ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 175 by ph10, Mon Jun 11 13:38:38 2007 UTC revision 1418 by ph10, Fri Dec 27 12:23:25 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4  Version 7.2 05-June-07  Version 8.35-RC1 xx-xxxx-201x
5    -----------------------------
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.
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.
15    3.  Got rid of some compiler warnings for potentially uninitialized variables
16        that show up only when compiled with -O2.
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.
25    Version 8.34 15-December-2013
26    -----------------------------
28    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
29        executable memory. Patch inspired by Carsten Klein.
31    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
32        this macro is never tested and has no effect, because the work to support
33        coverage involves only compiling and linking options and special targets in
34        the Makefile. The comment in config.h implied that defining the macro would
35        enable coverage support, which is totally false. There was also support for
36        setting this macro in the CMake files (my fault, I just copied it from
37        configure). SUPPORT_GCOV has now been removed.
39    3.  Make a small performance improvement in strlen16() and strlen32() in
40        pcretest.
42    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
43        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
44        these cases). There was also one in pcretest.c.
46    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
48    6.  In UTF mode, the code for checking whether a group could match an empty
49        string (which is used for indefinitely repeated groups to allow for
50        breaking an infinite loop) was broken when the group contained a repeated
51        negated single-character class with a character that occupied more than one
52        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
53        UTF-8 mode). The effect was undefined: the group might or might not be
54        deemed as matching an empty string, or the program might have crashed.
56    7.  The code for checking whether a group could match an empty string was not
57        recognizing that \h, \H, \v, \V, and \R must match a character.
59    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
60        an empty string. If it can, pcretest shows this in its information output.
62    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
63        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
64        matched by pcre_exec() without using JIT:
66        (a) If the rest of the pattern did not match after a maximal run of
67            grapheme clusters, the code for backing up to try with fewer of them
68            did not always back up over a full grapheme when characters that do not
69            have the modifier quality were involved, e.g. Hangul syllables.
71        (b) If the match point in a subject started with modifier character, and
72            there was no match, the code could incorrectly back up beyond the match
73            point, and potentially beyond the first character in the subject,
74            leading to a segfault or an incorrect match result.
76    10. A conditional group with an assertion condition could lead to PCRE
77        recording an incorrect first data item for a match if no other first data
78        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
79        first data item, and therefore matched "ca" after "c" instead of at the
80        start.
82    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
83        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
85    12. The source of pcregrep now includes z/OS-specific code so that it can be
86        compiled for z/OS as part of the special z/OS distribution.
88    13. Added the -T and -TM options to pcretest.
90    14. The code in pcre_compile.c for creating the table of named capturing groups
91        has been refactored. Instead of creating the table dynamically during the
92        actual compiling pass, the information is remembered during the pre-compile
93        pass (on the stack unless there are more than 20 named groups, in which
94        case malloc() is used) and the whole table is created before the actual
95        compile happens. This has simplified the code (it is now nearly 150 lines
96        shorter) and prepared the way for better handling of references to groups
97        with duplicate names.
99    15. A back reference to a named subpattern when there is more than one of the
100        same name now checks them in the order in which they appear in the pattern.
101        The first one that is set is used for the reference. Previously only the
102        first one was inspected. This change makes PCRE more compatible with Perl.
104    16. Unicode character properties were updated from Unicode 6.3.0.
106    17. The compile-time code for auto-possessification has been refactored, based
107        on a patch by Zoltan Herczeg. It now happens after instead of during
108        compilation. The code is cleaner, and more cases are handled. The option
109        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
110        options in pcretest are provided to set it. It can also be set by
111        (*NO_AUTO_POSSESS) at the start of a pattern.
113    18. The character VT has been added to the default ("C" locale) set of
114        characters that match \s and are generally treated as white space,
115        following this same change in Perl 5.18. There is now no difference between
116        "Perl space" and "POSIX space". Whether VT is treated as white space in
117        other locales depends on the locale.
119    19. The code for checking named groups as conditions, either for being set or
120        for being recursed, has been refactored (this is related to 14 and 15
121        above). Processing unduplicated named groups should now be as fast at
122        numerical groups, and processing duplicated groups should be faster than
123        before.
125    20. Two patches to the CMake build system, by Alexander Barkov:
127          (1) Replace the "source" command by "." in CMakeLists.txt because
128              "source" is a bash-ism.
130          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
131              without these the CMake build does not work on Solaris.
133    21. Perl has changed its handling of \8 and \9. If there is no previously
134        encountered capturing group of those numbers, they are treated as the
135        literal characters 8 and 9 instead of a binary zero followed by the
136        literals. PCRE now does the same.
138    22. Following Perl, added \o{} to specify codepoints in octal, making it
139        possible to specify values greater than 0777 and also making them
140        unambiguous.
142    23. Perl now gives an error for missing closing braces after \x{... instead of
143        treating the string as literal. PCRE now does the same.
145    24. RunTest used to grumble if an inappropriate test was selected explicitly,
146        but just skip it when running all tests. This make it awkward to run ranges
147        of tests when one of them was inappropriate. Now it just skips any
148        inappropriate tests, as it always did when running all tests.
150    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
151        character types such as \d or \w, too many callouts were inserted, and the
152        data that they returned was rubbish.
154    26. In UCP mode, \s was not matching two of the characters that Perl matches,
155        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
156        were matched by \h. The code has now been refactored so that the lists of
157        the horizontal and vertical whitespace characters used for \h and \v (which
158        are defined only in one place) are now also used for \s.
160    27. Add JIT support for the 64 bit TileGX architecture.
161        Patch by Jiong Wang (Tilera Corporation).
163    28. Possessive quantifiers for classes (both explicit and automatically
164        generated) now use special opcodes instead of wrapping in ONCE brackets.
166    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
167        (because it's meaningless), this was not happening when PCRE_CASELESS was
168        set. Not wrong, but inefficient.
170    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
171        properties for \w, \d, etc) is present in a test regex. Otherwise if the
172        test contains no characters greater than 255, Perl doesn't realise it
173        should be using Unicode semantics.
175    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
176        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
177        does in Unicode mode.
179    32. Added the "forbid" facility to pcretest so that putting tests into the
180        wrong test files can sometimes be quickly detected.
182    33. There is now a limit (default 250) on the depth of nesting of parentheses.
183        This limit is imposed to control the amount of system stack used at compile
184        time. It can be changed at build time by --with-parens-nest-limit=xxx or
185        the equivalent in CMake.
187    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
188        errors. Perl warns for these when in warning mode, but PCRE has no facility
189        for giving warnings.
191    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
192        However, this was not working for (?!) because it is optimized to (*FAIL),
193        for which PCRE does not allow quantifiers. The optimization is now disabled
194        when a quantifier follows (?!). I can't see any use for this, but it makes
195        things uniform.
197    36. Perl no longer allows group names to start with digits, so I have made this
198        change also in PCRE. It simplifies the code a bit.
200    37. In extended mode, Perl ignores spaces before a + that indicates a
201        possessive quantifier. PCRE allowed a space before the quantifier, but not
202        before the possessive +. It now does.
204    38. The use of \K (reset reported match start) within a repeated possessive
205        group such as (a\Kb)*+ was not working.
207    40. Document that the same character tables must be used at compile time and
208        run time, and that the facility to pass tables to pcre_exec() and
209        pcre_dfa_exec() is for use only with saved/restored patterns.
211    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
212        features for Builds with MSVC:
214        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
215           builds). The .rc files can be used to set FileDescription and many other
216           attributes.
218        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
219           This allows higher-level build scripts which want .pdb files to avoid
220           hard-coding the exact files needed."
222    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
223        mean "start of word" and "end of word", respectively, as a transition aid.
225    43. A minimizing repeat of a class containing codepoints greater than 255 in
226        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
227        compiled to use the heap for recursion.
229    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
230        is configured.
233    Version 8.33 28-May-2013
234    ------------------------
236    1.  Added 'U' to some constants that are compared to unsigned integers, to
237        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
238        variables that are added to signed variables, to ensure the result is
239        signed and can be negated.
241    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
242        CMake config files.
244    3.  Revise the creation of config.h.generic so that all boolean macros are
245        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
246        overriding via -D on the command line possible.
248    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
249        to unsigned int is reported to make a quite noticeable speed difference in
250        a specific Windows environment. Testing on Linux did also appear to show
251        some benefit (and it is clearly not harmful). Also fixed the definition of
252        Xop which should be unsigned.
254    5.  Related to (4), changing the definition of the intermediate variable cc
255        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
256        improvements.
258    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
259        unnecessary spaces.
261    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
262        and later.
264    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
266    9.  Optimizing fast_forward_start_bits in JIT.
268    10. Adding support for callouts in JIT, and fixing some issues revealed
269        during this work. Namely:
271        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
273        (b) Minimum length was not checked before the matching is started.
275    11. The value of capture_last that is passed to callouts was incorrect in some
276        cases when there was a capture on one path that was subsequently abandoned
277        after a backtrack. Also, the capture_last value is now reset after a
278        recursion, since all captures are also reset in this case.
280    12. The interpreter no longer returns the "too many substrings" error in the
281        case when an overflowing capture is in a branch that is subsequently
282        abandoned after a backtrack.
284    13. In the pathological case when an offset vector of size 2 is used, pcretest
285        now prints out the matched string after a yield of 0 or 1.
287    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
288        Only supported by the JIT compiler at the moment.
290    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
292    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
293        offset of the starting point of the matching process, provided the offsets
294        vector is large enough.
296    17. The \A escape now records a lookbehind value of 1, though its execution
297        does not actually inspect the previous character. This is to ensure that,
298        in partial multi-segment matching, at least one character from the old
299        segment is retained when a new segment is processed. Otherwise, if there
300        are no lookbehinds in the pattern, \A might match incorrectly at the start
301        of a new segment.
303    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
305    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
306        gives some modest performance improvement in 8-bit mode.
308    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
309        be expressed in certain programming languages using Universal Character
310        Names.
312    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
313        which points out that "non characters" are not "characters that may not
314        appear in Unicode strings" but rather "characters that are reserved for
315        internal use and have only local meaning".
317    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
318        there was a conditional group that depended on an assertion, if the
319        assertion was false, the callout that immediately followed the alternation
320        in the condition was skipped when pcre_exec() was used for matching.
322    23. Allow an explicit callout to be inserted before an assertion that is the
323        condition for a conditional group, for compatibility with automatic
324        callouts, which always insert a callout at this point.
326    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
327        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
329    25. (*PRUNE) is now supported by the JIT compiler.
