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

Diff of /code/trunk/ChangeLog

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

revision 98 by ph10, Mon Mar 5 15:55:02 2007 UTC revision 1366 by ph10, Sun Oct 6 18:34:33 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.1 05-Mar-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    
102    Version 8.33 28-May-2013
103    ------------------------
104    
105    1.  Added 'U' to some constants that are compared to unsigned integers, to
106        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
107        variables that are added to signed variables, to ensure the result is
108        signed and can be negated.
109    
110    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
111        CMake config files.
112    
113    3.  Revise the creation of config.h.generic so that all boolean macros are
114        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
115        overriding via -D on the command line possible.
116    
117    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
118        to unsigned int is reported to make a quite noticeable speed difference in
119        a specific Windows environment. Testing on Linux did also appear to show
120        some benefit (and it is clearly not harmful). Also fixed the definition of
121        Xop which should be unsigned.
122    
123    5.  Related to (4), changing the definition of the intermediate variable cc
124        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
125        improvements.
126    
127    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
128        unnecessary spaces.
129    
130    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
131        and later.
132    
133    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
134    
135    9.  Optimizing fast_forward_start_bits in JIT.
136    
137    10. Adding support for callouts in JIT, and fixing some issues revealed
138        during this work. Namely:
139    
140        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
141    
142        (b) Minimum length was not checked before the matching is started.
143    
144    11. The value of capture_last that is passed to callouts was incorrect in some
145        cases when there was a capture on one path that was subsequently abandoned
146        after a backtrack. Also, the capture_last value is now reset after a
147        recursion, since all captures are also reset in this case.
148    
149    12. The interpreter no longer returns the "too many substrings" error in the
150        case when an overflowing capture is in a branch that is subsequently
151        abandoned after a backtrack.
152    
153    13. In the pathological case when an offset vector of size 2 is used, pcretest
154        now prints out the matched string after a yield of 0 or 1.
155    
156    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
157        Only supported by the JIT compiler at the moment.
158    
159    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
160    
161    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
162        offset of the starting point of the matching process, provided the offsets
163        vector is large enough.
164    
165    17. The \A escape now records a lookbehind value of 1, though its execution
166        does not actually inspect the previous character. This is to ensure that,
167        in partial multi-segment matching, at least one character from the old
168        segment is retained when a new segment is processed. Otherwise, if there
169        are no lookbehinds in the pattern, \A might match incorrectly at the start
170        of a new segment.
171    
172    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
173    
174    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
175        gives some modest performance improvement in 8-bit mode.
176    
177    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
178        be expressed in certain programming languages using Universal Character
179        Names.
180    
181    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
182        which points out that "non characters" are not "characters that may not
183        appear in Unicode strings" but rather "characters that are reserved for
184        internal use and have only local meaning".
185    
186    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
187        there was a conditional group that depended on an assertion, if the
188        assertion was false, the callout that immediately followed the alternation
189        in the condition was skipped when pcre_exec() was used for matching.
190    
191    23. Allow an explicit callout to be inserted before an assertion that is the
192        condition for a conditional group, for compatibility with automatic
193        callouts, which always insert a callout at this point.
194    
195    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
196        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
197    
198    25. (*PRUNE) is now supported by the JIT compiler.
199    
200    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
201    
202    27. Fix the case where there are two or more SKIPs with arguments that may be
203        ignored.
204    
205    28. (*SKIP) is now supported by the JIT compiler.
206    
207    29. (*THEN) is now supported by the JIT compiler.
208    
209    30. Update RunTest with additional test selector options.
210    
211    31. The way PCRE handles backtracking verbs has been changed in two ways.
212    
213        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
214        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
215        backtracking. In some cases this makes it more Perl-compatible, but Perl's
216        rather obscure rules do not always do the same thing.
217    
218        (2) Previously, backtracking verbs were confined within assertions. This is
219        no longer the case for positive assertions, except for (*ACCEPT). Again,
220        this sometimes improves Perl compatibility, and sometimes does not.
221    
222    32. A number of tests that were in test 2 because Perl did things differently
223        have been moved to test 1, because either Perl or PCRE has changed, and
224        these tests are now compatible.
225    
226    32. Backtracking control verbs are now handled in the same way in JIT and
227        interpreter.
228    
229    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
230        contained a forward subroutine reference caused a compile error.
231    
232    34. Auto-detect and optimize limited repetitions in JIT.
233    
234    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
235        blocking (*UTF) etc.
236    
237    36. In the interpreter, maximizing pattern repetitions for characters and
238        character types now use tail recursion, which reduces stack usage.
239    
240    37. The value of the max lookbehind was not correctly preserved if a compiled
241        and saved regex was reloaded on a host of different endianness.
242    
243    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
244        of the compiled pattern block, expand the flags field from 16 to 32 bits
245        because it was almost full.
246    
247    39. Try madvise first before posix_madvise.
248    
249    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
250        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
251        original change locked out all matches of empty strings. This has been
252        changed so that one match of an empty string per line is recognized.
253        Subsequent searches on the same line (for colouring or for --only-matching,
254        for example) do not recognize empty strings.
255    
256    41. Applied a user patch to fix a number of spelling mistakes in comments.
257    
258    42. Data lines longer than 65536 caused pcretest to crash.
259    
260    43. Clarified the data type for length and startoffset arguments for pcre_exec
261        and pcre_dfa_exec in the function-specific man pages, where they were
262        explicitly stated to be in bytes, never having been updated. I also added
263        some clarification to the pcreapi man page.
264    
265    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
266        a segmentation fault.
267    
268    
269    Version 8.32 30-November-2012
270    -----------------------------
271    
272    1.  Improved JIT compiler optimizations for first character search and single
273        character iterators.
274    
275    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
276        Patch by Daniel Richard G.
277    
278    3.  Single character iterator optimizations in the JIT compiler.
279    
280    4.  Improved JIT compiler optimizations for character ranges.
281    
282    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
283        Reported by Giuseppe D'Angelo.
284    
285    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
286        of a line, was being set incorrectly in cases where .* appeared inside
287        atomic brackets at the start of a pattern, or where there was a subsequent
288        *PRUNE or *SKIP.
289    
290    7.  Improved instruction cache flush for POWER/PowerPC.
291        Patch by Daniel Richard G.
292    
293    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
294        grep:
295    
296        (a) There is now no limit to the number of patterns to be matched.
297    
298        (b) An error is given if a pattern is too long.
299    
300        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
301            are now supported.
302    
303        (d) --exclude-from and --include-from (multiple use) have been added.
304    
305        (e) Exclusions and inclusions now apply to all files and directories, not
306            just to those obtained from scanning a directory recursively.
307    
308        (f) Multiple uses of -f and --file-list are now supported.
309    
310        (g) In a Windows environment, the default for -d has been changed from
311            "read" (the GNU grep default) to "skip", because otherwise the presence
312            of a directory in the file list provokes an error.
313    
314        (h) The documentation has been revised and clarified in places.
315    
316    9.  Improve the matching speed of capturing brackets.
317    
318    10. Changed the meaning of \X so that it now matches a Unicode extended
319        grapheme cluster.
320    
321    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
322        out POSIX threads when JIT support is configured.
323    
324    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
325    
326    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
327        structure could go wrong in environments where size_t is not the same size
328        as int.
329    
330    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
331        set.
332    
333    15. The EBCDIC support had decayed; later updates to the code had included
334        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
335        general tidy up of EBCDIC-related issues, and the documentation was also
336        not quite right. There is now a test that can be run on ASCII systems to
337        check some of the EBCDIC-related things (but is it not a full test).
338    
339    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
340        in a small tidy to the code.
341    
342    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
343    
344    18. If the --only-matching (-o) option in pcregrep is specified multiple
345        times, each one causes appropriate output. For example, -o1 -o2 outputs the
346        substrings matched by the 1st and 2nd capturing parentheses. A separating
347        string can be specified by --om-separator (default empty).
348    
349    19. Improving the first n character searches.
350    
351    20. Turn case lists for horizontal and vertical white space into macros so that
352        they are defined only once.
353    
354    21. This set of changes together give more compatible Unicode case-folding
355        behaviour for characters that have more than one other case when UCP
356        support is available.
357    
358        (a) The Unicode property table now has offsets into a new table of sets of
359            three or more characters that are case-equivalent. The MultiStage2.py
360            script that generates these tables (the pcre_ucd.c file) now scans
361            CaseFolding.txt instead of UnicodeData.txt for character case
362            information.
363    
364        (b) The code for adding characters or ranges of characters to a character
365            class has been abstracted into a generalized function that also handles
366            case-independence. In UTF-mode with UCP support, this uses the new data
367            to handle characters with more than one other case.
368    
369        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
370            whose other case is greater than 256 are now correctly matched
371            caselessly. Previously, the high codepoint matched the low one, but not
372            vice versa.
373    
374        (d) The processing of \h, \H, \v, and \ in character classes now makes use
375            of the new class addition function, using character lists defined as
376            macros alongside the case definitions of 20 above.
377    
378        (e) Caseless back references now work with characters that have more than
379            one other case.
380    
381        (f) General caseless matching of characters with more than one other case
382            is supported.
383    
384    22. Unicode character properties were updated from Unicode 6.2.0
385    
386    23. Improved CMake support under Windows. Patch by Daniel Richard G.
387    
388    24. Add support for 32-bit character strings, and UTF-32
389    
390    25. Major JIT compiler update (code refactoring and bugfixing).
391        Experimental Sparc 32 support is added.
392    
393    26. Applied a modified version of Daniel Richard G's patch to create
394        pcre.h.generic and config.h.generic by "make" instead of in the
395        PrepareRelease script.
396    
397    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
398        pcre_compile.c when checking for a zero character.
399    
400    28. Introducing a native interface for JIT. Through this interface, the compiled
401        machine code can be directly executed. The purpose of this interface is to
402        provide fast pattern matching, so several sanity checks are not performed.
403        However, feature tests are still performed. The new interface provides
404        1.4x speedup compared to the old one.
405    
406    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
407        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
408        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
409    
410    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
411        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
412        the "old" RFC 2279). Instead, it ended up passing a negative length to
413        pcre_exec().
414    
415    31. Add support for GCC's visibility feature to hide internal functions.
416    
417    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
418        "unknown -C option" after outputting 0 or 1.
419    
420    33. There is now support for generating a code coverage report for the test
421        suite in environments where gcc is the compiler and lcov is installed. This
422        is mainly for the benefit of the developers.
423    
424    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
425        unaddressable using valgrind annotations, allowing valgrind to detect
426        invalid memory accesses. This is mainly for the benefit of the developers.
427    
428    25. (*UTF) can now be used to start a pattern in any of the three libraries.
429    
430    26. Give configure error if --enable-cpp but no C++ compiler found.
431    
432    
433    Version 8.31 06-July-2012
434    -------------------------
435    
436    1.  Fixing a wrong JIT test case and some compiler warnings.
437    
438    2.  Removed a bashism from the RunTest script.
439    
440    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
441        to unsigned type, result still unsigned" that was given by an MS compiler
442        on encountering the code "-sizeof(xxx)".
