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

Diff of /code/trunk/ChangeLog

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

revision 153 by ph10, Wed Apr 18 09:12:14 2007 UTC revision 1172 by ph10, Wed Oct 24 16:41:07 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.1 12-Mar-07  Version 8.32
5    ------------
6    
7    1.  Improved JIT compiler optimizations for first character search and single
8        character iterators.
9    
10    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
11        Patch by Daniel Richard G.
12    
13    3.  Single character iterator optimizations in the JIT compiler.
14    
15    4.  Improved JIT compiler optimizations for character ranges.
16    
17    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
18        Reported by Giuseppe D'Angelo.
19    
20    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
21        of a line, was being set incorrectly in cases where .* appeared inside
22        atomic brackets at the start of a pattern, or where there was a subsequent
23        *PRUNE or *SKIP.
24    
25    7.  Improved instruction cache flush for POWER/PowerPC.
26        Patch by Daniel Richard G.
27    
28    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
29        grep:
30    
31        (a) There is now no limit to the number of patterns to be matched.
32    
33        (b) An error is given if a pattern is too long.
34    
35        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
36            are now supported.
37    
38        (d) --exclude-from and --include-from (multiple use) have been added.
39    
40        (e) Exclusions and inclusions now apply to all files and directories, not
41            just to those obtained from scanning a directory recursively.
42    
43        (f) Multiple uses of -f and --file-list are now supported.
44    
45        (g) In a Windows environment, the default for -d has been changed from
46            "read" (the GNU grep default) to "skip", because otherwise the presence
47            of a directory in the file list provokes an error.
48    
49        (h) The documentation has been revised and clarified in places.
50    
51    9.  Improve the matching speed of capturing brackets.
52    
53    10. Changed the meaning of \X so that it now matches a Unicode extended
54        grapheme cluster.
55    
56    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
57        out POSIX threads when JIT support is configured.
58    
59    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
60    
61    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
62        structure could go wrong in environments where size_t is not the same size
63        as int.
64    
65    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
66        set.
67    
68    15. The EBCDIC support had decayed; later updates to the code had included
69        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
70        general tidy up of EBCDIC-related issues, and the documentation was also
71        not quite right. There is now a test that can be run on ASCII systems to
72        check some of the EBCDIC-related things (but is it not a full test).
73    
74    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
75        in a small tidy to the code.
76    
77    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
78    
79    18. If the --only-matching (-o) option in pcregrep is specified multiple
80        times, each one causes appropriate output. For example, -o1 -o2 outputs the
81        substrings matched by the 1st and 2nd capturing parentheses. A separating
82        string can be specified by --om-separator (default empty).
83    
84    19. Improving the first n character searches.
85    
86    20. Turn case lists for horizontal and vertical white space into macros so that
87        they are defined only once.
88    
89    21. This set of changes together give more compatible Unicode case-folding
90        behaviour for characters that have more than one other case when UCP
91        support is available.
92    
93        (a) The Unicode property table now has offsets into a new table of sets of
94            three or more characters that are case-equivalent. The MultiStage2.py
95            script that generates these tables (the pcre_ucd.c file) now scans
96            CaseFolding.txt instead of UnicodeData.txt for character case
97            information.
98    
99        (b) The code for adding characters or ranges of characters to a character
100            class has been abstracted into a generalized function that also handles
101            case-independence. In UTF-mode with UCP support, this uses the new data
102            to handle characters with more than one other case.
103    
104        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
105            whose other case is greater than 256 are now correctly matched
106            caselessly. Previously, the high codepoint matched the low one, but not
107            vice versa.
108    
109        (d) The processing of \h, \H, \v, and \ in character classes now makes use
110            of the new class addition function, using character lists defined as
111            macros alongside the case definitions of 20 above.
112    
113        (e) Caseless back references now work with characters that have more than
114            one other case.
115    
116        (f) General caseless matching of characters with more than one other case
117            is supported.
118    
119    22. Unicode character properties were updated from Unicode 6.2.0
120    
121    23. Improved CMake support under Windows. Patch by Daniel Richard G.
122    
123    24. Add support for 32-bit character strings, and UTF-32
124    
125    25. Major JIT compiler update (code refactoring and bugfixing).
126        Experimental Sparc 32 support is added.
127    
128    26. Applied a modified version of Daniel Richard G's patch to create
129        pcre.h.generic and config.h.generic by "make" instead of in the
130        PrepareRelease script.
131    
132    
133    Version 8.31 06-July-2012
134    -------------------------
135    
136    1.  Fixing a wrong JIT test case and some compiler warnings.
137    
138    2.  Removed a bashism from the RunTest script.
139    
140    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
141        to unsigned type, result still unsigned" that was given by an MS compiler
142        on encountering the code "-sizeof(xxx)".
143    
144    4.  Partial matching support is added to the JIT compiler.
145    
146    5.  Fixed several bugs concerned with partial matching of items that consist
147        of more than one character:
148    
149        (a) /^(..)\1/ did not partially match "aba" because checking references was
150            done on an "all or nothing" basis. This also applied to repeated
151            references.
152    
153        (b) \R did not give a hard partial match if \r was found at the end of the
154            subject.
155    
156        (c) \X did not give a hard partial match after matching one or more
157            characters at the end of the subject.
158    
159        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
160            a partial match for the string "\r".
161    
162        (e) When newline was set to CRLF, the metacharacter "." did not recognize
163            a partial match for a CR character at the end of the subject string.
164    
165    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
166        running pcretest, the text "(JIT)" added to the output whenever JIT is
167        actually used to run the match.
168    
169    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
170        or /S+[+] with a digit between 1 and 7.
171    
172    8.  OP_NOT now supports any UTF character not just single-byte ones.
173    
174    9.  (*MARK) control verb is now supported by the JIT compiler.
175    
176    10. The command "./RunTest list" lists the available tests without actually
177        running any of them. (Because I keep forgetting what they all are.)
178    
179    11. Add PCRE_INFO_MAXLOOKBEHIND.
180    
181    12. Applied a (slightly modified) user-supplied patch that improves performance
182        when the heap is used for recursion (compiled with --disable-stack-for-
183        recursion). Instead of malloc and free for each heap frame each time a
184        logical recursion happens, frames are retained on a chain and re-used where
185        possible. This sometimes gives as much as 30% improvement.
186    
187    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
188        call.
189    
190    14. As documented, (*COMMIT) is now confined to within a positive assertion.
191    
192    15. It is now possible to link pcretest with libedit as an alternative to
193        libreadline.
194    
195    16. (*COMMIT) control verb is now supported by the JIT compiler.
196    
197    17. The Unicode data tables have been updated to Unicode 6.1.0.
198    
199    18. Added --file-list option to pcregrep.
200    
201    19. Added binary file support to pcregrep, including the -a, --binary-files,
202        -I, and --text options.
203    
204    20. The madvise function is renamed for posix_madvise for QNX compatibility
205        reasons. Fixed by Giuseppe D'Angelo.
206    
207    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
208    
209    22. Changed the option for creating symbolic links for 16-bit man pages from
210        -s to -sf so that re-installing does not cause issues.
211    
212    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
213    
214    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
215        matches in certain environments (the workspace was not being correctly
216        retained). Also added to pcre_dfa_exec() a simple plausibility check on
217        some of the workspace data at the beginning of a restart.
218    
219    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
220        was not doing so when it should - probably a typo introduced by SVN 528
221        (change 8.10/14).
222    
223    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
224        \w+ when the character tables indicated that \x{c4} was a word character.
225        There were several related cases, all because the tests for doing a table
226        lookup were testing for characters less than 127 instead of 255.
227    
228    27. If a pattern contains capturing parentheses that are not used in a match,
229        their slots in the ovector are set to -1. For those that are higher than
230        any matched groups, this happens at the end of processing. In the case when
231        there were back references that the ovector was too small to contain
232        (causing temporary malloc'd memory to be used during matching), and the
233        highest capturing number was not used, memory off the end of the ovector
234        was incorrectly being set to -1. (It was using the size of the temporary
235        memory instead of the true size.)
236    
237    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
238        ovector size, it uses memory at the end of the block that it has got.
239    
240    29. Check for an overlong MARK name and give an error at compile time. The
241        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
242    
243    30. JIT compiler update.
244    
245    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
246        Rill for the patch.
247    
248    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
249    
250    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
251    
252    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
253        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
254        was enabled.
255    
256    35. Improve JIT code generation for greedy plus quantifier.
257    
258    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
259        1 to "aa" instead of to an empty string. The bug affected repeated groups
260        that could potentially match an empty string.
261    
262    37. Optimizing single character iterators in JIT.
263    
264    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
265        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
266        codepoints that are too big for the mode are faulted, and in a UTF mode,
267        disallowed codepoints are also faulted.
268    
269    39. If PCRE was compiled with UTF support, in three places in the DFA
270        matcher there was code that should only have been obeyed in UTF mode, but
271        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
272        processing when bytes with values greater than 127 were present. In 16-bit
273        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
274        both cases the values are those that cannot be the first data item in a UTF
275        character. The three items that might have provoked this were recursions,
276        possessively repeated groups, and atomic groups.
277    
278    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
279        and pcregrep, because some OS require shared objects to be explicitly
280        passed to ld, causing the link step to fail if they are not.
281    
282    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
283        mode, patterns that started with \h* or \R* might be incorrectly matched.
284    
285    
286    Version 8.30 04-February-2012
287    -----------------------------
288    
289    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
290        name is defined in ctype.h.
291    
292    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
293        only in quite long subpatterns.
294    
295    3.  Removed the function pcre_info(), which has been obsolete and deprecated
296        since it was replaced by pcre_fullinfo() in February 2000.