331    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
333    27. Fix the case where there are two or more SKIPs with arguments that may be
334        ignored.
336    28. (*SKIP) is now supported by the JIT compiler.
338    29. (*THEN) is now supported by the JIT compiler.
340    30. Update RunTest with additional test selector options.
342    31. The way PCRE handles backtracking verbs has been changed in two ways.
344        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
345        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
346        backtracking. In some cases this makes it more Perl-compatible, but Perl's
347        rather obscure rules do not always do the same thing.
349        (2) Previously, backtracking verbs were confined within assertions. This is
350        no longer the case for positive assertions, except for (*ACCEPT). Again,
351        this sometimes improves Perl compatibility, and sometimes does not.
353    32. A number of tests that were in test 2 because Perl did things differently
354        have been moved to test 1, because either Perl or PCRE has changed, and
355        these tests are now compatible.
357    32. Backtracking control verbs are now handled in the same way in JIT and
358        interpreter.
360    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
361        contained a forward subroutine reference caused a compile error.
363    34. Auto-detect and optimize limited repetitions in JIT.
365    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
366        blocking (*UTF) etc.
368    36. In the interpreter, maximizing pattern repetitions for characters and
369        character types now use tail recursion, which reduces stack usage.
371    37. The value of the max lookbehind was not correctly preserved if a compiled
372        and saved regex was reloaded on a host of different endianness.
374    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
375        of the compiled pattern block, expand the flags field from 16 to 32 bits
376        because it was almost full.
378    39. Try madvise first before posix_madvise.
380    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
381        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
382        original change locked out all matches of empty strings. This has been
383        changed so that one match of an empty string per line is recognized.
384        Subsequent searches on the same line (for colouring or for --only-matching,
385        for example) do not recognize empty strings.
387    41. Applied a user patch to fix a number of spelling mistakes in comments.
389    42. Data lines longer than 65536 caused pcretest to crash.
391    43. Clarified the data type for length and startoffset arguments for pcre_exec
392        and pcre_dfa_exec in the function-specific man pages, where they were
393        explicitly stated to be in bytes, never having been updated. I also added
394        some clarification to the pcreapi man page.
396    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
397        a segmentation fault.
400    Version 8.32 30-November-2012
401    -----------------------------
403    1.  Improved JIT compiler optimizations for first character search and single
404        character iterators.
406    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
407        Patch by Daniel Richard G.
409    3.  Single character iterator optimizations in the JIT compiler.
411    4.  Improved JIT compiler optimizations for character ranges.
413    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
414        Reported by Giuseppe D'Angelo.
416    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
417        of a line, was being set incorrectly in cases where .* appeared inside
418        atomic brackets at the start of a pattern, or where there was a subsequent
419        *PRUNE or *SKIP.
421    7.  Improved instruction cache flush for POWER/PowerPC.
422        Patch by Daniel Richard G.
424    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
425        grep:
427        (a) There is now no limit to the number of patterns to be matched.
429        (b) An error is given if a pattern is too long.
431        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
432            are now supported.
434        (d) --exclude-from and --include-from (multiple use) have been added.
436        (e) Exclusions and inclusions now apply to all files and directories, not
437            just to those obtained from scanning a directory recursively.
439        (f) Multiple uses of -f and --file-list are now supported.
441        (g) In a Windows environment, the default for -d has been changed from
442            "read" (the GNU grep default) to "skip", because otherwise the presence
443            of a directory in the file list provokes an error.
445        (h) The documentation has been revised and clarified in places.
447    9.  Improve the matching speed of capturing brackets.
449    10. Changed the meaning of \X so that it now matches a Unicode extended
450        grapheme cluster.
452    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
453        out POSIX threads when JIT support is configured.
455    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
457    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
458        structure could go wrong in environments where size_t is not the same size
459        as int.
461    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
462        set.
464    15. The EBCDIC support had decayed; later updates to the code had included
465        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
466        general tidy up of EBCDIC-related issues, and the documentation was also
467        not quite right. There is now a test that can be run on ASCII systems to
468        check some of the EBCDIC-related things (but is it not a full test).
470    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
471        in a small tidy to the code.
473    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
475    18. If the --only-matching (-o) option in pcregrep is specified multiple
476        times, each one causes appropriate output. For example, -o1 -o2 outputs the
477        substrings matched by the 1st and 2nd capturing parentheses. A separating
478        string can be specified by --om-separator (default empty).
480    19. Improving the first n character searches.
482    20. Turn case lists for horizontal and vertical white space into macros so that
483        they are defined only once.
485    21. This set of changes together give more compatible Unicode case-folding
486        behaviour for characters that have more than one other case when UCP
487        support is available.
489        (a) The Unicode property table now has offsets into a new table of sets of
490            three or more characters that are case-equivalent. The MultiStage2.py
491            script that generates these tables (the pcre_ucd.c file) now scans
492            CaseFolding.txt instead of UnicodeData.txt for character case
493            information.
495        (b) The code for adding characters or ranges of characters to a character
496            class has been abstracted into a generalized function that also handles
497            case-independence. In UTF-mode with UCP support, this uses the new data
498            to handle characters with more than one other case.
500        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
501            whose other case is greater than 256 are now correctly matched
502            caselessly. Previously, the high codepoint matched the low one, but not
503            vice versa.
505        (d) The processing of \h, \H, \v, and \ in character classes now makes use
506            of the new class addition function, using character lists defined as
507            macros alongside the case definitions of 20 above.
509        (e) Caseless back references now work with characters that have more than
510            one other case.
512        (f) General caseless matching of characters with more than one other case
513            is supported.
515    22. Unicode character properties were updated from Unicode 6.2.0
517    23. Improved CMake support under Windows. Patch by Daniel Richard G.
519    24. Add support for 32-bit character strings, and UTF-32
521    25. Major JIT compiler update (code refactoring and bugfixing).
522        Experimental Sparc 32 support is added.
524    26. Applied a modified version of Daniel Richard G's patch to create
525        pcre.h.generic and config.h.generic by "make" instead of in the
526        PrepareRelease script.
528    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
529        pcre_compile.c when checking for a zero character.
531    28. Introducing a native interface for JIT. Through this interface, the compiled
532        machine code can be directly executed. The purpose of this interface is to
533        provide fast pattern matching, so several sanity checks are not performed.
534        However, feature tests are still performed. The new interface provides
535        1.4x speedup compared to the old one.
537    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
538        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
539        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
541    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
542        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
543        the "old" RFC 2279). Instead, it ended up passing a negative length to
544        pcre_exec().
546    31. Add support for GCC's visibility feature to hide internal functions.
548    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
549        "unknown -C option" after outputting 0 or 1.
551    33. There is now support for generating a code coverage report for the test
552        suite in environments where gcc is the compiler and lcov is installed. This
553        is mainly for the benefit of the developers.
555    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
556        unaddressable using valgrind annotations, allowing valgrind to detect
557        invalid memory accesses. This is mainly for the benefit of the developers.
559    25. (*UTF) can now be used to start a pattern in any of the three libraries.
561    26. Give configure error if --enable-cpp but no C++ compiler found.
564    Version 8.31 06-July-2012
565    -------------------------
567    1.  Fixing a wrong JIT test case and some compiler warnings.
569    2.  Removed a bashism from the RunTest script.
571    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
572        to unsigned type, result still unsigned" that was given by an MS compiler
573        on encountering the code "-sizeof(xxx)".
575    4.  Partial matching support is added to the JIT compiler.
577    5.  Fixed several bugs concerned with partial matching of items that consist
578        of more than one character:
580        (a) /^(..)\1/ did not partially match "aba" because checking references was
581            done on an "all or nothing" basis. This also applied to repeated
582            references.
584        (b) \R did not give a hard partial match if \r was found at the end of the
585            subject.
587        (c) \X did not give a hard partial match after matching one or more
588            characters at the end of the subject.
590        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
591            a partial match for the string "\r".
593        (e) When newline was set to CRLF, the metacharacter "." did not recognize
594            a partial match for a CR character at the end of the subject string.
596    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
597        running pcretest, the text "(JIT)" added to the output whenever JIT is
598        actually used to run the match.
600    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
601        or /S+[+] with a digit between 1 and 7.
603    8.  OP_NOT now supports any UTF character not just single-byte ones.
605    9.  (*MARK) control verb is now supported by the JIT compiler.
607    10. The command "./RunTest list" lists the available tests without actually
608        running any of them. (Because I keep forgetting what they all are.)
612    12. Applied a (slightly modified) user-supplied patch that improves performance
613        when the heap is used for recursion (compiled with --disable-stack-for-
614        recursion). Instead of malloc and free for each heap frame each time a
615        logical recursion happens, frames are retained on a chain and re-used where
616        possible. This sometimes gives as much as 30% improvement.
618    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
619        call.
621    14. As documented, (*COMMIT) is now confined to within a positive assertion.
623    15. It is now possible to link pcretest with libedit as an alternative to
624        libreadline.
626    16. (*COMMIT) control verb is now supported by the JIT compiler.
628    17. The Unicode data tables have been updated to Unicode 6.1.0.
630    18. Added --file-list option to pcregrep.
632    19. Added binary file support to pcregrep, including the -a, --binary-files,
633        -I, and --text options.
635    20. The madvise function is renamed for posix_madvise for QNX compatibility
636        reasons. Fixed by Giuseppe D'Angelo.
638    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
640    22. Changed the option for creating symbolic links for 16-bit man pages from
641        -s to -sf so that re-installing does not cause issues.
643    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
645    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
646        matches in certain environments (the workspace was not being correctly
647        retained). Also added to pcre_dfa_exec() a simple plausibility check on
648        some of the workspace data at the beginning of a restart.
650    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
651        was not doing so when it should - probably a typo introduced by SVN 528
652        (change 8.10/14).
654    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
655        \w+ when the character tables indicated that \x{c4} was a word character.
656        There were several related cases, all because the tests for doing a table
657        lookup were testing for characters less than 127 instead of 255.
659    27. If a pattern contains capturing parentheses that are not used in a match,
660        their slots in the ovector are set to -1. For those that are higher than
661        any matched groups, this happens at the end of processing. In the case when
662        there were back references that the ovector was too small to contain
663        (causing temporary malloc'd memory to be used during matching), and the
664        highest capturing number was not used, memory off the end of the ovector
665        was incorrectly being set to -1. (It was using the size of the temporary
666        memory instead of the true size.)
668    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
669        ovector size, it uses memory at the end of the block that it has got.
671    29. Check for an overlong MARK name and give an error at compile time. The
672        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
674    30. JIT compiler update.
