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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.208  
changed lines
  Added in v.1372

  ViewVC Help
Powered by ViewVC 1.1.5