Preproc, '$MACRO.'), + (String, '.udo'), + (String, '"'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -362,25 +362,25 @@ def test_object_like_macro_definitions(lexer): body\\## ''') tokens = [ - (Comment.Preproc, u'# \tdefine'), - (Text, u' '), - (Comment.Preproc, u'MACRO'), - (Punctuation, u'#'), - (Comment.Preproc, u'macro_body'), - (Punctuation, u'#'), - (Text, u'\n'), - (Comment.Preproc, u'#define'), - (Comment.Multiline, u'/**/'), - (Text, u'\n'), - (Comment.Preproc, u'MACRO'), - (Comment.Multiline, u'/**/'), - (Text, u'\n'), - (Punctuation, u'#'), - (Comment.Preproc, u'\\#'), - (Comment.Preproc, u'macro\nbody'), - (Comment.Preproc, u'\\#'), - (Punctuation, u'#'), - (Text, u'\n') + (Comment.Preproc, '# \tdefine'), + (Text, ' '), + (Comment.Preproc, 'MACRO'), + (Punctuation, '#'), + (Comment.Preproc, 'macro_body'), + (Punctuation, '#'), + (Text, '\n'), + (Comment.Preproc, '#define'), + (Comment.Multiline, '/**/'), + (Text, '\n'), + (Comment.Preproc, 'MACRO'), + (Comment.Multiline, '/**/'), + (Text, '\n'), + (Punctuation, '#'), + (Comment.Preproc, '\\#'), + (Comment.Preproc, 'macro\nbody'), + (Comment.Preproc, '\\#'), + (Punctuation, '#'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -394,39 +394,39 @@ def test_function_like_macro_definitions(lexer): body\\## ''') tokens = [ - (Comment.Preproc, u'#define'), - (Text, u' '), - (Comment.Preproc, u'MACRO'), - (Punctuation, u'('), - (Comment.Preproc, u'ARG1'), - (Punctuation, u'#'), - (Comment.Preproc, u'ARG2'), - (Punctuation, u')'), - (Text, u' '), - (Punctuation, u'#'), - (Comment.Preproc, u'macro_body'), - (Punctuation, u'#'), - (Text, u'\n'), - (Comment.Preproc, u'#define'), - (Comment.Multiline, u'/**/'), - (Text, u'\n'), - (Comment.Preproc, u'MACRO'), - (Punctuation, u'('), - (Comment.Preproc, u'ARG1'), - (Punctuation, u"'"), - (Comment.Preproc, u'ARG2'), - (Punctuation, u"'"), - (Text, u' '), - (Comment.Preproc, u'ARG3'), - (Punctuation, u')'), - (Comment.Multiline, u'/**/'), - (Text, u'\n'), - (Punctuation, u'#'), - (Comment.Preproc, u'\\#'), - (Comment.Preproc, u'macro\nbody'), - (Comment.Preproc, u'\\#'), - (Punctuation, u'#'), - (Text, u'\n') + (Comment.Preproc, '#define'), + (Text, ' '), + (Comment.Preproc, 'MACRO'), + (Punctuation, '('), + (Comment.Preproc, 'ARG1'), + (Punctuation, '#'), + (Comment.Preproc, 'ARG2'), + (Punctuation, ')'), + (Text, ' '), + (Punctuation, '#'), + (Comment.Preproc, 'macro_body'), + (Punctuation, '#'), + (Text, '\n'), + (Comment.Preproc, '#define'), + (Comment.Multiline, '/**/'), + (Text, '\n'), + (Comment.Preproc, 'MACRO'), + (Punctuation, '('), + (Comment.Preproc, 'ARG1'), + (Punctuation, "'"), + (Comment.Preproc, 'ARG2'), + (Punctuation, "'"), + (Text, ' '), + (Comment.Preproc, 'ARG3'), + (Punctuation, ')'), + (Comment.Multiline, '/**/'), + (Text, '\n'), + (Punctuation, '#'), + (Comment.Preproc, '\\#'), + (Comment.Preproc, 'macro\nbody'), + (Comment.Preproc, '\\#'), + (Punctuation, '#'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -436,9 +436,9 @@ def test_macro_preprocessor_directives(lexer): fragment = directive + ' MACRO' tokens = [ (Comment.Preproc, directive), - (Text, u' '), - (Comment.Preproc, u'MACRO'), - (Text, u'\n') + (Text, ' '), + (Comment.Preproc, 'MACRO'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -453,18 +453,18 @@ def test_other_preprocessor_directives(lexer): @@ \t67890 ''') tokens = [ - (Comment.Preproc, u'#else'), - (Text, u'\n'), - (Comment.Preproc, u'#end'), - (Text, u'\n'), - (Comment.Preproc, u'#endif'), - (Text, u'\n'), - (Comment.Preproc, u'###'), - (Text, u'\n'), - (Comment.Preproc, u'@ \t12345'), - (Text, u'\n'), - (Comment.Preproc, u'@@ \t67890'), - (Text, u'\n') + (Comment.Preproc, '#else'), + (Text, '\n'), + (Comment.Preproc, '#end'), + (Text, '\n'), + (Comment.Preproc, '#endif'), + (Text, '\n'), + (Comment.Preproc, '###'), + (Text, '\n'), + (Comment.Preproc, '@ \t12345'), + (Text, '\n'), + (Comment.Preproc, '@@ \t67890'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens @@ -472,42 +472,42 @@ def test_other_preprocessor_directives(lexer): def test_function_like_macros(lexer): fragment = "$MACRO.(((x#y\\)))' \"(#'x)\\)x\\))\"# {{x\\))x)\\)(#'}});" tokens = [ - (Comment.Preproc, u'$MACRO.'), - (Punctuation, u'('), - (Comment.Preproc, u'('), - (Comment.Preproc, u'('), - (Comment.Preproc, u'x#y\\)'), - (Comment.Preproc, u')'), - (Comment.Preproc, u')'), - (Punctuation, u"'"), - (Comment.Preproc, u' '), - (String, u'"'), - (Error, u'('), - (Error, u'#'), - (Error, u"'"), - (String, u'x'), - (Error, u')'), - (Comment.Preproc, u'\\)'), - (String, u'x'), - (Comment.Preproc, u'\\)'), - (Error, u')'), - (String, u'"'), - (Punctuation, u'#'), - (Comment.Preproc, u' '), - (String, u'{{'), - (String, u'x'), - (Comment.Preproc, u'\\)'), - (Error, u')'), - (String, u'x'), - (Error, u')'), - (Comment.Preproc, u'\\)'), - (Error, u'('), - (Error, u'#'), - (Error, u"'"), - (String, u'}}'), - (Punctuation, u')'), - (Comment.Single, u';'), - (Text, u'\n') + (Comment.Preproc, '$MACRO.'), + (Punctuation, '('), + (Comment.Preproc, '('), + (Comment.Preproc, '('), + (Comment.Preproc, 'x#y\\)'), + (Comment.Preproc, ')'), + (Comment.Preproc, ')'), + (Punctuation, "'"), + (Comment.Preproc, ' '), + (String, '"'), + (Error, '('), + (Error, '#'), + (Error, "'"), + (String, 'x'), + (Error, ')'), + (Comment.Preproc, '\\)'), + (String, 'x'), + (Comment.Preproc, '\\)'), + (Error, ')'), + (String, '"'), + (Punctuation, '#'), + (Comment.Preproc, ' '), + (String, '{{'), + (String, 'x'), + (Comment.Preproc, '\\)'), + (Error, ')'), + (String, 'x'), + (Error, ')'), + (Comment.Preproc, '\\)'), + (Error, '('), + (Error, '#'), + (Error, "'"), + (String, '}}'), + (Punctuation, ')'), + (Comment.Single, ';'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens diff --git a/tests/test_data.py b/tests/test_data.py index 9724d235..22d4ee79 100644 --- a/tests/test_data.py +++ b/tests/test_data.py @@ -29,29 +29,29 @@ def lexer_yaml(): def test_basic_json(lexer_json): - fragment = u'{"foo": "bar", "foo2": [1, 2, 3]}\n' + fragment = '{"foo": "bar", "foo2": [1, 2, 3]}\n' tokens = [ - (Token.Punctuation, u'{'), - (Token.Name.Tag, u'"foo"'), - (Token.Punctuation, u':'), - (Token.Text, u' '), - (Token.Literal.String.Double, u'"bar"'), - (Token.Punctuation, u','), - (Token.Text, u' '), - (Token.Name.Tag, u'"foo2"'), - (Token.Punctuation, u':'), - (Token.Text, u' '), - (Token.Punctuation, u'['), - (Token.Literal.Number.Integer, u'1'), - (Token.Punctuation, u','), - (Token.Text, u' '), - (Token.Literal.Number.Integer, u'2'), - (Token.Punctuation, u','), - (Token.Text, u' '), - (Token.Literal.Number.Integer, u'3'), - (Token.Punctuation, u']'), - (Token.Punctuation, u'}'), - (Token.Text, u'\n'), + (Token.Punctuation, '{'), + (Token.Name.Tag, '"foo"'), + (Token.Punctuation, ':'), + (Token.Text, ' '), + (Token.Literal.String.Double, '"bar"'), + (Token.Punctuation, ','), + (Token.Text, ' '), + (Token.Name.Tag, '"foo2"'), + (Token.Punctuation, ':'), + (Token.Text, ' '), + (Token.Punctuation, '['), + (Token.Literal.Number.Integer, '1'), + (Token.Punctuation, ','), + (Token.Text, ' '), + (Token.Literal.Number.Integer, '2'), + (Token.Punctuation, ','), + (Token.Text, ' '), + (Token.Literal.Number.Integer, '3'), + (Token.Punctuation, ']'), + (Token.Punctuation, '}'), + (Token.Text, '\n'), ] assert list(lexer_json.get_tokens(fragment)) == tokens @@ -62,7 +62,7 @@ def test_json_escape_backtracking(lexer_json): # this test will hang and that's how we know it's broken :( fragment = r'{"\u00D0000\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\63CD' tokens = ( - [(Token.Punctuation, u'{'), + [(Token.Punctuation, '{'), (Token.Error, r'"'), (Token.Error, '\\'), (Token.Error, r'u'), @@ -87,27 +87,27 @@ def test_json_escape_backtracking(lexer_json): def test_basic_bare(lexer_bare): # This is the same as testBasic for JsonLexer above, except the # enclosing curly braces are removed. - fragment = u'"foo": "bar", "foo2": [1, 2, 3]\n' + fragment = '"foo": "bar", "foo2": [1, 2, 3]\n' tokens = [ - (Token.Name.Tag, u'"foo"'), - (Token.Punctuation, u':'), - (Token.Text, u' '), - (Token.Literal.String.Double, u'"bar"'), - (Token.Punctuation, u','), - (Token.Text, u' '), - (Token.Name.Tag, u'"foo2"'), - (Token.Punctuation, u':'), - (Token.Text, u' '), - (Token.Punctuation, u'['), - (Token.Literal.Number.Integer, u'1'), - (Token.Punctuation, u','), - (Token.Text, u' '), - (Token.Literal.Number.Integer, u'2'), - (Token.Punctuation, u','), - (Token.Text, u' '), - (Token.Literal.Number.Integer, u'3'), - (Token.Punctuation, u']'), - (Token.Text, u'\n'), + (Token.Name.Tag, '"foo"'), + (Token.Punctuation, ':'), + (Token.Text, ' '), + (Token.Literal.String.Double, '"bar"'), + (Token.Punctuation, ','), + (Token.Text, ' '), + (Token.Name.Tag, '"foo2"'), + (Token.Punctuation, ':'), + (Token.Text, ' '), + (Token.Punctuation, '['), + (Token.Literal.Number.Integer, '1'), + (Token.Punctuation, ','), + (Token.Text, ' '), + (Token.Literal.Number.Integer, '2'), + (Token.Punctuation, ','), + (Token.Text, ' '), + (Token.Literal.Number.Integer, '3'), + (Token.Punctuation, ']'), + (Token.Text, '\n'), ] assert list(lexer_bare.get_tokens(fragment)) == tokens @@ -139,14 +139,14 @@ def test_closing_curly_in_value(lexer_bare): def test_yaml(lexer_yaml): # Bug #1528: This previously parsed 'token # innocent' as a tag - fragment = u'here: token # innocent: comment\n' + fragment = 'here: token # innocent: comment\n' tokens = [ - (Token.Name.Tag, u'here'), - (Token.Punctuation, u':'), - (Token.Text, u' '), - (Token.Literal.Scalar.Plain, u'token'), - (Token.Text, u' '), - (Token.Comment.Single, u'# innocent: comment'), - (Token.Text, u'\n'), + (Token.Name.Tag, 'here'), + (Token.Punctuation, ':'), + (Token.Text, ' '), + (Token.Literal.Scalar.Plain, 'token'), + (Token.Text, ' '), + (Token.Comment.Single, '# innocent: comment'), + (Token.Text, '\n'), ] assert list(lexer_yaml.get_tokens(fragment)) == tokens diff --git a/tests/test_examplefiles.py b/tests/test_examplefiles.py index 22b5ced0..25b355ec 100644 --- a/tests/test_examplefiles.py +++ b/tests/test_examplefiles.py @@ -101,8 +101,8 @@ def test_examplefile(filename): text = text.strip(b'\n') + b'\n' try: text = text.decode('utf-8') - if text.startswith(u'\ufeff'): - text = text[len(u'\ufeff'):] + if text.startswith('\ufeff'): + text = text[len('\ufeff'):] except UnicodeError: text = text.decode('latin1') ntext = [] @@ -113,13 +113,13 @@ def test_examplefile(filename): ntext.append(val) assert type != Error, \ 'lexer %s generated error token for %s: %r at position %d' % \ - (lx, absfn, val, len(u''.join(ntext))) + (lx, absfn, val, len(''.join(ntext))) tokens.append((type, val)) t2 = time.time() STATS[os.path.basename(absfn)] = (len(text), 1000 * (t2 - t1), 1000 * (t2 - t1) / len(text)) - if u''.join(ntext) != text: - print('\n'.join(difflib.unified_diff(u''.join(ntext).splitlines(), + if ''.join(ntext) != text: + print('\n'.join(difflib.unified_diff(''.join(ntext).splitlines(), text.splitlines()))) raise AssertionError('round trip failed for ' + absfn) diff --git a/tests/test_ezhil.py b/tests/test_ezhil.py index 8047a30a..beca4c65 100644 --- a/tests/test_ezhil.py +++ b/tests/test_ezhil.py @@ -19,71 +19,71 @@ def lexer(): def test_sum(lexer): - fragment = u'1+3\n' + fragment = '1+3\n' tokens = [ - (Number.Integer, u'1'), - (Operator, u'+'), - (Number.Integer, u'3'), - (Text, u'\n'), + (Number.Integer, '1'), + (Operator, '+'), + (Number.Integer, '3'), + (Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_gcd_expr(lexer): - fragment = u'1^3+(5-5)*gcd(a,b)\n' + fragment = '1^3+(5-5)*gcd(a,b)\n' tokens = [ - (Token.Number.Integer, u'1'), - (Token.Operator, u'^'), - (Token.Literal.Number.Integer, u'3'), - (Token.Operator, u'+'), - (Token.Punctuation, u'('), - (Token.Literal.Number.Integer, u'5'), - (Token.Operator, u'-'), - (Token.Literal.Number.Integer, u'5'), - (Token.Punctuation, u')'), - (Token.Operator, u'*'), - (Token.Name, u'gcd'), - (Token.Punctuation, u'('), - (Token.Name, u'a'), - (Token.Operator, u','), - (Token.Name, u'b'), - (Token.Punctuation, u')'), - (Token.Text, u'\n') + (Token.Number.Integer, '1'), + (Token.Operator, '^'), + (Token.Literal.Number.Integer, '3'), + (Token.Operator, '+'), + (Token.Punctuation, '('), + (Token.Literal.Number.Integer, '5'), + (Token.Operator, '-'), + (Token.Literal.Number.Integer, '5'), + (Token.Punctuation, ')'), + (Token.Operator, '*'), + (Token.Name, 'gcd'), + (Token.Punctuation, '('), + (Token.Name, 'a'), + (Token.Operator, ','), + (Token.Name, 'b'), + (Token.Punctuation, ')'), + (Token.Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens def test_if_statement(lexer): - fragment = u"""@( 0 > 3 ) ஆனால் + fragment = """@( 0 > 3 ) ஆனால் பதிப்பி "wont print" முடி""" tokens = [ - (Token.Operator, u'@'), - (Token.Punctuation, u'('), - (Token.Text, u' '), - (Token.Literal.Number.Integer, u'0'), - (Token.Text, u' '), - (Token.Operator, u'>'), - (Token.Text, u' '), - (Token.Literal.Number.Integer, u'3'), - (Token.Text, u' '), - (Token.Punctuation, u')'), - (Token.Text, u' '), - (Token.Keyword, u'ஆனால்'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Keyword, u'பதிப்பி'), - (Token.Text, u' '), - (Token.Literal.String, u'"wont print"'), - (Token.Text, u'\n'), - (Token.Keyword, u'முடி'), - (Token.Text, u'\n') + (Token.Operator, '@'), + (Token.Punctuation, '('), + (Token.Text, ' '), + (Token.Literal.Number.Integer, '0'), + (Token.Text, ' '), + (Token.Operator, '>'), + (Token.Text, ' '), + (Token.Literal.Number.Integer, '3'), + (Token.Text, ' '), + (Token.Punctuation, ')'), + (Token.Text, ' '), + (Token.Keyword, 'ஆனால்'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Keyword, 'பதிப்பி'), + (Token.Text, ' '), + (Token.Literal.String, '"wont print"'), + (Token.Text, '\n'), + (Token.Keyword, 'முடி'), + (Token.Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens def test_function(lexer): - fragment = u"""# (C) முத்தையா அண்ணாமலை 2013, 2015 + fragment = """# (C) முத்தையா அண்ணாமலை 2013, 2015 நிரல்பாகம் gcd ( x, y ) மு = max(x,y) q = min(x,y) @@ -95,83 +95,83 @@ def test_function(lexer): முடி\n""" tokens = [ (Token.Comment.Single, - u'# (C) \u0bae\u0bc1\u0ba4\u0bcd\u0ba4\u0bc8\u0baf\u0bbe \u0b85' - u'\u0ba3\u0bcd\u0ba3\u0bbe\u0bae\u0bb2\u0bc8 2013, 2015\n'), - (Token.Keyword, u'நிரல்பாகம்'), - (Token.Text, u' '), - (Token.Name, u'gcd'), - (Token.Text, u' '), - (Token.Punctuation, u'('), - (Token.Text, u' '), - (Token.Name, u'x'), - (Token.Operator, u','), - (Token.Text, u' '), - (Token.Name, u'y'), - (Token.Text, u' '), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), - (Token.Name, u'\u0bae\u0bc1'), - (Token.Text, u' '), - (Token.Operator, u'='), - (Token.Text, u' '), - (Token.Name.Builtin, u'max'), - (Token.Punctuation, u'('), - (Token.Name, u'x'), - (Token.Operator, u','), - (Token.Name, u'y'), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Name, u'q'), - (Token.Text, u' '), - (Token.Operator, u'='), - (Token.Text, u' '), - (Token.Name.Builtin, u'min'), - (Token.Punctuation, u'('), - (Token.Name, u'x'), - (Token.Operator, u','), - (Token.Name, u'y'), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), - (Token.Text, u'\n'), - (Token.Operator, u'@'), - (Token.Punctuation, u'('), - (Token.Text, u' '), - (Token.Name, u'q'), - (Token.Text, u' '), - (Token.Operator, u'=='), - (Token.Text, u' '), - (Token.Literal.Number.Integer, u'0'), - (Token.Text, u' '), - (Token.Punctuation, u')'), - (Token.Text, u' '), - (Token.Keyword, u'ஆனால்'), - (Token.Text, u'\n'), - (Token.Text, u' '), - (Token.Keyword, u'பின்கொடு'), - (Token.Text, u' '), - (Token.Name, u'\u0bae\u0bc1'), - (Token.Text, u'\n'), - (Token.Keyword, u'முடி'), - (Token.Text, u'\n'), - (Token.Keyword, u'\u0baa\u0bbf\u0ba9\u0bcd\u0b95\u0bca\u0b9f\u0bc1'), - (Token.Text, u' '), - (Token.Name, u'gcd'), - (Token.Punctuation, u'('), - (Token.Text, u' '), - (Token.Name, u'\u0bae\u0bc1'), - (Token.Text, u' '), - (Token.Operator, u'-'), - (Token.Text, u' '), - (Token.Name, u'q'), - (Token.Text, u' '), - (Token.Operator, u','), - (Token.Text, u' '), - (Token.Name, u'q'), - (Token.Text, u' '), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), - (Token.Keyword, u'முடி'), # u'\u0bae\u0bc1\u0b9f\u0bbf'), - (Token.Text, u'\n') + '# (C) \u0bae\u0bc1\u0ba4\u0bcd\u0ba4\u0bc8\u0baf\u0bbe \u0b85' + '\u0ba3\u0bcd\u0ba3\u0bbe\u0bae\u0bb2\u0bc8 2013, 2015\n'), + (Token.Keyword, 'நிரல்பாகம்'), + (Token.Text, ' '), + (Token.Name, 'gcd'), + (Token.Text, ' '), + (Token.Punctuation, '('), + (Token.Text, ' '), + (Token.Name, 'x'), + (Token.Operator, ','), + (Token.Text, ' '), + (Token.Name, 'y'), + (Token.Text, ' '), + (Token.Punctuation, ')'), + (Token.Text, '\n'), + (Token.Name, '\u0bae\u0bc1'), + (Token.Text, ' '), + (Token.Operator, '='), + (Token.Text, ' '), + (Token.Name.Builtin, 'max'), + (Token.Punctuation, '('), + (Token.Name, 'x'), + (Token.Operator, ','), + (Token.Name, 'y'), + (Token.Punctuation, ')'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Name, 'q'), + (Token.Text, ' '), + (Token.Operator, '='), + (Token.Text, ' '), + (Token.Name.Builtin, 'min'), + (Token.Punctuation, '('), + (Token.Name, 'x'), + (Token.Operator, ','), + (Token.Name, 'y'), + (Token.Punctuation, ')'), + (Token.Text, '\n'), + (Token.Text, '\n'), + (Token.Operator, '@'), + (Token.Punctuation, '('), + (Token.Text, ' '), + (Token.Name, 'q'), + (Token.Text, ' '), + (Token.Operator, '=='), + (Token.Text, ' '), + (Token.Literal.Number.Integer, '0'), + (Token.Text, ' '), + (Token.Punctuation, ')'), + (Token.Text, ' '), + (Token.Keyword, 'ஆனால்'), + (Token.Text, '\n'), + (Token.Text, ' '), + (Token.Keyword, 'பின்கொடு'), + (Token.Text, ' '), + (Token.Name, '\u0bae\u0bc1'), + (Token.Text, '\n'), + (Token.Keyword, 'முடி'), + (Token.Text, '\n'), + (Token.Keyword, '\u0baa\u0bbf\u0ba9\u0bcd\u0b95\u0bca\u0b9f\u0bc1'), + (Token.Text, ' '), + (Token.Name, 'gcd'), + (Token.Punctuation, '('), + (Token.Text, ' '), + (Token.Name, '\u0bae\u0bc1'), + (Token.Text, ' '), + (Token.Operator, '-'), + (Token.Text, ' '), + (Token.Name, 'q'), + (Token.Text, ' '), + (Token.Operator, ','), + (Token.Text, ' '), + (Token.Name, 'q'), + (Token.Text, ' '), + (Token.Punctuation, ')'), + (Token.Text, '\n'), + (Token.Keyword, 'முடி'), # '\u0bae\u0bc1\u0b9f\u0bbf'), + (Token.Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens diff --git a/tests/test_grammar_notation.py b/tests/test_grammar_notation.py index 0d7e6865..9db46e62 100644 --- a/tests/test_grammar_notation.py +++ b/tests/test_grammar_notation.py @@ -19,21 +19,21 @@ def lexer_peg(): def test_peg_basic(lexer_peg): - fragment = u'rule<-("terminal"/nonterminal/[cls])*\n' + fragment = 'rule<-("terminal"/nonterminal/[cls])*\n' tokens = [ - (Token.Name.Class, u'rule'), - (Token.Operator, u'<-'), - (Token.Punctuation, u'('), - (Token.String.Double, u'"terminal"'), - (Token.Operator, u'/'), - (Token.Name.Class, u'nonterminal'), - (Token.Operator, u'/'), - (Token.Punctuation, u'['), - (Token.String, u'cls'), - (Token.Punctuation, u']'), - (Token.Punctuation, u')'), - (Token.Operator, u'*'), - (Token.Text, u'\n'), + (Token.Name.Class, 'rule'), + (Token.Operator, '<-'), + (Token.Punctuation, '('), + (Token.String.Double, '"terminal"'), + (Token.Operator, '/'), + (Token.Name.Class, 'nonterminal'), + (Token.Operator, '/'), + (Token.Punctuation, '['), + (Token.String, 'cls'), + (Token.Punctuation, ']'), + (Token.Punctuation, ')'), + (Token.Operator, '*'), + (Token.Text, '\n'), ] assert list(lexer_peg.get_tokens(fragment)) == tokens @@ -42,31 +42,31 @@ def test_peg_operators(lexer_peg): # see for example: # - https://github.com/gvanrossum/pegen # - https://nim-lang.org/docs/pegs.html - fragment = u"rule = 'a' | 'b'\n" + fragment = "rule = 'a' | 'b'\n" tokens = [ - (Token.Name.Class, u'rule'), - (Token.Text, u' '), - (Token.Operator, u'='), - (Token.Text, u' '), - (Token.String.Single, u"'a'"), - (Token.Text, u' '), - (Token.Operator, u'|'), - (Token.Text, u' '), - (Token.String.Single, u"'b'"), - (Token.Text, u'\n'), + (Token.Name.Class, 'rule'), + (Token.Text, ' '), + (Token.Operator, '='), + (Token.Text, ' '), + (Token.String.Single, "'a'"), + (Token.Text, ' '), + (Token.Operator, '|'), + (Token.Text, ' '), + (Token.String.Single, "'b'"), + (Token.Text, '\n'), ] assert list(lexer_peg.get_tokens(fragment)) == tokens - fragment = u"rule: 'a' ~ 'b'\n" + fragment = "rule: 'a' ~ 'b'\n" tokens = [ - (Token.Name.Class, u'rule'), - (Token.Operator, u':'), - (Token.Text, u' '), - (Token.String.Single, u"'a'"), - (Token.Text, u' '), - (Token.Operator, u'~'), - (Token.Text, u' '), - (Token.String.Single, u"'b'"), - (Token.Text, u'\n'), + (Token.Name.Class, 'rule'), + (Token.Operator, ':'), + (Token.Text, ' '), + (Token.String.Single, "'a'"), + (Token.Text, ' '), + (Token.Operator, '~'), + (Token.Text, ' '), + (Token.String.Single, "'b'"), + (Token.Text, '\n'), ] assert list(lexer_peg.get_tokens(fragment)) == tokens @@ -76,19 +76,19 @@ def test_peg_modified_strings(lexer_peg): # - http://textx.github.io/Arpeggio/ # - https://nim-lang.org/docs/pegs.html # - https://github.com/erikrose/parsimonious - fragment = u'~"regex" i"insensitive" "multimod"ilx ("not modified")\n' + fragment = '~"regex" i"insensitive" "multimod"ilx ("not modified")\n' tokens = [ # can't handle parsimonious-style regex while ~ is a cut operator - (Token.Operator, u'~'), - (Token.String.Double, u'"regex"'), - (Token.Text, u' '), - (Token.String.Double, u'i"insensitive"'), - (Token.Text, u' '), - (Token.String.Double, u'"multimod"ilx'), - (Token.Text, u' '), - (Token.Punctuation, u'('), - (Token.String.Double, u'"not modified"'), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), + (Token.Operator, '~'), + (Token.String.Double, '"regex"'), + (Token.Text, ' '), + (Token.String.Double, 'i"insensitive"'), + (Token.Text, ' '), + (Token.String.Double, '"multimod"ilx'), + (Token.Text, ' '), + (Token.Punctuation, '('), + (Token.String.Double, '"not modified"'), + (Token.Punctuation, ')'), + (Token.Text, '\n'), ] assert list(lexer_peg.get_tokens(fragment)) == tokens diff --git a/tests/test_html_formatter.py b/tests/test_html_formatter.py index 0a1b7a9a..7a1c563c 100644 --- a/tests/test_html_formatter.py +++ b/tests/test_html_formatter.py @@ -214,9 +214,9 @@ def test_get_background_style_defs_uses_multiple_css_prefixes(): def test_unicode_options(): - fmt = HtmlFormatter(title=u'Föö', - cssclass=u'bär', - cssstyles=u'div:before { content: \'bäz\' }', + fmt = HtmlFormatter(title='Föö', + cssclass='bär', + cssstyles='div:before { content: \'bäz\' }', encoding='utf-8') handle, pathname = tempfile.mkstemp('.html') with os.fdopen(handle, 'w+b') as tfile: diff --git a/tests/test_idris.py b/tests/test_idris.py index 8580bde4..42dcb3cb 100644 --- a/tests/test_idris.py +++ b/tests/test_idris.py @@ -18,48 +18,48 @@ def lexer(): yield IdrisLexer() def test_reserved_word(lexer): - fragment = u'namespace Foobar\n links : String\n links = "abc"' + fragment = 'namespace Foobar\n links : String\n links = "abc"' tokens = [ - (Keyword.Reserved, u'namespace'), - (Text, u' '), - (Keyword.Type, u'Foobar'), - (Text, u'\n'), - (Text, u' '), - (Name.Function, u'links'), - (Text, u' '), - (Operator.Word, u':'), - (Text, u' '), - (Keyword.Type, u'String'), - (Text, u'\n'), - (Text, u' '), - (Text, u' '), - (Text, u'links'), - (Text, u' '), - (Operator.Word, u'='), - (Text, u' '), - (Literal.String, u'"'), - (Literal.String, u'abc'), - (Literal.String, u'"'), - (Text, u'\n') + (Keyword.Reserved, 