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

Diff of /code/trunk/ChangeLog

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

revision 182 by ph10, Wed Jun 13 15:09:54 2007 UTC revision 1211 by ph10, Wed Nov 7 17:02:03 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.2 13-June-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    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
133        pcre_compile.c when checking for a zero character.
134    
135    28. Introducing a native interface for JIT. Through this interface, the compiled
136        machine code can be directly executed. The purpose of this interface is to
137        provide fast pattern matching, so several sanity checks are not performed.
138        However, feature tests are still performed. The new interface provides
139        1.4x speedup compared to the old one.
140    
141    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
142        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
143        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
144    
145    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
146        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
147        the "old" RFC 2279). Instead, it ended up passing a negative length to
148        pcre_exec().
149    
150    31. Add support for GCC's visibility feature to hide internal functions.
151    
152    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
153        "unknown -C option" after outputting 0 or 1.
154    
155    
156    Version 8.31 06-July-2012
157    -------------------------
158    
159    1.  Fixing a wrong JIT test case and some compiler warnings.
160    
161    2.  Removed a bashism from the RunTest script.
162    
163    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
164        to unsigned type, result still unsigned" that was given by an MS compiler
165        on encountering the code "-sizeof(xxx)".
166    
167    4.  Partial matching support is added to the JIT compiler.
168    
169    5.  Fixed several bugs concerned with partial matching of items that consist
170        of more than one character:
171    
172        (a) /^(..)\1/ did not partially match "aba" because checking references was
173            done on an "all or nothing" basis. This also applied to repeated
174            references.
175    
176        (b) \R did not give a hard partial match if \r was found at the end of the
177            subject.
178    
179        (c) \X did not give a hard partial match after matching one or more
180            characters at the end of the subject.
181    
182        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
183            a partial match for the string "\r".
184    
185        (e) When newline was set to CRLF, the metacharacter "." did not recognize
186            a partial match for a CR character at the end of the subject string.
187    
188    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
189        running pcretest, the text "(JIT)" added to the output whenever JIT is
190        actually used to run the match.
191    
192    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
193        or /S+[+] with a digit between 1 and 7.
194    
195    8.  OP_NOT now supports any UTF character not just single-byte ones.
196    
197    9.  (*MARK) control verb is now supported by the JIT compiler.
198    
199    10. The command "./RunTest list" lists the available tests without actually
200        running any of them. (Because I keep forgetting what they all are.)
201    
202    11. Add PCRE_INFO_MAXLOOKBEHIND.
203    
204    12. Applied a (slightly modified) user-supplied patch that improves performance
205        when the heap is used for recursion (compiled with --disable-stack-for-
206        recursion). Instead of malloc and free for each heap frame each time a
207        logical recursion happens, frames are retained on a chain and re-used where
208        possible. This sometimes gives as much as 30% improvement.
209    
210    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
211        call.
212    
213    14. As documented, (*COMMIT) is now confined to within a positive assertion.
214    
215    15. It is now possible to link pcretest with libedit as an alternative to
216        libreadline.
217    
218    16. (*COMMIT) control verb is now supported by the JIT compiler.
219    
220    17. The Unicode data tables have been updated to Unicode 6.1.0.
221    
222    18. Added --file-list option to pcregrep.
223    
224    19. Added binary file support to pcregrep, including the -a, --binary-files,
225        -I, and --text options.
226    
227    20. The madvise function is renamed for posix_madvise for QNX compatibility
228        reasons. Fixed by Giuseppe D'Angelo.
229    
230    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
231    
232    22. Changed the option for creating symbolic links for 16-bit man pages from
233        -s to -sf so that re-installing does not cause issues.
234    
235    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
236    
237    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
238        matches in certain environments (the workspace was not being correctly
239        retained). Also added to pcre_dfa_exec() a simple plausibility check on
240        some of the workspace data at the beginning of a restart.
241    
242    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
243        was not doing so when it should - probably a typo introduced by SVN 528
244        (change 8.10/14).
245    
246    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
247        \w+ when the character tables indicated that \x{c4} was a word character.
248        There were several related cases, all because the tests for doing a table
249        lookup were testing for characters less than 127 instead of 255.
250    
251    27. If a pattern contains capturing parentheses that are not used in a match,
252        their slots in the ovector are set to -1. For those that are higher than
253        any matched groups, this happens at the end of processing. In the case when
254        there were back references that the ovector was too small to contain
255        (causing temporary malloc'd memory to be used during matching), and the
256        highest capturing number was not used, memory off the end of the ovector
257        was incorrectly being set to -1. (It was using the size of the temporary
258        memory instead of the true size.)
259    
260    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
261        ovector size, it uses memory at the end of the block that it has got.
262    
263    29. Check for an overlong MARK name and give an error at compile time. The
264        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
265    
266    30. JIT compiler update.
267    
268    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
269        Rill for the patch.
270    
271    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
272    
273    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
274    
275    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
276        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
277        was enabled.
278    
279    35. Improve JIT code generation for greedy plus quantifier.
280    
281    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
282        1 to "aa" instead of to an empty string. The bug affected repeated groups
283        that could potentially match an empty string.
284    
285    37. Optimizing single character iterators in JIT.
286    
287    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
288        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
289        codepoints that are too big for the mode are faulted, and in a UTF mode,
290        disallowed codepoints are also faulted.
291    
292    39. If PCRE was compiled with UTF support, in three places in the DFA
293        matcher there was code that should only have been obeyed in UTF mode, but
294        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
295        processing when bytes with values greater than 127 were present. In 16-bit
296        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
297        both cases the values are those that cannot be the first data item in a UTF
298        character. The three items that might have provoked this were recursions,
299        possessively repeated groups, and atomic groups.
300    
301    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
302        and pcregrep, because some OS require shared objects to be explicitly
303        passed to ld, causing the link step to fail if they are not.
304    
305    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
306        mode, patterns that started with \h* or \R* might be incorrectly matched.
307    
308    
309    Version 8.30 04-February-2012
310    -----------------------------
311    
312    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
313        name is defined in ctype.h.
314    
315    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
316        only in quite long subpatterns.
317    
318    3.  Removed the function pcre_info(), which has been obsolete and deprecated
319        since it was replaced by pcre_fullinfo() in February 2000.
320    
321    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
322        match a (*MARK), and the match failed at the start of the subject, a
323        reference to memory before the start of the subject could occur. This bug
324        was introduced by fix 17 of release 8.21.
325    
326    5.  A reference to an unset group with zero minimum repetition was giving
327        totally wrong answers (in non-JavaScript-compatibility mode). For example,
328        /(another)?(\1?)test/ matched against "hello world test". This bug was
329        introduced in release 8.13.
330    
331    6.  Add support for 16-bit character strings (a large amount of work involving
332        many changes and refactorings).
333    
334    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
335        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
336        from a file.
337    
338    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
339        rounding is not applied in this particular case).
340    
341    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
342        if they appear, or are escaped, in patterns.
343    
344    10. Get rid of a number of -Wunused-but-set-variable warnings.
345    
346    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
347        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
348        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
349        also returns the mark "x". This bug applied to capturing parentheses,
350        non-capturing parentheses, and atomic parentheses. It also applied to some
351        assertions.
352    
353    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
354        information out of configure.ac instead of relying on pcre.h.generic, which
355        is not stored in the repository.
356    
357    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
358        -lreadline.
359    
360    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
361    
362    15. Applied Graycode's patch to put the top-level frame on the stack rather
363        than the heap when not using the stack for recursion. This gives a
364        performance improvement in many cases when recursion is not deep.
365    
366    16. Experimental code added to "pcretest -C" to output the stack frame size.
367    
368    
369    Version 8.21 12-Dec-2011
370    ------------------------
371    
372    1.  Updating the JIT compiler.
373    
374    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
375        are added as well.
376    
377    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
378        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
379        calling _pcre_jit_exec. Some extra comments are added.
380    
381    4.  (*MARK) settings inside atomic groups that do not contain any capturing
382        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
383        was introduced by change 18 for 8.20.
384    
385    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
386        ECMA-262 standard.
387    
388    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
389        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
390        This bug was probably introduced by change 9 of 8.13.
391    
392    7.  While fixing 6 above, I noticed that a number of other items were being
393        incorrectly rejected as "not fixed length". This arose partly because newer
394        opcodes had not been added to the fixed-length checking code. I have (a)
395        corrected the bug and added tests for these items, and (b) arranged for an
396        error to occur if an unknown opcode is encountered while checking for fixed
397        length instead of just assuming "not fixed length". The items that were
398        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
399        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
400        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
401    
402    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
403        being incorrectly compiled and would have given unpredicatble results.
404    
405    9.  A possessively repeated subpattern with minimum repeat count greater than
406        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
407        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
408        the first (A) could occur when it should not.
409    
410    10. Add a cast and remove a redundant test from the code.
411    
412    11. JIT should use pcre_malloc/pcre_free for allocation.
413    
414    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
415        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
416        is anything other than /usr, -L is still shown).
