/[pcre]/code/trunk/ChangeLog
ViewVC logotype

Diff of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 182 by ph10, Wed Jun 13 15:09:54 2007 UTC revision 1420 by zherczeg, Sun Dec 29 11:43:10 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.2 13-June-07  Version 8.35-RC1 xx-xxxx-201x
5    -----------------------------
6    
7    1.  A new flag is set, when property checks are present in an XCLASS.
8        When this flag is not set, PCRE can perform certain optimizations
9        such as studying these XCLASS-es.
10    
11    2.  The auto-possessification of character sets were improved: a normal
12        and an extended character set can be compared now. Furthermore
13        the JIT compiler optimizes more character set checks.
14    
15    3.  Got rid of some compiler warnings for potentially uninitialized variables
16        that show up only when compiled with -O2.
17    
18    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
19        of a match later then the end of the match. The pcretest program was not
20        handling the case sensibly - it was outputting from the start to the next
21        binary zero. It now reports this situation in a message, and outputs the
22        text from the end to the start.
23    
24    5.  Fast forward search is improved in JIT. Instead of the first three
25        characters, any three characters with fixed position can be searched.
26        Search order: first, last, middle.
27    
28    6.  Improve character range checks in JIT. Characters are read by an inprecise
29        function now, which returns with an unknown value if the character code is
30        above a certain treshold (e.g: 256). The only limitation is that the value
31        must be bigger than the treshold as well. This function is useful, when
32        the characters above the treshold are handled in the same way.
33    
34    
35    Version 8.34 15-December-2013
36    -----------------------------
37    
38    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
39        executable memory. Patch inspired by Carsten Klein.
40    
41    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
42        this macro is never tested and has no effect, because the work to support
43        coverage involves only compiling and linking options and special targets in
44        the Makefile. The comment in config.h implied that defining the macro would
45        enable coverage support, which is totally false. There was also support for
46        setting this macro in the CMake files (my fault, I just copied it from
47        configure). SUPPORT_GCOV has now been removed.
48    
49    3.  Make a small performance improvement in strlen16() and strlen32() in
50        pcretest.
51    
52    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
53        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
54        these cases). There was also one in pcretest.c.
55    
56    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
57    
58    6.  In UTF mode, the code for checking whether a group could match an empty
59        string (which is used for indefinitely repeated groups to allow for
60        breaking an infinite loop) was broken when the group contained a repeated
61        negated single-character class with a character that occupied more than one
62        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
63        UTF-8 mode). The effect was undefined: the group might or might not be
64        deemed as matching an empty string, or the program might have crashed.
65    
66    7.  The code for checking whether a group could match an empty string was not
67        recognizing that \h, \H, \v, \V, and \R must match a character.
68    
69    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
70        an empty string. If it can, pcretest shows this in its information output.
71    
72    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
73        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
74        matched by pcre_exec() without using JIT:
75    
76        (a) If the rest of the pattern did not match after a maximal run of
77            grapheme clusters, the code for backing up to try with fewer of them
78            did not always back up over a full grapheme when characters that do not
79            have the modifier quality were involved, e.g. Hangul syllables.
80    
81        (b) If the match point in a subject started with modifier character, and
82            there was no match, the code could incorrectly back up beyond the match
83            point, and potentially beyond the first character in the subject,
84            leading to a segfault or an incorrect match result.
85    
86    10. A conditional group with an assertion condition could lead to PCRE
87        recording an incorrect first data item for a match if no other first data
88        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
89        first data item, and therefore matched "ca" after "c" instead of at the
90        start.
91    
92    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
93        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
94    
95    12. The source of pcregrep now includes z/OS-specific code so that it can be
96        compiled for z/OS as part of the special z/OS distribution.
97    
98    13. Added the -T and -TM options to pcretest.
99    
100    14. The code in pcre_compile.c for creating the table of named capturing groups
101        has been refactored. Instead of creating the table dynamically during the
102        actual compiling pass, the information is remembered during the pre-compile
103        pass (on the stack unless there are more than 20 named groups, in which
104        case malloc() is used) and the whole table is created before the actual
105        compile happens. This has simplified the code (it is now nearly 150 lines
106        shorter) and prepared the way for better handling of references to groups
107        with duplicate names.
108    
109    15. A back reference to a named subpattern when there is more than one of the
110        same name now checks them in the order in which they appear in the pattern.
111        The first one that is set is used for the reference. Previously only the
112        first one was inspected. This change makes PCRE more compatible with Perl.
113    
114    16. Unicode character properties were updated from Unicode 6.3.0.
115    
116    17. The compile-time code for auto-possessification has been refactored, based
117        on a patch by Zoltan Herczeg. It now happens after instead of during
118        compilation. The code is cleaner, and more cases are handled. The option
119        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
120        options in pcretest are provided to set it. It can also be set by
121        (*NO_AUTO_POSSESS) at the start of a pattern.
122    
123    18. The character VT has been added to the default ("C" locale) set of
124        characters that match \s and are generally treated as white space,
125        following this same change in Perl 5.18. There is now no difference between
126        "Perl space" and "POSIX space". Whether VT is treated as white space in
127        other locales depends on the locale.
128    
129    19. The code for checking named groups as conditions, either for being set or
130        for being recursed, has been refactored (this is related to 14 and 15
131        above). Processing unduplicated named groups should now be as fast at
132        numerical groups, and processing duplicated groups should be faster than
133        before.
134    
135    20. Two patches to the CMake build system, by Alexander Barkov:
136    
137          (1) Replace the "source" command by "." in CMakeLists.txt because
138              "source" is a bash-ism.
139    
140          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
141              without these the CMake build does not work on Solaris.
142    
143    21. Perl has changed its handling of \8 and \9. If there is no previously
144        encountered capturing group of those numbers, they are treated as the
145        literal characters 8 and 9 instead of a binary zero followed by the
146        literals. PCRE now does the same.
147    
148    22. Following Perl, added \o{} to specify codepoints in octal, making it
149        possible to specify values greater than 0777 and also making them
150        unambiguous.
151    
152    23. Perl now gives an error for missing closing braces after \x{... instead of
153        treating the string as literal. PCRE now does the same.
154    
155    24. RunTest used to grumble if an inappropriate test was selected explicitly,
156        but just skip it when running all tests. This make it awkward to run ranges
157        of tests when one of them was inappropriate. Now it just skips any
158        inappropriate tests, as it always did when running all tests.
159    
160    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
161        character types such as \d or \w, too many callouts were inserted, and the
162        data that they returned was rubbish.
163    
164    26. In UCP mode, \s was not matching two of the characters that Perl matches,
165        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
166        were matched by \h. The code has now been refactored so that the lists of
167        the horizontal and vertical whitespace characters used for \h and \v (which
168        are defined only in one place) are now also used for \s.
169    
170    27. Add JIT support for the 64 bit TileGX architecture.
171        Patch by Jiong Wang (Tilera Corporation).
172    
173    28. Possessive quantifiers for classes (both explicit and automatically
174        generated) now use special opcodes instead of wrapping in ONCE brackets.
175    
176    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
177        (because it's meaningless), this was not happening when PCRE_CASELESS was
178        set. Not wrong, but inefficient.
179    
180    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
181        properties for \w, \d, etc) is present in a test regex. Otherwise if the
182        test contains no characters greater than 255, Perl doesn't realise it
183        should be using Unicode semantics.
184    
185    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
186        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
187        does in Unicode mode.
188    
189    32. Added the "forbid" facility to pcretest so that putting tests into the
190        wrong test files can sometimes be quickly detected.
191    
192    33. There is now a limit (default 250) on the depth of nesting of parentheses.
193        This limit is imposed to control the amount of system stack used at compile
194        time. It can be changed at build time by --with-parens-nest-limit=xxx or
195        the equivalent in CMake.
196    
197    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
198        errors. Perl warns for these when in warning mode, but PCRE has no facility
199        for giving warnings.
200    
201    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
202        However, this was not working for (?!) because it is optimized to (*FAIL),
203        for which PCRE does not allow quantifiers. The optimization is now disabled
204        when a quantifier follows (?!). I can't see any use for this, but it makes
205        things uniform.
206    
207    36. Perl no longer allows group names to start with digits, so I have made this
208        change also in PCRE. It simplifies the code a bit.
209    
210    37. In extended mode, Perl ignores spaces before a + that indicates a
211        possessive quantifier. PCRE allowed a space before the quantifier, but not
212        before the possessive +. It now does.
213    
214    38. The use of \K (reset reported match start) within a repeated possessive
215        group such as (a\Kb)*+ was not working.
216    
217    40. Document that the same character tables must be used at compile time and
218        run time, and that the facility to pass tables to pcre_exec() and
219        pcre_dfa_exec() is for use only with saved/restored patterns.
220    
221    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
222        features for Builds with MSVC:
223    
224        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
225           builds). The .rc files can be used to set FileDescription and many other
226           attributes.
227    
228        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
229           This allows higher-level build scripts which want .pdb files to avoid
230           hard-coding the exact files needed."
231    
232    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
233        mean "start of word" and "end of word", respectively, as a transition aid.
234    
235    43. A minimizing repeat of a class containing codepoints greater than 255 in
236        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
237        compiled to use the heap for recursion.
238    
239    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
240        is configured.
241    
242    
243    Version 8.33 28-May-2013
244    ------------------------
245    
246    1.  Added 'U' to some constants that are compared to unsigned integers, to
247        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
248        variables that are added to signed variables, to ensure the result is
249        signed and can be negated.
250    
251    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
252        CMake config files.
253    
254    3.  Revise the creation of config.h.generic so that all boolean macros are
255        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
256        overriding via -D on the command line possible.
257    
258    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
259        to unsigned int is reported to make a quite noticeable speed difference in
260        a specific Windows environment. Testing on Linux did also appear to show
261        some benefit (and it is clearly not harmful). Also fixed the definition of
262        Xop which should be unsigned.
263    
264    5.  Related to (4), changing the definition of the intermediate variable cc
265        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
266        improvements.
267    
268    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
269        unnecessary spaces.
270    
271    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
272        and later.
273    
274    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
275    
276    9.  Optimizing fast_forward_start_bits in JIT.
277    
278    10. Adding support for callouts in JIT, and fixing some issues revealed
279        during this work. Namely:
280    
281        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
282    
283        (b) Minimum length was not checked before the matching is started.
284    
285    11. The value of capture_last that is passed to callouts was incorrect in some
286        cases when there was a capture on one path that was subsequently abandoned
287        after a backtrack. Also, the capture_last value is now reset after a
288        recursion, since all captures are also reset in this case.
289    
290    12. The interpreter no longer returns the "too many substrings" error in the
291        case when an overflowing capture is in a branch that is subsequently
292        abandoned after a backtrack.
293    
294    13. In the pathological case when an offset vector of size 2 is used, pcretest
295        now prints out the matched string after a yield of 0 or 1.
296    
297    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
298        Only supported by the JIT compiler at the moment.
299    
300    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
301    
302    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
303        offset of the starting point of the matching process, provided the offsets
304        vector is large enough.
305    
306    17. The \A escape now records a lookbehind value of 1, though its execution
307        does not actually inspect the previous character. This is to ensure that,
308        in partial multi-segment matching, at least one character from the old
309        segment is retained when a new segment is processed. Otherwise, if there
310        are no lookbehinds in the pattern, \A might match incorrectly at the start
311        of a new segment.