443    
444    4.  Partial matching support is added to the JIT compiler.
445    
446    5.  Fixed several bugs concerned with partial matching of items that consist
447        of more than one character:
448    
449        (a) /^(..)\1/ did not partially match "aba" because checking references was
450            done on an "all or nothing" basis. This also applied to repeated
451            references.
452    
453        (b) \R did not give a hard partial match if \r was found at the end of the
454            subject.
455    
456        (c) \X did not give a hard partial match after matching one or more
457            characters at the end of the subject.
458    
459        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
460            a partial match for the string "\r".
461    
462        (e) When newline was set to CRLF, the metacharacter "." did not recognize
463            a partial match for a CR character at the end of the subject string.
464    
465    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
466        running pcretest, the text "(JIT)" added to the output whenever JIT is
467        actually used to run the match.
468    
469    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
470        or /S+[+] with a digit between 1 and 7.
471    
472    8.  OP_NOT now supports any UTF character not just single-byte ones.
473    
474    9.  (*MARK) control verb is now supported by the JIT compiler.
475    
476    10. The command "./RunTest list" lists the available tests without actually
477        running any of them. (Because I keep forgetting what they all are.)
478    
479    11. Add PCRE_INFO_MAXLOOKBEHIND.
480    
481    12. Applied a (slightly modified) user-supplied patch that improves performance
482        when the heap is used for recursion (compiled with --disable-stack-for-
483        recursion). Instead of malloc and free for each heap frame each time a
484        logical recursion happens, frames are retained on a chain and re-used where
485        possible. This sometimes gives as much as 30% improvement.
486    
487    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
488        call.
489    
490    14. As documented, (*COMMIT) is now confined to within a positive assertion.
491    
492    15. It is now possible to link pcretest with libedit as an alternative to
493        libreadline.
494    
495    16. (*COMMIT) control verb is now supported by the JIT compiler.
496    
497    17. The Unicode data tables have been updated to Unicode 6.1.0.
498    
499    18. Added --file-list option to pcregrep.
500    
501    19. Added binary file support to pcregrep, including the -a, --binary-files,
502        -I, and --text options.
503    
504    20. The madvise function is renamed for posix_madvise for QNX compatibility
505        reasons. Fixed by Giuseppe D'Angelo.
506    
507    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
508    
509    22. Changed the option for creating symbolic links for 16-bit man pages from
510        -s to -sf so that re-installing does not cause issues.
511    
512    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
513    
514    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
515        matches in certain environments (the workspace was not being correctly
516        retained). Also added to pcre_dfa_exec() a simple plausibility check on
517        some of the workspace data at the beginning of a restart.
518    
519    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
520        was not doing so when it should - probably a typo introduced by SVN 528
521        (change 8.10/14).
522    
523    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
524        \w+ when the character tables indicated that \x{c4} was a word character.
525        There were several related cases, all because the tests for doing a table
526        lookup were testing for characters less than 127 instead of 255.
527    
528    27. If a pattern contains capturing parentheses that are not used in a match,
529        their slots in the ovector are set to -1. For those that are higher than
530        any matched groups, this happens at the end of processing. In the case when
531        there were back references that the ovector was too small to contain
532        (causing temporary malloc'd memory to be used during matching), and the
533        highest capturing number was not used, memory off the end of the ovector
534        was incorrectly being set to -1. (It was using the size of the temporary
535        memory instead of the true size.)
536    
537    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
538        ovector size, it uses memory at the end of the block that it has got.
539    
540    29. Check for an overlong MARK name and give an error at compile time. The
541        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
542    
543    30. JIT compiler update.
544    
545    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
546        Rill for the patch.
547    
548    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
549    
550    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
551    
552    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
553        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
554        was enabled.
555    
556    35. Improve JIT code generation for greedy plus quantifier.
557    
558    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
559        1 to "aa" instead of to an empty string. The bug affected repeated groups
560        that could potentially match an empty string.
561    
562    37. Optimizing single character iterators in JIT.
563    
564    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
565        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
566        codepoints that are too big for the mode are faulted, and in a UTF mode,
567        disallowed codepoints are also faulted.
568    
569    39. If PCRE was compiled with UTF support, in three places in the DFA
570        matcher there was code that should only have been obeyed in UTF mode, but
571        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
572        processing when bytes with values greater than 127 were present. In 16-bit
573        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
574        both cases the values are those that cannot be the first data item in a UTF
575        character. The three items that might have provoked this were recursions,
576        possessively repeated groups, and atomic groups.
577    
578    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
579        and pcregrep, because some OS require shared objects to be explicitly
580        passed to ld, causing the link step to fail if they are not.
581    
582    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
583        mode, patterns that started with \h* or \R* might be incorrectly matched.
584    
585    
586    Version 8.30 04-February-2012
587    -----------------------------
588    
589    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
590        name is defined in ctype.h.
591    
592    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
593        only in quite long subpatterns.
594    
595    3.  Removed the function pcre_info(), which has been obsolete and deprecated
596        since it was replaced by pcre_fullinfo() in February 2000.
597    
598    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
599        match a (*MARK), and the match failed at the start of the subject, a
600        reference to memory before the start of the subject could occur. This bug
601        was introduced by fix 17 of release 8.21.
602    
603    5.  A reference to an unset group with zero minimum repetition was giving
604        totally wrong answers (in non-JavaScript-compatibility mode). For example,
605        /(another)?(\1?)test/ matched against "hello world test". This bug was
606        introduced in release 8.13.
607    
608    6.  Add support for 16-bit character strings (a large amount of work involving
609        many changes and refactorings).
610    
611    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
612        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
613        from a file.
614    
615    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
616        rounding is not applied in this particular case).
617    
618    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
619        if they appear, or are escaped, in patterns.
620    
621    10. Get rid of a number of -Wunused-but-set-variable warnings.
622    
623    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
624        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
625        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
626        also returns the mark "x". This bug applied to capturing parentheses,
627        non-capturing parentheses, and atomic parentheses. It also applied to some
628        assertions.
629    
630    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
631        information out of configure.ac instead of relying on pcre.h.generic, which
632        is not stored in the repository.
633    
634    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
635        -lreadline.
636    
637    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
638    
639    15. Applied Graycode's patch to put the top-level frame on the stack rather
640        than the heap when not using the stack for recursion. This gives a
641        performance improvement in many cases when recursion is not deep.
642    
643    16. Experimental code added to "pcretest -C" to output the stack frame size.
644    
645    
646    Version 8.21 12-Dec-2011
647    ------------------------
648    
649    1.  Updating the JIT compiler.
650    
651    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
652        are added as well.
653    
654    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
655        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
656        calling _pcre_jit_exec. Some extra comments are added.
657    
658    4.  (*MARK) settings inside atomic groups that do not contain any capturing
659        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
660        was introduced by change 18 for 8.20.
661    
662    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
663        ECMA-262 standard.
664    
665    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
666        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
667        This bug was probably introduced by change 9 of 8.13.
668    
669    7.  While fixing 6 above, I noticed that a number of other items were being
670        incorrectly rejected as "not fixed length". This arose partly because newer
671        opcodes had not been added to the fixed-length checking code. I have (a)
672        corrected the bug and added tests for these items, and (b) arranged for an
673        error to occur if an unknown opcode is encountered while checking for fixed
674        length instead of just assuming "not fixed length". The items that were
675        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
676        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
677        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
678    
679    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
680        being incorrectly compiled and would have given unpredicatble results.
681    
682    9.  A possessively repeated subpattern with minimum repeat count greater than
683        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
684        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
685        the first (A) could occur when it should not.
686    
687    10. Add a cast and remove a redundant test from the code.
688    
689    11. JIT should use pcre_malloc/pcre_free for allocation.
690    
691    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
692        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
693        is anything other than /usr, -L is still shown).
694    
695    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
696    
697    14. Perl does not support \N without a following name in a [] class; PCRE now
698        also gives an error.
699    
700    15. If a forward reference was repeated with an upper limit of around 2000,
701        it caused the error "internal error: overran compiling workspace". The
702        maximum number of forward references (including repeats) was limited by the
703        internal workspace, and dependent on the LINK_SIZE. The code has been
704        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
705        the default depends on LINK_SIZE. There is a new upper limit (for safety)
706        of around 200,000 forward references. While doing this, I also speeded up
707        the filling in of repeated forward references.
708    
709    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
710        incorrectly expecting the subject to contain another "a" after the start.
711    
712    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
713        in the match, the SKIP should be ignored. This was not happening; instead
714        the SKIP was being treated as NOMATCH. For patterns such as
715        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
716        tested.
717    
718    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
719        now much more compatible with Perl, in particular in cases where the result
720        is a non-match for a non-anchored pattern. For example, if
721        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
722        "m", where previously it did not return a name. A side effect of this
723        change is that for partial matches, the last encountered mark name is
724        returned, as for non matches. A number of tests that were previously not
725        Perl-compatible have been moved into the Perl-compatible test files. The
726        refactoring has had the pleasing side effect of removing one argument from
727        the match() function, thus reducing its stack requirements.
728    
729    19. If the /S+ option was used in pcretest to study a pattern using JIT,
730        subsequent uses of /S (without +) incorrectly behaved like /S+.
731    
732    21. Retrieve executable code size support for the JIT compiler and fixing
733        some warnings.
734    
735    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
736        not work when the shorter character appeared right at the end of the
737        subject string.
738    
739    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
740        systems.
741    
742    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
743        output it when the /M option is used in pcretest.
744    
745    25. The CheckMan script was not being included in the distribution. Also, added
746        an explicit "perl" to run Perl scripts from the PrepareRelease script
747        because this is reportedly needed in Windows.
748    
749    26. If study data was being save in a file and studying had not found a set of
750        "starts with" bytes for the pattern, the data written to the file (though
751        never used) was taken from uninitialized memory and so caused valgrind to
752        complain.
753    
754    27. Updated RunTest.bat as provided by Sheri Pierce.
755    
756    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
757    
758    29. Computation of memory usage for the table of capturing group names was
759        giving an unnecessarily large value.
760    
761    
762    Version 8.20 21-Oct-2011
763    ------------------------
764    
765    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
766        a POSIX class. After further experiments with Perl, which convinced me that
767        Perl has bugs and confusions, a closing square bracket is no longer allowed
768        in a POSIX name. This bug also affected patterns with classes that started
769        with full stops.
770    
771    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
772        captured substring, but while checking the failing first alternative,
773        substring 1 is temporarily captured. If the output vector supplied to
774        pcre_exec() was not big enough for this capture, the yield of the function
775        was still zero ("insufficient space for captured substrings"). This cannot
776        be totally fixed without adding another stack variable, which seems a lot
777        of expense for a edge case. However, I have improved the situation in cases
778        such as /(a)(b)x|abc/ matched against "abc", where the return code
779        indicates that fewer than the maximum number of slots in the ovector have
780        been set.
781    
782    3.  Related to (2) above: when there are more back references in a pattern than
783        slots in the output vector, pcre_exec() uses temporary memory during
784        matching, and copies in the captures as far as possible afterwards. It was
785        using the entire output vector, but this conflicts with the specification
786        that only 2/3 is used for passing back captured substrings. Now it uses
787        only the first 2/3, for compatibility. This is, of course, another edge
788        case.