'namespace'), + (Text, ' '), + (Keyword.Type, 'Foobar'), + (Text, '\n'), + (Text, ' '), + (Name.Function, 'links'), + (Text, ' '), + (Operator.Word, ':'), + (Text, ' '), + (Keyword.Type, 'String'), + (Text, '\n'), + (Text, ' '), + (Text, ' '), + (Text, 'links'), + (Text, ' '), + (Operator.Word, '='), + (Text, ' '), + (Literal.String, '"'), + (Literal.String, 'abc'), + (Literal.String, '"'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens def test_compiler_directive(lexer): - fragment = u'%link C "object.o"\n%name Vect xs' + fragment = '%link C "object.o"\n%name Vect xs' tokens = [ - (Keyword.Reserved, u'%link'), - (Text, u' '), - (Keyword.Type, u'C'), - (Text, u' '), - (Literal.String, u'"'), - (Literal.String, u'object.o'), - (Literal.String, u'"'), - (Text, u'\n'), - (Keyword.Reserved, u'%name'), - (Text, u' '), - (Keyword.Type, u'Vect'), - (Text, u' '), - (Text, u'xs'), - (Text, u'\n') + (Keyword.Reserved, '%link'), + (Text, ' '), + (Keyword.Type, 'C'), + (Text, ' '), + (Literal.String, '"'), + (Literal.String, 'object.o'), + (Literal.String, '"'), + (Text, '\n'), + (Keyword.Reserved, '%name'), + (Text, ' '), + (Keyword.Type, 'Vect'), + (Text, ' '), + (Text, 'xs'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens diff --git a/tests/test_irc_formatter.py b/tests/test_irc_formatter.py index 6dc43b7d..f93bbad4 100644 --- a/tests/test_irc_formatter.py +++ b/tests/test_irc_formatter.py @@ -20,4 +20,4 @@ def test_correct_output(): houtfile = StringIO() hfmt.format(tokensource, houtfile) - assert u'\x0302lambda\x03 x: \x0302123\x03\n' == houtfile.getvalue() + assert '\x0302lambda\x03 x: \x0302123\x03\n' == houtfile.getvalue() diff --git a/tests/test_java.py b/tests/test_java.py index 467a3b72..3baec0ad 100644 --- a/tests/test_java.py +++ b/tests/test_java.py @@ -19,23 +19,23 @@ def lexer(): def test_enhanced_for(lexer): - fragment = u'label:\nfor(String var2: var1) {}\n' + fragment = 'label:\nfor(String var2: var1) {}\n' tokens = [ - (Name.Label, u'label:'), - (Text, u'\n'), - (Keyword, u'for'), - (Punctuation, u'('), - (Name, u'String'), - (Text, u' '), - (Name, u'var2'), - (Punctuation, u':'), - (Text, u' '), - (Name, u'var1'), - (Punctuation, u')'), - (Text, u' '), - (Punctuation, u'{'), - (Punctuation, u'}'), - (Text, u'\n'), + (Name.Label, 'label:'), + (Text, '\n'), + (Keyword, 'for'), + (Punctuation, '('), + (Name, 'String'), + (Text, ' '), + (Name, 'var2'), + (Punctuation, ':'), + (Text, ' '), + (Name, 'var1'), + (Punctuation, ')'), + (Text, ' '), + (Punctuation, '{'), + (Punctuation, '}'), + (Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens diff --git a/tests/test_julia.py b/tests/test_julia.py index e041377b..14bcdee0 100644 --- a/tests/test_julia.py +++ b/tests/test_julia.py @@ -22,38 +22,38 @@ def test_unicode(lexer): """ Test that unicode character, √, in an expression is recognized """ - fragment = u's = \u221a((1/n) * sum(count .^ 2) - mu .^2)\n' + fragment = 's = \u221a((1/n) * sum(count .^ 2) - mu .^2)\n' tokens = [ - (Token.Name, u's'), - (Token.Text, u' '), - (Token.Operator, u'='), - (Token.Text, u' '), - (Token.Operator, u'\u221a'), - (Token.Punctuation, u'('), - (Token.Punctuation, u'('), - (Token.Literal.Number.Integer, u'1'), - (Token.Operator, u'/'), - (Token.Name, u'n'), - (Token.Punctuation, u')'), - (Token.Text, u' '), - (Token.Operator, u'*'), - (Token.Text, u' '), - (Token.Name, u'sum'), - (Token.Punctuation, u'('), - (Token.Name, u'count'), - (Token.Text, u' '), - (Token.Operator, u'.^'), - (Token.Text, u' '), - (Token.Literal.Number.Integer, u'2'), - (Token.Punctuation, u')'), - (Token.Text, u' '), - (Token.Operator, u'-'), - (Token.Text, u' '), - (Token.Name, u'mu'), - (Token.Text, u' '), - (Token.Operator, u'.^'), - (Token.Literal.Number.Integer, u'2'), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), + (Token.Name, 's'), + (Token.Text, ' '), + (Token.Operator, '='), + (Token.Text, ' '), + (Token.Operator, '\u221a'), + (Token.Punctuation, '('), + (Token.Punctuation, '('), + (Token.Literal.Number.Integer, '1'), + (Token.Operator, '/'), + (Token.Name, 'n'), + (Token.Punctuation, ')'), + (Token.Text, ' '), + (Token.Operator, '*'), + (Token.Text, ' '), + (Token.Name, 'sum'), + (Token.Punctuation, '('), + (Token.Name, 'count'), + (Token.Text, ' '), + (Token.Operator, '.^'), + (Token.Text, ' '), + (Token.Literal.Number.Integer, '2'), + (Token.Punctuation, ')'), + (Token.Text, ' '), + (Token.Operator, '-'), + (Token.Text, ' '), + (Token.Name, 'mu'), + (Token.Text, ' '), + (Token.Operator, '.^'), + (Token.Literal.Number.Integer, '2'), + (Token.Punctuation, ')'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens diff --git a/tests/test_kotlin.py b/tests/test_kotlin.py index 2f0eb376..9b9e898d 100644 --- a/tests/test_kotlin.py +++ b/tests/test_kotlin.py @@ -19,115 +19,115 @@ def lexer(): def test_can_cope_with_backtick_names_in_functions(lexer): - fragment = u'fun `wo bble`' + fragment = 'fun `wo bble`' tokens = [ - (Keyword, u'fun'), - (Text, u' '), - (Name.Function, u'`wo bble`'), - (Text, u'\n') + (Keyword, 'fun'), + (Text, ' '), + (Name.Function, '`wo bble`'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens def test_can_cope_with_commas_and_dashes_in_backtick_Names(lexer): - fragment = u'fun `wo,-bble`' + fragment = 'fun `wo,-bble`' tokens = [ - (Keyword, u'fun'), - (Text, u' '), - (Name.Function, u'`wo,-bble`'), - (Text, u'\n') + (Keyword, 'fun'), + (Text, ' '), + (Name.Function, '`wo,-bble`'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens def test_can_cope_with_destructuring(lexer): - fragment = u'val (a, b) = ' + fragment = 'val (a, b) = ' tokens = [ - (Keyword, u'val'), - (Text, u' '), - (Punctuation, u'('), - (Name.Property, u'a'), - (Punctuation, u','), - (Text, u' '), - (Name.Property, u'b'), - (Punctuation, u')'), - (Text, u' '), - (Punctuation, u'='), - (Text, u' '), - (Text, u'\n') + (Keyword, 'val'), + (Text, ' '), + (Punctuation, '('), + (Name.Property, 'a'), + (Punctuation, ','), + (Text, ' '), + (Name.Property, 'b'), + (Punctuation, ')'), + (Text, ' '), + (Punctuation, '='), + (Text, ' '), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens def test_can_cope_generics_in_destructuring(lexer): - fragment = u'val (a: List<Something>, b: Set<Wobble>) =' + fragment = 'val (a: List<Something>, b: Set<Wobble>) =' tokens = [ - (Keyword, u'val'), - (Text, u' '), - (Punctuation, u'('), - (Name.Property, u'a'), - (Punctuation, u':'), - (Text, u' '), - (Name.Property, u'List'), - (Punctuation, u'<'), - (Name, u'Something'), - (Punctuation, u'>'), - (Punctuation, u','), - (Text, u' '), - (Name.Property, u'b'), - (Punctuation, u':'), - (Text, u' '), - (Name.Property, u'Set'), - (Punctuation, u'<'), - (Name, u'Wobble'), - (Punctuation, u'>'), - (Punctuation, u')'), - (Text, u' '), - (Punctuation, u'='), - (Text, u'\n') + (Keyword, 'val'), + (Text, ' '), + (Punctuation, '('), + (Name.Property, 'a'), + (Punctuation, ':'), + (Text, ' '), + (Name.Property, 'List'), + (Punctuation, '<'), + (Name, 'Something'), + (Punctuation, '>'), + (Punctuation, ','), + (Text, ' '), + (Name.Property, 'b'), + (Punctuation, ':'), + (Text, ' '), + (Name.Property, 'Set'), + (Punctuation, '<'), + (Name, 'Wobble'), + (Punctuation, '>'), + (Punctuation, ')'), + (Text, ' '), + (Punctuation, '='), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens def test_can_cope_with_generics(lexer): - fragment = u'inline fun <reified T : ContractState> VaultService.queryBy(): Vault.Page<T> {' + fragment = 'inline fun <reified T : ContractState> VaultService.queryBy(): Vault.Page<T> {' tokens = [ - (Keyword, u'inline fun'), - (Text, u' '), - (Punctuation, u'<'), - (Keyword, u'reified'), - (Text, u' '), - (Name, u'T'), - (Text, u' '), - (Punctuation, u':'), - (Text, u' '), - (Name, u'ContractState'), - (Punctuation, u'>'), - (Text, u' '), - (Name.Class, u'VaultService'), - (Punctuation, u'.'), - (Name.Function, u'queryBy'), - (Punctuation, u'('), - (Punctuation, u')'), - (Punctuation, u':'), - (Text, u' '), - (Name, u'Vault'), - (Punctuation, u'.'), - (Name, u'Page'), - (Punctuation, u'<'), - (Name, u'T'), - (Punctuation, u'>'), - (Text, u' '), - (Punctuation, u'{'), - (Text, u'\n') + (Keyword, 'inline fun'), + (Text, ' '), + (Punctuation, '<'), + (Keyword, 'reified'), + (Text, ' '), + (Name, 'T'), + (Text, ' '), + (Punctuation, ':'), + (Text, ' '), + (Name, 'ContractState'), + (Punctuation, '>'), + (Text, ' '), + (Name.Class, 'VaultService'), + (Punctuation, '.'), + (Name.Function, 'queryBy'), + (Punctuation, '('), + (Punctuation, ')'), + (Punctuation, ':'), + (Text, ' '), + (Name, 'Vault'), + (Punctuation, '.'), + (Name, 'Page'), + (Punctuation, '<'), + (Name, 'T'), + (Punctuation, '>'), + (Text, ' '), + (Punctuation, '{'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens def test_should_cope_with_multiline_comments(lexer): - fragment = u'"""\nthis\nis\na\ncomment"""' + fragment = '"""\nthis\nis\na\ncomment"""' tokens = [ - (String, u'"""\nthis\nis\na\ncomment"""'), - (Text, u'\n') + (String, '"""\nthis\nis\na\ncomment"""'), + (Text, '\n') ] assert list(lexer.get_tokens(fragment)) == tokens diff --git a/tests/test_objectiveclexer.py b/tests/test_objectiveclexer.py index 31f833cf..7264bad8 100644 --- a/tests/test_objectiveclexer.py +++ b/tests/test_objectiveclexer.py @@ -19,78 +19,78 @@ def lexer(): def test_literal_number_int(lexer): - fragment = u'@(1);\n' + fragment = '@(1);\n' expected = [ - (Token.Literal, u'@('), - (Token.Literal.Number.Integer, u'1'), - (Token.Literal, u')'), - (Token.Punctuation, u';'), - (Token.Text, u'\n'), + (Token.Literal, '@('), + (Token.Literal.Number.Integer, '1'), + (Token.Literal, ')'), + (Token.Punctuation, ';'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == expected def test_literal_number_expression(lexer): - fragment = u'@(1+2);\n' + fragment = '@(1+2);\n' expected = [ - (Token.Literal, u'@('), - (Token.Literal.Number.Integer, u'1'), - (Token.Operator, u'+'), - (Token.Literal.Number.Integer, u'2'), - (Token.Literal, u')'), - (Token.Punctuation, u';'), - (Token.Text, u'\n'), + (Token.Literal, '@('), + (Token.Literal.Number.Integer, '1'), + (Token.Operator, '+'), + (Token.Literal.Number.Integer, '2'), + (Token.Literal, ')'), + (Token.Punctuation, ';'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == expected def test_literal_number_nested_expression(lexer): - fragment = u'@(1+(2+3));\n' + fragment = '@(1+(2+3));\n' expected = [ - (Token.Literal, u'@('), - (Token.Literal.Number.Integer, u'1'), - (Token.Operator, u'+'), - (Token.Punctuation, u'('), - (Token.Literal.Number.Integer, u'2'), - (Token.Operator, u'+'), - (Token.Literal.Number.Integer, u'3'), - (Token.Punctuation, u')'), - (Token.Literal, u')'), - (Token.Punctuation, u';'), - (Token.Text, u'\n'), + (Token.Literal, '@('), + (Token.Literal.Number.Integer, '1'), + (Token.Operator, '+'), + (Token.Punctuation, '('), + (Token.Literal.Number.Integer, '2'), + (Token.Operator, '+'), + (Token.Literal.Number.Integer, '3'), + (Token.Punctuation, ')'), + (Token.Literal, ')'), + (Token.Punctuation, ';'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == expected def test_literal_number_bool(lexer): - fragment = u'@NO;\n' + fragment = '@NO;\n' expected = [ - (Token.Literal.Number, u'@NO'), - (Token.Punctuation, u';'), - (Token.Text, u'\n'), + (Token.Literal.Number, '@NO'), + (Token.Punctuation, ';'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == expected def test_literal_number_bool_expression(lexer): - fragment = u'@(YES);\n' + fragment = '@(YES);\n' expected = [ - (Token.Literal, u'@('), - (Token.Name.Builtin, u'YES'), - (Token.Literal, u')'), - (Token.Punctuation, u';'), - (Token.Text, u'\n'), + (Token.Literal, '@('), + (Token.Name.Builtin, 