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

Diff of /code/trunk/ChangeLog

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

revision 98 by ph10, Mon Mar 5 15:55:02 2007 UTC revision 1273 by ph10, Fri Mar 8 10:25:02 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.1 05-Mar-07  Version 8.33 xx-xxxx-201x
5    -------------------------
6    
7    1.  Added 'U' to some constants that are compared to unsigned integers, to
8        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
9        variables that are added to signed variables, to ensure the result is
10        signed and can be negated.
11    
12    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
13        CMake config files.
14    
15    3.  Revise the creation of config.h.generic so that all boolean macros are
16        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
17        overriding via -D on the command line possible.
18    
19    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
20        to unsigned int is reported to make a quite noticeable speed difference in
21        a specific Windows environment. Testing on Linux did also appear to show
22        some benefit (and it is clearly not harmful). Also fixed the definition of
23        Xop which should be unsigned.
24    
25    5.  Related to (4), changing the definition of the intermediate variable cc
26        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
27        improvements.
28    
29    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
30        unnecessary spaces.
31    
32    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
33        and later.
34    
35    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
36    
37    9.  Optimizing fast_forward_start_bits in JIT.
38    
39    10. Adding experimental support for callouts in JIT, and fixing some
40        issues revealed during this work. Namely:
41    
42        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
43    
44        (b) Minimum length was not checked before the matching is started.
45    
46    11. The value of capture_last that is passed to callouts was incorrect in some
47        cases when there was a capture on one path that was subsequently abandoned
48        after a backtrack. Also, the capture_last value is now reset after a
49        recursion, since all captures are also reset in this case.
50    
51    12. The interpreter no longer returns the "too many substrings" error in the
52        case when an overflowing capture is in a branch that is subsequently
53        abandoned after a backtrack.
54    
55    13. In the pathological case when an offset vector of size 2 is used, pcretest
56        now prints out the matched string after a yield of 0 or 1.
57    
58    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
59        Only supported by the JIT compiler at the moment.
60    
61    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
62    
63    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
64        offset of the starting point of the matching process, provided the offsets
65        vector is large enough.
66    
67    17. The \A escape now records a lookbehind value of 1, though its execution
68        does not actually inspect the previous character. This is to ensure that,
69        in partial multi-segment matching, at least one character from the old
70        segment is retained when a new segment is processed. Otherwise, if there
71        are no lookbehinds in the pattern, \A might match incorrectly at the start
72        of a new segment.
73    
74    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
75    
76    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
77        gives some modest performance improvement in 8-bit mode.
78    
79    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
80        be expressed in certain programming languages using Universal Character
81        Names.
82    
83    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
84        which points out that "non characters" are not "characters that may not
85        appear in Unicode strings" but rather "characters that are reserved for
86        internal use and have only local meaning".
87    
88    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
89        there was a conditional group that depended on an assertion, if the
90        assertion was false, the callout that immediately followed the alternation
91        in the condition was skipped when pcre_exec() was used for matching.
92    
93    23. Allow an explicit callout to be inserted before an assertion that is the
94        condition for a conditional group, for compatibility with automatic
95        callouts, which always insert a callout at this point.
96    
97    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
98        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
99    
100    25. (*PRUNE) is now supported by the JIT compiler.
101    
102    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
103    
104    
105    Version 8.32 30-November-2012
106    -----------------------------
107    
108    1.  Improved JIT compiler optimizations for first character search and single
109        character iterators.
110    
111    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
112        Patch by Daniel Richard G.
113    
114    3.  Single character iterator optimizations in the JIT compiler.
115    
116    4.  Improved JIT compiler optimizations for character ranges.
117    
118    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
119        Reported by Giuseppe D'Angelo.
120    
121    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
122        of a line, was being set incorrectly in cases where .* appeared inside
123        atomic brackets at the start of a pattern, or where there was a subsequent
124        *PRUNE or *SKIP.
125    
126    7.  Improved instruction cache flush for POWER/PowerPC.
127        Patch by Daniel Richard G.
128    
129    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
130        grep:
131    
132        (a) There is now no limit to the number of patterns to be matched.
133    
134        (b) An error is given if a pattern is too long.
135    
136        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
137            are now supported.
138    
139        (d) --exclude-from and --include-from (multiple use) have been added.
140    
141        (e) Exclusions and inclusions now apply to all files and directories, not
142            just to those obtained from scanning a directory recursively.
143    
144        (f) Multiple uses of -f and --file-list are now supported.