676    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
677        Rill for the patch.
679    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
681    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
683    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
684        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
685        was enabled.
687    35. Improve JIT code generation for greedy plus quantifier.
689    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
690        1 to "aa" instead of to an empty string. The bug affected repeated groups
691        that could potentially match an empty string.
693    37. Optimizing single character iterators in JIT.
695    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
696        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
697        codepoints that are too big for the mode are faulted, and in a UTF mode,
698        disallowed codepoints are also faulted.
700    39. If PCRE was compiled with UTF support, in three places in the DFA
701        matcher there was code that should only have been obeyed in UTF mode, but
702        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
703        processing when bytes with values greater than 127 were present. In 16-bit
704        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
705        both cases the values are those that cannot be the first data item in a UTF
706        character. The three items that might have provoked this were recursions,
707        possessively repeated groups, and atomic groups.
709    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
710        and pcregrep, because some OS require shared objects to be explicitly
711        passed to ld, causing the link step to fail if they are not.
713    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
714        mode, patterns that started with \h* or \R* might be incorrectly matched.
717    Version 8.30 04-February-2012
718    -----------------------------
720    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
721        name is defined in ctype.h.
723    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
724        only in quite long subpatterns.
726    3.  Removed the function pcre_info(), which has been obsolete and deprecated
727        since it was replaced by pcre_fullinfo() in February 2000.
729    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
730        match a (*MARK), and the match failed at the start of the subject, a
731        reference to memory before the start of the subject could occur. This bug
732        was introduced by fix 17 of release 8.21.
734    5.  A reference to an unset group with zero minimum repetition was giving
735        totally wrong answers (in non-JavaScript-compatibility mode). For example,
736        /(another)?(\1?)test/ matched against "hello world test". This bug was
737        introduced in release 8.13.
739    6.  Add support for 16-bit character strings (a large amount of work involving
740        many changes and refactorings).
742    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
743        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
744        from a file.
746    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
747        rounding is not applied in this particular case).
749    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
750        if they appear, or are escaped, in patterns.
752    10. Get rid of a number of -Wunused-but-set-variable warnings.
754    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
755        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
756        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
757        also returns the mark "x". This bug applied to capturing parentheses,
758        non-capturing parentheses, and atomic parentheses. It also applied to some
759        assertions.
761    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
762        information out of configure.ac instead of relying on pcre.h.generic, which
763        is not stored in the repository.
765    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
766        -lreadline.
768    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
770    15. Applied Graycode's patch to put the top-level frame on the stack rather
771        than the heap when not using the stack for recursion. This gives a
772        performance improvement in many cases when recursion is not deep.
774    16. Experimental code added to "pcretest -C" to output the stack frame size.
777    Version 8.21 12-Dec-2011
778    ------------------------
780    1.  Updating the JIT compiler.
782    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
783        are added as well.
785    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
786        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
787        calling _pcre_jit_exec. Some extra comments are added.
789    4.  (*MARK) settings inside atomic groups that do not contain any capturing
790        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
791        was introduced by change 18 for 8.20.
793    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
794        ECMA-262 standard.
796    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
797        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
798        This bug was probably introduced by change 9 of 8.13.
800    7.  While fixing 6 above, I noticed that a number of other items were being
801        incorrectly rejected as "not fixed length". This arose partly because newer
802        opcodes had not been added to the fixed-length checking code. I have (a)
803        corrected the bug and added tests for these items, and (b) arranged for an
804        error to occur if an unknown opcode is encountered while checking for fixed
805        length instead of just assuming "not fixed length". The items that were
806        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
807        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
808        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
810    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
811        being incorrectly compiled and would have given unpredicatble results.
813    9.  A possessively repeated subpattern with minimum repeat count greater than
814        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
815        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
816        the first (A) could occur when it should not.
818    10. Add a cast and remove a redundant test from the code.
820    11. JIT should use pcre_malloc/pcre_free for allocation.
822    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
823        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
824        is anything other than /usr, -L is still shown).
826    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
828    14. Perl does not support \N without a following name in a [] class; PCRE now
829        also gives an error.
831    15. If a forward reference was repeated with an upper limit of around 2000,
832        it caused the error "internal error: overran compiling workspace". The
833        maximum number of forward references (including repeats) was limited by the
834        internal workspace, and dependent on the LINK_SIZE. The code has been
835        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
836        the default depends on LINK_SIZE. There is a new upper limit (for safety)
837        of around 200,000 forward references. While doing this, I also speeded up
838        the filling in of repeated forward references.
840    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
841        incorrectly expecting the subject to contain another "a" after the start.
843    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
844        in the match, the SKIP should be ignored. This was not happening; instead
845        the SKIP was being treated as NOMATCH. For patterns such as
846        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
847        tested.
849    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
850        now much more compatible with Perl, in particular in cases where the result
851        is a non-match for a non-anchored pattern. For example, if
852        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
853        "m", where previously it did not return a name. A side effect of this
854        change is that for partial matches, the last encountered mark name is
855        returned, as for non matches. A number of tests that were previously not
856        Perl-compatible have been moved into the Perl-compatible test files. The
857        refactoring has had the pleasing side effect of removing one argument from
858        the match() function, thus reducing its stack requirements.
860    19. If the /S+ option was used in pcretest to study a pattern using JIT,
861        subsequent uses of /S (without +) incorrectly behaved like /S+.
863    21. Retrieve executable code size support for the JIT compiler and fixing
864        some warnings.
866    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
867        not work when the shorter character appeared right at the end of the
868        subject string.
870    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
871        systems.
873    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
874        output it when the /M option is used in pcretest.
876    25. The CheckMan script was not being included in the distribution. Also, added
877        an explicit "perl" to run Perl scripts from the PrepareRelease script
878        because this is reportedly needed in Windows.
880    26. If study data was being save in a file and studying had not found a set of
881        "starts with" bytes for the pattern, the data written to the file (though
882        never used) was taken from uninitialized memory and so caused valgrind to
883        complain.
885    27. Updated RunTest.bat as provided by Sheri Pierce.
887    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
889    29. Computation of memory usage for the table of capturing group names was
890        giving an unnecessarily large value.
893    Version 8.20 21-Oct-2011
894    ------------------------
896    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
897        a POSIX class. After further experiments with Perl, which convinced me that
898        Perl has bugs and confusions, a closing square bracket is no longer allowed
899        in a POSIX name. This bug also affected patterns with classes that started
900        with full stops.
902    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
903        captured substring, but while checking the failing first alternative,
904        substring 1 is temporarily captured. If the output vector supplied to
905        pcre_exec() was not big enough for this capture, the yield of the function
906        was still zero ("insufficient space for captured substrings"). This cannot
907        be totally fixed without adding another stack variable, which seems a lot
908        of expense for a edge case. However, I have improved the situation in cases
909        such as /(a)(b)x|abc/ matched against "abc", where the return code
910        indicates that fewer than the maximum number of slots in the ovector have
911        been set.
913    3.  Related to (2) above: when there are more back references in a pattern than
914        slots in the output vector, pcre_exec() uses temporary memory during
915        matching, and copies in the captures as far as possible afterwards. It was
916        using the entire output vector, but this conflicts with the specification
917        that only 2/3 is used for passing back captured substrings. Now it uses
918        only the first 2/3, for compatibility. This is, of course, another edge
919        case.
921    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
922        main code base, and can be used by building with --enable-jit. When this is
923        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
924        runtime --no-jit option is given.
926    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
927        ovector, the return from the function was zero, implying that there were
928        other matches that did not fit. The correct "exactly full" value is now
929        returned.
931    6.  If a subpattern that was called recursively or as a subroutine contained
932        (*PRUNE) or any other control that caused it to give a non-standard return,
933        invalid errors such as "Error -26 (nested recursion at the same subject
934        position)" or even infinite loops could occur.
936    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
937        computing the minimum length on reaching *ACCEPT, and so ended up with the
938        wrong value of 1 rather than 0. Further investigation indicates that
939        computing a minimum subject length in the presence of *ACCEPT is difficult
940        (think back references, subroutine calls), and so I have changed the code
941        so that no minimum is registered for a pattern that contains *ACCEPT.
943    8.  If (*THEN) was present in the first (true) branch of a conditional group,
944        it was not handled as intended. [But see 16 below.]
946    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
947        Sheri Pierce.
949    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
950        the first byte in a match must be "a".
952    11. Change 17 for 8.13 increased the recursion depth for patterns like
953        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
954        pattern contains any instances of (*THEN). If it does not, the old
955        optimizations are restored. It would be nice to do this on a per-group
956        basis, but at the moment that is not feasible.
958    12. In some environments, the output of pcretest -C is CRLF terminated. This
959        broke RunTest's code that checks for the link size. A single white space
960        character after the value is now allowed for.
962    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
963        For "fr", it uses the Windows-specific input and output files.
965    14. If (*THEN) appeared in a group that was called recursively or as a
966        subroutine, it did not work as intended. [But see next item.]
968    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
969        pattern fragments (but not containing any | characters). If A and B are
970        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
971        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
972        D. In other words, Perl considers parentheses that do not contain any |
973        characters to be part of a surrounding alternative, whereas PCRE was
974        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
975        differently. PCRE now behaves in the same way as Perl, except in the case
976        of subroutine/recursion calls such as (?1) which have in any case always
977        been different (but PCRE had them first :-).
979    16. Related to 15 above: Perl does not treat the | in a conditional group as
980        creating alternatives. Such a group is treated in the same way as an
981        ordinary group without any | characters when processing (*THEN). PCRE has
982        been changed to match Perl's behaviour.
984    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
985        RunGrepTest script failed.
987    18. Change 22 for version 13 caused atomic groups to use more stack. This is
988        inevitable for groups that contain captures, but it can lead to a lot of
989        stack use in large patterns. The old behaviour has been restored for atomic
990        groups that do not contain any capturing parentheses.
992    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
993        suppress the check for a minimum subject length at run time. (If it was
994        given to pcre_exec() or pcre_dfa_exec() it did work.)
996    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
997        fail to work when decoding hex characters in data strings in EBCDIC
998        environments.
1000    21. It appears that in at least one Mac OS environment, the isxdigit() function
1001        is implemented as a macro that evaluates to its argument more than once,
1002        contravening the C 90 Standard (I haven't checked a later standard). There
1003        was an instance in pcretest which caused it to go wrong when processing
1004        \x{...} escapes in subject strings. The has been rewritten to avoid using
1005        things like p++ in the argument of isxdigit().
1008    Version 8.13 16-Aug-2011
1009    ------------------------
1011    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1013    2.  Two minor typos in pcre_internal.h have been fixed.
