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

Diff of /code/trunk/ChangeLog

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

revision 127 by ph10, Mon Mar 19 11:44:45 2007 UTC revision 1272 by zherczeg, Thu Mar 7 11:30:01 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.1 12-Mar-07  Version 8.33 xx-xxxx-201x
5    -------------------------
6    
7    1.  Added 'U' to some constants that are compared to unsigned integers, to
8        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
9        variables that are added to signed variables, to ensure the result is
10        signed and can be negated.
11    
12    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
13        CMake config files.
14    
15    3.  Revise the creation of config.h.generic so that all boolean macros are
16        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
17        overriding via -D on the command line possible.
18    
19    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
20        to unsigned int is reported to make a quite noticeable speed difference in
21        a specific Windows environment. Testing on Linux did also appear to show
22        some benefit (and it is clearly not harmful). Also fixed the definition of
23        Xop which should be unsigned.
24    
25    5.  Related to (4), changing the definition of the intermediate variable cc
26        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
27        improvements.
28    
29    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
30        unnecessary spaces.
31    
32    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
33        and later.
34    
35    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
36    
37    9.  Optimizing fast_forward_start_bits in JIT.
38    
39    10. Adding experimental support for callouts in JIT, and fixing some
40        issues revealed during this work. Namely:
41    
42        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
43    
44        (b) Minimum length was not checked before the matching is started.
45    
46    11. The value of capture_last that is passed to callouts was incorrect in some
47        cases when there was a capture on one path that was subsequently abandoned
48        after a backtrack. Also, the capture_last value is now reset after a
49        recursion, since all captures are also reset in this case.
50    
51    12. The interpreter no longer returns the "too many substrings" error in the
52        case when an overflowing capture is in a branch that is subsequently
53        abandoned after a backtrack.
54    
55    13. In the pathological case when an offset vector of size 2 is used, pcretest
56        now prints out the matched string after a yield of 0 or 1.
57    
58    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
59        Only supported by the JIT compiler at the moment.
60    
61    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
62    
63    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
64        offset of the starting point of the matching process, provided the offsets
65        vector is large enough.
66    
67    17. The \A escape now records a lookbehind value of 1, though its execution
68        does not actually inspect the previous character. This is to ensure that,
69        in partial multi-segment matching, at least one character from the old
70        segment is retained when a new segment is processed. Otherwise, if there
71        are no lookbehinds in the pattern, \A might match incorrectly at the start
72        of a new segment.
73    
74    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
75    
76    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
77        gives some modest performance improvement in 8-bit mode.
78    
79    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
80        be expressed in certain programming languages using Universal Character
81        Names.
82    
83    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
84        which points out that "non characters" are not "characters that may not
85        appear in Unicode strings" but rather "characters that are reserved for
86        internal use and have only local meaning".
87    
88    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
89        there was a conditional group that depended on an assertion, if the
90        assertion was false, the callout that immediately followed the alternation
91        in the condition was skipped when pcre_exec() was used for matching.
92    
93    23. Allow an explicit callout to be inserted before an assertion that is the
94        condition for a conditional group, for compatibility with automatic
95        callouts, which always insert a callout at this point.
96    
97    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
98        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
99    
100    25. (*PRUNE) is now supported by the JIT compiler.
101    
102    
103    Version 8.32 30-November-2012
104    -----------------------------
105    
106    1.  Improved JIT compiler optimizations for first character search and single
107        character iterators.
108    
109    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
110        Patch by Daniel Richard G.
111    
112    3.  Single character iterator optimizations in the JIT compiler.
113    
114    4.  Improved JIT compiler optimizations for character ranges.
115    
116    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
117        Reported by Giuseppe D'Angelo.
118    
119    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
120        of a line, was being set incorrectly in cases where .* appeared inside
121        atomic brackets at the start of a pattern, or where there was a subsequent
122        *PRUNE or *SKIP.
123    
124    7.  Improved instruction cache flush for POWER/PowerPC.
125        Patch by Daniel Richard G.
126    
127    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
128        grep:
129    
130        (a) There is now no limit to the number of patterns to be matched.
131    
132        (b) An error is given if a pattern is too long.
133    
134        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
135            are now supported.
136    
137        (d) --exclude-from and --include-from (multiple use) have been added.
138    
139        (e) Exclusions and inclusions now apply to all files and directories, not
140            just to those obtained from scanning a directory recursively.
141    
142        (f) Multiple uses of -f and --file-list are now supported.
143    
144        (g) In a Windows environment, the default for -d has been changed from
145            "read" (the GNU grep default) to "skip", because otherwise the presence
146            of a directory in the file list provokes an error.
147    
148        (h) The documentation has been revised and clarified in places.
149    
150    9.  Improve the matching speed of capturing brackets.
151    
152    10. Changed the meaning of \X so that it now matches a Unicode extended
153        grapheme cluster.
154    
155    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
156        out POSIX threads when JIT support is configured.
157    
158    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
159    
160    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
161        structure could go wrong in environments where size_t is not the same size
162        as int.
163    
164    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
165        set.
166    
167    15. The EBCDIC support had decayed; later updates to the code had included
168        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
169        general tidy up of EBCDIC-related issues, and the documentation was also
170        not quite right. There is now a test that can be run on ASCII systems to
171        check some of the EBCDIC-related things (but is it not a full test).
172    
173    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
174        in a small tidy to the code.
175    
176    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
177    
178    18. If the --only-matching (-o) option in pcregrep is specified multiple
179        times, each one causes appropriate output. For example, -o1 -o2 outputs the
180        substrings matched by the 1st and 2nd capturing parentheses. A separating
181        string can be specified by --om-separator (default empty).
182    
183    19. Improving the first n character searches.
184    
185    20. Turn case lists for horizontal and vertical white space into macros so that
186        they are defined only once.
187    
188    21. This set of changes together give more compatible Unicode case-folding
189        behaviour for characters that have more than one other case when UCP
190        support is available.
191    
192        (a) The Unicode property table now has offsets into a new table of sets of
193            three or more characters that are case-equivalent. The MultiStage2.py
194            script that generates these tables (the pcre_ucd.c file) now scans
195            CaseFolding.txt instead of UnicodeData.txt for character case
196            information.
197    
198        (b) The code for adding characters or ranges of characters to a character
199            class has been abstracted into a generalized function that also handles
200            case-independence. In UTF-mode with UCP support, this uses the new data
201            to handle characters with more than one other case.
202    
203        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
204            whose other case is greater than 256 are now correctly matched
205            caselessly. Previously, the high codepoint matched the low one, but not
206            vice versa.
207    
208        (d) The processing of \h, \H, \v, and \ in character classes now makes use
209            of the new class addition function, using character lists defined as
210            macros alongside the case definitions of 20 above.
211    
212        (e) Caseless back references now work with characters that have more than
213            one other case.
214    
215        (f) General caseless matching of characters with more than one other case
216            is supported.
217    
218    22. Unicode character properties were updated from Unicode 6.2.0
219    
220    23. Improved CMake support under Windows. Patch by Daniel Richard G.
221    
222    24. Add support for 32-bit character strings, and UTF-32
223    
224    25. Major JIT compiler update (code refactoring and bugfixing).
225        Experimental Sparc 32 support is added.
226    
227    26. Applied a modified version of Daniel Richard G's patch to create
228        pcre.h.generic and config.h.generic by "make" instead of in the
229        PrepareRelease script.
230    
231    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
232        pcre_compile.c when checking for a zero character.
233    
234    28. Introducing a native interface for JIT. Through this interface, the compiled
235        machine code can be directly executed. The purpose of this interface is to
236        provide fast pattern matching, so several sanity checks are not performed.
237        However, feature tests are still performed. The new interface provides
238        1.4x speedup compared to the old one.
239    
240    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
241        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
242        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
243    
244    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
245        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
246        the "old" RFC 2279). Instead, it ended up passing a negative length to
247        pcre_exec().
248    
249    31. Add support for GCC's visibility feature to hide internal functions.
250    
251    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
252        "unknown -C option" after outputting 0 or 1.
253    
254    33. There is now support for generating a code coverage report for the test
255        suite in environments where gcc is the compiler and lcov is installed. This
256        is mainly for the benefit of the developers.
257    
258    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
259        unaddressable using valgrind annotations, allowing valgrind to detect
260        invalid memory accesses. This is mainly for the benefit of the developers.
261    
262    25. (*UTF) can now be used to start a pattern in any of the three libraries.
263    
264    26. Give configure error if --enable-cpp but no C++ compiler found.
265    
266    
267    Version 8.31 06-July-2012
268    -------------------------
269    
270    1.  Fixing a wrong JIT test case and some compiler warnings.
271    
272    2.  Removed a bashism from the RunTest script.
273    
274    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
275        to unsigned type, result still unsigned" that was given by an MS compiler
276        on encountering the code "-sizeof(xxx)".
277    
278    4.  Partial matching support is added to the JIT compiler.
279    
280    5.  Fixed several bugs concerned with partial matching of items that consist
281        of more than one character:
282    
283        (a) /^(..)\1/ did not partially match "aba" because checking references was
284            done on an "all or nothing" basis. This also applied to repeated
285            references.
286    
287        (b) \R did not give a hard partial match if \r was found at the end of the
288            subject.
289    
290        (c) \X did not give a hard partial match after matching one or more
291            characters at the end of the subject.
292    
293        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
294            a partial match for the string "\r".
295    
296        (e) When newline was set to CRLF, the metacharacter "." did not recognize
297            a partial match for a CR character at the end of the subject string.
298    
299    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
300        running pcretest, the text "(JIT)" added to the output whenever JIT is
301        actually used to run the match.
302    
303    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
304        or /S+[+] with a digit between 1 and 7.
305    
306    8.  OP_NOT now supports any UTF character not just single-byte ones.
307    
308    9.  (*MARK) control verb is now supported by the JIT compiler.
309    
310    10. The command "./RunTest list" lists the available tests without actually
311        running any of them. (Because I keep forgetting what they all are.)
312    
313    11. Add PCRE_INFO_MAXLOOKBEHIND.
314    
315    12. Applied a (slightly modified) user-supplied patch that improves performance
316        when the heap is used for recursion (compiled with --disable-stack-for-
317        recursion). Instead of malloc and free for each heap frame each time a
318        logical recursion happens, frames are retained on a chain and re-used where
319        possible. This sometimes gives as much as 30% improvement.
320    
321    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
322        call.