417    
418    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
419    
420    14. Perl does not support \N without a following name in a [] class; PCRE now
421        also gives an error.
422    
423    15. If a forward reference was repeated with an upper limit of around 2000,
424        it caused the error "internal error: overran compiling workspace". The
425        maximum number of forward references (including repeats) was limited by the
426        internal workspace, and dependent on the LINK_SIZE. The code has been
427        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
428        the default depends on LINK_SIZE. There is a new upper limit (for safety)
429        of around 200,000 forward references. While doing this, I also speeded up
430        the filling in of repeated forward references.
431    
432    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
433        incorrectly expecting the subject to contain another "a" after the start.
434    
435    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
436        in the match, the SKIP should be ignored. This was not happening; instead
437        the SKIP was being treated as NOMATCH. For patterns such as
438        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
439        tested.
440    
441    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
442        now much more compatible with Perl, in particular in cases where the result
443        is a non-match for a non-anchored pattern. For example, if
444        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
445        "m", where previously it did not return a name. A side effect of this
446        change is that for partial matches, the last encountered mark name is
447        returned, as for non matches. A number of tests that were previously not
448        Perl-compatible have been moved into the Perl-compatible test files. The
449        refactoring has had the pleasing side effect of removing one argument from
450        the match() function, thus reducing its stack requirements.
451    
452    19. If the /S+ option was used in pcretest to study a pattern using JIT,
453        subsequent uses of /S (without +) incorrectly behaved like /S+.
454    
455    21. Retrieve executable code size support for the JIT compiler and fixing
456        some warnings.
457    
458    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
459        not work when the shorter character appeared right at the end of the
460        subject string.
461    
462    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
463        systems.
464    
465    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
466        output it when the /M option is used in pcretest.
467    
468    25. The CheckMan script was not being included in the distribution. Also, added
469        an explicit "perl" to run Perl scripts from the PrepareRelease script
470        because this is reportedly needed in Windows.
471    
472    26. If study data was being save in a file and studying had not found a set of
473        "starts with" bytes for the pattern, the data written to the file (though
474        never used) was taken from uninitialized memory and so caused valgrind to
475        complain.
476    
477    27. Updated RunTest.bat as provided by Sheri Pierce.
478    
479    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
480    
481    29. Computation of memory usage for the table of capturing group names was
482        giving an unnecessarily large value.
483    
484    
485    Version 8.20 21-Oct-2011
486    ------------------------
487    
488    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
489        a POSIX class. After further experiments with Perl, which convinced me that
490        Perl has bugs and confusions, a closing square bracket is no longer allowed
491        in a POSIX name. This bug also affected patterns with classes that started
492        with full stops.
493    
494    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
495        captured substring, but while checking the failing first alternative,
496        substring 1 is temporarily captured. If the output vector supplied to
497        pcre_exec() was not big enough for this capture, the yield of the function
498        was still zero ("insufficient space for captured substrings"). This cannot
499        be totally fixed without adding another stack variable, which seems a lot
500        of expense for a edge case. However, I have improved the situation in cases
501        such as /(a)(b)x|abc/ matched against "abc", where the return code
502        indicates that fewer than the maximum number of slots in the ovector have
503        been set.
504    
505    3.  Related to (2) above: when there are more back references in a pattern than
506        slots in the output vector, pcre_exec() uses temporary memory during
507        matching, and copies in the captures as far as possible afterwards. It was
508        using the entire output vector, but this conflicts with the specification
509        that only 2/3 is used for passing back captured substrings. Now it uses
510        only the first 2/3, for compatibility. This is, of course, another edge
511        case.
512    
513    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
514        main code base, and can be used by building with --enable-jit. When this is
515        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
516        runtime --no-jit option is given.
517    
518    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
519        ovector, the return from the function was zero, implying that there were
520        other matches that did not fit. The correct "exactly full" value is now
521        returned.
522    
523    6.  If a subpattern that was called recursively or as a subroutine contained
524        (*PRUNE) or any other control that caused it to give a non-standard return,
525        invalid errors such as "Error -26 (nested recursion at the same subject
526        position)" or even infinite loops could occur.
527    
528    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
529        computing the minimum length on reaching *ACCEPT, and so ended up with the
530        wrong value of 1 rather than 0. Further investigation indicates that
531        computing a minimum subject length in the presence of *ACCEPT is difficult
532        (think back references, subroutine calls), and so I have changed the code
533        so that no minimum is registered for a pattern that contains *ACCEPT.
534    
535    8.  If (*THEN) was present in the first (true) branch of a conditional group,
536        it was not handled as intended. [But see 16 below.]
537    
538    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
539        Sheri Pierce.
540    
541    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
542        the first byte in a match must be "a".
543    
544    11. Change 17 for 8.13 increased the recursion depth for patterns like
545        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
546        pattern contains any instances of (*THEN). If it does not, the old
547        optimizations are restored. It would be nice to do this on a per-group
548        basis, but at the moment that is not feasible.
549    
550    12. In some environments, the output of pcretest -C is CRLF terminated. This
551        broke RunTest's code that checks for the link size. A single white space
552        character after the value is now allowed for.
553    
554    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
555        For "fr", it uses the Windows-specific input and output files.
556    
557    14. If (*THEN) appeared in a group that was called recursively or as a
558        subroutine, it did not work as intended. [But see next item.]
559    
560    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
561        pattern fragments (but not containing any | characters). If A and B are
562        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
563        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
564        D. In other words, Perl considers parentheses that do not contain any |
565        characters to be part of a surrounding alternative, whereas PCRE was
566        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
567        differently. PCRE now behaves in the same way as Perl, except in the case
568        of subroutine/recursion calls such as (?1) which have in any case always
569        been different (but PCRE had them first :-).
570    
571    16. Related to 15 above: Perl does not treat the | in a conditional group as
572        creating alternatives. Such a group is treated in the same way as an
573        ordinary group without any | characters when processing (*THEN). PCRE has
574        been changed to match Perl's behaviour.
575    
576    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
577        RunGrepTest script failed.
578    
579    18. Change 22 for version 13 caused atomic groups to use more stack. This is
580        inevitable for groups that contain captures, but it can lead to a lot of
581        stack use in large patterns. The old behaviour has been restored for atomic
582        groups that do not contain any capturing parentheses.
583    
584    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
585        suppress the check for a minimum subject length at run time. (If it was
586        given to pcre_exec() or pcre_dfa_exec() it did work.)
587    
588    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
589        fail to work when decoding hex characters in data strings in EBCDIC
590        environments.
591    
592    21. It appears that in at least one Mac OS environment, the isxdigit() function
593        is implemented as a macro that evaluates to its argument more than once,
594        contravening the C 90 Standard (I haven't checked a later standard). There
595        was an instance in pcretest which caused it to go wrong when processing
596        \x{...} escapes in subject strings. The has been rewritten to avoid using
597        things like p++ in the argument of isxdigit().
598    
599    
600    Version 8.13 16-Aug-2011
601    ------------------------
602    
603    1.  The Unicode data tables have been updated to Unicode 6.0.0.
604    
605    2.  Two minor typos in pcre_internal.h have been fixed.
606    
607    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
608        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
609        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
610    
611    4.  There were a number of related bugs in the code for matching backrefences
612        caselessly in UTF-8 mode when codes for the characters concerned were
613        different numbers of bytes. For example, U+023A and U+2C65 are an upper
614        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
615        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
616        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
617        2-byte code at the end of the subject (it thought there wasn't enough data
618        left).
619    
620    5.  Comprehensive information about what went wrong is now returned by
621        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
622        as the output vector has at least 2 elements. The offset of the start of
623        the failing character and a reason code are placed in the vector.
624    
625    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
626        now returned is for the first byte of the failing character, instead of the
627        last byte inspected. This is an incompatible change, but I hope it is small
628        enough not to be a problem. It makes the returned offset consistent with
629        pcre_exec() and pcre_dfa_exec().
630    
631    7.  pcretest now gives a text phrase as well as the error number when
632        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
633        failure, the offset and reason code are output.
634    
635    8.  When \R was used with a maximizing quantifier it failed to skip backwards
636        over a \r\n pair if the subsequent match failed. Instead, it just skipped
637        back over a single character (\n). This seems wrong (because it treated the
638        two characters as a single entity when going forwards), conflicts with the
639        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
640        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
641        has been changed.
642    
643    9.  Some internal refactoring has changed the processing so that the handling
644        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
645        time (the PCRE_DOTALL option was changed this way some time ago: version
646        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
647        which was always a bit of a fudge. It also means that there is one less
648        argument for the match() function, which reduces its stack requirements
649        slightly. This change also fixes an incompatibility with Perl: the pattern
650        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
651    
652    10. More internal refactoring has drastically reduced the number of recursive
653        calls to match() for possessively repeated groups such as (abc)++ when
654        using pcre_exec().
655    
656    11. While implementing 10, a number of bugs in the handling of groups were
657        discovered and fixed:
658    
659        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
660        (a|)*(?1) gave a compile-time internal error.
661        ((a|)+)+  did not notice that the outer group could match an empty string.
