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

Diff of /code/trunk/ChangeLog

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

revision 45 by nigel, Sat Feb 24 21:39:25 2007 UTC revision 1334 by ph10, Wed May 15 16:53:18 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.33 28-April-2013
5    --------------------------
6    
7    1.  Added 'U' to some constants that are compared to unsigned integers, to
8        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
9        variables that are added to signed variables, to ensure the result is
10        signed and can be negated.
11    
12    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
13        CMake config files.
14    
15    3.  Revise the creation of config.h.generic so that all boolean macros are
16        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
17        overriding via -D on the command line possible.
18    
19    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
20        to unsigned int is reported to make a quite noticeable speed difference in
21        a specific Windows environment. Testing on Linux did also appear to show
22        some benefit (and it is clearly not harmful). Also fixed the definition of
23        Xop which should be unsigned.
24    
25    5.  Related to (4), changing the definition of the intermediate variable cc
26        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
27        improvements.
28    
29    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
30        unnecessary spaces.
31    
32    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
33        and later.
34    
35    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
36    
37    9.  Optimizing fast_forward_start_bits in JIT.
38    
39    10. Adding experimental support for callouts in JIT, and fixing some
40        issues revealed during this work. Namely:
41    
42        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
43    
44        (b) Minimum length was not checked before the matching is started.
45    
46    11. The value of capture_last that is passed to callouts was incorrect in some
47        cases when there was a capture on one path that was subsequently abandoned
48        after a backtrack. Also, the capture_last value is now reset after a
49        recursion, since all captures are also reset in this case.
50    
51    12. The interpreter no longer returns the "too many substrings" error in the
52        case when an overflowing capture is in a branch that is subsequently
53        abandoned after a backtrack.
54    
55    13. In the pathological case when an offset vector of size 2 is used, pcretest
56        now prints out the matched string after a yield of 0 or 1.
57    
58    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
59        Only supported by the JIT compiler at the moment.
60    
61    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
62    
63    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
64        offset of the starting point of the matching process, provided the offsets
65        vector is large enough.
66    
67    17. The \A escape now records a lookbehind value of 1, though its execution
68        does not actually inspect the previous character. This is to ensure that,
69        in partial multi-segment matching, at least one character from the old
70        segment is retained when a new segment is processed. Otherwise, if there
71        are no lookbehinds in the pattern, \A might match incorrectly at the start
72        of a new segment.
73    
74    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
75    
76    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
77        gives some modest performance improvement in 8-bit mode.
78    
79    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
80        be expressed in certain programming languages using Universal Character
81        Names.
82    
83    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
84        which points out that "non characters" are not "characters that may not
85        appear in Unicode strings" but rather "characters that are reserved for
86        internal use and have only local meaning".
87    
88    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
89        there was a conditional group that depended on an assertion, if the
90        assertion was false, the callout that immediately followed the alternation
91        in the condition was skipped when pcre_exec() was used for matching.
92    
93    23. Allow an explicit callout to be inserted before an assertion that is the
94        condition for a conditional group, for compatibility with automatic
95        callouts, which always insert a callout at this point.
96    
97    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
98        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
99    
100    25. (*PRUNE) is now supported by the JIT compiler.
101    
102    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
103    
104    27. Fix the case where there are two or more SKIPs with arguments that may be
105        ignored.
106    
107    28. (*SKIP) is now supported by the JIT compiler.
108    
109    29. (*THEN) is now supported by the JIT compiler.
110    
111    30. Update RunTest with additional test selector options.
112    
113    31. The way PCRE handles backtracking verbs has been changed in two ways.
114    
115        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
116        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
117        backtracking. In some cases this makes it more Perl-compatible, but Perl's
118        rather obscure rules do not always do the same thing.
119    
120        (2) Previously, backtracking verbs were confined within assertions. This is
121        no longer the case for positive assertions, except for (*ACCEPT). Again,
122        this sometimes improves Perl compatibility, and sometimes does not.
123    
124    32. A number of tests that were in test 2 because Perl did things differently
125        have been moved to test 1, because either Perl or PCRE has changed, and
126        these tests are now compatible.
127    
128    32. Control verbs are handled in the same way in JIT and interpreter.
129    
130    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
131        contained a forward subroutine reference caused a compile error.
132    
133    34. Auto-detect and optimize limited repetitions in JIT.
134    
135    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
136        blocking (*UTF) etc.
137    
138    36. In the interpreter, maximizing pattern repetitions for characters and
139        character types now use tail recursion, which reduces stack usage.
140    
141    37. The value of the max lookbehind was not correctly preserved if a compiled
142        and saved regex was reloaded on a host of different endianness.
143    
144    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
145        of the compiled pattern block, expand the flags field from 16 to 32 bits
146        because it was almost full.
147    
148    39. Try madvise first before posix_madvise.
149    
150    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
151        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
152        original change locked out all matches of empty strings. This has been
153        changed so that one match of an empty string per line is recognized.
154        Subsequent searches on the same line (for colouring or for --only-matching,
155        for example) do not recognize empty strings.
156    
157    41. Applied a user patch to fix a number of spelling mistakes in comments.
158    
159    42. Data lines longer than 65536 caused pcretest to crash.
160    
161    43. Clarified the data type for length and startoffset arguments for pcre_exec
162        and pcre_dfa_exec in the function-specific man pages, where they were
163        explicitly stated to be in bytes, never having been updated. I also added
164        some clarification to the pcreapi man page.
165    
166    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
167        a segmentation fault.
168    
169    
170    Version 8.32 30-November-2012
171    -----------------------------
172    
173    1.  Improved JIT compiler optimizations for first character search and single
174        character iterators.
175    
176    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
177        Patch by Daniel Richard G.
178    
179    3.  Single character iterator optimizations in the JIT compiler.
180    
181    4.  Improved JIT compiler optimizations for character ranges.
182    
183    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
184        Reported by Giuseppe D'Angelo.
185    
186    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
187        of a line, was being set incorrectly in cases where .* appeared inside
188        atomic brackets at the start of a pattern, or where there was a subsequent
189        *PRUNE or *SKIP.
190    
191    7.  Improved instruction cache flush for POWER/PowerPC.
192        Patch by Daniel Richard G.
193    
194    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
195        grep:
196    
197        (a) There is now no limit to the number of patterns to be matched.
198    
199        (b) An error is given if a pattern is too long.
200    
201        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
202            are now supported.
203    
204        (d) --exclude-from and --include-from (multiple use) have been added.
205    
206        (e) Exclusions and inclusions now apply to all files and directories, not
207            just to those obtained from scanning a directory recursively.
208    
209        (f) Multiple uses of -f and --file-list are now supported.
210    
211        (g) In a Windows environment, the default for -d has been changed from
212            "read" (the GNU grep default) to "skip", because otherwise the presence
213            of a directory in the file list provokes an error.
214    
215        (h) The documentation has been revised and clarified in places.
216    
217    9.  Improve the matching speed of capturing brackets.
218    
219    10. Changed the meaning of \X so that it now matches a Unicode extended
220        grapheme cluster.
221    
222    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
223        out POSIX threads when JIT support is configured.
224    
225    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
226    
227    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
228        structure could go wrong in environments where size_t is not the same size
229        as int.
230    
231    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
232        set.
233    
234    15. The EBCDIC support had decayed; later updates to the code had included
235        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
236        general tidy up of EBCDIC-related issues, and the documentation was also
237        not quite right. There is now a test that can be run on ASCII systems to
238        check some of the EBCDIC-related things (but is it not a full test).
239    
240    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
241        in a small tidy to the code.
242    
243    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
244    
245    18. If the --only-matching (-o) option in pcregrep is specified multiple
246        times, each one causes appropriate output. For example, -o1 -o2 outputs the
247        substrings matched by the 1st and 2nd capturing parentheses. A separating
248        string can be specified by --om-separator (default empty).
249    
250    19. Improving the first n character searches.
251    
252    20. Turn case lists for horizontal and vertical white space into macros so that
253        they are defined only once.
254    
255    21. This set of changes together give more compatible Unicode case-folding
256        behaviour for characters that have more than one other case when UCP
257        support is available.
258    
259        (a) The Unicode property table now has offsets into a new table of sets of
260            three or more characters that are case-equivalent. The MultiStage2.py
261            script that generates these tables (the pcre_ucd.c file) now scans
262            CaseFolding.txt instead of UnicodeData.txt for character case
263            information.
264    
265        (b) The code for adding characters or ranges of characters to a character
266            class has been abstracted into a generalized function that also handles
267            case-independence. In UTF-mode with UCP support, this uses the new data
268            to handle characters with more than one other case.
269    
270        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
271            whose other case is greater than 256 are now correctly matched
272            caselessly. Previously, the high codepoint matched the low one, but not
273            vice versa.
274    
275        (d) The processing of \h, \H, \v, and \ in character classes now makes use
276            of the new class addition function, using character lists defined as
277            macros alongside the case definitions of 20 above.
278    
279        (e) Caseless back references now work with characters that have more than
280            one other case.
281    
282        (f) General caseless matching of characters with more than one other case
283            is supported.
284    
285    22. Unicode character properties were updated from Unicode 6.2.0
286    
287    23. Improved CMake support under Windows. Patch by Daniel Richard G.
288    
289    24. Add support for 32-bit character strings, and UTF-32
290    
291    25. Major JIT compiler update (code refactoring and bugfixing).
292        Experimental Sparc 32 support is added.
293    
294    26. Applied a modified version of Daniel Richard G's patch to create
295        pcre.h.generic and config.h.generic by "make" instead of in the
296        PrepareRelease script.
297    
298    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
299        pcre_compile.c when checking for a zero character.
300    
301    28. Introducing a native interface for JIT. Through this interface, the compiled
302        machine code can be directly executed. The purpose of this interface is to
303        provide fast pattern matching, so several sanity checks are not performed.
304        However, feature tests are still performed. The new interface provides
305        1.4x speedup compared to the old one.
306    
307    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
308        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
309        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
310    
311    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
312        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
313        the "old" RFC 2279). Instead, it ended up passing a negative length to
314        pcre_exec().
315    
316    31. Add support for GCC's visibility feature to hide internal functions.
317    
318    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
319        "unknown -C option" after outputting 0 or 1.
320    
321    33. There is now support for generating a code coverage report for the test
322        suite in environments where gcc is the compiler and lcov is installed. This
323        is mainly for the benefit of the developers.
324    
325    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
326        unaddressable using valgrind annotations, allowing valgrind to detect
327        invalid memory accesses. This is mainly for the benefit of the developers.
328    
329    25. (*UTF) can now be used to start a pattern in any of the three libraries.
330    
331    26. Give configure error if --enable-cpp but no C++ compiler found.
332    
333    
334    Version 8.31 06-July-2012
335    -------------------------
336    
337    1.  Fixing a wrong JIT test case and some compiler warnings.
338    
339    2.  Removed a bashism from the RunTest script.
340    
341    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
342        to unsigned type, result still unsigned" that was given by an MS compiler
343        on encountering the code "-sizeof(xxx)".
344    
345    4.  Partial matching support is added to the JIT compiler.
346    
347    5.  Fixed several bugs concerned with partial matching of items that consist
348        of more than one character:
349    
350        (a) /^(..)\1/ did not partially match "aba" because checking references was
351            done on an "all or nothing" basis. This also applied to repeated
352            references.
353    
354        (b) \R did not give a hard partial match if \r was found at the end of the
355            subject.
356    
357        (c) \X did not give a hard partial match after matching one or more
358            characters at the end of the subject.
359    
360        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
361            a partial match for the string "\r".
362    
363        (e) When newline was set to CRLF, the metacharacter "." did not recognize
364            a partial match for a CR character at the end of the subject string.
365    
366    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
367        running pcretest, the text "(JIT)" added to the output whenever JIT is
368        actually used to run the match.
369    
370    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
371        or /S+[+] with a digit between 1 and 7.
372    
373    8.  OP_NOT now supports any UTF character not just single-byte ones.
374    
375    9.  (*MARK) control verb is now supported by the JIT compiler.
376    
377    10. The command "./RunTest list" lists the available tests without actually
378        running any of them. (Because I keep forgetting what they all are.)
379    
380    11. Add PCRE_INFO_MAXLOOKBEHIND.
381    
382    12. Applied a (slightly modified) user-supplied patch that improves performance
383        when the heap is used for recursion (compiled with --disable-stack-for-
384        recursion). Instead of malloc and free for each heap frame each time a
385        logical recursion happens, frames are retained on a chain and re-used where
386        possible. This sometimes gives as much as 30% improvement.
387    
388    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
389        call.
390    
391    14. As documented, (*COMMIT) is now confined to within a positive assertion.
392    
393    15. It is now possible to link pcretest with libedit as an alternative to
394        libreadline.
395    
396    16. (*COMMIT) control verb is now supported by the JIT compiler.
397    
398    17. The Unicode data tables have been updated to Unicode 6.1.0.
399    
400    18. Added --file-list option to pcregrep.
401    
402    19. Added binary file support to pcregrep, including the -a, --binary-files,
403        -I, and --text options.
404    
405    20. The madvise function is renamed for posix_madvise for QNX compatibility
406        reasons. Fixed by Giuseppe D'Angelo.
407    
408    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
409    
410    22. Changed the option for creating symbolic links for 16-bit man pages from
411        -s to -sf so that re-installing does not cause issues.
412    
413    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
414    
415    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
416        matches in certain environments (the workspace was not being correctly
417        retained). Also added to pcre_dfa_exec() a simple plausibility check on
418        some of the workspace data at the beginning of a restart.
419    
420    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
421        was not doing so when it should - probably a typo introduced by SVN 528
422        (change 8.10/14).
423    
424    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
425        \w+ when the character tables indicated that \x{c4} was a word character.
426        There were several related cases, all because the tests for doing a table
427        lookup were testing for characters less than 127 instead of 255.
428    
429    27. If a pattern contains capturing parentheses that are not used in a match,
430        their slots in the ovector are set to -1. For those that are higher than
431        any matched groups, this happens at the end of processing. In the case when
432        there were back references that the ovector was too small to contain
433        (causing temporary malloc'd memory to be used during matching), and the
434        highest capturing number was not used, memory off the end of the ovector
435        was incorrectly being set to -1. (It was using the size of the temporary
436        memory instead of the true size.)
437    
438    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
439        ovector size, it uses memory at the end of the block that it has got.
440    
441    29. Check for an overlong MARK name and give an error at compile time. The
442        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
443    
444    30. JIT compiler update.
445    
446    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
447        Rill for the patch.
448    
449    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
450    
451    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
452    
453    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
454        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
455        was enabled.
456    
457    35. Improve JIT code generation for greedy plus quantifier.
458    
459    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
460        1 to "aa" instead of to an empty string. The bug affected repeated groups
461        that could potentially match an empty string.
462    
463    37. Optimizing single character iterators in JIT.
464    
465    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
466        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
467        codepoints that are too big for the mode are faulted, and in a UTF mode,
468        disallowed codepoints are also faulted.
469    
470    39. If PCRE was compiled with UTF support, in three places in the DFA
471        matcher there was code that should only have been obeyed in UTF mode, but
472        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
473        processing when bytes with values greater than 127 were present. In 16-bit
474        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
475        both cases the values are those that cannot be the first data item in a UTF
476        character. The three items that might have provoked this were recursions,
477        possessively repeated groups, and atomic groups.
478    
479    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
480        and pcregrep, because some OS require shared objects to be explicitly
481        passed to ld, causing the link step to fail if they are not.
482    
483    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
484        mode, patterns that started with \h* or \R* might be incorrectly matched.
485    
486    
487    Version 8.30 04-February-2012
488    -----------------------------
489    
490    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
491        name is defined in ctype.h.
492    
493    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
494        only in quite long subpatterns.
