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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.193  
changed lines
  Added in v.1271

  ViewVC Help
Powered by ViewVC 1.1.5