789    
790    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
791        main code base, and can be used by building with --enable-jit. When this is
792        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
793        runtime --no-jit option is given.
794    
795    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
796        ovector, the return from the function was zero, implying that there were
797        other matches that did not fit. The correct "exactly full" value is now
798        returned.
799    
800    6.  If a subpattern that was called recursively or as a subroutine contained
801        (*PRUNE) or any other control that caused it to give a non-standard return,
802        invalid errors such as "Error -26 (nested recursion at the same subject
803        position)" or even infinite loops could occur.
804    
805    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
806        computing the minimum length on reaching *ACCEPT, and so ended up with the
807        wrong value of 1 rather than 0. Further investigation indicates that
808        computing a minimum subject length in the presence of *ACCEPT is difficult
809        (think back references, subroutine calls), and so I have changed the code
810        so that no minimum is registered for a pattern that contains *ACCEPT.
811    
812    8.  If (*THEN) was present in the first (true) branch of a conditional group,
813        it was not handled as intended. [But see 16 below.]
814    
815    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
816        Sheri Pierce.
817    
818    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
819        the first byte in a match must be "a".
820    
821    11. Change 17 for 8.13 increased the recursion depth for patterns like
822        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
823        pattern contains any instances of (*THEN). If it does not, the old
824        optimizations are restored. It would be nice to do this on a per-group
825        basis, but at the moment that is not feasible.
826    
827    12. In some environments, the output of pcretest -C is CRLF terminated. This
828        broke RunTest's code that checks for the link size. A single white space
829        character after the value is now allowed for.
830    
831    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
832        For "fr", it uses the Windows-specific input and output files.
833    
834    14. If (*THEN) appeared in a group that was called recursively or as a
835        subroutine, it did not work as intended. [But see next item.]
836    
837    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
838        pattern fragments (but not containing any | characters). If A and B are
839        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
840        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
841        D. In other words, Perl considers parentheses that do not contain any |
842        characters to be part of a surrounding alternative, whereas PCRE was
843        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
844        differently. PCRE now behaves in the same way as Perl, except in the case
845        of subroutine/recursion calls such as (?1) which have in any case always
846        been different (but PCRE had them first :-).
847    
848    16. Related to 15 above: Perl does not treat the | in a conditional group as
849        creating alternatives. Such a group is treated in the same way as an
850        ordinary group without any | characters when processing (*THEN). PCRE has
851        been changed to match Perl's behaviour.
852    
853    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
854        RunGrepTest script failed.
855    
856    18. Change 22 for version 13 caused atomic groups to use more stack. This is
857        inevitable for groups that contain captures, but it can lead to a lot of
858        stack use in large patterns. The old behaviour has been restored for atomic
859        groups that do not contain any capturing parentheses.
860    
861    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
862        suppress the check for a minimum subject length at run time. (If it was
863        given to pcre_exec() or pcre_dfa_exec() it did work.)
864    
865    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
866        fail to work when decoding hex characters in data strings in EBCDIC
867        environments.
868    
869    21. It appears that in at least one Mac OS environment, the isxdigit() function
870        is implemented as a macro that evaluates to its argument more than once,
871        contravening the C 90 Standard (I haven't checked a later standard). There
872        was an instance in pcretest which caused it to go wrong when processing
873        \x{...} escapes in subject strings. The has been rewritten to avoid using
874        things like p++ in the argument of isxdigit().
875    
876    
877    Version 8.13 16-Aug-2011
878    ------------------------
879    
880    1.  The Unicode data tables have been updated to Unicode 6.0.0.
881    
882    2.  Two minor typos in pcre_internal.h have been fixed.
883    
884    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
885        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
886        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
887    
888    4.  There were a number of related bugs in the code for matching backrefences
889        caselessly in UTF-8 mode when codes for the characters concerned were
890        different numbers of bytes. For example, U+023A and U+2C65 are an upper
891        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
892        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
893        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
894        2-byte code at the end of the subject (it thought there wasn't enough data
895        left).
896    
897    5.  Comprehensive information about what went wrong is now returned by
898        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
899        as the output vector has at least 2 elements. The offset of the start of
900        the failing character and a reason code are placed in the vector.
901    
902    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
903        now returned is for the first byte of the failing character, instead of the
904        last byte inspected. This is an incompatible change, but I hope it is small
905        enough not to be a problem. It makes the returned offset consistent with
906        pcre_exec() and pcre_dfa_exec().
907    
908    7.  pcretest now gives a text phrase as well as the error number when
909        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
910        failure, the offset and reason code are output.
911    
912    8.  When \R was used with a maximizing quantifier it failed to skip backwards
913        over a \r\n pair if the subsequent match failed. Instead, it just skipped
914        back over a single character (\n). This seems wrong (because it treated the
915        two characters as a single entity when going forwards), conflicts with the
916        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
917        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
918        has been changed.
919    
920    9.  Some internal refactoring has changed the processing so that the handling
921        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
922        time (the PCRE_DOTALL option was changed this way some time ago: version
923        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
924        which was always a bit of a fudge. It also means that there is one less
925        argument for the match() function, which reduces its stack requirements
926        slightly. This change also fixes an incompatibility with Perl: the pattern
927        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
928    
929    10. More internal refactoring has drastically reduced the number of recursive
930        calls to match() for possessively repeated groups such as (abc)++ when
931        using pcre_exec().
932    
933    11. While implementing 10, a number of bugs in the handling of groups were
934        discovered and fixed:
935    
936        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
937        (a|)*(?1) gave a compile-time internal error.
938        ((a|)+)+  did not notice that the outer group could match an empty string.
939        (^a|^)+   was not marked as anchored.
940        (.*a|.*)+ was not marked as matching at start or after a newline.
941    
942    12. Yet more internal refactoring has removed another argument from the match()
943        function. Special calls to this function are now indicated by setting a
944        value in a variable in the "match data" data block.
945    
946    13. Be more explicit in pcre_study() instead of relying on "default" for
947        opcodes that mean there is no starting character; this means that when new
948        ones are added and accidentally left out of pcre_study(), testing should
949        pick them up.
950    
951    14. The -s option of pcretest has been documented for ages as being an old
952        synonym of -m (show memory usage). I have changed it to mean "force study
953        for every regex", that is, assume /S for every regex. This is similar to -i
954        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
955        using it. It makes it easier to run collections of tests with and without
956        study enabled, and thereby test pcre_study() more easily. All the standard
957        tests are now run with and without -s (but some patterns can be marked as
958        "never study" - see 20 below).
959    
960    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
961        restoration of the capturing data to the outer values was not happening
962        correctly.
963    
964    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
965        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
966        pattern had matched an empty string, and so incorrectly returned a no
967        match.
968    
969    17. There was optimizing code for the last branch of non-capturing parentheses,
970        and also for the obeyed branch of a conditional subexpression, which used
971        tail recursion to cut down on stack usage. Unfortunately, now that there is
972        the possibility of (*THEN) occurring in these branches, tail recursion is
973        no longer possible because the return has to be checked for (*THEN). These
974        two optimizations have therefore been removed. [But see 8.20/11 above.]
975    
976    18. If a pattern containing \R was studied, it was assumed that \R always
977        matched two bytes, thus causing the minimum subject length to be
978        incorrectly computed because \R can also match just one byte.
979    
980    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
981        was incorrectly computed.
982    
983    20. If /S is present twice on a test pattern in pcretest input, it now
984        *disables* studying, thereby overriding the use of -s on the command line
985        (see 14 above). This is necessary for one or two tests to keep the output
986        identical in both cases.
987    
988    21. When (*ACCEPT) was used in an assertion that matched an empty string and
989        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
990    
991    22. When an atomic group that contained a capturing parenthesis was
992        successfully matched, but the branch in which it appeared failed, the
993        capturing was not being forgotten if a higher numbered group was later
994        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
995        group 1 to "a", when in fact it should be unset. This applied to multi-
996        branched capturing and non-capturing groups, repeated or not, and also to
997        positive assertions (capturing in negative assertions does not happen
998        in PCRE) and also to nested atomic groups.
999    
1000    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1001        subject after a captured substring, to make it easier to tell which of a
1002        number of identical substrings has been captured.
1003    
1004    24. The way atomic groups are processed by pcre_exec() has been changed so that
1005        if they are repeated, backtracking one repetition now resets captured
1006        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1007        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1008        "aaa". Previously, it would have been "a". As part of this code
1009        refactoring, the way recursive calls are handled has also been changed.
1010    
1011    25. If an assertion condition captured any substrings, they were not passed
1012        back unless some other capturing happened later. For example, if
1013        (?(?=(a))a) was matched against "a", no capturing was returned.
1014    
1015    26. When studying a pattern that contained subroutine calls or assertions,
1016        the code for finding the minimum length of a possible match was handling
1017        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1018        group 1 called group 2 while simultaneously a separate group 2 called group
1019        1). A stack overflow occurred in this case. I have fixed this by limiting
1020        the recursion depth to 10.
1021    
1022    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1023        Fortmann. This supports explicit test numbers on the command line, and has
1024        argument validation and error reporting.
1025    
1026    28. An instance of \X with an unlimited repeat could fail if at any point the
1027        first character it looked at was a mark character.
1028    
1029    29. Some minor code refactoring concerning Unicode properties and scripts
1030        should reduce the stack requirement of match() slightly.
1031    
1032    30. Added the '=' option to pcretest to check the setting of unused capturing
1033        slots at the end of the pattern, which are documented as being -1, but are
1034        not included in the return count.
1035    
1036    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1037        compiled something random. Now it gives a compile-time error (as does
1038        Perl).
1039    
1040    32. A *MARK encountered during the processing of a positive assertion is now
1041        recorded and passed back (compatible with Perl).
1042    
1043    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1044        had alternative anchored branches, the search for a second match in a line
1045        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1046        matched the line "0102" twice. The same bug affected patterns that started
1047        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1048        twice.
1049    
1050    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1051        does, and because of capturing effects, quantifying parenthesized
1052        assertions may at times be useful. Quantifiers are now allowed for
1053        parenthesized assertions.
1054    
1055    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1056    
1057    36. \g was being checked for fancy things in a character class, when it should
1058        just be a literal "g".
1059    
1060    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1061        appearance of a nested POSIX class supersedes an apparent external class.
1062        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1063        unescaped square brackets may also appear as part of class names. For
1064        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1065        more like Perl. (But see 8.20/1 above.)
1066    
1067    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1068        was because it thought it was \N{name}, which is not supported).
1069    
1070    39. Add minix to OS list not supporting the -S option in pcretest.
1071    
1072    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1073        cannot analyze patterns in sufficient detail to catch mutual recursions
1074        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1075        subgroup is called recursively as a subpattern for a second time at the
1076        same position in the subject string. In previous releases this might have
1077        been caught by the recursion limit, or it might have run out of stack.