495    
496    3.  Removed the function pcre_info(), which has been obsolete and deprecated
497        since it was replaced by pcre_fullinfo() in February 2000.
498    
499    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
500        match a (*MARK), and the match failed at the start of the subject, a
501        reference to memory before the start of the subject could occur. This bug
502        was introduced by fix 17 of release 8.21.
503    
504    5.  A reference to an unset group with zero minimum repetition was giving
505        totally wrong answers (in non-JavaScript-compatibility mode). For example,
506        /(another)?(\1?)test/ matched against "hello world test". This bug was
507        introduced in release 8.13.
508    
509    6.  Add support for 16-bit character strings (a large amount of work involving
510        many changes and refactorings).
511    
512    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
513        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
514        from a file.
515    
516    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
517        rounding is not applied in this particular case).
518    
519    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
520        if they appear, or are escaped, in patterns.
521    
522    10. Get rid of a number of -Wunused-but-set-variable warnings.
523    
524    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
525        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
526        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
527        also returns the mark "x". This bug applied to capturing parentheses,
528        non-capturing parentheses, and atomic parentheses. It also applied to some
529        assertions.
530    
531    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
532        information out of configure.ac instead of relying on pcre.h.generic, which
533        is not stored in the repository.
534    
535    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
536        -lreadline.
537    
538    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
539    
540    15. Applied Graycode's patch to put the top-level frame on the stack rather
541        than the heap when not using the stack for recursion. This gives a
542        performance improvement in many cases when recursion is not deep.
543    
544    16. Experimental code added to "pcretest -C" to output the stack frame size.
545    
546    
547    Version 8.21 12-Dec-2011
548    ------------------------
549    
550    1.  Updating the JIT compiler.
551    
552    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
553        are added as well.
554    
555    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
556        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
557        calling _pcre_jit_exec. Some extra comments are added.
558    
559    4.  (*MARK) settings inside atomic groups that do not contain any capturing
560        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
561        was introduced by change 18 for 8.20.
562    
563    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
564        ECMA-262 standard.
565    
566    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
567        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
568        This bug was probably introduced by change 9 of 8.13.
569    
570    7.  While fixing 6 above, I noticed that a number of other items were being
571        incorrectly rejected as "not fixed length". This arose partly because newer
572        opcodes had not been added to the fixed-length checking code. I have (a)
573        corrected the bug and added tests for these items, and (b) arranged for an
574        error to occur if an unknown opcode is encountered while checking for fixed
575        length instead of just assuming "not fixed length". The items that were
576        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
577        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
578        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
579    
580    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
581        being incorrectly compiled and would have given unpredicatble results.
582    
583    9.  A possessively repeated subpattern with minimum repeat count greater than
584        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
585        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
586        the first (A) could occur when it should not.
587    
588    10. Add a cast and remove a redundant test from the code.
589    
590    11. JIT should use pcre_malloc/pcre_free for allocation.
591    
592    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
593        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
594        is anything other than /usr, -L is still shown).
595    
596    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
597    
598    14. Perl does not support \N without a following name in a [] class; PCRE now
599        also gives an error.
600    
601    15. If a forward reference was repeated with an upper limit of around 2000,
602        it caused the error "internal error: overran compiling workspace". The
603        maximum number of forward references (including repeats) was limited by the
604        internal workspace, and dependent on the LINK_SIZE. The code has been
605        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
606        the default depends on LINK_SIZE. There is a new upper limit (for safety)
607        of around 200,000 forward references. While doing this, I also speeded up
608        the filling in of repeated forward references.
609    
610    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
611        incorrectly expecting the subject to contain another "a" after the start.
612    
613    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
614        in the match, the SKIP should be ignored. This was not happening; instead
615        the SKIP was being treated as NOMATCH. For patterns such as
616        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
617        tested.
618    
619    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
620        now much more compatible with Perl, in particular in cases where the result
621        is a non-match for a non-anchored pattern. For example, if
622        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
623        "m", where previously it did not return a name. A side effect of this
624        change is that for partial matches, the last encountered mark name is
625        returned, as for non matches. A number of tests that were previously not
626        Perl-compatible have been moved into the Perl-compatible test files. The
627        refactoring has had the pleasing side effect of removing one argument from
628        the match() function, thus reducing its stack requirements.
629    
630    19. If the /S+ option was used in pcretest to study a pattern using JIT,
631        subsequent uses of /S (without +) incorrectly behaved like /S+.
632    
633    21. Retrieve executable code size support for the JIT compiler and fixing
634        some warnings.
635    
636    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
637        not work when the shorter character appeared right at the end of the
638        subject string.
639    
640    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
641        systems.
642    
643    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
644        output it when the /M option is used in pcretest.
645    
646    25. The CheckMan script was not being included in the distribution. Also, added
647        an explicit "perl" to run Perl scripts from the PrepareRelease script
648        because this is reportedly needed in Windows.
649    
650    26. If study data was being save in a file and studying had not found a set of
651        "starts with" bytes for the pattern, the data written to the file (though
652        never used) was taken from uninitialized memory and so caused valgrind to
653        complain.
654    
655    27. Updated RunTest.bat as provided by Sheri Pierce.
656    
657    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
658    
659    29. Computation of memory usage for the table of capturing group names was
660        giving an unnecessarily large value.
661    
662    
663    Version 8.20 21-Oct-2011
664    ------------------------
665    
666    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
667        a POSIX class. After further experiments with Perl, which convinced me that
668        Perl has bugs and confusions, a closing square bracket is no longer allowed
669        in a POSIX name. This bug also affected patterns with classes that started
670        with full stops.
671    
672    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
673        captured substring, but while checking the failing first alternative,
674        substring 1 is temporarily captured. If the output vector supplied to
675        pcre_exec() was not big enough for this capture, the yield of the function
676        was still zero ("insufficient space for captured substrings"). This cannot
677        be totally fixed without adding another stack variable, which seems a lot
678        of expense for a edge case. However, I have improved the situation in cases
679        such as /(a)(b)x|abc/ matched against "abc", where the return code
680        indicates that fewer than the maximum number of slots in the ovector have
681        been set.
682    
683    3.  Related to (2) above: when there are more back references in a pattern than
684        slots in the output vector, pcre_exec() uses temporary memory during
685        matching, and copies in the captures as far as possible afterwards. It was
686        using the entire output vector, but this conflicts with the specification
687        that only 2/3 is used for passing back captured substrings. Now it uses
688        only the first 2/3, for compatibility. This is, of course, another edge
689        case.
690    
691    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
692        main code base, and can be used by building with --enable-jit. When this is
693        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
694        runtime --no-jit option is given.
695    
696    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
697        ovector, the return from the function was zero, implying that there were
698        other matches that did not fit. The correct "exactly full" value is now
699        returned.
700    
701    6.  If a subpattern that was called recursively or as a subroutine contained
702        (*PRUNE) or any other control that caused it to give a non-standard return,
703        invalid errors such as "Error -26 (nested recursion at the same subject
704        position)" or even infinite loops could occur.
705    
706    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
707        computing the minimum length on reaching *ACCEPT, and so ended up with the
708        wrong value of 1 rather than 0. Further investigation indicates that
709        computing a minimum subject length in the presence of *ACCEPT is difficult
710        (think back references, subroutine calls), and so I have changed the code
711        so that no minimum is registered for a pattern that contains *ACCEPT.
712    
713    8.  If (*THEN) was present in the first (true) branch of a conditional group,
714        it was not handled as intended. [But see 16 below.]
715    
716    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
717        Sheri Pierce.
718    
719    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
720        the first byte in a match must be "a".
721    
722    11. Change 17 for 8.13 increased the recursion depth for patterns like
723        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
724        pattern contains any instances of (*THEN). If it does not, the old
725        optimizations are restored. It would be nice to do this on a per-group
726        basis, but at the moment that is not feasible.
727    
728    12. In some environments, the output of pcretest -C is CRLF terminated. This
729        broke RunTest's code that checks for the link size. A single white space
730        character after the value is now allowed for.
731    
732    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
733        For "fr", it uses the Windows-specific input and output files.
734    
735    14. If (*THEN) appeared in a group that was called recursively or as a
736        subroutine, it did not work as intended. [But see next item.]
737    
738    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
739        pattern fragments (but not containing any | characters). If A and B are
740        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
741        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
742        D. In other words, Perl considers parentheses that do not contain any |
743        characters to be part of a surrounding alternative, whereas PCRE was
744        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
745        differently. PCRE now behaves in the same way as Perl, except in the case
746        of subroutine/recursion calls such as (?1) which have in any case always
747        been different (but PCRE had them first :-).
748    
749    16. Related to 15 above: Perl does not treat the | in a conditional group as
750        creating alternatives. Such a group is treated in the same way as an
751        ordinary group without any | characters when processing (*THEN). PCRE has
752        been changed to match Perl's behaviour.
753    
754    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
755        RunGrepTest script failed.
756    
757    18. Change 22 for version 13 caused atomic groups to use more stack. This is
758        inevitable for groups that contain captures, but it can lead to a lot of
759        stack use in large patterns. The old behaviour has been restored for atomic
760        groups that do not contain any capturing parentheses.
761    
762    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
763        suppress the check for a minimum subject length at run time. (If it was
764        given to pcre_exec() or pcre_dfa_exec() it did work.)
765    
766    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
767        fail to work when decoding hex characters in data strings in EBCDIC
768        environments.
769    
770    21. It appears that in at least one Mac OS environment, the isxdigit() function
771        is implemented as a macro that evaluates to its argument more than once,
772        contravening the C 90 Standard (I haven't checked a later standard). There
773        was an instance in pcretest which caused it to go wrong when processing
774        \x{...} escapes in subject strings. The has been rewritten to avoid using
775        things like p++ in the argument of isxdigit().
776    
777    
778    Version 8.13 16-Aug-2011
779    ------------------------
780    
781    1.  The Unicode data tables have been updated to Unicode 6.0.0.
782    
783    2.  Two minor typos in pcre_internal.h have been fixed.
784    
785    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
786        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
787        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
788    
789    4.  There were a number of related bugs in the code for matching backrefences
790        caselessly in UTF-8 mode when codes for the characters concerned were
791        different numbers of bytes. For example, U+023A and U+2C65 are an upper
792        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
793        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
794        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
795        2-byte code at the end of the subject (it thought there wasn't enough data
796        left).
797    
798    5.  Comprehensive information about what went wrong is now returned by
799        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
800        as the output vector has at least 2 elements. The offset of the start of
801        the failing character and a reason code are placed in the vector.
802    
803    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
804        now returned is for the first byte of the failing character, instead of the
805        last byte inspected. This is an incompatible change, but I hope it is small
806        enough not to be a problem. It makes the returned offset consistent with
807        pcre_exec() and pcre_dfa_exec().
808    
809    7.  pcretest now gives a text phrase as well as the error number when
810        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
811        failure, the offset and reason code are output.
812    
813    8.  When \R was used with a maximizing quantifier it failed to skip backwards
814        over a \r\n pair if the subsequent match failed. Instead, it just skipped
815        back over a single character (\n). This seems wrong (because it treated the
816        two characters as a single entity when going forwards), conflicts with the
817        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
818        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
819        has been changed.
820    
821    9.  Some internal refactoring has changed the processing so that the handling
822        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
823        time (the PCRE_DOTALL option was changed this way some time ago: version
824        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
825        which was always a bit of a fudge. It also means that there is one less
826        argument for the match() function, which reduces its stack requirements
827        slightly. This change also fixes an incompatibility with Perl: the pattern
828        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
829    
830    10. More internal refactoring has drastically reduced the number of recursive
831        calls to match() for possessively repeated groups such as (abc)++ when
832        using pcre_exec().
833    
834    11. While implementing 10, a number of bugs in the handling of groups were
835        discovered and fixed:
836    
837        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
838        (a|)*(?1) gave a compile-time internal error.
839        ((a|)+)+  did not notice that the outer group could match an empty string.
840        (^a|^)+   was not marked as anchored.
841        (.*a|.*)+ was not marked as matching at start or after a newline.
842    
843    12. Yet more internal refactoring has removed another argument from the match()
844        function. Special calls to this function are now indicated by setting a
845        value in a variable in the "match data" data block.
846    
847    13. Be more explicit in pcre_study() instead of relying on "default" for
848        opcodes that mean there is no starting character; this means that when new
849        ones are added and accidentally left out of pcre_study(), testing should
850        pick them up.
851    
852    14. The -s option of pcretest has been documented for ages as being an old
853        synonym of -m (show memory usage). I have changed it to mean "force study
854        for every regex", that is, assume /S for every regex. This is similar to -i
855        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
856        using it. It makes it easier to run collections of tests with and without
857        study enabled, and thereby test pcre_study() more easily. All the standard
858        tests are now run with and without -s (but some patterns can be marked as
859        "never study" - see 20 below).
860    
861    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
862        restoration of the capturing data to the outer values was not happening
863        correctly.
864    
865    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
866        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
867        pattern had matched an empty string, and so incorrectly returned a no
868        match.
869    
870    17. There was optimizing code for the last branch of non-capturing parentheses,
871        and also for the obeyed branch of a conditional subexpression, which used
872        tail recursion to cut down on stack usage. Unfortunately, now that there is
873        the possibility of (*THEN) occurring in these branches, tail recursion is
874        no longer possible because the return has to be checked for (*THEN). These
875        two optimizations have therefore been removed. [But see 8.20/11 above.]
876    
877    18. If a pattern containing \R was studied, it was assumed that \R always
878        matched two bytes, thus causing the minimum subject length to be
879        incorrectly computed because \R can also match just one byte.
880    
881    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
882        was incorrectly computed.
883    
884    20. If /S is present twice on a test pattern in pcretest input, it now
885        *disables* studying, thereby overriding the use of -s on the command line
886        (see 14 above). This is necessary for one or two tests to keep the output
887        identical in both cases.
888    
889    21. When (*ACCEPT) was used in an assertion that matched an empty string and
890        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
891    
892    22. When an atomic group that contained a capturing parenthesis was
893        successfully matched, but the branch in which it appeared failed, the
894        capturing was not being forgotten if a higher numbered group was later
895        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
896        group 1 to "a", when in fact it should be unset. This applied to multi-
897        branched capturing and non-capturing groups, repeated or not, and also to
898        positive assertions (capturing in negative assertions does not happen
899        in PCRE) and also to nested atomic groups.
900    
901    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
902        subject after a captured substring, to make it easier to tell which of a
903        number of identical substrings has been captured.
904    
905    24. The way atomic groups are processed by pcre_exec() has been changed so that
906        if they are repeated, backtracking one repetition now resets captured
907        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
908        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
909        "aaa". Previously, it would have been "a". As part of this code
910        refactoring, the way recursive calls are handled has also been changed.
911    
912    25. If an assertion condition captured any substrings, they were not passed
913        back unless some other capturing happened later. For example, if
914        (?(?=(a))a) was matched against "a", no capturing was returned.
915    
916    26. When studying a pattern that contained subroutine calls or assertions,
917        the code for finding the minimum length of a possible match was handling
918        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
919        group 1 called group 2 while simultaneously a separate group 2 called group
920        1). A stack overflow occurred in this case. I have fixed this by limiting
921        the recursion depth to 10.
922    
923    27. Updated RunTest.bat in the distribution to the version supplied by Tom
924        Fortmann. This supports explicit test numbers on the command line, and has
925        argument validation and error reporting.
926    
927    28. An instance of \X with an unlimited repeat could fail if at any point the
928        first character it looked at was a mark character.
929    
930    29. Some minor code refactoring concerning Unicode properties and scripts
931        should reduce the stack requirement of match() slightly.
932    
933    30. Added the '=' option to pcretest to check the setting of unused capturing
934        slots at the end of the pattern, which are documented as being -1, but are
935        not included in the return count.