662        (^a|^)+   was not marked as anchored.
663        (.*a|.*)+ was not marked as matching at start or after a newline.
664    
665    12. Yet more internal refactoring has removed another argument from the match()
666        function. Special calls to this function are now indicated by setting a
667        value in a variable in the "match data" data block.
668    
669    13. Be more explicit in pcre_study() instead of relying on "default" for
670        opcodes that mean there is no starting character; this means that when new
671        ones are added and accidentally left out of pcre_study(), testing should
672        pick them up.
673    
674    14. The -s option of pcretest has been documented for ages as being an old
675        synonym of -m (show memory usage). I have changed it to mean "force study
676        for every regex", that is, assume /S for every regex. This is similar to -i
677        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
678        using it. It makes it easier to run collections of tests with and without
679        study enabled, and thereby test pcre_study() more easily. All the standard
680        tests are now run with and without -s (but some patterns can be marked as
681        "never study" - see 20 below).
682    
683    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
684        restoration of the capturing data to the outer values was not happening
685        correctly.
686    
687    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
688        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
689        pattern had matched an empty string, and so incorrectly returned a no
690        match.
691    
692    17. There was optimizing code for the last branch of non-capturing parentheses,
693        and also for the obeyed branch of a conditional subexpression, which used
694        tail recursion to cut down on stack usage. Unfortunately, now that there is
695        the possibility of (*THEN) occurring in these branches, tail recursion is
696        no longer possible because the return has to be checked for (*THEN). These
697        two optimizations have therefore been removed. [But see 8.20/11 above.]
698    
699    18. If a pattern containing \R was studied, it was assumed that \R always
700        matched two bytes, thus causing the minimum subject length to be
701        incorrectly computed because \R can also match just one byte.
702    
703    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
704        was incorrectly computed.
705    
706    20. If /S is present twice on a test pattern in pcretest input, it now
707        *disables* studying, thereby overriding the use of -s on the command line
708        (see 14 above). This is necessary for one or two tests to keep the output
709        identical in both cases.
710    
711    21. When (*ACCEPT) was used in an assertion that matched an empty string and
712        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
713    
714    22. When an atomic group that contained a capturing parenthesis was
715        successfully matched, but the branch in which it appeared failed, the
716        capturing was not being forgotten if a higher numbered group was later
717        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
718        group 1 to "a", when in fact it should be unset. This applied to multi-
719        branched capturing and non-capturing groups, repeated or not, and also to
720        positive assertions (capturing in negative assertions does not happen
721        in PCRE) and also to nested atomic groups.
722    
723    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
724        subject after a captured substring, to make it easier to tell which of a
725        number of identical substrings has been captured.
726    
727    24. The way atomic groups are processed by pcre_exec() has been changed so that
728        if they are repeated, backtracking one repetition now resets captured
729        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
730        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
731        "aaa". Previously, it would have been "a". As part of this code
732        refactoring, the way recursive calls are handled has also been changed.
733    
734    25. If an assertion condition captured any substrings, they were not passed
735        back unless some other capturing happened later. For example, if
736        (?(?=(a))a) was matched against "a", no capturing was returned.
737    
738    26. When studying a pattern that contained subroutine calls or assertions,
739        the code for finding the minimum length of a possible match was handling
740        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
741        group 1 called group 2 while simultaneously a separate group 2 called group
742        1). A stack overflow occurred in this case. I have fixed this by limiting
743        the recursion depth to 10.
744    
745    27. Updated RunTest.bat in the distribution to the version supplied by Tom
746        Fortmann. This supports explicit test numbers on the command line, and has
747        argument validation and error reporting.
748    
749    28. An instance of \X with an unlimited repeat could fail if at any point the
750        first character it looked at was a mark character.
751    
752    29. Some minor code refactoring concerning Unicode properties and scripts
753        should reduce the stack requirement of match() slightly.
754    
755    30. Added the '=' option to pcretest to check the setting of unused capturing
756        slots at the end of the pattern, which are documented as being -1, but are
757        not included in the return count.
758    
759    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
760        compiled something random. Now it gives a compile-time error (as does
761        Perl).
762    
763    32. A *MARK encountered during the processing of a positive assertion is now
764        recorded and passed back (compatible with Perl).
765    
766    33. If --only-matching or --colour was set on a pcregrep call whose pattern
767        had alternative anchored branches, the search for a second match in a line
768        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
769        matched the line "0102" twice. The same bug affected patterns that started
770        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
771        twice.
772    
773    34. Previously, PCRE did not allow quantification of assertions. However, Perl
774        does, and because of capturing effects, quantifying parenthesized
775        assertions may at times be useful. Quantifiers are now allowed for
776        parenthesized assertions.
777    
778    35. A minor code tidy in pcre_compile() when checking options for \R usage.
779    
780    36. \g was being checked for fancy things in a character class, when it should
781        just be a literal "g".
782    
783    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
784        appearance of a nested POSIX class supersedes an apparent external class.
785        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
786        unescaped square brackets may also appear as part of class names. For
787        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
788        more like Perl. (But see 8.20/1 above.)
789    
790    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
791        was because it thought it was \N{name}, which is not supported).
792    
793    39. Add minix to OS list not supporting the -S option in pcretest.
794    
795    40. PCRE tries to detect cases of infinite recursion at compile time, but it
796        cannot analyze patterns in sufficient detail to catch mutual recursions
797        such as ((?1))((?2)). There is now a runtime test that gives an error if a
798        subgroup is called recursively as a subpattern for a second time at the
799        same position in the subject string. In previous releases this might have
800        been caught by the recursion limit, or it might have run out of stack.
801    
802    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
803        happen only once. PCRE was, however incorrectly giving a compile time error
804        "recursive call could loop indefinitely" because it cannot analyze the
805        pattern in sufficient detail. The compile time test no longer happens when
806        PCRE is compiling a conditional subpattern, but actual runaway loops are
807        now caught at runtime (see 40 above).
808    
809    42. It seems that Perl allows any characters other than a closing parenthesis
810        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
811        has been changed to be the same.
812    
813    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
814        as not to get warnings when autogen.sh is called. Also changed
815        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
816    
817    44. To help people who use pcregrep to scan files containing exceedingly long
818        lines, the following changes have been made:
819    
820        (a) The default value of the buffer size parameter has been increased from
821            8K to 20K. (The actual buffer used is three times this size.)
822    
823        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
824            PCRE is built.
825    
826        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
827            to be set at run time.
828    
829        (d) Numerical values in pcregrep options can be followed by K or M, for
830            example --buffer-size=50K.
831    
832        (e) If a line being scanned overflows pcregrep's buffer, an error is now
833            given and the return code is set to 2.
834    
835    45. Add a pointer to the latest mark to the callout data block.
836    
837    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
838        partial match of an empty string instead of no match. This was specific to
839        the use of ".".
840    
841    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
842        complete match instead of a partial match. This bug was dependent on both
843        the PCRE_UTF8 and PCRE_DOTALL options being set.
844    
845    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
846        starting byte set, because \b was not being ignored.
847    
848    
849    Version 8.12 15-Jan-2011
850    ------------------------
851    
852    1.  Fixed some typos in the markup of the man pages, and wrote a script that
853        checks for such things as part of the documentation building process.
854    
855    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
856        --match-limit and --recursion-limit options (added for 8.11). In
857        particular, this made one of the standard tests fail. (The integer value
858        went into the wrong half of a long int.)
859    
860    3.  If the --colour option was given to pcregrep with -v (invert match), it
861        did strange things, either producing crazy output, or crashing. It should,
862        of course, ignore a request for colour when reporting lines that do not
863        match.
864    
865    4.  Another pcregrep bug caused similar problems if --colour was specified with
866        -M (multiline) and the pattern match finished with a line ending.
867    
868    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
869        matched in multiline mode, the following line was shown as part of the
870        match. This seems wrong, so I have changed it.
871    
872    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
873        the check for further matches in the same line (so they could be coloured)
874        to overrun the end of the current line. If another match was found, it was
875        incorrectly shown (and then shown again when found in the next line).
876    
877    7.  If pcregrep was compiled under Windows, there was a reference to the
878        function pcregrep_exit() before it was defined. I am assuming this was
879        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
880        reported by a user. I've moved the definition above the reference.
881    
882    
883    Version 8.11 10-Dec-2010
884    ------------------------
885    
886    1.  (*THEN) was not working properly if there were untried alternatives prior
887        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
888        backtracked to try for "b" instead of moving to the next alternative branch
889        at the same level (in this case, to look for "c"). The Perl documentation
890        is clear that when (*THEN) is backtracked onto, it goes to the "next
891        alternative in the innermost enclosing group".
892    
893    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
894        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
895        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
896        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
897        (*THEN).
898    
899    3.  If \s appeared in a character class, it removed the VT character from
900        the class, even if it had been included by some previous item, for example
901        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
902        of \s, but is part of the POSIX "space" class.)