312    
313    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
314    
315    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
316        gives some modest performance improvement in 8-bit mode.
317    
318    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
319        be expressed in certain programming languages using Universal Character
320        Names.
321    
322    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
323        which points out that "non characters" are not "characters that may not
324        appear in Unicode strings" but rather "characters that are reserved for
325        internal use and have only local meaning".
326    
327    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
328        there was a conditional group that depended on an assertion, if the
329        assertion was false, the callout that immediately followed the alternation
330        in the condition was skipped when pcre_exec() was used for matching.
331    
332    23. Allow an explicit callout to be inserted before an assertion that is the
333        condition for a conditional group, for compatibility with automatic
334        callouts, which always insert a callout at this point.
335    
336    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
337        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
338    
339    25. (*PRUNE) is now supported by the JIT compiler.
340    
341    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
342    
343    27. Fix the case where there are two or more SKIPs with arguments that may be
344        ignored.
345    
346    28. (*SKIP) is now supported by the JIT compiler.
347    
348    29. (*THEN) is now supported by the JIT compiler.
349    
350    30. Update RunTest with additional test selector options.
351    
352    31. The way PCRE handles backtracking verbs has been changed in two ways.
353    
354        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
355        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
356        backtracking. In some cases this makes it more Perl-compatible, but Perl's
357        rather obscure rules do not always do the same thing.
358    
359        (2) Previously, backtracking verbs were confined within assertions. This is
360        no longer the case for positive assertions, except for (*ACCEPT). Again,
361        this sometimes improves Perl compatibility, and sometimes does not.
362    
363    32. A number of tests that were in test 2 because Perl did things differently
364        have been moved to test 1, because either Perl or PCRE has changed, and
365        these tests are now compatible.
366    
367    32. Backtracking control verbs are now handled in the same way in JIT and
368        interpreter.
369    
370    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
371        contained a forward subroutine reference caused a compile error.
372    
373    34. Auto-detect and optimize limited repetitions in JIT.
374    
375    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
376        blocking (*UTF) etc.
377    
378    36. In the interpreter, maximizing pattern repetitions for characters and
379        character types now use tail recursion, which reduces stack usage.
380    
381    37. The value of the max lookbehind was not correctly preserved if a compiled
382        and saved regex was reloaded on a host of different endianness.
383    
384    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
385        of the compiled pattern block, expand the flags field from 16 to 32 bits
386        because it was almost full.
387    
388    39. Try madvise first before posix_madvise.
389    
390    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
391        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
392        original change locked out all matches of empty strings. This has been
393        changed so that one match of an empty string per line is recognized.
394        Subsequent searches on the same line (for colouring or for --only-matching,
395        for example) do not recognize empty strings.
396    
397    41. Applied a user patch to fix a number of spelling mistakes in comments.
398    
399    42. Data lines longer than 65536 caused pcretest to crash.
400    
401    43. Clarified the data type for length and startoffset arguments for pcre_exec
402        and pcre_dfa_exec in the function-specific man pages, where they were
403        explicitly stated to be in bytes, never having been updated. I also added
404        some clarification to the pcreapi man page.
405    
406    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
407        a segmentation fault.
408    
409    
410    Version 8.32 30-November-2012
411    -----------------------------
412    
413    1.  Improved JIT compiler optimizations for first character search and single
414        character iterators.
415    
416    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
417        Patch by Daniel Richard G.
418    
419    3.  Single character iterator optimizations in the JIT compiler.
420    
421    4.  Improved JIT compiler optimizations for character ranges.
422    
423    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
424        Reported by Giuseppe D'Angelo.
425    
426    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
427        of a line, was being set incorrectly in cases where .* appeared inside
428        atomic brackets at the start of a pattern, or where there was a subsequent
429        *PRUNE or *SKIP.
430    
431    7.  Improved instruction cache flush for POWER/PowerPC.
432        Patch by Daniel Richard G.
433    
434    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
435        grep:
436    
437        (a) There is now no limit to the number of patterns to be matched.
438    
439        (b) An error is given if a pattern is too long.
440    
441        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
442            are now supported.
443    
444        (d) --exclude-from and --include-from (multiple use) have been added.
445    
446        (e) Exclusions and inclusions now apply to all files and directories, not
447            just to those obtained from scanning a directory recursively.
448    
449        (f) Multiple uses of -f and --file-list are now supported.
450    
451        (g) In a Windows environment, the default for -d has been changed from
452            "read" (the GNU grep default) to "skip", because otherwise the presence
453            of a directory in the file list provokes an error.
454    
455        (h) The documentation has been revised and clarified in places.
456    
457    9.  Improve the matching speed of capturing brackets.
458    
459    10. Changed the meaning of \X so that it now matches a Unicode extended
460        grapheme cluster.
461    
462    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
463        out POSIX threads when JIT support is configured.
464    
465    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
466    
467    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
468        structure could go wrong in environments where size_t is not the same size
469        as int.
470    
471    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
472        set.
473    
474    15. The EBCDIC support had decayed; later updates to the code had included
475        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
476        general tidy up of EBCDIC-related issues, and the documentation was also
477        not quite right. There is now a test that can be run on ASCII systems to
478        check some of the EBCDIC-related things (but is it not a full test).
479    
480    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
481        in a small tidy to the code.
482    
483    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
484    
485    18. If the --only-matching (-o) option in pcregrep is specified multiple
486        times, each one causes appropriate output. For example, -o1 -o2 outputs the
487        substrings matched by the 1st and 2nd capturing parentheses. A separating
488        string can be specified by --om-separator (default empty).
489    
490    19. Improving the first n character searches.
491    
492    20. Turn case lists for horizontal and vertical white space into macros so that
493        they are defined only once.
494    
495    21. This set of changes together give more compatible Unicode case-folding
496        behaviour for characters that have more than one other case when UCP
497        support is available.
498    
499        (a) The Unicode property table now has offsets into a new table of sets of
500            three or more characters that are case-equivalent. The MultiStage2.py
501            script that generates these tables (the pcre_ucd.c file) now scans
502            CaseFolding.txt instead of UnicodeData.txt for character case
503            information.
504    
505        (b) The code for adding characters or ranges of characters to a character
506            class has been abstracted into a generalized function that also handles
507            case-independence. In UTF-mode with UCP support, this uses the new data
508            to handle characters with more than one other case.
509    
510        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
511            whose other case is greater than 256 are now correctly matched
512            caselessly. Previously, the high codepoint matched the low one, but not
513            vice versa.
514    
515        (d) The processing of \h, \H, \v, and \ in character classes now makes use
516            of the new class addition function, using character lists defined as
517            macros alongside the case definitions of 20 above.
518    
519        (e) Caseless back references now work with characters that have more than
520            one other case.
521    
522        (f) General caseless matching of characters with more than one other case
523            is supported.
524    
525    22. Unicode character properties were updated from Unicode 6.2.0
526    
527    23. Improved CMake support under Windows. Patch by Daniel Richard G.
528    
529    24. Add support for 32-bit character strings, and UTF-32
530    
531    25. Major JIT compiler update (code refactoring and bugfixing).
532        Experimental Sparc 32 support is added.
533    
534    26. Applied a modified version of Daniel Richard G's patch to create
535        pcre.h.generic and config.h.generic by "make" instead of in the
536        PrepareRelease script.
537    
538    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
539        pcre_compile.c when checking for a zero character.
540    
541    28. Introducing a native interface for JIT. Through this interface, the compiled
542        machine code can be directly executed. The purpose of this interface is to
543        provide fast pattern matching, so several sanity checks are not performed.
544        However, feature tests are still performed. The new interface provides
545        1.4x speedup compared to the old one.
546    
547    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
548        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
549        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
550    
551    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
552        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
553        the "old" RFC 2279). Instead, it ended up passing a negative length to
554        pcre_exec().
555    
556    31. Add support for GCC's visibility feature to hide internal functions.
557    
558    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
559        "unknown -C option" after outputting 0 or 1.
560    
561    33. There is now support for generating a code coverage report for the test
562        suite in environments where gcc is the compiler and lcov is installed. This
563        is mainly for the benefit of the developers.
564    
565    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
566        unaddressable using valgrind annotations, allowing valgrind to detect
567        invalid memory accesses. This is mainly for the benefit of the developers.
568    
569    25. (*UTF) can now be used to start a pattern in any of the three libraries.
570    
571    26. Give configure error if --enable-cpp but no C++ compiler found.
572    
573    
574    Version 8.31 06-July-2012
575    -------------------------
576    
577    1.  Fixing a wrong JIT test case and some compiler warnings.
578    
579    2.  Removed a bashism from the RunTest script.
580    
581    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
582        to unsigned type, result still unsigned" that was given by an MS compiler
583        on encountering the code "-sizeof(xxx)".
584    
585    4.  Partial matching support is added to the JIT compiler.
586    
587    5.  Fixed several bugs concerned with partial matching of items that consist
588        of more than one character:
589    
590        (a) /^(..)\1/ did not partially match "aba" because checking references was
591            done on an "all or nothing" basis. This also applied to repeated
592            references.
593    
594        (b) \R did not give a hard partial match if \r was found at the end of the
595            subject.
596    
597        (c) \X did not give a hard partial match after matching one or more
598            characters at the end of the subject.
599    
600        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
601            a partial match for the string "\r".
602    
603        (e) When newline was set to CRLF, the metacharacter "." did not recognize
604            a partial match for a CR character at the end of the subject string.
605    
606    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
607        running pcretest, the text "(JIT)" added to the output whenever JIT is
608        actually used to run the match.
609    
610    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
611        or /S+[+] with a digit between 1 and 7.
612    
613    8.  OP_NOT now supports any UTF character not just single-byte ones.
614    
615    9.  (*MARK) control verb is now supported by the JIT compiler.
616    
617    10. The command "./RunTest list" lists the available tests without actually
618        running any of them. (Because I keep forgetting what they all are.)
619    
620    11. Add PCRE_INFO_MAXLOOKBEHIND.
621    
622    12. Applied a (slightly modified) user-supplied patch that improves performance
623        when the heap is used for recursion (compiled with --disable-stack-for-
624        recursion). Instead of malloc and free for each heap frame each time a
625        logical recursion happens, frames are retained on a chain and re-used where
626        possible. This sometimes gives as much as 30% improvement.
627    
628    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
629        call.
630    
631    14. As documented, (*COMMIT) is now confined to within a positive assertion.
632    
633    15. It is now possible to link pcretest with libedit as an alternative to
634        libreadline.
635    
636    16. (*COMMIT) control verb is now supported by the JIT compiler.
637    
638    17. The Unicode data tables have been updated to Unicode 6.1.0.
639    
640    18. Added --file-list option to pcregrep.
641    
642    19. Added binary file support to pcregrep, including the -a, --binary-files,
643        -I, and --text options.
644    
645    20. The madvise function is renamed for posix_madvise for QNX compatibility
646        reasons. Fixed by Giuseppe D'Angelo.
647    
648    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
649    
650    22. Changed the option for creating symbolic links for 16-bit man pages from
651        -s to -sf so that re-installing does not cause issues.
652    
653    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
654    
655    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
656        matches in certain environments (the workspace was not being correctly
657        retained). Also added to pcre_dfa_exec() a simple plausibility check on
658        some of the workspace data at the beginning of a restart.