1078    
1079    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1080        happen only once. PCRE was, however incorrectly giving a compile time error
1081        "recursive call could loop indefinitely" because it cannot analyze the
1082        pattern in sufficient detail. The compile time test no longer happens when
1083        PCRE is compiling a conditional subpattern, but actual runaway loops are
1084        now caught at runtime (see 40 above).
1085    
1086    42. It seems that Perl allows any characters other than a closing parenthesis
1087        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1088        has been changed to be the same.
1089    
1090    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1091        as not to get warnings when autogen.sh is called. Also changed
1092        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1093    
1094    44. To help people who use pcregrep to scan files containing exceedingly long
1095        lines, the following changes have been made:
1096    
1097        (a) The default value of the buffer size parameter has been increased from
1098            8K to 20K. (The actual buffer used is three times this size.)
1099    
1100        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1101            PCRE is built.
1102    
1103        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1104            to be set at run time.
1105    
1106        (d) Numerical values in pcregrep options can be followed by K or M, for
1107            example --buffer-size=50K.
1108    
1109        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1110            given and the return code is set to 2.
1111    
1112    45. Add a pointer to the latest mark to the callout data block.
1113    
1114    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1115        partial match of an empty string instead of no match. This was specific to
1116        the use of ".".
1117    
1118    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1119        complete match instead of a partial match. This bug was dependent on both
1120        the PCRE_UTF8 and PCRE_DOTALL options being set.
1121    
1122    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1123        starting byte set, because \b was not being ignored.
1124    
1125    
1126    Version 8.12 15-Jan-2011
1127    ------------------------
1128    
1129    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1130        checks for such things as part of the documentation building process.
1131    
1132    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1133        --match-limit and --recursion-limit options (added for 8.11). In
1134        particular, this made one of the standard tests fail. (The integer value
1135        went into the wrong half of a long int.)
1136    
1137    3.  If the --colour option was given to pcregrep with -v (invert match), it
1138        did strange things, either producing crazy output, or crashing. It should,
1139        of course, ignore a request for colour when reporting lines that do not
1140        match.
1141    
1142    4.  Another pcregrep bug caused similar problems if --colour was specified with
1143        -M (multiline) and the pattern match finished with a line ending.
1144    
1145    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1146        matched in multiline mode, the following line was shown as part of the
1147        match. This seems wrong, so I have changed it.
1148    
1149    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1150        the check for further matches in the same line (so they could be coloured)
1151        to overrun the end of the current line. If another match was found, it was
1152        incorrectly shown (and then shown again when found in the next line).
1153    
1154    7.  If pcregrep was compiled under Windows, there was a reference to the
1155        function pcregrep_exit() before it was defined. I am assuming this was
1156        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1157        reported by a user. I've moved the definition above the reference.
1158    
1159    
1160    Version 8.11 10-Dec-2010
1161    ------------------------
1162    
1163    1.  (*THEN) was not working properly if there were untried alternatives prior
1164        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1165        backtracked to try for "b" instead of moving to the next alternative branch
1166        at the same level (in this case, to look for "c"). The Perl documentation
1167        is clear that when (*THEN) is backtracked onto, it goes to the "next
1168        alternative in the innermost enclosing group".
1169    
1170    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1171        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1172        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1173        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1174        (*THEN).
1175    
1176    3.  If \s appeared in a character class, it removed the VT character from
1177        the class, even if it had been included by some previous item, for example
1178        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1179        of \s, but is part of the POSIX "space" class.)
1180    
1181    4.  A partial match never returns an empty string (because you can always
1182        match an empty string at the end of the subject); however the checking for
1183        an empty string was starting at the "start of match" point. This has been
1184        changed to the "earliest inspected character" point, because the returned
1185        data for a partial match starts at this character. This means that, for
1186        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1187        (previously it gave "no match").
1188    
1189    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1190        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1191        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1192        has an implication that the given string is incomplete (because a partial
1193        match is preferred over a full match). For this reason, these items now
1194        give a partial match in this situation. [Aside: previously, the one case
1195        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1196        match rather than a full match, which was wrong by the old rules, but is
1197        now correct.]
1198    
1199    6.  There was a bug in the handling of #-introduced comments, recognized when
1200        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1201        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1202        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1203        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1204        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1205        places in pcre_compile().
1206    
1207    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1208        comments when looking ahead for named forward references to subpatterns,
1209        the only newline sequence it recognized was NL. It now handles newlines
1210        according to the set newline convention.
1211    
1212    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1213        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1214        cater for a lack of strerror(). These oversights have been fixed.
1215    
1216    9.  Added --match-limit and --recursion-limit to pcregrep.
1217    
1218    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1219    
1220    11. When the -o option was used, pcregrep was setting a return code of 1, even
1221        when matches were found, and --line-buffered was not being honoured.
1222    
1223    12. Added an optional parentheses number to the -o and --only-matching options
1224        of pcregrep.
1225    
1226    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1227        can match an empty string. The code to do it in pcretest and pcredemo
1228        needed fixing:
1229    
1230        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1231            only one byte after an empty string match just before CRLF (this case
1232            just got forgotten; "any" and "anycrlf" were OK).
1233    
1234        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1235            mode when an empty string match preceded an ASCII character followed by
1236            a non-ASCII character. (The code for advancing by one character rather
1237            than one byte was nonsense.)
1238    
1239        (c) The pcredemo.c sample program did not have any code at all to handle
1240            the cases when CRLF is a valid newline sequence.
1241    
1242    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1243        as a starting offset was within the subject string. There is now a new
1244        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1245        negative or greater than the length of the string. In order to test this,
1246        pcretest is extended to allow the setting of negative starting offsets.
1247    
1248    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1249        starting offset points to the beginning of a UTF-8 character was
1250        unnecessarily clumsy. I tidied it up.
1251    
1252    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1253        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1254    
1255    17. Nobody had reported that the --include_dir option, which was added in
1256        release 7.7 should have been called --include-dir (hyphen, not underscore)
1257        for compatibility with GNU grep. I have changed it to --include-dir, but
1258        left --include_dir as an undocumented synonym, and the same for
1259        --exclude-dir, though that is not available in GNU grep, at least as of
1260        release 2.5.4.
1261    
1262    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1263        characters from a string of bytes) have been redefined so as not to use
1264        loops, in order to improve performance in some environments. At the same
1265        time, I abstracted some of the common code into auxiliary macros to save
1266        repetition (this should not affect the compiled code).
1267    
1268    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1269        compile-time error is now given if \c is not followed by an ASCII
1270        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1271        different, and any byte value is allowed.)
1272    
1273    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1274        START_OPTIMIZE option, which is now allowed at compile time - but just
1275        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1276        to pcregrep and other applications that have no direct access to PCRE
1277        options. The new /Y option in pcretest sets this option when calling
1278        pcre_compile().
1279    
1280    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1281        back references. Groups containing recursive back references were forced to
1282        be atomic by that change, but in the case of named groups, the amount of
1283        memory required was incorrectly computed, leading to "Failed: internal
1284        error: code overflow". This has been fixed.
1285    
1286    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1287        pcretest.c, to avoid build problems in some Borland environments.
1288    
1289    
1290    Version 8.10 25-Jun-2010
1291    ------------------------
1292    
1293    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1294        THEN.
1295    
1296    2.  (*ACCEPT) was not working when inside an atomic group.
1297    
1298    3.  Inside a character class, \B is treated as a literal by default, but
1299        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1300        causes the error). The code is unchanged, but I tidied the documentation.
1301    
1302    4.  Inside a character class, PCRE always treated \R and \X as literals,
1303        whereas Perl faults them if its -w option is set. I have changed PCRE so
1304        that it faults them when PCRE_EXTRA is set.
1305    
1306    5.  Added support for \N, which always matches any character other than
1307        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1308    
1309    6.  When compiling pcregrep with newer versions of gcc which may have
1310        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1311        declared with attribute warn_unused_result" were given. Just casting the
1312        result to (void) does not stop the warnings; a more elaborate fudge is
1313        needed. I've used a macro to implement this.
1314    
1315    7.  Minor change to pcretest.c to avoid a compiler warning.
1316    
1317    8.  Added four artifical Unicode properties to help with an option to make
1318        \s etc use properties (see next item). The new properties are: Xan
1319        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1320    
1321    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1322        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1323        this option. Modified pcretest to add /W to test this facility. Added
1324        REG_UCP to make it available via the POSIX interface.
1325    
1326    10. Added --line-buffered to pcregrep.
1327    
1328    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1329        studied, and the match started with a letter with a code point greater than
1330        127 whose first byte was different to the first byte of the other case of
1331        the letter, the other case of this starting letter was not recognized
1332        (#976).
1333    
1334    12. If a pattern that was studied started with a repeated Unicode property
1335        test, for example, \p{Nd}+, there was the theoretical possibility of
1336        setting up an incorrect bitmap of starting bytes, but fortunately it could
1337        not have actually happened in practice until change 8 above was made (it
1338        added property types that matched character-matching opcodes).
1339    
1340    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1341        possible starting bytes for non-anchored patterns.
1342    
1343    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1344        \R, and also a number of cases that involve Unicode properties, both
1345        explicit and implicit when PCRE_UCP is set.
1346    
1347    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1348        input, it could crash or give wrong results if characters with values
1349        greater than 0xc0 were present in the subject string. (Detail: it assumed
1350        UTF-8 input when processing these items.)
1351    
1352    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1353        size_t is 64-bit (#991).
1354    
1355    17. Added a check for running out of memory when PCRE is compiled with
1356        --disable-stack-for-recursion (#990).
1357    
1358    18. If the last data line in a file for pcretest does not have a newline on
1359        the end, a newline was missing in the output.
1360    
1361    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1362        less than 128) in its various bitmaps. However, there is a facility for
1363        generating tables according to the current locale when PCRE is compiled. It
1364        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1365        characters, are recognized by isspace() and therefore were getting set in
1366        these tables, and indeed these tables seem to approximate to ISO 8859. This
1367        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1368        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1369        which of course cannot start UTF-8 characters. I have changed the code so
1370        that only real ASCII characters (less than 128) and the correct starting
1371        bytes for UTF-8 encodings are set for characters greater than 127 when in
1372        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1373        altogether.)
1374    
1375    20. Added the /T option to pcretest so as to be able to run tests with non-
1376        standard character tables, thus making it possible to include the tests
1377        used for 19 above in the standard set of tests.
1378    
1379    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1380        reference to a subpattern the other side of a comment that contains an
1381        opening parenthesis caused either an internal compiling error, or a
1382        reference to the wrong subpattern.
1383    
1384    
1385    Version 8.02 19-Mar-2010
1386    ------------------------
1387    
1388    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1389    
1390    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1391        configured.
1392    
1393    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1394        original author of that file, following a query about its status.
1395    
1396    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1397        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1398    
1399    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1400        quantifier applied to a forward-referencing subroutine call, could compile
1401        incorrect code or give the error "internal error: previously-checked
1402        referenced subpattern not found".
1403    
1404    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1405        variables to point to external functions. For these systems, therefore,
1406        pcre_malloc etc. are now initialized to local functions that call the
1407        relevant global functions.