936    
937    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
938        compiled something random. Now it gives a compile-time error (as does
939        Perl).
940    
941    32. A *MARK encountered during the processing of a positive assertion is now
942        recorded and passed back (compatible with Perl).
943    
944    33. If --only-matching or --colour was set on a pcregrep call whose pattern
945        had alternative anchored branches, the search for a second match in a line
946        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
947        matched the line "0102" twice. The same bug affected patterns that started
948        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
949        twice.
950    
951    34. Previously, PCRE did not allow quantification of assertions. However, Perl
952        does, and because of capturing effects, quantifying parenthesized
953        assertions may at times be useful. Quantifiers are now allowed for
954        parenthesized assertions.
955    
956    35. A minor code tidy in pcre_compile() when checking options for \R usage.
957    
958    36. \g was being checked for fancy things in a character class, when it should
959        just be a literal "g".
960    
961    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
962        appearance of a nested POSIX class supersedes an apparent external class.
963        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
964        unescaped square brackets may also appear as part of class names. For
965        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
966        more like Perl. (But see 8.20/1 above.)
967    
968    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
969        was because it thought it was \N{name}, which is not supported).
970    
971    39. Add minix to OS list not supporting the -S option in pcretest.
972    
973    40. PCRE tries to detect cases of infinite recursion at compile time, but it
974        cannot analyze patterns in sufficient detail to catch mutual recursions
975        such as ((?1))((?2)). There is now a runtime test that gives an error if a
976        subgroup is called recursively as a subpattern for a second time at the
977        same position in the subject string. In previous releases this might have
978        been caught by the recursion limit, or it might have run out of stack.
979    
980    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
981        happen only once. PCRE was, however incorrectly giving a compile time error
982        "recursive call could loop indefinitely" because it cannot analyze the
983        pattern in sufficient detail. The compile time test no longer happens when
984        PCRE is compiling a conditional subpattern, but actual runaway loops are
985        now caught at runtime (see 40 above).
986    
987    42. It seems that Perl allows any characters other than a closing parenthesis
988        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
989        has been changed to be the same.
990    
991    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
992        as not to get warnings when autogen.sh is called. Also changed
993        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
994    
995    44. To help people who use pcregrep to scan files containing exceedingly long
996        lines, the following changes have been made:
997    
998        (a) The default value of the buffer size parameter has been increased from
999            8K to 20K. (The actual buffer used is three times this size.)
1000    
1001        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1002            PCRE is built.
1003    
1004        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1005            to be set at run time.
1006    
1007        (d) Numerical values in pcregrep options can be followed by K or M, for
1008            example --buffer-size=50K.
1009    
1010        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1011            given and the return code is set to 2.
1012    
1013    45. Add a pointer to the latest mark to the callout data block.
1014    
1015    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1016        partial match of an empty string instead of no match. This was specific to
1017        the use of ".".
1018    
1019    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1020        complete match instead of a partial match. This bug was dependent on both
1021        the PCRE_UTF8 and PCRE_DOTALL options being set.
1022    
1023    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1024        starting byte set, because \b was not being ignored.
1025    
1026    
1027    Version 8.12 15-Jan-2011
1028    ------------------------
1029    
1030    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1031        checks for such things as part of the documentation building process.
1032    
1033    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1034        --match-limit and --recursion-limit options (added for 8.11). In
1035        particular, this made one of the standard tests fail. (The integer value
1036        went into the wrong half of a long int.)
1037    
1038    3.  If the --colour option was given to pcregrep with -v (invert match), it
1039        did strange things, either producing crazy output, or crashing. It should,
1040        of course, ignore a request for colour when reporting lines that do not
1041        match.
1042    
1043    4.  Another pcregrep bug caused similar problems if --colour was specified with
1044        -M (multiline) and the pattern match finished with a line ending.
1045    
1046    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1047        matched in multiline mode, the following line was shown as part of the
1048        match. This seems wrong, so I have changed it.
1049    
1050    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1051        the check for further matches in the same line (so they could be coloured)
1052        to overrun the end of the current line. If another match was found, it was
1053        incorrectly shown (and then shown again when found in the next line).
1054    
1055    7.  If pcregrep was compiled under Windows, there was a reference to the
1056        function pcregrep_exit() before it was defined. I am assuming this was
1057        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1058        reported by a user. I've moved the definition above the reference.
1059    
1060    
1061    Version 8.11 10-Dec-2010
1062    ------------------------
1063    
1064    1.  (*THEN) was not working properly if there were untried alternatives prior
1065        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1066        backtracked to try for "b" instead of moving to the next alternative branch
1067        at the same level (in this case, to look for "c"). The Perl documentation
1068        is clear that when (*THEN) is backtracked onto, it goes to the "next
1069        alternative in the innermost enclosing group".
1070    
1071    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1072        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1073        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1074        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1075        (*THEN).
1076    
1077    3.  If \s appeared in a character class, it removed the VT character from
1078        the class, even if it had been included by some previous item, for example
1079        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1080        of \s, but is part of the POSIX "space" class.)
1081    
1082    4.  A partial match never returns an empty string (because you can always
1083        match an empty string at the end of the subject); however the checking for
1084        an empty string was starting at the "start of match" point. This has been
1085        changed to the "earliest inspected character" point, because the returned
1086        data for a partial match starts at this character. This means that, for
1087        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1088        (previously it gave "no match").
1089    
1090    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1091        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1092        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1093        has an implication that the given string is incomplete (because a partial
1094        match is preferred over a full match). For this reason, these items now
1095        give a partial match in this situation. [Aside: previously, the one case
1096        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1097        match rather than a full match, which was wrong by the old rules, but is
1098        now correct.]
1099    
1100    6.  There was a bug in the handling of #-introduced comments, recognized when
1101        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1102        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1103        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1104        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1105        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1106        places in pcre_compile().
1107    
1108    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1109        comments when looking ahead for named forward references to subpatterns,
1110        the only newline sequence it recognized was NL. It now handles newlines
1111        according to the set newline convention.
1112    
1113    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1114        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1115        cater for a lack of strerror(). These oversights have been fixed.
1116    
1117    9.  Added --match-limit and --recursion-limit to pcregrep.
1118    
1119    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1120    
1121    11. When the -o option was used, pcregrep was setting a return code of 1, even
1122        when matches were found, and --line-buffered was not being honoured.
1123    
1124    12. Added an optional parentheses number to the -o and --only-matching options
1125        of pcregrep.
1126    
1127    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1128        can match an empty string. The code to do it in pcretest and pcredemo
1129        needed fixing:
1130    
1131        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1132            only one byte after an empty string match just before CRLF (this case
1133            just got forgotten; "any" and "anycrlf" were OK).
1134    
1135        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1136            mode when an empty string match preceded an ASCII character followed by
1137            a non-ASCII character. (The code for advancing by one character rather
1138            than one byte was nonsense.)
1139    
1140        (c) The pcredemo.c sample program did not have any code at all to handle
1141            the cases when CRLF is a valid newline sequence.
1142    
1143    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1144        as a starting offset was within the subject string. There is now a new
1145        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1146        negative or greater than the length of the string. In order to test this,
1147        pcretest is extended to allow the setting of negative starting offsets.
1148    
1149    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1150        starting offset points to the beginning of a UTF-8 character was
1151        unnecessarily clumsy. I tidied it up.
1152    
1153    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1154        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1155    
1156    17. Nobody had reported that the --include_dir option, which was added in
1157        release 7.7 should have been called --include-dir (hyphen, not underscore)
1158        for compatibility with GNU grep. I have changed it to --include-dir, but
1159        left --include_dir as an undocumented synonym, and the same for
1160        --exclude-dir, though that is not available in GNU grep, at least as of
1161        release 2.5.4.
1162    
1163    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1164        characters from a string of bytes) have been redefined so as not to use
1165        loops, in order to improve performance in some environments. At the same
1166        time, I abstracted some of the common code into auxiliary macros to save
1167        repetition (this should not affect the compiled code).
1168    
1169    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1170        compile-time error is now given if \c is not followed by an ASCII
1171        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1172        different, and any byte value is allowed.)
1173    
1174    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1175        START_OPTIMIZE option, which is now allowed at compile time - but just
1176        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1177        to pcregrep and other applications that have no direct access to PCRE
1178        options. The new /Y option in pcretest sets this option when calling
1179        pcre_compile().
1180    
1181    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1182        back references. Groups containing recursive back references were forced to
1183        be atomic by that change, but in the case of named groups, the amount of
1184        memory required was incorrectly computed, leading to "Failed: internal
1185        error: code overflow". This has been fixed.
1186    
1187    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1188        pcretest.c, to avoid build problems in some Borland environments.
1189    
1190    
1191    Version 8.10 25-Jun-2010
1192    ------------------------
1193    
1194    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1195        THEN.
1196    
1197    2.  (*ACCEPT) was not working when inside an atomic group.
1198    
1199    3.  Inside a character class, \B is treated as a literal by default, but
1200        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1201        causes the error). The code is unchanged, but I tidied the documentation.
1202    
1203    4.  Inside a character class, PCRE always treated \R and \X as literals,
1204        whereas Perl faults them if its -w option is set. I have changed PCRE so
1205        that it faults them when PCRE_EXTRA is set.
1206    
1207    5.  Added support for \N, which always matches any character other than
1208        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1209    
1210    6.  When compiling pcregrep with newer versions of gcc which may have
1211        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1212        declared with attribute warn_unused_result" were given. Just casting the
1213        result to (void) does not stop the warnings; a more elaborate fudge is
1214        needed. I've used a macro to implement this.
1215    
1216    7.  Minor change to pcretest.c to avoid a compiler warning.
1217    
1218    8.  Added four artifical Unicode properties to help with an option to make
1219        \s etc use properties (see next item). The new properties are: Xan
1220        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1221    
1222    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1223        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1224        this option. Modified pcretest to add /W to test this facility. Added
1225        REG_UCP to make it available via the POSIX interface.
1226    
1227    10. Added --line-buffered to pcregrep.
1228    
1229    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1230        studied, and the match started with a letter with a code point greater than
1231        127 whose first byte was different to the first byte of the other case of
1232        the letter, the other case of this starting letter was not recognized
1233        (#976).
1234    
1235    12. If a pattern that was studied started with a repeated Unicode property
1236        test, for example, \p{Nd}+, there was the theoretical possibility of
1237        setting up an incorrect bitmap of starting bytes, but fortunately it could
1238        not have actually happened in practice until change 8 above was made (it
1239        added property types that matched character-matching opcodes).
1240    
1241    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1242        possible starting bytes for non-anchored patterns.
1243    
1244    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1245        \R, and also a number of cases that involve Unicode properties, both
1246        explicit and implicit when PCRE_UCP is set.
1247    
1248    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1249        input, it could crash or give wrong results if characters with values
1250        greater than 0xc0 were present in the subject string. (Detail: it assumed
1251        UTF-8 input when processing these items.)
1252    
1253    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1254        size_t is 64-bit (#991).
1255    
1256    17. Added a check for running out of memory when PCRE is compiled with
1257        --disable-stack-for-recursion (#990).
1258    
1259    18. If the last data line in a file for pcretest does not have a newline on
1260        the end, a newline was missing in the output.
1261    
1262    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1263        less than 128) in its various bitmaps. However, there is a facility for
1264        generating tables according to the current locale when PCRE is compiled. It
1265        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1266        characters, are recognized by isspace() and therefore were getting set in
1267        these tables, and indeed these tables seem to approximate to ISO 8859. This
1268        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1269        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1270        which of course cannot start UTF-8 characters. I have changed the code so
1271        that only real ASCII characters (less than 128) and the correct starting
1272        bytes for UTF-8 encodings are set for characters greater than 127 when in
1273        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1274        altogether.)
1275    
1276    20. Added the /T option to pcretest so as to be able to run tests with non-
1277        standard character tables, thus making it possible to include the tests
1278        used for 19 above in the standard set of tests.
1279    
1280    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1281        reference to a subpattern the other side of a comment that contains an
1282        opening parenthesis caused either an internal compiling error, or a
1283        reference to the wrong subpattern.
1284    
1285    
1286    Version 8.02 19-Mar-2010
1287    ------------------------
1288    
1289    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1290    
1291    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1292        configured.
1293    
1294    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1295        original author of that file, following a query about its status.
1296    
1297    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1298        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1299    
1300    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1301        quantifier applied to a forward-referencing subroutine call, could compile
1302        incorrect code or give the error "internal error: previously-checked
1303        referenced subpattern not found".
1304    
1305    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1306        variables to point to external functions. For these systems, therefore,
1307        pcre_malloc etc. are now initialized to local functions that call the
1308        relevant global functions.
1309    
1310    7.  There were two entries missing in the vectors called coptable and poptable
1311        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1312        I've fixed the data, and added a kludgy way of testing at compile time that
1313        the lengths are correct (equal to the number of opcodes).
1314    
1315    8.  Following on from 7, I added a similar kludge to check the length of the
1316        eint vector in pcreposix.c.
1317    
1318    9.  Error texts for pcre_compile() are held as one long string to avoid too
1319        much relocation at load time. To find a text, the string is searched,
1320        counting zeros. There was no check for running off the end of the string,
1321        which could happen if a new error number was added without updating the
1322        string.
1323    
1324    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1325    
1326    11. \K was not working if it appeared in an atomic group or in a group that
1327        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1328        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1329        the assertion is positive, but not if it is negative.
1330    
1331    12. Change 11 fortuitously reduced the size of the stack frame used in the
1332        "match()" function of pcre_exec.c by one pointer. Forthcoming
1333        implementation of support for (*MARK) will need an extra pointer on the
1334        stack; I have reserved it now, so that the stack frame size does not
1335        decrease.
1336    
1337    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1338        item in branch that calls a recursion is a subroutine call - as in the
1339        second branch in the above example - was incorrectly given the compile-
1340        time error "recursive call could loop indefinitely" because pcre_compile()
1341        was not correctly checking the subroutine for matching a non-empty string.
1342    
1343    14. The checks for overrunning compiling workspace could trigger after an
1344        overrun had occurred. This is a "should never occur" error, but it can be
1345        triggered by pathological patterns such as hundreds of nested parentheses.
1346        The checks now trigger 100 bytes before the end of the workspace.
1347    
1348    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1349    
1350    
1351    Version 8.01 19-Jan-2010
1352    ------------------------
1353    
1354    1.  If a pattern contained a conditional subpattern with only one branch (in
1355        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1356        computed the wrong minimum data length (which is of course zero for such
1357        subpatterns). This could cause incorrect "no match" results.
1358    
1359    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1360        the pattern is reset in the first branch, pcre_compile() failed with
1361        "internal error: code overflow at offset...". This happened only when
1362        the reset was to the original external option setting. (An optimization
1363        abstracts leading options settings into an external setting, which was the
1364        cause of this.)
1365    
1366    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1367        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1368        assertion pattern did not match (meaning that the assertion was true), it
1369        was incorrectly treated as false if the SKIP had been reached during the
1370        matching. This also applied to assertions used as conditions.
1371    
1372    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1373        assertion subpattern, including such a pattern used as a condition,
1374        unpredictable results occurred, instead of the error return
1375        PCRE_ERROR_DFA_UITEM.
1376    
1377    5.  The C++ GlobalReplace function was not working like Perl for the special
1378        situation when an empty string is matched. It now does the fancy magic
1379        stuff that is necessary.
1380    
1381    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1382        removed. (These were left over from very, very early versions of PCRE.)
1383    
1384    7.  Some cosmetic changes to the code to make life easier when compiling it
1385        as part of something else:
1386    
1387        (a) Change DEBUG to PCRE_DEBUG.
1388    
1389        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1390            called "current" as "current_branch", to prevent a collision with the
1391            Linux macro when compiled as a kernel module.
1392    
1393        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1394            prevent a collision with the Linux macro when compiled as a kernel
1395            module.