145    
146        (g) In a Windows environment, the default for -d has been changed from
147            "read" (the GNU grep default) to "skip", because otherwise the presence
148            of a directory in the file list provokes an error.
149    
150        (h) The documentation has been revised and clarified in places.
151    
152    9.  Improve the matching speed of capturing brackets.
153    
154    10. Changed the meaning of \X so that it now matches a Unicode extended
155        grapheme cluster.
156    
157    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
158        out POSIX threads when JIT support is configured.
159    
160    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
161    
162    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
163        structure could go wrong in environments where size_t is not the same size
164        as int.
165    
166    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
167        set.
168    
169    15. The EBCDIC support had decayed; later updates to the code had included
170        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
171        general tidy up of EBCDIC-related issues, and the documentation was also
172        not quite right. There is now a test that can be run on ASCII systems to
173        check some of the EBCDIC-related things (but is it not a full test).
174    
175    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
176        in a small tidy to the code.
177    
178    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
179    
180    18. If the --only-matching (-o) option in pcregrep is specified multiple
181        times, each one causes appropriate output. For example, -o1 -o2 outputs the
182        substrings matched by the 1st and 2nd capturing parentheses. A separating
183        string can be specified by --om-separator (default empty).
184    
185    19. Improving the first n character searches.
186    
187    20. Turn case lists for horizontal and vertical white space into macros so that
188        they are defined only once.
189    
190    21. This set of changes together give more compatible Unicode case-folding
191        behaviour for characters that have more than one other case when UCP
192        support is available.
193    
194        (a) The Unicode property table now has offsets into a new table of sets of
195            three or more characters that are case-equivalent. The MultiStage2.py
196            script that generates these tables (the pcre_ucd.c file) now scans
197            CaseFolding.txt instead of UnicodeData.txt for character case
198            information.
199    
200        (b) The code for adding characters or ranges of characters to a character
201            class has been abstracted into a generalized function that also handles
202            case-independence. In UTF-mode with UCP support, this uses the new data
203            to handle characters with more than one other case.
204    
205        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
206            whose other case is greater than 256 are now correctly matched
207            caselessly. Previously, the high codepoint matched the low one, but not
208            vice versa.
209    
210        (d) The processing of \h, \H, \v, and \ in character classes now makes use
211            of the new class addition function, using character lists defined as
212            macros alongside the case definitions of 20 above.
213    
214        (e) Caseless back references now work with characters that have more than
215            one other case.
216    
217        (f) General caseless matching of characters with more than one other case
218            is supported.
219    
220    22. Unicode character properties were updated from Unicode 6.2.0
221    
222    23. Improved CMake support under Windows. Patch by Daniel Richard G.
223    
224    24. Add support for 32-bit character strings, and UTF-32
225    
226    25. Major JIT compiler update (code refactoring and bugfixing).
227        Experimental Sparc 32 support is added.
228    
229    26. Applied a modified version of Daniel Richard G's patch to create
230        pcre.h.generic and config.h.generic by "make" instead of in the
231        PrepareRelease script.
232    
233    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
234        pcre_compile.c when checking for a zero character.
235    
236    28. Introducing a native interface for JIT. Through this interface, the compiled
237        machine code can be directly executed. The purpose of this interface is to
238        provide fast pattern matching, so several sanity checks are not performed.
239        However, feature tests are still performed. The new interface provides
240        1.4x speedup compared to the old one.
241    
242    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
243        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
244        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
245    
246    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
247        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
248        the "old" RFC 2279). Instead, it ended up passing a negative length to
249        pcre_exec().
250    
251    31. Add support for GCC's visibility feature to hide internal functions.
252    
253    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
254        "unknown -C option" after outputting 0 or 1.
255    
256    33. There is now support for generating a code coverage report for the test
257        suite in environments where gcc is the compiler and lcov is installed. This
258        is mainly for the benefit of the developers.
259    
260    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
261        unaddressable using valgrind annotations, allowing valgrind to detect
262        invalid memory accesses. This is mainly for the benefit of the developers.
263    
264    25. (*UTF) can now be used to start a pattern in any of the three libraries.
265    
266    26. Give configure error if --enable-cpp but no C++ compiler found.
267    
268    
269    Version 8.31 06-July-2012
270    -------------------------
271    
272    1.  Fixing a wrong JIT test case and some compiler warnings.
273    
274    2.  Removed a bashism from the RunTest script.
275    
276    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
277        to unsigned type, result still unsigned" that was given by an MS compiler
278        on encountering the code "-sizeof(xxx)".