1408    
1409    7.  There were two entries missing in the vectors called coptable and poptable
1410        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1411        I've fixed the data, and added a kludgy way of testing at compile time that
1412        the lengths are correct (equal to the number of opcodes).
1413    
1414    8.  Following on from 7, I added a similar kludge to check the length of the
1415        eint vector in pcreposix.c.
1416    
1417    9.  Error texts for pcre_compile() are held as one long string to avoid too
1418        much relocation at load time. To find a text, the string is searched,
1419        counting zeros. There was no check for running off the end of the string,
1420        which could happen if a new error number was added without updating the
1421        string.
1422    
1423    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1424    
1425    11. \K was not working if it appeared in an atomic group or in a group that
1426        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1427        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1428        the assertion is positive, but not if it is negative.
1429    
1430    12. Change 11 fortuitously reduced the size of the stack frame used in the
1431        "match()" function of pcre_exec.c by one pointer. Forthcoming
1432        implementation of support for (*MARK) will need an extra pointer on the
1433        stack; I have reserved it now, so that the stack frame size does not
1434        decrease.
1435    
1436    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1437        item in branch that calls a recursion is a subroutine call - as in the
1438        second branch in the above example - was incorrectly given the compile-
1439        time error "recursive call could loop indefinitely" because pcre_compile()
1440        was not correctly checking the subroutine for matching a non-empty string.
1441    
1442    14. The checks for overrunning compiling workspace could trigger after an
1443        overrun had occurred. This is a "should never occur" error, but it can be
1444        triggered by pathological patterns such as hundreds of nested parentheses.
1445        The checks now trigger 100 bytes before the end of the workspace.
1446    
1447    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1448    
1449    
1450    Version 8.01 19-Jan-2010
1451    ------------------------
1452    
1453    1.  If a pattern contained a conditional subpattern with only one branch (in
1454        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1455        computed the wrong minimum data length (which is of course zero for such
1456        subpatterns). This could cause incorrect "no match" results.
1457    
1458    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1459        the pattern is reset in the first branch, pcre_compile() failed with
1460        "internal error: code overflow at offset...". This happened only when
1461        the reset was to the original external option setting. (An optimization
1462        abstracts leading options settings into an external setting, which was the
1463        cause of this.)
1464    
1465    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1466        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1467        assertion pattern did not match (meaning that the assertion was true), it
1468        was incorrectly treated as false if the SKIP had been reached during the
1469        matching. This also applied to assertions used as conditions.
1470    
1471    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1472        assertion subpattern, including such a pattern used as a condition,
1473        unpredictable results occurred, instead of the error return
1474        PCRE_ERROR_DFA_UITEM.
1475    
1476    5.  The C++ GlobalReplace function was not working like Perl for the special
1477        situation when an empty string is matched. It now does the fancy magic
1478        stuff that is necessary.
1479    
1480    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1481        removed. (These were left over from very, very early versions of PCRE.)
1482    
1483    7.  Some cosmetic changes to the code to make life easier when compiling it
1484        as part of something else:
1485    
1486        (a) Change DEBUG to PCRE_DEBUG.
1487    
1488        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1489            called "current" as "current_branch", to prevent a collision with the
1490            Linux macro when compiled as a kernel module.
1491    
1492        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1493            prevent a collision with the Linux macro when compiled as a kernel
1494            module.
1495    
1496    8.  In pcre_compile() there are some checks for integer overflows that used to
1497        cast potentially large values to (double). This has been changed to that
1498        when building, a check for int64_t is made, and if it is found, it is used
1499        instead, thus avoiding the use of floating point arithmetic. (There is no
1500        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1501        double.
1502    
1503    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1504        2005 (difference between two addresses compared to an unsigned value).
1505    
1506    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1507        custom one, because of the following reported problem in Windows:
1508    
1509          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1510              under Win32.
1511          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1512              therefore missing the function definition.
1513          - The compiler thus generates a "C" signature for the test function.
1514          - The linker fails to find the "C" function.
1515          - PCRE fails to configure if asked to do so against libbz2.
1516    
1517    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1518        messages were output:
1519    
1520          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1521          rerunning libtoolize, to keep the correct libtool macros in-tree.
1522          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1523    
1524        I have done both of these things.
1525    
1526    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1527        most of the time, it *can* run out if it is given a pattern that contains a
1528        runaway infinite recursion. I updated the discussion in the pcrestack man
1529        page.
1530    
1531    13. Now that we have gone to the x.xx style of version numbers, the minor
1532        version may start with zero. Using 08 or 09 is a bad idea because users
1533        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1534        interpreted as invalid octal numbers. I've updated the previous comment in
1535        configure.ac, and also added a check that gives an error if 08 or 09 are
1536        used.
1537    
1538    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1539        causing partial matching to fail when the end of the subject matched \W
1540        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1541    
1542    15. There were some discrepancies between the declarations in pcre_internal.h
1543        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1544        their definitions. The declarations used "const uschar *" and the
1545        definitions used USPTR. Even though USPTR is normally defined as "const
1546        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1547        reported that: "This difference in casting confuses some C++ compilers, for
1548        example, SunCC recognizes above declarations as different functions and
1549        generates broken code for hbpcre." I have changed the declarations to use
1550        USPTR.
1551    
1552    16. GNU libtool is named differently on some systems. The autogen.sh script now
1553        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1554        (FreeBSD).
1555    
1556    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1557        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1558        comment: "Figure out how to create a longlong from a string: strtoll and
1559        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1560        instance, but it only takes 2 args instead of 3!"
1561    
1562    18. A subtle bug concerned with back references has been fixed by a change of
1563        specification, with a corresponding code fix. A pattern such as
1564        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1565        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1566        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1567        same bug. Such groups have to be quantified to be useful, or contained
1568        inside another quantified group. (If there's no repetition, the reference
1569        can never match.) The problem arises because, having left the group and
1570        moved on to the rest of the pattern, a later failure that backtracks into
1571        the group uses the captured value from the final iteration of the group
1572        rather than the correct earlier one. I have fixed this in PCRE by forcing
1573        any group that contains a reference to itself to be an atomic group; that
1574        is, there cannot be any backtracking into it once it has completed. This is
1575        similar to recursive and subroutine calls.
1576    
1577    
1578    Version 8.00 19-Oct-09
1579    ----------------------
1580    
1581    1.  The table for translating pcre_compile() error codes into POSIX error codes
1582        was out-of-date, and there was no check on the pcre_compile() error code
1583        being within the table. This could lead to an OK return being given in
1584        error.
1585    
1586    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1587        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1588        in a Windows environment.
1589    
1590    3.  The pcregrep --count option prints the count for each file even when it is
1591        zero, as does GNU grep. However, pcregrep was also printing all files when
1592        --files-with-matches was added. Now, when both options are given, it prints
1593        counts only for those files that have at least one match. (GNU grep just
1594        prints the file name in this circumstance, but including the count seems
1595        more useful - otherwise, why use --count?) Also ensured that the
1596        combination -clh just lists non-zero counts, with no names.
1597    
1598    4.  The long form of the pcregrep -F option was incorrectly implemented as
1599        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1600        but it seems right to fix it, and I didn't think it was worth preserving
1601        the old behaviour.
1602    
1603    5.  The command line items --regex=pattern and --regexp=pattern were not
1604        recognized by pcregrep, which required --regex pattern or --regexp pattern
1605        (with a space rather than an '='). The man page documented the '=' forms,
1606        which are compatible with GNU grep; these now work.
1607    
1608    6.  No libpcreposix.pc file was created for pkg-config; there was just
1609        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1610    
1611    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1612        when UCP support is not needed, by modifying the Python script that
1613        generates it from Unicode data files. This should not matter if the module
1614        is correctly used as a library, but I received one complaint about 50K of
1615        unwanted data. My guess is that the person linked everything into his
1616        program rather than using a library. Anyway, it does no harm.
1617    
1618    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1619        was a minimum greater than 1 for a wide character in a possessive
1620        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1621        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1622        character. Chaos in the form of incorrect output or a compiling loop could
1623        result.
1624    
1625    9.  The restrictions on what a pattern can contain when partial matching is
1626        requested for pcre_exec() have been removed. All patterns can now be
1627        partially matched by this function. In addition, if there are at least two
1628        slots in the offset vector, the offset of the earliest inspected character
1629        for the match and the offset of the end of the subject are set in them when
1630        PCRE_ERROR_PARTIAL is returned.
1631    
1632    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1633        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1634        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1635        and may be more useful for multi-segment matching.
1636    
1637    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1638        used to be given if ever the end of the subject was reached; now it is
1639        given only if matching could not proceed because another character was
1640        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1641        string "Z", which now yields "no match" instead of "partial match". In the
1642        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1643        final character ended with (*FAIL).
1644    
1645    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1646        if the pattern had a "must contain" character that was already found in the
1647        earlier partial match, unless partial matching was again requested. For
1648        example, with the pattern /dog.(body)?/, the "must contain" character is
1649        "g". If the first part-match was for the string "dog", restarting with
1650        "sbody" failed. This bug has been fixed.
1651    
1652    13. The string returned by pcre_dfa_exec() after a partial match has been
1653        changed so that it starts at the first inspected character rather than the
1654        first character of the match. This makes a difference only if the pattern
1655        starts with a lookbehind assertion or \b or \B (\K is not supported by
1656        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1657        matching functions compatible, and I think it's the right thing to do.
1658    
1659    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1660        so that the demonstration program is easily available in environments where
1661        PCRE has not been installed from source.
1662    
1663    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1664        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1665        library.
1666    
1667    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1668        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1669        is not the first non-POSIX option to be added. Clearly some people find
1670        these options useful.
1671    
1672    17. If a caller to the POSIX matching function regexec() passes a non-zero
1673        value for nmatch with a NULL value for pmatch, the value of
1674        nmatch is forced to zero.
1675    
1676    18. RunGrepTest did not have a test for the availability of the -u option of
1677        the diff command, as RunTest does. It now checks in the same way as
1678        RunTest, and also checks for the -b option.
1679    
1680    19. If an odd number of negated classes containing just a single character
1681        interposed, within parentheses, between a forward reference to a named
1682        subpattern and the definition of the subpattern, compilation crashed with
1683        an internal error, complaining that it could not find the referenced
1684        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1685        [The bug was that it was starting one character too far in when skipping
1686        over the character class, thus treating the ] as data rather than
1687        terminating the class. This meant it could skip too much.]
1688    
1689    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1690        /g option in pcretest when the pattern contains \K, which makes it possible
1691        to have an empty string match not at the start, even when the pattern is
1692        anchored. Updated pcretest and pcredemo to use this option.
1693    
1694    21. If the maximum number of capturing subpatterns in a recursion was greater
1695        than the maximum at the outer level, the higher number was returned, but
1696        with unset values at the outer level. The correct (outer level) value is
1697        now given.
1698    
1699    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1700        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1701        make it do so. The string so far is captured, making this feature
1702        compatible with Perl.
1703    
1704    23. The tests have been re-organized, adding tests 11 and 12, to make it
1705        possible to check the Perl 5.10 features against Perl 5.10.