1396    
1397    8.  In pcre_compile() there are some checks for integer overflows that used to
1398        cast potentially large values to (double). This has been changed to that
1399        when building, a check for int64_t is made, and if it is found, it is used
1400        instead, thus avoiding the use of floating point arithmetic. (There is no
1401        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1402        double.
1403    
1404    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1405        2005 (difference between two addresses compared to an unsigned value).
1406    
1407    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1408        custom one, because of the following reported problem in Windows:
1409    
1410          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1411              under Win32.
1412          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1413              therefore missing the function definition.
1414          - The compiler thus generates a "C" signature for the test function.
1415          - The linker fails to find the "C" function.
1416          - PCRE fails to configure if asked to do so against libbz2.
1417    
1418    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1419        messages were output:
1420    
1421          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1422          rerunning libtoolize, to keep the correct libtool macros in-tree.
1423          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1424    
1425        I have done both of these things.
1426    
1427    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1428        most of the time, it *can* run out if it is given a pattern that contains a
1429        runaway infinite recursion. I updated the discussion in the pcrestack man
1430        page.
1431    
1432    13. Now that we have gone to the x.xx style of version numbers, the minor
1433        version may start with zero. Using 08 or 09 is a bad idea because users
1434        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1435        interpreted as invalid octal numbers. I've updated the previous comment in
1436        configure.ac, and also added a check that gives an error if 08 or 09 are
1437        used.
1438    
1439    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1440        causing partial matching to fail when the end of the subject matched \W
1441        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1442    
1443    15. There were some discrepancies between the declarations in pcre_internal.h
1444        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1445        their definitions. The declarations used "const uschar *" and the
1446        definitions used USPTR. Even though USPTR is normally defined as "const
1447        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1448        reported that: "This difference in casting confuses some C++ compilers, for
1449        example, SunCC recognizes above declarations as different functions and
1450        generates broken code for hbpcre." I have changed the declarations to use
1451        USPTR.
1452    
1453    16. GNU libtool is named differently on some systems. The autogen.sh script now
1454        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1455        (FreeBSD).
1456    
1457    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1458        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1459        comment: "Figure out how to create a longlong from a string: strtoll and
1460        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1461        instance, but it only takes 2 args instead of 3!"
1462    
1463    18. A subtle bug concerned with back references has been fixed by a change of
1464        specification, with a corresponding code fix. A pattern such as
1465        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1466        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1467        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1468        same bug. Such groups have to be quantified to be useful, or contained
1469        inside another quantified group. (If there's no repetition, the reference
1470        can never match.) The problem arises because, having left the group and
1471        moved on to the rest of the pattern, a later failure that backtracks into
1472        the group uses the captured value from the final iteration of the group
1473        rather than the correct earlier one. I have fixed this in PCRE by forcing
1474        any group that contains a reference to itself to be an atomic group; that
1475        is, there cannot be any backtracking into it once it has completed. This is
1476        similar to recursive and subroutine calls.
1477    
1478    
1479    Version 8.00 19-Oct-09
1480    ----------------------
1481    
1482    1.  The table for translating pcre_compile() error codes into POSIX error codes
1483        was out-of-date, and there was no check on the pcre_compile() error code
1484        being within the table. This could lead to an OK return being given in
1485        error.
1486    
1487    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1488        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1489        in a Windows environment.
1490    
1491    3.  The pcregrep --count option prints the count for each file even when it is
1492        zero, as does GNU grep. However, pcregrep was also printing all files when
1493        --files-with-matches was added. Now, when both options are given, it prints
1494        counts only for those files that have at least one match. (GNU grep just
1495        prints the file name in this circumstance, but including the count seems
1496        more useful - otherwise, why use --count?) Also ensured that the
1497        combination -clh just lists non-zero counts, with no names.
1498    
1499    4.  The long form of the pcregrep -F option was incorrectly implemented as
1500        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1501        but it seems right to fix it, and I didn't think it was worth preserving
1502        the old behaviour.
1503    
1504    5.  The command line items --regex=pattern and --regexp=pattern were not
1505        recognized by pcregrep, which required --regex pattern or --regexp pattern
1506        (with a space rather than an '='). The man page documented the '=' forms,
1507        which are compatible with GNU grep; these now work.
1508    
1509    6.  No libpcreposix.pc file was created for pkg-config; there was just
1510        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1511    
1512    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1513        when UCP support is not needed, by modifying the Python script that
1514        generates it from Unicode data files. This should not matter if the module
1515        is correctly used as a library, but I received one complaint about 50K of
1516        unwanted data. My guess is that the person linked everything into his
1517        program rather than using a library. Anyway, it does no harm.
1518    
1519    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1520        was a minimum greater than 1 for a wide character in a possessive
1521        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1522        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1523        character. Chaos in the form of incorrect output or a compiling loop could
1524        result.
1525    
1526    9.  The restrictions on what a pattern can contain when partial matching is
1527        requested for pcre_exec() have been removed. All patterns can now be
1528        partially matched by this function. In addition, if there are at least two
1529        slots in the offset vector, the offset of the earliest inspected character
1530        for the match and the offset of the end of the subject are set in them when
1531        PCRE_ERROR_PARTIAL is returned.
1532    
1533    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1534        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1535        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1536        and may be more useful for multi-segment matching.
1537    
1538    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1539        used to be given if ever the end of the subject was reached; now it is
1540        given only if matching could not proceed because another character was
1541        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1542        string "Z", which now yields "no match" instead of "partial match". In the
1543        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1544        final character ended with (*FAIL).
1545    
1546    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1547        if the pattern had a "must contain" character that was already found in the
1548        earlier partial match, unless partial matching was again requested. For
1549        example, with the pattern /dog.(body)?/, the "must contain" character is
1550        "g". If the first part-match was for the string "dog", restarting with
1551        "sbody" failed. This bug has been fixed.
1552    
1553    13. The string returned by pcre_dfa_exec() after a partial match has been
1554        changed so that it starts at the first inspected character rather than the
1555        first character of the match. This makes a difference only if the pattern
1556        starts with a lookbehind assertion or \b or \B (\K is not supported by
1557        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1558        matching functions compatible, and I think it's the right thing to do.
1559    
1560    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1561        so that the demonstration program is easily available in environments where
1562        PCRE has not been installed from source.
1563    
1564    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1565        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1566        library.
1567    
1568    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1569        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1570        is not the first non-POSIX option to be added. Clearly some people find
1571        these options useful.
1572    
1573    17. If a caller to the POSIX matching function regexec() passes a non-zero
1574        value for nmatch with a NULL value for pmatch, the value of
1575        nmatch is forced to zero.
1576    
1577    18. RunGrepTest did not have a test for the availability of the -u option of
1578        the diff command, as RunTest does. It now checks in the same way as
1579        RunTest, and also checks for the -b option.
1580    
1581    19. If an odd number of negated classes containing just a single character
1582        interposed, within parentheses, between a forward reference to a named
1583        subpattern and the definition of the subpattern, compilation crashed with
1584        an internal error, complaining that it could not find the referenced
1585        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1586        [The bug was that it was starting one character too far in when skipping
1587        over the character class, thus treating the ] as data rather than
1588        terminating the class. This meant it could skip too much.]
1589    
1590    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1591        /g option in pcretest when the pattern contains \K, which makes it possible
1592        to have an empty string match not at the start, even when the pattern is
1593        anchored. Updated pcretest and pcredemo to use this option.
1594    
1595    21. If the maximum number of capturing subpatterns in a recursion was greater
1596        than the maximum at the outer level, the higher number was returned, but
1597        with unset values at the outer level. The correct (outer level) value is
1598        now given.
1599    
1600    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1601        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1602        make it do so. The string so far is captured, making this feature
1603        compatible with Perl.
1604    
1605    23. The tests have been re-organized, adding tests 11 and 12, to make it
1606        possible to check the Perl 5.10 features against Perl 5.10.
1607    
1608    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1609        pattern matches a fixed length string. PCRE did not allow this; now it
1610        does. Neither allows recursion.
1611    
1612    25. I finally figured out how to implement a request to provide the minimum
1613        length of subject string that was needed in order to match a given pattern.
1614        (It was back references and recursion that I had previously got hung up
1615        on.) This code has now been added to pcre_study(); it finds a lower bound
1616        to the length of subject needed. It is not necessarily the greatest lower
1617        bound, but using it to avoid searching strings that are too short does give
1618        some useful speed-ups. The value is available to calling programs via
1619        pcre_fullinfo().
1620    
1621    26. While implementing 25, I discovered to my embarrassment that pcretest had
1622        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1623        study optimizations had never been tested with that matching function.
1624        Oops. What is worse, even when it was passed study data, there was a bug in
1625        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1626        were also very few tests of studied patterns with pcre_dfa_exec().
1627    
1628    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1629        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1630        on the other side of the coin, they are no longer allowed to have different
1631        names, because these cannot be distinguished in PCRE, and this has caused
1632        confusion. (This is a difference from Perl.)
1633    
1634    28. When duplicate subpattern names are present (necessarily with different
1635        numbers, as required by 27 above), and a test is made by name in a
1636        conditional pattern, either for a subpattern having been matched, or for
1637        recursion in such a pattern, all the associated numbered subpatterns are
1638        tested, and the overall condition is true if the condition is true for any
1639        one of them. This is the way Perl works, and is also more like the way
1640        testing by number works.
1641    
1642    
1643    Version 7.9 11-Apr-09
1644    ---------------------
1645    
1646    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1647        (pcretest), all targets were linked against these libraries. This included
1648        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1649        libraries. This caused unwanted dependencies to be created. This problem
1650        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1651        pcretest is linked with readline.
1652    
1653    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1654        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1655        moved outside it again, because FALSE and TRUE are already defined in AIX,
1656        but BOOL is not.
1657    
1658    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1659        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1660    
1661    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1662        hyphens) following file names and line numbers when outputting matching
1663        lines. This is not true; no spaces are inserted. I have also clarified the
1664        wording for the --colour (or --color) option.
1665    
1666    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1667        was not coloured; this is different to GNU grep, so I have changed it to be
1668        the same.
1669    
1670    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1671        each matching line was coloured. Now it goes on to look for further matches
1672        of any of the test patterns, which is the same behaviour as GNU grep.
1673    
1674    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1675        doesn't make sense to accept an empty string match in pcregrep, so I have
1676        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1677        seems to be how GNU grep behaves. [But see later change 40 for release
1678        8.33.]
1679    
1680    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1681        start or after a newline", because the conditional assertion was not being
1682        correctly handled. The rule now is that both the assertion and what follows
1683        in the first alternative must satisfy the test.
1684    
1685    9.  If auto-callout was enabled in a pattern with a conditional group whose
1686        condition was an assertion, PCRE could crash during matching, both with
1687        pcre_exec() and pcre_dfa_exec().
1688    
1689    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1690        used for matching.
1691    
1692    11. Unicode property support in character classes was not working for
1693        characters (bytes) greater than 127 when not in UTF-8 mode.
1694    
1695    12. Added the -M command line option to pcretest.
1696    
1697    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1698    
1699    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1700    
1701    16. Added comments and documentation about mis-use of no_arg in the C++
1702        wrapper.
1703    
1704    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1705        from Martin Jerabek that uses macro names for all relevant character and
1706        string constants.
1707    
1708    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1709        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1710        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1711        these, but not everybody uses configure.
1712    
1713    19. A conditional group that had only one branch was not being correctly
1714        recognized as an item that could match an empty string. This meant that an
1715        enclosing group might also not be so recognized, causing infinite looping
1716        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1717        with the subject "ab", where knowledge that the repeated group can match
1718        nothing is needed in order to break the loop.
1719    
1720    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1721        exec(), but without supplying a callout function, matching went wrong.
1722    
1723    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1724        leak if the size of the offset vector was greater than 30. When the vector
1725        is smaller, the saved offsets during recursion go onto a local stack
1726        vector, but for larger vectors malloc() is used. It was failing to free
1727        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1728        error, in fact).
1729    
1730    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1731        heapframe that is used only when UTF-8 support is enabled. This caused no
1732        problem, but was untidy.
1733    
1734    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1735        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1736        included within another project.
1737    
1738    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1739        slightly modified by me:
1740    
1741          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1742              not building pcregrep.
1743    
1744          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1745              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1746    
1747    25. Forward references, both numeric and by name, in patterns that made use of
1748        duplicate group numbers, could behave incorrectly or give incorrect errors,
1749        because when scanning forward to find the reference group, PCRE was not
1750        taking into account the duplicate group numbers. A pattern such as
1751        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1752    
1753    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1754        the feature of a custom pointer more persuasive (as requested by a user).
1755    
1756    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1757        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1758        pre-defined.
1759    
1760    28. Added support for (*UTF8) at the start of a pattern.
1761    
1762    29. Arrange for flags added by the "release type" setting in CMake to be shown
1763        in the configuration summary.
1764    
1765    
1766    Version 7.8 05-Sep-08
1767    ---------------------
1768    
1769    1.  Replaced UCP searching code with optimized version as implemented for Ad
1770        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1771        stage table and inline lookup instead of a function, giving speed ups of 2
1772        to 5 times on some simple patterns that I tested. Permission was given to
1773        distribute the MultiStage2.py script that generates the tables (it's not in
1774        the tarball, but is in the Subversion repository).
1775    
1776    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1777        scripts.
1778    
1779    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1780        a group with a zero qualifier. The result of the study could be incorrect,
1781        or the function might crash, depending on the pattern.
1782    
1783    4.  Caseless matching was not working for non-ASCII characters in back
1784        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1785        It now works when Unicode Property Support is available.
1786    
1787    5.  In pcretest, an escape such as \x{de} in the data was always generating
1788        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1789        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1790        truncation.
1791    
1792    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1793    
1794    7.  Added two (int) casts to pcregrep when printing the difference of two
1795        pointers, in case they are 64-bit values.
1796    
1797    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1798        test 2 if it fails.
1799    
1800    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1801        and a #define of that name to empty if it is not externally set. This is to
1802        allow users of MSVC to set it if necessary.
1803    
1804    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1805        the convenience functions in the pcre_get.c source file.
1806    
1807    11. An option change at the start of a pattern that had top-level alternatives
1808        could cause overwriting and/or a crash. This command provoked a crash in
1809        some environments:
1810    
1811          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1812    
1813        This potential security problem was recorded as CVE-2008-2371.
1814    
1815    12. For a pattern where the match had to start at the beginning or immediately
1816        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1817        pcre_dfa_exec() could read past the end of the passed subject if there was
1818        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1819        pcretest so that it places the subject at the end of its malloc-ed buffer.
1820    
1821    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1822        exec() might read past the end of the data buffer in UTF-8 mode.
1823    
1824    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1825        the data contained the byte 0x85 as part of a UTF-8 character within its
1826        first line. This applied both to normal and DFA matching.
1827    
1828    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1829        /^[^d]*?$/8 failed to match "abc".
1830    
1831    16. Added a missing copyright notice to pcrecpp_internal.h.
1832    
1833    17. Make it more clear in the documentation that values returned from
1834        pcre_exec() in ovector are byte offsets, not character counts.
1835    
1836    18. Tidied a few places to stop certain compilers from issuing warnings.
1837    
1838    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1839        supplied by Stefan Weber. I made a further small update for 7.8 because
1840        there is a change of source arrangements: the pcre_searchfuncs.c module is
1841        replaced by pcre_ucd.c.
1842    
1843    
1844    Version 7.7 07-May-08
1845    ---------------------
1846    
1847    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1848        a string to a long long, pretend we don't even have a long long." This is
1849        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1850    
1851    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1852        pre-7.6 versions, which defined a global no_arg variable instead of putting
1853        it in the RE class. (See also #8 below.)