659    
660    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
661        was not doing so when it should - probably a typo introduced by SVN 528
662        (change 8.10/14).
663    
664    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
665        \w+ when the character tables indicated that \x{c4} was a word character.
666        There were several related cases, all because the tests for doing a table
667        lookup were testing for characters less than 127 instead of 255.
668    
669    27. If a pattern contains capturing parentheses that are not used in a match,
670        their slots in the ovector are set to -1. For those that are higher than
671        any matched groups, this happens at the end of processing. In the case when
672        there were back references that the ovector was too small to contain
673        (causing temporary malloc'd memory to be used during matching), and the
674        highest capturing number was not used, memory off the end of the ovector
675        was incorrectly being set to -1. (It was using the size of the temporary
676        memory instead of the true size.)
677    
678    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
679        ovector size, it uses memory at the end of the block that it has got.
680    
681    29. Check for an overlong MARK name and give an error at compile time. The
682        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
683    
684    30. JIT compiler update.
685    
686    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
687        Rill for the patch.
688    
689    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
690    
691    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
692    
693    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
694        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
695        was enabled.
696    
697    35. Improve JIT code generation for greedy plus quantifier.
698    
699    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
700        1 to "aa" instead of to an empty string. The bug affected repeated groups
701        that could potentially match an empty string.
702    
703    37. Optimizing single character iterators in JIT.
704    
705    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
706        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
707        codepoints that are too big for the mode are faulted, and in a UTF mode,
708        disallowed codepoints are also faulted.
709    
710    39. If PCRE was compiled with UTF support, in three places in the DFA
711        matcher there was code that should only have been obeyed in UTF mode, but
712        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
713        processing when bytes with values greater than 127 were present. In 16-bit
714        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
715        both cases the values are those that cannot be the first data item in a UTF
716        character. The three items that might have provoked this were recursions,
717        possessively repeated groups, and atomic groups.
718    
719    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
720        and pcregrep, because some OS require shared objects to be explicitly
721        passed to ld, causing the link step to fail if they are not.
722    
723    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
724        mode, patterns that started with \h* or \R* might be incorrectly matched.
725    
726    
727    Version 8.30 04-February-2012
728    -----------------------------
729    
730    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
731        name is defined in ctype.h.
732    
733    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
734        only in quite long subpatterns.
735    
736    3.  Removed the function pcre_info(), which has been obsolete and deprecated
737        since it was replaced by pcre_fullinfo() in February 2000.
738    
739    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
740        match a (*MARK), and the match failed at the start of the subject, a
741        reference to memory before the start of the subject could occur. This bug
742        was introduced by fix 17 of release 8.21.
743    
744    5.  A reference to an unset group with zero minimum repetition was giving
745        totally wrong answers (in non-JavaScript-compatibility mode). For example,
746        /(another)?(\1?)test/ matched against "hello world test". This bug was
747        introduced in release 8.13.
748    
749    6.  Add support for 16-bit character strings (a large amount of work involving
750        many changes and refactorings).
751    
752    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
753        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
754        from a file.
755    
756    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
757        rounding is not applied in this particular case).
758    
759    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
760        if they appear, or are escaped, in patterns.
761    
762    10. Get rid of a number of -Wunused-but-set-variable warnings.
763    
764    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
765        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
766        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
767        also returns the mark "x". This bug applied to capturing parentheses,
768        non-capturing parentheses, and atomic parentheses. It also applied to some
769        assertions.
770    
771    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
772        information out of configure.ac instead of relying on pcre.h.generic, which
773        is not stored in the repository.
774    
775    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
776        -lreadline.
777    
778    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
779    
780    15. Applied Graycode's patch to put the top-level frame on the stack rather
781        than the heap when not using the stack for recursion. This gives a
782        performance improvement in many cases when recursion is not deep.
783    
784    16. Experimental code added to "pcretest -C" to output the stack frame size.
785    
786    
787    Version 8.21 12-Dec-2011
788    ------------------------
789    
790    1.  Updating the JIT compiler.
791    
792    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
793        are added as well.
794    
795    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
796        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
797        calling _pcre_jit_exec. Some extra comments are added.
798    
799    4.  (*MARK) settings inside atomic groups that do not contain any capturing
800        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
801        was introduced by change 18 for 8.20.
802    
803    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
804        ECMA-262 standard.
805    
806    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
807        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
808        This bug was probably introduced by change 9 of 8.13.
809    
810    7.  While fixing 6 above, I noticed that a number of other items were being
811        incorrectly rejected as "not fixed length". This arose partly because newer
812        opcodes had not been added to the fixed-length checking code. I have (a)
813        corrected the bug and added tests for these items, and (b) arranged for an
814        error to occur if an unknown opcode is encountered while checking for fixed
815        length instead of just assuming "not fixed length". The items that were
816        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
817        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
818        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
819    
820    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
821        being incorrectly compiled and would have given unpredicatble results.
822    
823    9.  A possessively repeated subpattern with minimum repeat count greater than
824        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
825        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
826        the first (A) could occur when it should not.
827    
828    10. Add a cast and remove a redundant test from the code.
829    
830    11. JIT should use pcre_malloc/pcre_free for allocation.
831    
832    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
833        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
834        is anything other than /usr, -L is still shown).
835    
836    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
837    
838    14. Perl does not support \N without a following name in a [] class; PCRE now
839        also gives an error.
840    
841    15. If a forward reference was repeated with an upper limit of around 2000,
842        it caused the error "internal error: overran compiling workspace". The
843        maximum number of forward references (including repeats) was limited by the
844        internal workspace, and dependent on the LINK_SIZE. The code has been
845        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
846        the default depends on LINK_SIZE. There is a new upper limit (for safety)
847        of around 200,000 forward references. While doing this, I also speeded up
848        the filling in of repeated forward references.
849    
850    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
851        incorrectly expecting the subject to contain another "a" after the start.
852    
853    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
854        in the match, the SKIP should be ignored. This was not happening; instead
855        the SKIP was being treated as NOMATCH. For patterns such as
856        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
857        tested.
858    
859    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
860        now much more compatible with Perl, in particular in cases where the result
861        is a non-match for a non-anchored pattern. For example, if
862        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
863        "m", where previously it did not return a name. A side effect of this
864        change is that for partial matches, the last encountered mark name is
865        returned, as for non matches. A number of tests that were previously not
866        Perl-compatible have been moved into the Perl-compatible test files. The
867        refactoring has had the pleasing side effect of removing one argument from
868        the match() function, thus reducing its stack requirements.
869    
870    19. If the /S+ option was used in pcretest to study a pattern using JIT,
871        subsequent uses of /S (without +) incorrectly behaved like /S+.
872    
873    21. Retrieve executable code size support for the JIT compiler and fixing
874        some warnings.
875    
876    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
877        not work when the shorter character appeared right at the end of the
878        subject string.
879    
880    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
881        systems.
882    
883    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
884        output it when the /M option is used in pcretest.
885    
886    25. The CheckMan script was not being included in the distribution. Also, added
887        an explicit "perl" to run Perl scripts from the PrepareRelease script
888        because this is reportedly needed in Windows.
889    
890    26. If study data was being save in a file and studying had not found a set of
891        "starts with" bytes for the pattern, the data written to the file (though
892        never used) was taken from uninitialized memory and so caused valgrind to
893        complain.
894    
895    27. Updated RunTest.bat as provided by Sheri Pierce.
896    
897    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
898    
899    29. Computation of memory usage for the table of capturing group names was
900        giving an unnecessarily large value.
901    
902    
903    Version 8.20 21-Oct-2011
904    ------------------------
905    
906    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
907        a POSIX class. After further experiments with Perl, which convinced me that
908        Perl has bugs and confusions, a closing square bracket is no longer allowed
909        in a POSIX name. This bug also affected patterns with classes that started
910        with full stops.
911    
912    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
913        captured substring, but while checking the failing first alternative,
914        substring 1 is temporarily captured. If the output vector supplied to
915        pcre_exec() was not big enough for this capture, the yield of the function
916        was still zero ("insufficient space for captured substrings"). This cannot
917        be totally fixed without adding another stack variable, which seems a lot
918        of expense for a edge case. However, I have improved the situation in cases
919        such as /(a)(b)x|abc/ matched against "abc", where the return code
920        indicates that fewer than the maximum number of slots in the ovector have
921        been set.
922    
923    3.  Related to (2) above: when there are more back references in a pattern than
924        slots in the output vector, pcre_exec() uses temporary memory during
925        matching, and copies in the captures as far as possible afterwards. It was
926        using the entire output vector, but this conflicts with the specification
927        that only 2/3 is used for passing back captured substrings. Now it uses
928        only the first 2/3, for compatibility. This is, of course, another edge
929        case.
930    
931    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
932        main code base, and can be used by building with --enable-jit. When this is
933        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
934        runtime --no-jit option is given.
935    
936    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
937        ovector, the return from the function was zero, implying that there were
938        other matches that did not fit. The correct "exactly full" value is now
939        returned.
940    
941    6.  If a subpattern that was called recursively or as a subroutine contained
942        (*PRUNE) or any other control that caused it to give a non-standard return,
943        invalid errors such as "Error -26 (nested recursion at the same subject
944        position)" or even infinite loops could occur.
945    
946    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
947        computing the minimum length on reaching *ACCEPT, and so ended up with the
948        wrong value of 1 rather than 0. Further investigation indicates that
949        computing a minimum subject length in the presence of *ACCEPT is difficult
950        (think back references, subroutine calls), and so I have changed the code
951        so that no minimum is registered for a pattern that contains *ACCEPT.
952    
953    8.  If (*THEN) was present in the first (true) branch of a conditional group,
954        it was not handled as intended. [But see 16 below.]
955    
956    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
957        Sheri Pierce.
958    
959    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
960        the first byte in a match must be "a".
961    
962    11. Change 17 for 8.13 increased the recursion depth for patterns like
963        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
964        pattern contains any instances of (*THEN). If it does not, the old
965        optimizations are restored. It would be nice to do this on a per-group
966        basis, but at the moment that is not feasible.
967    
968    12. In some environments, the output of pcretest -C is CRLF terminated. This
969        broke RunTest's code that checks for the link size. A single white space
970        character after the value is now allowed for.
971    
972    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
973        For "fr", it uses the Windows-specific input and output files.
974    
975    14. If (*THEN) appeared in a group that was called recursively or as a
976        subroutine, it did not work as intended. [But see next item.]
977    
978    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
979        pattern fragments (but not containing any | characters). If A and B are
980        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
981        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
982        D. In other words, Perl considers parentheses that do not contain any |
983        characters to be part of a surrounding alternative, whereas PCRE was
984        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
985        differently. PCRE now behaves in the same way as Perl, except in the case
986        of subroutine/recursion calls such as (?1) which have in any case always
987        been different (but PCRE had them first :-).
988    
989    16. Related to 15 above: Perl does not treat the | in a conditional group as
990        creating alternatives. Such a group is treated in the same way as an
991        ordinary group without any | characters when processing (*THEN). PCRE has
992        been changed to match Perl's behaviour.
993    
994    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
995        RunGrepTest script failed.
996    
997    18. Change 22 for version 13 caused atomic groups to use more stack. This is
998        inevitable for groups that contain captures, but it can lead to a lot of
999        stack use in large patterns. The old behaviour has been restored for atomic
1000        groups that do not contain any capturing parentheses.