297    
298    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
299        match a (*MARK), and the match failed at the start of the subject, a
300        reference to memory before the start of the subject could occur. This bug
301        was introduced by fix 17 of release 8.21.
302    
303    5.  A reference to an unset group with zero minimum repetition was giving
304        totally wrong answers (in non-JavaScript-compatibility mode). For example,
305        /(another)?(\1?)test/ matched against "hello world test". This bug was
306        introduced in release 8.13.
307    
308    6.  Add support for 16-bit character strings (a large amount of work involving
309        many changes and refactorings).
310    
311    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
312        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
313        from a file.
314    
315    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
316        rounding is not applied in this particular case).
317    
318    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
319        if they appear, or are escaped, in patterns.
320    
321    10. Get rid of a number of -Wunused-but-set-variable warnings.
322    
323    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
324        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
325        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
326        also returns the mark "x". This bug applied to capturing parentheses,
327        non-capturing parentheses, and atomic parentheses. It also applied to some
328        assertions.
329    
330    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
331        information out of configure.ac instead of relying on pcre.h.generic, which
332        is not stored in the repository.
333    
334    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
335        -lreadline.
336    
337    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
338    
339    15. Applied Graycode's patch to put the top-level frame on the stack rather
340        than the heap when not using the stack for recursion. This gives a
341        performance improvement in many cases when recursion is not deep.
342    
343    16. Experimental code added to "pcretest -C" to output the stack frame size.
344    
345    
346    Version 8.21 12-Dec-2011
347    ------------------------
348    
349    1.  Updating the JIT compiler.
350    
351    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
352        are added as well.
353    
354    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
355        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
356        calling _pcre_jit_exec. Some extra comments are added.
357    
358    4.  (*MARK) settings inside atomic groups that do not contain any capturing
359        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
360        was introduced by change 18 for 8.20.
361    
362    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
363        ECMA-262 standard.
364    
365    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
366        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
367        This bug was probably introduced by change 9 of 8.13.
368    
369    7.  While fixing 6 above, I noticed that a number of other items were being
370        incorrectly rejected as "not fixed length". This arose partly because newer
371        opcodes had not been added to the fixed-length checking code. I have (a)
372        corrected the bug and added tests for these items, and (b) arranged for an
373        error to occur if an unknown opcode is encountered while checking for fixed
374        length instead of just assuming "not fixed length". The items that were
375        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
376        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
377        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
378    
379    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
380        being incorrectly compiled and would have given unpredicatble results.
381    
382    9.  A possessively repeated subpattern with minimum repeat count greater than
383        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
384        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
385        the first (A) could occur when it should not.
386    
387    10. Add a cast and remove a redundant test from the code.
388    
389    11. JIT should use pcre_malloc/pcre_free for allocation.
390    
391    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
392        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
393        is anything other than /usr, -L is still shown).
394    
395    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
396    
397    14. Perl does not support \N without a following name in a [] class; PCRE now
398        also gives an error.
399    
400    15. If a forward reference was repeated with an upper limit of around 2000,
401        it caused the error "internal error: overran compiling workspace". The
402        maximum number of forward references (including repeats) was limited by the
403        internal workspace, and dependent on the LINK_SIZE. The code has been
404        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
405        the default depends on LINK_SIZE. There is a new upper limit (for safety)
406        of around 200,000 forward references. While doing this, I also speeded up
407        the filling in of repeated forward references.
408    
409    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
410        incorrectly expecting the subject to contain another "a" after the start.
411    
412    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
413        in the match, the SKIP should be ignored. This was not happening; instead
414        the SKIP was being treated as NOMATCH. For patterns such as
415        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
416        tested.
417    
418    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
419        now much more compatible with Perl, in particular in cases where the result
420        is a non-match for a non-anchored pattern. For example, if
421        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
422        "m", where previously it did not return a name. A side effect of this
423        change is that for partial matches, the last encountered mark name is
424        returned, as for non matches. A number of tests that were previously not
425        Perl-compatible have been moved into the Perl-compatible test files. The
426        refactoring has had the pleasing side effect of removing one argument from
427        the match() function, thus reducing its stack requirements.
428    
429    19. If the /S+ option was used in pcretest to study a pattern using JIT,
430        subsequent uses of /S (without +) incorrectly behaved like /S+.
431    
432    21. Retrieve executable code size support for the JIT compiler and fixing
433        some warnings.
434    
435    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
436        not work when the shorter character appeared right at the end of the
437        subject string.
438    
439    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
440        systems.
441    
442    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
443        output it when the /M option is used in pcretest.
444    
445    25. The CheckMan script was not being included in the distribution. Also, added
446        an explicit "perl" to run Perl scripts from the PrepareRelease script
447        because this is reportedly needed in Windows.
448    
449    26. If study data was being save in a file and studying had not found a set of
450        "starts with" bytes for the pattern, the data written to the file (though
451        never used) was taken from uninitialized memory and so caused valgrind to
452        complain.
453    
454    27. Updated RunTest.bat as provided by Sheri Pierce.
455    
456    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
457    
458    29. Computation of memory usage for the table of capturing group names was
459        giving an unnecessarily large value.
460    
461    
462    Version 8.20 21-Oct-2011
463    ------------------------
464    
465    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
466        a POSIX class. After further experiments with Perl, which convinced me that
467        Perl has bugs and confusions, a closing square bracket is no longer allowed
468        in a POSIX name. This bug also affected patterns with classes that started
469        with full stops.
470    
471    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
472        captured substring, but while checking the failing first alternative,
473        substring 1 is temporarily captured. If the output vector supplied to
474        pcre_exec() was not big enough for this capture, the yield of the function
475        was still zero ("insufficient space for captured substrings"). This cannot
476        be totally fixed without adding another stack variable, which seems a lot
477        of expense for a edge case. However, I have improved the situation in cases
478        such as /(a)(b)x|abc/ matched against "abc", where the return code
479        indicates that fewer than the maximum number of slots in the ovector have
480        been set.
481    
482    3.  Related to (2) above: when there are more back references in a pattern than
483        slots in the output vector, pcre_exec() uses temporary memory during
484        matching, and copies in the captures as far as possible afterwards. It was
485        using the entire output vector, but this conflicts with the specification
486        that only 2/3 is used for passing back captured substrings. Now it uses
487        only the first 2/3, for compatibility. This is, of course, another edge
488        case.
489    
490    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
491        main code base, and can be used by building with --enable-jit. When this is
492        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
493        runtime --no-jit option is given.
494    
495    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
496        ovector, the return from the function was zero, implying that there were
497        other matches that did not fit. The correct "exactly full" value is now
498        returned.
499    
500    6.  If a subpattern that was called recursively or as a subroutine contained
501        (*PRUNE) or any other control that caused it to give a non-standard return,
502        invalid errors such as "Error -26 (nested recursion at the same subject
503        position)" or even infinite loops could occur.
504    
505    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
506        computing the minimum length on reaching *ACCEPT, and so ended up with the
507        wrong value of 1 rather than 0. Further investigation indicates that
508        computing a minimum subject length in the presence of *ACCEPT is difficult
509        (think back references, subroutine calls), and so I have changed the code
510        so that no minimum is registered for a pattern that contains *ACCEPT.
511    
512    8.  If (*THEN) was present in the first (true) branch of a conditional group,
513        it was not handled as intended. [But see 16 below.]
514    
515    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
516        Sheri Pierce.
517    
518    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
519        the first byte in a match must be "a".
520    
521    11. Change 17 for 8.13 increased the recursion depth for patterns like
522        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
523        pattern contains any instances of (*THEN). If it does not, the old
524        optimizations are restored. It would be nice to do this on a per-group
525        basis, but at the moment that is not feasible.
526    
527    12. In some environments, the output of pcretest -C is CRLF terminated. This
528        broke RunTest's code that checks for the link size. A single white space
529        character after the value is now allowed for.
530    
531    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
532        For "fr", it uses the Windows-specific input and output files.
533    
534    14. If (*THEN) appeared in a group that was called recursively or as a
535        subroutine, it did not work as intended. [But see next item.]
536    
537    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
538        pattern fragments (but not containing any | characters). If A and B are
539        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
540        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
541        D. In other words, Perl considers parentheses that do not contain any |
542        characters to be part of a surrounding alternative, whereas PCRE was
543        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
544        differently. PCRE now behaves in the same way as Perl, except in the case
545        of subroutine/recursion calls such as (?1) which have in any case always
546        been different (but PCRE had them first :-).
547    
548    16. Related to 15 above: Perl does not treat the | in a conditional group as
549        creating alternatives. Such a group is treated in the same way as an
550        ordinary group without any | characters when processing (*THEN). PCRE has
551        been changed to match Perl's behaviour.
552    
553    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
554        RunGrepTest script failed.
555    
556    18. Change 22 for version 13 caused atomic groups to use more stack. This is
557        inevitable for groups that contain captures, but it can lead to a lot of
558        stack use in large patterns. The old behaviour has been restored for atomic
559        groups that do not contain any capturing parentheses.
560    
561    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
562        suppress the check for a minimum subject length at run time. (If it was
563        given to pcre_exec() or pcre_dfa_exec() it did work.)
564    
565    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
566        fail to work when decoding hex characters in data strings in EBCDIC
567        environments.
568    
569    21. It appears that in at least one Mac OS environment, the isxdigit() function
570        is implemented as a macro that evaluates to its argument more than once,
571        contravening the C 90 Standard (I haven't checked a later standard). There
572        was an instance in pcretest which caused it to go wrong when processing
573        \x{...} escapes in subject strings. The has been rewritten to avoid using
574        things like p++ in the argument of isxdigit().