1015    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1016        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1017        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1019    4.  There were a number of related bugs in the code for matching backrefences
1020        caselessly in UTF-8 mode when codes for the characters concerned were
1021        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1022        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1023        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1024        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1025        2-byte code at the end of the subject (it thought there wasn't enough data
1026        left).
1028    5.  Comprehensive information about what went wrong is now returned by
1029        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1030        as the output vector has at least 2 elements. The offset of the start of
1031        the failing character and a reason code are placed in the vector.
1033    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1034        now returned is for the first byte of the failing character, instead of the
1035        last byte inspected. This is an incompatible change, but I hope it is small
1036        enough not to be a problem. It makes the returned offset consistent with
1037        pcre_exec() and pcre_dfa_exec().
1039    7.  pcretest now gives a text phrase as well as the error number when
1040        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1041        failure, the offset and reason code are output.
1043    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1044        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1045        back over a single character (\n). This seems wrong (because it treated the
1046        two characters as a single entity when going forwards), conflicts with the
1047        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1048        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1049        has been changed.
1051    9.  Some internal refactoring has changed the processing so that the handling
1052        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1053        time (the PCRE_DOTALL option was changed this way some time ago: version
1054        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1055        which was always a bit of a fudge. It also means that there is one less
1056        argument for the match() function, which reduces its stack requirements
1057        slightly. This change also fixes an incompatibility with Perl: the pattern
1058        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1060    10. More internal refactoring has drastically reduced the number of recursive
1061        calls to match() for possessively repeated groups such as (abc)++ when
1062        using pcre_exec().
1064    11. While implementing 10, a number of bugs in the handling of groups were
1065        discovered and fixed:
1067        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1068        (a|)*(?1) gave a compile-time internal error.
1069        ((a|)+)+  did not notice that the outer group could match an empty string.
1070        (^a|^)+   was not marked as anchored.
1071        (.*a|.*)+ was not marked as matching at start or after a newline.
1073    12. Yet more internal refactoring has removed another argument from the match()
1074        function. Special calls to this function are now indicated by setting a
1075        value in a variable in the "match data" data block.
1077    13. Be more explicit in pcre_study() instead of relying on "default" for
1078        opcodes that mean there is no starting character; this means that when new
1079        ones are added and accidentally left out of pcre_study(), testing should
1080        pick them up.
1082    14. The -s option of pcretest has been documented for ages as being an old
1083        synonym of -m (show memory usage). I have changed it to mean "force study
1084        for every regex", that is, assume /S for every regex. This is similar to -i
1085        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1086        using it. It makes it easier to run collections of tests with and without
1087        study enabled, and thereby test pcre_study() more easily. All the standard
1088        tests are now run with and without -s (but some patterns can be marked as
1089        "never study" - see 20 below).
1091    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1092        restoration of the capturing data to the outer values was not happening
1093        correctly.
1095    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1096        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1097        pattern had matched an empty string, and so incorrectly returned a no
1098        match.
1100    17. There was optimizing code for the last branch of non-capturing parentheses,
1101        and also for the obeyed branch of a conditional subexpression, which used
1102        tail recursion to cut down on stack usage. Unfortunately, now that there is
1103        the possibility of (*THEN) occurring in these branches, tail recursion is
1104        no longer possible because the return has to be checked for (*THEN). These
1105        two optimizations have therefore been removed. [But see 8.20/11 above.]
1107    18. If a pattern containing \R was studied, it was assumed that \R always
1108        matched two bytes, thus causing the minimum subject length to be
1109        incorrectly computed because \R can also match just one byte.
1111    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1112        was incorrectly computed.
1114    20. If /S is present twice on a test pattern in pcretest input, it now
1115        *disables* studying, thereby overriding the use of -s on the command line
1116        (see 14 above). This is necessary for one or two tests to keep the output
1117        identical in both cases.
1119    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1120        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1122    22. When an atomic group that contained a capturing parenthesis was
1123        successfully matched, but the branch in which it appeared failed, the
1124        capturing was not being forgotten if a higher numbered group was later
1125        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1126        group 1 to "a", when in fact it should be unset. This applied to multi-
1127        branched capturing and non-capturing groups, repeated or not, and also to
1128        positive assertions (capturing in negative assertions does not happen
1129        in PCRE) and also to nested atomic groups.
1131    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1132        subject after a captured substring, to make it easier to tell which of a
1133        number of identical substrings has been captured.
1135    24. The way atomic groups are processed by pcre_exec() has been changed so that
1136        if they are repeated, backtracking one repetition now resets captured
1137        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1138        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1139        "aaa". Previously, it would have been "a". As part of this code
1140        refactoring, the way recursive calls are handled has also been changed.
1142    25. If an assertion condition captured any substrings, they were not passed
1143        back unless some other capturing happened later. For example, if
1144        (?(?=(a))a) was matched against "a", no capturing was returned.
1146    26. When studying a pattern that contained subroutine calls or assertions,
1147        the code for finding the minimum length of a possible match was handling
1148        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1149        group 1 called group 2 while simultaneously a separate group 2 called group
1150        1). A stack overflow occurred in this case. I have fixed this by limiting
1151        the recursion depth to 10.
1153    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1154        Fortmann. This supports explicit test numbers on the command line, and has
1155        argument validation and error reporting.
1157    28. An instance of \X with an unlimited repeat could fail if at any point the
1158        first character it looked at was a mark character.
1160    29. Some minor code refactoring concerning Unicode properties and scripts
1161        should reduce the stack requirement of match() slightly.
1163    30. Added the '=' option to pcretest to check the setting of unused capturing
1164        slots at the end of the pattern, which are documented as being -1, but are
1165        not included in the return count.
1167    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1168        compiled something random. Now it gives a compile-time error (as does
1169        Perl).
1171    32. A *MARK encountered during the processing of a positive assertion is now
1172        recorded and passed back (compatible with Perl).
1174    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1175        had alternative anchored branches, the search for a second match in a line
1176        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1177        matched the line "0102" twice. The same bug affected patterns that started
1178        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1179        twice.
1181    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1182        does, and because of capturing effects, quantifying parenthesized
1183        assertions may at times be useful. Quantifiers are now allowed for
1184        parenthesized assertions.
1186    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1188    36. \g was being checked for fancy things in a character class, when it should
1189        just be a literal "g".
1191    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1192        appearance of a nested POSIX class supersedes an apparent external class.
1193        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1194        unescaped square brackets may also appear as part of class names. For
1195        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1196        more like Perl. (But see 8.20/1 above.)
1198    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1199        was because it thought it was \N{name}, which is not supported).
1201    39. Add minix to OS list not supporting the -S option in pcretest.
1203    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1204        cannot analyze patterns in sufficient detail to catch mutual recursions
1205        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1206        subgroup is called recursively as a subpattern for a second time at the
1207        same position in the subject string. In previous releases this might have
1208        been caught by the recursion limit, or it might have run out of stack.
1210    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1211        happen only once. PCRE was, however incorrectly giving a compile time error
1212        "recursive call could loop indefinitely" because it cannot analyze the
1213        pattern in sufficient detail. The compile time test no longer happens when
1214        PCRE is compiling a conditional subpattern, but actual runaway loops are
1215        now caught at runtime (see 40 above).
1217    42. It seems that Perl allows any characters other than a closing parenthesis
1218        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1219        has been changed to be the same.
1221    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1222        as not to get warnings when autogen.sh is called. Also changed
1223        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1225    44. To help people who use pcregrep to scan files containing exceedingly long
1226        lines, the following changes have been made:
1228        (a) The default value of the buffer size parameter has been increased from
1229            8K to 20K. (The actual buffer used is three times this size.)
1231        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1232            PCRE is built.
1234        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1235            to be set at run time.
1237        (d) Numerical values in pcregrep options can be followed by K or M, for
1238            example --buffer-size=50K.
1240        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1241            given and the return code is set to 2.
1243    45. Add a pointer to the latest mark to the callout data block.
1245    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1246        partial match of an empty string instead of no match. This was specific to
1247        the use of ".".
1249    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1250        complete match instead of a partial match. This bug was dependent on both
1251        the PCRE_UTF8 and PCRE_DOTALL options being set.
1253    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1254        starting byte set, because \b was not being ignored.
1257    Version 8.12 15-Jan-2011
1258    ------------------------
1260    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1261        checks for such things as part of the documentation building process.
1263    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1264        --match-limit and --recursion-limit options (added for 8.11). In
1265        particular, this made one of the standard tests fail. (The integer value
1266        went into the wrong half of a long int.)
1268    3.  If the --colour option was given to pcregrep with -v (invert match), it
1269        did strange things, either producing crazy output, or crashing. It should,
1270        of course, ignore a request for colour when reporting lines that do not
1271        match.
1273    4.  Another pcregrep bug caused similar problems if --colour was specified with
1274        -M (multiline) and the pattern match finished with a line ending.
1276    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1277        matched in multiline mode, the following line was shown as part of the
1278        match. This seems wrong, so I have changed it.
1280    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1281        the check for further matches in the same line (so they could be coloured)
1282        to overrun the end of the current line. If another match was found, it was
1283        incorrectly shown (and then shown again when found in the next line).
1285    7.  If pcregrep was compiled under Windows, there was a reference to the
1286        function pcregrep_exit() before it was defined. I am assuming this was
1287        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1288        reported by a user. I've moved the definition above the reference.
1291    Version 8.11 10-Dec-2010
1292    ------------------------
1294    1.  (*THEN) was not working properly if there were untried alternatives prior
1295        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1296        backtracked to try for "b" instead of moving to the next alternative branch
1297        at the same level (in this case, to look for "c"). The Perl documentation
1298        is clear that when (*THEN) is backtracked onto, it goes to the "next
1299        alternative in the innermost enclosing group".
1301    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1302        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1303        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1304        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1305        (*THEN).
1307    3.  If \s appeared in a character class, it removed the VT character from
1308        the class, even if it had been included by some previous item, for example
1309        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1310        of \s, but is part of the POSIX "space" class.)
1312    4.  A partial match never returns an empty string (because you can always
1313        match an empty string at the end of the subject); however the checking for
1314        an empty string was starting at the "start of match" point. This has been
1315        changed to the "earliest inspected character" point, because the returned
1316        data for a partial match starts at this character. This means that, for
1317        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1318        (previously it gave "no match").
1320    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1321        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1322        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1323        has an implication that the given string is incomplete (because a partial
1324        match is preferred over a full match). For this reason, these items now
1325        give a partial match in this situation. [Aside: previously, the one case
1326        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1327        match rather than a full match, which was wrong by the old rules, but is
1328        now correct.]