1001    
1002    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1003        suppress the check for a minimum subject length at run time. (If it was
1004        given to pcre_exec() or pcre_dfa_exec() it did work.)
1005    
1006    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1007        fail to work when decoding hex characters in data strings in EBCDIC
1008        environments.
1009    
1010    21. It appears that in at least one Mac OS environment, the isxdigit() function
1011        is implemented as a macro that evaluates to its argument more than once,
1012        contravening the C 90 Standard (I haven't checked a later standard). There
1013        was an instance in pcretest which caused it to go wrong when processing
1014        \x{...} escapes in subject strings. The has been rewritten to avoid using
1015        things like p++ in the argument of isxdigit().
1016    
1017    
1018    Version 8.13 16-Aug-2011
1019    ------------------------
1020    
1021    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1022    
1023    2.  Two minor typos in pcre_internal.h have been fixed.
1024    
1025    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1026        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1027        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1028    
1029    4.  There were a number of related bugs in the code for matching backrefences
1030        caselessly in UTF-8 mode when codes for the characters concerned were
1031        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1032        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1033        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1034        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1035        2-byte code at the end of the subject (it thought there wasn't enough data
1036        left).
1037    
1038    5.  Comprehensive information about what went wrong is now returned by
1039        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1040        as the output vector has at least 2 elements. The offset of the start of
1041        the failing character and a reason code are placed in the vector.
1042    
1043    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1044        now returned is for the first byte of the failing character, instead of the
1045        last byte inspected. This is an incompatible change, but I hope it is small
1046        enough not to be a problem. It makes the returned offset consistent with
1047        pcre_exec() and pcre_dfa_exec().
1048    
1049    7.  pcretest now gives a text phrase as well as the error number when
1050        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1051        failure, the offset and reason code are output.
1052    
1053    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1054        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1055        back over a single character (\n). This seems wrong (because it treated the
1056        two characters as a single entity when going forwards), conflicts with the
1057        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1058        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1059        has been changed.
1060    
1061    9.  Some internal refactoring has changed the processing so that the handling
1062        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1063        time (the PCRE_DOTALL option was changed this way some time ago: version
1064        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1065        which was always a bit of a fudge. It also means that there is one less
1066        argument for the match() function, which reduces its stack requirements
1067        slightly. This change also fixes an incompatibility with Perl: the pattern
1068        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1069    
1070    10. More internal refactoring has drastically reduced the number of recursive
1071        calls to match() for possessively repeated groups such as (abc)++ when
1072        using pcre_exec().
1073    
1074    11. While implementing 10, a number of bugs in the handling of groups were
1075        discovered and fixed:
1076    
1077        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1078        (a|)*(?1) gave a compile-time internal error.
1079        ((a|)+)+  did not notice that the outer group could match an empty string.
1080        (^a|^)+   was not marked as anchored.
1081        (.*a|.*)+ was not marked as matching at start or after a newline.
1082    
1083    12. Yet more internal refactoring has removed another argument from the match()
1084        function. Special calls to this function are now indicated by setting a
1085        value in a variable in the "match data" data block.
1086    
1087    13. Be more explicit in pcre_study() instead of relying on "default" for
1088        opcodes that mean there is no starting character; this means that when new
1089        ones are added and accidentally left out of pcre_study(), testing should
1090        pick them up.
1091    
1092    14. The -s option of pcretest has been documented for ages as being an old
1093        synonym of -m (show memory usage). I have changed it to mean "force study
1094        for every regex", that is, assume /S for every regex. This is similar to -i
1095        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1096        using it. It makes it easier to run collections of tests with and without
1097        study enabled, and thereby test pcre_study() more easily. All the standard
1098        tests are now run with and without -s (but some patterns can be marked as
1099        "never study" - see 20 below).
1100    
1101    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1102        restoration of the capturing data to the outer values was not happening
1103        correctly.
1104    
1105    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1106        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1107        pattern had matched an empty string, and so incorrectly returned a no
1108        match.
1109    
1110    17. There was optimizing code for the last branch of non-capturing parentheses,
1111        and also for the obeyed branch of a conditional subexpression, which used
1112        tail recursion to cut down on stack usage. Unfortunately, now that there is
1113        the possibility of (*THEN) occurring in these branches, tail recursion is
1114        no longer possible because the return has to be checked for (*THEN). These
1115        two optimizations have therefore been removed. [But see 8.20/11 above.]
1116    
1117    18. If a pattern containing \R was studied, it was assumed that \R always
1118        matched two bytes, thus causing the minimum subject length to be
1119        incorrectly computed because \R can also match just one byte.
1120    
1121    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1122        was incorrectly computed.
1123    
1124    20. If /S is present twice on a test pattern in pcretest input, it now
1125        *disables* studying, thereby overriding the use of -s on the command line
1126        (see 14 above). This is necessary for one or two tests to keep the output
1127        identical in both cases.
1128    
1129    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1130        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1131    
1132    22. When an atomic group that contained a capturing parenthesis was
1133        successfully matched, but the branch in which it appeared failed, the
1134        capturing was not being forgotten if a higher numbered group was later
1135        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1136        group 1 to "a", when in fact it should be unset. This applied to multi-
1137        branched capturing and non-capturing groups, repeated or not, and also to
1138        positive assertions (capturing in negative assertions does not happen
1139        in PCRE) and also to nested atomic groups.
1140    
1141    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1142        subject after a captured substring, to make it easier to tell which of a
1143        number of identical substrings has been captured.
1144    
1145    24. The way atomic groups are processed by pcre_exec() has been changed so that
1146        if they are repeated, backtracking one repetition now resets captured
1147        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1148        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1149        "aaa". Previously, it would have been "a". As part of this code
1150        refactoring, the way recursive calls are handled has also been changed.
1151    
1152    25. If an assertion condition captured any substrings, they were not passed
1153        back unless some other capturing happened later. For example, if
1154        (?(?=(a))a) was matched against "a", no capturing was returned.
1155    
1156    26. When studying a pattern that contained subroutine calls or assertions,
1157        the code for finding the minimum length of a possible match was handling
1158        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1159        group 1 called group 2 while simultaneously a separate group 2 called group
1160        1). A stack overflow occurred in this case. I have fixed this by limiting
1161        the recursion depth to 10.
1162    
1163    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1164        Fortmann. This supports explicit test numbers on the command line, and has
1165        argument validation and error reporting.
1166    
1167    28. An instance of \X with an unlimited repeat could fail if at any point the
1168        first character it looked at was a mark character.
1169    
1170    29. Some minor code refactoring concerning Unicode properties and scripts
1171        should reduce the stack requirement of match() slightly.
1172    
1173    30. Added the '=' option to pcretest to check the setting of unused capturing
1174        slots at the end of the pattern, which are documented as being -1, but are
1175        not included in the return count.
1176    
1177    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1178        compiled something random. Now it gives a compile-time error (as does
1179        Perl).
1180    
1181    32. A *MARK encountered during the processing of a positive assertion is now
1182        recorded and passed back (compatible with Perl).
1183    
1184    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1185        had alternative anchored branches, the search for a second match in a line
1186        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1187        matched the line "0102" twice. The same bug affected patterns that started
1188        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1189        twice.
1190    
1191    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1192        does, and because of capturing effects, quantifying parenthesized
1193        assertions may at times be useful. Quantifiers are now allowed for
1194        parenthesized assertions.
1195    
1196    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1197    
1198    36. \g was being checked for fancy things in a character class, when it should
1199        just be a literal "g".
1200    
1201    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1202        appearance of a nested POSIX class supersedes an apparent external class.
1203        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1204        unescaped square brackets may also appear as part of class names. For
1205        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1206        more like Perl. (But see 8.20/1 above.)
1207    
1208    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1209        was because it thought it was \N{name}, which is not supported).
1210    
1211    39. Add minix to OS list not supporting the -S option in pcretest.
1212    
1213    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1214        cannot analyze patterns in sufficient detail to catch mutual recursions
1215        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1216        subgroup is called recursively as a subpattern for a second time at the
1217        same position in the subject string. In previous releases this might have
1218        been caught by the recursion limit, or it might have run out of stack.
1219    
1220    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1221        happen only once. PCRE was, however incorrectly giving a compile time error
1222        "recursive call could loop indefinitely" because it cannot analyze the
1223        pattern in sufficient detail. The compile time test no longer happens when
1224        PCRE is compiling a conditional subpattern, but actual runaway loops are
1225        now caught at runtime (see 40 above).
1226    
1227    42. It seems that Perl allows any characters other than a closing parenthesis
1228        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1229        has been changed to be the same.
1230    
1231    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1232        as not to get warnings when autogen.sh is called. Also changed
1233        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1234    
1235    44. To help people who use pcregrep to scan files containing exceedingly long
1236        lines, the following changes have been made:
1237    
1238        (a) The default value of the buffer size parameter has been increased from
1239            8K to 20K. (The actual buffer used is three times this size.)
1240    
1241        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1242            PCRE is built.
1243    
1244        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1245            to be set at run time.
1246    
1247        (d) Numerical values in pcregrep options can be followed by K or M, for
1248            example --buffer-size=50K.
1249    
1250        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1251            given and the return code is set to 2.
1252    
1253    45. Add a pointer to the latest mark to the callout data block.
1254    
1255    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1256        partial match of an empty string instead of no match. This was specific to
1257        the use of ".".
1258    
1259    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1260        complete match instead of a partial match. This bug was dependent on both
1261        the PCRE_UTF8 and PCRE_DOTALL options being set.
1262    
1263    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1264        starting byte set, because \b was not being ignored.
1265    
1266    
1267    Version 8.12 15-Jan-2011
1268    ------------------------
1269    
1270    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1271        checks for such things as part of the documentation building process.
1272    
1273    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1274        --match-limit and --recursion-limit options (added for 8.11). In
1275        particular, this made one of the standard tests fail. (The integer value
1276        went into the wrong half of a long int.)
1277    
1278    3.  If the --colour option was given to pcregrep with -v (invert match), it
1279        did strange things, either producing crazy output, or crashing. It should,
1280        of course, ignore a request for colour when reporting lines that do not
1281        match.
1282    
1283    4.  Another pcregrep bug caused similar problems if --colour was specified with
1284        -M (multiline) and the pattern match finished with a line ending.
1285    
1286    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1287        matched in multiline mode, the following line was shown as part of the
1288        match. This seems wrong, so I have changed it.
1289    
1290    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1291        the check for further matches in the same line (so they could be coloured)
1292        to overrun the end of the current line. If another match was found, it was
1293        incorrectly shown (and then shown again when found in the next line).
1294    
1295    7.  If pcregrep was compiled under Windows, there was a reference to the
1296        function pcregrep_exit() before it was defined. I am assuming this was
1297        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1298        reported by a user. I've moved the definition above the reference.
1299    
1300    
1301    Version 8.11 10-Dec-2010
1302    ------------------------
1303    
1304    1.  (*THEN) was not working properly if there were untried alternatives prior
1305        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1306        backtracked to try for "b" instead of moving to the next alternative branch
1307        at the same level (in this case, to look for "c"). The Perl documentation
1308        is clear that when (*THEN) is backtracked onto, it goes to the "next
1309        alternative in the innermost enclosing group".
1310    
1311    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1312        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1313        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1314        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1315        (*THEN).