575    
576    
577    Version 8.13 16-Aug-2011
578    ------------------------
579    
580    1.  The Unicode data tables have been updated to Unicode 6.0.0.
581    
582    2.  Two minor typos in pcre_internal.h have been fixed.
583    
584    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
585        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
586        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
587    
588    4.  There were a number of related bugs in the code for matching backrefences
589        caselessly in UTF-8 mode when codes for the characters concerned were
590        different numbers of bytes. For example, U+023A and U+2C65 are an upper
591        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
592        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
593        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
594        2-byte code at the end of the subject (it thought there wasn't enough data
595        left).
596    
597    5.  Comprehensive information about what went wrong is now returned by
598        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
599        as the output vector has at least 2 elements. The offset of the start of
600        the failing character and a reason code are placed in the vector.
601    
602    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
603        now returned is for the first byte of the failing character, instead of the
604        last byte inspected. This is an incompatible change, but I hope it is small
605        enough not to be a problem. It makes the returned offset consistent with
606        pcre_exec() and pcre_dfa_exec().
607    
608    7.  pcretest now gives a text phrase as well as the error number when
609        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
610        failure, the offset and reason code are output.
611    
612    8.  When \R was used with a maximizing quantifier it failed to skip backwards
613        over a \r\n pair if the subsequent match failed. Instead, it just skipped
614        back over a single character (\n). This seems wrong (because it treated the
615        two characters as a single entity when going forwards), conflicts with the
616        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
617        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
618        has been changed.
619    
620    9.  Some internal refactoring has changed the processing so that the handling
621        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
622        time (the PCRE_DOTALL option was changed this way some time ago: version
623        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
624        which was always a bit of a fudge. It also means that there is one less
625        argument for the match() function, which reduces its stack requirements
626        slightly. This change also fixes an incompatibility with Perl: the pattern
627        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
628    
629    10. More internal refactoring has drastically reduced the number of recursive
630        calls to match() for possessively repeated groups such as (abc)++ when
631        using pcre_exec().
632    
633    11. While implementing 10, a number of bugs in the handling of groups were
634        discovered and fixed:
635    
636        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
637        (a|)*(?1) gave a compile-time internal error.
638        ((a|)+)+  did not notice that the outer group could match an empty string.
639        (^a|^)+   was not marked as anchored.
640        (.*a|.*)+ was not marked as matching at start or after a newline.
641    
642    12. Yet more internal refactoring has removed another argument from the match()
643        function. Special calls to this function are now indicated by setting a
644        value in a variable in the "match data" data block.
645    
646    13. Be more explicit in pcre_study() instead of relying on "default" for
647        opcodes that mean there is no starting character; this means that when new
648        ones are added and accidentally left out of pcre_study(), testing should
649        pick them up.
650    
651    14. The -s option of pcretest has been documented for ages as being an old
652        synonym of -m (show memory usage). I have changed it to mean "force study
653        for every regex", that is, assume /S for every regex. This is similar to -i
654        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
655        using it. It makes it easier to run collections of tests with and without
656        study enabled, and thereby test pcre_study() more easily. All the standard
657        tests are now run with and without -s (but some patterns can be marked as
658        "never study" - see 20 below).
659    
660    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
661        restoration of the capturing data to the outer values was not happening
662        correctly.
663    
664    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
665        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
666        pattern had matched an empty string, and so incorrectly returned a no
667        match.
668    
669    17. There was optimizing code for the last branch of non-capturing parentheses,
670        and also for the obeyed branch of a conditional subexpression, which used
671        tail recursion to cut down on stack usage. Unfortunately, now that there is
672        the possibility of (*THEN) occurring in these branches, tail recursion is
673        no longer possible because the return has to be checked for (*THEN). These
674        two optimizations have therefore been removed. [But see 8.20/11 above.]
675    
676    18. If a pattern containing \R was studied, it was assumed that \R always
677        matched two bytes, thus causing the minimum subject length to be
678        incorrectly computed because \R can also match just one byte.
679    
680    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
681        was incorrectly computed.
682    
683    20. If /S is present twice on a test pattern in pcretest input, it now
684        *disables* studying, thereby overriding the use of -s on the command line
685        (see 14 above). This is necessary for one or two tests to keep the output
686        identical in both cases.
687    
688    21. When (*ACCEPT) was used in an assertion that matched an empty string and
689        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
690    
691    22. When an atomic group that contained a capturing parenthesis was
692        successfully matched, but the branch in which it appeared failed, the
693        capturing was not being forgotten if a higher numbered group was later
694        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
695        group 1 to "a", when in fact it should be unset. This applied to multi-
696        branched capturing and non-capturing groups, repeated or not, and also to
697        positive assertions (capturing in negative assertions does not happen
698        in PCRE) and also to nested atomic groups.
699    
700    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
701        subject after a captured substring, to make it easier to tell which of a
702        number of identical substrings has been captured.
703    
704    24. The way atomic groups are processed by pcre_exec() has been changed so that
705        if they are repeated, backtracking one repetition now resets captured
706        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
707        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
708        "aaa". Previously, it would have been "a". As part of this code
709        refactoring, the way recursive calls are handled has also been changed.
710    
711    25. If an assertion condition captured any substrings, they were not passed
712        back unless some other capturing happened later. For example, if
713        (?(?=(a))a) was matched against "a", no capturing was returned.
714    
715    26. When studying a pattern that contained subroutine calls or assertions,
716        the code for finding the minimum length of a possible match was handling
717        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
718        group 1 called group 2 while simultaneously a separate group 2 called group
719        1). A stack overflow occurred in this case. I have fixed this by limiting
720        the recursion depth to 10.
721    
722    27. Updated RunTest.bat in the distribution to the version supplied by Tom
723        Fortmann. This supports explicit test numbers on the command line, and has
724        argument validation and error reporting.
725    
726    28. An instance of \X with an unlimited repeat could fail if at any point the
727        first character it looked at was a mark character.
728    
729    29. Some minor code refactoring concerning Unicode properties and scripts
730        should reduce the stack requirement of match() slightly.
731    
732    30. Added the '=' option to pcretest to check the setting of unused capturing
733        slots at the end of the pattern, which are documented as being -1, but are
734        not included in the return count.
735    
736    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
737        compiled something random. Now it gives a compile-time error (as does
738        Perl).
739    
740    32. A *MARK encountered during the processing of a positive assertion is now
741        recorded and passed back (compatible with Perl).
742    
743    33. If --only-matching or --colour was set on a pcregrep call whose pattern
744        had alternative anchored branches, the search for a second match in a line
745        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
746        matched the line "0102" twice. The same bug affected patterns that started
747        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
748        twice.
749    
750    34. Previously, PCRE did not allow quantification of assertions. However, Perl
751        does, and because of capturing effects, quantifying parenthesized
752        assertions may at times be useful. Quantifiers are now allowed for
753        parenthesized assertions.
754    
755    35. A minor code tidy in pcre_compile() when checking options for \R usage.
756    
757    36. \g was being checked for fancy things in a character class, when it should
758        just be a literal "g".
759    
760    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
761        appearance of a nested POSIX class supersedes an apparent external class.
762        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
763        unescaped square brackets may also appear as part of class names. For
764        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
765        more like Perl. (But see 8.20/1 above.)
766    
767    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
768        was because it thought it was \N{name}, which is not supported).
769    
770    39. Add minix to OS list not supporting the -S option in pcretest.
771    
772    40. PCRE tries to detect cases of infinite recursion at compile time, but it
773        cannot analyze patterns in sufficient detail to catch mutual recursions
774        such as ((?1))((?2)). There is now a runtime test that gives an error if a
775        subgroup is called recursively as a subpattern for a second time at the
776        same position in the subject string. In previous releases this might have
777        been caught by the recursion limit, or it might have run out of stack.
778    
779    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
780        happen only once. PCRE was, however incorrectly giving a compile time error
781        "recursive call could loop indefinitely" because it cannot analyze the
782        pattern in sufficient detail. The compile time test no longer happens when
783        PCRE is compiling a conditional subpattern, but actual runaway loops are
784        now caught at runtime (see 40 above).
785    
786    42. It seems that Perl allows any characters other than a closing parenthesis
787        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
788        has been changed to be the same.
789    
790    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
791        as not to get warnings when autogen.sh is called. Also changed
792        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
793    
794    44. To help people who use pcregrep to scan files containing exceedingly long
795        lines, the following changes have been made:
796    
797        (a) The default value of the buffer size parameter has been increased from
798            8K to 20K. (The actual buffer used is three times this size.)
799    
800        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
801            PCRE is built.
802    
803        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
804            to be set at run time.
805    
806        (d) Numerical values in pcregrep options can be followed by K or M, for
807            example --buffer-size=50K.
808    
809        (e) If a line being scanned overflows pcregrep's buffer, an error is now
810            given and the return code is set to 2.
811    
812    45. Add a pointer to the latest mark to the callout data block.
813    
814    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
815        partial match of an empty string instead of no match. This was specific to
816        the use of ".".
817    
818    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
819        complete match instead of a partial match. This bug was dependent on both
820        the PCRE_UTF8 and PCRE_DOTALL options being set.
821    
822    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
823        starting byte set, because \b was not being ignored.
824    
825    
826    Version 8.12 15-Jan-2011
827    ------------------------
828    
829    1.  Fixed some typos in the markup of the man pages, and wrote a script that
830        checks for such things as part of the documentation building process.
831    
832    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
833        --match-limit and --recursion-limit options (added for 8.11). In
834        particular, this made one of the standard tests fail. (The integer value
835        went into the wrong half of a long int.)