1330    6.  There was a bug in the handling of #-introduced comments, recognized when
1331        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1332        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1333        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1334        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1335        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1336        places in pcre_compile().
1338    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1339        comments when looking ahead for named forward references to subpatterns,
1340        the only newline sequence it recognized was NL. It now handles newlines
1341        according to the set newline convention.
1343    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1344        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1345        cater for a lack of strerror(). These oversights have been fixed.
1347    9.  Added --match-limit and --recursion-limit to pcregrep.
1349    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1351    11. When the -o option was used, pcregrep was setting a return code of 1, even
1352        when matches were found, and --line-buffered was not being honoured.
1354    12. Added an optional parentheses number to the -o and --only-matching options
1355        of pcregrep.
1357    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1358        can match an empty string. The code to do it in pcretest and pcredemo
1359        needed fixing:
1361        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1362            only one byte after an empty string match just before CRLF (this case
1363            just got forgotten; "any" and "anycrlf" were OK).
1365        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1366            mode when an empty string match preceded an ASCII character followed by
1367            a non-ASCII character. (The code for advancing by one character rather
1368            than one byte was nonsense.)
1370        (c) The pcredemo.c sample program did not have any code at all to handle
1371            the cases when CRLF is a valid newline sequence.
1373    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1374        as a starting offset was within the subject string. There is now a new
1375        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1376        negative or greater than the length of the string. In order to test this,
1377        pcretest is extended to allow the setting of negative starting offsets.
1379    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1380        starting offset points to the beginning of a UTF-8 character was
1381        unnecessarily clumsy. I tidied it up.
1383    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1384        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1386    17. Nobody had reported that the --include_dir option, which was added in
1387        release 7.7 should have been called --include-dir (hyphen, not underscore)
1388        for compatibility with GNU grep. I have changed it to --include-dir, but
1389        left --include_dir as an undocumented synonym, and the same for
1390        --exclude-dir, though that is not available in GNU grep, at least as of
1391        release 2.5.4.
1393    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1394        characters from a string of bytes) have been redefined so as not to use
1395        loops, in order to improve performance in some environments. At the same
1396        time, I abstracted some of the common code into auxiliary macros to save
1397        repetition (this should not affect the compiled code).
1399    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1400        compile-time error is now given if \c is not followed by an ASCII
1401        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1402        different, and any byte value is allowed.)
1404    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1405        START_OPTIMIZE option, which is now allowed at compile time - but just
1406        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1407        to pcregrep and other applications that have no direct access to PCRE
1408        options. The new /Y option in pcretest sets this option when calling
1409        pcre_compile().
1411    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1412        back references. Groups containing recursive back references were forced to
1413        be atomic by that change, but in the case of named groups, the amount of
1414        memory required was incorrectly computed, leading to "Failed: internal
1415        error: code overflow". This has been fixed.
1417    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1418        pcretest.c, to avoid build problems in some Borland environments.
1421    Version 8.10 25-Jun-2010
1422    ------------------------
1424    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1425        THEN.
1427    2.  (*ACCEPT) was not working when inside an atomic group.
1429    3.  Inside a character class, \B is treated as a literal by default, but
1430        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1431        causes the error). The code is unchanged, but I tidied the documentation.
1433    4.  Inside a character class, PCRE always treated \R and \X as literals,
1434        whereas Perl faults them if its -w option is set. I have changed PCRE so
1435        that it faults them when PCRE_EXTRA is set.
1437    5.  Added support for \N, which always matches any character other than
1438        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1440    6.  When compiling pcregrep with newer versions of gcc which may have
1441        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1442        declared with attribute warn_unused_result" were given. Just casting the
1443        result to (void) does not stop the warnings; a more elaborate fudge is
1444        needed. I've used a macro to implement this.
1446    7.  Minor change to pcretest.c to avoid a compiler warning.
1448    8.  Added four artifical Unicode properties to help with an option to make
1449        \s etc use properties (see next item). The new properties are: Xan
1450        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1452    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1453        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1454        this option. Modified pcretest to add /W to test this facility. Added
1455        REG_UCP to make it available via the POSIX interface.
1457    10. Added --line-buffered to pcregrep.
1459    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1460        studied, and the match started with a letter with a code point greater than
1461        127 whose first byte was different to the first byte of the other case of
1462        the letter, the other case of this starting letter was not recognized
1463        (#976).
1465    12. If a pattern that was studied started with a repeated Unicode property
1466        test, for example, \p{Nd}+, there was the theoretical possibility of
1467        setting up an incorrect bitmap of starting bytes, but fortunately it could
1468        not have actually happened in practice until change 8 above was made (it
1469        added property types that matched character-matching opcodes).
1471    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1472        possible starting bytes for non-anchored patterns.
1474    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1475        \R, and also a number of cases that involve Unicode properties, both
1476        explicit and implicit when PCRE_UCP is set.
1478    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1479        input, it could crash or give wrong results if characters with values
1480        greater than 0xc0 were present in the subject string. (Detail: it assumed
1481        UTF-8 input when processing these items.)
1483    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1484        size_t is 64-bit (#991).
1486    17. Added a check for running out of memory when PCRE is compiled with
1487        --disable-stack-for-recursion (#990).
1489    18. If the last data line in a file for pcretest does not have a newline on
1490        the end, a newline was missing in the output.
1492    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1493        less than 128) in its various bitmaps. However, there is a facility for
1494        generating tables according to the current locale when PCRE is compiled. It
1495        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1496        characters, are recognized by isspace() and therefore were getting set in
1497        these tables, and indeed these tables seem to approximate to ISO 8859. This
1498        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1499        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1500        which of course cannot start UTF-8 characters. I have changed the code so
1501        that only real ASCII characters (less than 128) and the correct starting
1502        bytes for UTF-8 encodings are set for characters greater than 127 when in
1503        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1504        altogether.)
1506    20. Added the /T option to pcretest so as to be able to run tests with non-
1507        standard character tables, thus making it possible to include the tests
1508        used for 19 above in the standard set of tests.
1510    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1511        reference to a subpattern the other side of a comment that contains an
1512        opening parenthesis caused either an internal compiling error, or a
1513        reference to the wrong subpattern.
1516    Version 8.02 19-Mar-2010
1517    ------------------------
1519    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1521    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1522        configured.
1524    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1525        original author of that file, following a query about its status.
1527    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1528        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1530    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1531        quantifier applied to a forward-referencing subroutine call, could compile
1532        incorrect code or give the error "internal error: previously-checked
1533        referenced subpattern not found".
1535    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1536        variables to point to external functions. For these systems, therefore,
1537        pcre_malloc etc. are now initialized to local functions that call the
1538        relevant global functions.
1540    7.  There were two entries missing in the vectors called coptable and poptable
1541        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1542        I've fixed the data, and added a kludgy way of testing at compile time that
1543        the lengths are correct (equal to the number of opcodes).
1545    8.  Following on from 7, I added a similar kludge to check the length of the
1546        eint vector in pcreposix.c.
1548    9.  Error texts for pcre_compile() are held as one long string to avoid too
1549        much relocation at load time. To find a text, the string is searched,
1550        counting zeros. There was no check for running off the end of the string,
1551        which could happen if a new error number was added without updating the
1552        string.
1554    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1556    11. \K was not working if it appeared in an atomic group or in a group that
1557        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1558        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1559        the assertion is positive, but not if it is negative.
1561    12. Change 11 fortuitously reduced the size of the stack frame used in the
1562        "match()" function of pcre_exec.c by one pointer. Forthcoming
1563        implementation of support for (*MARK) will need an extra pointer on the
1564        stack; I have reserved it now, so that the stack frame size does not
1565        decrease.
1567    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1568        item in branch that calls a recursion is a subroutine call - as in the
1569        second branch in the above example - was incorrectly given the compile-
1570        time error "recursive call could loop indefinitely" because pcre_compile()
1571        was not correctly checking the subroutine for matching a non-empty string.
1573    14. The checks for overrunning compiling workspace could trigger after an
1574        overrun had occurred. This is a "should never occur" error, but it can be
1575        triggered by pathological patterns such as hundreds of nested parentheses.
1576        The checks now trigger 100 bytes before the end of the workspace.
1578    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1581    Version 8.01 19-Jan-2010
1582    ------------------------
1584    1.  If a pattern contained a conditional subpattern with only one branch (in
1585        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1586        computed the wrong minimum data length (which is of course zero for such
1587        subpatterns). This could cause incorrect "no match" results.
1589    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1590        the pattern is reset in the first branch, pcre_compile() failed with
1591        "internal error: code overflow at offset...". This happened only when
1592        the reset was to the original external option setting. (An optimization
1593        abstracts leading options settings into an external setting, which was the
1594        cause of this.)
1596    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1597        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1598        assertion pattern did not match (meaning that the assertion was true), it
1599        was incorrectly treated as false if the SKIP had been reached during the
1600        matching. This also applied to assertions used as conditions.
1602    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1603        assertion subpattern, including such a pattern used as a condition,
1604        unpredictable results occurred, instead of the error return
1607    5.  The C++ GlobalReplace function was not working like Perl for the special
1608        situation when an empty string is matched. It now does the fancy magic
1609        stuff that is necessary.
1611    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1612        removed. (These were left over from very, very early versions of PCRE.)
1614    7.  Some cosmetic changes to the code to make life easier when compiling it
1615        as part of something else:
1617        (a) Change DEBUG to PCRE_DEBUG.
1619        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1620            called "current" as "current_branch", to prevent a collision with the
1621            Linux macro when compiled as a kernel module.
1623        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1624            prevent a collision with the Linux macro when compiled as a kernel
1625            module.
1627    8.  In pcre_compile() there are some checks for integer overflows that used to
1628        cast potentially large values to (double). This has been changed to that
1629        when building, a check for int64_t is made, and if it is found, it is used
1630        instead, thus avoiding the use of floating point arithmetic. (There is no
1631        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1632        double.
1634    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1635        2005 (difference between two addresses compared to an unsigned value).
1637    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1638        custom one, because of the following reported problem in Windows:
1640          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1641              under Win32.
1642          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1643              therefore missing the function definition.
1644          - The compiler thus generates a "C" signature for the test function.
1645          - The linker fails to find the "C" function.
1646          - PCRE fails to configure if asked to do so against libbz2.
1648    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1649        messages were output:
1651          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1652          rerunning libtoolize, to keep the correct libtool macros in-tree.
1653          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1655        I have done both of these things.
