Implement PCRE2_EXTRA_ALT_BSUX to support ECMAscript 6's \u{hhh..} syntax.

git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@1070 6239d852-aaf2-0410-a92c-79f79f948069

4 files changed, 49 insertions, 23 deletions

diff --git a/testdata/testinput2 b/testdata/testinput2
index a6e44dd..433f9f8 100644
--- a/testdata/testinput2
+++ b/testdata/testinput2
@@ -2408,13 +2408,13 @@
 \= Expect no match
     cat
 
-/(\3)(\1)(a)/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/(\3)(\1)(a)/allow_empty_class,match_unset_backref,dupnames
     cat
 
 /TA]/
     The ACTA] comes
 
-/TA]/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/TA]/allow_empty_class,match_unset_backref,dupnames
     The ACTA] comes
 
 /(?2)[]a()b](abc)/
@@ -2446,25 +2446,25 @@
 
 /a[^]b/
 
-/a[]b/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/a[]b/allow_empty_class,match_unset_backref,dupnames
 \= Expect no match
     ab
 
-/a[]+b/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/a[]+b/allow_empty_class,match_unset_backref,dupnames
 \= Expect no match
     ab
 
-/a[]*+b/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/a[]*+b/allow_empty_class,match_unset_backref,dupnames
 \= Expect no match
     ab
 
-/a[^]b/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/a[^]b/allow_empty_class,match_unset_backref,dupnames
     aXb
     a\nb
 \= Expect no match
     ab
 
-/a[^]+b/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/a[^]+b/allow_empty_class,match_unset_backref,dupnames
     aXb
     a\nX\nXb
 \= Expect no match
@@ -2903,10 +2903,10 @@
     xxxxabcde\=ps
     xxxxabcde\=ph
 
-/(\3)(\1)(a)/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/(\3)(\1)(a)/allow_empty_class,match_unset_backref,dupnames
     cat
 
-/(\3)(\1)(a)/I,alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/(\3)(\1)(a)/I,allow_empty_class,match_unset_backref,dupnames
     cat
 
 /(\3)(\1)(a)/I
@@ -3418,6 +3418,14 @@
     aU0041z
 \= Expect no match
     aAz
+    
+/^\u{7a}/alt_bsux
+    u{7a}
+\= Expect no match
+    zoo 
+
+/^\u{7a}/extra_alt_bsux
+    zoo 
 
 /(?(?=c)c|d)++Y/B
 
diff --git a/testdata/testinput5 b/testdata/testinput5
index 3e7701e..b48d27a 100644
--- a/testdata/testinput5
+++ b/testdata/testinput5
@@ -333,13 +333,13 @@
 
 /[[:a\x{100}b:]]/utf
 
-/a[^]b/utf,alt_bsux,allow_empty_class,match_unset_backref
+/a[^]b/utf,allow_empty_class,match_unset_backref
     a\x{1234}b
     a\nb
 \= Expect no match
     ab
 
-/a[^]+b/utf,alt_bsux,allow_empty_class,match_unset_backref
+/a[^]+b/utf,allow_empty_class,match_unset_backref
     aXb
     a\nX\nX\x{1234}b
 \= Expect no match
@@ -814,6 +814,9 @@
 
 /\ud800/utf,alt_bsux,allow_empty_class,match_unset_backref
 
+/^\u{0000000000010ffff}/utf,extra_alt_bsux
+    \x{10ffff}
+
 /^a+[a\x{200}]/B,utf
     aa
 
diff --git a/testdata/testoutput2 b/testdata/testoutput2
index 6628457..d0e289d 100644
--- a/testdata/testoutput2
+++ b/testdata/testoutput2
@@ -8774,7 +8774,7 @@ No match
     cat
 No match
 
-/(\3)(\1)(a)/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/(\3)(\1)(a)/allow_empty_class,match_unset_backref,dupnames
     cat
  0: a
  1: 
@@ -8785,7 +8785,7 @@ No match
     The ACTA] comes
  0: TA]
 
-/TA]/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/TA]/allow_empty_class,match_unset_backref,dupnames
     The ACTA] comes
  0: TA]
 
@@ -8833,22 +8833,22 @@ Failed: error 106 at offset 4: missing terminating ] for character class
 /a[^]b/
 Failed: error 106 at offset 5: missing terminating ] for character class
 
-/a[]b/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/a[]b/allow_empty_class,match_unset_backref,dupnames
 \= Expect no match
     ab
 No match
 
-/a[]+b/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/a[]+b/allow_empty_class,match_unset_backref,dupnames
 \= Expect no match
     ab
 No match
 
-/a[]*+b/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/a[]*+b/allow_empty_class,match_unset_backref,dupnames
 \= Expect no match
     ab
 No match
 
-/a[^]b/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/a[^]b/allow_empty_class,match_unset_backref,dupnames
     aXb
  0: aXb
     a\nb
@@ -8857,7 +8857,7 @@ No match
     ab
 No match
 
-/a[^]+b/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/a[^]+b/allow_empty_class,match_unset_backref,dupnames
     aXb
  0: aXb
     a\nX\nXb
@@ -9971,17 +9971,17 @@ Partial match: abca
     xxxxabcde\=ph
 Partial match: abcde
 
-/(\3)(\1)(a)/alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/(\3)(\1)(a)/allow_empty_class,match_unset_backref,dupnames
     cat
  0: a
  1: 
  2: 
  3: a
 
-/(\3)(\1)(a)/I,alt_bsux,allow_empty_class,match_unset_backref,dupnames
+/(\3)(\1)(a)/I,allow_empty_class,match_unset_backref,dupnames
 Capture group count = 3
 Max back reference = 3
-Options: alt_bsux allow_empty_class dupnames match_unset_backref
+Options: allow_empty_class dupnames match_unset_backref
 Last code unit = 'a'
 Subject length lower bound = 1
     cat
@@ -11364,6 +11364,17 @@ No match
 \= Expect no match
     aAz
 No match
+    
+/^\u{7a}/alt_bsux
+    u{7a}
+ 0: u{7a}
+\= Expect no match
+    zoo 
+No match
+
+/^\u{7a}/extra_alt_bsux
+    zoo 
+ 0: z
 
 /(?(?=c)c|d)++Y/B
 ------------------------------------------------------------------
diff --git a/testdata/testoutput5 b/testdata/testoutput5
index d2f9352..2bcb39f 100644
--- a/testdata/testoutput5
+++ b/testdata/testoutput5
@@ -798,7 +798,7 @@ No match
 /[[:a\x{100}b:]]/utf
 Failed: error 130 at offset 3: unknown POSIX class name
 
-/a[^]b/utf,alt_bsux,allow_empty_class,match_unset_backref
+/a[^]b/utf,allow_empty_class,match_unset_backref
     a\x{1234}b
  0: a\x{1234}b
     a\nb
@@ -807,7 +807,7 @@ Failed: error 130 at offset 3: unknown POSIX class name
     ab
 No match
 
-/a[^]+b/utf,alt_bsux,allow_empty_class,match_unset_backref
+/a[^]+b/utf,allow_empty_class,match_unset_backref
     aXb
  0: aXb
     a\nX\nX\x{1234}b
@@ -1734,6 +1734,10 @@ No match
 /\ud800/utf,alt_bsux,allow_empty_class,match_unset_backref
 Failed: error 173 at offset 6: disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
 
+/^\u{0000000000010ffff}/utf,extra_alt_bsux
+    \x{10ffff}
+ 0: \x{10ffff}
+
 /^a+[a\x{200}]/B,utf
 ------------------------------------------------------------------
         Bra