903    
904    4.  A partial match never returns an empty string (because you can always
905        match an empty string at the end of the subject); however the checking for
906        an empty string was starting at the "start of match" point. This has been
907        changed to the "earliest inspected character" point, because the returned
908        data for a partial match starts at this character. This means that, for
909        example, /(?<=abc)def/ gives a partial match for the subject "abc"
910        (previously it gave "no match").
911    
912    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
913        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
914        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
915        has an implication that the given string is incomplete (because a partial
916        match is preferred over a full match). For this reason, these items now
917        give a partial match in this situation. [Aside: previously, the one case
918        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
919        match rather than a full match, which was wrong by the old rules, but is
920        now correct.]
921    
922    6.  There was a bug in the handling of #-introduced comments, recognized when
923        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
924        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
925        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
926        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
927        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
928        places in pcre_compile().
929    
930    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
931        comments when looking ahead for named forward references to subpatterns,
932        the only newline sequence it recognized was NL. It now handles newlines
933        according to the set newline convention.
934    
935    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
936        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
937        cater for a lack of strerror(). These oversights have been fixed.
938    
939    9.  Added --match-limit and --recursion-limit to pcregrep.
940    
941    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
942    
943    11. When the -o option was used, pcregrep was setting a return code of 1, even
944        when matches were found, and --line-buffered was not being honoured.
945    
946    12. Added an optional parentheses number to the -o and --only-matching options
947        of pcregrep.
948    
949    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
950        can match an empty string. The code to do it in pcretest and pcredemo
951        needed fixing:
952    
953        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
954            only one byte after an empty string match just before CRLF (this case
955            just got forgotten; "any" and "anycrlf" were OK).
956    
957        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
958            mode when an empty string match preceded an ASCII character followed by
959            a non-ASCII character. (The code for advancing by one character rather
960            than one byte was nonsense.)
961    
962        (c) The pcredemo.c sample program did not have any code at all to handle
963            the cases when CRLF is a valid newline sequence.
964    
965    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
966        as a starting offset was within the subject string. There is now a new
967        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
968        negative or greater than the length of the string. In order to test this,
969        pcretest is extended to allow the setting of negative starting offsets.
970    
971    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
972        starting offset points to the beginning of a UTF-8 character was
973        unnecessarily clumsy. I tidied it up.
974    
975    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
976        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
977    
978    17. Nobody had reported that the --include_dir option, which was added in
979        release 7.7 should have been called --include-dir (hyphen, not underscore)
980        for compatibility with GNU grep. I have changed it to --include-dir, but
981        left --include_dir as an undocumented synonym, and the same for
982        --exclude-dir, though that is not available in GNU grep, at least as of
983        release 2.5.4.
984    
985    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
986        characters from a string of bytes) have been redefined so as not to use
987        loops, in order to improve performance in some environments. At the same
988        time, I abstracted some of the common code into auxiliary macros to save
989        repetition (this should not affect the compiled code).
990    
991    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
992        compile-time error is now given if \c is not followed by an ASCII
993        character, that is, a byte less than 128. (In EBCDIC mode, the code is
994        different, and any byte value is allowed.)
995    
996    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
997        START_OPTIMIZE option, which is now allowed at compile time - but just
998        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
999        to pcregrep and other applications that have no direct access to PCRE
1000        options. The new /Y option in pcretest sets this option when calling
1001        pcre_compile().
1002    
1003    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1004        back references. Groups containing recursive back references were forced to
1005        be atomic by that change, but in the case of named groups, the amount of
1006        memory required was incorrectly computed, leading to "Failed: internal
1007        error: code overflow". This has been fixed.
1008    
1009    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1010        pcretest.c, to avoid build problems in some Borland environments.
1011    
1012    
1013    Version 8.10 25-Jun-2010
1014    ------------------------
1015    
1016    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1017        THEN.
1018    
1019    2.  (*ACCEPT) was not working when inside an atomic group.
1020    
1021    3.  Inside a character class, \B is treated as a literal by default, but
1022        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1023        causes the error). The code is unchanged, but I tidied the documentation.
1024    
1025    4.  Inside a character class, PCRE always treated \R and \X as literals,
1026        whereas Perl faults them if its -w option is set. I have changed PCRE so
1027        that it faults them when PCRE_EXTRA is set.
1028    
1029    5.  Added support for \N, which always matches any character other than
1030        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1031    
1032    6.  When compiling pcregrep with newer versions of gcc which may have
1033        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1034        declared with attribute warn_unused_result" were given. Just casting the
1035        result to (void) does not stop the warnings; a more elaborate fudge is
1036        needed. I've used a macro to implement this.
1037    
1038    7.  Minor change to pcretest.c to avoid a compiler warning.
1039    
1040    8.  Added four artifical Unicode properties to help with an option to make
1041        \s etc use properties (see next item). The new properties are: Xan
1042        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1043    
1044    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1045        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1046        this option. Modified pcretest to add /W to test this facility. Added
1047        REG_UCP to make it available via the POSIX interface.
1048    
1049    10. Added --line-buffered to pcregrep.
1050    
1051    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1052        studied, and the match started with a letter with a code point greater than
1053        127 whose first byte was different to the first byte of the other case of
1054        the letter, the other case of this starting letter was not recognized
1055        (#976).
1056    
1057    12. If a pattern that was studied started with a repeated Unicode property
1058        test, for example, \p{Nd}+, there was the theoretical possibility of
1059        setting up an incorrect bitmap of starting bytes, but fortunately it could
1060        not have actually happened in practice until change 8 above was made (it
1061        added property types that matched character-matching opcodes).
1062    
1063    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1064        possible starting bytes for non-anchored patterns.
1065    
1066    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1067        \R, and also a number of cases that involve Unicode properties, both
1068        explicit and implicit when PCRE_UCP is set.
1069    
1070    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1071        input, it could crash or give wrong results if characters with values
1072        greater than 0xc0 were present in the subject string. (Detail: it assumed
1073        UTF-8 input when processing these items.)
1074    
1075    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1076        size_t is 64-bit (#991).
1077    
1078    17. Added a check for running out of memory when PCRE is compiled with
1079        --disable-stack-for-recursion (#990).
1080    
1081    18. If the last data line in a file for pcretest does not have a newline on
1082        the end, a newline was missing in the output.
1083    
1084    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1085        less than 128) in its various bitmaps. However, there is a facility for
1086        generating tables according to the current locale when PCRE is compiled. It
1087        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1088        characters, are recognized by isspace() and therefore were getting set in
1089        these tables, and indeed these tables seem to approximate to ISO 8859. This
1090        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1091        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1092        which of course cannot start UTF-8 characters. I have changed the code so
1093        that only real ASCII characters (less than 128) and the correct starting
1094        bytes for UTF-8 encodings are set for characters greater than 127 when in
1095        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1096        altogether.)
1097    
1098    20. Added the /T option to pcretest so as to be able to run tests with non-
1099        standard character tables, thus making it possible to include the tests
1100        used for 19 above in the standard set of tests.
1101    
1102    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1103        reference to a subpattern the other side of a comment that contains an
1104        opening parenthesis caused either an internal compiling error, or a
1105        reference to the wrong subpattern.
1106    
1107    
1108    Version 8.02 19-Mar-2010
1109    ------------------------
1110    
1111    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1112    
1113    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1114        configured.
1115    
1116    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1117        original author of that file, following a query about its status.
1118    
1119    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1120        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1121    
1122    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1123        quantifier applied to a forward-referencing subroutine call, could compile
1124        incorrect code or give the error "internal error: previously-checked
1125        referenced subpattern not found".
1126    
1127    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1128        variables to point to external functions. For these systems, therefore,
1129        pcre_malloc etc. are now initialized to local functions that call the
1130        relevant global functions.
1131    
1132    7.  There were two entries missing in the vectors called coptable and poptable
1133        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1134        I've fixed the data, and added a kludgy way of testing at compile time that
1135        the lengths are correct (equal to the number of opcodes).
1136    
1137    8.  Following on from 7, I added a similar kludge to check the length of the
1138        eint vector in pcreposix.c.
1139    
1140    9.  Error texts for pcre_compile() are held as one long string to avoid too
1141        much relocation at load time. To find a text, the string is searched,
1142        counting zeros. There was no check for running off the end of the string,
1143        which could happen if a new error number was added without updating the
1144        string.
1145    
1146    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1147    
1148    11. \K was not working if it appeared in an atomic group or in a group that
1149        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1150        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1151        the assertion is positive, but not if it is negative.
1152    
1153    12. Change 11 fortuitously reduced the size of the stack frame used in the
1154        "match()" function of pcre_exec.c by one pointer. Forthcoming
1155        implementation of support for (*MARK) will need an extra pointer on the
1156        stack; I have reserved it now, so that the stack frame size does not
1157        decrease.
1158    
1159    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1160        item in branch that calls a recursion is a subroutine call - as in the
1161        second branch in the above example - was incorrectly given the compile-
1162        time error "recursive call could loop indefinitely" because pcre_compile()
1163        was not correctly checking the subroutine for matching a non-empty string.