1657    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1658        most of the time, it *can* run out if it is given a pattern that contains a
1659        runaway infinite recursion. I updated the discussion in the pcrestack man
1660        page.
1662    13. Now that we have gone to the x.xx style of version numbers, the minor
1663        version may start with zero. Using 08 or 09 is a bad idea because users
1664        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1665        interpreted as invalid octal numbers. I've updated the previous comment in
1666        configure.ac, and also added a check that gives an error if 08 or 09 are
1667        used.
1669    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1670        causing partial matching to fail when the end of the subject matched \W
1671        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1673    15. There were some discrepancies between the declarations in pcre_internal.h
1674        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1675        their definitions. The declarations used "const uschar *" and the
1676        definitions used USPTR. Even though USPTR is normally defined as "const
1677        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1678        reported that: "This difference in casting confuses some C++ compilers, for
1679        example, SunCC recognizes above declarations as different functions and
1680        generates broken code for hbpcre." I have changed the declarations to use
1681        USPTR.
1683    16. GNU libtool is named differently on some systems. The autogen.sh script now
1684        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1685        (FreeBSD).
1687    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1688        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1689        comment: "Figure out how to create a longlong from a string: strtoll and
1690        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1691        instance, but it only takes 2 args instead of 3!"
1693    18. A subtle bug concerned with back references has been fixed by a change of
1694        specification, with a corresponding code fix. A pattern such as
1695        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1696        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1697        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1698        same bug. Such groups have to be quantified to be useful, or contained
1699        inside another quantified group. (If there's no repetition, the reference
1700        can never match.) The problem arises because, having left the group and
1701        moved on to the rest of the pattern, a later failure that backtracks into
1702        the group uses the captured value from the final iteration of the group
1703        rather than the correct earlier one. I have fixed this in PCRE by forcing
1704        any group that contains a reference to itself to be an atomic group; that
1705        is, there cannot be any backtracking into it once it has completed. This is
1706        similar to recursive and subroutine calls.
1709    Version 8.00 19-Oct-09
1710    ----------------------
1712    1.  The table for translating pcre_compile() error codes into POSIX error codes
1713        was out-of-date, and there was no check on the pcre_compile() error code
1714        being within the table. This could lead to an OK return being given in
1715        error.
1717    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1718        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1719        in a Windows environment.
1721    3.  The pcregrep --count option prints the count for each file even when it is
1722        zero, as does GNU grep. However, pcregrep was also printing all files when
1723        --files-with-matches was added. Now, when both options are given, it prints
1724        counts only for those files that have at least one match. (GNU grep just
1725        prints the file name in this circumstance, but including the count seems
1726        more useful - otherwise, why use --count?) Also ensured that the
1727        combination -clh just lists non-zero counts, with no names.
1729    4.  The long form of the pcregrep -F option was incorrectly implemented as
1730        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1731        but it seems right to fix it, and I didn't think it was worth preserving
1732        the old behaviour.
1734    5.  The command line items --regex=pattern and --regexp=pattern were not
1735        recognized by pcregrep, which required --regex pattern or --regexp pattern
1736        (with a space rather than an '='). The man page documented the '=' forms,
1737        which are compatible with GNU grep; these now work.
1739    6.  No libpcreposix.pc file was created for pkg-config; there was just
1740        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1742    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1743        when UCP support is not needed, by modifying the Python script that
1744        generates it from Unicode data files. This should not matter if the module
1745        is correctly used as a library, but I received one complaint about 50K of
1746        unwanted data. My guess is that the person linked everything into his
1747        program rather than using a library. Anyway, it does no harm.
1749    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1750        was a minimum greater than 1 for a wide character in a possessive
1751        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1752        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1753        character. Chaos in the form of incorrect output or a compiling loop could
1754        result.
1756    9.  The restrictions on what a pattern can contain when partial matching is
1757        requested for pcre_exec() have been removed. All patterns can now be
1758        partially matched by this function. In addition, if there are at least two
1759        slots in the offset vector, the offset of the earliest inspected character
1760        for the match and the offset of the end of the subject are set in them when
1761        PCRE_ERROR_PARTIAL is returned.
1763    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1764        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1765        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1766        and may be more useful for multi-segment matching.
1768    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1769        used to be given if ever the end of the subject was reached; now it is
1770        given only if matching could not proceed because another character was
1771        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1772        string "Z", which now yields "no match" instead of "partial match". In the
1773        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1774        final character ended with (*FAIL).
1776    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1777        if the pattern had a "must contain" character that was already found in the
1778        earlier partial match, unless partial matching was again requested. For
1779        example, with the pattern /dog.(body)?/, the "must contain" character is
1780        "g". If the first part-match was for the string "dog", restarting with
1781        "sbody" failed. This bug has been fixed.
1783    13. The string returned by pcre_dfa_exec() after a partial match has been
1784        changed so that it starts at the first inspected character rather than the
1785        first character of the match. This makes a difference only if the pattern
1786        starts with a lookbehind assertion or \b or \B (\K is not supported by
1787        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1788        matching functions compatible, and I think it's the right thing to do.
1790    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1791        so that the demonstration program is easily available in environments where
1792        PCRE has not been installed from source.
1794    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1795        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1796        library.
1798    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1799        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1800        is not the first non-POSIX option to be added. Clearly some people find
1801        these options useful.
1803    17. If a caller to the POSIX matching function regexec() passes a non-zero
1804        value for nmatch with a NULL value for pmatch, the value of
1805        nmatch is forced to zero.
1807    18. RunGrepTest did not have a test for the availability of the -u option of
1808        the diff command, as RunTest does. It now checks in the same way as
1809        RunTest, and also checks for the -b option.
1811    19. If an odd number of negated classes containing just a single character
1812        interposed, within parentheses, between a forward reference to a named
1813        subpattern and the definition of the subpattern, compilation crashed with
1814        an internal error, complaining that it could not find the referenced
1815        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1816        [The bug was that it was starting one character too far in when skipping
1817        over the character class, thus treating the ] as data rather than
1818        terminating the class. This meant it could skip too much.]
1820    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1821        /g option in pcretest when the pattern contains \K, which makes it possible
1822        to have an empty string match not at the start, even when the pattern is
1823        anchored. Updated pcretest and pcredemo to use this option.
1825    21. If the maximum number of capturing subpatterns in a recursion was greater
1826        than the maximum at the outer level, the higher number was returned, but
1827        with unset values at the outer level. The correct (outer level) value is
1828        now given.
1830    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1831        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1832        make it do so. The string so far is captured, making this feature
1833        compatible with Perl.
1835    23. The tests have been re-organized, adding tests 11 and 12, to make it
1836        possible to check the Perl 5.10 features against Perl 5.10.
1838    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1839        pattern matches a fixed length string. PCRE did not allow this; now it
1840        does. Neither allows recursion.
1842    25. I finally figured out how to implement a request to provide the minimum
1843        length of subject string that was needed in order to match a given pattern.
1844        (It was back references and recursion that I had previously got hung up
1845        on.) This code has now been added to pcre_study(); it finds a lower bound
1846        to the length of subject needed. It is not necessarily the greatest lower
1847        bound, but using it to avoid searching strings that are too short does give
1848        some useful speed-ups. The value is available to calling programs via
1849        pcre_fullinfo().
1851    26. While implementing 25, I discovered to my embarrassment that pcretest had
1852        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1853        study optimizations had never been tested with that matching function.
1854        Oops. What is worse, even when it was passed study data, there was a bug in
1855        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1856        were also very few tests of studied patterns with pcre_dfa_exec().
1858    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1859        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1860        on the other side of the coin, they are no longer allowed to have different
1861        names, because these cannot be distinguished in PCRE, and this has caused
1862        confusion. (This is a difference from Perl.)
1864    28. When duplicate subpattern names are present (necessarily with different
1865        numbers, as required by 27 above), and a test is made by name in a
1866        conditional pattern, either for a subpattern having been matched, or for
1867        recursion in such a pattern, all the associated numbered subpatterns are
1868        tested, and the overall condition is true if the condition is true for any
1869        one of them. This is the way Perl works, and is also more like the way
1870        testing by number works.
1873    Version 7.9 11-Apr-09
1874    ---------------------
1876    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1877        (pcretest), all targets were linked against these libraries. This included
1878        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1879        libraries. This caused unwanted dependencies to be created. This problem
1880        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1881        pcretest is linked with readline.
1883    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1884        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1885        moved outside it again, because FALSE and TRUE are already defined in AIX,
1886        but BOOL is not.
1888    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1889        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1891    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1892        hyphens) following file names and line numbers when outputting matching
1893        lines. This is not true; no spaces are inserted. I have also clarified the
1894        wording for the --colour (or --color) option.
1896    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1897        was not coloured; this is different to GNU grep, so I have changed it to be
1898        the same.
1900    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1901        each matching line was coloured. Now it goes on to look for further matches
1902        of any of the test patterns, which is the same behaviour as GNU grep.
1904    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1905        doesn't make sense to accept an empty string match in pcregrep, so I have
1906        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1907        seems to be how GNU grep behaves. [But see later change 40 for release
1908        8.33.]
1910    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1911        start or after a newline", because the conditional assertion was not being
1912        correctly handled. The rule now is that both the assertion and what follows
1913        in the first alternative must satisfy the test.
1915    9.  If auto-callout was enabled in a pattern with a conditional group whose
1916        condition was an assertion, PCRE could crash during matching, both with
1917        pcre_exec() and pcre_dfa_exec().
1919    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1920        used for matching.
1922    11. Unicode property support in character classes was not working for
1923        characters (bytes) greater than 127 when not in UTF-8 mode.
1925    12. Added the -M command line option to pcretest.
1927    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1929    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1931    16. Added comments and documentation about mis-use of no_arg in the C++
1932        wrapper.
1934    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1935        from Martin Jerabek that uses macro names for all relevant character and
1936        string constants.
1938    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1939        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1940        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1941        these, but not everybody uses configure.
1943    19. A conditional group that had only one branch was not being correctly
1944        recognized as an item that could match an empty string. This meant that an
1945        enclosing group might also not be so recognized, causing infinite looping
1946        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1947        with the subject "ab", where knowledge that the repeated group can match
1948        nothing is needed in order to break the loop.
1950    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1951        exec(), but without supplying a callout function, matching went wrong.
1953    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1954        leak if the size of the offset vector was greater than 30. When the vector
1955        is smaller, the saved offsets during recursion go onto a local stack
1956        vector, but for larger vectors malloc() is used. It was failing to free
1957        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1958        error, in fact).
1960    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1961        heapframe that is used only when UTF-8 support is enabled. This caused no
1962        problem, but was untidy.