836    
837    3.  If the --colour option was given to pcregrep with -v (invert match), it
838        did strange things, either producing crazy output, or crashing. It should,
839        of course, ignore a request for colour when reporting lines that do not
840        match.
841    
842    4.  Another pcregrep bug caused similar problems if --colour was specified with
843        -M (multiline) and the pattern match finished with a line ending.
844    
845    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
846        matched in multiline mode, the following line was shown as part of the
847        match. This seems wrong, so I have changed it.
848    
849    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
850        the check for further matches in the same line (so they could be coloured)
851        to overrun the end of the current line. If another match was found, it was
852        incorrectly shown (and then shown again when found in the next line).
853    
854    7.  If pcregrep was compiled under Windows, there was a reference to the
855        function pcregrep_exit() before it was defined. I am assuming this was
856        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
857        reported by a user. I've moved the definition above the reference.
858    
859    
860    Version 8.11 10-Dec-2010
861    ------------------------
862    
863    1.  (*THEN) was not working properly if there were untried alternatives prior
864        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
865        backtracked to try for "b" instead of moving to the next alternative branch
866        at the same level (in this case, to look for "c"). The Perl documentation
867        is clear that when (*THEN) is backtracked onto, it goes to the "next
868        alternative in the innermost enclosing group".
869    
870    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
871        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
872        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
873        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
874        (*THEN).
875    
876    3.  If \s appeared in a character class, it removed the VT character from
877        the class, even if it had been included by some previous item, for example
878        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
879        of \s, but is part of the POSIX "space" class.)
880    
881    4.  A partial match never returns an empty string (because you can always
882        match an empty string at the end of the subject); however the checking for
883        an empty string was starting at the "start of match" point. This has been
884        changed to the "earliest inspected character" point, because the returned
885        data for a partial match starts at this character. This means that, for
886        example, /(?<=abc)def/ gives a partial match for the subject "abc"
887        (previously it gave "no match").
888    
889    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
890        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
891        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
892        has an implication that the given string is incomplete (because a partial
893        match is preferred over a full match). For this reason, these items now
894        give a partial match in this situation. [Aside: previously, the one case
895        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
896        match rather than a full match, which was wrong by the old rules, but is
897        now correct.]
898    
899    6.  There was a bug in the handling of #-introduced comments, recognized when
900        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
901        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
902        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
903        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
904        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
905        places in pcre_compile().
906    
907    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
908        comments when looking ahead for named forward references to subpatterns,
909        the only newline sequence it recognized was NL. It now handles newlines
910        according to the set newline convention.
911    
912    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
913        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
914        cater for a lack of strerror(). These oversights have been fixed.
915    
916    9.  Added --match-limit and --recursion-limit to pcregrep.
917    
918    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
919    
920    11. When the -o option was used, pcregrep was setting a return code of 1, even
921        when matches were found, and --line-buffered was not being honoured.
922    
923    12. Added an optional parentheses number to the -o and --only-matching options
924        of pcregrep.
925    
926    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
927        can match an empty string. The code to do it in pcretest and pcredemo
928        needed fixing:
929    
930        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
931            only one byte after an empty string match just before CRLF (this case
932            just got forgotten; "any" and "anycrlf" were OK).
933    
934        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
935            mode when an empty string match preceded an ASCII character followed by
936            a non-ASCII character. (The code for advancing by one character rather
937            than one byte was nonsense.)
938    
939        (c) The pcredemo.c sample program did not have any code at all to handle
940            the cases when CRLF is a valid newline sequence.
941    
942    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
943        as a starting offset was within the subject string. There is now a new
944        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
945        negative or greater than the length of the string. In order to test this,
946        pcretest is extended to allow the setting of negative starting offsets.
947    
948    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
949        starting offset points to the beginning of a UTF-8 character was
950        unnecessarily clumsy. I tidied it up.
951    
952    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
953        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
954    
955    17. Nobody had reported that the --include_dir option, which was added in
956        release 7.7 should have been called --include-dir (hyphen, not underscore)
957        for compatibility with GNU grep. I have changed it to --include-dir, but
958        left --include_dir as an undocumented synonym, and the same for
959        --exclude-dir, though that is not available in GNU grep, at least as of
960        release 2.5.4.
961    
962    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
963        characters from a string of bytes) have been redefined so as not to use
964        loops, in order to improve performance in some environments. At the same
965        time, I abstracted some of the common code into auxiliary macros to save
966        repetition (this should not affect the compiled code).
967    
968    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
969        compile-time error is now given if \c is not followed by an ASCII
970        character, that is, a byte less than 128. (In EBCDIC mode, the code is
971        different, and any byte value is allowed.)
972    
973    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
974        START_OPTIMIZE option, which is now allowed at compile time - but just
975        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
976        to pcregrep and other applications that have no direct access to PCRE
977        options. The new /Y option in pcretest sets this option when calling
978        pcre_compile().
979    
980    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
981        back references. Groups containing recursive back references were forced to
982        be atomic by that change, but in the case of named groups, the amount of
983        memory required was incorrectly computed, leading to "Failed: internal
984        error: code overflow". This has been fixed.
985    
986    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
987        pcretest.c, to avoid build problems in some Borland environments.
988    
989    
990    Version 8.10 25-Jun-2010
991    ------------------------
992    
993    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
994        THEN.
995    
996    2.  (*ACCEPT) was not working when inside an atomic group.
997    
998    3.  Inside a character class, \B is treated as a literal by default, but
999        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1000        causes the error). The code is unchanged, but I tidied the documentation.
1001    
1002    4.  Inside a character class, PCRE always treated \R and \X as literals,
1003        whereas Perl faults them if its -w option is set. I have changed PCRE so
1004        that it faults them when PCRE_EXTRA is set.
1005    
1006    5.  Added support for \N, which always matches any character other than
1007        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1008    
1009    6.  When compiling pcregrep with newer versions of gcc which may have
1010        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1011        declared with attribute warn_unused_result" were given. Just casting the
1012        result to (void) does not stop the warnings; a more elaborate fudge is
1013        needed. I've used a macro to implement this.
1014    
1015    7.  Minor change to pcretest.c to avoid a compiler warning.
1016    
1017    8.  Added four artifical Unicode properties to help with an option to make
1018        \s etc use properties (see next item). The new properties are: Xan
1019        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1020    
1021    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1022        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1023        this option. Modified pcretest to add /W to test this facility. Added
1024        REG_UCP to make it available via the POSIX interface.
1025    
1026    10. Added --line-buffered to pcregrep.
1027    
1028    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1029        studied, and the match started with a letter with a code point greater than
1030        127 whose first byte was different to the first byte of the other case of
1031        the letter, the other case of this starting letter was not recognized
1032        (#976).
1033    
1034    12. If a pattern that was studied started with a repeated Unicode property
1035        test, for example, \p{Nd}+, there was the theoretical possibility of
1036        setting up an incorrect bitmap of starting bytes, but fortunately it could
1037        not have actually happened in practice until change 8 above was made (it
1038        added property types that matched character-matching opcodes).
1039    
1040    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1041        possible starting bytes for non-anchored patterns.
1042    
1043    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1044        \R, and also a number of cases that involve Unicode properties, both
1045        explicit and implicit when PCRE_UCP is set.
1046    
1047    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1048        input, it could crash or give wrong results if characters with values
1049        greater than 0xc0 were present in the subject string. (Detail: it assumed
1050        UTF-8 input when processing these items.)
1051    
1052    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1053        size_t is 64-bit (#991).
1054    
1055    17. Added a check for running out of memory when PCRE is compiled with
1056        --disable-stack-for-recursion (#990).
1057    
1058    18. If the last data line in a file for pcretest does not have a newline on
1059        the end, a newline was missing in the output.
1060    
1061    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1062        less than 128) in its various bitmaps. However, there is a facility for
1063        generating tables according to the current locale when PCRE is compiled. It
1064        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1065        characters, are recognized by isspace() and therefore were getting set in
1066        these tables, and indeed these tables seem to approximate to ISO 8859. This
1067        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1068        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1069        which of course cannot start UTF-8 characters. I have changed the code so
1070        that only real ASCII characters (less than 128) and the correct starting
1071        bytes for UTF-8 encodings are set for characters greater than 127 when in
1072        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1073        altogether.)
1074    
1075    20. Added the /T option to pcretest so as to be able to run tests with non-
1076        standard character tables, thus making it possible to include the tests
1077        used for 19 above in the standard set of tests.
1078    
1079    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1080        reference to a subpattern the other side of a comment that contains an
1081        opening parenthesis caused either an internal compiling error, or a
1082        reference to the wrong subpattern.
1083    
1084    
1085    Version 8.02 19-Mar-2010
1086    ------------------------
1087    
1088    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1089    
1090    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1091        configured.
1092    
1093    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1094        original author of that file, following a query about its status.
1095    
1096    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1097        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1098    
1099    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1100        quantifier applied to a forward-referencing subroutine call, could compile
1101        incorrect code or give the error "internal error: previously-checked
1102        referenced subpattern not found".
1103    
1104    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1105        variables to point to external functions. For these systems, therefore,
1106        pcre_malloc etc. are now initialized to local functions that call the
1107        relevant global functions.
1108    
1109    7.  There were two entries missing in the vectors called coptable and poptable
1110        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1111        I've fixed the data, and added a kludgy way of testing at compile time that
1112        the lengths are correct (equal to the number of opcodes).
1113    
1114    8.  Following on from 7, I added a similar kludge to check the length of the
1115        eint vector in pcreposix.c.
1116    
1117    9.  Error texts for pcre_compile() are held as one long string to avoid too
1118        much relocation at load time. To find a text, the string is searched,
1119        counting zeros. There was no check for running off the end of the string,
1120        which could happen if a new error number was added without updating the
1121        string.