323    
324    14. As documented, (*COMMIT) is now confined to within a positive assertion.
325    
326    15. It is now possible to link pcretest with libedit as an alternative to
327        libreadline.
328    
329    16. (*COMMIT) control verb is now supported by the JIT compiler.
330    
331    17. The Unicode data tables have been updated to Unicode 6.1.0.
332    
333    18. Added --file-list option to pcregrep.
334    
335    19. Added binary file support to pcregrep, including the -a, --binary-files,
336        -I, and --text options.
337    
338    20. The madvise function is renamed for posix_madvise for QNX compatibility
339        reasons. Fixed by Giuseppe D'Angelo.
340    
341    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
342    
343    22. Changed the option for creating symbolic links for 16-bit man pages from
344        -s to -sf so that re-installing does not cause issues.
345    
346    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
347    
348    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
349        matches in certain environments (the workspace was not being correctly
350        retained). Also added to pcre_dfa_exec() a simple plausibility check on
351        some of the workspace data at the beginning of a restart.
352    
353    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
354        was not doing so when it should - probably a typo introduced by SVN 528
355        (change 8.10/14).
356    
357    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
358        \w+ when the character tables indicated that \x{c4} was a word character.
359        There were several related cases, all because the tests for doing a table
360        lookup were testing for characters less than 127 instead of 255.
361    
362    27. If a pattern contains capturing parentheses that are not used in a match,
363        their slots in the ovector are set to -1. For those that are higher than
364        any matched groups, this happens at the end of processing. In the case when
365        there were back references that the ovector was too small to contain
366        (causing temporary malloc'd memory to be used during matching), and the
367        highest capturing number was not used, memory off the end of the ovector
368        was incorrectly being set to -1. (It was using the size of the temporary
369        memory instead of the true size.)
370    
371    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
372        ovector size, it uses memory at the end of the block that it has got.
373    
374    29. Check for an overlong MARK name and give an error at compile time. The
375        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
376    
377    30. JIT compiler update.
378    
379    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
380        Rill for the patch.
381    
382    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
383    
384    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
385    
386    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
387        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
388        was enabled.
389    
390    35. Improve JIT code generation for greedy plus quantifier.
391    
392    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
393        1 to "aa" instead of to an empty string. The bug affected repeated groups
394        that could potentially match an empty string.
395    
396    37. Optimizing single character iterators in JIT.
397    
398    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
399        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
400        codepoints that are too big for the mode are faulted, and in a UTF mode,
401        disallowed codepoints are also faulted.
402    
403    39. If PCRE was compiled with UTF support, in three places in the DFA
404        matcher there was code that should only have been obeyed in UTF mode, but
405        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
406        processing when bytes with values greater than 127 were present. In 16-bit
407        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
408        both cases the values are those that cannot be the first data item in a UTF
409        character. The three items that might have provoked this were recursions,
410        possessively repeated groups, and atomic groups.
411    
412    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
413        and pcregrep, because some OS require shared objects to be explicitly
414        passed to ld, causing the link step to fail if they are not.
415    
416    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
417        mode, patterns that started with \h* or \R* might be incorrectly matched.
418    
419    
420    Version 8.30 04-February-2012
421    -----------------------------
422    
423    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
424        name is defined in ctype.h.
425    
426    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
427        only in quite long subpatterns.
428    
429    3.  Removed the function pcre_info(), which has been obsolete and deprecated
430        since it was replaced by pcre_fullinfo() in February 2000.
431    
432    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
433        match a (*MARK), and the match failed at the start of the subject, a
434        reference to memory before the start of the subject could occur. This bug
435        was introduced by fix 17 of release 8.21.
436    
437    5.  A reference to an unset group with zero minimum repetition was giving
438        totally wrong answers (in non-JavaScript-compatibility mode). For example,
439        /(another)?(\1?)test/ matched against "hello world test". This bug was
440        introduced in release 8.13.
441    
442    6.  Add support for 16-bit character strings (a large amount of work involving
443        many changes and refactorings).
444    
445    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
446        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
447        from a file.
448    
449    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
450        rounding is not applied in this particular case).
451    
452    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
453        if they appear, or are escaped, in patterns.
454    
455    10. Get rid of a number of -Wunused-but-set-variable warnings.
456    
457    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
458        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
459        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
460        also returns the mark "x". This bug applied to capturing parentheses,
461        non-capturing parentheses, and atomic parentheses. It also applied to some
462        assertions.
463    
464    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
465        information out of configure.ac instead of relying on pcre.h.generic, which
466        is not stored in the repository.
467    
468    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
469        -lreadline.
470    
471    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
472    
473    15. Applied Graycode's patch to put the top-level frame on the stack rather
474        than the heap when not using the stack for recursion. This gives a
475        performance improvement in many cases when recursion is not deep.
476    
477    16. Experimental code added to "pcretest -C" to output the stack frame size.
478    
479    
480    Version 8.21 12-Dec-2011
481    ------------------------
482    
483    1.  Updating the JIT compiler.
484    
485    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
486        are added as well.
487    
488    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
489        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
490        calling _pcre_jit_exec. Some extra comments are added.
491    
492    4.  (*MARK) settings inside atomic groups that do not contain any capturing
493        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
494        was introduced by change 18 for 8.20.
495    
496    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
497        ECMA-262 standard.
498    
499    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
500        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
501        This bug was probably introduced by change 9 of 8.13.
502    
503    7.  While fixing 6 above, I noticed that a number of other items were being
504        incorrectly rejected as "not fixed length". This arose partly because newer
505        opcodes had not been added to the fixed-length checking code. I have (a)
506        corrected the bug and added tests for these items, and (b) arranged for an
507        error to occur if an unknown opcode is encountered while checking for fixed
508        length instead of just assuming "not fixed length". The items that were
509        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
510        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
511        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
512    
513    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
514        being incorrectly compiled and would have given unpredicatble results.
515    
516    9.  A possessively repeated subpattern with minimum repeat count greater than
517        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
518        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
519        the first (A) could occur when it should not.
520    
521    10. Add a cast and remove a redundant test from the code.
522    
523    11. JIT should use pcre_malloc/pcre_free for allocation.
524    
525    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
526        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
527        is anything other than /usr, -L is still shown).
528    
529    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
530    
531    14. Perl does not support \N without a following name in a [] class; PCRE now
532        also gives an error.
533    
534    15. If a forward reference was repeated with an upper limit of around 2000,
535        it caused the error "internal error: overran compiling workspace". The
536        maximum number of forward references (including repeats) was limited by the
537        internal workspace, and dependent on the LINK_SIZE. The code has been
538        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
539        the default depends on LINK_SIZE. There is a new upper limit (for safety)
540        of around 200,000 forward references. While doing this, I also speeded up
541        the filling in of repeated forward references.
542    
543    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
544        incorrectly expecting the subject to contain another "a" after the start.
545    
546    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
547        in the match, the SKIP should be ignored. This was not happening; instead
548        the SKIP was being treated as NOMATCH. For patterns such as
549        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
550        tested.
551    
552    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
553        now much more compatible with Perl, in particular in cases where the result
554        is a non-match for a non-anchored pattern. For example, if
555        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
556        "m", where previously it did not return a name. A side effect of this
557        change is that for partial matches, the last encountered mark name is
558        returned, as for non matches. A number of tests that were previously not
559        Perl-compatible have been moved into the Perl-compatible test files. The
560        refactoring has had the pleasing side effect of removing one argument from
561        the match() function, thus reducing its stack requirements.
562    
563    19. If the /S+ option was used in pcretest to study a pattern using JIT,
564        subsequent uses of /S (without +) incorrectly behaved like /S+.
565    
566    21. Retrieve executable code size support for the JIT compiler and fixing
567        some warnings.
568    
569    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
570        not work when the shorter character appeared right at the end of the
571        subject string.
572    
573    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
574        systems.
575    
576    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
577        output it when the /M option is used in pcretest.
578    
579    25. The CheckMan script was not being included in the distribution. Also, added
580        an explicit "perl" to run Perl scripts from the PrepareRelease script
581        because this is reportedly needed in Windows.
582    
583    26. If study data was being save in a file and studying had not found a set of
584        "starts with" bytes for the pattern, the data written to the file (though
585        never used) was taken from uninitialized memory and so caused valgrind to
586        complain.
587    
588    27. Updated RunTest.bat as provided by Sheri Pierce.
589    
590    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
591    
592    29. Computation of memory usage for the table of capturing group names was
593        giving an unnecessarily large value.
594    
595    
596    Version 8.20 21-Oct-2011
597    ------------------------
598    
599    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
600        a POSIX class. After further experiments with Perl, which convinced me that
601        Perl has bugs and confusions, a closing square bracket is no longer allowed
602        in a POSIX name. This bug also affected patterns with classes that started
603        with full stops.
604    
605    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
606        captured substring, but while checking the failing first alternative,
607        substring 1 is temporarily captured. If the output vector supplied to
608        pcre_exec() was not big enough for this capture, the yield of the function
609        was still zero ("insufficient space for captured substrings"). This cannot
610        be totally fixed without adding another stack variable, which seems a lot
611        of expense for a edge case. However, I have improved the situation in cases
612        such as /(a)(b)x|abc/ matched against "abc", where the return code
613        indicates that fewer than the maximum number of slots in the ovector have
614        been set.
615    
616    3.  Related to (2) above: when there are more back references in a pattern than
617        slots in the output vector, pcre_exec() uses temporary memory during
618        matching, and copies in the captures as far as possible afterwards. It was
619        using the entire output vector, but this conflicts with the specification
620        that only 2/3 is used for passing back captured substrings. Now it uses
621        only the first 2/3, for compatibility. This is, of course, another edge
622        case.
623    
624    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
625        main code base, and can be used by building with --enable-jit. When this is
626        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
627        runtime --no-jit option is given.
628    
629    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
630        ovector, the return from the function was zero, implying that there were
631        other matches that did not fit. The correct "exactly full" value is now
632        returned.
633    
634    6.  If a subpattern that was called recursively or as a subroutine contained
635        (*PRUNE) or any other control that caused it to give a non-standard return,
636        invalid errors such as "Error -26 (nested recursion at the same subject
637        position)" or even infinite loops could occur.
638    
639    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
640        computing the minimum length on reaching *ACCEPT, and so ended up with the
641        wrong value of 1 rather than 0. Further investigation indicates that
642        computing a minimum subject length in the presence of *ACCEPT is difficult
643        (think back references, subroutine calls), and so I have changed the code
644        so that no minimum is registered for a pattern that contains *ACCEPT.