1964    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1965        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1966        included within another project.
1968    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1969        slightly modified by me:
1971          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1972              not building pcregrep.
1974          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1975              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1977    25. Forward references, both numeric and by name, in patterns that made use of
1978        duplicate group numbers, could behave incorrectly or give incorrect errors,
1979        because when scanning forward to find the reference group, PCRE was not
1980        taking into account the duplicate group numbers. A pattern such as
1981        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1983    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1984        the feature of a custom pointer more persuasive (as requested by a user).
1986    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1987        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1988        pre-defined.
1990    28. Added support for (*UTF8) at the start of a pattern.
1992    29. Arrange for flags added by the "release type" setting in CMake to be shown
1993        in the configuration summary.
1996    Version 7.8 05-Sep-08
1997    ---------------------
1999    1.  Replaced UCP searching code with optimized version as implemented for Ad
2000        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2001        stage table and inline lookup instead of a function, giving speed ups of 2
2002        to 5 times on some simple patterns that I tested. Permission was given to
2003        distribute the MultiStage2.py script that generates the tables (it's not in
2004        the tarball, but is in the Subversion repository).
2006    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2007        scripts.
2009    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2010        a group with a zero qualifier. The result of the study could be incorrect,
2011        or the function might crash, depending on the pattern.
2013    4.  Caseless matching was not working for non-ASCII characters in back
2014        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2015        It now works when Unicode Property Support is available.
2017    5.  In pcretest, an escape such as \x{de} in the data was always generating
2018        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2019        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2020        truncation.
2022    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2024    7.  Added two (int) casts to pcregrep when printing the difference of two
2025        pointers, in case they are 64-bit values.
2027    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
2028        test 2 if it fails.
2030    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2031        and a #define of that name to empty if it is not externally set. This is to
2032        allow users of MSVC to set it if necessary.
2034    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2035        the convenience functions in the pcre_get.c source file.
2037    11. An option change at the start of a pattern that had top-level alternatives
2038        could cause overwriting and/or a crash. This command provoked a crash in
2039        some environments:
2041          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2043        This potential security problem was recorded as CVE-2008-2371.
2045    12. For a pattern where the match had to start at the beginning or immediately
2046        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2047        pcre_dfa_exec() could read past the end of the passed subject if there was
2048        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2049        pcretest so that it places the subject at the end of its malloc-ed buffer.
2051    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2052        exec() might read past the end of the data buffer in UTF-8 mode.
2054    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2055        the data contained the byte 0x85 as part of a UTF-8 character within its
2056        first line. This applied both to normal and DFA matching.
2058    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2059        /^[^d]*?$/8 failed to match "abc".
2061    16. Added a missing copyright notice to pcrecpp_internal.h.
2063    17. Make it more clear in the documentation that values returned from
2064        pcre_exec() in ovector are byte offsets, not character counts.
2066    18. Tidied a few places to stop certain compilers from issuing warnings.
2068    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2069        supplied by Stefan Weber. I made a further small update for 7.8 because
2070        there is a change of source arrangements: the pcre_searchfuncs.c module is
2071        replaced by pcre_ucd.c.
2074    Version 7.7 07-May-08
2075    ---------------------
2077    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2078        a string to a long long, pretend we don't even have a long long." This is
2079        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2081    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2082        pre-7.6 versions, which defined a global no_arg variable instead of putting
2083        it in the RE class. (See also #8 below.)
2085    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2086        Lopes.
2088    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2090        (1) The include/exclude patterns were being applied to the whole pathnames
2091            of files, instead of just to the final components.
2093        (2) If there was more than one level of directory, the subdirectories were
2094            skipped unless they satisfied the include/exclude conditions. This is
2095            inconsistent with GNU grep (and could even be seen as contrary to the
2096            pcregrep specification - which I improved to make it absolutely clear).
2097            The action now is always to scan all levels of directory, and just
2098            apply the include/exclude patterns to regular files.
2100    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2101        --exclude_dir in the tests to avoid scanning .svn directories.
2103    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2104        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2105        doesn't support NULs in patterns.
2107    7.  Added some missing "const"s to declarations of static tables in
2108        pcre_compile.c and pcre_dfa_exec.c.
2110    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2111        caused by fix #2  above. (Subsequently also a second patch to fix the
2112        first patch. And a third patch - this was a messy problem.)
2114    9.  Applied Craig's patch to remove the use of push_back().
2116    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2117        matching function regexec().
2119    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2120        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2121        references. PCRE supports relative numbers with this syntax (I don't think
2122        Oniguruma does).
2124    12. Previously, a group with a zero repeat such as (...){0} was completely
2125        omitted from the compiled regex. However, this means that if the group
2126        was called as a subroutine from elsewhere in the pattern, things went wrong
2127        (an internal error was given). Such groups are now left in the compiled
2128        pattern, with a new opcode that causes them to be skipped at execution
2129        time.
2131    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2132        to the way PCRE behaves:
2134        (a) A lone ] character is dis-allowed (Perl treats it as data).
2136        (b) A back reference to an unmatched subpattern matches an empty string
2137            (Perl fails the current match path).
2139        (c) A data ] in a character class must be notated as \] because if the
2140            first data character in a class is ], it defines an empty class. (In
2141            Perl it is not possible to have an empty class.) The empty class []
2142            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2143            The negative empty class [^] matches any one character, independently
2144            of the DOTALL setting.
2146    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2147        non-existent subpattern following a character class starting with ']' and
2148        containing () gave an internal compiling error instead of "reference to
2149        non-existent subpattern". Fortunately, when the pattern did exist, the
2150        compiled code was correct. (When scanning forwards to check for the
2151        existence of the subpattern, it was treating the data ']' as terminating
2152        the class, so got the count wrong. When actually compiling, the reference
2153        was subsequently set up correctly.)
2155    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2156        it was being rejected as not supported by pcre_dfa_exec(), even though
2157        other assertions are supported. I have made pcre_dfa_exec() support
2158        (*FAIL).
2160    16. The implementation of 13c above involved the invention of a new opcode,
2161        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2162        cannot be changed at match time, I realized I could make a small
2163        improvement to matching performance by compiling OP_ALLANY instead of
2164        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2165        on the OP_ANY path.
2167    17. Compiling pcretest on Windows with readline support failed without the
2168        following two fixes: (1) Make the unistd.h include conditional on
2169        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2171    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2172        ncurses library to be included for pcretest when ReadLine support is
2173        requested, but also to allow for it to be overridden. This patch came from
2174        Daniel Bergström.
2176    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2177        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2178        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2179        spotting this.
2182    Version 7.6 28-Jan-08
2183    ---------------------
2185    1.  A character class containing a very large number of characters with
2186        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2187        overflow.
2189    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2190        HAVE_LONG_LONG is not defined.
2192    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2193        bring them up to date and include new features. This patch includes:
2195        - Fixed PH's badly added libz and libbz2 support.
2196        - Fixed a problem with static linking.
2197        - Added pcredemo. [But later removed - see 7 below.]
2198        - Fixed dftables problem and added an option.
2199        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2200            HAVE_LONG_LONG.
2201        - Added readline support for pcretest.
2202        - Added an listing of the option settings after cmake has run.
2204    4.  A user submitted a patch to Makefile that makes it easy to create
2205        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2206        Makefile.am that cause it to include this special target, without
2207        affecting anything else. Note that the same mingw target plus all
2208        the other distribution libraries and programs are now supported
2209        when configuring with CMake (see 6 below) instead of with
2210        Configure/Make.
2212    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2213        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2214        exported in the Windows port". It has not yet been confirmed that the patch
2215        solves the problem, but it does no harm.
2217    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2218        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2219        with CMake, and also correct the comment about stack recursion.
2221    7.  Remove the automatic building of pcredemo from the ./configure system and
2222        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2223        of a program that users should build themselves after PCRE is installed, so
2224        building it automatically is not really right. What is more, it gave
2225        trouble in some build environments.
2227    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2230    Version 7.5 10-Jan-08
2231    ---------------------
2233    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2234        values in parens when parsing an RE using the C++ wrapper."
2236    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2237        Characters greater than 255 were excluded from the class instead of being
2238        included.
2240    3.  The same bug as (2) above applied to negated POSIX classes such as
2241        [:^space:].
2243    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2244        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2245        I have changed it.
2247    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2248        first named subpattern) and a construct such as (?&a) would reference the
2249        first named subpattern whose name started with "a" (in other words, the
2250        length check was missing). Both these problems are fixed. "Subpattern name
2251        expected" is now given for (?&) (a zero-length name), and this patch also
2252        makes it give the same error for \k'' (previously it complained that that
2253        was a reference to a non-existent subpattern).
2255    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2256        this is right because (?- can be followed by option settings as well as by
2257        digits. I have, however, made the messages clearer.
2259    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2260        than the number used in the conditional) now cause a compile-time error.
2261        This is actually not compatible with Perl, which accepts such patterns, but
2262        treats the conditional as always being FALSE (as PCRE used to), but it
2263        seems to me that giving a diagnostic is better.
2265    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2266        and messages.
2268    9.  Fix two occurrences of "backslash" in comments that should have been
2269        "backspace".
2271    10. Remove two redundant lines of code that can never be obeyed (their function
2272        was moved elsewhere).
2274    11. The program that makes PCRE's Unicode character property table had a bug
2275        which caused it to generate incorrect table entries for sequences of
2276        characters that have the same character type, but are in different scripts.
2277        It amalgamated them into a single range, with the script of the first of
2278        them. In other words, some characters were in the wrong script. There were
2279        thirteen such cases, affecting characters in the following ranges:
2281          U+002b0 - U+002c1
2282          U+0060c - U+0060d
2283          U+0061e - U+00612
2284          U+0064b - U+0065e
2285          U+0074d - U+0076d
2286          U+01800 - U+01805
2287          U+01d00 - U+01d77
2288          U+01d9b - U+01dbf
2289          U+0200b - U+0200f
2290          U+030fc - U+030fe
2291          U+03260 - U+0327f
2292          U+0fb46 - U+0fbb1
2293          U+10450 - U+1049d
2295    12. The -o option (show only the matching part of a line) for pcregrep was not
2296        compatible with GNU grep in that, if there was more than one match in a
2297        line, it showed only the first of them. It now behaves in the same way as
2298        GNU grep.
2300    13. If the -o and -v options were combined for pcregrep, it printed a blank
2301        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2302        does the same. The return code can be used to tell if there were any
2303        non-matching lines.
2305    14. Added --file-offsets and --line-offsets to pcregrep.