1316    
1317    3.  If \s appeared in a character class, it removed the VT character from
1318        the class, even if it had been included by some previous item, for example
1319        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1320        of \s, but is part of the POSIX "space" class.)
1321    
1322    4.  A partial match never returns an empty string (because you can always
1323        match an empty string at the end of the subject); however the checking for
1324        an empty string was starting at the "start of match" point. This has been
1325        changed to the "earliest inspected character" point, because the returned
1326        data for a partial match starts at this character. This means that, for
1327        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1328        (previously it gave "no match").
1329    
1330    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1331        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1332        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1333        has an implication that the given string is incomplete (because a partial
1334        match is preferred over a full match). For this reason, these items now
1335        give a partial match in this situation. [Aside: previously, the one case
1336        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1337        match rather than a full match, which was wrong by the old rules, but is
1338        now correct.]
1339    
1340    6.  There was a bug in the handling of #-introduced comments, recognized when
1341        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1342        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1343        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1344        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1345        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1346        places in pcre_compile().
1347    
1348    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1349        comments when looking ahead for named forward references to subpatterns,
1350        the only newline sequence it recognized was NL. It now handles newlines
1351        according to the set newline convention.
1352    
1353    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1354        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1355        cater for a lack of strerror(). These oversights have been fixed.
1356    
1357    9.  Added --match-limit and --recursion-limit to pcregrep.
1358    
1359    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1360    
1361    11. When the -o option was used, pcregrep was setting a return code of 1, even
1362        when matches were found, and --line-buffered was not being honoured.
1363    
1364    12. Added an optional parentheses number to the -o and --only-matching options
1365        of pcregrep.
1366    
1367    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1368        can match an empty string. The code to do it in pcretest and pcredemo
1369        needed fixing:
1370    
1371        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1372            only one byte after an empty string match just before CRLF (this case
1373            just got forgotten; "any" and "anycrlf" were OK).
1374    
1375        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1376            mode when an empty string match preceded an ASCII character followed by
1377            a non-ASCII character. (The code for advancing by one character rather
1378            than one byte was nonsense.)
1379    
1380        (c) The pcredemo.c sample program did not have any code at all to handle
1381            the cases when CRLF is a valid newline sequence.
1382    
1383    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1384        as a starting offset was within the subject string. There is now a new
1385        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1386        negative or greater than the length of the string. In order to test this,
1387        pcretest is extended to allow the setting of negative starting offsets.
1388    
1389    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1390        starting offset points to the beginning of a UTF-8 character was
1391        unnecessarily clumsy. I tidied it up.
1392    
1393    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1394        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1395    
1396    17. Nobody had reported that the --include_dir option, which was added in
1397        release 7.7 should have been called --include-dir (hyphen, not underscore)
1398        for compatibility with GNU grep. I have changed it to --include-dir, but
1399        left --include_dir as an undocumented synonym, and the same for
1400        --exclude-dir, though that is not available in GNU grep, at least as of
1401        release 2.5.4.
1402    
1403    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1404        characters from a string of bytes) have been redefined so as not to use
1405        loops, in order to improve performance in some environments. At the same
1406        time, I abstracted some of the common code into auxiliary macros to save
1407        repetition (this should not affect the compiled code).
1408    
1409    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1410        compile-time error is now given if \c is not followed by an ASCII
1411        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1412        different, and any byte value is allowed.)
1413    
1414    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1415        START_OPTIMIZE option, which is now allowed at compile time - but just
1416        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1417        to pcregrep and other applications that have no direct access to PCRE
1418        options. The new /Y option in pcretest sets this option when calling
1419        pcre_compile().
1420    
1421    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1422        back references. Groups containing recursive back references were forced to
1423        be atomic by that change, but in the case of named groups, the amount of
1424        memory required was incorrectly computed, leading to "Failed: internal
1425        error: code overflow". This has been fixed.
1426    
1427    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1428        pcretest.c, to avoid build problems in some Borland environments.
1429    
1430    
1431    Version 8.10 25-Jun-2010
1432    ------------------------
1433    
1434    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1435        THEN.
1436    
1437    2.  (*ACCEPT) was not working when inside an atomic group.
1438    
1439    3.  Inside a character class, \B is treated as a literal by default, but
1440        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1441        causes the error). The code is unchanged, but I tidied the documentation.
1442    
1443    4.  Inside a character class, PCRE always treated \R and \X as literals,
1444        whereas Perl faults them if its -w option is set. I have changed PCRE so
1445        that it faults them when PCRE_EXTRA is set.
1446    
1447    5.  Added support for \N, which always matches any character other than
1448        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1449    
1450    6.  When compiling pcregrep with newer versions of gcc which may have
1451        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1452        declared with attribute warn_unused_result" were given. Just casting the
1453        result to (void) does not stop the warnings; a more elaborate fudge is
1454        needed. I've used a macro to implement this.
1455    
1456    7.  Minor change to pcretest.c to avoid a compiler warning.
1457    
1458    8.  Added four artifical Unicode properties to help with an option to make
1459        \s etc use properties (see next item). The new properties are: Xan
1460        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1461    
1462    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1463        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1464        this option. Modified pcretest to add /W to test this facility. Added
1465        REG_UCP to make it available via the POSIX interface.
1466    
1467    10. Added --line-buffered to pcregrep.
1468    
1469    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1470        studied, and the match started with a letter with a code point greater than
1471        127 whose first byte was different to the first byte of the other case of
1472        the letter, the other case of this starting letter was not recognized
1473        (#976).
1474    
1475    12. If a pattern that was studied started with a repeated Unicode property
1476        test, for example, \p{Nd}+, there was the theoretical possibility of
1477        setting up an incorrect bitmap of starting bytes, but fortunately it could
1478        not have actually happened in practice until change 8 above was made (it
1479        added property types that matched character-matching opcodes).
1480    
1481    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1482        possible starting bytes for non-anchored patterns.
1483    
1484    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1485        \R, and also a number of cases that involve Unicode properties, both
1486        explicit and implicit when PCRE_UCP is set.
1487    
1488    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1489        input, it could crash or give wrong results if characters with values
1490        greater than 0xc0 were present in the subject string. (Detail: it assumed
1491        UTF-8 input when processing these items.)
1492    
1493    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1494        size_t is 64-bit (#991).
1495    
1496    17. Added a check for running out of memory when PCRE is compiled with
1497        --disable-stack-for-recursion (#990).
1498    
1499    18. If the last data line in a file for pcretest does not have a newline on
1500        the end, a newline was missing in the output.
1501    
1502    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1503        less than 128) in its various bitmaps. However, there is a facility for
1504        generating tables according to the current locale when PCRE is compiled. It
1505        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1506        characters, are recognized by isspace() and therefore were getting set in
1507        these tables, and indeed these tables seem to approximate to ISO 8859. This
1508        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1509        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1510        which of course cannot start UTF-8 characters. I have changed the code so
1511        that only real ASCII characters (less than 128) and the correct starting
1512        bytes for UTF-8 encodings are set for characters greater than 127 when in
1513        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1514        altogether.)
1515    
1516    20. Added the /T option to pcretest so as to be able to run tests with non-
1517        standard character tables, thus making it possible to include the tests
1518        used for 19 above in the standard set of tests.
1519    
1520    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1521        reference to a subpattern the other side of a comment that contains an
1522        opening parenthesis caused either an internal compiling error, or a
1523        reference to the wrong subpattern.
1524    
1525    
1526    Version 8.02 19-Mar-2010
1527    ------------------------
1528    
1529    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1530    
1531    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1532        configured.
1533    
1534    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1535        original author of that file, following a query about its status.
1536    
1537    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1538        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1539    
1540    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1541        quantifier applied to a forward-referencing subroutine call, could compile
1542        incorrect code or give the error "internal error: previously-checked
1543        referenced subpattern not found".
1544    
1545    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1546        variables to point to external functions. For these systems, therefore,
1547        pcre_malloc etc. are now initialized to local functions that call the
1548        relevant global functions.
1549    
1550    7.  There were two entries missing in the vectors called coptable and poptable
1551        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1552        I've fixed the data, and added a kludgy way of testing at compile time that
1553        the lengths are correct (equal to the number of opcodes).
1554    
1555    8.  Following on from 7, I added a similar kludge to check the length of the
1556        eint vector in pcreposix.c.
1557    
1558    9.  Error texts for pcre_compile() are held as one long string to avoid too
1559        much relocation at load time. To find a text, the string is searched,
1560        counting zeros. There was no check for running off the end of the string,
1561        which could happen if a new error number was added without updating the
1562        string.
1563    
1564    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1565    
1566    11. \K was not working if it appeared in an atomic group or in a group that
1567        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1568        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1569        the assertion is positive, but not if it is negative.
1570    
1571    12. Change 11 fortuitously reduced the size of the stack frame used in the
1572        "match()" function of pcre_exec.c by one pointer. Forthcoming
1573        implementation of support for (*MARK) will need an extra pointer on the
1574        stack; I have reserved it now, so that the stack frame size does not
1575        decrease.
1576    
1577    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1578        item in branch that calls a recursion is a subroutine call - as in the
1579        second branch in the above example - was incorrectly given the compile-
1580        time error "recursive call could loop indefinitely" because pcre_compile()
1581        was not correctly checking the subroutine for matching a non-empty string.
1582    
1583    14. The checks for overrunning compiling workspace could trigger after an
1584        overrun had occurred. This is a "should never occur" error, but it can be
1585        triggered by pathological patterns such as hundreds of nested parentheses.
1586        The checks now trigger 100 bytes before the end of the workspace.
1587    
1588    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1589    
1590    
1591    Version 8.01 19-Jan-2010
1592    ------------------------
1593    
1594    1.  If a pattern contained a conditional subpattern with only one branch (in
1595        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1596        computed the wrong minimum data length (which is of course zero for such
1597        subpatterns). This could cause incorrect "no match" results.
1598    
1599    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1600        the pattern is reset in the first branch, pcre_compile() failed with
1601        "internal error: code overflow at offset...". This happened only when
1602        the reset was to the original external option setting. (An optimization
1603        abstracts leading options settings into an external setting, which was the
1604        cause of this.)
1605    
1606    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1607        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1608        assertion pattern did not match (meaning that the assertion was true), it
1609        was incorrectly treated as false if the SKIP had been reached during the
1610        matching. This also applied to assertions used as conditions.
1611    
1612    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1613        assertion subpattern, including such a pattern used as a condition,
1614        unpredictable results occurred, instead of the error return
1615        PCRE_ERROR_DFA_UITEM.
1616    
1617    5.  The C++ GlobalReplace function was not working like Perl for the special
1618        situation when an empty string is matched. It now does the fancy magic
1619        stuff that is necessary.
1620    
1621    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1622        removed. (These were left over from very, very early versions of PCRE.)
1623    
1624    7.  Some cosmetic changes to the code to make life easier when compiling it
1625        as part of something else:
1626    
1627        (a) Change DEBUG to PCRE_DEBUG.
1628    
1629        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1630            called "current" as "current_branch", to prevent a collision with the
1631            Linux macro when compiled as a kernel module.
1632    
1633        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1634            prevent a collision with the Linux macro when compiled as a kernel
1635            module.