1164    
1165    14. The checks for overrunning compiling workspace could trigger after an
1166        overrun had occurred. This is a "should never occur" error, but it can be
1167        triggered by pathological patterns such as hundreds of nested parentheses.
1168        The checks now trigger 100 bytes before the end of the workspace.
1169    
1170    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1171    
1172    
1173    Version 8.01 19-Jan-2010
1174    ------------------------
1175    
1176    1.  If a pattern contained a conditional subpattern with only one branch (in
1177        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1178        computed the wrong minimum data length (which is of course zero for such
1179        subpatterns). This could cause incorrect "no match" results.
1180    
1181    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1182        the pattern is reset in the first branch, pcre_compile() failed with
1183        "internal error: code overflow at offset...". This happened only when
1184        the reset was to the original external option setting. (An optimization
1185        abstracts leading options settings into an external setting, which was the
1186        cause of this.)
1187    
1188    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1189        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1190        assertion pattern did not match (meaning that the assertion was true), it
1191        was incorrectly treated as false if the SKIP had been reached during the
1192        matching. This also applied to assertions used as conditions.
1193    
1194    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1195        assertion subpattern, including such a pattern used as a condition,
1196        unpredictable results occurred, instead of the error return
1197        PCRE_ERROR_DFA_UITEM.
1198    
1199    5.  The C++ GlobalReplace function was not working like Perl for the special
1200        situation when an empty string is matched. It now does the fancy magic
1201        stuff that is necessary.
1202    
1203    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1204        removed. (These were left over from very, very early versions of PCRE.)
1205    
1206    7.  Some cosmetic changes to the code to make life easier when compiling it
1207        as part of something else:
1208    
1209        (a) Change DEBUG to PCRE_DEBUG.
1210    
1211        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1212            called "current" as "current_branch", to prevent a collision with the
1213            Linux macro when compiled as a kernel module.
1214    
1215        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1216            prevent a collision with the Linux macro when compiled as a kernel
1217            module.
1218    
1219    8.  In pcre_compile() there are some checks for integer overflows that used to
1220        cast potentially large values to (double). This has been changed to that
1221        when building, a check for int64_t is made, and if it is found, it is used
1222        instead, thus avoiding the use of floating point arithmetic. (There is no
1223        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1224        double.
1225    
1226    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1227        2005 (difference between two addresses compared to an unsigned value).
1228    
1229    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1230        custom one, because of the following reported problem in Windows:
1231    
1232          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1233              under Win32.
1234          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1235              therefore missing the function definition.
1236          - The compiler thus generates a "C" signature for the test function.
1237          - The linker fails to find the "C" function.
1238          - PCRE fails to configure if asked to do so against libbz2.
1239    
1240    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1241        messages were output:
1242    
1243          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1244          rerunning libtoolize, to keep the correct libtool macros in-tree.
1245          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1246    
1247        I have done both of these things.
1248    
1249    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1250        most of the time, it *can* run out if it is given a pattern that contains a
1251        runaway infinite recursion. I updated the discussion in the pcrestack man
1252        page.
1253    
1254    13. Now that we have gone to the x.xx style of version numbers, the minor
1255        version may start with zero. Using 08 or 09 is a bad idea because users
1256        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1257        interpreted as invalid octal numbers. I've updated the previous comment in
1258        configure.ac, and also added a check that gives an error if 08 or 09 are
1259        used.
1260    
1261    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1262        causing partial matching to fail when the end of the subject matched \W
1263        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1264    
1265    15. There were some discrepancies between the declarations in pcre_internal.h
1266        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1267        their definitions. The declarations used "const uschar *" and the
1268        definitions used USPTR. Even though USPTR is normally defined as "const
1269        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1270        reported that: "This difference in casting confuses some C++ compilers, for
1271        example, SunCC recognizes above declarations as different functions and
1272        generates broken code for hbpcre." I have changed the declarations to use
1273        USPTR.
1274    
1275    16. GNU libtool is named differently on some systems. The autogen.sh script now
1276        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1277        (FreeBSD).
1278    
1279    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1280        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1281        comment: "Figure out how to create a longlong from a string: strtoll and
1282        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1283        instance, but it only takes 2 args instead of 3!"
1284    
1285    18. A subtle bug concerned with back references has been fixed by a change of
1286        specification, with a corresponding code fix. A pattern such as
1287        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1288        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1289        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1290        same bug. Such groups have to be quantified to be useful, or contained
1291        inside another quantified group. (If there's no repetition, the reference
1292        can never match.) The problem arises because, having left the group and
1293        moved on to the rest of the pattern, a later failure that backtracks into
1294        the group uses the captured value from the final iteration of the group
1295        rather than the correct earlier one. I have fixed this in PCRE by forcing
1296        any group that contains a reference to itself to be an atomic group; that
1297        is, there cannot be any backtracking into it once it has completed. This is
1298        similar to recursive and subroutine calls.
1299    
1300    
1301    Version 8.00 19-Oct-09
1302    ----------------------
1303    
1304    1.  The table for translating pcre_compile() error codes into POSIX error codes
1305        was out-of-date, and there was no check on the pcre_compile() error code
1306        being within the table. This could lead to an OK return being given in
1307        error.
1308    
1309    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1310        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1311        in a Windows environment.
1312    
1313    3.  The pcregrep --count option prints the count for each file even when it is
1314        zero, as does GNU grep. However, pcregrep was also printing all files when
1315        --files-with-matches was added. Now, when both options are given, it prints
1316        counts only for those files that have at least one match. (GNU grep just
1317        prints the file name in this circumstance, but including the count seems
1318        more useful - otherwise, why use --count?) Also ensured that the
1319        combination -clh just lists non-zero counts, with no names.
1320    
1321    4.  The long form of the pcregrep -F option was incorrectly implemented as
1322        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1323        but it seems right to fix it, and I didn't think it was worth preserving
1324        the old behaviour.
1325    
1326    5.  The command line items --regex=pattern and --regexp=pattern were not
1327        recognized by pcregrep, which required --regex pattern or --regexp pattern
1328        (with a space rather than an '='). The man page documented the '=' forms,
1329        which are compatible with GNU grep; these now work.
1330    
1331    6.  No libpcreposix.pc file was created for pkg-config; there was just
1332        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1333    
1334    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1335        when UCP support is not needed, by modifying the Python script that
1336        generates it from Unicode data files. This should not matter if the module
1337        is correctly used as a library, but I received one complaint about 50K of
1338        unwanted data. My guess is that the person linked everything into his
1339        program rather than using a library. Anyway, it does no harm.
1340    
1341    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1342        was a minimum greater than 1 for a wide character in a possessive
1343        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1344        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1345        character. Chaos in the form of incorrect output or a compiling loop could
1346        result.
1347    
1348    9.  The restrictions on what a pattern can contain when partial matching is
1349        requested for pcre_exec() have been removed. All patterns can now be
1350        partially matched by this function. In addition, if there are at least two
1351        slots in the offset vector, the offset of the earliest inspected character
1352        for the match and the offset of the end of the subject are set in them when
1353        PCRE_ERROR_PARTIAL is returned.
1354    
1355    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1356        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1357        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1358        and may be more useful for multi-segment matching.
1359    
1360    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1361        used to be given if ever the end of the subject was reached; now it is
1362        given only if matching could not proceed because another character was
1363        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1364        string "Z", which now yields "no match" instead of "partial match". In the
1365        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1366        final character ended with (*FAIL).
1367    
1368    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1369        if the pattern had a "must contain" character that was already found in the
1370        earlier partial match, unless partial matching was again requested. For
1371        example, with the pattern /dog.(body)?/, the "must contain" character is
1372        "g". If the first part-match was for the string "dog", restarting with
1373        "sbody" failed. This bug has been fixed.
1374    
1375    13. The string returned by pcre_dfa_exec() after a partial match has been
1376        changed so that it starts at the first inspected character rather than the
1377        first character of the match. This makes a difference only if the pattern
1378        starts with a lookbehind assertion or \b or \B (\K is not supported by
1379        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1380        matching functions compatible, and I think it's the right thing to do.
1381    
1382    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1383        so that the demonstration program is easily available in environments where
1384        PCRE has not been installed from source.
1385    
1386    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1387        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1388        library.
1389    
1390    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1391        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1392        is not the first non-POSIX option to be added. Clearly some people find
1393        these options useful.
1394    
1395    17. If a caller to the POSIX matching function regexec() passes a non-zero
1396        value for nmatch with a NULL value for pmatch, the value of
1397        nmatch is forced to zero.
1398    
1399    18. RunGrepTest did not have a test for the availability of the -u option of
1400        the diff command, as RunTest does. It now checks in the same way as
1401        RunTest, and also checks for the -b option.
1402    
1403    19. If an odd number of negated classes containing just a single character
1404        interposed, within parentheses, between a forward reference to a named
1405        subpattern and the definition of the subpattern, compilation crashed with
1406        an internal error, complaining that it could not find the referenced
1407        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1408        [The bug was that it was starting one character too far in when skipping
1409        over the character class, thus treating the ] as data rather than
1410        terminating the class. This meant it could skip too much.]