1854    
1855    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1856        Lopes.
1857    
1858    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1859    
1860        (1) The include/exclude patterns were being applied to the whole pathnames
1861            of files, instead of just to the final components.
1862    
1863        (2) If there was more than one level of directory, the subdirectories were
1864            skipped unless they satisfied the include/exclude conditions. This is
1865            inconsistent with GNU grep (and could even be seen as contrary to the
1866            pcregrep specification - which I improved to make it absolutely clear).
1867            The action now is always to scan all levels of directory, and just
1868            apply the include/exclude patterns to regular files.
1869    
1870    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1871        --exclude_dir in the tests to avoid scanning .svn directories.
1872    
1873    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1874        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1875        doesn't support NULs in patterns.
1876    
1877    7.  Added some missing "const"s to declarations of static tables in
1878        pcre_compile.c and pcre_dfa_exec.c.
1879    
1880    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1881        caused by fix #2  above. (Subsequently also a second patch to fix the
1882        first patch. And a third patch - this was a messy problem.)
1883    
1884    9.  Applied Craig's patch to remove the use of push_back().
1885    
1886    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1887        matching function regexec().
1888    
1889    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1890        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1891        references. PCRE supports relative numbers with this syntax (I don't think
1892        Oniguruma does).
1893    
1894    12. Previously, a group with a zero repeat such as (...){0} was completely
1895        omitted from the compiled regex. However, this means that if the group
1896        was called as a subroutine from elsewhere in the pattern, things went wrong
1897        (an internal error was given). Such groups are now left in the compiled
1898        pattern, with a new opcode that causes them to be skipped at execution
1899        time.
1900    
1901    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1902        to the way PCRE behaves:
1903    
1904        (a) A lone ] character is dis-allowed (Perl treats it as data).
1905    
1906        (b) A back reference to an unmatched subpattern matches an empty string
1907            (Perl fails the current match path).
1908    
1909        (c) A data ] in a character class must be notated as \] because if the
1910            first data character in a class is ], it defines an empty class. (In
1911            Perl it is not possible to have an empty class.) The empty class []
1912            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1913            The negative empty class [^] matches any one character, independently
1914            of the DOTALL setting.
1915    
1916    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1917        non-existent subpattern following a character class starting with ']' and
1918        containing () gave an internal compiling error instead of "reference to
1919        non-existent subpattern". Fortunately, when the pattern did exist, the
1920        compiled code was correct. (When scanning forwards to check for the
1921        existence of the subpattern, it was treating the data ']' as terminating
1922        the class, so got the count wrong. When actually compiling, the reference
1923        was subsequently set up correctly.)
1924    
1925    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1926        it was being rejected as not supported by pcre_dfa_exec(), even though
1927        other assertions are supported. I have made pcre_dfa_exec() support
1928        (*FAIL).
1929    
1930    16. The implementation of 13c above involved the invention of a new opcode,
1931        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1932        cannot be changed at match time, I realized I could make a small
1933        improvement to matching performance by compiling OP_ALLANY instead of
1934        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1935        on the OP_ANY path.
1936    
1937    17. Compiling pcretest on Windows with readline support failed without the
1938        following two fixes: (1) Make the unistd.h include conditional on
1939        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1940    
1941    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1942        ncurses library to be included for pcretest when ReadLine support is
1943        requested, but also to allow for it to be overridden. This patch came from
1944        Daniel Bergström.
1945    
1946    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1947        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1948        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1949        spotting this.
1950    
1951    
1952    Version 7.6 28-Jan-08
1953    ---------------------
1954    
1955    1.  A character class containing a very large number of characters with
1956        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1957        overflow.
1958    
1959    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1960        HAVE_LONG_LONG is not defined.
1961    
1962    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1963        bring them up to date and include new features. This patch includes:
1964    
1965        - Fixed PH's badly added libz and libbz2 support.
1966        - Fixed a problem with static linking.
1967        - Added pcredemo. [But later removed - see 7 below.]
1968        - Fixed dftables problem and added an option.
1969        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1970            HAVE_LONG_LONG.
1971        - Added readline support for pcretest.
1972        - Added an listing of the option settings after cmake has run.
1973    
1974    4.  A user submitted a patch to Makefile that makes it easy to create
1975        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1976        Makefile.am that cause it to include this special target, without
1977        affecting anything else. Note that the same mingw target plus all
1978        the other distribution libraries and programs are now supported
1979        when configuring with CMake (see 6 below) instead of with
1980        Configure/Make.
1981    
1982    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1983        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1984        exported in the Windows port". It has not yet been confirmed that the patch
1985        solves the problem, but it does no harm.
1986    
1987    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1988        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1989        with CMake, and also correct the comment about stack recursion.
1990    
1991    7.  Remove the automatic building of pcredemo from the ./configure system and
1992        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1993        of a program that users should build themselves after PCRE is installed, so
1994        building it automatically is not really right. What is more, it gave
1995        trouble in some build environments.
1996    
1997    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1998    
1999    
2000    Version 7.5 10-Jan-08
2001    ---------------------
2002    
2003    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2004        values in parens when parsing an RE using the C++ wrapper."
2005    
2006    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2007        Characters greater than 255 were excluded from the class instead of being
2008        included.
2009    
2010    3.  The same bug as (2) above applied to negated POSIX classes such as
2011        [:^space:].
2012    
2013    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2014        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2015        I have changed it.
2016    
2017    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2018        first named subpattern) and a construct such as (?&a) would reference the
2019        first named subpattern whose name started with "a" (in other words, the
2020        length check was missing). Both these problems are fixed. "Subpattern name
2021        expected" is now given for (?&) (a zero-length name), and this patch also
2022        makes it give the same error for \k'' (previously it complained that that
2023        was a reference to a non-existent subpattern).
2024    
2025    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2026        this is right because (?- can be followed by option settings as well as by
2027        digits. I have, however, made the messages clearer.
2028    
2029    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2030        than the number used in the conditional) now cause a compile-time error.
2031        This is actually not compatible with Perl, which accepts such patterns, but
2032        treats the conditional as always being FALSE (as PCRE used to), but it
2033        seems to me that giving a diagnostic is better.
2034    
2035    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2036        and messages.
2037    
2038    9.  Fix two occurrences of "backslash" in comments that should have been
2039        "backspace".
2040    
2041    10. Remove two redundant lines of code that can never be obeyed (their function
2042        was moved elsewhere).
2043    
2044    11. The program that makes PCRE's Unicode character property table had a bug
2045        which caused it to generate incorrect table entries for sequences of
2046        characters that have the same character type, but are in different scripts.
2047        It amalgamated them into a single range, with the script of the first of
2048        them. In other words, some characters were in the wrong script. There were
2049        thirteen such cases, affecting characters in the following ranges:
2050    
2051          U+002b0 - U+002c1
2052          U+0060c - U+0060d
2053          U+0061e - U+00612
2054          U+0064b - U+0065e
2055          U+0074d - U+0076d
2056          U+01800 - U+01805
2057          U+01d00 - U+01d77
2058          U+01d9b - U+01dbf
2059          U+0200b - U+0200f
2060          U+030fc - U+030fe
2061          U+03260 - U+0327f
2062          U+0fb46 - U+0fbb1
2063          U+10450 - U+1049d
2064    
2065    12. The -o option (show only the matching part of a line) for pcregrep was not
2066        compatible with GNU grep in that, if there was more than one match in a
2067        line, it showed only the first of them. It now behaves in the same way as
2068        GNU grep.
2069    
2070    13. If the -o and -v options were combined for pcregrep, it printed a blank
2071        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2072        does the same. The return code can be used to tell if there were any
2073        non-matching lines.
2074    
2075    14. Added --file-offsets and --line-offsets to pcregrep.
2076    
2077    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2078        infinitely looping recursion. The bug was that positive lookaheads were not
2079        being skipped when checking for a possible empty match (negative lookaheads
2080        and both kinds of lookbehind were skipped).
2081    
2082    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2083        inclusion of <windows.h> to before rather than after the definition of
2084        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2085    
2086    17. Specifying a possessive quantifier with a specific limit for a Unicode
2087        character property caused pcre_compile() to compile bad code, which led at
2088        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2089        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2090        caused the error; without that there was no problem.
2091    
2092    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2093    
2094    19. Added --enable-pcretest-libreadline.
2095    
2096    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2097        RE::GlobalReplace(). As a result, the number of replacements returned was
2098        double what it should be. I removed one of the increments, but Craig sent a
2099        later patch that removed the other one (the right fix) and added unit tests
2100        that check the return values (which was not done before).
2101    
2102    21. Several CMake things:
2103    
2104        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2105            the names libpcre and libpcreposix, not just pcre and pcreposix.
2106    
2107        (2) The above change means that pcretest and pcregrep are now correctly
2108            linked with the newly-built libraries, not previously installed ones.
2109    
2110        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2111    
2112    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2113        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2114        UTF-8 newline character). The key issue is that the pattern starts .*;
2115        this means that the match must be either at the beginning, or after a
2116        newline. The bug was in the code for advancing after a failed match and
2117        checking that the new position followed a newline. It was not taking
2118        account of UTF-8 characters correctly.
2119    
2120    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2121        character classes. PCRE was not treating the sequence [:...:] as a
2122        character class unless the ... were all letters. Perl, however, seems to
2123        allow any characters between [: and :], though of course it rejects as
2124        unknown any "names" that contain non-letters, because all the known class
2125        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2126        for example, whereas PCRE did not - it did not recognize a POSIX character
2127        class. This seemed a bit dangerous, so the code has been changed to be
2128        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2129        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2130        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2131        Perl does, and where it didn't before.
2132    
2133    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2134        Windows environments %n is disabled by default.
2135    
2136    
2137    Version 7.4 21-Sep-07
2138    ---------------------
2139    
2140    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2141        means that a class such as [\s] counted as "explicit reference to CR or
2142        LF". That isn't really right - the whole point of the change was to try to
2143        help when there was an actual mention of one of the two characters. So now
2144        the change happens only if \r or \n (or a literal CR or LF) character is
2145        encountered.
2146    
2147    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2148        of both had grown to the point where there were only 3 bits left.
2149        Fortunately, there was spare space in the data structure, and so I have
2150        moved the internal flags into a new 16-bit field to free up more option
2151        bits.
2152    
2153    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2154        but did not set the internal JCHANGED flag - either of these is enough to
2155        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2156        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2157        start sets both bits.
2158    
2159    4.  Added options (at build time, compile time, exec time) to change \R from
2160        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2161    
2162    5.  doc/pcresyntax.html was missing from the distribution.
2163    
2164    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2165        compatibility, even though it is no longer used.
2166    
2167    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2168        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2169        windows.h file is present (where different names are used). [This was
2170        reversed later after testing - see 16 below.]
2171    
2172    8.  Changed all #include <config.h> to #include "config.h". There were also
2173        some further <pcre.h> cases that I changed to "pcre.h".
2174    
2175    9.  When pcregrep was used with the --colour option, it missed the line ending
2176        sequence off the lines that it output.
2177    
2178    10. It was pointed out to me that arrays of string pointers cause lots of
2179        relocations when a shared library is dynamically loaded. A technique of
2180        using a single long string with a table of offsets can drastically reduce
2181        these. I have refactored PCRE in four places to do this. The result is
2182        dramatic:
2183    
2184          Originally:                          290
2185          After changing UCP table:            187
2186          After changing error message table:   43
2187          After changing table of "verbs"       36
2188          After changing table of Posix names   22
2189    
2190        Thanks to the folks working on Gregex for glib for this insight.
2191    
2192    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2193        unicode-properties was also set.
2194    
2195    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2196    
2197    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2198        checked only for CRLF.
2199    
2200    14. Added casts to pcretest.c to avoid compiler warnings.
2201    
2202    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2203    
2204    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2205        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2206        entirely. This removes changes made in 7 above.
2207    
2208    17. The CMake files have been updated, and there is now more information about
2209        building with CMake in the NON-UNIX-USE document.
2210    
2211    
2212    Version 7.3 28-Aug-07
2213    ---------------------
2214    
2215     1. In the rejigging of the build system that eventually resulted in 7.1, the
2216        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2217        brackets there is not right, since it causes compilers to look for an
2218        installed pcre.h, not the version that is in the source that is being
2219        compiled (which of course may be different). I have changed it back to:
2220    
2221          #include "pcre.h"
2222    
2223        I have a vague recollection that the change was concerned with compiling in
2224        different directories, but in the new build system, that is taken care of
2225        by the VPATH setting the Makefile.
2226    
2227     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2228        when the subject happened to end in the byte 0x85 (e.g. if the last
2229        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2230        characters but of course it shouldn't be taken as a newline when it is part
2231        of another character. The bug was that, for an unlimited repeat of . in
2232        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2233        characters when looking for a newline.
2234    
2235     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2236    
2237     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2238        in debug output.
2239    
2240     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2241        long printing in the pcrecpp unittest when running under MinGW.
2242    
2243     6. ESC_K was left out of the EBCDIC table.
2244    
2245     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2246        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2247        limit also applies to "virtual nesting" when a pattern is recursive, and in
2248        this case 1000 isn't so big. I have been able to remove this limit at the
2249        expense of backing off one optimization in certain circumstances. Normally,
2250        when pcre_exec() would call its internal match() function recursively and
2251        immediately return the result unconditionally, it uses a "tail recursion"
2252        feature to save stack. However, when a subpattern that can match an empty
2253        string has an unlimited repetition quantifier, it no longer makes this
2254        optimization. That gives it a stack frame in which to save the data for
2255        checking that an empty string has been matched. Previously this was taken
2256        from the 1000-entry workspace that had been reserved. So now there is no
2257        explicit limit, but more stack is used.
2258    
2259     8. Applied Daniel's patches to solve problems with the import/export magic
2260        syntax that is required for Windows, and which was going wrong for the
2261        pcreposix and pcrecpp parts of the library. These were overlooked when this
2262        problem was solved for the main library.
2263    
2264     9. There were some crude static tests to avoid integer overflow when computing
2265        the size of patterns that contain repeated groups with explicit upper
2266        limits. As the maximum quantifier is 65535, the maximum group length was
2267        set at 30,000 so that the product of these two numbers did not overflow a
2268        32-bit integer. However, it turns out that people want to use groups that
2269        are longer than 30,000 bytes (though not repeat them that many times).
2270        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2271        made it possible to implement the integer overflow checks in a much more
2272        dynamic way, which I have now done. The artificial limitation on group
2273        length has been removed - we now have only the limit on the total length of
2274        the compiled pattern, which depends on the LINK_SIZE setting.
2275    
2276    10. Fixed a bug in the documentation for get/copy named substring when
2277        duplicate names are permitted. If none of the named substrings are set, the
2278        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2279        empty string.
2280    
2281    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2282        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2283        because the ] is interpreted as the first data character and the
2284        terminating ] is not found. PCRE has been made compatible with Perl in this
2285        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2286        cause memory overwriting.
2287    
2288    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2289        string has been matched (to stop an infinite loop). It was not recognizing
2290        a conditional subpattern that could match an empty string if that
2291        subpattern was within another subpattern. For example, it looped when
2292        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2293        condition was not nested. This bug has been fixed.
2294    
2295    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2296        past the start of the subject in the presence of bytes with the top bit
2297        set, for example "\x8aBCD".
2298    
2299    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2300        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2301    
2302    14. Optimized (?!) to (*FAIL).
2303    
2304    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2305        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2306        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2307        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2308        does: it's just the validity check that is more restrictive.
2309    
2310    16. Inserted checks for integer overflows during escape sequence (backslash)
2311        processing, and also fixed erroneous offset values for syntax errors during
2312        backslash processing.
2313    
2314    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2315        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2316    
2317    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2318        caused an overrun.