1706    
1707    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1708        pattern matches a fixed length string. PCRE did not allow this; now it
1709        does. Neither allows recursion.
1710    
1711    25. I finally figured out how to implement a request to provide the minimum
1712        length of subject string that was needed in order to match a given pattern.
1713        (It was back references and recursion that I had previously got hung up
1714        on.) This code has now been added to pcre_study(); it finds a lower bound
1715        to the length of subject needed. It is not necessarily the greatest lower
1716        bound, but using it to avoid searching strings that are too short does give
1717        some useful speed-ups. The value is available to calling programs via
1718        pcre_fullinfo().
1719    
1720    26. While implementing 25, I discovered to my embarrassment that pcretest had
1721        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1722        study optimizations had never been tested with that matching function.
1723        Oops. What is worse, even when it was passed study data, there was a bug in
1724        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1725        were also very few tests of studied patterns with pcre_dfa_exec().
1726    
1727    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1728        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1729        on the other side of the coin, they are no longer allowed to have different
1730        names, because these cannot be distinguished in PCRE, and this has caused
1731        confusion. (This is a difference from Perl.)
1732    
1733    28. When duplicate subpattern names are present (necessarily with different
1734        numbers, as required by 27 above), and a test is made by name in a
1735        conditional pattern, either for a subpattern having been matched, or for
1736        recursion in such a pattern, all the associated numbered subpatterns are
1737        tested, and the overall condition is true if the condition is true for any
1738        one of them. This is the way Perl works, and is also more like the way
1739        testing by number works.
1740    
1741    
1742    Version 7.9 11-Apr-09
1743    ---------------------
1744    
1745    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1746        (pcretest), all targets were linked against these libraries. This included
1747        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1748        libraries. This caused unwanted dependencies to be created. This problem
1749        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1750        pcretest is linked with readline.
1751    
1752    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1753        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1754        moved outside it again, because FALSE and TRUE are already defined in AIX,
1755        but BOOL is not.
1756    
1757    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1758        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1759    
1760    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1761        hyphens) following file names and line numbers when outputting matching
1762        lines. This is not true; no spaces are inserted. I have also clarified the
1763        wording for the --colour (or --color) option.
1764    
1765    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1766        was not coloured; this is different to GNU grep, so I have changed it to be
1767        the same.
1768    
1769    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1770        each matching line was coloured. Now it goes on to look for further matches
1771        of any of the test patterns, which is the same behaviour as GNU grep.
1772    
1773    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1774        doesn't make sense to accept an empty string match in pcregrep, so I have
1775        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1776        seems to be how GNU grep behaves. [But see later change 40 for release
1777        8.33.]
1778    
1779    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1780        start or after a newline", because the conditional assertion was not being
1781        correctly handled. The rule now is that both the assertion and what follows
1782        in the first alternative must satisfy the test.
1783    
1784    9.  If auto-callout was enabled in a pattern with a conditional group whose
1785        condition was an assertion, PCRE could crash during matching, both with
1786        pcre_exec() and pcre_dfa_exec().
1787    
1788    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1789        used for matching.
1790    
1791    11. Unicode property support in character classes was not working for
1792        characters (bytes) greater than 127 when not in UTF-8 mode.
1793    
1794    12. Added the -M command line option to pcretest.
1795    
1796    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1797    
1798    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1799    
1800    16. Added comments and documentation about mis-use of no_arg in the C++
1801        wrapper.
1802    
1803    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1804        from Martin Jerabek that uses macro names for all relevant character and
1805        string constants.
1806    
1807    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1808        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1809        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1810        these, but not everybody uses configure.
1811    
1812    19. A conditional group that had only one branch was not being correctly
1813        recognized as an item that could match an empty string. This meant that an
1814        enclosing group might also not be so recognized, causing infinite looping
1815        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1816        with the subject "ab", where knowledge that the repeated group can match
1817        nothing is needed in order to break the loop.
1818    
1819    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1820        exec(), but without supplying a callout function, matching went wrong.
1821    
1822    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1823        leak if the size of the offset vector was greater than 30. When the vector
1824        is smaller, the saved offsets during recursion go onto a local stack
1825        vector, but for larger vectors malloc() is used. It was failing to free
1826        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1827        error, in fact).
1828    
1829    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1830        heapframe that is used only when UTF-8 support is enabled. This caused no
1831        problem, but was untidy.
1832    
1833    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1834        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1835        included within another project.
1836    
1837    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1838        slightly modified by me:
1839    
1840          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1841              not building pcregrep.
1842    
1843          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1844              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1845    
1846    25. Forward references, both numeric and by name, in patterns that made use of
1847        duplicate group numbers, could behave incorrectly or give incorrect errors,
1848        because when scanning forward to find the reference group, PCRE was not
1849        taking into account the duplicate group numbers. A pattern such as
1850        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1851    
1852    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1853        the feature of a custom pointer more persuasive (as requested by a user).
1854    
1855    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1856        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1857        pre-defined.
1858    
1859    28. Added support for (*UTF8) at the start of a pattern.
1860    
1861    29. Arrange for flags added by the "release type" setting in CMake to be shown
1862        in the configuration summary.
1863    
1864    
1865    Version 7.8 05-Sep-08
1866    ---------------------
1867    
1868    1.  Replaced UCP searching code with optimized version as implemented for Ad
1869        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1870        stage table and inline lookup instead of a function, giving speed ups of 2
1871        to 5 times on some simple patterns that I tested. Permission was given to
1872        distribute the MultiStage2.py script that generates the tables (it's not in
1873        the tarball, but is in the Subversion repository).
1874    
1875    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1876        scripts.
1877    
1878    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1879        a group with a zero qualifier. The result of the study could be incorrect,
1880        or the function might crash, depending on the pattern.
1881    
1882    4.  Caseless matching was not working for non-ASCII characters in back
1883        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1884        It now works when Unicode Property Support is available.
1885    
1886    5.  In pcretest, an escape such as \x{de} in the data was always generating
1887        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1888        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1889        truncation.
1890    
1891    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1892    
1893    7.  Added two (int) casts to pcregrep when printing the difference of two
1894        pointers, in case they are 64-bit values.
1895    
1896    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1897        test 2 if it fails.
1898    
1899    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1900        and a #define of that name to empty if it is not externally set. This is to
1901        allow users of MSVC to set it if necessary.
1902    
1903    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1904        the convenience functions in the pcre_get.c source file.
1905    
1906    11. An option change at the start of a pattern that had top-level alternatives
1907        could cause overwriting and/or a crash. This command provoked a crash in
1908        some environments:
1909    
1910          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1911    
1912        This potential security problem was recorded as CVE-2008-2371.
1913    
1914    12. For a pattern where the match had to start at the beginning or immediately
1915        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1916        pcre_dfa_exec() could read past the end of the passed subject if there was
1917        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1918        pcretest so that it places the subject at the end of its malloc-ed buffer.
1919    
1920    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1921        exec() might read past the end of the data buffer in UTF-8 mode.
1922    
1923    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1924        the data contained the byte 0x85 as part of a UTF-8 character within its
1925        first line. This applied both to normal and DFA matching.
1926    
1927    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1928        /^[^d]*?$/8 failed to match "abc".
1929    
1930    16. Added a missing copyright notice to pcrecpp_internal.h.
1931    
1932    17. Make it more clear in the documentation that values returned from
1933        pcre_exec() in ovector are byte offsets, not character counts.
1934    
1935    18. Tidied a few places to stop certain compilers from issuing warnings.
1936    
1937    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1938        supplied by Stefan Weber. I made a further small update for 7.8 because
1939        there is a change of source arrangements: the pcre_searchfuncs.c module is
1940        replaced by pcre_ucd.c.
1941    
1942    
1943    Version 7.7 07-May-08
1944    ---------------------
1945    
1946    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1947        a string to a long long, pretend we don't even have a long long." This is
1948        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1949    
1950    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1951        pre-7.6 versions, which defined a global no_arg variable instead of putting
1952        it in the RE class. (See also #8 below.)
1953    
1954    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1955        Lopes.
1956    
1957    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1958    
1959        (1) The include/exclude patterns were being applied to the whole pathnames
1960            of files, instead of just to the final components.
1961    
1962        (2) If there was more than one level of directory, the subdirectories were
1963            skipped unless they satisfied the include/exclude conditions. This is
1964            inconsistent with GNU grep (and could even be seen as contrary to the
1965            pcregrep specification - which I improved to make it absolutely clear).
1966            The action now is always to scan all levels of directory, and just
1967            apply the include/exclude patterns to regular files.
1968    
1969    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1970        --exclude_dir in the tests to avoid scanning .svn directories.
1971    
1972    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1973        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1974        doesn't support NULs in patterns.
1975    
1976    7.  Added some missing "const"s to declarations of static tables in
1977        pcre_compile.c and pcre_dfa_exec.c.
1978    
1979    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1980        caused by fix #2  above. (Subsequently also a second patch to fix the
1981        first patch. And a third patch - this was a messy problem.)
1982    
1983    9.  Applied Craig's patch to remove the use of push_back().
1984    
1985    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1986        matching function regexec().
1987    
1988    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1989        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1990        references. PCRE supports relative numbers with this syntax (I don't think
1991        Oniguruma does).
1992    
1993    12. Previously, a group with a zero repeat such as (...){0} was completely
1994        omitted from the compiled regex. However, this means that if the group
1995        was called as a subroutine from elsewhere in the pattern, things went wrong
1996        (an internal error was given). Such groups are now left in the compiled
1997        pattern, with a new opcode that causes them to be skipped at execution
1998        time.
1999    
2000    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2001        to the way PCRE behaves:
2002    
2003        (a) A lone ] character is dis-allowed (Perl treats it as data).
2004    
2005        (b) A back reference to an unmatched subpattern matches an empty string
2006            (Perl fails the current match path).
2007    
2008        (c) A data ] in a character class must be notated as \] because if the
2009            first data character in a class is ], it defines an empty class. (In
2010            Perl it is not possible to have an empty class.) The empty class []
2011            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2012            The negative empty class [^] matches any one character, independently
2013            of the DOTALL setting.
2014    
2015    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2016        non-existent subpattern following a character class starting with ']' and
2017        containing () gave an internal compiling error instead of "reference to
2018        non-existent subpattern". Fortunately, when the pattern did exist, the
2019        compiled code was correct. (When scanning forwards to check for the
2020        existence of the subpattern, it was treating the data ']' as terminating
2021        the class, so got the count wrong. When actually compiling, the reference
2022        was subsequently set up correctly.)
2023    
2024    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2025        it was being rejected as not supported by pcre_dfa_exec(), even though
2026        other assertions are supported. I have made pcre_dfa_exec() support
2027        (*FAIL).
2028    
2029    16. The implementation of 13c above involved the invention of a new opcode,
2030        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2031        cannot be changed at match time, I realized I could make a small
2032        improvement to matching performance by compiling OP_ALLANY instead of
2033        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2034        on the OP_ANY path.