1411    
1412    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1413        /g option in pcretest when the pattern contains \K, which makes it possible
1414        to have an empty string match not at the start, even when the pattern is
1415        anchored. Updated pcretest and pcredemo to use this option.
1416    
1417    21. If the maximum number of capturing subpatterns in a recursion was greater
1418        than the maximum at the outer level, the higher number was returned, but
1419        with unset values at the outer level. The correct (outer level) value is
1420        now given.
1421    
1422    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1423        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1424        make it do so. The string so far is captured, making this feature
1425        compatible with Perl.
1426    
1427    23. The tests have been re-organized, adding tests 11 and 12, to make it
1428        possible to check the Perl 5.10 features against Perl 5.10.
1429    
1430    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1431        pattern matches a fixed length string. PCRE did not allow this; now it
1432        does. Neither allows recursion.
1433    
1434    25. I finally figured out how to implement a request to provide the minimum
1435        length of subject string that was needed in order to match a given pattern.
1436        (It was back references and recursion that I had previously got hung up
1437        on.) This code has now been added to pcre_study(); it finds a lower bound
1438        to the length of subject needed. It is not necessarily the greatest lower
1439        bound, but using it to avoid searching strings that are too short does give
1440        some useful speed-ups. The value is available to calling programs via
1441        pcre_fullinfo().
1442    
1443    26. While implementing 25, I discovered to my embarrassment that pcretest had
1444        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1445        study optimizations had never been tested with that matching function.
1446        Oops. What is worse, even when it was passed study data, there was a bug in
1447        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1448        were also very few tests of studied patterns with pcre_dfa_exec().
1449    
1450    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1451        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1452        on the other side of the coin, they are no longer allowed to have different
1453        names, because these cannot be distinguished in PCRE, and this has caused
1454        confusion. (This is a difference from Perl.)
1455    
1456    28. When duplicate subpattern names are present (necessarily with different
1457        numbers, as required by 27 above), and a test is made by name in a
1458        conditional pattern, either for a subpattern having been matched, or for
1459        recursion in such a pattern, all the associated numbered subpatterns are
1460        tested, and the overall condition is true if the condition is true for any
1461        one of them. This is the way Perl works, and is also more like the way
1462        testing by number works.
1463    
1464    
1465    Version 7.9 11-Apr-09
1466    ---------------------
1467    
1468    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1469        (pcretest), all targets were linked against these libraries. This included
1470        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1471        libraries. This caused unwanted dependencies to be created. This problem
1472        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1473        pcretest is linked with readline.
1474    
1475    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1476        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1477        moved outside it again, because FALSE and TRUE are already defined in AIX,
1478        but BOOL is not.
1479    
1480    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1481        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1482    
1483    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1484        hyphens) following file names and line numbers when outputting matching
1485        lines. This is not true; no spaces are inserted. I have also clarified the
1486        wording for the --colour (or --color) option.
1487    
1488    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1489        was not coloured; this is different to GNU grep, so I have changed it to be
1490        the same.
1491    
1492    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1493        each matching line was coloured. Now it goes on to look for further matches
1494        of any of the test patterns, which is the same behaviour as GNU grep.
1495    
1496    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1497        doesn't make sense to accept an empty string match in pcregrep, so I have
1498        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1499        seems to be how GNU grep behaves.
1500    
1501    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1502        start or after a newline", because the conditional assertion was not being
1503        correctly handled. The rule now is that both the assertion and what follows
1504        in the first alternative must satisfy the test.
1505    
1506    9.  If auto-callout was enabled in a pattern with a conditional group whose
1507        condition was an assertion, PCRE could crash during matching, both with
1508        pcre_exec() and pcre_dfa_exec().
1509    
1510    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1511        used for matching.
1512    
1513    11. Unicode property support in character classes was not working for
1514        characters (bytes) greater than 127 when not in UTF-8 mode.
1515    
1516    12. Added the -M command line option to pcretest.
1517    
1518    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1519    
1520    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1521    
1522    16. Added comments and documentation about mis-use of no_arg in the C++
1523        wrapper.
1524    
1525    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1526        from Martin Jerabek that uses macro names for all relevant character and
1527        string constants.
1528    
1529    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1530        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1531        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1532        these, but not everybody uses configure.
1533    
1534    19. A conditional group that had only one branch was not being correctly
1535        recognized as an item that could match an empty string. This meant that an
1536        enclosing group might also not be so recognized, causing infinite looping
1537        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1538        with the subject "ab", where knowledge that the repeated group can match
1539        nothing is needed in order to break the loop.
1540    
1541    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1542        exec(), but without supplying a callout function, matching went wrong.
1543    
1544    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1545        leak if the size of the offset vector was greater than 30. When the vector
1546        is smaller, the saved offsets during recursion go onto a local stack
1547        vector, but for larger vectors malloc() is used. It was failing to free
1548        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1549        error, in fact).
1550    
1551    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1552        heapframe that is used only when UTF-8 support is enabled. This caused no
1553        problem, but was untidy.
1554    
1555    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1556        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1557        included within another project.
1558    
1559    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1560        slightly modified by me:
1561    
1562          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1563              not building pcregrep.
1564    
1565          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1566              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1567    
1568    25. Forward references, both numeric and by name, in patterns that made use of
1569        duplicate group numbers, could behave incorrectly or give incorrect errors,
1570        because when scanning forward to find the reference group, PCRE was not
1571        taking into account the duplicate group numbers. A pattern such as
1572        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1573    
1574    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1575        the feature of a custom pointer more persuasive (as requested by a user).
1576    
1577    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1578        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1579        pre-defined.
1580    
1581    28. Added support for (*UTF8) at the start of a pattern.
1582    
1583    29. Arrange for flags added by the "release type" setting in CMake to be shown
1584        in the configuration summary.
1585    
1586    
1587    Version 7.8 05-Sep-08
1588    ---------------------
1589    
1590    1.  Replaced UCP searching code with optimized version as implemented for Ad
1591        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1592        stage table and inline lookup instead of a function, giving speed ups of 2
1593        to 5 times on some simple patterns that I tested. Permission was given to
1594        distribute the MultiStage2.py script that generates the tables (it's not in
1595        the tarball, but is in the Subversion repository).
1596    
1597    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1598        scripts.
1599    
1600    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1601        a group with a zero qualifier. The result of the study could be incorrect,
1602        or the function might crash, depending on the pattern.
1603    
1604    4.  Caseless matching was not working for non-ASCII characters in back
1605        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1606        It now works when Unicode Property Support is available.
1607    
1608    5.  In pcretest, an escape such as \x{de} in the data was always generating
1609        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1610        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1611        truncation.
1612    
1613    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1614    
1615    7.  Added two (int) casts to pcregrep when printing the difference of two
1616        pointers, in case they are 64-bit values.
1617    
1618    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1619        test 2 if it fails.
1620    
1621    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1622        and a #define of that name to empty if it is not externally set. This is to
1623        allow users of MSVC to set it if necessary.
1624    
1625    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1626        the convenience functions in the pcre_get.c source file.
1627    
1628    11. An option change at the start of a pattern that had top-level alternatives
1629        could cause overwriting and/or a crash. This command provoked a crash in
1630        some environments:
1631    
1632          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1633    
1634        This potential security problem was recorded as CVE-2008-2371.
1635    
1636    12. For a pattern where the match had to start at the beginning or immediately
1637        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1638        pcre_dfa_exec() could read past the end of the passed subject if there was
1639        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1640        pcretest so that it places the subject at the end of its malloc-ed buffer.
1641    
1642    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1643        exec() might read past the end of the data buffer in UTF-8 mode.
1644    
1645    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1646        the data contained the byte 0x85 as part of a UTF-8 character within its
1647        first line. This applied both to normal and DFA matching.
1648    
1649    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1650        /^[^d]*?$/8 failed to match "abc".
1651    
1652    16. Added a missing copyright notice to pcrecpp_internal.h.
1653    
1654    17. Make it more clear in the documentation that values returned from
1655        pcre_exec() in ovector are byte offsets, not character counts.
1656    
1657    18. Tidied a few places to stop certain compilers from issuing warnings.
1658    
1659    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1660        supplied by Stefan Weber. I made a further small update for 7.8 because
1661        there is a change of source arrangements: the pcre_searchfuncs.c module is
1662        replaced by pcre_ucd.c.
1663    
1664    
1665    Version 7.7 07-May-08
1666    ---------------------
1667    
1668    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1669        a string to a long long, pretend we don't even have a long long." This is
1670        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1671    
1672    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1673        pre-7.6 versions, which defined a global no_arg variable instead of putting
1674        it in the RE class. (See also #8 below.)
1675    
1676    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1677        Lopes.
1678    
1679    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1680    
1681        (1) The include/exclude patterns were being applied to the whole pathnames
1682            of files, instead of just to the final components.
1683    
1684        (2) If there was more than one level of directory, the subdirectories were
1685            skipped unless they satisfied the include/exclude conditions. This is
1686            inconsistent with GNU grep (and could even be seen as contrary to the
1687            pcregrep specification - which I improved to make it absolutely clear).