2319    
2320    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2321        something other than just ASCII characters) inside a group that had an
2322        unlimited repeat caused a loop at compile time (while checking to see
2323        whether the group could match an empty string).
2324    
2325    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2326        [\P{Any}] did so. (Error in the code for printing property names.)
2327    
2328    21. An orphan \E inside a character class could cause a crash.
2329    
2330    22. A repeated capturing bracket such as (A)? could cause a wild memory
2331        reference during compilation.
2332    
2333    23. There are several functions in pcre_compile() that scan along a compiled
2334        expression for various reasons (e.g. to see if it's fixed length for look
2335        behind). There were bugs in these functions when a repeated \p or \P was
2336        present in the pattern. These operators have additional parameters compared
2337        with \d, etc, and these were not being taken into account when moving along
2338        the compiled data. Specifically:
2339    
2340        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2341            length.
2342    
2343        (b) An item such as \pL+ within a repeated group could cause crashes or
2344            loops.
2345    
2346        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2347            "reference to non-existent subpattern" error.
2348    
2349        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2350    
2351    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2352        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2353    
2354    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2355    
2356    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2357        character were causing crashes (broken optimization).
2358    
2359    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2360        \p or \P) caused a compile-time loop.
2361    
2362    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2363        break. For example, the unstudied pattern [\r\n]A does not match the string
2364        "\r\nA" because change 7.0/46 below moves the current point on by two
2365        characters after failing to match at the start. However, the pattern \nA
2366        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2367        the same is true. There doesn't seem any very clean way out of this, but
2368        what I have chosen to do makes the common cases work: PCRE now takes note
2369        of whether there can be an explicit match for \r or \n anywhere in the
2370        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2371        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2372        pattern has explicit CR or LF references.
2373    
2374    29. Added (*CR) etc for changing newline setting at start of pattern.
2375    
2376    
2377    Version 7.2 19-Jun-07
2378    ---------------------
2379    
2380     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2381        which is apparently normally available under Windows.
2382    
2383     2. Re-jig the pcregrep tests with different newline settings in an attempt
2384        to make them independent of the local environment's newline setting.
2385    
2386     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2387    
2388     4. Some of the "internals" tests were previously cut out when the link size
2389        was not 2, because the output contained actual offsets. The recent new
2390        "Z" feature of pcretest means that these can be cut out, making the tests
2391        usable with all link sizes.
2392    
2393     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2394        stack recursion. This gives a massive performance boost under BSD, but just
2395        a small improvement under Linux. However, it saves one field in the frame
2396        in all cases.
2397    
2398     6. Added more features from the forthcoming Perl 5.10:
2399    
2400        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2401            recursion call. It refers to the nth most recently opened parentheses.
2402    
2403        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2404            to be opened parentheses.
2405    
2406        (c) Conditions that refer to capturing parentheses can be specified
2407            relatively, for example, (?(-2)... or (?(+3)...
2408    
2409        (d) \K resets the start of the current match so that everything before
2410            is not part of it.
2411    
2412        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2413    
2414        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2415            reference syntax.
2416    
2417        (g) (?| introduces a group in which the numbering of parentheses in each
2418            alternative starts with the same number.
2419    
2420        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2421    
2422     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2423        PCRE_INFO_JCHANGED.
2424    
2425     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2426        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2427        for detecting groups that can match an empty string.
2428    
2429     9. A pattern with a very large number of alternatives (more than several
2430        hundred) was running out of internal workspace during the pre-compile
2431        phase, where pcre_compile() figures out how much memory will be needed. A
2432        bit of new cunning has reduced the workspace needed for groups with
2433        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2434        workspace instead of running out of the 4096 that are available.
2435    
2436    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2437    
2438    11. Applied patch from Google to remove an optimization that didn't quite work.
2439        The report of the bug said:
2440    
2441          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2442          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2443          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2444    
2445    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2446        it matched the wrong number of bytes.
2447    
2448    
2449    Version 7.1 24-Apr-07
2450    ---------------------
2451    
2452     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2453        that is more "standard", making use of automake and other Autotools. There
2454        is some re-arrangement of the files and adjustment of comments consequent
2455        on this.
2456    
2457     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2458        for recursive directory scanning broke on some systems because the files
2459        are not scanned in any specific order and on different systems the order
2460        was different. A call to "sort" has been inserted into RunGrepTest for the
2461        approprate test as a short-term fix. In the longer term there may be an
2462        alternative.
2463    
2464     3. I had an email from Eric Raymond about problems translating some of PCRE's
2465        man pages to HTML (despite the fact that I distribute HTML pages, some
2466        people do their own conversions for various reasons). The problems
2467        concerned the use of low-level troff macros .br and .in. I have therefore
2468        removed all such uses from the man pages (some were redundant, some could
2469        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2470        HTML has been updated to handle .nf/.fi and to complain if it encounters
2471        .br or .in.
2472    
2473     4. Updated comments in configure.ac that get placed in config.h.in and also
2474        arranged for config.h to be included in the distribution, with the name
2475        config.h.generic, for the benefit of those who have to compile without
2476        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2477    
2478     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2479        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2480        makevp.bat for the current PCRE, using the additional files
2481        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2482    
2483     6. A Windows user reported a minor discrepancy with test 2, which turned out
2484        to be caused by a trailing space on an input line that had got lost in his
2485        copy. The trailing space was an accident, so I've just removed it.
2486    
2487     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2488        that is needed.
2489    
2490     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2491        as "const" (a) because they are and (b) because it helps the PHP
2492        maintainers who have recently made a script to detect big data structures
2493        in the php code that should be moved to the .rodata section. I remembered
2494        to update Builducptable as well, so it won't revert if ucptable.h is ever
2495        re-created.
2496    
2497     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2498        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2499        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2500        support is not required. This saves 1.5-2K of code, which is important in
2501        some applications.
2502    
2503        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2504        so as not to refer to the tables, even though these functions will never be
2505        called when UTF-8 support is disabled. Otherwise there are problems with a
2506        shared library.
2507    
2508    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2509    
2510        (a) It was defining its arguments as char * instead of void *.
2511    
2512        (b) It was assuming that all moves were upwards in memory; this was true
2513            a long time ago when I wrote it, but is no longer the case.
2514    
2515        The emulated memove() is provided for those environments that have neither
2516        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2517        is clearly not the case, as these two bugs were recently reported.
2518    
2519    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2520        and Detrail to create the HTML documentation, the .txt form of the man
2521        pages, and it removes trailing spaces from listed files. It also creates
2522        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2523        case, it wraps all the #defines with #ifndefs. This script should be run
2524        before "make dist".
2525    
2526    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2527        with Unicode property support.
2528    
2529        (a) For a maximizing quantifier, if the two different cases of the
2530            character were of different lengths in their UTF-8 codings (there are
2531            some cases like this - I found 11), and the matching function had to
2532            back up over a mixture of the two cases, it incorrectly assumed they
2533            were both the same length.
2534    
2535        (b) When PCRE was configured to use the heap rather than the stack for
2536            recursion during matching, it was not correctly preserving the data for
2537            the other case of a UTF-8 character when checking ahead for a match
2538            while processing a minimizing repeat. If the check also involved
2539            matching a wide character, but failed, corruption could cause an
2540            erroneous result when trying to check for a repeat of the original
2541            character.
2542    
2543    13. Some tidying changes to the testing mechanism:
2544    
2545        (a) The RunTest script now detects the internal link size and whether there
2546            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2547            values substituted by "configure". (The RunGrepTest script already did
2548            this for UTF-8.) The configure.ac script no longer substitutes the
2549            relevant variables.
2550    
2551        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2552            with length and offset values. This means that the output is different
2553            for different internal link sizes. Test 2 is skipped for link sizes
2554            other than 2 because of this, bypassing the problem. Unfortunately,
2555            there was also a test in test 3 (the locale tests) that used /B and
2556            failed for link sizes other than 2. Rather than cut the whole test out,
2557            I have added a new /Z option to pcretest that replaces the length and
2558            offset values with spaces. This is now used to make test 3 independent
2559            of link size. (Test 2 will be tidied up later.)
2560    
2561    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2562        segmentation fault instead of returning the appropriate error message.
2563    
2564    15. In multiline mode when the newline sequence was set to "any", the pattern
2565        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2566        This doesn't seem right; it now treats the CRLF combination as the line
2567        ending, and so does not match in that case. It's only a pattern such as ^$
2568        that would hit this one: something like ^ABC$ would have failed after \r
2569        and then tried again after \r\n.
2570    
2571    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2572        in an attempt to make files that differ only in their line terminators
2573        compare equal. This works on Linux.
2574    
2575    17. Under certain error circumstances pcregrep might try to free random memory
2576        as it exited. This is now fixed, thanks to valgrind.
2577    
2578    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2579        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2580        was because its rules for how to advance for /g after matching an empty
2581        string at the end of a line did not allow for this case. They now check for
2582        it specially.
2583    
2584    20. pcretest is supposed to handle patterns and data of any length, by
2585        extending its buffers when necessary. It was getting this wrong when the
2586        buffer for a data line had to be extended.
2587    
2588    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2589        CRLF as a newline sequence.
2590    
2591    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2592        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2593        I have nevertheless tidied it up.
2594    
2595    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2596    
2597    24. Added a man page for pcre-config.
2598    
2599    
2600    Version 7.0 19-Dec-06
2601    ---------------------
2602    
2603     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2604        moving to gcc 4.1.1.
2605    
2606     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2607        sys/time.h, which is documented as needed for this function. It doesn't
2608        seem to matter on Linux, but it showed up on some releases of OS X.
2609    
2610     3. It seems that there are systems where bytes whose values are greater than
2611        127 match isprint() in the "C" locale. The "C" locale should be the
2612        default when a C program starts up. In most systems, only ASCII printing
2613        characters match isprint(). This difference caused the output from pcretest
2614        to vary, making some of the tests fail. I have changed pcretest so that:
2615    
2616        (a) When it is outputting text in the compiled version of a pattern, bytes
2617            other than 32-126 are always shown as hex escapes.
2618    
2619        (b) When it is outputting text that is a matched part of a subject string,
2620            it does the same, unless a different locale has been set for the match
2621            (using the /L modifier). In this case, it uses isprint() to decide.
2622    
2623     4. Fixed a major bug that caused incorrect computation of the amount of memory
2624        required for a compiled pattern when options that changed within the
2625        pattern affected the logic of the preliminary scan that determines the
2626        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2627        that the computed length was too small. The symptoms of this bug were
2628        either the PCRE error "internal error: code overflow" from pcre_compile(),
2629        or a glibc crash with a message such as "pcretest: free(): invalid next
2630        size (fast)". Examples of patterns that provoked this bug (shown in
2631        pcretest format) are:
2632    
2633          /(?-x: )/x
2634          /(?x)(?-x: \s*#\s*)/
2635          /((?i)[\x{c0}])/8
2636          /(?i:[\x{c0}])/8
2637    
2638        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2639        is now done differently.
2640    
2641     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2642        wrapper classes; (b) implement a new function in the C++ scanner that is
2643        more efficient than the old way of doing things because it avoids levels of
2644        recursion in the regex matching; (c) add a paragraph to the documentation
2645        for the FullMatch() function.
2646    
2647     6. The escape sequence \n was being treated as whatever was defined as
2648        "newline". Not only was this contrary to the documentation, which states
2649        that \n is character 10 (hex 0A), but it also went horribly wrong when
2650        "newline" was defined as CRLF. This has been fixed.
2651    
2652     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2653        was being set to -1 for the "end of line" case (supposedly a value that no
2654        character can have). Though this value is never used (the check for end of
2655        line is "zero bytes in current character"), it caused compiler complaints.
2656        I've changed it to 0xffffffff.
2657    
2658     8. In pcre_version.c, the version string was being built by a sequence of
2659        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2660        string (as it is for production releases) called a macro with an empty
2661        argument. The C standard says the result of this is undefined. The gcc
2662        compiler treats it as an empty string (which was what was wanted) but it is
2663        reported that Visual C gives an error. The source has been hacked around to
2664        avoid this problem.
2665    
2666     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2667        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2668        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2669        of them did).
2670    
2671    10. Originally, pcretest opened its input and output without "b"; then I was
2672        told that "b" was needed in some environments, so it was added for release
2673        5.0 to both the input and output. (It makes no difference on Unix-like
2674        systems.) Later I was told that it is wrong for the input on Windows. I've
2675        now abstracted the modes into two macros, to make it easier to fiddle with
2676        them, and removed "b" from the input mode under Windows.
2677    
2678    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2679    
2680    12. Added -help and --help to pcretest as an official way of being reminded
2681        of the options.
2682    
2683    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2684        and pcrecpp.cc because they annoy compilers at high warning levels.
2685    
2686    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2687    
2688    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2689        scripts are not C programs :-) and which was not noticed because it works
2690        on Linux.
2691    
2692    16. pcretest is supposed to handle any length of pattern and data line (as one
2693        line or as a continued sequence of lines) by extending its input buffer if
2694        necessary. This feature was broken for very long pattern lines, leading to
2695        a string of junk being passed to pcre_compile() if the pattern was longer
2696        than about 50K.
2697    
2698    17. I have done a major re-factoring of the way pcre_compile() computes the
2699        amount of memory needed for a compiled pattern. Previously, there was code
2700        that made a preliminary scan of the pattern in order to do this. That was
2701        OK when PCRE was new, but as the facilities have expanded, it has become
2702        harder and harder to keep it in step with the real compile phase, and there
2703        have been a number of bugs (see for example, 4 above). I have now found a
2704        cunning way of running the real compile function in a "fake" mode that
2705        enables it to compute how much memory it would need, while actually only
2706        ever using a few hundred bytes of working memory and without too many
2707        tests of the mode. This should make future maintenance and development
2708        easier. A side effect of this work is that the limit of 200 on the nesting
2709        depth of parentheses has been removed (though this was never a serious
2710        limitation, I suspect). However, there is a downside: pcre_compile() now
2711        runs more slowly than before (30% or more, depending on the pattern). I
2712        hope this isn't a big issue. There is no effect on runtime performance.
2713    
2714    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2715        newline (only possible for the last line of a file) and it was a
2716        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2717    
2718    19. Added additional timing features to pcretest. (1) The -tm option now times
2719        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2720        separate command line item, by a number that specifies the number of
2721        repeats to use when timing. The default is 50000; this gives better
2722        precision, but takes uncomfortably long for very large patterns.
2723    
2724    20. Extended pcre_study() to be more clever in cases where a branch of a
2725        subpattern has no definite first character. For example, (a*|b*)[cd] would
2726        previously give no result from pcre_study(). Now it recognizes that the
2727        first character must be a, b, c, or d.
2728    
2729    21. There was an incorrect error "recursive call could loop indefinitely" if
2730        a subpattern (or the entire pattern) that was being tested for matching an
2731        empty string contained only one non-empty item after a nested subpattern.
2732        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2733        incorrectly, because the \d was being skipped in the check.
2734    
2735    22. The pcretest program now has a new pattern option /B and a command line
2736        option -b, which is equivalent to adding /B to every pattern. This causes
2737        it to show the compiled bytecode, without the additional information that
2738        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2739        is the same as /B/I).
2740    
2741    23. A new optimization is now able automatically to treat some sequences such
2742        as a*b as a*+b. More specifically, if something simple (such as a character
2743        or a simple class like \d) has an unlimited quantifier, and is followed by
2744        something that cannot possibly match the quantified thing, the quantifier
2745        is automatically "possessified".
2746    
2747    24. A recursive reference to a subpattern whose number was greater than 39
2748        went wrong under certain circumstances in UTF-8 mode. This bug could also
2749        have affected the operation of pcre_study().