645    
646    8.  If (*THEN) was present in the first (true) branch of a conditional group,
647        it was not handled as intended. [But see 16 below.]
648    
649    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
650        Sheri Pierce.
651    
652    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
653        the first byte in a match must be "a".
654    
655    11. Change 17 for 8.13 increased the recursion depth for patterns like
656        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
657        pattern contains any instances of (*THEN). If it does not, the old
658        optimizations are restored. It would be nice to do this on a per-group
659        basis, but at the moment that is not feasible.
660    
661    12. In some environments, the output of pcretest -C is CRLF terminated. This
662        broke RunTest's code that checks for the link size. A single white space
663        character after the value is now allowed for.
664    
665    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
666        For "fr", it uses the Windows-specific input and output files.
667    
668    14. If (*THEN) appeared in a group that was called recursively or as a
669        subroutine, it did not work as intended. [But see next item.]
670    
671    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
672        pattern fragments (but not containing any | characters). If A and B are
673        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
674        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
675        D. In other words, Perl considers parentheses that do not contain any |
676        characters to be part of a surrounding alternative, whereas PCRE was
677        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
678        differently. PCRE now behaves in the same way as Perl, except in the case
679        of subroutine/recursion calls such as (?1) which have in any case always
680        been different (but PCRE had them first :-).
681    
682    16. Related to 15 above: Perl does not treat the | in a conditional group as
683        creating alternatives. Such a group is treated in the same way as an
684        ordinary group without any | characters when processing (*THEN). PCRE has
685        been changed to match Perl's behaviour.
686    
687    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
688        RunGrepTest script failed.
689    
690    18. Change 22 for version 13 caused atomic groups to use more stack. This is
691        inevitable for groups that contain captures, but it can lead to a lot of
692        stack use in large patterns. The old behaviour has been restored for atomic
693        groups that do not contain any capturing parentheses.
694    
695    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
696        suppress the check for a minimum subject length at run time. (If it was
697        given to pcre_exec() or pcre_dfa_exec() it did work.)
698    
699    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
700        fail to work when decoding hex characters in data strings in EBCDIC
701        environments.
702    
703    21. It appears that in at least one Mac OS environment, the isxdigit() function
704        is implemented as a macro that evaluates to its argument more than once,
705        contravening the C 90 Standard (I haven't checked a later standard). There
706        was an instance in pcretest which caused it to go wrong when processing
707        \x{...} escapes in subject strings. The has been rewritten to avoid using
708        things like p++ in the argument of isxdigit().
709    
710    
711    Version 8.13 16-Aug-2011
712    ------------------------
713    
714    1.  The Unicode data tables have been updated to Unicode 6.0.0.
715    
716    2.  Two minor typos in pcre_internal.h have been fixed.
717    
718    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
719        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
720        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
721    
722    4.  There were a number of related bugs in the code for matching backrefences
723        caselessly in UTF-8 mode when codes for the characters concerned were
724        different numbers of bytes. For example, U+023A and U+2C65 are an upper
725        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
726        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
727        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
728        2-byte code at the end of the subject (it thought there wasn't enough data
729        left).
730    
731    5.  Comprehensive information about what went wrong is now returned by
732        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
733        as the output vector has at least 2 elements. The offset of the start of
734        the failing character and a reason code are placed in the vector.
735    
736    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
737        now returned is for the first byte of the failing character, instead of the
738        last byte inspected. This is an incompatible change, but I hope it is small
739        enough not to be a problem. It makes the returned offset consistent with
740        pcre_exec() and pcre_dfa_exec().
741    
742    7.  pcretest now gives a text phrase as well as the error number when
743        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
744        failure, the offset and reason code are output.
745    
746    8.  When \R was used with a maximizing quantifier it failed to skip backwards
747        over a \r\n pair if the subsequent match failed. Instead, it just skipped
748        back over a single character (\n). This seems wrong (because it treated the
749        two characters as a single entity when going forwards), conflicts with the
750        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
751        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
752        has been changed.
753    
754    9.  Some internal refactoring has changed the processing so that the handling
755        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
756        time (the PCRE_DOTALL option was changed this way some time ago: version
757        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
758        which was always a bit of a fudge. It also means that there is one less
759        argument for the match() function, which reduces its stack requirements
760        slightly. This change also fixes an incompatibility with Perl: the pattern
761        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
762    
763    10. More internal refactoring has drastically reduced the number of recursive
764        calls to match() for possessively repeated groups such as (abc)++ when
765        using pcre_exec().
766    
767    11. While implementing 10, a number of bugs in the handling of groups were
768        discovered and fixed:
769    
770        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
771        (a|)*(?1) gave a compile-time internal error.
772        ((a|)+)+  did not notice that the outer group could match an empty string.
773        (^a|^)+   was not marked as anchored.
774        (.*a|.*)+ was not marked as matching at start or after a newline.
775    
776    12. Yet more internal refactoring has removed another argument from the match()
777        function. Special calls to this function are now indicated by setting a
778        value in a variable in the "match data" data block.
779    
780    13. Be more explicit in pcre_study() instead of relying on "default" for
781        opcodes that mean there is no starting character; this means that when new
782        ones are added and accidentally left out of pcre_study(), testing should
783        pick them up.
784    
785    14. The -s option of pcretest has been documented for ages as being an old
786        synonym of -m (show memory usage). I have changed it to mean "force study
787        for every regex", that is, assume /S for every regex. This is similar to -i
788        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
789        using it. It makes it easier to run collections of tests with and without
790        study enabled, and thereby test pcre_study() more easily. All the standard
791        tests are now run with and without -s (but some patterns can be marked as
792        "never study" - see 20 below).
793    
794    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
795        restoration of the capturing data to the outer values was not happening
796        correctly.
797    
798    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
799        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
800        pattern had matched an empty string, and so incorrectly returned a no
801        match.
802    
803    17. There was optimizing code for the last branch of non-capturing parentheses,
804        and also for the obeyed branch of a conditional subexpression, which used
805        tail recursion to cut down on stack usage. Unfortunately, now that there is
806        the possibility of (*THEN) occurring in these branches, tail recursion is
807        no longer possible because the return has to be checked for (*THEN). These
808        two optimizations have therefore been removed. [But see 8.20/11 above.]
809    
810    18. If a pattern containing \R was studied, it was assumed that \R always
811        matched two bytes, thus causing the minimum subject length to be
812        incorrectly computed because \R can also match just one byte.
813    
814    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
815        was incorrectly computed.
816    
817    20. If /S is present twice on a test pattern in pcretest input, it now
818        *disables* studying, thereby overriding the use of -s on the command line
819        (see 14 above). This is necessary for one or two tests to keep the output
820        identical in both cases.
821    
822    21. When (*ACCEPT) was used in an assertion that matched an empty string and
823        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
824    
825    22. When an atomic group that contained a capturing parenthesis was
826        successfully matched, but the branch in which it appeared failed, the
827        capturing was not being forgotten if a higher numbered group was later
828        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
829        group 1 to "a", when in fact it should be unset. This applied to multi-
830        branched capturing and non-capturing groups, repeated or not, and also to
831        positive assertions (capturing in negative assertions does not happen
832        in PCRE) and also to nested atomic groups.
833    
834    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
835        subject after a captured substring, to make it easier to tell which of a
836        number of identical substrings has been captured.
837    
838    24. The way atomic groups are processed by pcre_exec() has been changed so that
839        if they are repeated, backtracking one repetition now resets captured
840        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
841        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
842        "aaa". Previously, it would have been "a". As part of this code
843        refactoring, the way recursive calls are handled has also been changed.
844    
845    25. If an assertion condition captured any substrings, they were not passed
846        back unless some other capturing happened later. For example, if
847        (?(?=(a))a) was matched against "a", no capturing was returned.
848    
849    26. When studying a pattern that contained subroutine calls or assertions,
850        the code for finding the minimum length of a possible match was handling
851        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
852        group 1 called group 2 while simultaneously a separate group 2 called group
853        1). A stack overflow occurred in this case. I have fixed this by limiting
854        the recursion depth to 10.
855    
856    27. Updated RunTest.bat in the distribution to the version supplied by Tom
857        Fortmann. This supports explicit test numbers on the command line, and has
858        argument validation and error reporting.
859    
860    28. An instance of \X with an unlimited repeat could fail if at any point the
861        first character it looked at was a mark character.
862    
863    29. Some minor code refactoring concerning Unicode properties and scripts
864        should reduce the stack requirement of match() slightly.
865    
866    30. Added the '=' option to pcretest to check the setting of unused capturing
867        slots at the end of the pattern, which are documented as being -1, but are
868        not included in the return count.
869    
870    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
871        compiled something random. Now it gives a compile-time error (as does
872        Perl).
873    
874    32. A *MARK encountered during the processing of a positive assertion is now
875        recorded and passed back (compatible with Perl).
876    
877    33. If --only-matching or --colour was set on a pcregrep call whose pattern
878        had alternative anchored branches, the search for a second match in a line
879        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
880        matched the line "0102" twice. The same bug affected patterns that started
881        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
882        twice.
883    
884    34. Previously, PCRE did not allow quantification of assertions. However, Perl
885        does, and because of capturing effects, quantifying parenthesized
886        assertions may at times be useful. Quantifiers are now allowed for
887        parenthesized assertions.
888    
889    35. A minor code tidy in pcre_compile() when checking options for \R usage.
890    
891    36. \g was being checked for fancy things in a character class, when it should
892        just be a literal "g".
893    
894    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
895        appearance of a nested POSIX class supersedes an apparent external class.
896        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
897        unescaped square brackets may also appear as part of class names. For
898        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
899        more like Perl. (But see 8.20/1 above.)
900    
901    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
902        was because it thought it was \N{name}, which is not supported).
903    
904    39. Add minix to OS list not supporting the -S option in pcretest.
905    
906    40. PCRE tries to detect cases of infinite recursion at compile time, but it
907        cannot analyze patterns in sufficient detail to catch mutual recursions
908        such as ((?1))((?2)). There is now a runtime test that gives an error if a
909        subgroup is called recursively as a subpattern for a second time at the
910        same position in the subject string. In previous releases this might have
911        been caught by the recursion limit, or it might have run out of stack.
