summaryrefslogtreecommitdiff
path: root/gtkdoc-scan.in
Commit message (Collapse)AuthorAgeFilesLines
* Meson: Fix running tools uninstalledXavier Claessens2019-08-151-1/+5
| | | | | This fix getting directories when builddir!=srcdir which is always the case with Meson.
* tools: cleanup options handlingStefan Sauer2017-06-221-4/+2
| | | | Use the 'required' flags for mandatory args. Drop some dest flags where not needed.
* logging: add back support for GTKDOC_TRACEStefan Sauer2017-04-261-1/+3
| | | | | Like before in perl one can set this envar to get verbose log output. The output is meant for developers.
* scan: split into wrapper and moduleStefan Sauer2017-04-101-874/+26
|
* scan: Convert to Python.Jussi Pakkanen2017-04-091-783/+696
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=780789
* code-style: change run() to be upper-caseStefan Sauer2016-11-231-2/+2
|
* gtkdoc: add/update my copyrightStefan Sauer2016-11-231-0/+1
|
* scan: only copy new sections file if it changedStefan Sauer2016-05-311-1/+1
| | | | Avoids needless rebuilds when people use --rebuild-sections.
* scan: remove special case for '_' macroStefan Sauer2016-03-241-39/+36
| | | | | | As discussed in #753052, the special case for '_' is not very useful. The '_' macro is not part of the library API anyway. Further this is inconsitent w.r.t the _N() macros and others. Hence remove the special case.
* rebase,scan: refactor into a 'modulino' and add a test skelletonStefan Sauer2015-06-171-116/+120
|
* gtkdoc-scan: make HTML output deterministicPeter De Wachter2015-05-201-1/+1
| | | | Sort symbols by name before printing them to the decl file.
* gtkdoc-scan: fix regex for get_type() functionsRyan Lortie2015-04-171-2/+2
| | | | | | | | | | | | | | | | | | | | | | The regexp /(void|)/ was apparently trying to look for "(void)" or "()", but in fact, the parens here were only acting as (redundant) grouping, so the expression would match any string containing either "void" or "", which is every string. It turns out that we don't want to look for parens anyway, since they are already stripped out above. We do want to look for exactly "void" or "", however, so add ^ and $. Unfortunately, some code above also replaces all whitespace surrounding newlines with a single space character, and a trailing newline is left at the end of the declaration by another regular expression above, resulting in seeing "void " at this point. Fix that expression up to also clean up the newline at the end, by adding /s as a regexp flag. This fixes the inappropriate matching of non-void functions, such as g_io_extension_get_type(). https://bugzilla.gnome.org/show_bug.cgi?id=746118
* scanner: special case G_DECLARE_*Xavier Claessens2015-03-131-0/+24
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=743879
* gtkdoc-scan: function prototypes can have multiple decoratorsStefan Sauer2015-01-061-1/+1
| | | | | Instead of allowing none or one, allow any number of decorator matches at the end of prototypes. Fixes #741941
* scan: Ignore enumeration fields with DEPRECATED in their nameEmmanuele Bassi2014-07-091-1/+2
| | | | | | | GParamFlags should not be declared deprecated just because it has a value called G_PARAM_DEPRECATED. https://bugzilla.gnome.org/show_bug.cgi?id=730658
* scan: Ignore macro definitions with DEPRECATED in their nameEmmanuele Bassi2014-07-091-2/+4
| | | | | | These should not be automatically deprecated. https://bugzilla.gnome.org/show_bug.cgi?id=731417
* gtkdoc-scan.in: allow ordinary structs and enums to be documentedThomas Wood2014-04-221-3/+3
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=657444
* scan: add missing /oStefan Sauer2014-03-061-1/+1
| | | | While /o is not so much recommened, we're still using it. Lets be atleast consitent.
* gtkdoc-scan: Fix use of uninitialised value in trace loggingPhilip Withnall2014-02-101-2/+2
| | | | | | | | | | | Fixes: > Use of uninitialized value $6 in concatenation (.) or string at > /opt/gnome3/build/bin/gtkdoc-scan line 552, <INPUT> line X. This was caused by re-using the numbered match variables from one regexp after running another regexp. https://bugzilla.gnome.org/show_bug.cgi?id=724002
* scan: tweak the matching of typesStefan Sauer2014-02-061-16/+17
| | | | Join the prefix + type name to make the code easier to read.
* gtkdoc-scan: track doc-comemnts in headers and use for section-docsStefan Sauer2014-02-061-9/+41
| | | | | We track which symbols are docuemnted in the headers and use that to place gobject related entries in the appropriate subsection.
* scan: apply IGNORE_DECORATORS to variables as wellStefan Sauer2014-01-311-1/+1
| | | | Fix detecting global exported variables.
* Fix the hack in the previous commitMatthias Clasen2014-01-231-4/+4
| | | | | | | With 0.1 bump whenever we see _DEPRECATED, we can now enter the nesting calculation with a counter value that is between 0 and 1. Therefore, change the checks to look for >=1 instead of >0. This fixes parsing of an #ifdef in gtkstock.h.