1636    
1637    8.  In pcre_compile() there are some checks for integer overflows that used to
1638        cast potentially large values to (double). This has been changed to that
1639        when building, a check for int64_t is made, and if it is found, it is used
1640        instead, thus avoiding the use of floating point arithmetic. (There is no
1641        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1642        double.
1643    
1644    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1645        2005 (difference between two addresses compared to an unsigned value).
1646    
1647    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1648        custom one, because of the following reported problem in Windows:
1649    
1650          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1651              under Win32.
1652          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1653              therefore missing the function definition.
1654          - The compiler thus generates a "C" signature for the test function.
1655          - The linker fails to find the "C" function.
1656          - PCRE fails to configure if asked to do so against libbz2.
1657    
1658    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1659        messages were output:
1660    
1661          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1662          rerunning libtoolize, to keep the correct libtool macros in-tree.
1663          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1664    
1665        I have done both of these things.
1666    
1667    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1668        most of the time, it *can* run out if it is given a pattern that contains a
1669        runaway infinite recursion. I updated the discussion in the pcrestack man
1670        page.
1671    
1672    13. Now that we have gone to the x.xx style of version numbers, the minor
1673        version may start with zero. Using 08 or 09 is a bad idea because users
1674        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1675        interpreted as invalid octal numbers. I've updated the previous comment in
1676        configure.ac, and also added a check that gives an error if 08 or 09 are
1677        used.
1678    
1679    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1680        causing partial matching to fail when the end of the subject matched \W
1681        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1682    
1683    15. There were some discrepancies between the declarations in pcre_internal.h
1684        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1685        their definitions. The declarations used "const uschar *" and the
1686        definitions used USPTR. Even though USPTR is normally defined as "const
1687        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1688        reported that: "This difference in casting confuses some C++ compilers, for
1689        example, SunCC recognizes above declarations as different functions and
1690        generates broken code for hbpcre." I have changed the declarations to use
1691        USPTR.
1692    
1693    16. GNU libtool is named differently on some systems. The autogen.sh script now
1694        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1695        (FreeBSD).
1696    
1697    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1698        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1699        comment: "Figure out how to create a longlong from a string: strtoll and
1700        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1701        instance, but it only takes 2 args instead of 3!"
1702    
1703    18. A subtle bug concerned with back references has been fixed by a change of
1704        specification, with a corresponding code fix. A pattern such as
1705        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1706        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1707        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1708        same bug. Such groups have to be quantified to be useful, or contained
1709        inside another quantified group. (If there's no repetition, the reference
1710        can never match.) The problem arises because, having left the group and
1711        moved on to the rest of the pattern, a later failure that backtracks into
1712        the group uses the captured value from the final iteration of the group
1713        rather than the correct earlier one. I have fixed this in PCRE by forcing
1714        any group that contains a reference to itself to be an atomic group; that
1715        is, there cannot be any backtracking into it once it has completed. This is
1716        similar to recursive and subroutine calls.
1717    
1718    
1719    Version 8.00 19-Oct-09
1720    ----------------------
1721    
1722    1.  The table for translating pcre_compile() error codes into POSIX error codes
1723        was out-of-date, and there was no check on the pcre_compile() error code
1724        being within the table. This could lead to an OK return being given in
1725        error.
1726    
1727    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1728        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1729        in a Windows environment.
1730    
1731    3.  The pcregrep --count option prints the count for each file even when it is
1732        zero, as does GNU grep. However, pcregrep was also printing all files when
1733        --files-with-matches was added. Now, when both options are given, it prints
1734        counts only for those files that have at least one match. (GNU grep just
1735        prints the file name in this circumstance, but including the count seems
1736        more useful - otherwise, why use --count?) Also ensured that the
1737        combination -clh just lists non-zero counts, with no names.
1738    
1739    4.  The long form of the pcregrep -F option was incorrectly implemented as
1740        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1741        but it seems right to fix it, and I didn't think it was worth preserving
1742        the old behaviour.
1743    
1744    5.  The command line items --regex=pattern and --regexp=pattern were not
1745        recognized by pcregrep, which required --regex pattern or --regexp pattern
1746        (with a space rather than an '='). The man page documented the '=' forms,
1747        which are compatible with GNU grep; these now work.
1748    
1749    6.  No libpcreposix.pc file was created for pkg-config; there was just
1750        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1751    
1752    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1753        when UCP support is not needed, by modifying the Python script that
1754        generates it from Unicode data files. This should not matter if the module
1755        is correctly used as a library, but I received one complaint about 50K of
1756        unwanted data. My guess is that the person linked everything into his
1757        program rather than using a library. Anyway, it does no harm.
1758    
1759    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1760        was a minimum greater than 1 for a wide character in a possessive
1761        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1762        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1763        character. Chaos in the form of incorrect output or a compiling loop could
1764        result.
1765    
1766    9.  The restrictions on what a pattern can contain when partial matching is
1767        requested for pcre_exec() have been removed. All patterns can now be
1768        partially matched by this function. In addition, if there are at least two
1769        slots in the offset vector, the offset of the earliest inspected character
1770        for the match and the offset of the end of the subject are set in them when
1771        PCRE_ERROR_PARTIAL is returned.
1772    
1773    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1774        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1775        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1776        and may be more useful for multi-segment matching.
1777    
1778    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1779        used to be given if ever the end of the subject was reached; now it is
1780        given only if matching could not proceed because another character was
1781        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1782        string "Z", which now yields "no match" instead of "partial match". In the
1783        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1784        final character ended with (*FAIL).
1785    
1786    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1787        if the pattern had a "must contain" character that was already found in the
1788        earlier partial match, unless partial matching was again requested. For
1789        example, with the pattern /dog.(body)?/, the "must contain" character is
1790        "g". If the first part-match was for the string "dog", restarting with
1791        "sbody" failed. This bug has been fixed.
1792    
1793    13. The string returned by pcre_dfa_exec() after a partial match has been
1794        changed so that it starts at the first inspected character rather than the
1795        first character of the match. This makes a difference only if the pattern
1796        starts with a lookbehind assertion or \b or \B (\K is not supported by
1797        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1798        matching functions compatible, and I think it's the right thing to do.
1799    
1800    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1801        so that the demonstration program is easily available in environments where
1802        PCRE has not been installed from source.
1803    
1804    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1805        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1806        library.
1807    
1808    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1809        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1810        is not the first non-POSIX option to be added. Clearly some people find
1811        these options useful.
1812    
1813    17. If a caller to the POSIX matching function regexec() passes a non-zero
1814        value for nmatch with a NULL value for pmatch, the value of
1815        nmatch is forced to zero.
1816    
1817    18. RunGrepTest did not have a test for the availability of the -u option of
1818        the diff command, as RunTest does. It now checks in the same way as
1819        RunTest, and also checks for the -b option.
1820    
1821    19. If an odd number of negated classes containing just a single character
1822        interposed, within parentheses, between a forward reference to a named
1823        subpattern and the definition of the subpattern, compilation crashed with
1824        an internal error, complaining that it could not find the referenced
1825        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1826        [The bug was that it was starting one character too far in when skipping
1827        over the character class, thus treating the ] as data rather than
1828        terminating the class. This meant it could skip too much.]
1829    
1830    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1831        /g option in pcretest when the pattern contains \K, which makes it possible
1832        to have an empty string match not at the start, even when the pattern is
1833        anchored. Updated pcretest and pcredemo to use this option.
1834    
1835    21. If the maximum number of capturing subpatterns in a recursion was greater
1836        than the maximum at the outer level, the higher number was returned, but
1837        with unset values at the outer level. The correct (outer level) value is
1838        now given.
1839    
1840    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1841        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1842        make it do so. The string so far is captured, making this feature
1843        compatible with Perl.
1844    
1845    23. The tests have been re-organized, adding tests 11 and 12, to make it
1846        possible to check the Perl 5.10 features against Perl 5.10.
1847    
1848    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1849        pattern matches a fixed length string. PCRE did not allow this; now it
1850        does. Neither allows recursion.
1851    
1852    25. I finally figured out how to implement a request to provide the minimum
1853        length of subject string that was needed in order to match a given pattern.
1854        (It was back references and recursion that I had previously got hung up
1855        on.) This code has now been added to pcre_study(); it finds a lower bound
1856        to the length of subject needed. It is not necessarily the greatest lower
1857        bound, but using it to avoid searching strings that are too short does give
1858        some useful speed-ups. The value is available to calling programs via
1859        pcre_fullinfo().
1860    
1861    26. While implementing 25, I discovered to my embarrassment that pcretest had
1862        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1863        study optimizations had never been tested with that matching function.
1864        Oops. What is worse, even when it was passed study data, there was a bug in
1865        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1866        were also very few tests of studied patterns with pcre_dfa_exec().
1867    
1868    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1869        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1870        on the other side of the coin, they are no longer allowed to have different
1871        names, because these cannot be distinguished in PCRE, and this has caused
1872        confusion. (This is a difference from Perl.)
1873    
1874    28. When duplicate subpattern names are present (necessarily with different
1875        numbers, as required by 27 above), and a test is made by name in a
1876        conditional pattern, either for a subpattern having been matched, or for
1877        recursion in such a pattern, all the associated numbered subpatterns are
1878        tested, and the overall condition is true if the condition is true for any
1879        one of them. This is the way Perl works, and is also more like the way
1880        testing by number works.
1881    
1882    
1883    Version 7.9 11-Apr-09
1884    ---------------------
1885    
1886    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1887        (pcretest), all targets were linked against these libraries. This included
1888        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1889        libraries. This caused unwanted dependencies to be created. This problem
1890        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1891        pcretest is linked with readline.
1892    
1893    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1894        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1895        moved outside it again, because FALSE and TRUE are already defined in AIX,
1896        but BOOL is not.
1897    
1898    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1899        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1900    
1901    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1902        hyphens) following file names and line numbers when outputting matching
1903        lines. This is not true; no spaces are inserted. I have also clarified the
1904        wording for the --colour (or --color) option.
1905    
1906    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1907        was not coloured; this is different to GNU grep, so I have changed it to be
1908        the same.
1909    
1910    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1911        each matching line was coloured. Now it goes on to look for further matches
1912        of any of the test patterns, which is the same behaviour as GNU grep.
1913    
1914    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1915        doesn't make sense to accept an empty string match in pcregrep, so I have
1916        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1917        seems to be how GNU grep behaves. [But see later change 40 for release
1918        8.33.]
1919    
1920    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1921        start or after a newline", because the conditional assertion was not being
1922        correctly handled. The rule now is that both the assertion and what follows
1923        in the first alternative must satisfy the test.
1924    
1925    9.  If auto-callout was enabled in a pattern with a conditional group whose
1926        condition was an assertion, PCRE could crash during matching, both with
1927        pcre_exec() and pcre_dfa_exec().
1928    
1929    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1930        used for matching.
1931    
1932    11. Unicode property support in character classes was not working for
1933        characters (bytes) greater than 127 when not in UTF-8 mode.
1934    
1935    12. Added the -M command line option to pcretest.
1936    
1937    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1938    
1939    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1940    
1941    16. Added comments and documentation about mis-use of no_arg in the C++
1942        wrapper.
1943    
1944    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1945        from Martin Jerabek that uses macro names for all relevant character and
1946        string constants.
1947    
1948    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1949        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1950        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1951        these, but not everybody uses configure.