912    
913    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
914        happen only once. PCRE was, however incorrectly giving a compile time error
915        "recursive call could loop indefinitely" because it cannot analyze the
916        pattern in sufficient detail. The compile time test no longer happens when
917        PCRE is compiling a conditional subpattern, but actual runaway loops are
918        now caught at runtime (see 40 above).
919    
920    42. It seems that Perl allows any characters other than a closing parenthesis
921        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
922        has been changed to be the same.
923    
924    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
925        as not to get warnings when autogen.sh is called. Also changed
926        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
927    
928    44. To help people who use pcregrep to scan files containing exceedingly long
929        lines, the following changes have been made:
930    
931        (a) The default value of the buffer size parameter has been increased from
932            8K to 20K. (The actual buffer used is three times this size.)
933    
934        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
935            PCRE is built.
936    
937        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
938            to be set at run time.
939    
940        (d) Numerical values in pcregrep options can be followed by K or M, for
941            example --buffer-size=50K.
942    
943        (e) If a line being scanned overflows pcregrep's buffer, an error is now
944            given and the return code is set to 2.
945    
946    45. Add a pointer to the latest mark to the callout data block.
947    
948    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
949        partial match of an empty string instead of no match. This was specific to
950        the use of ".".
951    
952    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
953        complete match instead of a partial match. This bug was dependent on both
954        the PCRE_UTF8 and PCRE_DOTALL options being set.
955    
956    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
957        starting byte set, because \b was not being ignored.
958    
959    
960    Version 8.12 15-Jan-2011
961    ------------------------
962    
963    1.  Fixed some typos in the markup of the man pages, and wrote a script that
964        checks for such things as part of the documentation building process.
965    
966    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
967        --match-limit and --recursion-limit options (added for 8.11). In
968        particular, this made one of the standard tests fail. (The integer value
969        went into the wrong half of a long int.)
970    
971    3.  If the --colour option was given to pcregrep with -v (invert match), it
972        did strange things, either producing crazy output, or crashing. It should,
973        of course, ignore a request for colour when reporting lines that do not
974        match.
975    
976    4.  Another pcregrep bug caused similar problems if --colour was specified with
977        -M (multiline) and the pattern match finished with a line ending.
978    
979    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
980        matched in multiline mode, the following line was shown as part of the
981        match. This seems wrong, so I have changed it.
982    
983    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
984        the check for further matches in the same line (so they could be coloured)
985        to overrun the end of the current line. If another match was found, it was
986        incorrectly shown (and then shown again when found in the next line).
987    
988    7.  If pcregrep was compiled under Windows, there was a reference to the
989        function pcregrep_exit() before it was defined. I am assuming this was
990        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
991        reported by a user. I've moved the definition above the reference.
992    
993    
994    Version 8.11 10-Dec-2010
995    ------------------------
996    
997    1.  (*THEN) was not working properly if there were untried alternatives prior
998        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
999        backtracked to try for "b" instead of moving to the next alternative branch
1000        at the same level (in this case, to look for "c"). The Perl documentation
1001        is clear that when (*THEN) is backtracked onto, it goes to the "next
1002        alternative in the innermost enclosing group".
1003    
1004    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1005        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1006        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1007        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1008        (*THEN).
1009    
1010    3.  If \s appeared in a character class, it removed the VT character from
1011        the class, even if it had been included by some previous item, for example
1012        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1013        of \s, but is part of the POSIX "space" class.)
1014    
1015    4.  A partial match never returns an empty string (because you can always
1016        match an empty string at the end of the subject); however the checking for
1017        an empty string was starting at the "start of match" point. This has been
1018        changed to the "earliest inspected character" point, because the returned
1019        data for a partial match starts at this character. This means that, for
1020        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1021        (previously it gave "no match").
1022    
1023    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1024        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1025        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1026        has an implication that the given string is incomplete (because a partial
1027        match is preferred over a full match). For this reason, these items now
1028        give a partial match in this situation. [Aside: previously, the one case
1029        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1030        match rather than a full match, which was wrong by the old rules, but is
1031        now correct.]
1032    
1033    6.  There was a bug in the handling of #-introduced comments, recognized when
1034        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1035        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1036        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1037        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1038        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1039        places in pcre_compile().
1040    
1041    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1042        comments when looking ahead for named forward references to subpatterns,
1043        the only newline sequence it recognized was NL. It now handles newlines
1044        according to the set newline convention.
1045    
1046    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1047        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1048        cater for a lack of strerror(). These oversights have been fixed.
1049    
1050    9.  Added --match-limit and --recursion-limit to pcregrep.
1051    
1052    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1053    
1054    11. When the -o option was used, pcregrep was setting a return code of 1, even
1055        when matches were found, and --line-buffered was not being honoured.
1056    
1057    12. Added an optional parentheses number to the -o and --only-matching options
1058        of pcregrep.
1059    
1060    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1061        can match an empty string. The code to do it in pcretest and pcredemo
1062        needed fixing:
1063    
1064        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1065            only one byte after an empty string match just before CRLF (this case
1066            just got forgotten; "any" and "anycrlf" were OK).
1067    
1068        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1069            mode when an empty string match preceded an ASCII character followed by
1070            a non-ASCII character. (The code for advancing by one character rather
1071            than one byte was nonsense.)
1072    
1073        (c) The pcredemo.c sample program did not have any code at all to handle
1074            the cases when CRLF is a valid newline sequence.
1075    
1076    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1077        as a starting offset was within the subject string. There is now a new
1078        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1079        negative or greater than the length of the string. In order to test this,
1080        pcretest is extended to allow the setting of negative starting offsets.
1081    
1082    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1083        starting offset points to the beginning of a UTF-8 character was
1084        unnecessarily clumsy. I tidied it up.
1085    
1086    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1087        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1088    
1089    17. Nobody had reported that the --include_dir option, which was added in
1090        release 7.7 should have been called --include-dir (hyphen, not underscore)
1091        for compatibility with GNU grep. I have changed it to --include-dir, but
1092        left --include_dir as an undocumented synonym, and the same for
1093        --exclude-dir, though that is not available in GNU grep, at least as of
1094        release 2.5.4.
1095    
1096    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1097        characters from a string of bytes) have been redefined so as not to use
1098        loops, in order to improve performance in some environments. At the same
1099        time, I abstracted some of the common code into auxiliary macros to save
1100        repetition (this should not affect the compiled code).
1101    
1102    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1103        compile-time error is now given if \c is not followed by an ASCII
1104        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1105        different, and any byte value is allowed.)
1106    
1107    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1108        START_OPTIMIZE option, which is now allowed at compile time - but just
1109        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1110        to pcregrep and other applications that have no direct access to PCRE
1111        options. The new /Y option in pcretest sets this option when calling
1112        pcre_compile().
1113    
1114    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1115        back references. Groups containing recursive back references were forced to
1116        be atomic by that change, but in the case of named groups, the amount of
1117        memory required was incorrectly computed, leading to "Failed: internal
1118        error: code overflow". This has been fixed.
1119    
1120    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1121        pcretest.c, to avoid build problems in some Borland environments.
1122    
1123    
1124    Version 8.10 25-Jun-2010
1125    ------------------------
1126    
1127    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1128        THEN.
1129    
1130    2.  (*ACCEPT) was not working when inside an atomic group.
1131    
1132    3.  Inside a character class, \B is treated as a literal by default, but
1133        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1134        causes the error). The code is unchanged, but I tidied the documentation.
1135    
1136    4.  Inside a character class, PCRE always treated \R and \X as literals,
1137        whereas Perl faults them if its -w option is set. I have changed PCRE so
1138        that it faults them when PCRE_EXTRA is set.
1139    
1140    5.  Added support for \N, which always matches any character other than
1141        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1142    
1143    6.  When compiling pcregrep with newer versions of gcc which may have
1144        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1145        declared with attribute warn_unused_result" were given. Just casting the
1146        result to (void) does not stop the warnings; a more elaborate fudge is
1147        needed. I've used a macro to implement this.
1148    
1149    7.  Minor change to pcretest.c to avoid a compiler warning.
1150    
1151    8.  Added four artifical Unicode properties to help with an option to make
1152        \s etc use properties (see next item). The new properties are: Xan
1153        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1154    
1155    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1156        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1157        this option. Modified pcretest to add /W to test this facility. Added
1158        REG_UCP to make it available via the POSIX interface.
1159    
1160    10. Added --line-buffered to pcregrep.
1161    
1162    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1163        studied, and the match started with a letter with a code point greater than
1164        127 whose first byte was different to the first byte of the other case of
1165        the letter, the other case of this starting letter was not recognized
1166        (#976).
1167    
1168    12. If a pattern that was studied started with a repeated Unicode property
1169        test, for example, \p{Nd}+, there was the theoretical possibility of
1170        setting up an incorrect bitmap of starting bytes, but fortunately it could
1171        not have actually happened in practice until change 8 above was made (it
1172        added property types that matched character-matching opcodes).
1173    
1174    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1175        possible starting bytes for non-anchored patterns.
1176    
1177    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1178        \R, and also a number of cases that involve Unicode properties, both
1179        explicit and implicit when PCRE_UCP is set.
1180    
1181    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1182        input, it could crash or give wrong results if characters with values
1183        greater than 0xc0 were present in the subject string. (Detail: it assumed
1184        UTF-8 input when processing these items.)
1185    
1186    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1187        size_t is 64-bit (#991).
1188    
1189    17. Added a check for running out of memory when PCRE is compiled with
1190        --disable-stack-for-recursion (#990).
1191    
1192    18. If the last data line in a file for pcretest does not have a newline on
1193        the end, a newline was missing in the output.
1194    
1195    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1196        less than 128) in its various bitmaps. However, there is a facility for
1197        generating tables according to the current locale when PCRE is compiled. It
1198        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1199        characters, are recognized by isspace() and therefore were getting set in
1200        these tables, and indeed these tables seem to approximate to ISO 8859. This
1201        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1202        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1203        which of course cannot start UTF-8 characters. I have changed the code so
1204        that only real ASCII characters (less than 128) and the correct starting
1205        bytes for UTF-8 encodings are set for characters greater than 127 when in
1206        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1207        altogether.)
1208    
1209    20. Added the /T option to pcretest so as to be able to run tests with non-
1210        standard character tables, thus making it possible to include the tests
1211        used for 19 above in the standard set of tests.
1212    
1213    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1214        reference to a subpattern the other side of a comment that contains an
1215        opening parenthesis caused either an internal compiling error, or a
1216        reference to the wrong subpattern.