'YES'), + (Token.Literal, ')'), + (Token.Punctuation, ';'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == expected def test_module_import(lexer): - fragment = u'@import ModuleA;\n' + fragment = '@import ModuleA;\n' expected = [ - (Token.Keyword, u'@import'), - (Token.Text, u' '), - (Token.Name, u'ModuleA'), - (Token.Punctuation, u';'), - (Token.Text, u'\n'), + (Token.Keyword, '@import'), + (Token.Text, ' '), + (Token.Name, 'ModuleA'), + (Token.Punctuation, ';'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == expected diff --git a/tests/test_praat.py b/tests/test_praat.py index e3997671..cced8983 100644 --- a/tests/test_praat.py +++ b/tests/test_praat.py @@ -19,187 +19,195 @@ def lexer(): def test_numeric_assignment(lexer): - fragment = u'var = -15e4\n' + fragment = 'var = -15e4\n' tokens = [ - (Token.Text, u'var'), - (Token.Text, u' '), - (Token.Operator, u'='), - (Token.Text, u' '), - (Token.Operator, u'-'), - (Token.Literal.Number, u'15e4'), - (Token.Text, u'\n'), + (Token.Text, 'var'), + (Token.Text, ' '), + (Token.Operator, '='), + (Token.Text, ' '), + (Token.Operator, '-'), + (Token.Literal.Number, '15e4'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def testStringAssignment(lexer): - fragment = u'var$ = "foo"\n' + fragment = 'var$ = "foo"\n' tokens = [ - (Token.Text, u'var$'), - (Token.Text, u' '), - (Token.Operator, u'='), - (Token.Text, u' '), - (Token.Literal.String, u'"'), - (Token.Literal.String, u'foo'), - (Token.Literal.String, u'"'), - (Token.Text, u'\n'), + (Token.Text, 'var$'), + (Token.Text, ' '), + (Token.Operator, '='), + (Token.Text, ' '), + (Token.Literal.String, '"'), + (Token.Literal.String, 'foo'), + (Token.Literal.String, '"'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_string_escaped_quotes(lexer): - fragment = u'"it said ""foo"""\n' + fragment = '"it said ""foo"""\n' tokens = [ - (Token.Literal.String, u'"'), - (Token.Literal.String, u'it said '), - (Token.Literal.String, u'"'), - (Token.Literal.String, u'"'), - (Token.Literal.String, u'foo'), - (Token.Literal.String, u'"'), - (Token.Literal.String, u'"'), - (Token.Literal.String, u'"'), - (Token.Text, u'\n'), + (Token.Literal.String, '"'), + (Token.Literal.String, 'it said '), + (Token.Literal.String, '"'), + (Token.Literal.String, '"'), + (Token.Literal.String, 'foo'), + (Token.Literal.String, '"'), + (Token.Literal.String, '"'), + (Token.Literal.String, '"'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_function_call(lexer): - fragment = u'selected("Sound", i+(a*b))\n' + fragment = 'selected("Sound", i+(a*b))\n' tokens = [ - (Token.Name.Function, u'selected'), - (Token.Punctuation, u'('), - (Token.Literal.String, u'"'), - (Token.Literal.String, u'Sound'), - (Token.Literal.String, u'"'), - (Token.Punctuation, u','), - (Token.Text, u' '), - (Token.Text, u'i'), - (Token.Operator, u'+'), - (Token.Text, u'('), - (Token.Text, u'a'), - (Token.Operator, u'*'), - (Token.Text, u'b'), - (Token.Text, u')'), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), + (Token.Name.Function, 'selected'), + (Token.Punctuation, '('), + (Token.Literal.String, '"'), + (Token.Literal.String, 'Sound'), + (Token.Literal.String, '"'), + (Token.Punctuation, ','), + (Token.Text, ' '), + (Token.Text, 'i'), + (Token.Operator, '+'), + (Token.Text, '('), + (Token.Text, 'a'), + (Token.Operator, '*'), + (Token.Text, 'b'), + (Token.Text, ')'), + (Token.Punctuation, ')'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_broken_unquoted_string(lexer): - fragment = u'printline string\n... \'interpolated\' string\n' + fragment = 'printline string\n... \'interpolated\' string\n' tokens = [ - (Token.Keyword, u'printline'), - (Token.Text, u' '), - (Token.Literal.String, u'string'), - (Token.Text, u'\n'), - (Token.Punctuation, u'...'), - (Token.Text, u' '), - (Token.Literal.String.Interpol, u"'interpolated'"), - (Token.Text, u' '), - (Token.Literal.String, u'string'), - (Token.Text, u'\n'), + (Token.Keyword, 'printline'), + (Token.Text, ' '), + (Token.Literal.String, 'string'), + (Token.Text, '\n'), + (Token.Punctuation, '...'), + (Token.Text, ' '), + (Token.Literal.String.Interpol, "'interpolated'"), + (Token.Text, ' '), + (Token.Literal.String, 'string'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_inline_if(lexer): - fragment = u'var = if true == 1 then -1 else 0 fi' + fragment = 'var = if true == 1 then -1 else 0 fi' tokens = [ - (Token.Text, u'var'), - (Token.Text, u' '), - (Token.Operator, u'='), - (Token.Text, u' '), - (Token.Keyword, u'if'), - (Token.Text, u' '), - (Token.Text, u'true'), - (Token.Text, u' '), - (Token.Operator, u'=='), - (Token.Text, u' '), - (Token.Literal.Number, u'1'), - (Token.Text, u' '), - (Token.Keyword, u'then'), - (Token.Text, u' '), - (Token.Operator, u'-'), - (Token.Literal.Number, u'1'), - (Token.Text, u' '), - (Token.Keyword, u'else'), - (Token.Text, u' '), - (Token.Literal.Number, u'0'), - (Token.Text, u' '), - (Token.Keyword, u'fi'), - (Token.Text, u'\n'), + (Token.Text, 'var'), + (Token.Text, ' '), + (Token.Operator, '='), + (Token.Text, ' '), + (Token.Keyword, 'if'), + (Token.Text, ' '), + (Token.Text, 'true'), + (Token.Text, ' '), + (Token.Operator, '=='), + (Token.Text, ' '), + (Token.Literal.Number, '1'), + (Token.Text, ' '), + (Token.Keyword, 'then'), + (Token.Text, ' '), + (Token.Operator, '-'), + (Token.Literal.Number, '1'), + (Token.Text, ' '), + (Token.Keyword, 'else'), + (Token.Text, ' '), + (Token.Literal.Number, '0'), + (Token.Text, ' '), + (Token.Keyword, 'fi'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens + def test_interpolation_boundary(lexer): - fragment = u'"\'" + "\'"' + fragment = '"\'" + "\'"' tokens = [ - (Token.Literal.String, u'"'), - (Token.Literal.String, u"'"), - (Token.Literal.String, u'"'), - (Token.Text, u' '), - (Token.Operator, u'+'), - (Token.Text, u' '), - (Token.Literal.String, u'"'), - (Token.Literal.String, u"'"), - (Token.Literal.String, u'"'), - (Token.Text, u'\n'), + (Token.Literal.String, '"'), + (Token.Literal.String, "'"), + (Token.Literal.String, '"'), + (Token.Text, ' '), + (Token.Operator, '+'), + (Token.Text, ' '), + (Token.Literal.String, '"'), + (Token.Literal.String, "'"), + (Token.Literal.String, '"'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens + def test_interpolated_numeric_indexed(lexer): - fragment = u"'a[3]'" + fragment = "'a[3]'" tokens = [ - (Token.Literal.String.Interpol, u"'a[3]'"), - (Token.Text, u'\n'), + (Token.Literal.String.Interpol, "'a[3]'"), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens + def test_interpolated_numeric_hash(lexer): - fragment = u"'a[\"b\"]'" + fragment = "'a[\"b\"]'" tokens = [ - (Token.Literal.String.Interpol, u"'a[\"b\"]'"), - (Token.Text, u'\n'), + (Token.Literal.String.Interpol, "'a[\"b\"]'"), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens + def test_interpolated_string_indexed(lexer): - fragment = u"'a$[3]'" + fragment = "'a$[3]'" tokens = [ - (Token.Literal.String.Interpol, u"'a$[3]'"), - (Token.Text, u'\n'), + (Token.Literal.String.Interpol, "'a$[3]'"), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens + def test_interpolated_string_hash(lexer): - fragment = u"'a$[\"b\"]'" + fragment = "'a$[\"b\"]'" tokens = [ - (Token.Literal.String.Interpol, u"'a$[\"b\"]'"), - (Token.Text, u'\n'), + (Token.Literal.String.Interpol, "'a$[\"b\"]'"), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens + def test_interpolated_numeric_with_precision(lexer): - fragment = u"'a:3'" + fragment = "'a:3'" tokens = [ - (Token.Literal.String.Interpol, u"'a:3'"), - (Token.Text, u'\n'), + (Token.Literal.String.Interpol, "'a:3'"), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens + def test_interpolated_indexed_numeric_with_precision(lexer): - fragment = u"'a[3]:3'" + fragment = "'a[3]:3'" tokens = [ - (Token.Literal.String.Interpol, u"'a[3]:3'"), - (Token.Text, u'\n'), + (Token.Literal.String.Interpol, "'a[3]:3'"), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens + def test_interpolated_local_numeric_with_precision(lexer): - fragment = u"'a.a:3'" + fragment = "'a.a:3'" tokens = [ - (Token.Literal.String.Interpol, u"'a.a:3'"), - (Token.Text, u'\n'), + (Token.Literal.String.Interpol, "'a.a:3'"), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens diff --git a/tests/test_promql.py b/tests/test_promql.py index cd02a57a..3b3dce91 100644 --- a/tests/test_promql.py +++ b/tests/test_promql.py @@ -19,7 +19,7 @@ def lexer(): def test_metric(lexer): - fragment = u"go_gc_duration_seconds" + fragment = "go_gc_duration_seconds" tokens = [ (Token.Name.Variable, "go_gc_duration_seconds"), (Token.Text.Whitespace, "\n"), @@ -28,7 +28,7 @@ def test_metric(lexer): def test_metric_one_label(lexer): - fragment = u'go_gc_duration_seconds{instance="localhost:9090"}' + fragment = 'go_gc_duration_seconds{instance="localhost:9090"}' tokens = [ (Token.Name.Variable, "go_gc_duration_seconds"), (Token.Punctuation, "{"), @@ -44,7 +44,7 @@ def test_metric_one_label(lexer): def test_metric_multiple_labels(lexer): - fragment = u'go_gc_duration_seconds{instance="localhost:9090",job="alertmanager"}' + fragment = 'go_gc_duration_seconds{instance="localhost:9090",job="alertmanager"}' tokens = [ (Token.Name.Variable, "go_gc_duration_seconds"), (Token.Punctuation, "{"), @@ -66,7 +66,7 @@ def test_metric_multiple_labels(lexer): def test_metric_multiple_labels_with_spaces(lexer): - fragment = u'go_gc_duration_seconds{ instance="localhost:9090", job="alertmanager" }' + fragment = 'go_gc_duration_seconds{ instance="localhost:9090", job="alertmanager" }' tokens = [ (Token.Name.Variable, "go_gc_duration_seconds"), (Token.Punctuation, "{"), @@ -91,7 +91,7 @@ def test_metric_multiple_labels_with_spaces(lexer): def test_expression_and_comment(lexer): - fragment = u'go_gc_duration_seconds{instance="localhost:9090"} # single comment\n' + fragment = 'go_gc_duration_seconds{instance="localhost:9090"} # single comment\n' tokens = [ (Token.Name.Variable, "go_gc_duration_seconds"), (Token.Punctuation, "{"), @@ -109,7 +109,7 @@ def test_expression_and_comment(lexer): def test_function_delta(lexer): - fragment = u'delta(cpu_temp_celsius{host="zeus"}[2h])' + fragment = 'delta(cpu_temp_celsius{host="zeus"}[2h])' tokens = [ (Token.Keyword.Reserved, "delta"), (Token.Operator, "("), @@ -131,7 +131,7 @@ def test_function_delta(lexer): def test_function_sum_with_args(lexer): - fragment = u"sum by (app, proc) (instance_memory_usage_bytes)\n" + fragment = 'sum by (app, proc) (instance_memory_usage_bytes)\n' tokens = [ (Token.Keyword, "sum"), (Token.Text.Whitespace, " "), @@ -153,7 +153,7 @@ def test_function_sum_with_args(lexer): def test_function_multi_line(lexer): - fragment = u"""label_replace( + fragment = """label_replace( sum by (instance) ( irate(node_disk_read_bytes_total[2m]) ) / 1024 / 1024, @@ -229,7 +229,7 @@ def test_function_multi_line(lexer): def test_function_multi_line_with_offset(lexer): - fragment = u"""label_replace( + fragment = """label_replace( avg by(instance) (irate(node_cpu_seconds_total{mode = "idle"}[5m] offset 3s) ) * 100, diff --git a/tests/test_python.py b/tests/test_python.py index 8e53677b..73993936 100644 --- a/tests/test_python.py +++ b/tests/test_python.py @@ -61,57 +61,57 @@ def test_needs_name(lexer3): """ Tests that '@' is recognized as an Operator """ - fragment = u'S = (H @ beta - r).T @ inv(H @ V @ H.T) @ (H @ beta - r)\n' + fragment = 'S = (H @ beta - r).T @ inv(H @ V @ H.T) @ (H @ beta - r)\n' tokens = [ - (Token.Name, u'S'), - (Token.Text, u' '), - (Token.Operator, u'='), - (Token.Text, u' '), - (Token.Punctuation, u'('), - (Token.Name, u'H'), - (Token.Text, u' '), - (Token.Operator, u'@'), - (Token.Text, u' '), - (Token.Name, u'beta'), - (Token.Text, u' '), - (Token.Operator, u'-'), - (Token.Text, u' '), - (Token.Name, u'r'), - (Token.Punctuation, u')'), - (Token.Operator, u'.'), - (Token.Name, u'T'), - (Token.Text, u' '), - (Token.Operator, u'@'), - (Token.Text, u' '), - (Token.Name, u'inv'), - (Token.Punctuation, u'('), - (Token.Name, u'H'), - (Token.Text, u' '), - (Token.Operator, u'@'), - (Token.Text, u' '), - (Token.Name, u'V'), - (Token.Text, u' '), - (Token.Operator, u'@'), - (Token.Text, u' '), - (Token.Name, u'H'), - (Token.Operator, u'.'), - (Token.Name, u'T'), - (Token.Punctuation, u')'), - (Token.Text, u' '), - (Token.Operator, u'@'), - (Token.Text, u' '), - (Token.Punctuation, u'('), - (Token.Name, u'H'), - (Token.Text, u' '), - (Token.Operator, u'@'), - (Token.Text, u' '), - (Token.Name, u'beta'), - (Token.Text, u' '), - (Token.Operator, u'-'), - (Token.Text, u' '), - (Token.Name, u'r'), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), + (Token.Name, 'S'), + (Token.Text, ' '), + (Token.Operator, '='), + (Token.Text, ' '), + (Token.Punctuation, '('), + (Token.Name, 'H'), + (Token.Text, ' '), + (Token.Operator, '@'), + (Token.Text, ' '), + (Token.Name, 'beta'), + (Token.Text, ' '), + (Token.Operator, '-'), + (Token.Text, ' '), + (Token.Name, 'r'), + (Token.Punctuation, ')'), + (Token.Operator, '.'), + (Token.Name, 'T'), + (Token.Text, ' '), + (Token.Operator, '@'), + (Token.Text, ' '), + (Token.Name, 'inv'), + (Token.Punctuation, '('), + (Token.Name, 'H'), + (Token.Text, ' '), + (Token.Operator, '@'), + (Token.Text, ' '), + (Token.Name, 'V'), + (Token.Text, ' '), + (Token.Operator, '@'), + (Token.Text, ' '), + (Token.Name, 'H'), + (Token.Operator, '.'), + (Token.Name, 'T'), + (Token.Punctuation, ')'), + (Token.Text, ' '), + (Token.Operator, '@'), + (Token.Text, ' '), + (Token.Punctuation, '('), + (Token.Name, 'H'), + (Token.Text, ' '), + (Token.Operator, '@'), + (Token.Text, ' '), + (Token.Name, 'beta'), + (Token.Text, ' '), + (Token.Operator, '-'), + (Token.Text, ' '), + (Token.Name, 'r'), + (Token.Punctuation, ')'), + (Token.Text, '\n'), ] assert list(lexer3.get_tokens(fragment)) == tokens @@ -121,18 +121,18 @@ def test_pep_515(lexer3): Tests that the lexer can parse numeric literals with underscores """ fragments = ( - (Token.Literal.Number.Integer, u'1_000_000'), - (Token.Literal.Number.Float, u'1_000.000_001'), - (Token.Literal.Number.Float, u'1_000e1_000j'), - (Token.Literal.Number.Hex, u'0xCAFE_F00D'), - (Token.Literal.Number.Bin, u'0b_0011_1111_0100_1110'), - (Token.Literal.Number.Oct, u'0o_777_123'), + (Token.Literal.Number.Integer, '1_000_000'), + (Token.Literal.Number.Float, '1_000.000_001'), + (Token.Literal.Number.Float, '1_000e1_000j'), + (Token.Literal.Number.Hex, '0xCAFE_F00D'), + (Token.Literal.Number.Bin, '0b_0011_1111_0100_1110'), + (Token.Literal.Number.Oct, '0o_777_123'), ) for token, fragment in fragments: tokens = [ (token, fragment), - (Token.Text, u'\n'), + (Token.Text, '\n'), ] assert list(lexer3.get_tokens(fragment)) == tokens @@ -141,7 +141,7 @@ def test_walrus_operator(lexer3): """ Tests that ':=' is recognized as an Operator """ - fragment = u'if (a := 2) > 4:' + fragment = 'if (a := 2) > 4:' tokens = [ (Token.Keyword, 'if'), (Token.Text, ' '), @@ -191,7 +191,7 @@ def test_fstring(lexer3): (Token.Literal.String.Interpol, '}'), (Token.Literal.String.Single, '.'), (Token.Literal.String.Single, "'"), - (Token.Text, u'\n') + (Token.Text, '\n') ] ), ( "f'He said his name is {name!r}.'\n", diff --git a/tests/test_qbasiclexer.py b/tests/test_qbasiclexer.py index fb721437..83077e7e 100644 --- a/tests/test_qbasiclexer.py +++ b/tests/test_qbasiclexer.py @@ -19,23 +19,23 @@ def lexer(): def test_keywords_with_dollar(lexer): - fragment = u'DIM x\nx = RIGHT$("abc", 1)\n' + fragment = 'DIM x\nx = RIGHT$("abc", 1)\n' expected = [ - (Token.Keyword.Declaration, u'DIM'), - (Token.Text.Whitespace, u' '), - (Token.Name.Variable.Global, u'x'), - (Token.Text, u'\n'), - (Token.Name.Variable.Global, u'x'), - (Token.Text.Whitespace, u' '), - (Token.Operator, u'='), - (Token.Text.Whitespace, u' '), - (Token.Keyword.Reserved, u'RIGHT$'), - (Token.Punctuation, u'('), - (Token.Literal.String.Double, u'"abc"'), - (Token.Punctuation, u','), - (Token.Text.Whitespace, u' '), - (Token.Literal.Number.Integer.Long, u'1'), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), + (Token.Keyword.Declaration, 'DIM'), + (Token.Text.Whitespace, ' '), + (Token.Name.Variable.Global, 'x'), + (Token.Text, '\n'), + (Token.Name.Variable.Global, 'x'), + (Token.Text.Whitespace, ' '), + (Token.Operator, '='), + (Token.Text.Whitespace, ' '), + (Token.Keyword.Reserved, 'RIGHT$'), + (Token.Punctuation, '('), + (Token.Literal.String.Double, '"abc"'), + (Token.Punctuation, ','), + (Token.Text.Whitespace, ' '), + (Token.Literal.Number.Integer.Long, '1'), + (Token.Punctuation, ')'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == expected diff --git a/tests/test_r.py b/tests/test_r.py index 2814acd7..c243652b 100644 --- a/tests/test_r.py +++ b/tests/test_r.py @@ -19,76 +19,76 @@ def lexer(): def test_call(lexer): - fragment = u'f(1, a)\n' + fragment = 'f(1, a)\n' tokens = [ - (Name.Function, u'f'), - (Punctuation, u'('), - (Token.Literal.Number, u'1'), - (Punctuation, u','), - (Token.Text, u' '), - (Token.Name, u'a'), - (Punctuation, u')'), - (Token.Text, u'\n'), + (Name.Function, 'f'), + (Punctuation, '('), + (Token.Literal.Number, '1'), + (Punctuation, ','), + (Token.Text, ' '), + (Token.Name, 'a'), + (Punctuation, ')'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_name1(lexer): - fragment = u'._a_2.c' + fragment = '._a_2.c' tokens = [ - (Name, u'._a_2.c'), - (Token.Text, u'\n'), + (Name, '._a_2.c'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_name2(lexer): # Invalid names are valid if backticks are used - fragment = u'`.1 blah`' + fragment = '`.1 blah`' tokens = [ - (Name, u'`.1 blah`'), - (Token.Text, u'\n'), + (Name, '`.1 blah`'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_name3(lexer): # Internal backticks can be escaped - fragment = u'`.1 \\` blah`' + fragment = '`.1 \\` blah`' tokens = [ - (Name, u'`.1 \\` blah`'), - (Token.Text, u'\n'), + (Name, '`.1 \\` blah`'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_custom_operator(lexer): - fragment = u'7 % and % 8' + fragment = '7 % and % 8' tokens = [ - (Token.Literal.Number, u'7'), - (Token.Text, u' '), - (Token.Operator, u'% and %'), - (Token.Text, u' '), - (Token.Literal.Number, u'8'), - (Token.Text, u'\n'), + (Token.Literal.Number, '7'), + (Token.Text, ' '), + (Token.Operator, '% and %'), + (Token.Text, ' '), + (Token.Literal.Number, '8'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_indexing(lexer): - fragment = u'a[1]' + fragment = 'a[1]' tokens = [ - (Token.Name, u'a'), - (Token.Punctuation, u'['), - (Token.Literal.Number, u'1'), - (Token.Punctuation, u']'), - (Token.Text, u'\n'), + (Token.Name, 'a'), + (Token.Punctuation, '['), + (Token.Literal.Number, '1'), + (Token.Punctuation, ']'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_dot_name(lexer): - fragment = u'. <- 1' + fragment = '. <- 1' tokens = [ (Token.Name, '.'), (Token.Text, ' '), @@ -101,12 +101,12 @@ def test_dot_name(lexer): def test_dot_indexing(lexer): - fragment = u'.[1]' + fragment = '.[1]' tokens = [ - (Token.Name, u'.'), - (Token.Punctuation, u'['), - (Token.Literal.Number, u'1'), - (Token.Punctuation, u']'), - (Token.Text, u'\n'), + (Token.Name, '.'), + (Token.Punctuation, '['), + (Token.Literal.Number, '1'), + (Token.Punctuation, ']'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens diff --git a/tests/test_regexlexer.py b/tests/test_regexlexer.py index 8e55696c..4029db1d 100644 --- a/tests/test_regexlexer.py +++ b/tests/test_regexlexer.py @@ -48,7 +48,7 @@ def test_tuple(lexer): def test_multiline(lexer): toks = list(lexer.get_tokens_unprocessed('a\ne')) assert toks == [ - (0, Text.Root, 'a'), (1, Text, u'\n'), (2, Text.Root, 'e')] + (0, Text.Root, 'a'), (1, Text, '\n'), (2, Text.Root, 'e')] def test_default(lexer): diff --git a/tests/test_rtf_formatter.py b/tests/test_rtf_formatter.py index 6f60d06e..dc4fffb4 100644 --- a/tests/test_rtf_formatter.py +++ b/tests/test_rtf_formatter.py @@ -27,10 +27,10 @@ def _build_message(*args, **kwargs): result = _escape(kwargs.get('result', '')) if string is None: - string = (u"The expected output of '{t}'\n" - u"\t\tShould be '{expected}'\n" - u"\t\tActually outputs '{result}'\n" - u"\t(WARNING: Partial Output of Result!)") + string = ("The expected output of '{t}'\n" + "\t\tShould be '{expected}'\n" + "\t\tActually outputs '{result}'\n" + "\t(WARNING: Partial Output of Result!)") end = -len(_escape(foot)) start = end - len(expected) @@ -51,31 +51,31 @@ def format_rtf(t): def test_rtf_header(): - t = u'' + t = '' result = format_rtf(t) expected = r'{\rtf1\ansi\uc0' - msg = (u"RTF documents are expected to start with '{expected}'\n" - u"\t\tStarts intead with '{result}'\n" - u"\t(WARNING: Partial Output of Result!)".format( + msg = ("RTF documents are expected to start with '{expected}'\n" + "\t\tStarts intead with '{result}'\n" + "\t(WARNING: Partial Output of Result!)".format( expected=expected, result=result[:len(expected)])) assert result.startswith(expected), msg def test_rtf_footer(): - t = u'' + t = '' result = format_rtf(t) expected = '' - msg = (u"RTF documents are expected to end with '{expected}'\n" - u"\t\tEnds intead with '{result}'\n" - u"\t(WARNING: Partial Output of Result!)".format( + msg = ("RTF documents are expected to end with '{expected}'\n" + "\t\tEnds intead with '{result}'\n" + "\t(WARNING: Partial Output of Result!)".format( expected=_escape(expected), result=_escape(result[-len(expected):]))) assert result.endswith(expected+foot), msg def test_ascii_characters(): - t = u'a b c d ~' + t = 'a b c d ~' result = format_rtf(t) expected = (r'a b c d ~') msg = _build_message(t=t, result=result, expected=expected) @@ -83,7 +83,7 @@ def test_ascii_characters(): def test_escape_characters(): - t = u'\\ {{' + t = '\\ {{' result = format_rtf(t) expected = r'\\ \{\{' msg = _build_message(t=t, result=result, expected=expected) @@ -91,7 +91,7 @@ def test_escape_characters(): def test_single_characters(): - t = u'â € ¤ каждой' + t = 'â € ¤ каждой' result = format_rtf(t) expected = (r'{\u226} {\u8364} {\u164} ' r'{\u1082}{\u1072}{\u1078}{\u1076}{\u1086}{\u1081}') @@ -100,7 +100,7 @@ def test_single_characters(): def test_double_characters(): - t = u'က 힣 ↕ ↕︎ 鼖' + t = 'က 힣 ↕ ↕︎ 鼖' result = format_rtf(t) expected = (r'{\u4096} {\u55203} {\u8597} ' r'{\u8597}{\u65038} {\u55422}{\u56859}') diff --git a/tests/test_ruby.py b/tests/test_ruby.py index 86a9ee77..24e3bef1 100644 --- a/tests/test_ruby.py +++ b/tests/test_ruby.py @@ -19,131 +19,131 @@ def lexer(): def test_range_syntax1(lexer): - fragment = u'1..3\n' + fragment = '1..3\n' tokens = [ - (Number.Integer, u'1'), - (Operator, u'..'), - (Number.Integer, u'3'), - (Text, u'\n'), + (Number.Integer, '1'), + (Operator, '..'), + (Number.Integer, '3'), + (Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_range_syntax2(lexer): - fragment = u'1...3\n' + fragment = '1...3\n' tokens = [ - (Number.Integer, u'1'), - (Operator, u'...'), - (Number.Integer, u'3'), - (Text, u'\n'), + (Number.Integer, '1'), + (Operator, '...'), + (Number.Integer, '3'), + (Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_range_syntax3(lexer): - fragment = u'1 .. 3\n' + fragment = '1 .. 