2035    
2036    17. Compiling pcretest on Windows with readline support failed without the
2037        following two fixes: (1) Make the unistd.h include conditional on
2038        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2039    
2040    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2041        ncurses library to be included for pcretest when ReadLine support is
2042        requested, but also to allow for it to be overridden. This patch came from
2043        Daniel Bergström.
2044    
2045    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2046        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2047        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2048        spotting this.
2049    
2050    
2051    Version 7.6 28-Jan-08
2052    ---------------------
2053    
2054    1.  A character class containing a very large number of characters with
2055        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2056        overflow.
2057    
2058    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2059        HAVE_LONG_LONG is not defined.
2060    
2061    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2062        bring them up to date and include new features. This patch includes:
2063    
2064        - Fixed PH's badly added libz and libbz2 support.
2065        - Fixed a problem with static linking.
2066        - Added pcredemo. [But later removed - see 7 below.]
2067        - Fixed dftables problem and added an option.
2068        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2069            HAVE_LONG_LONG.
2070        - Added readline support for pcretest.
2071        - Added an listing of the option settings after cmake has run.
2072    
2073    4.  A user submitted a patch to Makefile that makes it easy to create
2074        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2075        Makefile.am that cause it to include this special target, without
2076        affecting anything else. Note that the same mingw target plus all
2077        the other distribution libraries and programs are now supported
2078        when configuring with CMake (see 6 below) instead of with
2079        Configure/Make.
2080    
2081    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2082        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2083        exported in the Windows port". It has not yet been confirmed that the patch
2084        solves the problem, but it does no harm.
2085    
2086    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2087        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2088        with CMake, and also correct the comment about stack recursion.
2089    
2090    7.  Remove the automatic building of pcredemo from the ./configure system and
2091        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2092        of a program that users should build themselves after PCRE is installed, so
2093        building it automatically is not really right. What is more, it gave
2094        trouble in some build environments.
2095    
2096    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2097    
2098    
2099    Version 7.5 10-Jan-08
2100    ---------------------
2101    
2102    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2103        values in parens when parsing an RE using the C++ wrapper."
2104    
2105    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2106        Characters greater than 255 were excluded from the class instead of being
2107        included.
2108    
2109    3.  The same bug as (2) above applied to negated POSIX classes such as
2110        [:^space:].
2111    
2112    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2113        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2114        I have changed it.
2115    
2116    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2117        first named subpattern) and a construct such as (?&a) would reference the
2118        first named subpattern whose name started with "a" (in other words, the
2119        length check was missing). Both these problems are fixed. "Subpattern name
2120        expected" is now given for (?&) (a zero-length name), and this patch also
2121        makes it give the same error for \k'' (previously it complained that that
2122        was a reference to a non-existent subpattern).
2123    
2124    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2125        this is right because (?- can be followed by option settings as well as by
2126        digits. I have, however, made the messages clearer.
2127    
2128    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2129        than the number used in the conditional) now cause a compile-time error.
2130        This is actually not compatible with Perl, which accepts such patterns, but
2131        treats the conditional as always being FALSE (as PCRE used to), but it
2132        seems to me that giving a diagnostic is better.
2133    
2134    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2135        and messages.
2136    
2137    9.  Fix two occurrences of "backslash" in comments that should have been
2138        "backspace".
2139    
2140    10. Remove two redundant lines of code that can never be obeyed (their function
2141        was moved elsewhere).
2142    
2143    11. The program that makes PCRE's Unicode character property table had a bug
2144        which caused it to generate incorrect table entries for sequences of
2145        characters that have the same character type, but are in different scripts.
2146        It amalgamated them into a single range, with the script of the first of
2147        them. In other words, some characters were in the wrong script. There were
2148        thirteen such cases, affecting characters in the following ranges:
2149    
2150          U+002b0 - U+002c1
2151          U+0060c - U+0060d
2152          U+0061e - U+00612
2153          U+0064b - U+0065e
2154          U+0074d - U+0076d
2155          U+01800 - U+01805
2156          U+01d00 - U+01d77
2157          U+01d9b - U+01dbf
2158          U+0200b - U+0200f
2159          U+030fc - U+030fe
2160          U+03260 - U+0327f
2161          U+0fb46 - U+0fbb1
2162          U+10450 - U+1049d
2163    
2164    12. The -o option (show only the matching part of a line) for pcregrep was not
2165        compatible with GNU grep in that, if there was more than one match in a
2166        line, it showed only the first of them. It now behaves in the same way as
2167        GNU grep.
2168    
2169    13. If the -o and -v options were combined for pcregrep, it printed a blank
2170        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2171        does the same. The return code can be used to tell if there were any
2172        non-matching lines.
2173    
2174    14. Added --file-offsets and --line-offsets to pcregrep.
2175    
2176    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2177        infinitely looping recursion. The bug was that positive lookaheads were not
2178        being skipped when checking for a possible empty match (negative lookaheads
2179        and both kinds of lookbehind were skipped).
2180    
2181    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2182        inclusion of <windows.h> to before rather than after the definition of
2183        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2184    
2185    17. Specifying a possessive quantifier with a specific limit for a Unicode
2186        character property caused pcre_compile() to compile bad code, which led at
2187        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2188        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2189        caused the error; without that there was no problem.
2190    
2191    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2192    
2193    19. Added --enable-pcretest-libreadline.
2194    
2195    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2196        RE::GlobalReplace(). As a result, the number of replacements returned was
2197        double what it should be. I removed one of the increments, but Craig sent a
2198        later patch that removed the other one (the right fix) and added unit tests
2199        that check the return values (which was not done before).
2200    
2201    21. Several CMake things:
2202    
2203        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2204            the names libpcre and libpcreposix, not just pcre and pcreposix.
2205    
2206        (2) The above change means that pcretest and pcregrep are now correctly
2207            linked with the newly-built libraries, not previously installed ones.
2208    
2209        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2210    
2211    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2212        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2213        UTF-8 newline character). The key issue is that the pattern starts .*;
2214        this means that the match must be either at the beginning, or after a
2215        newline. The bug was in the code for advancing after a failed match and
2216        checking that the new position followed a newline. It was not taking
2217        account of UTF-8 characters correctly.
2218    
2219    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2220        character classes. PCRE was not treating the sequence [:...:] as a
2221        character class unless the ... were all letters. Perl, however, seems to
2222        allow any characters between [: and :], though of course it rejects as
2223        unknown any "names" that contain non-letters, because all the known class
2224        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2225        for example, whereas PCRE did not - it did not recognize a POSIX character
2226        class. This seemed a bit dangerous, so the code has been changed to be
2227        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2228        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2229        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2230        Perl does, and where it didn't before.
2231    
2232    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2233        Windows environments %n is disabled by default.
2234    
2235    
2236    Version 7.4 21-Sep-07
2237    ---------------------
2238    
2239    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2240        means that a class such as [\s] counted as "explicit reference to CR or
2241        LF". That isn't really right - the whole point of the change was to try to
2242        help when there was an actual mention of one of the two characters. So now
2243        the change happens only if \r or \n (or a literal CR or LF) character is
2244        encountered.
2245    
2246    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2247        of both had grown to the point where there were only 3 bits left.
2248        Fortunately, there was spare space in the data structure, and so I have
2249        moved the internal flags into a new 16-bit field to free up more option
2250        bits.
2251    
2252    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2253        but did not set the internal JCHANGED flag - either of these is enough to
2254        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2255        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2256        start sets both bits.
2257    
2258    4.  Added options (at build time, compile time, exec time) to change \R from
2259        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2260    
2261    5.  doc/pcresyntax.html was missing from the distribution.
2262    
2263    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2264        compatibility, even though it is no longer used.
2265    
2266    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2267        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2268        windows.h file is present (where different names are used). [This was
2269        reversed later after testing - see 16 below.]
2270    
2271    8.  Changed all #include <config.h> to #include "config.h". There were also
2272        some further <pcre.h> cases that I changed to "pcre.h".
2273    
2274    9.  When pcregrep was used with the --colour option, it missed the line ending
2275        sequence off the lines that it output.
2276    
2277    10. It was pointed out to me that arrays of string pointers cause lots of
2278        relocations when a shared library is dynamically loaded. A technique of
2279        using a single long string with a table of offsets can drastically reduce
2280        these. I have refactored PCRE in four places to do this. The result is
2281        dramatic:
2282    
2283          Originally:                          290
2284          After changing UCP table:            187
2285          After changing error message table:   43
2286          After changing table of "verbs"       36
2287          After changing table of Posix names   22
2288    
2289        Thanks to the folks working on Gregex for glib for this insight.
2290    
2291    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2292        unicode-properties was also set.
2293    
2294    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2295    
2296    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2297        checked only for CRLF.
2298    
2299    14. Added casts to pcretest.c to avoid compiler warnings.
2300    
2301    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2302    
2303    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2304        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2305        entirely. This removes changes made in 7 above.
2306    
2307    17. The CMake files have been updated, and there is now more information about
2308        building with CMake in the NON-UNIX-USE document.
2309    
2310    
2311    Version 7.3 28-Aug-07
2312    ---------------------
2313    
2314     1. In the rejigging of the build system that eventually resulted in 7.1, the
2315        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2316        brackets there is not right, since it causes compilers to look for an
2317        installed pcre.h, not the version that is in the source that is being
2318        compiled (which of course may be different). I have changed it back to:
2319    
2320          #include "pcre.h"
2321    
2322        I have a vague recollection that the change was concerned with compiling in
2323        different directories, but in the new build system, that is taken care of
2324        by the VPATH setting the Makefile.
2325    
2326     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2327        when the subject happened to end in the byte 0x85 (e.g. if the last
2328        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2329        characters but of course it shouldn't be taken as a newline when it is part
2330        of another character. The bug was that, for an unlimited repeat of . in
2331        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2332        characters when looking for a newline.
2333    
2334     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2335    
2336     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2337        in debug output.
2338    
2339     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2340        long printing in the pcrecpp unittest when running under MinGW.
2341    
2342     6. ESC_K was left out of the EBCDIC table.
2343    
2344     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2345        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2346        limit also applies to "virtual nesting" when a pattern is recursive, and in
2347        this case 1000 isn't so big. I have been able to remove this limit at the
2348        expense of backing off one optimization in certain circumstances. Normally,
2349        when pcre_exec() would call its internal match() function recursively and
2350        immediately return the result unconditionally, it uses a "tail recursion"
2351        feature to save stack. However, when a subpattern that can match an empty
2352        string has an unlimited repetition quantifier, it no longer makes this
2353        optimization. That gives it a stack frame in which to save the data for
2354        checking that an empty string has been matched. Previously this was taken
2355        from the 1000-entry workspace that had been reserved. So now there is no
2356        explicit limit, but more stack is used.
2357    
2358     8. Applied Daniel's patches to solve problems with the import/export magic
2359        syntax that is required for Windows, and which was going wrong for the
2360        pcreposix and pcrecpp parts of the library. These were overlooked when this
2361        problem was solved for the main library.
2362    
2363     9. There were some crude static tests to avoid integer overflow when computing
2364        the size of patterns that contain repeated groups with explicit upper
2365        limits. As the maximum quantifier is 65535, the maximum group length was
2366        set at 30,000 so that the product of these two numbers did not overflow a
2367        32-bit integer. However, it turns out that people want to use groups that
2368        are longer than 30,000 bytes (though not repeat them that many times).