1217    
1218    
1219    Version 8.02 19-Mar-2010
1220    ------------------------
1221    
1222    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1223    
1224    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1225        configured.
1226    
1227    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1228        original author of that file, following a query about its status.
1229    
1230    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1231        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1232    
1233    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1234        quantifier applied to a forward-referencing subroutine call, could compile
1235        incorrect code or give the error "internal error: previously-checked
1236        referenced subpattern not found".
1237    
1238    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1239        variables to point to external functions. For these systems, therefore,
1240        pcre_malloc etc. are now initialized to local functions that call the
1241        relevant global functions.
1242    
1243    7.  There were two entries missing in the vectors called coptable and poptable
1244        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1245        I've fixed the data, and added a kludgy way of testing at compile time that
1246        the lengths are correct (equal to the number of opcodes).
1247    
1248    8.  Following on from 7, I added a similar kludge to check the length of the
1249        eint vector in pcreposix.c.
1250    
1251    9.  Error texts for pcre_compile() are held as one long string to avoid too
1252        much relocation at load time. To find a text, the string is searched,
1253        counting zeros. There was no check for running off the end of the string,
1254        which could happen if a new error number was added without updating the
1255        string.
1256    
1257    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1258    
1259    11. \K was not working if it appeared in an atomic group or in a group that
1260        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1261        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1262        the assertion is positive, but not if it is negative.
1263    
1264    12. Change 11 fortuitously reduced the size of the stack frame used in the
1265        "match()" function of pcre_exec.c by one pointer. Forthcoming
1266        implementation of support for (*MARK) will need an extra pointer on the
1267        stack; I have reserved it now, so that the stack frame size does not
1268        decrease.
1269    
1270    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1271        item in branch that calls a recursion is a subroutine call - as in the
1272        second branch in the above example - was incorrectly given the compile-
1273        time error "recursive call could loop indefinitely" because pcre_compile()
1274        was not correctly checking the subroutine for matching a non-empty string.
1275    
1276    14. The checks for overrunning compiling workspace could trigger after an
1277        overrun had occurred. This is a "should never occur" error, but it can be
1278        triggered by pathological patterns such as hundreds of nested parentheses.
1279        The checks now trigger 100 bytes before the end of the workspace.
1280    
1281    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1282    
1283    
1284    Version 8.01 19-Jan-2010
1285    ------------------------
1286    
1287    1.  If a pattern contained a conditional subpattern with only one branch (in
1288        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1289        computed the wrong minimum data length (which is of course zero for such
1290        subpatterns). This could cause incorrect "no match" results.
1291    
1292    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1293        the pattern is reset in the first branch, pcre_compile() failed with
1294        "internal error: code overflow at offset...". This happened only when
1295        the reset was to the original external option setting. (An optimization
1296        abstracts leading options settings into an external setting, which was the
1297        cause of this.)
1298    
1299    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1300        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1301        assertion pattern did not match (meaning that the assertion was true), it
1302        was incorrectly treated as false if the SKIP had been reached during the
1303        matching. This also applied to assertions used as conditions.
1304    
1305    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1306        assertion subpattern, including such a pattern used as a condition,
1307        unpredictable results occurred, instead of the error return
1308        PCRE_ERROR_DFA_UITEM.
1309    
1310    5.  The C++ GlobalReplace function was not working like Perl for the special
1311        situation when an empty string is matched. It now does the fancy magic
1312        stuff that is necessary.
1313    
1314    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1315        removed. (These were left over from very, very early versions of PCRE.)
1316    
1317    7.  Some cosmetic changes to the code to make life easier when compiling it
1318        as part of something else:
1319    
1320        (a) Change DEBUG to PCRE_DEBUG.
1321    
1322        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1323            called "current" as "current_branch", to prevent a collision with the
1324            Linux macro when compiled as a kernel module.
1325    
1326        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1327            prevent a collision with the Linux macro when compiled as a kernel
1328            module.
1329    
1330    8.  In pcre_compile() there are some checks for integer overflows that used to
1331        cast potentially large values to (double). This has been changed to that
1332        when building, a check for int64_t is made, and if it is found, it is used
1333        instead, thus avoiding the use of floating point arithmetic. (There is no
1334        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1335        double.
1336    
1337    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1338        2005 (difference between two addresses compared to an unsigned value).
1339    
1340    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1341        custom one, because of the following reported problem in Windows:
1342    
1343          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1344              under Win32.
1345          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1346              therefore missing the function definition.
1347          - The compiler thus generates a "C" signature for the test function.
1348          - The linker fails to find the "C" function.
1349          - PCRE fails to configure if asked to do so against libbz2.
1350    
1351    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1352        messages were output:
1353    
1354          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1355          rerunning libtoolize, to keep the correct libtool macros in-tree.
1356          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1357    
1358        I have done both of these things.
1359    
1360    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1361        most of the time, it *can* run out if it is given a pattern that contains a
1362        runaway infinite recursion. I updated the discussion in the pcrestack man
1363        page.
1364    
1365    13. Now that we have gone to the x.xx style of version numbers, the minor
1366        version may start with zero. Using 08 or 09 is a bad idea because users
1367        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1368        interpreted as invalid octal numbers. I've updated the previous comment in
1369        configure.ac, and also added a check that gives an error if 08 or 09 are
1370        used.
1371    
1372    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1373        causing partial matching to fail when the end of the subject matched \W
1374        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1375    
1376    15. There were some discrepancies between the declarations in pcre_internal.h
1377        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1378        their definitions. The declarations used "const uschar *" and the
1379        definitions used USPTR. Even though USPTR is normally defined as "const
1380        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1381        reported that: "This difference in casting confuses some C++ compilers, for
1382        example, SunCC recognizes above declarations as different functions and
1383        generates broken code for hbpcre." I have changed the declarations to use
1384        USPTR.
1385    
1386    16. GNU libtool is named differently on some systems. The autogen.sh script now
1387        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1388        (FreeBSD).
1389    
1390    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1391        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1392        comment: "Figure out how to create a longlong from a string: strtoll and
1393        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1394        instance, but it only takes 2 args instead of 3!"
1395    
1396    18. A subtle bug concerned with back references has been fixed by a change of
1397        specification, with a corresponding code fix. A pattern such as
1398        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1399        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1400        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1401        same bug. Such groups have to be quantified to be useful, or contained
1402        inside another quantified group. (If there's no repetition, the reference
1403        can never match.) The problem arises because, having left the group and
1404        moved on to the rest of the pattern, a later failure that backtracks into
1405        the group uses the captured value from the final iteration of the group
1406        rather than the correct earlier one. I have fixed this in PCRE by forcing
1407        any group that contains a reference to itself to be an atomic group; that
1408        is, there cannot be any backtracking into it once it has completed. This is
1409        similar to recursive and subroutine calls.
1410    
1411    
1412    Version 8.00 19-Oct-09
1413    ----------------------
1414    
1415    1.  The table for translating pcre_compile() error codes into POSIX error codes
1416        was out-of-date, and there was no check on the pcre_compile() error code
1417        being within the table. This could lead to an OK return being given in
1418        error.
1419    
1420    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1421        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1422        in a Windows environment.
1423    
1424    3.  The pcregrep --count option prints the count for each file even when it is
1425        zero, as does GNU grep. However, pcregrep was also printing all files when
1426        --files-with-matches was added. Now, when both options are given, it prints
1427        counts only for those files that have at least one match. (GNU grep just
1428        prints the file name in this circumstance, but including the count seems
1429        more useful - otherwise, why use --count?) Also ensured that the
1430        combination -clh just lists non-zero counts, with no names.
1431    
1432    4.  The long form of the pcregrep -F option was incorrectly implemented as
1433        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1434        but it seems right to fix it, and I didn't think it was worth preserving
1435        the old behaviour.
1436    
1437    5.  The command line items --regex=pattern and --regexp=pattern were not
1438        recognized by pcregrep, which required --regex pattern or --regexp pattern
1439        (with a space rather than an '='). The man page documented the '=' forms,
1440        which are compatible with GNU grep; these now work.
1441    
1442    6.  No libpcreposix.pc file was created for pkg-config; there was just
1443        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1444    
1445    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1446        when UCP support is not needed, by modifying the Python script that
1447        generates it from Unicode data files. This should not matter if the module
1448        is correctly used as a library, but I received one complaint about 50K of
1449        unwanted data. My guess is that the person linked everything into his
1450        program rather than using a library. Anyway, it does no harm.
1451    
1452    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1453        was a minimum greater than 1 for a wide character in a possessive
1454        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1455        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1456        character. Chaos in the form of incorrect output or a compiling loop could
1457        result.
1458    
1459    9.  The restrictions on what a pattern can contain when partial matching is
1460        requested for pcre_exec() have been removed. All patterns can now be
1461        partially matched by this function. In addition, if there are at least two
1462        slots in the offset vector, the offset of the earliest inspected character
1463        for the match and the offset of the end of the subject are set in them when
1464        PCRE_ERROR_PARTIAL is returned.
1465    
1466    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1467        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1468        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1469        and may be more useful for multi-segment matching.
1470    
1471    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1472        used to be given if ever the end of the subject was reached; now it is
1473        given only if matching could not proceed because another character was
1474        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1475        string "Z", which now yields "no match" instead of "partial match". In the
1476        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1477        final character ended with (*FAIL).
1478    
1479    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1480        if the pattern had a "must contain" character that was already found in the
1481        earlier partial match, unless partial matching was again requested. For
1482        example, with the pattern /dog.(body)?/, the "must contain" character is
1483        "g". If the first part-match was for the string "dog", restarting with
1484        "sbody" failed. This bug has been fixed.
1485    
1486    13. The string returned by pcre_dfa_exec() after a partial match has been
1487        changed so that it starts at the first inspected character rather than the
1488        first character of the match. This makes a difference only if the pattern
1489        starts with a lookbehind assertion or \b or \B (\K is not supported by
1490        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1491        matching functions compatible, and I think it's the right thing to do.
1492    
1493    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1494        so that the demonstration program is easily available in environments where
1495        PCRE has not been installed from source.
1496    
1497    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1498        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1499        library.
1500    
1501    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1502        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1503        is not the first non-POSIX option to be added. Clearly some people find
1504        these options useful.
1505    
1506    17. If a caller to the POSIX matching function regexec() passes a non-zero
1507        value for nmatch with a NULL value for pmatch, the value of
1508        nmatch is forced to zero.