1952    
1953    19. A conditional group that had only one branch was not being correctly
1954        recognized as an item that could match an empty string. This meant that an
1955        enclosing group might also not be so recognized, causing infinite looping
1956        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1957        with the subject "ab", where knowledge that the repeated group can match
1958        nothing is needed in order to break the loop.
1959    
1960    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1961        exec(), but without supplying a callout function, matching went wrong.
1962    
1963    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1964        leak if the size of the offset vector was greater than 30. When the vector
1965        is smaller, the saved offsets during recursion go onto a local stack
1966        vector, but for larger vectors malloc() is used. It was failing to free
1967        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1968        error, in fact).
1969    
1970    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1971        heapframe that is used only when UTF-8 support is enabled. This caused no
1972        problem, but was untidy.
1973    
1974    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1975        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1976        included within another project.
1977    
1978    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1979        slightly modified by me:
1980    
1981          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1982              not building pcregrep.
1983    
1984          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1985              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1986    
1987    25. Forward references, both numeric and by name, in patterns that made use of
1988        duplicate group numbers, could behave incorrectly or give incorrect errors,
1989        because when scanning forward to find the reference group, PCRE was not
1990        taking into account the duplicate group numbers. A pattern such as
1991        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1992    
1993    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1994        the feature of a custom pointer more persuasive (as requested by a user).
1995    
1996    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1997        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1998        pre-defined.
1999    
2000    28. Added support for (*UTF8) at the start of a pattern.
2001    
2002    29. Arrange for flags added by the "release type" setting in CMake to be shown
2003        in the configuration summary.
2004    
2005    
2006    Version 7.8 05-Sep-08
2007    ---------------------
2008    
2009    1.  Replaced UCP searching code with optimized version as implemented for Ad
2010        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2011        stage table and inline lookup instead of a function, giving speed ups of 2
2012        to 5 times on some simple patterns that I tested. Permission was given to
2013        distribute the MultiStage2.py script that generates the tables (it's not in
2014        the tarball, but is in the Subversion repository).
2015    
2016    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2017        scripts.
2018    
2019    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2020        a group with a zero qualifier. The result of the study could be incorrect,
2021        or the function might crash, depending on the pattern.
2022    
2023    4.  Caseless matching was not working for non-ASCII characters in back
2024        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2025        It now works when Unicode Property Support is available.
2026    
2027    5.  In pcretest, an escape such as \x{de} in the data was always generating
2028        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2029        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2030        truncation.
2031    
2032    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2033    
2034    7.  Added two (int) casts to pcregrep when printing the difference of two
2035        pointers, in case they are 64-bit values.
2036    
2037    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
2038        test 2 if it fails.
2039    
2040    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2041        and a #define of that name to empty if it is not externally set. This is to
2042        allow users of MSVC to set it if necessary.
2043    
2044    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2045        the convenience functions in the pcre_get.c source file.
2046    
2047    11. An option change at the start of a pattern that had top-level alternatives
2048        could cause overwriting and/or a crash. This command provoked a crash in
2049        some environments:
2050    
2051          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2052    
2053        This potential security problem was recorded as CVE-2008-2371.
2054    
2055    12. For a pattern where the match had to start at the beginning or immediately
2056        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2057        pcre_dfa_exec() could read past the end of the passed subject if there was
2058        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2059        pcretest so that it places the subject at the end of its malloc-ed buffer.
2060    
2061    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2062        exec() might read past the end of the data buffer in UTF-8 mode.
2063    
2064    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2065        the data contained the byte 0x85 as part of a UTF-8 character within its
2066        first line. This applied both to normal and DFA matching.
2067    
2068    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2069        /^[^d]*?$/8 failed to match "abc".
2070    
2071    16. Added a missing copyright notice to pcrecpp_internal.h.
2072    
2073    17. Make it more clear in the documentation that values returned from
2074        pcre_exec() in ovector are byte offsets, not character counts.
2075    
2076    18. Tidied a few places to stop certain compilers from issuing warnings.
2077    
2078    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2079        supplied by Stefan Weber. I made a further small update for 7.8 because
2080        there is a change of source arrangements: the pcre_searchfuncs.c module is
2081        replaced by pcre_ucd.c.
2082    
2083    
2084    Version 7.7 07-May-08
2085    ---------------------
2086    
2087    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2088        a string to a long long, pretend we don't even have a long long." This is
2089        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2090    
2091    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2092        pre-7.6 versions, which defined a global no_arg variable instead of putting
2093        it in the RE class. (See also #8 below.)
2094    
2095    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2096        Lopes.
2097    
2098    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2099    
2100        (1) The include/exclude patterns were being applied to the whole pathnames
2101            of files, instead of just to the final components.
2102    
2103        (2) If there was more than one level of directory, the subdirectories were
2104            skipped unless they satisfied the include/exclude conditions. This is
2105            inconsistent with GNU grep (and could even be seen as contrary to the
2106            pcregrep specification - which I improved to make it absolutely clear).
2107            The action now is always to scan all levels of directory, and just
2108            apply the include/exclude patterns to regular files.
2109    
2110    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2111        --exclude_dir in the tests to avoid scanning .svn directories.
2112    
2113    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2114        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2115        doesn't support NULs in patterns.
2116    
2117    7.  Added some missing "const"s to declarations of static tables in
2118        pcre_compile.c and pcre_dfa_exec.c.
2119    
2120    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2121        caused by fix #2  above. (Subsequently also a second patch to fix the
2122        first patch. And a third patch - this was a messy problem.)
2123    
2124    9.  Applied Craig's patch to remove the use of push_back().
2125    
2126    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2127        matching function regexec().
2128    
2129    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2130        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2131        references. PCRE supports relative numbers with this syntax (I don't think
2132        Oniguruma does).
2133    
2134    12. Previously, a group with a zero repeat such as (...){0} was completely
2135        omitted from the compiled regex. However, this means that if the group
2136        was called as a subroutine from elsewhere in the pattern, things went wrong
2137        (an internal error was given). Such groups are now left in the compiled
2138        pattern, with a new opcode that causes them to be skipped at execution
2139        time.
2140    
2141    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2142        to the way PCRE behaves:
2143    
2144        (a) A lone ] character is dis-allowed (Perl treats it as data).
2145    
2146        (b) A back reference to an unmatched subpattern matches an empty string
2147            (Perl fails the current match path).
2148    
2149        (c) A data ] in a character class must be notated as \] because if the
2150            first data character in a class is ], it defines an empty class. (In
2151            Perl it is not possible to have an empty class.) The empty class []
2152            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2153            The negative empty class [^] matches any one character, independently
2154            of the DOTALL setting.
2155    
2156    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2157        non-existent subpattern following a character class starting with ']' and
2158        containing () gave an internal compiling error instead of "reference to
2159        non-existent subpattern". Fortunately, when the pattern did exist, the
2160        compiled code was correct. (When scanning forwards to check for the
2161        existence of the subpattern, it was treating the data ']' as terminating
2162        the class, so got the count wrong. When actually compiling, the reference
2163        was subsequently set up correctly.)
2164    
2165    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2166        it was being rejected as not supported by pcre_dfa_exec(), even though
2167        other assertions are supported. I have made pcre_dfa_exec() support
2168        (*FAIL).
2169    
2170    16. The implementation of 13c above involved the invention of a new opcode,
2171        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2172        cannot be changed at match time, I realized I could make a small
2173        improvement to matching performance by compiling OP_ALLANY instead of
2174        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2175        on the OP_ANY path.
2176    
2177    17. Compiling pcretest on Windows with readline support failed without the
2178        following two fixes: (1) Make the unistd.h include conditional on
2179        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2180    
2181    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2182        ncurses library to be included for pcretest when ReadLine support is
2183        requested, but also to allow for it to be overridden. This patch came from
2184        Daniel Bergström.
2185    
2186    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2187        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2188        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2189        spotting this.
2190    
2191    
2192    Version 7.6 28-Jan-08
2193    ---------------------
2194    
2195    1.  A character class containing a very large number of characters with
2196        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2197        overflow.
2198    
2199    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2200        HAVE_LONG_LONG is not defined.
2201    
2202    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2203        bring them up to date and include new features. This patch includes:
2204    
2205        - Fixed PH's badly added libz and libbz2 support.
2206        - Fixed a problem with static linking.
2207        - Added pcredemo. [But later removed - see 7 below.]
2208        - Fixed dftables problem and added an option.
2209        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2210            HAVE_LONG_LONG.
2211        - Added readline support for pcretest.
2212        - Added an listing of the option settings after cmake has run.
2213    
2214    4.  A user submitted a patch to Makefile that makes it easy to create
2215        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2216        Makefile.am that cause it to include this special target, without
2217        affecting anything else. Note that the same mingw target plus all
2218        the other distribution libraries and programs are now supported
2219        when configuring with CMake (see 6 below) instead of with
2220        Configure/Make.
2221    
2222    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2223        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2224        exported in the Windows port". It has not yet been confirmed that the patch
2225        solves the problem, but it does no harm.
2226    
2227    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2228        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2229        with CMake, and also correct the comment about stack recursion.
2230    
2231    7.  Remove the automatic building of pcredemo from the ./configure system and
2232        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2233        of a program that users should build themselves after PCRE is installed, so
2234        building it automatically is not really right. What is more, it gave
2235        trouble in some build environments.
2236    
2237    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2238    
2239    
2240    Version 7.5 10-Jan-08
2241    ---------------------
2242    
2243    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2244        values in parens when parsing an RE using the C++ wrapper."
2245    
2246    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2247        Characters greater than 255 were excluded from the class instead of being
2248        included.
2249    
2250    3.  The same bug as (2) above applied to negated POSIX classes such as
2251        [:^space:].
2252    
2253    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2254        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2255        I have changed it.
2256    
2257    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2258        first named subpattern) and a construct such as (?&a) would reference the
2259        first named subpattern whose name started with "a" (in other words, the
2260        length check was missing). Both these problems are fixed. "Subpattern name
2261        expected" is now given for (?&) (a zero-length name), and this patch also
2262        makes it give the same error for \k'' (previously it complained that that
2263        was a reference to a non-existent subpattern).
2264    
2265    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2266        this is right because (?- can be followed by option settings as well as by
2267        digits. I have, however, made the messages clearer.
2268    
2269    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2270        than the number used in the conditional) now cause a compile-time error.
2271        This is actually not compatible with Perl, which accepts such patterns, but
2272        treats the conditional as always being FALSE (as PCRE used to), but it
2273        seems to me that giving a diagnostic is better.
2274    
2275    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2276        and messages.
2277    
2278    9.  Fix two occurrences of "backslash" in comments that should have been
2279        "backspace".
2280    
2281    10. Remove two redundant lines of code that can never be obeyed (their function
2282        was moved elsewhere).
2283    
2284    11. The program that makes PCRE's Unicode character property table had a bug
2285        which caused it to generate incorrect table entries for sequences of
2286        characters that have the same character type, but are in different scripts.