1122    
1123    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1124    
1125    11. \K was not working if it appeared in an atomic group or in a group that
1126        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1127        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1128        the assertion is positive, but not if it is negative.
1129    
1130    12. Change 11 fortuitously reduced the size of the stack frame used in the
1131        "match()" function of pcre_exec.c by one pointer. Forthcoming
1132        implementation of support for (*MARK) will need an extra pointer on the
1133        stack; I have reserved it now, so that the stack frame size does not
1134        decrease.
1135    
1136    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1137        item in branch that calls a recursion is a subroutine call - as in the
1138        second branch in the above example - was incorrectly given the compile-
1139        time error "recursive call could loop indefinitely" because pcre_compile()
1140        was not correctly checking the subroutine for matching a non-empty string.
1141    
1142    14. The checks for overrunning compiling workspace could trigger after an
1143        overrun had occurred. This is a "should never occur" error, but it can be
1144        triggered by pathological patterns such as hundreds of nested parentheses.
1145        The checks now trigger 100 bytes before the end of the workspace.
1146    
1147    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1148    
1149    
1150    Version 8.01 19-Jan-2010
1151    ------------------------
1152    
1153    1.  If a pattern contained a conditional subpattern with only one branch (in
1154        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1155        computed the wrong minimum data length (which is of course zero for such
1156        subpatterns). This could cause incorrect "no match" results.
1157    
1158    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1159        the pattern is reset in the first branch, pcre_compile() failed with
1160        "internal error: code overflow at offset...". This happened only when
1161        the reset was to the original external option setting. (An optimization
1162        abstracts leading options settings into an external setting, which was the
1163        cause of this.)
1164    
1165    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1166        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1167        assertion pattern did not match (meaning that the assertion was true), it
1168        was incorrectly treated as false if the SKIP had been reached during the
1169        matching. This also applied to assertions used as conditions.
1170    
1171    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1172        assertion subpattern, including such a pattern used as a condition,
1173        unpredictable results occurred, instead of the error return
1174        PCRE_ERROR_DFA_UITEM.
1175    
1176    5.  The C++ GlobalReplace function was not working like Perl for the special
1177        situation when an empty string is matched. It now does the fancy magic
1178        stuff that is necessary.
1179    
1180    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1181        removed. (These were left over from very, very early versions of PCRE.)
1182    
1183    7.  Some cosmetic changes to the code to make life easier when compiling it
1184        as part of something else:
1185    
1186        (a) Change DEBUG to PCRE_DEBUG.
1187    
1188        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1189            called "current" as "current_branch", to prevent a collision with the
1190            Linux macro when compiled as a kernel module.
1191    
1192        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1193            prevent a collision with the Linux macro when compiled as a kernel
1194            module.
1195    
1196    8.  In pcre_compile() there are some checks for integer overflows that used to
1197        cast potentially large values to (double). This has been changed to that
1198        when building, a check for int64_t is made, and if it is found, it is used
1199        instead, thus avoiding the use of floating point arithmetic. (There is no
1200        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1201        double.
1202    
1203    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1204        2005 (difference between two addresses compared to an unsigned value).
1205    
1206    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1207        custom one, because of the following reported problem in Windows:
1208    
1209          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1210              under Win32.
1211          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1212              therefore missing the function definition.
1213          - The compiler thus generates a "C" signature for the test function.
1214          - The linker fails to find the "C" function.
1215          - PCRE fails to configure if asked to do so against libbz2.
1216    
1217    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1218        messages were output:
1219    
1220          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1221          rerunning libtoolize, to keep the correct libtool macros in-tree.
1222          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1223    
1224        I have done both of these things.
1225    
1226    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1227        most of the time, it *can* run out if it is given a pattern that contains a
1228        runaway infinite recursion. I updated the discussion in the pcrestack man
1229        page.
1230    
1231    13. Now that we have gone to the x.xx style of version numbers, the minor
1232        version may start with zero. Using 08 or 09 is a bad idea because users
1233        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1234        interpreted as invalid octal numbers. I've updated the previous comment in
1235        configure.ac, and also added a check that gives an error if 08 or 09 are
1236        used.
1237    
1238    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1239        causing partial matching to fail when the end of the subject matched \W
1240        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1241    
1242    15. There were some discrepancies between the declarations in pcre_internal.h
1243        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1244        their definitions. The declarations used "const uschar *" and the
1245        definitions used USPTR. Even though USPTR is normally defined as "const
1246        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1247        reported that: "This difference in casting confuses some C++ compilers, for
1248        example, SunCC recognizes above declarations as different functions and
1249        generates broken code for hbpcre." I have changed the declarations to use
1250        USPTR.
1251    
1252    16. GNU libtool is named differently on some systems. The autogen.sh script now
1253        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1254        (FreeBSD).
1255    
1256    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1257        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1258        comment: "Figure out how to create a longlong from a string: strtoll and
1259        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1260        instance, but it only takes 2 args instead of 3!"
1261    
1262    18. A subtle bug concerned with back references has been fixed by a change of
1263        specification, with a corresponding code fix. A pattern such as
1264        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1265        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1266        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1267        same bug. Such groups have to be quantified to be useful, or contained
1268        inside another quantified group. (If there's no repetition, the reference
1269        can never match.) The problem arises because, having left the group and
1270        moved on to the rest of the pattern, a later failure that backtracks into
1271        the group uses the captured value from the final iteration of the group
1272        rather than the correct earlier one. I have fixed this in PCRE by forcing
1273        any group that contains a reference to itself to be an atomic group; that
1274        is, there cannot be any backtracking into it once it has completed. This is
1275        similar to recursive and subroutine calls.
1276    
1277    
1278    Version 8.00 19-Oct-09
1279    ----------------------
1280    
1281    1.  The table for translating pcre_compile() error codes into POSIX error codes
1282        was out-of-date, and there was no check on the pcre_compile() error code
1283        being within the table. This could lead to an OK return being given in
1284        error.
1285    
1286    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1287        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1288        in a Windows environment.
1289    
1290    3.  The pcregrep --count option prints the count for each file even when it is
1291        zero, as does GNU grep. However, pcregrep was also printing all files when
1292        --files-with-matches was added. Now, when both options are given, it prints
1293        counts only for those files that have at least one match. (GNU grep just
1294        prints the file name in this circumstance, but including the count seems
1295        more useful - otherwise, why use --count?) Also ensured that the
1296        combination -clh just lists non-zero counts, with no names.
1297    
1298    4.  The long form of the pcregrep -F option was incorrectly implemented as
1299        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1300        but it seems right to fix it, and I didn't think it was worth preserving
1301        the old behaviour.
1302    
1303    5.  The command line items --regex=pattern and --regexp=pattern were not
1304        recognized by pcregrep, which required --regex pattern or --regexp pattern
1305        (with a space rather than an '='). The man page documented the '=' forms,
1306        which are compatible with GNU grep; these now work.
1307    
1308    6.  No libpcreposix.pc file was created for pkg-config; there was just
1309        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1310    
1311    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1312        when UCP support is not needed, by modifying the Python script that
1313        generates it from Unicode data files. This should not matter if the module
1314        is correctly used as a library, but I received one complaint about 50K of
1315        unwanted data. My guess is that the person linked everything into his
1316        program rather than using a library. Anyway, it does no harm.
1317    
1318    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1319        was a minimum greater than 1 for a wide character in a possessive
1320        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1321        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1322        character. Chaos in the form of incorrect output or a compiling loop could
1323        result.
1324    
1325    9.  The restrictions on what a pattern can contain when partial matching is
1326        requested for pcre_exec() have been removed. All patterns can now be
1327        partially matched by this function. In addition, if there are at least two
1328        slots in the offset vector, the offset of the earliest inspected character
1329        for the match and the offset of the end of the subject are set in them when
1330        PCRE_ERROR_PARTIAL is returned.
1331    
1332    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1333        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1334        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1335        and may be more useful for multi-segment matching.
1336    
1337    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1338        used to be given if ever the end of the subject was reached; now it is
1339        given only if matching could not proceed because another character was
1340        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1341        string "Z", which now yields "no match" instead of "partial match". In the
1342        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1343        final character ended with (*FAIL).
1344    
1345    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1346        if the pattern had a "must contain" character that was already found in the
1347        earlier partial match, unless partial matching was again requested. For
1348        example, with the pattern /dog.(body)?/, the "must contain" character is
1349        "g". If the first part-match was for the string "dog", restarting with
1350        "sbody" failed. This bug has been fixed.
1351    
1352    13. The string returned by pcre_dfa_exec() after a partial match has been
1353        changed so that it starts at the first inspected character rather than the
1354        first character of the match. This makes a difference only if the pattern
1355        starts with a lookbehind assertion or \b or \B (\K is not supported by
1356        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1357        matching functions compatible, and I think it's the right thing to do.
1358    
1359    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1360        so that the demonstration program is easily available in environments where
1361        PCRE has not been installed from source.
1362    
1363    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1364        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1365        library.
1366    
1367    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1368        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1369        is not the first non-POSIX option to be added. Clearly some people find
1370        these options useful.
1371    
1372    17. If a caller to the POSIX matching function regexec() passes a non-zero
1373        value for nmatch with a NULL value for pmatch, the value of
1374        nmatch is forced to zero.