1509    
1510    18. RunGrepTest did not have a test for the availability of the -u option of
1511        the diff command, as RunTest does. It now checks in the same way as
1512        RunTest, and also checks for the -b option.
1513    
1514    19. If an odd number of negated classes containing just a single character
1515        interposed, within parentheses, between a forward reference to a named
1516        subpattern and the definition of the subpattern, compilation crashed with
1517        an internal error, complaining that it could not find the referenced
1518        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1519        [The bug was that it was starting one character too far in when skipping
1520        over the character class, thus treating the ] as data rather than
1521        terminating the class. This meant it could skip too much.]
1522    
1523    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1524        /g option in pcretest when the pattern contains \K, which makes it possible
1525        to have an empty string match not at the start, even when the pattern is
1526        anchored. Updated pcretest and pcredemo to use this option.
1527    
1528    21. If the maximum number of capturing subpatterns in a recursion was greater
1529        than the maximum at the outer level, the higher number was returned, but
1530        with unset values at the outer level. The correct (outer level) value is
1531        now given.
1532    
1533    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1534        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1535        make it do so. The string so far is captured, making this feature
1536        compatible with Perl.
1537    
1538    23. The tests have been re-organized, adding tests 11 and 12, to make it
1539        possible to check the Perl 5.10 features against Perl 5.10.
1540    
1541    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1542        pattern matches a fixed length string. PCRE did not allow this; now it
1543        does. Neither allows recursion.
1544    
1545    25. I finally figured out how to implement a request to provide the minimum
1546        length of subject string that was needed in order to match a given pattern.
1547        (It was back references and recursion that I had previously got hung up
1548        on.) This code has now been added to pcre_study(); it finds a lower bound
1549        to the length of subject needed. It is not necessarily the greatest lower
1550        bound, but using it to avoid searching strings that are too short does give
1551        some useful speed-ups. The value is available to calling programs via
1552        pcre_fullinfo().
1553    
1554    26. While implementing 25, I discovered to my embarrassment that pcretest had
1555        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1556        study optimizations had never been tested with that matching function.
1557        Oops. What is worse, even when it was passed study data, there was a bug in
1558        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1559        were also very few tests of studied patterns with pcre_dfa_exec().
1560    
1561    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1562        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1563        on the other side of the coin, they are no longer allowed to have different
1564        names, because these cannot be distinguished in PCRE, and this has caused
1565        confusion. (This is a difference from Perl.)
1566    
1567    28. When duplicate subpattern names are present (necessarily with different
1568        numbers, as required by 27 above), and a test is made by name in a
1569        conditional pattern, either for a subpattern having been matched, or for
1570        recursion in such a pattern, all the associated numbered subpatterns are
1571        tested, and the overall condition is true if the condition is true for any
1572        one of them. This is the way Perl works, and is also more like the way
1573        testing by number works.
1574    
1575    
1576    Version 7.9 11-Apr-09
1577    ---------------------
1578    
1579    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1580        (pcretest), all targets were linked against these libraries. This included
1581        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1582        libraries. This caused unwanted dependencies to be created. This problem
1583        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1584        pcretest is linked with readline.
1585    
1586    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1587        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1588        moved outside it again, because FALSE and TRUE are already defined in AIX,
1589        but BOOL is not.
1590    
1591    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1592        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1593    
1594    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1595        hyphens) following file names and line numbers when outputting matching
1596        lines. This is not true; no spaces are inserted. I have also clarified the
1597        wording for the --colour (or --color) option.
1598    
1599    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1600        was not coloured; this is different to GNU grep, so I have changed it to be
1601        the same.
1602    
1603    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1604        each matching line was coloured. Now it goes on to look for further matches
1605        of any of the test patterns, which is the same behaviour as GNU grep.
1606    
1607    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1608        doesn't make sense to accept an empty string match in pcregrep, so I have
1609        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1610        seems to be how GNU grep behaves.
1611    
1612    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1613        start or after a newline", because the conditional assertion was not being
1614        correctly handled. The rule now is that both the assertion and what follows
1615        in the first alternative must satisfy the test.
1616    
1617    9.  If auto-callout was enabled in a pattern with a conditional group whose
1618        condition was an assertion, PCRE could crash during matching, both with
1619        pcre_exec() and pcre_dfa_exec().
1620    
1621    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1622        used for matching.
1623    
1624    11. Unicode property support in character classes was not working for
1625        characters (bytes) greater than 127 when not in UTF-8 mode.
1626    
1627    12. Added the -M command line option to pcretest.
1628    
1629    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1630    
1631    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1632    
1633    16. Added comments and documentation about mis-use of no_arg in the C++
1634        wrapper.
1635    
1636    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1637        from Martin Jerabek that uses macro names for all relevant character and
1638        string constants.
1639    
1640    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1641        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1642        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1643        these, but not everybody uses configure.
1644    
1645    19. A conditional group that had only one branch was not being correctly
1646        recognized as an item that could match an empty string. This meant that an
1647        enclosing group might also not be so recognized, causing infinite looping
1648        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1649        with the subject "ab", where knowledge that the repeated group can match
1650        nothing is needed in order to break the loop.
1651    
1652    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1653        exec(), but without supplying a callout function, matching went wrong.
1654    
1655    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1656        leak if the size of the offset vector was greater than 30. When the vector
1657        is smaller, the saved offsets during recursion go onto a local stack
1658        vector, but for larger vectors malloc() is used. It was failing to free
1659        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1660        error, in fact).
1661    
1662    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1663        heapframe that is used only when UTF-8 support is enabled. This caused no
1664        problem, but was untidy.
1665    
1666    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1667        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1668        included within another project.
1669    
1670    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1671        slightly modified by me:
1672    
1673          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1674              not building pcregrep.
1675    
1676          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1677              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1678    
1679    25. Forward references, both numeric and by name, in patterns that made use of
1680        duplicate group numbers, could behave incorrectly or give incorrect errors,
1681        because when scanning forward to find the reference group, PCRE was not
1682        taking into account the duplicate group numbers. A pattern such as
1683        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1684    
1685    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1686        the feature of a custom pointer more persuasive (as requested by a user).
1687    
1688    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1689        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1690        pre-defined.
1691    
1692    28. Added support for (*UTF8) at the start of a pattern.
1693    
1694    29. Arrange for flags added by the "release type" setting in CMake to be shown
1695        in the configuration summary.
1696    
1697    
1698    Version 7.8 05-Sep-08
1699    ---------------------
1700    
1701    1.  Replaced UCP searching code with optimized version as implemented for Ad
1702        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1703        stage table and inline lookup instead of a function, giving speed ups of 2
1704        to 5 times on some simple patterns that I tested. Permission was given to
1705        distribute the MultiStage2.py script that generates the tables (it's not in
1706        the tarball, but is in the Subversion repository).
1707    
1708    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1709        scripts.
1710    
1711    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1712        a group with a zero qualifier. The result of the study could be incorrect,
1713        or the function might crash, depending on the pattern.
1714    
1715    4.  Caseless matching was not working for non-ASCII characters in back
1716        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1717        It now works when Unicode Property Support is available.
1718    
1719    5.  In pcretest, an escape such as \x{de} in the data was always generating
1720        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1721        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1722        truncation.
1723    
1724    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1725    
1726    7.  Added two (int) casts to pcregrep when printing the difference of two
1727        pointers, in case they are 64-bit values.
1728    
1729    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1730        test 2 if it fails.
1731    
1732    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1733        and a #define of that name to empty if it is not externally set. This is to
1734        allow users of MSVC to set it if necessary.
1735    
1736    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1737        the convenience functions in the pcre_get.c source file.
1738    
1739    11. An option change at the start of a pattern that had top-level alternatives
1740        could cause overwriting and/or a crash. This command provoked a crash in
1741        some environments:
1742    
1743          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1744    
1745        This potential security problem was recorded as CVE-2008-2371.
1746    
1747    12. For a pattern where the match had to start at the beginning or immediately
1748        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1749        pcre_dfa_exec() could read past the end of the passed subject if there was
1750        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1751        pcretest so that it places the subject at the end of its malloc-ed buffer.
1752    
1753    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1754        exec() might read past the end of the data buffer in UTF-8 mode.
1755    
1756    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1757        the data contained the byte 0x85 as part of a UTF-8 character within its
1758        first line. This applied both to normal and DFA matching.
1759    
1760    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1761        /^[^d]*?$/8 failed to match "abc".
1762    
1763    16. Added a missing copyright notice to pcrecpp_internal.h.
1764    
1765    17. Make it more clear in the documentation that values returned from
1766        pcre_exec() in ovector are byte offsets, not character counts.
1767    
1768    18. Tidied a few places to stop certain compilers from issuing warnings.
1769    
1770    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1771        supplied by Stefan Weber. I made a further small update for 7.8 because
1772        there is a change of source arrangements: the pcre_searchfuncs.c module is
1773        replaced by pcre_ucd.c.
1774    
1775    
1776    Version 7.7 07-May-08
1777    ---------------------
1778    
1779    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1780        a string to a long long, pretend we don't even have a long long." This is
1781        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1782    
1783    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1784        pre-7.6 versions, which defined a global no_arg variable instead of putting
1785        it in the RE class. (See also #8 below.)
1786    
1787    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1788        Lopes.
1789    
1790    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1791    
1792        (1) The include/exclude patterns were being applied to the whole pathnames
1793            of files, instead of just to the final components.
1794    
1795        (2) If there was more than one level of directory, the subdirectories were
1796            skipped unless they satisfied the include/exclude conditions. This is
1797            inconsistent with GNU grep (and could even be seen as contrary to the
1798            pcregrep specification - which I improved to make it absolutely clear).
1799            The action now is always to scan all levels of directory, and just
1800            apply the include/exclude patterns to regular files.
1801    
1802    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1803        --exclude_dir in the tests to avoid scanning .svn directories.
1804    
1805    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1806        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1807        doesn't support NULs in patterns.
1808    
1809    7.  Added some missing "const"s to declarations of static tables in
1810        pcre_compile.c and pcre_dfa_exec.c.
1811    
1812    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1813        caused by fix #2  above. (Subsequently also a second patch to fix the
1814        first patch. And a third patch - this was a messy problem.)
1815    
1816    9.  Applied Craig's patch to remove the use of push_back().
1817    
1818    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1819        matching function regexec().