2287        It amalgamated them into a single range, with the script of the first of
2288        them. In other words, some characters were in the wrong script. There were
2289        thirteen such cases, affecting characters in the following ranges:
2290    
2291          U+002b0 - U+002c1
2292          U+0060c - U+0060d
2293          U+0061e - U+00612
2294          U+0064b - U+0065e
2295          U+0074d - U+0076d
2296          U+01800 - U+01805
2297          U+01d00 - U+01d77
2298          U+01d9b - U+01dbf
2299          U+0200b - U+0200f
2300          U+030fc - U+030fe
2301          U+03260 - U+0327f
2302          U+0fb46 - U+0fbb1
2303          U+10450 - U+1049d
2304    
2305    12. The -o option (show only the matching part of a line) for pcregrep was not
2306        compatible with GNU grep in that, if there was more than one match in a
2307        line, it showed only the first of them. It now behaves in the same way as
2308        GNU grep.
2309    
2310    13. If the -o and -v options were combined for pcregrep, it printed a blank
2311        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2312        does the same. The return code can be used to tell if there were any
2313        non-matching lines.
2314    
2315    14. Added --file-offsets and --line-offsets to pcregrep.
2316    
2317    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2318        infinitely looping recursion. The bug was that positive lookaheads were not
2319        being skipped when checking for a possible empty match (negative lookaheads
2320        and both kinds of lookbehind were skipped).
2321    
2322    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2323        inclusion of <windows.h> to before rather than after the definition of
2324        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2325    
2326    17. Specifying a possessive quantifier with a specific limit for a Unicode
2327        character property caused pcre_compile() to compile bad code, which led at
2328        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2329        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2330        caused the error; without that there was no problem.
2331    
2332    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2333    
2334    19. Added --enable-pcretest-libreadline.
2335    
2336    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2337        RE::GlobalReplace(). As a result, the number of replacements returned was
2338        double what it should be. I removed one of the increments, but Craig sent a
2339        later patch that removed the other one (the right fix) and added unit tests
2340        that check the return values (which was not done before).
2341    
2342    21. Several CMake things:
2343    
2344        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2345            the names libpcre and libpcreposix, not just pcre and pcreposix.
2346    
2347        (2) The above change means that pcretest and pcregrep are now correctly
2348            linked with the newly-built libraries, not previously installed ones.
2349    
2350        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2351    
2352    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2353        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2354        UTF-8 newline character). The key issue is that the pattern starts .*;
2355        this means that the match must be either at the beginning, or after a
2356        newline. The bug was in the code for advancing after a failed match and
2357        checking that the new position followed a newline. It was not taking
2358        account of UTF-8 characters correctly.
2359    
2360    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2361        character classes. PCRE was not treating the sequence [:...:] as a
2362        character class unless the ... were all letters. Perl, however, seems to
2363        allow any characters between [: and :], though of course it rejects as
2364        unknown any "names" that contain non-letters, because all the known class
2365        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2366        for example, whereas PCRE did not - it did not recognize a POSIX character
2367        class. This seemed a bit dangerous, so the code has been changed to be
2368        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2369        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2370        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2371        Perl does, and where it didn't before.
2372    
2373    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2374        Windows environments %n is disabled by default.
2375    
2376    
2377    Version 7.4 21-Sep-07
2378    ---------------------
2379    
2380    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2381        means that a class such as [\s] counted as "explicit reference to CR or
2382        LF". That isn't really right - the whole point of the change was to try to
2383        help when there was an actual mention of one of the two characters. So now
2384        the change happens only if \r or \n (or a literal CR or LF) character is
2385        encountered.
2386    
2387    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2388        of both had grown to the point where there were only 3 bits left.
2389        Fortunately, there was spare space in the data structure, and so I have
2390        moved the internal flags into a new 16-bit field to free up more option
2391        bits.
2392    
2393    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2394        but did not set the internal JCHANGED flag - either of these is enough to
2395        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2396        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2397        start sets both bits.
2398    
2399    4.  Added options (at build time, compile time, exec time) to change \R from
2400        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2401    
2402    5.  doc/pcresyntax.html was missing from the distribution.
2403    
2404    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2405        compatibility, even though it is no longer used.
2406    
2407    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2408        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2409        windows.h file is present (where different names are used). [This was
2410        reversed later after testing - see 16 below.]
2411    
2412    8.  Changed all #include <config.h> to #include "config.h". There were also
2413        some further <pcre.h> cases that I changed to "pcre.h".
2414    
2415    9.  When pcregrep was used with the --colour option, it missed the line ending
2416        sequence off the lines that it output.
2417    
2418    10. It was pointed out to me that arrays of string pointers cause lots of
2419        relocations when a shared library is dynamically loaded. A technique of
2420        using a single long string with a table of offsets can drastically reduce
2421        these. I have refactored PCRE in four places to do this. The result is
2422        dramatic:
2423    
2424          Originally:                          290
2425          After changing UCP table:            187
2426          After changing error message table:   43
2427          After changing table of "verbs"       36
2428          After changing table of Posix names   22
2429    
2430        Thanks to the folks working on Gregex for glib for this insight.
2431    
2432    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2433        unicode-properties was also set.
2434    
2435    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2436    
2437    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2438        checked only for CRLF.
2439    
2440    14. Added casts to pcretest.c to avoid compiler warnings.
2441    
2442    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2443    
2444    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2445        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2446        entirely. This removes changes made in 7 above.
2447    
2448    17. The CMake files have been updated, and there is now more information about
2449        building with CMake in the NON-UNIX-USE document.
2450    
2451    
2452    Version 7.3 28-Aug-07
2453    ---------------------
2454    
2455     1. In the rejigging of the build system that eventually resulted in 7.1, the
2456        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2457        brackets there is not right, since it causes compilers to look for an
2458        installed pcre.h, not the version that is in the source that is being
2459        compiled (which of course may be different). I have changed it back to:
2460    
2461          #include "pcre.h"
2462    
2463        I have a vague recollection that the change was concerned with compiling in
2464        different directories, but in the new build system, that is taken care of
2465        by the VPATH setting the Makefile.
2466    
2467     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2468        when the subject happened to end in the byte 0x85 (e.g. if the last
2469        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2470        characters but of course it shouldn't be taken as a newline when it is part
2471        of another character. The bug was that, for an unlimited repeat of . in
2472        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2473        characters when looking for a newline.
2474    
2475     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2476    
2477     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2478        in debug output.
2479    
2480     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2481        long printing in the pcrecpp unittest when running under MinGW.
2482    
2483     6. ESC_K was left out of the EBCDIC table.
2484    
2485     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2486        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2487        limit also applies to "virtual nesting" when a pattern is recursive, and in
2488        this case 1000 isn't so big. I have been able to remove this limit at the
2489        expense of backing off one optimization in certain circumstances. Normally,
2490        when pcre_exec() would call its internal match() function recursively and
2491        immediately return the result unconditionally, it uses a "tail recursion"
2492        feature to save stack. However, when a subpattern that can match an empty
2493        string has an unlimited repetition quantifier, it no longer makes this
2494        optimization. That gives it a stack frame in which to save the data for
2495        checking that an empty string has been matched. Previously this was taken
2496        from the 1000-entry workspace that had been reserved. So now there is no
2497        explicit limit, but more stack is used.
2498    
2499     8. Applied Daniel's patches to solve problems with the import/export magic
2500        syntax that is required for Windows, and which was going wrong for the
2501        pcreposix and pcrecpp parts of the library. These were overlooked when this
2502        problem was solved for the main library.
2503    
2504     9. There were some crude static tests to avoid integer overflow when computing
2505        the size of patterns that contain repeated groups with explicit upper
2506        limits. As the maximum quantifier is 65535, the maximum group length was
2507        set at 30,000 so that the product of these two numbers did not overflow a
2508        32-bit integer. However, it turns out that people want to use groups that
2509        are longer than 30,000 bytes (though not repeat them that many times).
2510        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2511        made it possible to implement the integer overflow checks in a much more
2512        dynamic way, which I have now done. The artificial limitation on group
2513        length has been removed - we now have only the limit on the total length of
2514        the compiled pattern, which depends on the LINK_SIZE setting.
2515    
2516    10. Fixed a bug in the documentation for get/copy named substring when
2517        duplicate names are permitted. If none of the named substrings are set, the
2518        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2519        empty string.
2520    
2521    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2522        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2523        because the ] is interpreted as the first data character and the
2524        terminating ] is not found. PCRE has been made compatible with Perl in this
2525        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2526        cause memory overwriting.
2527    
2528    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2529        string has been matched (to stop an infinite loop). It was not recognizing
2530        a conditional subpattern that could match an empty string if that
2531        subpattern was within another subpattern. For example, it looped when
2532        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2533        condition was not nested. This bug has been fixed.
2534    
2535    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2536        past the start of the subject in the presence of bytes with the top bit
2537        set, for example "\x8aBCD".
2538    
2539    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2540        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2541    
2542    14. Optimized (?!) to (*FAIL).
2543    
2544    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2545        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2546        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2547        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2548        does: it's just the validity check that is more restrictive.
2549    
2550    16. Inserted checks for integer overflows during escape sequence (backslash)
2551        processing, and also fixed erroneous offset values for syntax errors during
2552        backslash processing.
2553    
2554    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2555        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2556    
2557    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2558        caused an overrun.
2559    
2560    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2561        something other than just ASCII characters) inside a group that had an
2562        unlimited repeat caused a loop at compile time (while checking to see
2563        whether the group could match an empty string).
2564    
2565    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2566        [\P{Any}] did so. (Error in the code for printing property names.)
2567    
2568    21. An orphan \E inside a character class could cause a crash.
2569    
2570    22. A repeated capturing bracket such as (A)? could cause a wild memory
2571        reference during compilation.
2572    
2573    23. There are several functions in pcre_compile() that scan along a compiled
2574        expression for various reasons (e.g. to see if it's fixed length for look
2575        behind). There were bugs in these functions when a repeated \p or \P was
2576        present in the pattern. These operators have additional parameters compared
2577        with \d, etc, and these were not being taken into account when moving along
2578        the compiled data. Specifically:
2579    
2580        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2581            length.
2582    
2583        (b) An item such as \pL+ within a repeated group could cause crashes or
2584            loops.
2585    
2586        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2587            "reference to non-existent subpattern" error.
2588    
2589        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2590    
2591    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2592        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2593    
2594    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2595    
2596    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2597        character were causing crashes (broken optimization).
2598    
2599    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2600        \p or \P) caused a compile-time loop.
2601    
2602    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2603        break. For example, the unstudied pattern [\r\n]A does not match the string
2604        "\r\nA" because change 7.0/46 below moves the current point on by two
2605        characters after failing to match at the start. However, the pattern \nA
2606        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2607        the same is true. There doesn't seem any very clean way out of this, but
2608        what I have chosen to do makes the common cases work: PCRE now takes note
2609        of whether there can be an explicit match for \r or \n anywhere in the
2610        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2611        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2612        pattern has explicit CR or LF references.
2613    
2614    29. Added (*CR) etc for changing newline setting at start of pattern.
2615    
2616    
2617    Version 7.2 19-Jun-07
2618  ---------------------  ---------------------
2619    
2620   1. If the fr_FR locale cannot be found for test 3, try the "french" locale,   1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
# Line 69  Version 7.2 13-June-07 Line 2682  Version 7.2 13-June-07
2682        pcrecpp::RE("a*?").FullMatch("aaa") does not, and        pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2683        pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.        pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2684    
2685    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2686        it matched the wrong number of bytes.
2687    
2688    
2689  Version 7.1 24-Apr-07  Version 7.1 24-Apr-07
2690  ---------------------  ---------------------

Legend:
Removed from v.182  
changed lines
  Added in v.1420

  ViewVC Help
Powered by ViewVC 1.1.5