3\n' tokens = [ - (Number.Integer, u'1'), - (Text, u' '), - (Operator, u'..'), - (Text, u' '), - (Number.Integer, u'3'), - (Text, u'\n'), + (Number.Integer, '1'), + (Text, ' '), + (Operator, '..'), + (Text, ' '), + (Number.Integer, '3'), + (Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_interpolation_nested_curly(lexer): fragment = ( - u'"A#{ (3..5).group_by { |x| x/2}.map ' - u'do |k,v| "#{k}" end.join }" + "Z"\n') + '"A#{ (3..5).group_by { |x| x/2}.map ' + 'do |k,v| "#{k}" end.join }" + "Z"\n') tokens = [ - (Token.Literal.String.Double, u'"'), - (Token.Literal.String.Double, u'A'), - (Token.Literal.String.Interpol, u'#{'), - (Token.Text, u' '), - (Token.Punctuation, u'('), - (Token.Literal.Number.Integer, u'3'), - (Token.Operator, u'..'), - (Token.Literal.Number.Integer, u'5'), - (Token.Punctuation, u')'), - (Token.Operator, u'.'), - (Token.Name, u'group_by'), - (Token.Text, u' '), - (Token.Literal.String.Interpol, u'{'), - (Token.Text, u' '), - (Token.Operator, u'|'), - (Token.Name, u'x'), - (Token.Operator, u'|'), - (Token.Text, u' '), - (Token.Name, u'x'), - (Token.Operator, u'/'), - (Token.Literal.Number.Integer, u'2'), - (Token.Literal.String.Interpol, u'}'), - (Token.Operator, u'.'), - (Token.Name, u'map'), - (Token.Text, u' '), - (Token.Keyword, u'do'), - (Token.Text, u' '), - (Token.Operator, u'|'), - (Token.Name, u'k'), - (Token.Punctuation, u','), - (Token.Name, u'v'), - (Token.Operator, u'|'), - (Token.Text, u' '), - (Token.Literal.String.Double, u'"'), - (Token.Literal.String.Interpol, u'#{'), - (Token.Name, u'k'), - (Token.Literal.String.Interpol, u'}'), - (Token.Literal.String.Double, u'"'), - (Token.Text, u' '), - (Token.Keyword, u'end'), - (Token.Operator, u'.'), - (Token.Name, u'join'), - (Token.Text, u' '), - (Token.Literal.String.Interpol, u'}'), - (Token.Literal.String.Double, u'"'), - (Token.Text, u' '), - (Token.Operator, u'+'), - (Token.Text, u' '), - (Token.Literal.String.Double, u'"'), - (Token.Literal.String.Double, u'Z'), - (Token.Literal.String.Double, u'"'), - (Token.Text, u'\n'), + (Token.Literal.String.Double, '"'), + (Token.Literal.String.Double, 'A'), + (Token.Literal.String.Interpol, '#{'), + (Token.Text, ' '), + (Token.Punctuation, '('), + (Token.Literal.Number.Integer, '3'), + (Token.Operator, '..'), + (Token.Literal.Number.Integer, '5'), + (Token.Punctuation, ')'), + (Token.Operator, '.'), + (Token.Name, 'group_by'), + (Token.Text, ' '), + (Token.Literal.String.Interpol, '{'), + (Token.Text, ' '), + (Token.Operator, '|'), + (Token.Name, 'x'), + (Token.Operator, '|'), + (Token.Text, ' '), + (Token.Name, 'x'), + (Token.Operator, '/'), + (Token.Literal.Number.Integer, '2'), + (Token.Literal.String.Interpol, '}'), + (Token.Operator, '.'), + (Token.Name, 'map'), + (Token.Text, ' '), + (Token.Keyword, 'do'), + (Token.Text, ' '), + (Token.Operator, '|'), + (Token.Name, 'k'), + (Token.Punctuation, ','), + (Token.Name, 'v'), + (Token.Operator, '|'), + (Token.Text, ' '), + (Token.Literal.String.Double, '"'), + (Token.Literal.String.Interpol, '#{'), + (Token.Name, 'k'), + (Token.Literal.String.Interpol, '}'), + (Token.Literal.String.Double, '"'), + (Token.Text, ' '), + (Token.Keyword, 'end'), + (Token.Operator, '.'), + (Token.Name, 'join'), + (Token.Text, ' '), + (Token.Literal.String.Interpol, '}'), + (Token.Literal.String.Double, '"'), + (Token.Text, ' '), + (Token.Operator, '+'), + (Token.Text, ' '), + (Token.Literal.String.Double, '"'), + (Token.Literal.String.Double, 'Z'), + (Token.Literal.String.Double, '"'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_operator_methods(lexer): - fragment = u'x.==4\n' + fragment = 'x.==4\n' tokens = [ - (Token.Name, u'x'), - (Token.Operator, u'.'), - (Token.Name.Operator, u'=='), - (Token.Literal.Number.Integer, u'4'), - (Token.Text, u'\n'), + (Token.Name, 'x'), + (Token.Operator, '.'), + (Token.Name.Operator, '=='), + (Token.Literal.Number.Integer, '4'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_escaped_bracestring(lexer): - fragment = u'str.gsub(%r{\\\\\\\\}, "/")\n' + fragment = 'str.gsub(%r{\\\\\\\\}, "/")\n' tokens = [ - (Token.Name, u'str'), - (Token.Operator, u'.'), - (Token.Name, u'gsub'), - (Token.Punctuation, u'('), - (Token.Literal.String.Regex, u'%r{'), - (Token.Literal.String.Regex, u'\\\\'), - (Token.Literal.String.Regex, u'\\\\'), - (Token.Literal.String.Regex, u'}'), - (Token.Punctuation, u','), - (Token.Text, u' '), - (Token.Literal.String.Double, u'"'), - (Token.Literal.String.Double, u'/'), - (Token.Literal.String.Double, u'"'), - (Token.Punctuation, u')'), - (Token.Text, u'\n'), + (Token.Name, 'str'), + (Token.Operator, '.'), + (Token.Name, 'gsub'), + (Token.Punctuation, '('), + (Token.Literal.String.Regex, '%r{'), + (Token.Literal.String.Regex, '\\\\'), + (Token.Literal.String.Regex, '\\\\'), + (Token.Literal.String.Regex, '}'), + (Token.Punctuation, ','), + (Token.Text, ' '), + (Token.Literal.String.Double, '"'), + (Token.Literal.String.Double, '/'), + (Token.Literal.String.Double, '"'), + (Token.Punctuation, ')'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens diff --git a/tests/test_shell.py b/tests/test_shell.py index 79f78b3a..b3600a4e 100644 --- a/tests/test_shell.py +++ b/tests/test_shell.py @@ -34,194 +34,194 @@ def lexer_powershell_session(): def test_curly_no_escape_and_quotes(lexer_bash): - fragment = u'echo "${a//["b"]/}"\n' + fragment = 'echo "${a//["b"]/}"\n' tokens = [ - (Token.Name.Builtin, u'echo'), - (Token.Text, u' '), - (Token.Literal.String.Double, u'"'), - (Token.String.Interpol, u'${'), - (Token.Name.Variable, u'a'), - (Token.Punctuation, u'//['), - (Token.Literal.String.Double, u'"b"'), - (Token.Punctuation, u']/'), - (Token.String.Interpol, u'}'), - (Token.Literal.String.Double, u'"'), - (Token.Text, u'\n'), + (Token.Name.Builtin, 'echo'), + (Token.Text, ' '), + (Token.Literal.String.Double, '"'), + (Token.String.Interpol, '${'), + (Token.Name.Variable, 'a'), + (Token.Punctuation, '//['), + (Token.Literal.String.Double, '"b"'), + (Token.Punctuation, ']/'), + (Token.String.Interpol, '}'), + (Token.Literal.String.Double, '"'), + (Token.Text, '\n'), ] assert list(lexer_bash.get_tokens(fragment)) == tokens def test_curly_with_escape(lexer_bash): - fragment = u'echo ${a//[\\"]/}\n' + fragment = 'echo ${a//[\\"]/}\n' tokens = [ - (Token.Name.Builtin, u'echo'), - (Token.Text, u' '), - (Token.String.Interpol, u'${'), - (Token.Name.Variable, u'a'), - (Token.Punctuation, u'//['), - (Token.Literal.String.Escape, u'\\"'), - (Token.Punctuation, u']/'), - (Token.String.Interpol, u'}'), - (Token.Text, u'\n'), + (Token.Name.Builtin, 'echo'), + (Token.Text, ' '), + (Token.String.Interpol, '${'), + (Token.Name.Variable, 'a'), + (Token.Punctuation, '//['), + (Token.Literal.String.Escape, '\\"'), + (Token.Punctuation, ']/'), + (Token.String.Interpol, '}'), + (Token.Text, '\n'), ] assert list(lexer_bash.get_tokens(fragment)) == tokens def test_parsed_single(lexer_bash): - fragment = u"a=$'abc\\''\n" + fragment = "a=$'abc\\''\n" tokens = [ - (Token.Name.Variable, u'a'), - (Token.Operator, u'='), - (Token.Literal.String.Single, u"$'abc\\''"), - (Token.Text, u'\n'), + (Token.Name.Variable, 'a'), + (Token.Operator, '='), + (Token.Literal.String.Single, "$'abc\\''"), + (Token.Text, '\n'), ] assert list(lexer_bash.get_tokens(fragment)) == tokens def test_short_variable_names(lexer_bash): - fragment = u'x="$"\ny="$_"\nz="$abc"\n' + fragment = 'x="$"\ny="$_"\nz="$abc"\n' tokens = [ # single lone $ - (Token.Name.Variable, u'x'), - (Token.Operator, u'='), - (Token.Literal.String.Double, u'"'), - (Token.Text, u'$'), - (Token.Literal.String.Double, u'"'), - (Token.Text, u'\n'), + (Token.Name.Variable, 'x'), + (Token.Operator, '='), + (Token.Literal.String.Double, '"'), + (Token.Text, '$'), + (Token.Literal.String.Double, '"'), + (Token.Text, '\n'), # single letter shell var - (Token.Name.Variable, u'y'), - (Token.Operator, u'='), - (Token.Literal.String.Double, u'"'), - (Token.Name.Variable, u'$_'), - (Token.Literal.String.Double, u'"'), - (Token.Text, u'\n'), + (Token.Name.Variable, 'y'), + (Token.Operator, '='), + (Token.Literal.String.Double, '"'), + (Token.Name.Variable, '$_'), + (Token.Literal.String.Double, '"'), + (Token.Text, '\n'), # multi-letter user var - (Token.Name.Variable, u'z'), - (Token.Operator, u'='), - (Token.Literal.String.Double, u'"'), - (Token.Name.Variable, u'$abc'), - (Token.Literal.String.Double, u'"'), - (Token.Text, u'\n'), + (Token.Name.Variable, 'z'), + (Token.Operator, '='), + (Token.Literal.String.Double, '"'), + (Token.Name.Variable, '$abc'), + (Token.Literal.String.Double, '"'), + (Token.Text, '\n'), ] assert list(lexer_bash.get_tokens(fragment)) == tokens def test_array_nums(lexer_bash): - fragment = u'a=(1 2 3)\n' + fragment = 'a=(1 2 3)\n' tokens = [ - (Token.Name.Variable, u'a'), - (Token.Operator, u'='), - (Token.Operator, u'('), - (Token.Literal.Number, u'1'), - (Token.Text, u' '), - (Token.Literal.Number, u'2'), - (Token.Text, u' '), - (Token.Literal.Number, u'3'), - (Token.Operator, u')'), - (Token.Text, u'\n'), + (Token.Name.Variable, 'a'), + (Token.Operator, '='), + (Token.Operator, '('), + (Token.Literal.Number, '1'), + (Token.Text, ' '), + (Token.Literal.Number, '2'), + (Token.Text, ' '), + (Token.Literal.Number, '3'), + (Token.Operator, ')'), + (Token.Text, '\n'), ] assert list(lexer_bash.get_tokens(fragment)) == tokens def test_end_of_line_nums(lexer_bash): - fragment = u'a=1\nb=2 # comment\n' + fragment = 'a=1\nb=2 # comment\n' tokens = [ - (Token.Name.Variable, u'a'), - (Token.Operator, u'='), - (Token.Literal.Number, u'1'), - (Token.Text, u'\n'), - (Token.Name.Variable, u'b'), - (Token.Operator, u'='), - (Token.Literal.Number, u'2'), - (Token.Text, u' '), - (Token.Comment.Single, u'# comment\n'), + (Token.Name.Variable, 'a'), + (Token.Operator, '='), + (Token.Literal.Number, '1'), + (Token.Text, '\n'), + (Token.Name.Variable, 'b'), + (Token.Operator, '='), + (Token.Literal.Number, '2'), + (Token.Text, ' '), + (Token.Comment.Single, '# comment\n'), ] assert list(lexer_bash.get_tokens(fragment)) == tokens def test_newline_in_echo(lexer_session): - fragment = u'$ echo \\\nhi\nhi\n' + fragment = '$ echo \\\nhi\nhi\n' tokens = [ - (Token.Text, u''), - (Token.Generic.Prompt, u'$'), - (Token.Text, u' '), - (Token.Name.Builtin, u'echo'), - (Token.Text, u' '), - (Token.Literal.String.Escape, u'\\\n'), - (Token.Text, u'hi'), - (Token.Text, u'\n'), - (Token.Generic.Output, u'hi\n'), + (Token.Text, ''), + (Token.Generic.Prompt, '$'), + (Token.Text, ' '), + (Token.Name.Builtin, 'echo'), + (Token.Text, ' '), + (Token.Literal.String.Escape, '\\\n'), + (Token.Text, 'hi'), + (Token.Text, '\n'), + (Token.Generic.Output, 'hi\n'), ] assert list(lexer_session.get_tokens(fragment)) == tokens def test_msdos_gt_only(lexer_msdos): - fragment = u'> py\nhi\n' + fragment = '> py\nhi\n' tokens = [ - (Token.Text, u''), - (Token.Generic.Prompt, u'>'), - (Token.Text, u' '), - (Token.Text, u'py'), - (Token.Text, u''), - (Token.Text, u'\n'), - (Token.Generic.Output, u'hi\n'), + (Token.Text, ''), + (Token.Generic.Prompt, '>'), + (Token.Text, ' '), + (Token.Text, 'py'), + (Token.Text, ''), + (Token.Text, '\n'), + (Token.Generic.Output, 'hi\n'), ] assert list(lexer_msdos.get_tokens(fragment)) == tokens def test_powershell_session(lexer_powershell_session): - fragment = u'PS C:\\> Get-ChildItem\n' + fragment = 'PS C:\\> Get-ChildItem\n' tokens = [ - (Token.Name.Builtin, u''), - (Token.Generic.Prompt, u'PS C:\\> '), - (Token.Name.Builtin, u'Get-ChildItem'), - (Token.Text, u'\n') + (Token.Name.Builtin, ''), + (Token.Generic.Prompt, 'PS C:\\> '), + (Token.Name.Builtin, 'Get-ChildItem'), + (Token.Text, '\n') ] assert list(lexer_powershell_session.get_tokens(fragment)) == tokens - fragment = u'PS> Get-ChildItem\n' + fragment = 'PS> Get-ChildItem\n' tokens = [ - (Token.Name.Builtin, u''), - (Token.Generic.Prompt, u'PS> '), - (Token.Name.Builtin, u'Get-ChildItem'), - (Token.Text, u'\n') + (Token.Name.Builtin, ''), + (Token.Generic.Prompt, 'PS> '), + (Token.Name.Builtin, 'Get-ChildItem'), + (Token.Text, '\n') ] assert list(lexer_powershell_session.get_tokens(fragment)) == tokens - fragment = u'PS > Get-ChildItem\n' + fragment = 'PS > Get-ChildItem\n' tokens = [ - (Token.Name.Builtin, u''), - (Token.Generic.Prompt, u'PS > '), - (Token.Name.Builtin, u'Get-ChildItem'), - (Token.Text, u'\n') + (Token.Name.Builtin, ''), + (Token.Generic.Prompt, 'PS > '), + (Token.Name.Builtin, 'Get-ChildItem'), + (Token.Text, '\n') ] assert list(lexer_powershell_session.get_tokens(fragment)) == tokens def test_powershell_remoting_session(lexer_powershell_session): - fragment = u'[Long-NetBIOS-Hostname]: PS C:\\> Get-ChildItem\n' + fragment = '[Long-NetBIOS-Hostname]: PS C:\\> Get-ChildItem\n' tokens = [ - (Token.Name.Builtin, u''), - (Token.Generic.Prompt, u'[Long-NetBIOS-Hostname]: PS C:\\> '), - (Token.Name.Builtin, u'Get-ChildItem'), - (Token.Text, u'\n') + (Token.Name.Builtin, ''), + (Token.Generic.Prompt, '[Long-NetBIOS-Hostname]: PS C:\\> '), + (Token.Name.Builtin, 'Get-ChildItem'), + (Token.Text, '\n') ] assert list(lexer_powershell_session.get_tokens(fragment)) == tokens def test_virtualenv(lexer_session): - fragment = u'(env) [~/project]$ foo -h\n' + fragment = '(env) [~/project]$ foo -h\n' tokens = [ - (Token.Text, u''), - (Token.Generic.Prompt.VirtualEnv, u'(env)'), - (Token.Text, u''), - (Token.Text, u' '), - (Token.Text, u''), - (Token.Generic.Prompt, u'[~/project]$'), - (Token.Text, u' '), - (Token.Text, u'foo'), - (Token.Text, u' '), - (Token.Text, u'-h'), - (Token.Text, u'\n'), + (Token.Text, ''), + (Token.Generic.Prompt.VirtualEnv, '(env)'), + (Token.Text, ''), + (Token.Text, ' '), + (Token.Text, ''), + (Token.Generic.Prompt, '[~/project]$'), + (Token.Text, ' '), + (Token.Text, 'foo'), + (Token.Text, ' '), + (Token.Text, '-h'), + (Token.Text, '\n'), ] assert list(lexer_session.get_tokens(fragment)) == tokens diff --git a/tests/test_smarty.py b/tests/test_smarty.py index 2b45abee..caf23622 100644 --- a/tests/test_smarty.py +++ b/tests/test_smarty.py @@ -19,21 +19,21 @@ def lexer(): def test_nested_curly(lexer): - fragment = u'{templateFunction param={anotherFunction} param2=$something}\n' + fragment = '{templateFunction param={anotherFunction} param2=$something}\n' tokens = [ - (Token.Comment.Preproc, u'{'), - (Token.Name.Function, u'templateFunction'), - (Token.Text, u' '), - (Token.Name.Attribute, u'param'), - (Token.Operator, u'='), - (Token.Comment.Preproc, u'{'), - (Token.Name.Attribute, u'anotherFunction'), - (Token.Comment.Preproc, u'}'), - (Token.Text, u' '), - (Token.Name.Attribute, u'param2'), - (Token.Operator, u'='), - (Token.Name.Variable, u'$something'), - (Token.Comment.Preproc, u'}'), - (Token.Other, u'\n'), + (Token.Comment.Preproc, '{'), + (Token.Name.Function, 'templateFunction'), + (Token.Text, ' '), + (Token.Name.Attribute, 'param'), + (Token.Operator, '='), + (Token.Comment.Preproc, '{'), + (Token.Name.Attribute, 'anotherFunction'), + (Token.Comment.Preproc, '}'), + (Token.Text, ' '), + (Token.Name.Attribute, 'param2'), + (Token.Operator, '='), + (Token.Name.Variable, '$something'), + (Token.Comment.Preproc, '}'), + (Token.Other, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens diff --git a/tests/test_sql.py b/tests/test_sql.py index ead06de7..bd71d1d0 100644 --- a/tests/test_sql.py +++ b/tests/test_sql.py @@ -62,7 +62,7 @@ def test_can_lex_integer(lexer): def test_can_lex_names(lexer): _assert_are_tokens_of_type(lexer, - u'thingy thingy123 _thingy _ _123 Ähnliches Müll #temp1 ##temp2', + 'thingy thingy123 _thingy _ _123 Ähnliches Müll #temp1 ##temp2', Name) diff --git a/tests/test_textfmts.py b/tests/test_textfmts.py index f4ce9b33..b8e25b51 100644 --- a/tests/test_textfmts.py +++ b/tests/test_textfmts.py @@ -19,62 +19,62 @@ def lexer(): def test_http_status_line(lexer): - fragment = u'HTTP/1.1 200 OK\n' + fragment = 'HTTP/1.1 200 OK\n' tokens = [ - (Token.Keyword.Reserved, u'HTTP'), - (Token.Operator, u'/'), - (Token.Number, u'1.1'), - (Token.Text, u' '), - (Token.Number, u'200'), - (Token.Text, u' '), - (Token.Name.Exception, u'OK'), - (Token.Text, u'\n'), + (Token.Keyword.Reserved, 'HTTP'), + (Token.Operator, '/'), + (Token.Number, '1.1'), + (Token.Text, ' '), + (Token.Number, '200'), + (Token.Text, ' '), + (Token.Name.Exception, 'OK'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_http_status_line_without_reason_phrase(lexer): - fragment = u'HTTP/1.1 200\n' + fragment = 'HTTP/1.1 200\n' tokens = [ - (Token.Keyword.Reserved, u'HTTP'), - (Token.Operator, u'/'), - (Token.Number, u'1.1'), - (Token.Text, u' '), - (Token.Number, u'200'), - (Token.Text, u'\n'), + (Token.Keyword.Reserved, 'HTTP'), + (Token.Operator, '/'), + (Token.Number, '1.1'), + (Token.Text, ' '), + (Token.Number, '200'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_http_status_line_without_reason_phrase_rfc_7230(lexer): - fragment = u'HTTP/1.1 200 \n' + fragment = 'HTTP/1.1 200 \n' tokens = [ - (Token.Keyword.Reserved, u'HTTP'), - (Token.Operator, u'/'), - (Token.Number, u'1.1'), - (Token.Text, u' '), - (Token.Number, u'200'), - (Token.Text, u' '), - (Token.Text, u'\n'), + (Token.Keyword.Reserved, 'HTTP'), + (Token.Operator, '/'), + (Token.Number, '1.1'), + (Token.Text, ' '), + (Token.Number, '200'), + (Token.Text, ' '), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens def test_application_xml(lexer): - fragment = u'GET / HTTP/1.0\nContent-Type: application/xml\n\n<foo>\n' + fragment = 'GET / HTTP/1.0\nContent-Type: application/xml\n\n<foo>\n' tokens = [ - (Token.Name.Tag, u'<foo'), - (Token.Name.Tag, u'>'), - (Token.Text, u'\n'), + (Token.Name.Tag, '<foo'), + (Token.Name.Tag, '>'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment))[-len(tokens):] == tokens def test_application_calendar_xml(lexer): - fragment = u'GET / HTTP/1.0\nContent-Type: application/calendar+xml\n\n<foo>\n' + fragment = 'GET / HTTP/1.0\nContent-Type: application/calendar+xml\n\n<foo>\n' tokens = [ - (Token.Name.Tag, u'<foo'), - (Token.Name.Tag, u'>'), - (Token.Text, u'\n'), + (Token.Name.Tag, '<foo'), + (Token.Name.Tag, '>'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment))[-len(tokens):] == tokens diff --git a/tests/test_usd.py b/tests/test_usd.py index 954820b5..8edebeab 100644..100755 --- a/tests/test_usd.py +++ b/tests/test_usd.py @@ -217,66 +217,66 @@ class Features(_Common): self.assertEqual( [ - (Keyword.Token, u"custom"), - (Whitespace, u" "), - (Keyword.Type, u"int[]"), - (Whitespace, u" "), - (Name.Attribute, u"foo"), - (Whitespace, u" "), - (Operator, u"="), - (Whitespace, u" "), - (Punctuation, u"["), - (Number, u"8"), - (Punctuation, u","), - (Whitespace, u" "), - (Number, u"10"), - (Punctuation, u","), - (Whitespace, u" "), - (Number, u"14"), - (Punctuation, u"]"), - (Whitespace, u"\n"), - (Keyword.Token, u"custom"), - (Whitespace, u" "), - (Keyword.Type, u"int[]"), - (Whitespace, u" "), - (Name.Attribute, u"foo"), - (Text, u"."), - (Name.Keyword.Tokens, u"timeSamples"), - (Whitespace, u" "), - (Operator, u"="), - (Whitespace, u" "), - (Punctuation, u"{"), - (Whitespace, u"\n "), - (Number, u"1"), - (Punctuation, u":"), - (Whitespace, u" "), - (Punctuation, u"["), - (Number, u"8"), - (Punctuation, u","), - (Whitespace, u" "), - (Number, u"0"), - (Punctuation, u","), - (Whitespace, u" "), - (Number, u"14"), - (Punctuation, u"]"), - (Punctuation, u","), - (Whitespace, u"\n "), - (Number, u"2"), - (Punctuation, u":"), - (Whitespace, u" "), - (Punctuation, u"["), - (Number, u"-8"), - (Punctuation, u","), - (Whitespace, u" "), - (Number, u"0"), - (Punctuation, u","), - (Whitespace, u" "), - (Number, u"14"), - (Punctuation, u"]"), - (Punctuation, u","), - (Whitespace, u"\n"), - (Punctuation, u"}"), - (Whitespace, u"\n"), + (Keyword.Token, "custom"), + (Whitespace, " "), + (Keyword.Type, "int[]"), + (Whitespace, " "), + (Name.Attribute, "foo"), + (Whitespace, " "), + (Operator, "="), + (Whitespace, " "), + (Punctuation, "["), + (Number, "8"), + (Punctuation, ","), + (Whitespace, " "), + (Number, "10"), + (Punctuation, ","), + (Whitespace, " "), + (Number, "14"), + (Punctuation, "]"), + (Whitespace, "\n"), + (Keyword.Token, "custom"), + (Whitespace, " "), + (Keyword.Type, "int[]"), + (Whitespace, " "), + (Name.Attribute, "foo"), + (Text, "."), + (Name.Keyword.Tokens, "timeSamples"), + (Whitespace, " "), + (Operator, "="), + (Whitespace, " "), + (Punctuation, "{"), + (Whitespace, "\n "), + (Number, "1"), + (Punctuation, ":"), + (Whitespace, " "), + (Punctuation, "["), + (Number, "8"), + (Punctuation, ","), + (Whitespace, " "), + (Number, "0"), + (Punctuation, ","), + (Whitespace, " "), + (Number, "14"), + (Punctuation, "]"), + (Punctuation, ","), + (Whitespace, "\n "), + (Number, "2"), + (Punctuation, ":"), + (Whitespace, " "), + (Punctuation, "["), + (Number, "-8"), + (Punctuation, ","), + (Whitespace, " "), + (Number, "0"), + (Punctuation, ","), + (Whitespace, " "), + (Number, "14"), + (Punctuation, "]"), + (Punctuation, ","), + (Whitespace, "\n"), + (Punctuation, "}"), + (Whitespace, "\n"), ], self._get(timesamples), ) @@ -292,8 +292,8 @@ class Features(_Common): self.assertEqual( [ - (String, u'"""\ncustom int[] foo = [8, 10, 14]\n"""'), - (Whitespace, u"\n"), + (String, '"""\ncustom int[] foo = [8, 10, 14]\n"""'), + (Whitespace, "\n"), ], self._get(code), ) @@ -304,23 +304,23 @@ class Features(_Common): self.assertEqual( [ - (Number, u"8"), - (Whitespace, u" "), - (Number, u"8.0123312132"), - (Punctuation, u","), - (Whitespace, u" "), - (Number, u"-4"), - (Whitespace, u" "), - (Number, u"-14.123"), - (Whitespace, u" "), - (Number, u"1e10"), - (Whitespace, u" "), - (Number, u"0.1e10"), - (Whitespace, u" "), - (Number, u"10.123e+10"), - (Whitespace, u" "), - (Number, u"0.123e-14"), - (Whitespace, u"\n"), + (Number, "8"), + (Whitespace, " "), + (Number, "8.0123312132"), + (Punctuation, ","), + (Whitespace, " "), + (Number, "-4"), + (Whitespace, " "), + (Number, "-14.123"), + (Whitespace, " "), + (Number, "1e10"), + (Whitespace, " "), + (Number, "0.1e10"), + (Whitespace, " "), + (Number, "10.123e+10"), + (Whitespace, " "), + (Number, "0.123e-14"), + (Whitespace, "\n"), ], self._get(code), ) @@ -349,86 +349,86 @@ class Features(_Common): self.assertEqual( [ - (Keyword.Tokens, u"def"), - (Whitespace, u" "), - (Text, u"Xform"), - (Whitespace, u" "), - (String, u'"BottleMedical"'), - (Whitespace, u" "), - (Punctuation, u"("), - (Whitespace, u"\n "), - (Name.Builtins, u"kind"), - (Whitespace, u" "), - (Operator, u"="), - (Whitespace, u" "), - (String, u'"prop"'), - (Whitespace, u"\n "), - (Keyword.Tokens, u"payload"), - (Whitespace, u" "), - (Operator, u"="), - (Whitespace, u" "), - (String.Interpol, u"@./BottleMedical_payload.usd@"), - (Name.Namespace, u"</BottleMedical>"), - (Whitespace, u"\n "), - (Keyword.Tokens, u"variants"), - (Whitespace, u" "), - (Operator, u"="), - (Whitespace, u" "), - (Punctuation, u"{"), - (Whitespace, u"\n "), - (Keyword.Type, u"string"), - (Whitespace, u" "), - (Name.Attribute, u"modelingVariant"), - (Whitespace, u" "), - (Operator, u"="), - (Whitespace, u" "), - (String, u'"LiquidBottleLg"'), - (Whitespace, u"\n "), - (Keyword.Type, u"string"), - (Whitespace, u" "), - (Name.Attribute, u"shadingComplexity"), - (Whitespace, u" "), - (Operator, u"="), - (Whitespace, u" "), - (String, u'"full"'), - (Whitespace, u"\n "), - (Punctuation, u"}"), - (Whitespace, u"\n "), - (Keyword.Type, u"add"), - (Text.Whitespace, u" "), - (Name.Attribute, u"variantSets"), - (Text.Whitespace, u" "), - (Operator, u"="), - (Whitespace, u" "), - (Punctuation, u"["), - (String, u'"modelingVariant"'), - (Punctuation, u","), - (Whitespace, u" "), - (String, u'"shadingComplexity"'), - (Punctuation, u"]"), - (Whitespace, u"\n"), - (Punctuation, u")"), - (Whitespace, u"\n"), - (Punctuation, u"{"), - (Whitespace, u"\n "), - (Keyword.Tokens, u"variantSet"), - (Whitespace, u" "), - (String, u'"modelingVariant"'), - (Whitespace, u" "), - (Operator, u"="), - (Whitespace, u" "), - (Punctuation, u"{"), - (Whitespace, u"\n "), - (String, u'"ALL_VARIANTS"'), - (Whitespace, u" "), - (Punctuation, u"{"), - (Whitespace, u"\n "), - (Punctuation, u"}"), - (Whitespace, u"\n "), - (Punctuation, u"}"), - (Whitespace, u"\n"), - (Punctuation, u"}"), - (Whitespace, u"\n"), + (Keyword.Tokens, "def"), + (Whitespace, " "), + (Text, "Xform"), + (Whitespace, " "), + (String, '"BottleMedical"'), + (Whitespace, " "), + (Punctuation, "("), + (Whitespace, "\n "), + (Name.Builtins, "kind"), + (Whitespace, " "), + (Operator, "="), + (Whitespace, " "), + (String, '"prop"'), + (Whitespace, "\n "), + (Keyword.Tokens, "payload"), + (Whitespace, " "), + (Operator, "="), + (Whitespace, " "), + (String.Interpol, "@./BottleMedical_payload.usd@"), + (Name.Namespace, "</BottleMedical>"), + (Whitespace, "\n "), + (Keyword.Tokens, "variants"), + (Whitespace, " "), + (Operator, "="), + (Whitespace, " "), + (Punctuation, "{"), + (Whitespace, "\n "), + (Keyword.Type, "string"), + (Whitespace, " "), + (Name.Attribute, "modelingVariant"), + (Whitespace, " "), + (Operator, "="), + (Whitespace, " "), + (String, '"LiquidBottleLg"'), + (Whitespace, "\n "), + (Keyword.Type, "string"), + (Whitespace, " "), + (Name.Attribute, "shadingComplexity"), + (Whitespace, " "), + (Operator, "="), + (Whitespace, " "), + (String, '"full"'), + (Whitespace, "\n "), + (Punctuation, "}"), + (Whitespace, "\n "), + (Keyword.Type, "add"), + (Text.Whitespace, " "), + (Name.Attribute, "variantSets"), + (Text.Whitespace, " "), + (Operator, "="), + (Whitespace, " "), + (Punctuation, "["), + (String, '"modelingVariant"'), + (Punctuation, ","), + (Whitespace, " "), + (String, '"shadingComplexity"'), + (Punctuation, "]"), + (Whitespace, "\n"), + (Punctuation, ")"), + (Whitespace, "\n"), + (Punctuation, "{"), + (Whitespace, "\n "), + (Keyword.Tokens, "variantSet"), + (Whitespace, " "), + (String, '"modelingVariant"'), + (Whitespace, " "), + (Operator, "="), + (Whitespace, " "), + (Punctuation, "{"), + (Whitespace, "\n "), + (String, '"ALL_VARIANTS"'), + (Whitespace, " "), + (Punctuation, "{"), + (Whitespace, "\n "), + (Punctuation, "}"), + (Whitespace, "\n "), + (Punctuation, "}"), + (Whitespace, "\n"), + (Punctuation, "}"), + (Whitespace, "\n"), ], self._get(code), ) @@ -438,13 +438,13 @@ class Features(_Common): code = '"Some \'text"' self.assertEqual( - [(String, code), (Whitespace, u"\n")], self._get(code), + [(String, code), (Whitespace, "\n")], self._get(code), ) def test_string_multiple_line(self): """Check that different multi-line strings work correctly.""" code1 = textwrap.dedent( - u'''\ + '''\ """ Some text multiline """''' @@ -455,7 +455,7 @@ class Features(_Common): ) code2 = textwrap.dedent( - u'''\ + '''\ """Some text multiline """''' ) @@ -465,7 +465,7 @@ class Features(_Common): ) code3 = textwrap.dedent( - u'''\ + '''\ """ Some text multiline"""''' ) @@ -491,32 +491,32 @@ class EdgeCases(_Common): self.assertEqual( [ - (Keyword.Type, u"float[]"), - (Whitespace, u" "), - (Name.Attribute, u"primvars:skel:jointWeights"), - (Whitespace, u" "), - (Operator, u"="), - (Whitespace, u" "), - (Punctuation, u"["), - (Number, u"1"), - (Punctuation, u"]"), - (Whitespace, u" "), - (Punctuation, u"("), - (Whitespace, u"\n "), - (Name.Builtins, u"elementSize"), - (Whitespace, u" "), - (Operator, u"="), - (Whitespace, u" "), - (Number, u"1"), - (Whitespace, u"\n "), - (Name.Builtins, u"interpolation"), - (Whitespace, u" "), - (Operator, u"="), - (Whitespace, u" "), - (String, u'"constant"'), - (Whitespace, u"\n"), - (Punctuation, u")"), - (Whitespace, u"\n"), + (Keyword.Type, "float[]"), + (Whitespace, " "), + (Name.Attribute, "primvars:skel:jointWeights"), + (Whitespace, " "), + (Operator, "="), + (Whitespace, " "), + (Punctuation, "["), + (Number, "1"), + (Punctuation, "]"), + (Whitespace, " "), + (Punctuation, "("), + (Whitespace, "\n "), + (Name.Builtins, "elementSize"), + (Whitespace, " "), + (Operator, "="), + (Whitespace, " "), + (Number, "1"), + (Whitespace, "\n "), + (Name.Builtins, "interpolation"), + (Whitespace, " "), + (Operator, "="), + (Whitespace, " "), + (String, '"constant"'), + (Whitespace, "\n"), + (Punctuation, ")"), + (Whitespace, "\n"), ], self._get(code), ) @@ -527,14 +527,14 @@ class EdgeCases(_Common): self.assertEqual( [ - (String.Interpol, u"@firststring@"), - (Whitespace, u" "), - (Text, u"something"), - (Whitespace, u" "), - (Text, u"else"), - (Whitespace, u" "), - (String.Interpol, u"@secondstring@"), - (Whitespace, u"\n"), + (String.Interpol, "@firststring@"), + (Whitespace, " "), + (Text, "something"), + (Whitespace, " "), + (Text, "else"), + (Whitespace, " "), + (String.Interpol, "@secondstring@"), + (Whitespace, "\n"), ], self._get(at_sign), ) @@ -543,14 +543,14 @@ class EdgeCases(_Common): self.assertEqual( [ - (String, u"'firststring'"), - (Whitespace, u" "), - (Text, u"something"), - (Whitespace, u" "), - (Text, u"else"), - (Whitespace, u" "), - (String, u"'secondstring'"), - (Whitespace, u"\n"), + (String, "'firststring'"), + (Whitespace, " "), + (Text, "something"), + (Whitespace, " "), + (Text, "else"), + (Whitespace, " "), + (String, "'secondstring'"), + (Whitespace, "\n"), ], self._get(single), ) @@ -559,14 +559,14 @@ class EdgeCases(_Common): self.assertEqual( [ - (String, u"'firststring'"), - (Whitespace, u" "), - (Text, u"something"), - (Whitespace, u" "), - (Text, u"else"), - (Whitespace, u" "), - (String, u"'secondstring'"), - (Whitespace, u"\n"), + (String, "'firststring'"), + (Whitespace, " "), + (Text, "something"), + (Whitespace, " "), + (Text, "else"), + (Whitespace, " "), + (String, "'secondstring'"), + (Whitespace, "\n"), ], self._get(double), ) diff --git a/tests/test_util.py b/tests/test_util.py index 81b3b054..b90397ea 100644 --- a/tests/test_util.py +++ b/tests/test_util.py @@ -127,23 +127,6 @@ def test_xml(): assert not util.looks_like_xml('<html>') -def test_unirange(): - first_non_bmp = u'\U00010000' - r = re.compile(util.unirange(0x10000, 0x20000)) - m = r.match(first_non_bmp) - assert m - assert m.end() == len(first_non_bmp) - assert not r.match(u'\uffff') - assert not r.match(u'xxx') - # Tests that end is inclusive - r = re.compile(util.unirange(0x10000, 0x10000) + '+') - # Tests that the plus works for the entire unicode point, if narrow - # build - m = r.match(first_non_bmp * 2) - assert m - assert m.end() == len(first_non_bmp) * 2 - - def test_format_lines(): lst = ['cat', 'dog'] output = util.format_lines('var', lst) @@ -173,8 +156,8 @@ def test_duplicates_removed_nonconsecutive(): def test_guess_decode(): # UTF-8 should be decoded as UTF-8 - s = util.guess_decode(u'\xff'.encode('utf-8')) - assert s == (u'\xff', 'utf-8') + s = util.guess_decode('\xff'.encode('utf-8')) + assert s == ('\xff', 'utf-8') # otherwise, it could be latin1 or the locale encoding... import locale @@ -186,11 +169,11 @@ def test_guess_decode_from_terminal(): class Term: encoding = 'utf-7' - s = util.guess_decode_from_terminal(u'\xff'.encode('utf-7'), Term) - assert s == (u'\xff', 'utf-7') + s = util.guess_decode_from_terminal('\xff'.encode('utf-7'), Term) + assert s == ('\xff', 'utf-7') - s = util.guess_decode_from_terminal(u'\xff'.encode('utf-8'), Term) - assert s == (u'\xff', 'utf-8') + s = util.guess_decode_from_terminal('\xff'.encode('utf-8'), Term) + assert s == ('\xff', 'utf-8') def test_console_ansiformat(): diff --git a/tests/test_whiley.py b/tests/test_whiley.py index e844dafb..c895e17f 100644 --- a/tests/test_whiley.py +++ b/tests/test_whiley.py @@ -19,13 +19,13 @@ def lexer(): def test_whiley_operator(lexer): - fragment = u'123 \u2200 x\n' + fragment = '123 \u2200 x\n' tokens = [ - (Token.Literal.Number.Integer, u'123'), - (Token.Text, u' '), - (Token.Operator, u'\u2200'), - (Token.Text, u' '), - (Token.Name, u'x'), - (Token.Text, u'\n'), + (Token.Literal.Number.Integer, '123'), + (Token.Text, ' '), + (Token.Operator, '\u2200'), + (Token.Text, ' '), + (Token.Name, 'x'), + (Token.Text, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens diff --git a/tests/test_yang.py b/tests/test_yang.py index 1de57d05..e2449b59 100644 --- a/tests/test_yang.py +++ b/tests/test_yang.py @@ -20,13 +20,13 @@ def test_namespace_1(lexer): """ Namespace `urn:test:std:yang` should not be explicitly highlighted """ - fragment = u'namespace urn:test:std:yang;\n' + fragment = 'namespace urn:test:std:yang;\n' tokens = [ - (Token.Keyword, u'namespace'), - (Token.Text.Whitespace, u' '), - (Token.Name.Variable, u'urn:test:std:yang'), - (Token.Punctuation, u';'), - (Token.Text.Whitespace, u'\n'), + (Token.Keyword, 'namespace'), + (Token.Text.Whitespace, ' '), + (Token.Name.Variable, 'urn:test:std:yang'), + (Token.Punctuation, ';'), + (Token.Text.Whitespace, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens @@ -34,15 +34,15 @@ def test_namespace_2(lexer): """ namespace-prefix `yang` should be explicitly highlighted """ - fragment = u'type yang:counter64;\n' + fragment = 'type yang:counter64;\n' tokens = [ - (Token.Keyword, u'type'), - (Token.Text.Whitespace, u' '), - (Token.Name.Namespace, u'yang'), - (Token.Punctuation, u':'), - (Token.Name.Variable, u'counter64'), - (Token.Punctuation, u';'), - (Token.Text.Whitespace, u'\n'), + (Token.Keyword, 'type'), + (Token.Text.Whitespace, ' '), + (Token.Name.Namespace, 'yang'), + (Token.Punctuation, ':'), + (Token.Name.Variable, 'counter64'), + (Token.Punctuation, ';'), + (Token.Text.Whitespace, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens @@ -50,13 +50,13 @@ def test_revision_date(lexer): """ Revision-date `2020-08-03` should be explicitly highlighted """ - fragment = u'revision 2020-03-08{\n' + fragment = 'revision 2020-03-08{\n' tokens = [ - (Token.Keyword, u'revision'), - (Token.Text.Whitespace, u' '), - (Token.Name.Label, u'2020-03-08'), - (Token.Punctuation, u'{'), - (Token.Text.Whitespace, u'\n'), + (Token.Keyword, 'revision'), + (Token.Text.Whitespace, ' '), + (Token.Name.Label, '2020-03-08'), + (Token.Punctuation, '{'), + (Token.Text.Whitespace, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens @@ -64,13 +64,13 @@ def test_integer_value(lexer): """ Integer value `5` should be explicitly highlighted """ - fragment = u'value 5;\n' + fragment = 'value 5;\n' tokens = [ - (Token.Keyword, u'value'), - (Token.Text.Whitespace, u' '), - (Token.Number.Integer, u'5'), - (Token.Punctuation, u';'), - (Token.Text.Whitespace, u'\n'), + (Token.Keyword, 'value'), + (Token.Text.Whitespace, ' '), + (Token.Number.Integer, '5'), + (Token.Punctuation, ';'), + (Token.Text.Whitespace, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens @@ -78,13 +78,13 @@ def test_string_value(lexer): """ String value `"5"` should be not explicitly highlighted """ - fragment = u'value "5";\n' + fragment = 'value "5";\n' tokens = [ - (Token.Keyword, u'value'), - (Token.Text.Whitespace, u' '), - (Token.String.Double, u'"5"'), - (Token.Punctuation, u';'), - (Token.Text.Whitespace, u'\n'), + (Token.Keyword, 'value'), + (Token.Text.Whitespace, ' '), + (Token.String.Double, '"5"'), + (Token.Punctuation, ';'), + (Token.Text.Whitespace, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens @@ -92,12 +92,12 @@ def test_float_value(lexer): """ Float value `1.1` should be explicitly highlighted """ - fragment = u'yang-version 1.1;\n' + fragment = 'yang-version 1.1;\n' tokens = [ - (Token.Keyword, u'yang-version'), - (Token.Text.Whitespace, u' '), - (Token.Number.Float, u'1.1'), - (Token.Punctuation, u';'), - (Token.Text.Whitespace, u'\n'), + (Token.Keyword, 'yang-version'), + (Token.Text.Whitespace, ' '), + (Token.Number.Float, '1.1'), + (Token.Punctuation, ';'), + (Token.Text.Whitespace, '\n'), ] assert list(lexer.get_tokens(fragment)) == tokens |