2307    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2308        infinitely looping recursion. The bug was that positive lookaheads were not
2309        being skipped when checking for a possible empty match (negative lookaheads
2310        and both kinds of lookbehind were skipped).
2312    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2313        inclusion of <windows.h> to before rather than after the definition of
2314        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2316    17. Specifying a possessive quantifier with a specific limit for a Unicode
2317        character property caused pcre_compile() to compile bad code, which led at
2318        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2319        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2320        caused the error; without that there was no problem.
2322    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2324    19. Added --enable-pcretest-libreadline.
2326    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2327        RE::GlobalReplace(). As a result, the number of replacements returned was
2328        double what it should be. I removed one of the increments, but Craig sent a
2329        later patch that removed the other one (the right fix) and added unit tests
2330        that check the return values (which was not done before).
2332    21. Several CMake things:
2334        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2335            the names libpcre and libpcreposix, not just pcre and pcreposix.
2337        (2) The above change means that pcretest and pcregrep are now correctly
2338            linked with the newly-built libraries, not previously installed ones.
2342    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2343        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2344        UTF-8 newline character). The key issue is that the pattern starts .*;
2345        this means that the match must be either at the beginning, or after a
2346        newline. The bug was in the code for advancing after a failed match and
2347        checking that the new position followed a newline. It was not taking
2348        account of UTF-8 characters correctly.
2350    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2351        character classes. PCRE was not treating the sequence [:...:] as a
2352        character class unless the ... were all letters. Perl, however, seems to
2353        allow any characters between [: and :], though of course it rejects as
2354        unknown any "names" that contain non-letters, because all the known class
2355        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2356        for example, whereas PCRE did not - it did not recognize a POSIX character
2357        class. This seemed a bit dangerous, so the code has been changed to be
2358        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2359        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2360        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2361        Perl does, and where it didn't before.
2363    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2364        Windows environments %n is disabled by default.
2367    Version 7.4 21-Sep-07
2368    ---------------------
2370    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2371        means that a class such as [\s] counted as "explicit reference to CR or
2372        LF". That isn't really right - the whole point of the change was to try to
2373        help when there was an actual mention of one of the two characters. So now
2374        the change happens only if \r or \n (or a literal CR or LF) character is
2375        encountered.
2377    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2378        of both had grown to the point where there were only 3 bits left.
2379        Fortunately, there was spare space in the data structure, and so I have
2380        moved the internal flags into a new 16-bit field to free up more option
2381        bits.
2383    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2384        but did not set the internal JCHANGED flag - either of these is enough to
2385        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2386        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2387        start sets both bits.
2389    4.  Added options (at build time, compile time, exec time) to change \R from
2390        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2392    5.  doc/pcresyntax.html was missing from the distribution.
2394    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2395        compatibility, even though it is no longer used.
2397    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2398        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2399        windows.h file is present (where different names are used). [This was
2400        reversed later after testing - see 16 below.]
2402    8.  Changed all #include <config.h> to #include "config.h". There were also
2403        some further <pcre.h> cases that I changed to "pcre.h".
2405    9.  When pcregrep was used with the --colour option, it missed the line ending
2406        sequence off the lines that it output.
2408    10. It was pointed out to me that arrays of string pointers cause lots of
2409        relocations when a shared library is dynamically loaded. A technique of
2410        using a single long string with a table of offsets can drastically reduce
2411        these. I have refactored PCRE in four places to do this. The result is
2412        dramatic:
2414          Originally:                          290
2415          After changing UCP table:            187
2416          After changing error message table:   43
2417          After changing table of "verbs"       36
2418          After changing table of Posix names   22
2420        Thanks to the folks working on Gregex for glib for this insight.
2422    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2423        unicode-properties was also set.
2425    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2427    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2428        checked only for CRLF.
2430    14. Added casts to pcretest.c to avoid compiler warnings.
2432    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2434    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2435        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2436        entirely. This removes changes made in 7 above.
2438    17. The CMake files have been updated, and there is now more information about
2439        building with CMake in the NON-UNIX-USE document.
2442    Version 7.3 28-Aug-07
2443    ---------------------
2445     1. In the rejigging of the build system that eventually resulted in 7.1, the
2446        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2447        brackets there is not right, since it causes compilers to look for an
2448        installed pcre.h, not the version that is in the source that is being
2449        compiled (which of course may be different). I have changed it back to:
2451          #include "pcre.h"
2453        I have a vague recollection that the change was concerned with compiling in
2454        different directories, but in the new build system, that is taken care of
2455        by the VPATH setting the Makefile.
2457     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2458        when the subject happened to end in the byte 0x85 (e.g. if the last
2459        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2460        characters but of course it shouldn't be taken as a newline when it is part
2461        of another character. The bug was that, for an unlimited repeat of . in
2462        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2463        characters when looking for a newline.
2465     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2467     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2468        in debug output.
2470     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2471        long printing in the pcrecpp unittest when running under MinGW.
2473     6. ESC_K was left out of the EBCDIC table.
2475     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2476        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2477        limit also applies to "virtual nesting" when a pattern is recursive, and in
2478        this case 1000 isn't so big. I have been able to remove this limit at the
2479        expense of backing off one optimization in certain circumstances. Normally,
2480        when pcre_exec() would call its internal match() function recursively and
2481        immediately return the result unconditionally, it uses a "tail recursion"
2482        feature to save stack. However, when a subpattern that can match an empty
2483        string has an unlimited repetition quantifier, it no longer makes this
2484        optimization. That gives it a stack frame in which to save the data for
2485        checking that an empty string has been matched. Previously this was taken
2486        from the 1000-entry workspace that had been reserved. So now there is no
2487        explicit limit, but more stack is used.
2489     8. Applied Daniel's patches to solve problems with the import/export magic
2490        syntax that is required for Windows, and which was going wrong for the
2491        pcreposix and pcrecpp parts of the library. These were overlooked when this
2492        problem was solved for the main library.
2494     9. There were some crude static tests to avoid integer overflow when computing
2495        the size of patterns that contain repeated groups with explicit upper
2496        limits. As the maximum quantifier is 65535, the maximum group length was
2497        set at 30,000 so that the product of these two numbers did not overflow a
2498        32-bit integer. However, it turns out that people want to use groups that
2499        are longer than 30,000 bytes (though not repeat them that many times).
2500        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2501        made it possible to implement the integer overflow checks in a much more
2502        dynamic way, which I have now done. The artificial limitation on group
2503        length has been removed - we now have only the limit on the total length of
2504        the compiled pattern, which depends on the LINK_SIZE setting.
2506    10. Fixed a bug in the documentation for get/copy named substring when
2507        duplicate names are permitted. If none of the named substrings are set, the
2508        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2509        empty string.
2511    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2512        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2513        because the ] is interpreted as the first data character and the
2514        terminating ] is not found. PCRE has been made compatible with Perl in this
2515        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2516        cause memory overwriting.
2518    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2519        string has been matched (to stop an infinite loop). It was not recognizing
2520        a conditional subpattern that could match an empty string if that
2521        subpattern was within another subpattern. For example, it looped when
2522        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2523        condition was not nested. This bug has been fixed.
2525    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2526        past the start of the subject in the presence of bytes with the top bit
2527        set, for example "\x8aBCD".
2529    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2530        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2532    14. Optimized (?!) to (*FAIL).
2534    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2535        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2536        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2537        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2538        does: it's just the validity check that is more restrictive.
2540    16. Inserted checks for integer overflows during escape sequence (backslash)
2541        processing, and also fixed erroneous offset values for syntax errors during
2542        backslash processing.
2544    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2545        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2547    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2548        caused an overrun.
2550    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2551        something other than just ASCII characters) inside a group that had an
2552        unlimited repeat caused a loop at compile time (while checking to see
2553        whether the group could match an empty string).
2555    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2556        [\P{Any}] did so. (Error in the code for printing property names.)
2558    21. An orphan \E inside a character class could cause a crash.
2560    22. A repeated capturing bracket such as (A)? could cause a wild memory
2561        reference during compilation.
2563    23. There are several functions in pcre_compile() that scan along a compiled
2564        expression for various reasons (e.g. to see if it's fixed length for look
2565        behind). There were bugs in these functions when a repeated \p or \P was
2566        present in the pattern. These operators have additional parameters compared
2567        with \d, etc, and these were not being taken into account when moving along
2568        the compiled data. Specifically:
2570        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2571            length.
2573        (b) An item such as \pL+ within a repeated group could cause crashes or
2574            loops.
2576        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2577            "reference to non-existent subpattern" error.
2579        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2581    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2582        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2584    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2586    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2587        character were causing crashes (broken optimization).
2589    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2590        \p or \P) caused a compile-time loop.
2592    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2593        break. For example, the unstudied pattern [\r\n]A does not match the string
2594        "\r\nA" because change 7.0/46 below moves the current point on by two
2595        characters after failing to match at the start. However, the pattern \nA
2596        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2597        the same is true. There doesn't seem any very clean way out of this, but
2598        what I have chosen to do makes the common cases work: PCRE now takes note
2599        of whether there can be an explicit match for \r or \n anywhere in the
2600        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2601        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2602        pattern has explicit CR or LF references.
2604    29. Added (*CR) etc for changing newline setting at start of pattern.
2607    Version 7.2 19-Jun-07
2608  ---------------------  ---------------------
2610   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 40  Version 7.2 05-June-07 Line 2643  Version 7.2 05-June-07
2644      (f) \g{name} is another synonym - part of Perl 5.10's unification of      (f) \g{name} is another synonym - part of Perl 5.10's unification of
2645          reference syntax.          reference syntax.
2647      (g) (?| introduces a group in which the numbering of parentheses in each      (g) (?| introduces a group in which the numbering of parentheses in each
2648          alternative starts with the same number.          alternative starts with the same number.
2650        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2652   7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and   7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
# Line 58  Version 7.2 05-June-07 Line 2663  Version 7.2 05-June-07
2663      alternatives. The 1000-alternative test pattern now uses 12 bytes of      alternatives. The 1000-alternative test pattern now uses 12 bytes of
2664      workspace instead of running out of the 4096 that are available.      workspace instead of running out of the 4096 that are available.
2666    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2668    11. Applied patch from Google to remove an optimization that didn't quite work.
2669        The report of the bug said:
2671          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2672          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2673          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2675    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2676        it matched the wrong number of bytes.
2679  Version 7.1 24-Apr-07  Version 7.1 24-Apr-07
2680  ---------------------  ---------------------

Removed from v.175  
changed lines
  Added in v.1418

  ViewVC Help
Powered by ViewVC 1.1.5