1375    
1376    18. RunGrepTest did not have a test for the availability of the -u option of
1377        the diff command, as RunTest does. It now checks in the same way as
1378        RunTest, and also checks for the -b option.
1379    
1380    19. If an odd number of negated classes containing just a single character
1381        interposed, within parentheses, between a forward reference to a named
1382        subpattern and the definition of the subpattern, compilation crashed with
1383        an internal error, complaining that it could not find the referenced
1384        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1385        [The bug was that it was starting one character too far in when skipping
1386        over the character class, thus treating the ] as data rather than
1387        terminating the class. This meant it could skip too much.]
1388    
1389    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1390        /g option in pcretest when the pattern contains \K, which makes it possible
1391        to have an empty string match not at the start, even when the pattern is
1392        anchored. Updated pcretest and pcredemo to use this option.
1393    
1394    21. If the maximum number of capturing subpatterns in a recursion was greater
1395        than the maximum at the outer level, the higher number was returned, but
1396        with unset values at the outer level. The correct (outer level) value is
1397        now given.
1398    
1399    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1400        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1401        make it do so. The string so far is captured, making this feature
1402        compatible with Perl.
1403    
1404    23. The tests have been re-organized, adding tests 11 and 12, to make it
1405        possible to check the Perl 5.10 features against Perl 5.10.
1406    
1407    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1408        pattern matches a fixed length string. PCRE did not allow this; now it
1409        does. Neither allows recursion.
1410    
1411    25. I finally figured out how to implement a request to provide the minimum
1412        length of subject string that was needed in order to match a given pattern.
1413        (It was back references and recursion that I had previously got hung up
1414        on.) This code has now been added to pcre_study(); it finds a lower bound
1415        to the length of subject needed. It is not necessarily the greatest lower
1416        bound, but using it to avoid searching strings that are too short does give
1417        some useful speed-ups. The value is available to calling programs via
1418        pcre_fullinfo().
1419    
1420    26. While implementing 25, I discovered to my embarrassment that pcretest had
1421        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1422        study optimizations had never been tested with that matching function.
1423        Oops. What is worse, even when it was passed study data, there was a bug in
1424        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1425        were also very few tests of studied patterns with pcre_dfa_exec().
1426    
1427    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1428        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1429        on the other side of the coin, they are no longer allowed to have different
1430        names, because these cannot be distinguished in PCRE, and this has caused
1431        confusion. (This is a difference from Perl.)
1432    
1433    28. When duplicate subpattern names are present (necessarily with different
1434        numbers, as required by 27 above), and a test is made by name in a
1435        conditional pattern, either for a subpattern having been matched, or for
1436        recursion in such a pattern, all the associated numbered subpatterns are
1437        tested, and the overall condition is true if the condition is true for any
1438        one of them. This is the way Perl works, and is also more like the way
1439        testing by number works.
1440    
1441    
1442    Version 7.9 11-Apr-09
1443    ---------------------
1444    
1445    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1446        (pcretest), all targets were linked against these libraries. This included
1447        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1448        libraries. This caused unwanted dependencies to be created. This problem
1449        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1450        pcretest is linked with readline.
1451    
1452    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1453        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1454        moved outside it again, because FALSE and TRUE are already defined in AIX,
1455        but BOOL is not.
1456    
1457    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1458        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1459    
1460    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1461        hyphens) following file names and line numbers when outputting matching
1462        lines. This is not true; no spaces are inserted. I have also clarified the
1463        wording for the --colour (or --color) option.
1464    
1465    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1466        was not coloured; this is different to GNU grep, so I have changed it to be
1467        the same.
1468    
1469    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1470        each matching line was coloured. Now it goes on to look for further matches
1471        of any of the test patterns, which is the same behaviour as GNU grep.
1472    
1473    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1474        doesn't make sense to accept an empty string match in pcregrep, so I have
1475        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1476        seems to be how GNU grep behaves.
1477    
1478    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1479        start or after a newline", because the conditional assertion was not being
1480        correctly handled. The rule now is that both the assertion and what follows
1481        in the first alternative must satisfy the test.
1482    
1483    9.  If auto-callout was enabled in a pattern with a conditional group whose
1484        condition was an assertion, PCRE could crash during matching, both with
1485        pcre_exec() and pcre_dfa_exec().
1486    
1487    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1488        used for matching.
1489    
1490    11. Unicode property support in character classes was not working for
1491        characters (bytes) greater than 127 when not in UTF-8 mode.
1492    
1493    12. Added the -M command line option to pcretest.
1494    
1495    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1496    
1497    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1498    
1499    16. Added comments and documentation about mis-use of no_arg in the C++
1500        wrapper.
1501    
1502    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1503        from Martin Jerabek that uses macro names for all relevant character and
1504        string constants.
1505    
1506    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1507        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1508        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1509        these, but not everybody uses configure.
1510    
1511    19. A conditional group that had only one branch was not being correctly
1512        recognized as an item that could match an empty string. This meant that an
1513        enclosing group might also not be so recognized, causing infinite looping
1514        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1515        with the subject "ab", where knowledge that the repeated group can match
1516        nothing is needed in order to break the loop.
1517    
1518    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1519        exec(), but without supplying a callout function, matching went wrong.
1520    
1521    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1522        leak if the size of the offset vector was greater than 30. When the vector
1523        is smaller, the saved offsets during recursion go onto a local stack
1524        vector, but for larger vectors malloc() is used. It was failing to free
1525        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1526        error, in fact).
1527    
1528    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1529        heapframe that is used only when UTF-8 support is enabled. This caused no
1530        problem, but was untidy.
1531    
1532    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1533        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1534        included within another project.
1535    
1536    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1537        slightly modified by me:
1538    
1539          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1540              not building pcregrep.
1541    
1542          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1543              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1544    
1545    25. Forward references, both numeric and by name, in patterns that made use of
1546        duplicate group numbers, could behave incorrectly or give incorrect errors,
1547        because when scanning forward to find the reference group, PCRE was not
1548        taking into account the duplicate group numbers. A pattern such as
1549        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1550    
1551    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1552        the feature of a custom pointer more persuasive (as requested by a user).
1553    
1554    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1555        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1556        pre-defined.
1557    
1558    28. Added support for (*UTF8) at the start of a pattern.
1559    
1560    29. Arrange for flags added by the "release type" setting in CMake to be shown
1561        in the configuration summary.
1562    
1563    
1564    Version 7.8 05-Sep-08
1565    ---------------------
1566    
1567    1.  Replaced UCP searching code with optimized version as implemented for Ad
1568        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1569        stage table and inline lookup instead of a function, giving speed ups of 2
1570        to 5 times on some simple patterns that I tested. Permission was given to
1571        distribute the MultiStage2.py script that generates the tables (it's not in
1572        the tarball, but is in the Subversion repository).
1573    
1574    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1575        scripts.
1576    
1577    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1578        a group with a zero qualifier. The result of the study could be incorrect,
1579        or the function might crash, depending on the pattern.
1580    
1581    4.  Caseless matching was not working for non-ASCII characters in back
1582        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1583        It now works when Unicode Property Support is available.
1584    
1585    5.  In pcretest, an escape such as \x{de} in the data was always generating
1586        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1587        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1588        truncation.
1589    
1590    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1591    
1592    7.  Added two (int) casts to pcregrep when printing the difference of two
1593        pointers, in case they are 64-bit values.
1594    
1595    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1596        test 2 if it fails.
1597    
1598    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1599        and a #define of that name to empty if it is not externally set. This is to
1600        allow users of MSVC to set it if necessary.
1601    
1602    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1603        the convenience functions in the pcre_get.c source file.
1604    
1605    11. An option change at the start of a pattern that had top-level alternatives
1606        could cause overwriting and/or a crash. This command provoked a crash in
1607        some environments:
1608    
1609          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1610    
1611        This potential security problem was recorded as CVE-2008-2371.
1612    
1613    12. For a pattern where the match had to start at the beginning or immediately
1614        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1615        pcre_dfa_exec() could read past the end of the passed subject if there was
1616        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1617        pcretest so that it places the subject at the end of its malloc-ed buffer.
1618    
1619    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1620        exec() might read past the end of the data buffer in UTF-8 mode.
1621    
1622    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1623        the data contained the byte 0x85 as part of a UTF-8 character within its
1624        first line. This applied both to normal and DFA matching.
1625    
1626    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1627        /^[^d]*?$/8 failed to match "abc".
1628    
1629    16. Added a missing copyright notice to pcrecpp_internal.h.
1630    
1631    17. Make it more clear in the documentation that values returned from
1632        pcre_exec() in ovector are byte offsets, not character counts.
1633    
1634    18. Tidied a few places to stop certain compilers from issuing warnings.
1635    
1636    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1637        supplied by Stefan Weber. I made a further small update for 7.8 because
1638        there is a change of source arrangements: the pcre_searchfuncs.c module is
1639        replaced by pcre_ucd.c.
1640    
1641    
1642    Version 7.7 07-May-08
1643    ---------------------
1644    
1645    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1646        a string to a long long, pretend we don't even have a long long." This is
1647        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1648    
1649    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1650        pre-7.6 versions, which defined a global no_arg variable instead of putting
1651        it in the RE class. (See also #8 below.)
1652    
1653    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1654        Lopes.
1655    
1656    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1657    
1658        (1) The include/exclude patterns were being applied to the whole pathnames
1659            of files, instead of just to the final components.
1660    
1661        (2) If there was more than one level of directory, the subdirectories were
1662            skipped unless they satisfied the include/exclude conditions. This is
1663            inconsistent with GNU grep (and could even be seen as contrary to the
1664            pcregrep specification - which I improved to make it absolutely clear).
1665            The action now is always to scan all levels of directory, and just
1666            apply the include/exclude patterns to regular files.
1667    
1668    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1669        --exclude_dir in the tests to avoid scanning .svn directories.
1670    
1671    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1672        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1673        doesn't support NULs in patterns.