1820    
1821    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1822        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1823        references. PCRE supports relative numbers with this syntax (I don't think
1824        Oniguruma does).
1825    
1826    12. Previously, a group with a zero repeat such as (...){0} was completely
1827        omitted from the compiled regex. However, this means that if the group
1828        was called as a subroutine from elsewhere in the pattern, things went wrong
1829        (an internal error was given). Such groups are now left in the compiled
1830        pattern, with a new opcode that causes them to be skipped at execution
1831        time.
1832    
1833    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1834        to the way PCRE behaves:
1835    
1836        (a) A lone ] character is dis-allowed (Perl treats it as data).
1837    
1838        (b) A back reference to an unmatched subpattern matches an empty string
1839            (Perl fails the current match path).
1840    
1841        (c) A data ] in a character class must be notated as \] because if the
1842            first data character in a class is ], it defines an empty class. (In
1843            Perl it is not possible to have an empty class.) The empty class []
1844            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1845            The negative empty class [^] matches any one character, independently
1846            of the DOTALL setting.
1847    
1848    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1849        non-existent subpattern following a character class starting with ']' and
1850        containing () gave an internal compiling error instead of "reference to
1851        non-existent subpattern". Fortunately, when the pattern did exist, the
1852        compiled code was correct. (When scanning forwards to check for the
1853        existencd of the subpattern, it was treating the data ']' as terminating
1854        the class, so got the count wrong. When actually compiling, the reference
1855        was subsequently set up correctly.)
1856    
1857    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1858        it was being rejected as not supported by pcre_dfa_exec(), even though
1859        other assertions are supported. I have made pcre_dfa_exec() support
1860        (*FAIL).
1861    
1862    16. The implementation of 13c above involved the invention of a new opcode,
1863        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1864        cannot be changed at match time, I realized I could make a small
1865        improvement to matching performance by compiling OP_ALLANY instead of
1866        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1867        on the OP_ANY path.
1868    
1869    17. Compiling pcretest on Windows with readline support failed without the
1870        following two fixes: (1) Make the unistd.h include conditional on
1871        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1872    
1873    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1874        ncurses library to be included for pcretest when ReadLine support is
1875        requested, but also to allow for it to be overridden. This patch came from
1876        Daniel Bergström.
1877    
1878    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1879        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1880        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1881        spotting this.
1882    
1883    
1884    Version 7.6 28-Jan-08
1885    ---------------------
1886    
1887    1.  A character class containing a very large number of characters with
1888        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1889        overflow.
1890    
1891    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1892        HAVE_LONG_LONG is not defined.
1893    
1894    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1895        bring them up to date and include new features. This patch includes:
1896    
1897        - Fixed PH's badly added libz and libbz2 support.
1898        - Fixed a problem with static linking.
1899        - Added pcredemo. [But later removed - see 7 below.]
1900        - Fixed dftables problem and added an option.
1901        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1902            HAVE_LONG_LONG.
1903        - Added readline support for pcretest.
1904        - Added an listing of the option settings after cmake has run.
1905    
1906    4.  A user submitted a patch to Makefile that makes it easy to create
1907        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1908        Makefile.am that cause it to include this special target, without
1909        affecting anything else. Note that the same mingw target plus all
1910        the other distribution libraries and programs are now supported
1911        when configuring with CMake (see 6 below) instead of with
1912        Configure/Make.
1913    
1914    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1915        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1916        exported in the Windows port". It has not yet been confirmed that the patch
1917        solves the problem, but it does no harm.
1918    
1919    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1920        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1921        with CMake, and also correct the comment about stack recursion.
1922    
1923    7.  Remove the automatic building of pcredemo from the ./configure system and
1924        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1925        of a program that users should build themselves after PCRE is installed, so
1926        building it automatically is not really right. What is more, it gave
1927        trouble in some build environments.
1928    
1929    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1930    
1931    
1932    Version 7.5 10-Jan-08
1933    ---------------------
1934    
1935    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1936        values in parens when parsing an RE using the C++ wrapper."
1937    
1938    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1939        Characters greater than 255 were excluded from the class instead of being
1940        included.
1941    
1942    3.  The same bug as (2) above applied to negated POSIX classes such as
1943        [:^space:].
1944    
1945    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1946        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1947        I have changed it.
1948    
1949    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1950        first named subpattern) and a construct such as (?&a) would reference the
1951        first named subpattern whose name started with "a" (in other words, the
1952        length check was missing). Both these problems are fixed. "Subpattern name
1953        expected" is now given for (?&) (a zero-length name), and this patch also
1954        makes it give the same error for \k'' (previously it complained that that
1955        was a reference to a non-existent subpattern).
1956    
1957    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1958        this is right because (?- can be followed by option settings as well as by
1959        digits. I have, however, made the messages clearer.
1960    
1961    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1962        than the number used in the conditional) now cause a compile-time error.
1963        This is actually not compatible with Perl, which accepts such patterns, but
1964        treats the conditional as always being FALSE (as PCRE used to), but it
1965        seems to me that giving a diagnostic is better.
1966    
1967    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1968        and messages.
1969    
1970    9.  Fix two occurrences of "backslash" in comments that should have been
1971        "backspace".
1972    
1973    10. Remove two redundant lines of code that can never be obeyed (their function
1974        was moved elsewhere).
1975    
1976    11. The program that makes PCRE's Unicode character property table had a bug
1977        which caused it to generate incorrect table entries for sequences of
1978        characters that have the same character type, but are in different scripts.
1979        It amalgamated them into a single range, with the script of the first of
1980        them. In other words, some characters were in the wrong script. There were
1981        thirteen such cases, affecting characters in the following ranges:
1982    
1983          U+002b0 - U+002c1
1984          U+0060c - U+0060d
1985          U+0061e - U+00612
1986          U+0064b - U+0065e
1987          U+0074d - U+0076d
1988          U+01800 - U+01805
1989          U+01d00 - U+01d77
1990          U+01d9b - U+01dbf
1991          U+0200b - U+0200f
1992          U+030fc - U+030fe
1993          U+03260 - U+0327f
1994          U+0fb46 - U+0fbb1
1995          U+10450 - U+1049d
1996    
1997    12. The -o option (show only the matching part of a line) for pcregrep was not
1998        compatible with GNU grep in that, if there was more than one match in a
1999        line, it showed only the first of them. It now behaves in the same way as
2000        GNU grep.
2001    
2002    13. If the -o and -v options were combined for pcregrep, it printed a blank
2003        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2004        does the same. The return code can be used to tell if there were any
2005        non-matching lines.
2006    
2007    14. Added --file-offsets and --line-offsets to pcregrep.
2008    
2009    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2010        infinitely looping recursion. The bug was that positive lookaheads were not
2011        being skipped when checking for a possible empty match (negative lookaheads
2012        and both kinds of lookbehind were skipped).
2013    
2014    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2015        inclusion of <windows.h> to before rather than after the definition of
2016        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2017    
2018    17. Specifying a possessive quantifier with a specific limit for a Unicode
2019        character property caused pcre_compile() to compile bad code, which led at
2020        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2021        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2022        caused the error; without that there was no problem.
2023    
2024    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2025    
2026    19. Added --enable-pcretest-libreadline.
2027    
2028    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2029        RE::GlobalReplace(). As a result, the number of replacements returned was
2030        double what it should be. I removed one of the increments, but Craig sent a
2031        later patch that removed the other one (the right fix) and added unit tests
2032        that check the return values (which was not done before).
2033    
2034    21. Several CMake things:
2035    
2036        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2037            the names libpcre and libpcreposix, not just pcre and pcreposix.
2038    
2039        (2) The above change means that pcretest and pcregrep are now correctly
2040            linked with the newly-built libraries, not previously installed ones.
2041    
2042        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2043    
2044    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2045        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2046        UTF-8 newline character). The key issue is that the pattern starts .*;
2047        this means that the match must be either at the beginning, or after a
2048        newline. The bug was in the code for advancing after a failed match and
2049        checking that the new position followed a newline. It was not taking
2050        account of UTF-8 characters correctly.
2051    
2052    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2053        character classes. PCRE was not treating the sequence [:...:] as a
2054        character class unless the ... were all letters. Perl, however, seems to
2055        allow any characters between [: and :], though of course it rejects as
2056        unknown any "names" that contain non-letters, because all the known class
2057        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2058        for example, whereas PCRE did not - it did not recognize a POSIX character
2059        class. This seemed a bit dangerous, so the code has been changed to be
2060        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2061        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2062        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2063        Perl does, and where it didn't before.
2064    
2065    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2066        Windows environments %n is disabled by default.
2067    
2068    
2069    Version 7.4 21-Sep-07
2070    ---------------------
2071    
2072    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2073        means that a class such as [\s] counted as "explicit reference to CR or
2074        LF". That isn't really right - the whole point of the change was to try to
2075        help when there was an actual mention of one of the two characters. So now
2076        the change happens only if \r or \n (or a literal CR or LF) character is
2077        encountered.
2078    
2079    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2080        of both had grown to the point where there were only 3 bits left.
2081        Fortunately, there was spare space in the data structure, and so I have
2082        moved the internal flags into a new 16-bit field to free up more option
2083        bits.
2084    
2085    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2086        but did not set the internal JCHANGED flag - either of these is enough to
2087        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2088        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2089        start sets both bits.
2090    
2091    4.  Added options (at build time, compile time, exec time) to change \R from
2092        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2093    
2094    5.  doc/pcresyntax.html was missing from the distribution.
2095    
2096    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2097        compatibility, even though it is no longer used.
2098    
2099    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2100        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2101        windows.h file is present (where different names are used). [This was
2102        reversed later after testing - see 16 below.]
2103    
2104    8.  Changed all #include <config.h> to #include "config.h". There were also
2105        some further <pcre.h> cases that I changed to "pcre.h".
2106    
2107    9.  When pcregrep was used with the --colour option, it missed the line ending
2108        sequence off the lines that it output.
2109    
2110    10. It was pointed out to me that arrays of string pointers cause lots of
2111        relocations when a shared library is dynamically loaded. A technique of
2112        using a single long string with a table of offsets can drastically reduce
2113        these. I have refactored PCRE in four places to do this. The result is
2114        dramatic:
2115    
2116          Originally:                          290
2117          After changing UCP table:            187
2118          After changing error message table:   43
2119          After changing table of "verbs"       36
2120          After changing table of Posix names   22
2121    
2122        Thanks to the folks working on Gregex for glib for this insight.
2123    
2124    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2125        unicode-properties was also set.
2126    
2127    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2128    
2129    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2130        checked only for CRLF.