1688            The action now is always to scan all levels of directory, and just
1689            apply the include/exclude patterns to regular files.
1690    
1691    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1692        --exclude_dir in the tests to avoid scanning .svn directories.
1693    
1694    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1695        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1696        doesn't support NULs in patterns.
1697    
1698    7.  Added some missing "const"s to declarations of static tables in
1699        pcre_compile.c and pcre_dfa_exec.c.
1700    
1701    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1702        caused by fix #2  above. (Subsequently also a second patch to fix the
1703        first patch. And a third patch - this was a messy problem.)
1704    
1705    9.  Applied Craig's patch to remove the use of push_back().
1706    
1707    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1708        matching function regexec().
1709    
1710    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1711        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1712        references. PCRE supports relative numbers with this syntax (I don't think
1713        Oniguruma does).
1714    
1715    12. Previously, a group with a zero repeat such as (...){0} was completely
1716        omitted from the compiled regex. However, this means that if the group
1717        was called as a subroutine from elsewhere in the pattern, things went wrong
1718        (an internal error was given). Such groups are now left in the compiled
1719        pattern, with a new opcode that causes them to be skipped at execution
1720        time.
1721    
1722    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1723        to the way PCRE behaves:
1724    
1725        (a) A lone ] character is dis-allowed (Perl treats it as data).
1726    
1727        (b) A back reference to an unmatched subpattern matches an empty string
1728            (Perl fails the current match path).
1729    
1730        (c) A data ] in a character class must be notated as \] because if the
1731            first data character in a class is ], it defines an empty class. (In
1732            Perl it is not possible to have an empty class.) The empty class []
1733            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1734            The negative empty class [^] matches any one character, independently
1735            of the DOTALL setting.
1736    
1737    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1738        non-existent subpattern following a character class starting with ']' and
1739        containing () gave an internal compiling error instead of "reference to
1740        non-existent subpattern". Fortunately, when the pattern did exist, the
1741        compiled code was correct. (When scanning forwards to check for the
1742        existencd of the subpattern, it was treating the data ']' as terminating
1743        the class, so got the count wrong. When actually compiling, the reference
1744        was subsequently set up correctly.)
1745    
1746    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1747        it was being rejected as not supported by pcre_dfa_exec(), even though
1748        other assertions are supported. I have made pcre_dfa_exec() support
1749        (*FAIL).
1750    
1751    16. The implementation of 13c above involved the invention of a new opcode,
1752        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1753        cannot be changed at match time, I realized I could make a small
1754        improvement to matching performance by compiling OP_ALLANY instead of
1755        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1756        on the OP_ANY path.
1757    
1758    17. Compiling pcretest on Windows with readline support failed without the
1759        following two fixes: (1) Make the unistd.h include conditional on
1760        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1761    
1762    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1763        ncurses library to be included for pcretest when ReadLine support is
1764        requested, but also to allow for it to be overridden. This patch came from
1765        Daniel Bergström.
1766    
1767    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1768        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1769        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1770        spotting this.
1771    
1772    
1773    Version 7.6 28-Jan-08
1774    ---------------------
1775    
1776    1.  A character class containing a very large number of characters with
1777        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1778        overflow.
1779    
1780    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1781        HAVE_LONG_LONG is not defined.
1782    
1783    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1784        bring them up to date and include new features. This patch includes:
1785    
1786        - Fixed PH's badly added libz and libbz2 support.
1787        - Fixed a problem with static linking.
1788        - Added pcredemo. [But later removed - see 7 below.]
1789        - Fixed dftables problem and added an option.
1790        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1791            HAVE_LONG_LONG.
1792        - Added readline support for pcretest.
1793        - Added an listing of the option settings after cmake has run.
1794    
1795    4.  A user submitted a patch to Makefile that makes it easy to create
1796        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1797        Makefile.am that cause it to include this special target, without
1798        affecting anything else. Note that the same mingw target plus all
1799        the other distribution libraries and programs are now supported
1800        when configuring with CMake (see 6 below) instead of with
1801        Configure/Make.
1802    
1803    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1804        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1805        exported in the Windows port". It has not yet been confirmed that the patch
1806        solves the problem, but it does no harm.
1807    
1808    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1809        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1810        with CMake, and also correct the comment about stack recursion.
1811    
1812    7.  Remove the automatic building of pcredemo from the ./configure system and
1813        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1814        of a program that users should build themselves after PCRE is installed, so
1815        building it automatically is not really right. What is more, it gave
1816        trouble in some build environments.
1817    
1818    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1819    
1820    
1821    Version 7.5 10-Jan-08
1822    ---------------------
1823    
1824    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1825        values in parens when parsing an RE using the C++ wrapper."
1826    
1827    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1828        Characters greater than 255 were excluded from the class instead of being
1829        included.
1830    
1831    3.  The same bug as (2) above applied to negated POSIX classes such as
1832        [:^space:].
1833    
1834    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1835        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1836        I have changed it.
1837    
1838    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1839        first named subpattern) and a construct such as (?&a) would reference the
1840        first named subpattern whose name started with "a" (in other words, the
1841        length check was missing). Both these problems are fixed. "Subpattern name
1842        expected" is now given for (?&) (a zero-length name), and this patch also
1843        makes it give the same error for \k'' (previously it complained that that
1844        was a reference to a non-existent subpattern).
1845    
1846    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1847        this is right because (?- can be followed by option settings as well as by
1848        digits. I have, however, made the messages clearer.
1849    
1850    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1851        than the number used in the conditional) now cause a compile-time error.
1852        This is actually not compatible with Perl, which accepts such patterns, but
1853        treats the conditional as always being FALSE (as PCRE used to), but it
1854        seems to me that giving a diagnostic is better.
1855    
1856    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1857        and messages.
1858    
1859    9.  Fix two occurrences of "backslash" in comments that should have been
1860        "backspace".
1861    
1862    10. Remove two redundant lines of code that can never be obeyed (their function
1863        was moved elsewhere).
1864    
1865    11. The program that makes PCRE's Unicode character property table had a bug
1866        which caused it to generate incorrect table entries for sequences of
1867        characters that have the same character type, but are in different scripts.
1868        It amalgamated them into a single range, with the script of the first of
1869        them. In other words, some characters were in the wrong script. There were
1870        thirteen such cases, affecting characters in the following ranges:
1871    
1872          U+002b0 - U+002c1
1873          U+0060c - U+0060d
1874          U+0061e - U+00612
1875          U+0064b - U+0065e
1876          U+0074d - U+0076d
1877          U+01800 - U+01805
1878          U+01d00 - U+01d77
1879          U+01d9b - U+01dbf
1880          U+0200b - U+0200f
1881          U+030fc - U+030fe
1882          U+03260 - U+0327f
1883          U+0fb46 - U+0fbb1
1884          U+10450 - U+1049d
1885    
1886    12. The -o option (show only the matching part of a line) for pcregrep was not
1887        compatible with GNU grep in that, if there was more than one match in a
1888        line, it showed only the first of them. It now behaves in the same way as
1889        GNU grep.
1890    
1891    13. If the -o and -v options were combined for pcregrep, it printed a blank
1892        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1893        does the same. The return code can be used to tell if there were any
1894        non-matching lines.
1895    
1896    14. Added --file-offsets and --line-offsets to pcregrep.
1897    
1898    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1899        infinitely looping recursion. The bug was that positive lookaheads were not
1900        being skipped when checking for a possible empty match (negative lookaheads
1901        and both kinds of lookbehind were skipped).
1902    
1903    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1904        inclusion of <windows.h> to before rather than after the definition of
1905        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1906    
1907    17. Specifying a possessive quantifier with a specific limit for a Unicode
1908        character property caused pcre_compile() to compile bad code, which led at
1909        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1910        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1911        caused the error; without that there was no problem.
1912    
1913    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1914    
1915    19. Added --enable-pcretest-libreadline.
1916    
1917    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1918        RE::GlobalReplace(). As a result, the number of replacements returned was
1919        double what it should be. I removed one of the increments, but Craig sent a
1920        later patch that removed the other one (the right fix) and added unit tests
1921        that check the return values (which was not done before).
1922    
1923    21. Several CMake things:
1924    
1925        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1926            the names libpcre and libpcreposix, not just pcre and pcreposix.
1927    
1928        (2) The above change means that pcretest and pcregrep are now correctly
1929            linked with the newly-built libraries, not previously installed ones.
1930    
1931        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1932    
1933    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1934        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1935        UTF-8 newline character). The key issue is that the pattern starts .*;
1936        this means that the match must be either at the beginning, or after a
1937        newline. The bug was in the code for advancing after a failed match and
1938        checking that the new position followed a newline. It was not taking
1939        account of UTF-8 characters correctly.