2750    
2751    25. Realized that a little bit of performance could be had by replacing
2752        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2753    
2754    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2755    
2756    27. Possessive quantifiers such as a++ were previously implemented by turning
2757        them into atomic groups such as ($>a+). Now they have their own opcodes,
2758        which improves performance. This includes the automatically created ones
2759        from 23 above.
2760    
2761    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2762        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2763        the first matched character to be a colon. This applied both to named and
2764        numbered groups.
2765    
2766    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2767    
2768    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2769        building PCRE on Windows easier, so I have included it in the distribution.
2770    
2771    31. There is now a check in pcretest against a ridiculously large number being
2772        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2773        loop, the loop is abandoned.
2774    
2775    32. Forward references to subpatterns in conditions such as (?(2)...) where
2776        subpattern 2 is defined later cause pcre_compile() to search forwards in
2777        the pattern for the relevant set of parentheses. This search went wrong
2778        when there were unescaped parentheses in a character class, parentheses
2779        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2780    
2781    33. "Subroutine" calls and backreferences were previously restricted to
2782        referencing subpatterns earlier in the regex. This restriction has now
2783        been removed.
2784    
2785    34. Added a number of extra features that are going to be in Perl 5.10. On the
2786        whole, these are just syntactic alternatives for features that PCRE had
2787        previously implemented using the Python syntax or my own invention. The
2788        other formats are all retained for compatibility.
2789    
2790        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2791            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2792            also .NET compatible.
2793    
2794        (b) A recursion or subroutine call to a named group can now be defined as
2795            (?&name) as well as (?P>name).
2796    
2797        (c) A backreference to a named group can now be defined as \k<name> or
2798            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2799            5.10, are also .NET compatible.
2800    
2801        (d) A conditional reference to a named group can now use the syntax
2802            (?(<name>) or (?('name') as well as (?(name).
2803    
2804        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2805            groups (named and numbered) that are never evaluated inline, but can be
2806            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2807            is always false. There may be only one alternative in such a group.
2808    
2809        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2810            as the simple (?(R). The condition is true only if the most recent
2811            recursion is that of the given number or name. It does not search out
2812            through the entire recursion stack.
2813    
2814        (g) The escape \gN or \g{N} has been added, where N is a positive or
2815            negative number, specifying an absolute or relative reference.
2816    
2817    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2818        some "unreachable code" warnings.
2819    
2820    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2821        things, this adds five new scripts.
2822    
2823    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2824        There were also incompatibilities regarding the handling of \Q..\E inside
2825        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2826        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2827    
2828    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2829        matches an empty string, and forcibly breaks the loop. There were bugs in
2830        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2831        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2832        separate and independent bugs (that affected different cases) have been
2833        fixed.
2834    
2835    39. Refactored the code to abolish the use of different opcodes for small
2836        capturing bracket numbers. This is a tidy that I avoided doing when I
2837        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2838        The new approach is not only tidier, it makes it possible to reduce the
2839        memory needed to fix the previous bug (38).
2840    
2841    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2842        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2843        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2844        mode.
2845    
2846    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2847        report.
2848    
2849    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2850        copy construction and assignment in the C++ wrapper.
2851    
2852    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2853        couple of bugs that could have given wrong results in the "--newline=crlf"
2854        case.
2855    
2856    44. Added a number of casts and did some reorganization of signed/unsigned int
2857        variables following suggestions from Dair Grant. Also renamed the variable
2858        "this" as "item" because it is a C++ keyword.
2859    
2860    45. Arranged for dftables to add
2861    
2862          #include "pcre_internal.h"
2863    
2864        to pcre_chartables.c because without it, gcc 4.x may remove the array
2865        definition from the final binary if PCRE is built into a static library and
2866        dead code stripping is activated.
2867    
2868    46. For an unanchored pattern, if a match attempt fails at the start of a
2869        newline sequence, and the newline setting is CRLF or ANY, and the next two
2870        characters are CRLF, advance by two characters instead of one.
2871    
2872    
2873    Version 6.7 04-Jul-06
2874    ---------------------
2875    
2876     1. In order to handle tests when input lines are enormously long, pcretest has
2877        been re-factored so that it automatically extends its buffers when
2878        necessary. The code is crude, but this _is_ just a test program. The
2879        default size has been increased from 32K to 50K.
2880    
2881     2. The code in pcre_study() was using the value of the re argument before
2882        testing it for NULL. (Of course, in any sensible call of the function, it
2883        won't be NULL.)
2884    
2885     3. The memmove() emulation function in pcre_internal.h, which is used on
2886        systems that lack both memmove() and bcopy() - that is, hardly ever -
2887        was missing a "static" storage class specifier.
2888    
2889     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2890        containing an extended class (one that cannot be represented by a bitmap
2891        because it contains high-valued characters or Unicode property items, e.g.
2892        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2893        pattern, but PCRE should not loop if you do not (it no longer does).
2894        [Detail: two cases were found: (a) a repeated subpattern containing an
2895        extended class; (b) a recursive reference to a subpattern that followed a
2896        previous extended class. It wasn't skipping over the extended class
2897        correctly when UTF-8 mode was not set.]
2898    
2899     5. A negated single-character class was not being recognized as fixed-length
2900        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2901        compile error "lookbehind assertion is not fixed length".
2902    
2903     6. The RunPerlTest auxiliary script was showing an unexpected difference
2904        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2905        write a Perl script that can interpret lines of an input file either as
2906        byte characters or as UTF-8, which is what "perltest" was being required to
2907        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2908        can't do is switch easily at run time between having the "use utf8;" pragma
2909        or not. In the end, I fudged it by using the RunPerlTest script to insert
2910        "use utf8;" explicitly for the UTF-8 tests.
2911    
2912     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2913        the end of the subject string, contrary to the documentation and to what
2914        Perl does. This was true of both matching functions. Now it matches only at
2915        the start of the subject and immediately after *internal* newlines.
2916    
2917     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2918        a pointer to an int instead of a pointer to an unsigned long int. This
2919        caused problems on 64-bit systems.
2920    
2921     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2922        instance of the 'standard' template library not being so standard".
2923    
2924    10. There was no check on the number of named subpatterns nor the maximum
2925        length of a subpattern name. The product of these values is used to compute
2926        the size of the memory block for a compiled pattern. By supplying a very
2927        long subpattern name and a large number of named subpatterns, the size
2928        computation could be caused to overflow. This is now prevented by limiting
2929        the length of names to 32 characters, and the number of named subpatterns
2930        to 10,000.
2931    
2932    11. Subpatterns that are repeated with specific counts have to be replicated in
2933        the compiled pattern. The size of memory for this was computed from the
2934        length of the subpattern and the repeat count. The latter is limited to
2935        65535, but there was no limit on the former, meaning that integer overflow
2936        could in principle occur. The compiled length of a repeated subpattern is
2937        now limited to 30,000 bytes in order to prevent this.
2938    
2939    12. Added the optional facility to have named substrings with the same name.
2940    
2941    13. Added the ability to use a named substring as a condition, using the
2942        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2943        are numbers (not recommended). Forward references are permitted.
2944    
2945    14. Added forward references in named backreferences (if you see what I mean).
2946    
2947    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2948        pattern could run off the end of the subject. For example, the pattern
2949        "(?s)(.{1,5})"8 did this with the subject "ab".
2950    
2951    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2952        PCRE_CASELESS was set when matching characters that were quantified with ?
2953        or *.
2954    
2955    17. A character class other than a single negated character that had a minimum
2956        but no maximum quantifier - for example [ab]{6,} - was not handled
2957        correctly by pce_dfa_exec(). It would match only one character.
2958    
2959    18. A valid (though odd) pattern that looked like a POSIX character
2960        class but used an invalid character after [ (for example [[,abc,]]) caused
2961        pcre_compile() to give the error "Failed: internal error: code overflow" or
2962        in some cases to crash with a glibc free() error. This could even happen if
2963        the pattern terminated after [[ but there just happened to be a sequence of
2964        letters, a binary zero, and a closing ] in the memory that followed.
2965    
2966    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2967        over the years. Originally (before any Unicode support), just the bottom 8
2968        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2969        output from "man perlunicode" includes this:
2970    
2971          The regular expression compiler produces polymorphic opcodes.  That
2972          is, the pattern adapts to the data and automatically switches to
2973          the Unicode character scheme when presented with Unicode data--or
2974          instead uses a traditional byte scheme when presented with byte
2975          data.
2976    
2977        Sadly, a wide octal escape does not cause a switch, and in a string with
2978        no other multibyte characters, these octal escapes are treated as before.
2979        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2980        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2981        Unicode string.
2982    
2983        I have not perpetrated such confusion in PCRE. Up till now, it took just
2984        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2985        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2986        translate to the appropriate multibyte character.
2987    
2988    29. Applied some refactoring to reduce the number of warnings from Microsoft
2989        and Borland compilers. This has included removing the fudge introduced
2990        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2991        a warning about an unused variable.
2992    
2993    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2994        characters since release 4.0, because Perl (from release 5.004) does not.
2995        [Or at least, is documented not to: some releases seem to be in conflict
2996        with the documentation.] However, when a pattern was studied with
2997        pcre_study() and all its branches started with \s, PCRE still included VT
2998        as a possible starting character. Of course, this did no harm; it just
2999        caused an unnecessary match attempt.
3000    
3001    22. Removed a now-redundant internal flag bit that recorded the fact that case
3002        dependency changed within the pattern. This was once needed for "required
3003        byte" processing, but is no longer used. This recovers a now-scarce options
3004        bit. Also moved the least significant internal flag bit to the most-
3005        significant bit of the word, which was not previously used (hangover from
3006        the days when it was an int rather than a uint) to free up another bit for
3007        the future.
3008    
3009    23. Added support for CRLF line endings as well as CR and LF. As well as the
3010        default being selectable at build time, it can now be changed at runtime
3011        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
3012        specify that it is scanning data with non-default line endings.
3013    
3014    24. Changed the definition of CXXLINK to make it agree with the definition of
3015        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
3016    
3017    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
3018        recursions. This makes a big different to stack usage for some patterns.
3019    
3020    26. If a subpattern containing a named recursion or subroutine reference such
3021        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
3022        the space required for the compiled pattern went wrong and gave too small a
3023        value. Depending on the environment, this could lead to "Failed: internal
3024        error: code overflow at offset 49" or "glibc detected double free or
3025        corruption" errors.
3026    
3027    27. Applied patches from Google (a) to support the new newline modes and (b) to
3028        advance over multibyte UTF-8 characters in GlobalReplace.
3029    
3030    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
3031        difference for some implementation of PCRE in some Windows version.
3032    
3033    29. Added some extra testing facilities to pcretest:
3034    
3035        \q<number>   in a data line sets the "match limit" value
3036        \Q<number>   in a data line sets the "match recursion limt" value
3037        -S <number>  sets the stack size, where <number> is in megabytes
3038    
3039        The -S option isn't available for Windows.
3040    
3041    
3042    Version 6.6 06-Feb-06
3043    ---------------------
3044    
3045     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
3046        in pcreposix.h. I have copied the definition from pcre.h.
3047    
3048     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
3049        because pcre.h is no longer a built file.
3050    
3051     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
3052        not normally included in the compiled code.
3053    
3054    
3055    Version 6.5 01-Feb-06
3056    ---------------------
3057    
3058     1. When using the partial match feature with pcre_dfa_exec(), it was not
3059        anchoring the second and subsequent partial matches at the new starting
3060        point. This could lead to incorrect results. For example, with the pattern
3061        /1234/, partially matching against "123" and then "a4" gave a match.
3062    
3063     2. Changes to pcregrep:
3064    
3065        (a) All non-match returns from pcre_exec() were being treated as failures
3066            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
3067            error message is output. Some extra information is given for the
3068            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
3069            probably the only errors that are likely to be caused by users (by
3070            specifying a regex that has nested indefinite repeats, for instance).
3071            If there are more than 20 of these errors, pcregrep is abandoned.
3072    
3073        (b) A binary zero was treated as data while matching, but terminated the
3074            output line if it was written out. This has been fixed: binary zeroes
3075            are now no different to any other data bytes.
3076    
3077        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
3078            used to set a locale for matching. The --locale=xxxx long option has
3079            been added (no short equivalent) to specify a locale explicitly on the
3080            pcregrep command, overriding the environment variables.
3081    
3082        (d) When -B was used with -n, some line numbers in the output were one less
3083            than they should have been.
3084    
3085        (e) Added the -o (--only-matching) option.
3086    
3087        (f) If -A or -C was used with -c (count only), some lines of context were
3088            accidentally printed for the final match.
3089    
3090        (g) Added the -H (--with-filename) option.
3091    
3092        (h) The combination of options -rh failed to suppress file names for files
3093            that were found from directory arguments.
3094    
3095        (i) Added the -D (--devices) and -d (--directories) options.
3096    
3097        (j) Added the -F (--fixed-strings) option.
3098    
3099        (k) Allow "-" to be used as a file name for -f as well as for a data file.
3100    
3101        (l) Added the --colo(u)r option.
3102    
3103        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
3104            is not present by default.
3105    
3106     3. A nasty bug was discovered in the handling of recursive patterns, that is,
3107        items such as (?R) or (?1), when the recursion could match a number of
3108        alternatives. If it matched one of the alternatives, but subsequently,
3109        outside the recursion, there was a failure, the code tried to back up into
3110        the recursion. However, because of the way PCRE is implemented, this is not
3111        possible, and the result was an incorrect result from the match.
3112    
3113        In order to prevent this happening, the specification of recursion has
3114        been changed so that all such subpatterns are automatically treated as
3115        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
3116    
3117     4. I had overlooked the fact that, in some locales, there are characters for
3118        which isalpha() is true but neither isupper() nor islower() are true. In
3119        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
3120        and ordfeminine) are like this. This affected the treatment of \w and \W
3121        when they appeared in character classes, but not when they appeared outside
3122        a character class. The bit map for "word" characters is now created
3123        separately from the results of isalnum() instead of just taking it from the
3124        upper, lower, and digit maps. (Plus the underscore character, of course.)
3125    
3126     5. The above bug also affected the handling of POSIX character classes such as
3127        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
3128        permanent tables. Instead, the bit maps for such a class were previously
3129        created as the appropriate unions of the upper, lower, and digit bitmaps.
3130        Now they are created by subtraction from the [[:word:]] class, which has
3131        its own bitmap.
3132    
3133     6. The [[:blank:]] character class matches horizontal, but not vertical space.
3134        It is created by subtracting the vertical space characters (\x09, \x0a,
3135        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
3136        subtraction was done in the overall bitmap for a character class, meaning
3137        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
3138        be recognized. This bug has been fixed.
3139    
3140     7. Patches from the folks at Google:
3141    
3142          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
3143          real life, but is still worth protecting against".
3144    
3145          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
3146          regular expressions".
3147    
3148          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
3149          have it.
3150    
3151          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
3152          "configure" and the latter not, in order to fix a problem somebody had
3153          with compiling the Arg class on HP-UX.
3154    
3155          (e) Improve the error-handling of the C++ wrapper a little bit.
3156    
3157          (f) New tests for checking recursion limiting.
3158    
3159     8. The pcre_memmove() function, which is used only if the environment does not
3160        have a standard memmove() function (and is therefore rarely compiled),
3161        contained two bugs: (a) use of int instead of size_t, and (b) it was not
3162        returning a result (though PCRE never actually uses the result).
3163    
3164     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
3165        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
3166        returned instead of calling malloc() with an overflowing number that would
3167        most likely cause subsequent chaos.
3168    
3169    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
3170    
3171    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
3172        with this option is matched, the nmatch and pmatch options of regexec() are
3173        ignored.
3174    
3175    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
3176        provided in case anyone wants to the the POSIX interface with UTF-8
3177        strings.