1674    
1675    7.  Added some missing "const"s to declarations of static tables in
1676        pcre_compile.c and pcre_dfa_exec.c.
1677    
1678    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1679        caused by fix #2  above. (Subsequently also a second patch to fix the
1680        first patch. And a third patch - this was a messy problem.)
1681    
1682    9.  Applied Craig's patch to remove the use of push_back().
1683    
1684    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1685        matching function regexec().
1686    
1687    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1688        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1689        references. PCRE supports relative numbers with this syntax (I don't think
1690        Oniguruma does).
1691    
1692    12. Previously, a group with a zero repeat such as (...){0} was completely
1693        omitted from the compiled regex. However, this means that if the group
1694        was called as a subroutine from elsewhere in the pattern, things went wrong
1695        (an internal error was given). Such groups are now left in the compiled
1696        pattern, with a new opcode that causes them to be skipped at execution
1697        time.
1698    
1699    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1700        to the way PCRE behaves:
1701    
1702        (a) A lone ] character is dis-allowed (Perl treats it as data).
1703    
1704        (b) A back reference to an unmatched subpattern matches an empty string
1705            (Perl fails the current match path).
1706    
1707        (c) A data ] in a character class must be notated as \] because if the
1708            first data character in a class is ], it defines an empty class. (In
1709            Perl it is not possible to have an empty class.) The empty class []
1710            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1711            The negative empty class [^] matches any one character, independently
1712            of the DOTALL setting.
1713    
1714    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1715        non-existent subpattern following a character class starting with ']' and
1716        containing () gave an internal compiling error instead of "reference to
1717        non-existent subpattern". Fortunately, when the pattern did exist, the
1718        compiled code was correct. (When scanning forwards to check for the
1719        existencd of the subpattern, it was treating the data ']' as terminating
1720        the class, so got the count wrong. When actually compiling, the reference
1721        was subsequently set up correctly.)
1722    
1723    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1724        it was being rejected as not supported by pcre_dfa_exec(), even though
1725        other assertions are supported. I have made pcre_dfa_exec() support
1726        (*FAIL).
1727    
1728    16. The implementation of 13c above involved the invention of a new opcode,
1729        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1730        cannot be changed at match time, I realized I could make a small
1731        improvement to matching performance by compiling OP_ALLANY instead of
1732        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1733        on the OP_ANY path.
1734    
1735    17. Compiling pcretest on Windows with readline support failed without the
1736        following two fixes: (1) Make the unistd.h include conditional on
1737        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1738    
1739    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1740        ncurses library to be included for pcretest when ReadLine support is
1741        requested, but also to allow for it to be overridden. This patch came from
1742        Daniel Bergström.
1743    
1744    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1745        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1746        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1747        spotting this.
1748    
1749    
1750    Version 7.6 28-Jan-08
1751    ---------------------
1752    
1753    1.  A character class containing a very large number of characters with
1754        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1755        overflow.
1756    
1757    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1758        HAVE_LONG_LONG is not defined.
1759    
1760    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1761        bring them up to date and include new features. This patch includes:
1762    
1763        - Fixed PH's badly added libz and libbz2 support.
1764        - Fixed a problem with static linking.
1765        - Added pcredemo. [But later removed - see 7 below.]
1766        - Fixed dftables problem and added an option.
1767        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1768            HAVE_LONG_LONG.
1769        - Added readline support for pcretest.
1770        - Added an listing of the option settings after cmake has run.
1771    
1772    4.  A user submitted a patch to Makefile that makes it easy to create
1773        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1774        Makefile.am that cause it to include this special target, without
1775        affecting anything else. Note that the same mingw target plus all
1776        the other distribution libraries and programs are now supported
1777        when configuring with CMake (see 6 below) instead of with
1778        Configure/Make.
1779    
1780    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1781        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1782        exported in the Windows port". It has not yet been confirmed that the patch
1783        solves the problem, but it does no harm.
1784    
1785    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1786        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1787        with CMake, and also correct the comment about stack recursion.
1788    
1789    7.  Remove the automatic building of pcredemo from the ./configure system and
1790        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1791        of a program that users should build themselves after PCRE is installed, so
1792        building it automatically is not really right. What is more, it gave
1793        trouble in some build environments.
1794    
1795    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1796    
1797    
1798    Version 7.5 10-Jan-08
1799    ---------------------
1800    
1801    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1802        values in parens when parsing an RE using the C++ wrapper."
1803    
1804    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1805        Characters greater than 255 were excluded from the class instead of being
1806        included.
1807    
1808    3.  The same bug as (2) above applied to negated POSIX classes such as
1809        [:^space:].
1810    
1811    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1812        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1813        I have changed it.
1814    
1815    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1816        first named subpattern) and a construct such as (?&a) would reference the
1817        first named subpattern whose name started with "a" (in other words, the
1818        length check was missing). Both these problems are fixed. "Subpattern name
1819        expected" is now given for (?&) (a zero-length name), and this patch also
1820        makes it give the same error for \k'' (previously it complained that that
1821        was a reference to a non-existent subpattern).
1822    
1823    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1824        this is right because (?- can be followed by option settings as well as by
1825        digits. I have, however, made the messages clearer.
1826    
1827    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1828        than the number used in the conditional) now cause a compile-time error.
1829        This is actually not compatible with Perl, which accepts such patterns, but
1830        treats the conditional as always being FALSE (as PCRE used to), but it
1831        seems to me that giving a diagnostic is better.
1832    
1833    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1834        and messages.
1835    
1836    9.  Fix two occurrences of "backslash" in comments that should have been
1837        "backspace".
1838    
1839    10. Remove two redundant lines of code that can never be obeyed (their function
1840        was moved elsewhere).
1841    
1842    11. The program that makes PCRE's Unicode character property table had a bug
1843        which caused it to generate incorrect table entries for sequences of
1844        characters that have the same character type, but are in different scripts.
1845        It amalgamated them into a single range, with the script of the first of
1846        them. In other words, some characters were in the wrong script. There were
1847        thirteen such cases, affecting characters in the following ranges:
1848    
1849          U+002b0 - U+002c1
1850          U+0060c - U+0060d
1851          U+0061e - U+00612
1852          U+0064b - U+0065e
1853          U+0074d - U+0076d
1854          U+01800 - U+01805
1855          U+01d00 - U+01d77
1856          U+01d9b - U+01dbf
1857          U+0200b - U+0200f
1858          U+030fc - U+030fe
1859          U+03260 - U+0327f
1860          U+0fb46 - U+0fbb1
1861          U+10450 - U+1049d
1862    
1863    12. The -o option (show only the matching part of a line) for pcregrep was not
1864        compatible with GNU grep in that, if there was more than one match in a
1865        line, it showed only the first of them. It now behaves in the same way as
1866        GNU grep.
1867    
1868    13. If the -o and -v options were combined for pcregrep, it printed a blank
1869        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1870        does the same. The return code can be used to tell if there were any
1871        non-matching lines.
1872    
1873    14. Added --file-offsets and --line-offsets to pcregrep.
1874    
1875    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1876        infinitely looping recursion. The bug was that positive lookaheads were not
1877        being skipped when checking for a possible empty match (negative lookaheads
1878        and both kinds of lookbehind were skipped).
1879    
1880    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1881        inclusion of <windows.h> to before rather than after the definition of
1882        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1883    
1884    17. Specifying a possessive quantifier with a specific limit for a Unicode
1885        character property caused pcre_compile() to compile bad code, which led at
1886        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1887        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1888        caused the error; without that there was no problem.
1889    
1890    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1891    
1892    19. Added --enable-pcretest-libreadline.
1893    
1894    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1895        RE::GlobalReplace(). As a result, the number of replacements returned was
1896        double what it should be. I removed one of the increments, but Craig sent a
1897        later patch that removed the other one (the right fix) and added unit tests
1898        that check the return values (which was not done before).
1899    
1900    21. Several CMake things:
1901    
1902        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1903            the names libpcre and libpcreposix, not just pcre and pcreposix.
1904    
1905        (2) The above change means that pcretest and pcregrep are now correctly
1906            linked with the newly-built libraries, not previously installed ones.
1907    
1908        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1909    
1910    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1911        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1912        UTF-8 newline character). The key issue is that the pattern starts .*;
1913        this means that the match must be either at the beginning, or after a
1914        newline. The bug was in the code for advancing after a failed match and
1915        checking that the new position followed a newline. It was not taking
1916        account of UTF-8 characters correctly.
1917    
1918    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1919        character classes. PCRE was not treating the sequence [:...:] as a
1920        character class unless the ... were all letters. Perl, however, seems to
1921        allow any characters between [: and :], though of course it rejects as
1922        unknown any "names" that contain non-letters, because all the known class
1923        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1924        for example, whereas PCRE did not - it did not recognize a POSIX character
1925        class. This seemed a bit dangerous, so the code has been changed to be
1926        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1927        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1928        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1929        Perl does, and where it didn't before.
1930    
1931    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1932        Windows environments %n is disabled by default.
1933    
1934    
1935    Version 7.4 21-Sep-07
1936    ---------------------
1937    
1938    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1939        means that a class such as [\s] counted as "explicit reference to CR or
1940        LF". That isn't really right - the whole point of the change was to try to
1941        help when there was an actual mention of one of the two characters. So now
1942        the change happens only if \r or \n (or a literal CR or LF) character is
1943        encountered.
1944    
1945    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1946        of both had grown to the point where there were only 3 bits left.
1947        Fortunately, there was spare space in the data structure, and so I have
1948        moved the internal flags into a new 16-bit field to free up more option
1949        bits.