1940    
1941    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1942        character classes. PCRE was not treating the sequence [:...:] as a
1943        character class unless the ... were all letters. Perl, however, seems to
1944        allow any characters between [: and :], though of course it rejects as
1945        unknown any "names" that contain non-letters, because all the known class
1946        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1947        for example, whereas PCRE did not - it did not recognize a POSIX character
1948        class. This seemed a bit dangerous, so the code has been changed to be
1949        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1950        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1951        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1952        Perl does, and where it didn't before.
1953    
1954    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1955        Windows environments %n is disabled by default.
1956    
1957    
1958    Version 7.4 21-Sep-07
1959    ---------------------
1960    
1961    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1962        means that a class such as [\s] counted as "explicit reference to CR or
1963        LF". That isn't really right - the whole point of the change was to try to
1964        help when there was an actual mention of one of the two characters. So now
1965        the change happens only if \r or \n (or a literal CR or LF) character is
1966        encountered.
1967    
1968    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1969        of both had grown to the point where there were only 3 bits left.
1970        Fortunately, there was spare space in the data structure, and so I have
1971        moved the internal flags into a new 16-bit field to free up more option
1972        bits.
1973    
1974    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1975        but did not set the internal JCHANGED flag - either of these is enough to
1976        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1977        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1978        start sets both bits.
1979    
1980    4.  Added options (at build time, compile time, exec time) to change \R from
1981        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1982    
1983    5.  doc/pcresyntax.html was missing from the distribution.
1984    
1985    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1986        compatibility, even though it is no longer used.
1987    
1988    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1989        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1990        windows.h file is present (where different names are used). [This was
1991        reversed later after testing - see 16 below.]
1992    
1993    8.  Changed all #include <config.h> to #include "config.h". There were also
1994        some further <pcre.h> cases that I changed to "pcre.h".
1995    
1996    9.  When pcregrep was used with the --colour option, it missed the line ending
1997        sequence off the lines that it output.
1998    
1999    10. It was pointed out to me that arrays of string pointers cause lots of
2000        relocations when a shared library is dynamically loaded. A technique of
2001        using a single long string with a table of offsets can drastically reduce
2002        these. I have refactored PCRE in four places to do this. The result is
2003        dramatic:
2004    
2005          Originally:                          290
2006          After changing UCP table:            187
2007          After changing error message table:   43
2008          After changing table of "verbs"       36
2009          After changing table of Posix names   22
2010    
2011        Thanks to the folks working on Gregex for glib for this insight.
2012    
2013    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2014        unicode-properties was also set.
2015    
2016    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2017    
2018    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2019        checked only for CRLF.
2020    
2021    14. Added casts to pcretest.c to avoid compiler warnings.
2022    
2023    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2024    
2025    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2026        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2027        entirely. This removes changes made in 7 above.
2028    
2029    17. The CMake files have been updated, and there is now more information about
2030        building with CMake in the NON-UNIX-USE document.
2031    
2032    
2033    Version 7.3 28-Aug-07
2034    ---------------------
2035    
2036     1. In the rejigging of the build system that eventually resulted in 7.1, the
2037        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2038        brackets there is not right, since it causes compilers to look for an
2039        installed pcre.h, not the version that is in the source that is being
2040        compiled (which of course may be different). I have changed it back to:
2041    
2042          #include "pcre.h"
2043    
2044        I have a vague recollection that the change was concerned with compiling in
2045        different directories, but in the new build system, that is taken care of
2046        by the VPATH setting the Makefile.
2047    
2048     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2049        when the subject happened to end in the byte 0x85 (e.g. if the last
2050        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2051        characters but of course it shouldn't be taken as a newline when it is part
2052        of another character. The bug was that, for an unlimited repeat of . in
2053        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2054        characters when looking for a newline.
2055    
2056     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2057    
2058     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2059        in debug output.
2060    
2061     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2062        long printing in the pcrecpp unittest when running under MinGW.
2063    
2064     6. ESC_K was left out of the EBCDIC table.
2065    
2066     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2067        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2068        limit also applies to "virtual nesting" when a pattern is recursive, and in
2069        this case 1000 isn't so big. I have been able to remove this limit at the
2070        expense of backing off one optimization in certain circumstances. Normally,
2071        when pcre_exec() would call its internal match() function recursively and
2072        immediately return the result unconditionally, it uses a "tail recursion"
2073        feature to save stack. However, when a subpattern that can match an empty
2074        string has an unlimited repetition quantifier, it no longer makes this
2075        optimization. That gives it a stack frame in which to save the data for
2076        checking that an empty string has been matched. Previously this was taken
2077        from the 1000-entry workspace that had been reserved. So now there is no
2078        explicit limit, but more stack is used.
2079    
2080     8. Applied Daniel's patches to solve problems with the import/export magic
2081        syntax that is required for Windows, and which was going wrong for the
2082        pcreposix and pcrecpp parts of the library. These were overlooked when this
2083        problem was solved for the main library.
2084    
2085     9. There were some crude static tests to avoid integer overflow when computing
2086        the size of patterns that contain repeated groups with explicit upper
2087        limits. As the maximum quantifier is 65535, the maximum group length was
2088        set at 30,000 so that the product of these two numbers did not overflow a
2089        32-bit integer. However, it turns out that people want to use groups that
2090        are longer than 30,000 bytes (though not repeat them that many times).
2091        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2092        made it possible to implement the integer overflow checks in a much more
2093        dynamic way, which I have now done. The artificial limitation on group
2094        length has been removed - we now have only the limit on the total length of
2095        the compiled pattern, which depends on the LINK_SIZE setting.
2096    
2097    10. Fixed a bug in the documentation for get/copy named substring when
2098        duplicate names are permitted. If none of the named substrings are set, the
2099        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2100        empty string.
2101    
2102    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2103        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2104        because the ] is interpreted as the first data character and the
2105        terminating ] is not found. PCRE has been made compatible with Perl in this
2106        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2107        cause memory overwriting.
2108    
2109    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2110        string has been matched (to stop an infinite loop). It was not recognizing
2111        a conditional subpattern that could match an empty string if that
2112        subpattern was within another subpattern. For example, it looped when
2113        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2114        condition was not nested. This bug has been fixed.
2115    
2116    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2117        past the start of the subject in the presence of bytes with the top bit
2118        set, for example "\x8aBCD".
2119    
2120    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2121        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2122    
2123    14. Optimized (?!) to (*FAIL).
2124    
2125    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2126        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2127        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2128        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2129        does: it's just the validity check that is more restrictive.
2130    
2131    16. Inserted checks for integer overflows during escape sequence (backslash)
2132        processing, and also fixed erroneous offset values for syntax errors during
2133        backslash processing.
2134    
2135    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2136        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2137    
2138    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2139        caused an overrun.
2140    
2141    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2142        something other than just ASCII characters) inside a group that had an
2143        unlimited repeat caused a loop at compile time (while checking to see
2144        whether the group could match an empty string).
2145    
2146    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2147        [\P{Any}] did so. (Error in the code for printing property names.)
2148    
2149    21. An orphan \E inside a character class could cause a crash.
2150    
2151    22. A repeated capturing bracket such as (A)? could cause a wild memory
2152        reference during compilation.
2153    
2154    23. There are several functions in pcre_compile() that scan along a compiled
2155        expression for various reasons (e.g. to see if it's fixed length for look
2156        behind). There were bugs in these functions when a repeated \p or \P was
2157        present in the pattern. These operators have additional parameters compared
2158        with \d, etc, and these were not being taken into account when moving along
2159        the compiled data. Specifically:
2160    
2161        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2162            length.
2163    
2164        (b) An item such as \pL+ within a repeated group could cause crashes or
2165            loops.
2166    
2167        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2168            "reference to non-existent subpattern" error.
2169    
2170        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2171    
2172    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2173        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2174    
2175    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2176    
2177    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2178        character were causing crashes (broken optimization).
2179    
2180    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2181        \p or \P) caused a compile-time loop.
2182    
2183    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2184        break. For example, the unstudied pattern [\r\n]A does not match the string
2185        "\r\nA" because change 7.0/46 below moves the current point on by two
2186        characters after failing to match at the start. However, the pattern \nA
2187        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2188        the same is true. There doesn't seem any very clean way out of this, but
2189        what I have chosen to do makes the common cases work: PCRE now takes note
2190        of whether there can be an explicit match for \r or \n anywhere in the
2191        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2192        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2193        pattern has explicit CR or LF references.
2194    
2195    29. Added (*CR) etc for changing newline setting at start of pattern.
2196    
2197    
2198    Version 7.2 19-Jun-07
2199  ---------------------  ---------------------
2200    
2201   1. If the fr_FR locale cannot be found for test 3, try the "french" locale,   1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
# Line 69  Version 7.2 13-June-07 Line 2263  Version 7.2 13-June-07
2263        pcrecpp::RE("a*?").FullMatch("aaa") does not, and        pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2264        pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.        pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2265    
2266    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2267        it matched the wrong number of bytes.
2268    
2269    
2270  Version 7.1 24-Apr-07  Version 7.1 24-Apr-07
2271  ---------------------  ---------------------

Legend:
Removed from v.182  
changed lines
  Added in v.1211

  ViewVC Help
Powered by ViewVC 1.1.5