3178    
3179    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
3180        C++ linking (needed for some HP-UX environments).
3181    
3182    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
3183        (unused parameter) and in the pcre_printint() function (omitted "default"
3184        switch label when the default is to do nothing).
3185    
3186    15. Added some code to make it possible, when PCRE is compiled as a C++
3187        library, to replace subject pointers for pcre_exec() with a smart pointer
3188        class, thus making it possible to process discontinuous strings.
3189    
3190    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
3191        much the same function. They were added by different people who were trying
3192        to make PCRE easy to compile on non-Unix systems. It has been suggested
3193        that PCRE_EXPORT be abolished now that there is more automatic apparatus
3194        for compiling on Windows systems. I have therefore replaced it with
3195        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
3196        defaults to "extern" for C or "extern C" for C++, which works fine on
3197        Unix-like systems. It is now possible to override the value of PCRE_DATA_
3198        SCOPE with something explicit in config.h. In addition:
3199    
3200        (a) pcreposix.h still had just "extern" instead of either of these macros;
3201            I have replaced it with PCRE_DATA_SCOPE.
3202    
3203        (b) Functions such as _pcre_xclass(), which are internal to the library,
3204            but external in the C sense, all had PCRE_EXPORT in their definitions.
3205            This is apparently wrong for the Windows case, so I have removed it.
3206            (It makes no difference on Unix-like systems.)
3207    
3208    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
3209        of recursive calls to match(). This is different to MATCH_LIMIT because
3210        that limits the total number of calls to match(), not all of which increase
3211        the depth of recursion. Limiting the recursion depth limits the amount of
3212        stack (or heap if NO_RECURSE is set) that is used. The default can be set
3213        when PCRE is compiled, and changed at run time. A patch from Google adds
3214        this functionality to the C++ interface.
3215    
3216    18. Changes to the handling of Unicode character properties:
3217    
3218        (a) Updated the table to Unicode 4.1.0.
3219    
3220        (b) Recognize characters that are not in the table as "Cn" (undefined).
3221    
3222        (c) I revised the way the table is implemented to a much improved format
3223            which includes recognition of ranges. It now supports the ranges that
3224            are defined in UnicodeData.txt, and it also amalgamates other
3225            characters into ranges. This has reduced the number of entries in the
3226            table from around 16,000 to around 3,000, thus reducing its size
3227            considerably. I realized I did not need to use a tree structure after
3228            all - a binary chop search is just as efficient. Having reduced the
3229            number of entries, I extended their size from 6 bytes to 8 bytes to
3230            allow for more data.
3231    
3232        (d) Added support for Unicode script names via properties such as \p{Han}.
3233    
3234    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
3235        matching that character.
3236    
3237    20. When matching a repeated Unicode property with a minimum greater than zero,
3238        (for example \pL{2,}), PCRE could look past the end of the subject if it
3239        reached it while seeking the minimum number of characters. This could
3240        happen only if some of the characters were more than one byte long, because
3241        there is a check for at least the minimum number of bytes.
3242    
3243    21. Refactored the implementation of \p and \P so as to be more general, to
3244        allow for more different types of property in future. This has changed the
3245        compiled form incompatibly. Anybody with saved compiled patterns that use
3246        \p or \P will have to recompile them.
3247    
3248    22. Added "Any" and "L&" to the supported property types.
3249    
3250    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
3251        but give a compile time error if the value is greater than 0xff.
3252    
3253    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
3254        accidentally not being installed or uninstalled.
3255    
3256    25. The pcre.h file was built from pcre.h.in, but the only changes that were
3257        made were to insert the current release number. This seemed silly, because
3258        it made things harder for people building PCRE on systems that don't run
3259        "configure". I have turned pcre.h into a distributed file, no longer built
3260        by "configure", with the version identification directly included. There is
3261        no longer a pcre.h.in file.
3262    
3263        However, this change necessitated a change to the pcre-config script as
3264        well. It is built from pcre-config.in, and one of the substitutions was the
3265        release number. I have updated configure.ac so that ./configure now finds
3266        the release number by grepping pcre.h.
3267    
3268    26. Added the ability to run the tests under valgrind.
3269    
3270    
3271    Version 6.4 05-Sep-05
3272    ---------------------
3273    
3274     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3275        "--" to be printed when multiple files were scanned, even when none of the
3276        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3277        consider it to be a bug, and have restored the previous behaviour.
3278    
3279     2. A couple of code tidies to get rid of compiler warnings.
3280    
3281     3. The pcretest program used to cheat by referring to symbols in the library
3282        whose names begin with _pcre_. These are internal symbols that are not
3283        really supposed to be visible externally, and in some environments it is
3284        possible to suppress them. The cheating is now confined to including
3285        certain files from the library's source, which is a bit cleaner.
3286    
3287     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3288        file's purpose clearer.
3289    
3290     5. Reorganized pcre_ucp_findchar().
3291    
3292    
3293    Version 6.3 15-Aug-05
3294    ---------------------
3295    
3296     1. The file libpcre.pc.in did not have general read permission in the tarball.
3297    
3298     2. There were some problems when building without C++ support:
3299    
3300        (a) If C++ support was not built, "make install" and "make test" still
3301            tried to test it.
3302    
3303        (b) There were problems when the value of CXX was explicitly set. Some
3304            changes have been made to try to fix these, and ...
3305    
3306        (c) --disable-cpp can now be used to explicitly disable C++ support.
3307    
3308        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3309            backslash in a target when C++ was disabled. This confuses some
3310            versions of "make", apparently. Using an intermediate variable solves
3311            this. (Same for CPP_LOBJ.)
3312    
3313     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3314        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3315        necessary on certain architectures.
3316    
3317     4. Added a setting of -export-symbols-regex to the link command to remove
3318        those symbols that are exported in the C sense, but actually are local
3319        within the library, and not documented. Their names all begin with
3320        "_pcre_". This is not a perfect job, because (a) we have to except some
3321        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3322        available (and never for static libraries). I have made a note to try to
3323        find a way round (a) in the future.
3324    
3325    
3326    Version 6.2 01-Aug-05
3327    ---------------------
3328    
3329     1. There was no test for integer overflow of quantifier values. A construction
3330        such as {1111111111111111} would give undefined results. What is worse, if
3331        a minimum quantifier for a parenthesized subpattern overflowed and became
3332        negative, the calculation of the memory size went wrong. This could have
3333        led to memory overwriting.
3334    
3335     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3336    
3337     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3338        operating environments where this matters.
3339    
3340     4. Applied Giuseppe Maxia's patch to add additional features for controlling
3341        PCRE options from within the C++ wrapper.
3342    
3343     5. Named capturing subpatterns were not being correctly counted when a pattern
3344        was compiled. This caused two problems: (a) If there were more than 100
3345        such subpatterns, the calculation of the memory needed for the whole
3346        compiled pattern went wrong, leading to an overflow error. (b) Numerical
3347        back references of the form \12, where the number was greater than 9, were
3348        not recognized as back references, even though there were sufficient
3349        previous subpatterns.
3350    
3351     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3352        versions of gcc, e.g. 2.95.4.
3353    
3354    
3355    Version 6.1 21-Jun-05
3356    ---------------------
3357    
3358     1. There was one reference to the variable "posix" in pcretest.c that was not
3359        surrounded by "#if !defined NOPOSIX".
3360    
3361     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3362        the cross-check on the old pcre_info() function, for the benefit of the
3363        cut-down version of PCRE that is currently imported into Exim.
3364    
3365     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3366        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3367        patterns that start (?i) but I don't think that matters. The use of (?i) is
3368        just an example; this all applies to the other options as well.
3369    
3370     4. Since libtool seems to echo the compile commands it is issuing, the output
3371        from "make" can be reduced a bit by putting "@" in front of each libtool
3372        compile command.
3373    
3374     5. Patch from the folks at Google for configure.in to be a bit more thorough
3375        in checking for a suitable C++ installation before trying to compile the
3376        C++ stuff. This should fix a reported problem when a compiler was present,
3377        but no suitable headers.
3378    
3379     6. The man pages all had just "PCRE" as their title. I have changed them to
3380        be the relevant file name. I have also arranged that these names are
3381        retained in the file doc/pcre.txt, which is a concatenation in text format
3382        of all the man pages except the little individual ones for each function.
3383    
3384     7. The NON-UNIX-USE file had not been updated for the different set of source
3385        files that come with release 6. I also added a few comments about the C++
3386        wrapper.
3387    
3388    
3389    Version 6.0 07-Jun-05
3390    ---------------------
3391    
3392     1. Some minor internal re-organization to help with my DFA experiments.
3393    
3394     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3395        didn't matter for the library itself when fully configured, but did matter
3396        when compiling without UCP support, or within Exim, where the ucp files are
3397        not imported.
3398    
3399     3. Refactoring of the library code to split up the various functions into
3400        different source modules. The addition of the new DFA matching code (see
3401        below) to a single monolithic source would have made it really too
3402        unwieldy, quite apart from causing all the code to be include in a
3403        statically linked application, when only some functions are used. This is
3404        relevant even without the DFA addition now that patterns can be compiled in
3405        one application and matched in another.
3406    
3407        The downside of splitting up is that there have to be some external
3408        functions and data tables that are used internally in different modules of
3409        the library but which are not part of the API. These have all had their
3410        names changed to start with "_pcre_" so that they are unlikely to clash
3411        with other external names.
3412    
3413     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3414        a different (DFA) algorithm. Although it is slower than the original
3415        function, it does have some advantages for certain types of matching
3416        problem.
3417    
3418     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3419        including restarting after a partial match.
3420    
3421     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3422        defined when compiling for Windows was sent to me. I have put it into the
3423        code, though I have no means of testing or verifying it.
3424    
3425     7. Added the pcre_refcount() auxiliary function.
3426    
3427     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3428        match before or at the first newline in the subject string. In pcretest,
3429        the /f option on a pattern can be used to set this.
3430    
3431     9. A repeated \w when used in UTF-8 mode with characters greater than 256
3432        would behave wrongly. This has been present in PCRE since release 4.0.
3433    
3434    10. A number of changes to the pcregrep command:
3435    
3436        (a) Refactored how -x works; insert ^(...)$ instead of setting
3437            PCRE_ANCHORED and checking the length, in preparation for adding
3438            something similar for -w.
3439    
3440        (b) Added the -w (match as a word) option.
3441    
3442        (c) Refactored the way lines are read and buffered so as to have more
3443            than one at a time available.
3444    
3445        (d) Implemented a pcregrep test script.
3446    
3447        (e) Added the -M (multiline match) option. This allows patterns to match
3448            over several lines of the subject. The buffering ensures that at least
3449            8K, or the rest of the document (whichever is the shorter) is available
3450            for matching (and similarly the previous 8K for lookbehind assertions).
3451    
3452        (f) Changed the --help output so that it now says
3453    
3454              -w, --word-regex(p)
3455    
3456            instead of two lines, one with "regex" and the other with "regexp"
3457            because that confused at least one person since the short forms are the
3458            same. (This required a bit of code, as the output is generated
3459            automatically from a table. It wasn't just a text change.)
3460    
3461        (g) -- can be used to terminate pcregrep options if the next thing isn't an
3462            option but starts with a hyphen. Could be a pattern or a path name
3463            starting with a hyphen, for instance.
3464    
3465        (h) "-" can be given as a file name to represent stdin.
3466    
3467        (i) When file names are being printed, "(standard input)" is used for
3468            the standard input, for compatibility with GNU grep. Previously
3469            "<stdin>" was used.
3470    
3471        (j) The option --label=xxx can be used to supply a name to be used for
3472            stdin when file names are being printed. There is no short form.
3473    
3474        (k) Re-factored the options decoding logic because we are going to add
3475            two more options that take data. Such options can now be given in four
3476            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3477    
3478        (l) Added the -A, -B, and -C options for requesting that lines of context
3479            around matches be printed.
3480    
3481        (m) Added the -L option to print the names of files that do not contain
3482            any matching lines, that is, the complement of -l.
3483    
3484        (n) The return code is 2 if any file cannot be opened, but pcregrep does
3485            continue to scan other files.
3486    
3487        (o) The -s option was incorrectly implemented. For compatibility with other
3488            greps, it now suppresses the error message for a non-existent or non-
3489            accessible file (but not the return code). There is a new option called
3490            -q that suppresses the output of matching lines, which was what -s was
3491            previously doing.
3492    
3493        (p) Added --include and --exclude options to specify files for inclusion
3494            and exclusion when recursing.
3495    
3496    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3497        Hopefully, it now does.
3498    
3499    12. Missing cast in pcre_study().
3500    
3501    13. Added an "uninstall" target to the makefile.
3502    
3503    14. Replaced "extern" in the function prototypes in Makefile.in with
3504        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3505        world, but is set differently for Windows.
3506    
3507    15. Added a second compiling function called pcre_compile2(). The only
3508        difference is that it has an extra argument, which is a pointer to an
3509        integer error code. When there is a compile-time failure, this is set
3510        non-zero, in addition to the error test pointer being set to point to an
3511        error message. The new argument may be NULL if no error number is required
3512        (but then you may as well call pcre_compile(), which is now just a
3513        wrapper). This facility is provided because some applications need a
3514        numeric error indication, but it has also enabled me to tidy up the way
3515        compile-time errors are handled in the POSIX wrapper.
3516    
3517    16. Added VPATH=.libs to the makefile; this should help when building with one
3518        prefix path and installing with another. (Or so I'm told by someone who
3519        knows more about this stuff than I do.)
3520    
3521    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3522        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3523        match everything, including newlines. This is not POSIX-compatible, but
3524        somebody wanted the feature. From pcretest it can be activated by using
3525        both the P and the s flags.
3526    
3527    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3528    
3529    19. libpcre.pc was being incorrectly installed as executable.
3530    
3531    20. A couple of places in pcretest check for end-of-line by looking for '\n';
3532        it now also looks for '\r' so that it will work unmodified on Windows.
3533    
3534    21. Added Google's contributed C++ wrapper to the distribution.
3535    
3536    22. Added some untidy missing memory free() calls in pcretest, to keep
3537        Electric Fence happy when testing.
3538    
3539    
3540    
3541    Version 5.0 13-Sep-04
3542    ---------------------
3543    
3544     1. Internal change: literal characters are no longer packed up into items
3545        containing multiple characters in a single byte-string. Each character
3546        is now matched using a separate opcode. However, there may be more than one
3547        byte in the character in UTF-8 mode.
3548    
3549     2. The pcre_callout_block structure has two new fields: pattern_position and
3550        next_item_length. These contain the offset in the pattern to the next match
3551        item, and its length, respectively.
3552    
3553     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3554        insertion of callouts before each pattern item. Added the /C option to
3555        pcretest to make use of this.
3556    
3557     4. On the advice of a Windows user, the lines
3558    
3559          #if defined(_WIN32) || defined(WIN32)
3560          _setmode( _fileno( stdout ), 0x8000 );
3561          #endif  /* defined(_WIN32) || defined(WIN32) */
3562    
3563        have been added to the source of pcretest. This apparently does useful
3564        magic in relation to line terminators.
3565    
3566     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3567        for the benefit of those environments where the "b" makes a difference.
3568    
3569     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3570        to know about it. I have put a hack into configure.in that adds in code
3571        to set GCC=yes if CC=icc. This seems to end up at a point in the
3572        generated configure script that is early enough to affect the setting of
3573        compiler options, which is what is needed, but I have no means of testing
3574        whether it really works. (The user who reported this had patched the
3575        generated configure script, which of course I cannot do.)
3576    
3577        LATER: After change 22 below (new libtool files), the configure script
3578        seems to know about icc (and also ecc). Therefore, I have commented out
3579        this hack in configure.in.
3580    
3581     7. Added support for pkg-config (2 patches were sent in).
3582    
3583     8. Negated POSIX character classes that used a combination of internal tables
3584     &nb