* gtkdoc-scan: recognise attribute-based deprecationsRyan Lortie2014-01-231-0/+12
| | | | | | | | | | | If we see something that looks like a macro for an attribute-based deprecation (ie: some string matching /_DEPRECATED/) then mark the next single declaration as <DEPRECATED/> in the -decl.txt. This avoids the warning for missing deprecation guards. Add Deprecated: tags to the doc-blobs of the existing tests to verify that we don't get the warnings. Fixes: #624001
* gtkdoc-scan: Ignore decorators containing parenthesesCarlos Garcia Campos2014-01-151-14/+14
| | | | | | | | Remove the \b after the IGNORE_DECORATORS regexp to allow passing a decorator containing parentheses like in deprecation macros such as G_DEPRECATED_FOR(). https://bugzilla.gnome.org/show_bug.cgi?id=711598
* gtkdoc-scan: don't add the title until all declarations have been addedAlexander Kanavin2013-08-091-1/+5
| | | | | | | | Previously, adding "SomeSymbol" using AddSymbolToList() didn't work if <TITLE>SomeSymbol</TITLE> was already there, due to regexp in AddSymbolToList. I fixed the order, rather than the regexp. https://bugzilla.gnome.org/show_bug.cgi?id=705633
* scan: also try full name to allow to disambiguateStefan Sauer2012-07-231-0/+5
|
* gtkdoc-scan: revert the sorting of symbolsStefan Sauer2012-05-231-3/+0
| | | | | Developers can arrange symbols by shuffling them inside the header file. Fixes #676685.
* Ignore more decorators on functionsMatthias Clasen2011-10-091-1/+1
| | | | | | | | | | gtk-doc was explicitly ignoring only G_GNUC_*, which is good enough for G_GNUC_MALLOC, G_GNUC_PRINTF etc. But for deprecation annotations, we want to be able to change these on a per-module basis, so we need to be able to wrap G_GNUC_DEPRECATED in, say, a GTK_DEPRECATED macro. And it would be good if gtk-doc would not stumble over it. https://bugzilla.gnome.org/show_bug.cgi?id=624001
* scan: better matching of typesStefan Sauer2011-09-071-6/+6
|
* scan: detect XxxYyyyPrivate structs and move to standard sectionStefan Sauer2011-08-261-8/+15
|
* scan: improve gtype detectionStefan Sauer2011-08-261-31/+42
| | | | | Run code that detects a gtype and moves the standard symbols to the standard section in a loop to e.g. handle enums defined in the same header.
* scan: one blank line between the sections is enoughStefan Sauer2011-08-261-1/+1
| | | | The last commit that added sorting also added an extran blank line.
* scan: make the generated section file more deterministicStefan Kost2011-08-251-18/+23
| | | | Sort the section and the symbols with the section.
* scan,mkdb: handle short,long for (external) variablesStefan Kost2011-08-221-2/+2
| | | | Improve the parser to handle more type variants. Add a test. Fixes #656946
* scan,mkdb: more improvements on extern and global variable parsingStefan Kost2011-08-221-3/+15
| | | | Don't drop type details in gtkdoc-mkdb. Improve the layout and add more tests.
* scan: improve parsing of externsStefan Kost2011-08-181-3/+3
| | | | | An extern of the form "extern const type const name" was not recognised. Add a test for this as well. Fixes #656773
* scan: improve the brace countingDavid Nečas2011-06-201-0/+3
| | | | Eliminate groups of braces on the same line.
* scan: only end structs/unions with '}' at the start of the lineStefan Kost2011-05-011-1/+1
| | | | Fix preliminary ending of structs/unions. Add a test.
* indent: convert tabs to spaces in perl filesStefan Kost2011-04-151-410/+410
|
* scan: add one more trace statementStefan Kost2011-04-051-1/+4
|
* scan: turn #print into new @TRACE@Stefan Kost2011-03-291-52/+52
|
* scan: use LogWarning moreStefan Kost2011-03-291-2/+2
|
* Makefile.am,scan: clarify --igrnore-headersStefan Kost2010-11-261-1/+2
| | | | Aparently ignore-headers can be used to skip whole dirs as well.
* scan: track forward decls and thus allow skipping duplicate symbolsStefan Kost2010-09-201-11/+20
|
* scan: use var for match to avoid invalidating it befroe later accessStefan Kost2010-09-201-2/+3
|
* scan: attempt to handle skipping redeclarations of symbolsStefan Kost2010-09-201-30/+44
| | | | | This unfortunately does not work yet. If a header has multiple different declarations of a symbol, we would like to only use the first.
* scan: provide a way to skip unparsable codeStefan Kost2010-09-131-0/+16
| | | | | Allow to use #ifndef __GTK_DOC_IGNORE__ to mark unparseable code. Fixes #512565.
* scan: improve handling of internal symbolsStefan Kost2010-07-141-22/+29
| | | | | | We need to fully parse them, but omit them from the docs. We did that for macros, but we also need to handle functions that way. Use an extra flag-var instead of extra '-skip' states. Add tests.
* scan: allow static inline functions in headers. Fixes #481811Stefan Kost2010-07-131-3/+14
| | | | Also fix the test case - G_INLINE_FUNC != static inline
View Lorry Controller Status