2131    
2132    14. Added casts to pcretest.c to avoid compiler warnings.
2133    
2134    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2135    
2136    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2137        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2138        entirely. This removes changes made in 7 above.
2139    
2140    17. The CMake files have been updated, and there is now more information about
2141        building with CMake in the NON-UNIX-USE document.
2142    
2143    
2144    Version 7.3 28-Aug-07
2145    ---------------------
2146    
2147     1. In the rejigging of the build system that eventually resulted in 7.1, the
2148        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2149        brackets there is not right, since it causes compilers to look for an
2150        installed pcre.h, not the version that is in the source that is being
2151        compiled (which of course may be different). I have changed it back to:
2152    
2153          #include "pcre.h"
2154    
2155        I have a vague recollection that the change was concerned with compiling in
2156        different directories, but in the new build system, that is taken care of
2157        by the VPATH setting the Makefile.
2158    
2159     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2160        when the subject happened to end in the byte 0x85 (e.g. if the last
2161        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2162        characters but of course it shouldn't be taken as a newline when it is part
2163        of another character. The bug was that, for an unlimited repeat of . in
2164        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2165        characters when looking for a newline.
2166    
2167     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2168    
2169     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2170        in debug output.
2171    
2172     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2173        long printing in the pcrecpp unittest when running under MinGW.
2174    
2175     6. ESC_K was left out of the EBCDIC table.
2176    
2177     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2178        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2179        limit also applies to "virtual nesting" when a pattern is recursive, and in
2180        this case 1000 isn't so big. I have been able to remove this limit at the
2181        expense of backing off one optimization in certain circumstances. Normally,
2182        when pcre_exec() would call its internal match() function recursively and
2183        immediately return the result unconditionally, it uses a "tail recursion"
2184        feature to save stack. However, when a subpattern that can match an empty
2185        string has an unlimited repetition quantifier, it no longer makes this
2186        optimization. That gives it a stack frame in which to save the data for
2187        checking that an empty string has been matched. Previously this was taken
2188        from the 1000-entry workspace that had been reserved. So now there is no
2189        explicit limit, but more stack is used.
2190    
2191     8. Applied Daniel's patches to solve problems with the import/export magic
2192        syntax that is required for Windows, and which was going wrong for the
2193        pcreposix and pcrecpp parts of the library. These were overlooked when this
2194        problem was solved for the main library.
2195    
2196     9. There were some crude static tests to avoid integer overflow when computing
2197        the size of patterns that contain repeated groups with explicit upper
2198        limits. As the maximum quantifier is 65535, the maximum group length was
2199        set at 30,000 so that the product of these two numbers did not overflow a
2200        32-bit integer. However, it turns out that people want to use groups that
2201        are longer than 30,000 bytes (though not repeat them that many times).
2202        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2203        made it possible to implement the integer overflow checks in a much more
2204        dynamic way, which I have now done. The artificial limitation on group
2205        length has been removed - we now have only the limit on the total length of
2206        the compiled pattern, which depends on the LINK_SIZE setting.
2207    
2208    10. Fixed a bug in the documentation for get/copy named substring when
2209        duplicate names are permitted. If none of the named substrings are set, the
2210        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2211        empty string.
2212    
2213    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2214        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2215        because the ] is interpreted as the first data character and the
2216        terminating ] is not found. PCRE has been made compatible with Perl in this
2217        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2218        cause memory overwriting.
2219    
2220    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2221        string has been matched (to stop an infinite loop). It was not recognizing
2222        a conditional subpattern that could match an empty string if that
2223        subpattern was within another subpattern. For example, it looped when
2224        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2225        condition was not nested. This bug has been fixed.
2226    
2227    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2228        past the start of the subject in the presence of bytes with the top bit
2229        set, for example "\x8aBCD".
2230    
2231    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2232        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2233    
2234    14. Optimized (?!) to (*FAIL).
2235    
2236    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2237        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2238        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2239        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2240        does: it's just the validity check that is more restrictive.
2241    
2242    16. Inserted checks for integer overflows during escape sequence (backslash)
2243        processing, and also fixed erroneous offset values for syntax errors during
2244        backslash processing.
2245    
2246    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2247        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2248    
2249    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2250        caused an overrun.
2251    
2252    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2253        something other than just ASCII characters) inside a group that had an
2254        unlimited repeat caused a loop at compile time (while checking to see
2255        whether the group could match an empty string).
2256    
2257    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2258        [\P{Any}] did so. (Error in the code for printing property names.)
2259    
2260    21. An orphan \E inside a character class could cause a crash.
2261    
2262    22. A repeated capturing bracket such as (A)? could cause a wild memory
2263        reference during compilation.
2264    
2265    23. There are several functions in pcre_compile() that scan along a compiled
2266        expression for various reasons (e.g. to see if it's fixed length for look
2267        behind). There were bugs in these functions when a repeated \p or \P was
2268        present in the pattern. These operators have additional parameters compared
2269        with \d, etc, and these were not being taken into account when moving along
2270        the compiled data. Specifically:
2271    
2272        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2273            length.
2274    
2275        (b) An item such as \pL+ within a repeated group could cause crashes or
2276            loops.
2277    
2278        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2279            "reference to non-existent subpattern" error.
2280    
2281        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2282    
2283    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2284        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2285    
2286    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2287    
2288    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2289        character were causing crashes (broken optimization).
2290    
2291    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2292        \p or \P) caused a compile-time loop.
2293    
2294    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2295        break. For example, the unstudied pattern [\r\n]A does not match the string
2296        "\r\nA" because change 7.0/46 below moves the current point on by two
2297        characters after failing to match at the start. However, the pattern \nA
2298        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2299        the same is true. There doesn't seem any very clean way out of this, but
2300        what I have chosen to do makes the common cases work: PCRE now takes note
2301        of whether there can be an explicit match for \r or \n anywhere in the
2302        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2303        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2304        pattern has explicit CR or LF references.
2305    
2306    29. Added (*CR) etc for changing newline setting at start of pattern.
2307    
2308    
2309    Version 7.2 19-Jun-07
2310    ---------------------
2311    
2312     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2313        which is apparently normally available under Windows.
2314    
2315     2. Re-jig the pcregrep tests with different newline settings in an attempt
2316        to make them independent of the local environment's newline setting.
2317    
2318     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2319    
2320     4. Some of the "internals" tests were previously cut out when the link size
2321        was not 2, because the output contained actual offsets. The recent new
2322        "Z" feature of pcretest means that these can be cut out, making the tests
2323        usable with all link sizes.
2324    
2325     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2326        stack recursion. This gives a massive performance boost under BSD, but just
2327        a small improvement under Linux. However, it saves one field in the frame
2328        in all cases.
2329    
2330     6. Added more features from the forthcoming Perl 5.10:
2331    
2332        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2333            recursion call. It refers to the nth most recently opened parentheses.
2334    
2335        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2336            to be opened parentheses.
2337    
2338        (c) Conditions that refer to capturing parentheses can be specified
2339            relatively, for example, (?(-2)... or (?(+3)...
2340    
2341        (d) \K resets the start of the current match so that everything before
2342            is not part of it.
2343    
2344        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2345    
2346        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2347            reference syntax.
2348    
2349        (g) (?| introduces a group in which the numbering of parentheses in each
2350            alternative starts with the same number.
2351    
2352        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2353    
2354     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2355        PCRE_INFO_JCHANGED.
2356    
2357     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2358        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2359        for detecting groups that can match an empty string.
2360    
2361     9. A pattern with a very large number of alternatives (more than several
2362        hundred) was running out of internal workspace during the pre-compile
2363        phase, where pcre_compile() figures out how much memory will be needed. A
2364        bit of new cunning has reduced the workspace needed for groups with
2365        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2366        workspace instead of running out of the 4096 that are available.
2367    
2368    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2369    
2370    11. Applied patch from Google to remove an optimization that didn't quite work.
2371        The report of the bug said:
2372    
2373          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2374          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2375          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2376    
2377    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2378        it matched the wrong number of bytes.
2379    
2380    
2381    Version 7.1 24-Apr-07
2382  ---------------------  ---------------------
2383    
2384   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
# Line 33  Version 7.1 12-Mar-07 Line 2410  Version 7.1 12-Mar-07
2410   5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan   5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2411      Weber: (1) pcre_internal.h was missing some function renames; (2) updated      Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2412      makevp.bat for the current PCRE, using the additional files      makevp.bat for the current PCRE, using the additional files
2413      makevp-compile.txt, makevp-linklib.txt, and pcregexp.pas.      makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2414    
2415   6. A Windows user reported a minor discrepancy with test 2, which turned out   6. A Windows user reported a minor discrepancy with test 2, which turned out
2416      to be caused by a trailing space on an input line that had got lost in his      to be caused by a trailing space on an input line that had got lost in his
# Line 116  Version 7.1 12-Mar-07 Line 2493  Version 7.1 12-Mar-07
2493  14. If erroroffset was passed as NULL to pcre_compile, it provoked a  14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2494      segmentation fault instead of returning the appropriate error message.      segmentation fault instead of returning the appropriate error message.
2495    
2496    15. In multiline mode when the newline sequence was set to "any", the pattern
2497        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2498        This doesn't seem right; it now treats the CRLF combination as the line
2499        ending, and so does not match in that case. It's only a pattern such as ^$
2500        that would hit this one: something like ^ABC$ would have failed after \r
2501        and then tried again after \r\n.
2502    
2503    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2504        in an attempt to make files that differ only in their line terminators
2505        compare equal. This works on Linux.
2506    
2507    17. Under certain error circumstances pcregrep might try to free random memory
2508        as it exited. This is now fixed, thanks to valgrind.
2509    
2510    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2511        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2512        was because its rules for how to advance for /g after matching an empty
2513        string at the end of a line did not allow for this case. They now check for
2514        it specially.
2515    
2516    20. pcretest is supposed to handle patterns and data of any length, by
2517        extending its buffers when necessary. It was getting this wrong when the
2518        buffer for a data line had to be extended.
2519    
2520    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2521        CRLF as a newline sequence.
2522    
2523    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2524        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2525        I have nevertheless tidied it up.
2526    
2527    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2528    
2529    24. Added a man page for pcre-config.
2530    
2531    
2532  Version 7.0 19-Dec-06  Version 7.0 19-Dec-06
2533  ---------------------  ---------------------

Legend:
Removed from v.127  
changed lines
  Added in v.1272

  ViewVC Help
Powered by ViewVC 1.1.5