1950    
1951    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1952        but did not set the internal JCHANGED flag - either of these is enough to
1953        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1954        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1955        start sets both bits.
1956    
1957    4.  Added options (at build time, compile time, exec time) to change \R from
1958        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1959    
1960    5.  doc/pcresyntax.html was missing from the distribution.
1961    
1962    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1963        compatibility, even though it is no longer used.
1964    
1965    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1966        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1967        windows.h file is present (where different names are used). [This was
1968        reversed later after testing - see 16 below.]
1969    
1970    8.  Changed all #include <config.h> to #include "config.h". There were also
1971        some further <pcre.h> cases that I changed to "pcre.h".
1972    
1973    9.  When pcregrep was used with the --colour option, it missed the line ending
1974        sequence off the lines that it output.
1975    
1976    10. It was pointed out to me that arrays of string pointers cause lots of
1977        relocations when a shared library is dynamically loaded. A technique of
1978        using a single long string with a table of offsets can drastically reduce
1979        these. I have refactored PCRE in four places to do this. The result is
1980        dramatic:
1981    
1982          Originally:                          290
1983          After changing UCP table:            187
1984          After changing error message table:   43
1985          After changing table of "verbs"       36
1986          After changing table of Posix names   22
1987    
1988        Thanks to the folks working on Gregex for glib for this insight.
1989    
1990    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1991        unicode-properties was also set.
1992    
1993    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1994    
1995    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1996        checked only for CRLF.
1997    
1998    14. Added casts to pcretest.c to avoid compiler warnings.
1999    
2000    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2001    
2002    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2003        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2004        entirely. This removes changes made in 7 above.
2005    
2006    17. The CMake files have been updated, and there is now more information about
2007        building with CMake in the NON-UNIX-USE document.
2008    
2009    
2010    Version 7.3 28-Aug-07
2011    ---------------------
2012    
2013     1. In the rejigging of the build system that eventually resulted in 7.1, the
2014        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2015        brackets there is not right, since it causes compilers to look for an
2016        installed pcre.h, not the version that is in the source that is being
2017        compiled (which of course may be different). I have changed it back to:
2018    
2019          #include "pcre.h"
2020    
2021        I have a vague recollection that the change was concerned with compiling in
2022        different directories, but in the new build system, that is taken care of
2023        by the VPATH setting the Makefile.
2024    
2025     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2026        when the subject happened to end in the byte 0x85 (e.g. if the last
2027        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2028        characters but of course it shouldn't be taken as a newline when it is part
2029        of another character. The bug was that, for an unlimited repeat of . in
2030        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2031        characters when looking for a newline.
2032    
2033     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2034    
2035     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2036        in debug output.
2037    
2038     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2039        long printing in the pcrecpp unittest when running under MinGW.
2040    
2041     6. ESC_K was left out of the EBCDIC table.
2042    
2043     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2044        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2045        limit also applies to "virtual nesting" when a pattern is recursive, and in
2046        this case 1000 isn't so big. I have been able to remove this limit at the
2047        expense of backing off one optimization in certain circumstances. Normally,
2048        when pcre_exec() would call its internal match() function recursively and
2049        immediately return the result unconditionally, it uses a "tail recursion"
2050        feature to save stack. However, when a subpattern that can match an empty
2051        string has an unlimited repetition quantifier, it no longer makes this
2052        optimization. That gives it a stack frame in which to save the data for
2053        checking that an empty string has been matched. Previously this was taken
2054        from the 1000-entry workspace that had been reserved. So now there is no
2055        explicit limit, but more stack is used.
2056    
2057     8. Applied Daniel's patches to solve problems with the import/export magic
2058        syntax that is required for Windows, and which was going wrong for the
2059        pcreposix and pcrecpp parts of the library. These were overlooked when this
2060        problem was solved for the main library.
2061    
2062     9. There were some crude static tests to avoid integer overflow when computing
2063        the size of patterns that contain repeated groups with explicit upper
2064        limits. As the maximum quantifier is 65535, the maximum group length was
2065        set at 30,000 so that the product of these two numbers did not overflow a
2066        32-bit integer. However, it turns out that people want to use groups that
2067        are longer than 30,000 bytes (though not repeat them that many times).
2068        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2069        made it possible to implement the integer overflow checks in a much more
2070        dynamic way, which I have now done. The artificial limitation on group
2071        length has been removed - we now have only the limit on the total length of
2072        the compiled pattern, which depends on the LINK_SIZE setting.
2073    
2074    10. Fixed a bug in the documentation for get/copy named substring when
2075        duplicate names are permitted. If none of the named substrings are set, the
2076        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2077        empty string.
2078    
2079    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2080        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2081        because the ] is interpreted as the first data character and the
2082        terminating ] is not found. PCRE has been made compatible with Perl in this
2083        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2084        cause memory overwriting.
2085    
2086    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2087        string has been matched (to stop an infinite loop). It was not recognizing
2088        a conditional subpattern that could match an empty string if that
2089        subpattern was within another subpattern. For example, it looped when
2090        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2091        condition was not nested. This bug has been fixed.
2092    
2093    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2094        past the start of the subject in the presence of bytes with the top bit
2095        set, for example "\x8aBCD".
2096    
2097    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2098        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2099    
2100    14. Optimized (?!) to (*FAIL).
2101    
2102    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2103        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2104        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2105        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2106        does: it's just the validity check that is more restrictive.
2107    
2108    16. Inserted checks for integer overflows during escape sequence (backslash)
2109        processing, and also fixed erroneous offset values for syntax errors during
2110        backslash processing.
2111    
2112    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2113        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2114    
2115    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2116        caused an overrun.
2117    
2118    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2119        something other than just ASCII characters) inside a group that had an
2120        unlimited repeat caused a loop at compile time (while checking to see
2121        whether the group could match an empty string).
2122    
2123    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2124        [\P{Any}] did so. (Error in the code for printing property names.)
2125    
2126    21. An orphan \E inside a character class could cause a crash.
2127    
2128    22. A repeated capturing bracket such as (A)? could cause a wild memory
2129        reference during compilation.
2130    
2131    23. There are several functions in pcre_compile() that scan along a compiled
2132        expression for various reasons (e.g. to see if it's fixed length for look
2133        behind). There were bugs in these functions when a repeated \p or \P was
2134        present in the pattern. These operators have additional parameters compared
2135        with \d, etc, and these were not being taken into account when moving along
2136        the compiled data. Specifically:
2137    
2138        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2139            length.
2140    
2141        (b) An item such as \pL+ within a repeated group could cause crashes or
2142            loops.
2143    
2144        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2145            "reference to non-existent subpattern" error.
2146    
2147        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2148    
2149    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2150        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2151    
2152    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2153    
2154    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2155        character were causing crashes (broken optimization).
2156    
2157    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2158        \p or \P) caused a compile-time loop.
2159    
2160    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2161        break. For example, the unstudied pattern [\r\n]A does not match the string
2162        "\r\nA" because change 7.0/46 below moves the current point on by two
2163        characters after failing to match at the start. However, the pattern \nA
2164        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2165        the same is true. There doesn't seem any very clean way out of this, but
2166        what I have chosen to do makes the common cases work: PCRE now takes note
2167        of whether there can be an explicit match for \r or \n anywhere in the
2168        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2169        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2170        pattern has explicit CR or LF references.
2171    
2172    29. Added (*CR) etc for changing newline setting at start of pattern.
2173    
2174    
2175    Version 7.2 19-Jun-07
2176    ---------------------
2177    
2178     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2179        which is apparently normally available under Windows.
2180    
2181     2. Re-jig the pcregrep tests with different newline settings in an attempt
2182        to make them independent of the local environment's newline setting.
2183    
2184     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2185    
2186     4. Some of the "internals" tests were previously cut out when the link size
2187        was not 2, because the output contained actual offsets. The recent new
2188        "Z" feature of pcretest means that these can be cut out, making the tests
2189        usable with all link sizes.
2190    
2191     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2192        stack recursion. This gives a massive performance boost under BSD, but just
2193        a small improvement under Linux. However, it saves one field in the frame
2194        in all cases.
2195    
2196     6. Added more features from the forthcoming Perl 5.10:
2197    
2198        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2199            recursion call. It refers to the nth most recently opened parentheses.
2200    
2201        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2202            to be opened parentheses.
2203    
2204        (c) Conditions that refer to capturing parentheses can be specified
2205            relatively, for example, (?(-2)... or (?(+3)...
2206    
2207        (d) \K resets the start of the current match so that everything before
2208            is not part of it.
2209    
2210        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2211    
2212        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2213            reference syntax.
2214    
2215        (g) (?| introduces a group in which the numbering of parentheses in each
2216            alternative starts with the same number.
2217    
2218        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2219    
2220     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2221        PCRE_INFO_JCHANGED.
2222    
2223     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2224        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2225        for detecting groups that can match an empty string.
2226    
2227     9. A pattern with a very large number of alternatives (more than several
2228        hundred) was running out of internal workspace during the pre-compile
2229        phase, where pcre_compile() figures out how much memory will be needed. A
2230        bit of new cunning has reduced the workspace needed for groups with
2231        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2232        workspace instead of running out of the 4096 that are available.
2233    
2234    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2235    
2236    11. Applied patch from Google to remove an optimization that didn't quite work.
2237        The report of the bug said:
2238    
2239          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2240          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2241          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2242    
2243    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2244        it matched the wrong number of bytes.
2245    
2246    
2247    Version 7.1 24-Apr-07
2248  ---------------------  ---------------------
2249    
2250   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one

Legend:
Removed from v.153  
changed lines
  Added in v.1172

  ViewVC Help
Powered by ViewVC 1.1.5