2369        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2370        made it possible to implement the integer overflow checks in a much more
2371        dynamic way, which I have now done. The artificial limitation on group
2372        length has been removed - we now have only the limit on the total length of
2373        the compiled pattern, which depends on the LINK_SIZE setting.
2374    
2375    10. Fixed a bug in the documentation for get/copy named substring when
2376        duplicate names are permitted. If none of the named substrings are set, the
2377        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2378        empty string.
2379    
2380    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2381        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2382        because the ] is interpreted as the first data character and the
2383        terminating ] is not found. PCRE has been made compatible with Perl in this
2384        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2385        cause memory overwriting.
2386    
2387    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2388        string has been matched (to stop an infinite loop). It was not recognizing
2389        a conditional subpattern that could match an empty string if that
2390        subpattern was within another subpattern. For example, it looped when
2391        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2392        condition was not nested. This bug has been fixed.
2393    
2394    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2395        past the start of the subject in the presence of bytes with the top bit
2396        set, for example "\x8aBCD".
2397    
2398    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2399        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2400    
2401    14. Optimized (?!) to (*FAIL).
2402    
2403    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2404        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2405        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2406        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2407        does: it's just the validity check that is more restrictive.
2408    
2409    16. Inserted checks for integer overflows during escape sequence (backslash)
2410        processing, and also fixed erroneous offset values for syntax errors during
2411        backslash processing.
2412    
2413    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2414        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2415    
2416    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2417        caused an overrun.
2418    
2419    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2420        something other than just ASCII characters) inside a group that had an
2421        unlimited repeat caused a loop at compile time (while checking to see
2422        whether the group could match an empty string).
2423    
2424    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2425        [\P{Any}] did so. (Error in the code for printing property names.)
2426    
2427    21. An orphan \E inside a character class could cause a crash.
2428    
2429    22. A repeated capturing bracket such as (A)? could cause a wild memory
2430        reference during compilation.
2431    
2432    23. There are several functions in pcre_compile() that scan along a compiled
2433        expression for various reasons (e.g. to see if it's fixed length for look
2434        behind). There were bugs in these functions when a repeated \p or \P was
2435        present in the pattern. These operators have additional parameters compared
2436        with \d, etc, and these were not being taken into account when moving along
2437        the compiled data. Specifically:
2438    
2439        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2440            length.
2441    
2442        (b) An item such as \pL+ within a repeated group could cause crashes or
2443            loops.
2444    
2445        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2446            "reference to non-existent subpattern" error.
2447    
2448        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2449    
2450    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2451        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2452    
2453    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2454    
2455    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2456        character were causing crashes (broken optimization).
2457    
2458    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2459        \p or \P) caused a compile-time loop.
2460    
2461    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2462        break. For example, the unstudied pattern [\r\n]A does not match the string
2463        "\r\nA" because change 7.0/46 below moves the current point on by two
2464        characters after failing to match at the start. However, the pattern \nA
2465        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2466        the same is true. There doesn't seem any very clean way out of this, but
2467        what I have chosen to do makes the common cases work: PCRE now takes note
2468        of whether there can be an explicit match for \r or \n anywhere in the
2469        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2470        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2471        pattern has explicit CR or LF references.
2472    
2473    29. Added (*CR) etc for changing newline setting at start of pattern.
2474    
2475    
2476    Version 7.2 19-Jun-07
2477    ---------------------
2478    
2479     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2480        which is apparently normally available under Windows.
2481    
2482     2. Re-jig the pcregrep tests with different newline settings in an attempt
2483        to make them independent of the local environment's newline setting.
2484    
2485     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2486    
2487     4. Some of the "internals" tests were previously cut out when the link size
2488        was not 2, because the output contained actual offsets. The recent new
2489        "Z" feature of pcretest means that these can be cut out, making the tests
2490        usable with all link sizes.
2491    
2492     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2493        stack recursion. This gives a massive performance boost under BSD, but just
2494        a small improvement under Linux. However, it saves one field in the frame
2495        in all cases.
2496    
2497     6. Added more features from the forthcoming Perl 5.10:
2498    
2499        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2500            recursion call. It refers to the nth most recently opened parentheses.
2501    
2502        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2503            to be opened parentheses.
2504    
2505        (c) Conditions that refer to capturing parentheses can be specified
2506            relatively, for example, (?(-2)... or (?(+3)...
2507    
2508        (d) \K resets the start of the current match so that everything before
2509            is not part of it.
2510    
2511        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2512    
2513        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2514            reference syntax.
2515    
2516        (g) (?| introduces a group in which the numbering of parentheses in each
2517            alternative starts with the same number.
2518    
2519        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2520    
2521     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2522        PCRE_INFO_JCHANGED.
2523    
2524     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2525        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2526        for detecting groups that can match an empty string.
2527    
2528     9. A pattern with a very large number of alternatives (more than several
2529        hundred) was running out of internal workspace during the pre-compile
2530        phase, where pcre_compile() figures out how much memory will be needed. A
2531        bit of new cunning has reduced the workspace needed for groups with
2532        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2533        workspace instead of running out of the 4096 that are available.
2534    
2535    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2536    
2537    11. Applied patch from Google to remove an optimization that didn't quite work.
2538        The report of the bug said:
2539    
2540          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2541          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2542          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2543    
2544    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2545        it matched the wrong number of bytes.
2546    
2547    
2548    Version 7.1 24-Apr-07
2549  ---------------------  ---------------------
2550    
2551   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2552      that is more "standard", making use of automake and other autotools.      that is more "standard", making use of automake and other Autotools. There
2553        is some re-arrangement of the files and adjustment of comments consequent
2554        on this.
2555    
2556     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2557        for recursive directory scanning broke on some systems because the files
2558        are not scanned in any specific order and on different systems the order
2559        was different. A call to "sort" has been inserted into RunGrepTest for the
2560        approprate test as a short-term fix. In the longer term there may be an
2561        alternative.
2562    
2563     3. I had an email from Eric Raymond about problems translating some of PCRE's
2564        man pages to HTML (despite the fact that I distribute HTML pages, some
2565        people do their own conversions for various reasons). The problems
2566        concerned the use of low-level troff macros .br and .in. I have therefore
2567        removed all such uses from the man pages (some were redundant, some could
2568        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2569        HTML has been updated to handle .nf/.fi and to complain if it encounters
2570        .br or .in.
2571    
2572     4. Updated comments in configure.ac that get placed in config.h.in and also
2573        arranged for config.h to be included in the distribution, with the name
2574        config.h.generic, for the benefit of those who have to compile without
2575        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2576    
2577     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2578        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2579        makevp.bat for the current PCRE, using the additional files
2580        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2581    
2582     6. A Windows user reported a minor discrepancy with test 2, which turned out
2583        to be caused by a trailing space on an input line that had got lost in his
2584        copy. The trailing space was an accident, so I've just removed it.
2585    
2586     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2587        that is needed.
2588    
2589     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2590        as "const" (a) because they are and (b) because it helps the PHP
2591        maintainers who have recently made a script to detect big data structures
2592        in the php code that should be moved to the .rodata section. I remembered
2593        to update Builducptable as well, so it won't revert if ucptable.h is ever
2594        re-created.
2595    
2596     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2597        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2598        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2599        support is not required. This saves 1.5-2K of code, which is important in
2600        some applications.
2601    
2602        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2603        so as not to refer to the tables, even though these functions will never be
2604        called when UTF-8 support is disabled. Otherwise there are problems with a
2605        shared library.
2606    
2607    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2608    
2609        (a) It was defining its arguments as char * instead of void *.
2610    
2611        (b) It was assuming that all moves were upwards in memory; this was true
2612            a long time ago when I wrote it, but is no longer the case.
2613    
2614        The emulated memove() is provided for those environments that have neither
2615        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2616        is clearly not the case, as these two bugs were recently reported.
2617    
2618    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2619        and Detrail to create the HTML documentation, the .txt form of the man
2620        pages, and it removes trailing spaces from listed files. It also creates
2621        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2622        case, it wraps all the #defines with #ifndefs. This script should be run
2623        before "make dist".
2624    
2625    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2626        with Unicode property support.
2627    
2628        (a) For a maximizing quantifier, if the two different cases of the
2629            character were of different lengths in their UTF-8 codings (there are
2630            some cases like this - I found 11), and the matching function had to
2631            back up over a mixture of the two cases, it incorrectly assumed they
2632            were both the same length.
2633    
2634        (b) When PCRE was configured to use the heap rather than the stack for
2635            recursion during matching, it was not correctly preserving the data for
2636            the other case of a UTF-8 character when checking ahead for a match
2637            while processing a minimizing repeat. If the check also involved
2638            matching a wide character, but failed, corruption could cause an
2639            erroneous result when trying to check for a repeat of the original
2640            character.
2641    
2642    13. Some tidying changes to the testing mechanism:
2643    
2644        (a) The RunTest script now detects the internal link size and whether there
2645            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2646            values substituted by "configure". (The RunGrepTest script already did
2647            this for UTF-8.) The configure.ac script no longer substitutes the
2648            relevant variables.
2649    
2650        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2651            with length and offset values. This means that the output is different
2652            for different internal link sizes. Test 2 is skipped for link sizes
2653            other than 2 because of this, bypassing the problem. Unfortunately,
2654            there was also a test in test 3 (the locale tests) that used /B and
2655            failed for link sizes other than 2. Rather than cut the whole test out,
2656            I have added a new /Z option to pcretest that replaces the length and
2657            offset values with spaces. This is now used to make test 3 independent
2658            of link size. (Test 2 will be tidied up later.)
2659    
2660    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2661        segmentation fault instead of returning the appropriate error message.
2662    
2663    15. In multiline mode when the newline sequence was set to "any", the pattern
2664        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2665        This doesn't seem right; it now treats the CRLF combination as the line
2666        ending, and so does not match in that case. It's only a pattern such as ^$
2667        that would hit this one: something like ^ABC$ would have failed after \r
2668        and then tried again after \r\n.
2669    
2670    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2671        in an attempt to make files that differ only in their line terminators
2672        compare equal. This works on Linux.
2673    
2674    17. Under certain error circumstances pcregrep might try to free random memory
2675        as it exited. This is now fixed, thanks to valgrind.
2676    
2677    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2678        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2679        was because its rules for how to advance for /g after matching an empty
2680        string at the end of a line did not allow for this case. They now check for
2681        it specially.
2682    
2683    20. pcretest is supposed to handle patterns and data of any length, by
2684        extending its buffers when necessary. It was getting this wrong when the
2685        buffer for a data line had to be extended.
2686    
2687    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2688        CRLF as a newline sequence.
2689    
2690    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2691        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2692        I have nevertheless tidied it up.
2693    
2694    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2695    
2696    24. Added a man page for pcre-config.
2697    
2698    
2699  Version 7.0 19-Dec-06  Version 7.0 19-Dec-06

Legend:
Removed from v.98  
changed lines
  Added in v.1366

  ViewVC Help
Powered by ViewVC 1.1.5