279    
280    4.  Partial matching support is added to the JIT compiler.
281    
282    5.  Fixed several bugs concerned with partial matching of items that consist
283        of more than one character:
284    
285        (a) /^(..)\1/ did not partially match "aba" because checking references was
286            done on an "all or nothing" basis. This also applied to repeated
287            references.
288    
289        (b) \R did not give a hard partial match if \r was found at the end of the
290            subject.
291    
292        (c) \X did not give a hard partial match after matching one or more
293            characters at the end of the subject.
294    
295        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
296            a partial match for the string "\r".
297    
298        (e) When newline was set to CRLF, the metacharacter "." did not recognize
299            a partial match for a CR character at the end of the subject string.
300    
301    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
302        running pcretest, the text "(JIT)" added to the output whenever JIT is
303        actually used to run the match.
304    
305    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
306        or /S+[+] with a digit between 1 and 7.
307    
308    8.  OP_NOT now supports any UTF character not just single-byte ones.
309    
310    9.  (*MARK) control verb is now supported by the JIT compiler.
311    
312    10. The command "./RunTest list" lists the available tests without actually
313        running any of them. (Because I keep forgetting what they all are.)
314    
315    11. Add PCRE_INFO_MAXLOOKBEHIND.
316    
317    12. Applied a (slightly modified) user-supplied patch that improves performance
318        when the heap is used for recursion (compiled with --disable-stack-for-
319        recursion). Instead of malloc and free for each heap frame each time a
320        logical recursion happens, frames are retained on a chain and re-used where
321        possible. This sometimes gives as much as 30% improvement.
322    
323    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
324        call.
325    
326    14. As documented, (*COMMIT) is now confined to within a positive assertion.
327    
328    15. It is now possible to link pcretest with libedit as an alternative to
329        libreadline.
330    
331    16. (*COMMIT) control verb is now supported by the JIT compiler.
332    
333    17. The Unicode data tables have been updated to Unicode 6.1.0.
334    
335    18. Added --file-list option to pcregrep.
336    
337    19. Added binary file support to pcregrep, including the -a, --binary-files,
338        -I, and --text options.
339    
340    20. The madvise function is renamed for posix_madvise for QNX compatibility
341        reasons. Fixed by Giuseppe D'Angelo.
342    
343    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
344    
345    22. Changed the option for creating symbolic links for 16-bit man pages from
346        -s to -sf so that re-installing does not cause issues.
347    
348    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
349    
350    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
351        matches in certain environments (the workspace was not being correctly
352        retained). Also added to pcre_dfa_exec() a simple plausibility check on
353        some of the workspace data at the beginning of a restart.
354    
355    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
356        was not doing so when it should - probably a typo introduced by SVN 528
357        (change 8.10/14).
358    
359    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
360        \w+ when the character tables indicated that \x{c4} was a word character.
361        There were several related cases, all because the tests for doing a table
362        lookup were testing for characters less than 127 instead of 255.
363    
364    27. If a pattern contains capturing parentheses that are not used in a match,
365        their slots in the ovector are set to -1. For those that are higher than
366        any matched groups, this happens at the end of processing. In the case when
367        there were back references that the ovector was too small to contain
368        (causing temporary malloc'd memory to be used during matching), and the
369        highest capturing number was not used, memory off the end of the ovector
370        was incorrectly being set to -1. (It was using the size of the temporary
371        memory instead of the true size.)
372    
373    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
374        ovector size, it uses memory at the end of the block that it has got.
375    
376    29. Check for an overlong MARK name and give an error at compile time. The
377        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
378    
379    30. JIT compiler update.
380    
381    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
382        Rill for the patch.
383    
384    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
385    
386    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
387    
388    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
389        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
390        was enabled.
391    
392    35. Improve JIT code generation for greedy plus quantifier.
393    
394    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
395        1 to "aa" instead of to an empty string. The bug affected repeated groups
396        that could potentially match an empty string.
397    
398    37. Optimizing single character iterators in JIT.
399    
400    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
401        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
402        codepoints that are too big for the mode are faulted, and in a UTF mode,
403        disallowed codepoints are also faulted.
404    
405    39. If PCRE was compiled with UTF support, in three places in the DFA
406        matcher there was code that should only have been obeyed in UTF mode, but
407        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
408        processing when bytes with values greater than 127 were present. In 16-bit
409        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
410        both cases the values are those that cannot be the first data item in a UTF
411        character. The three items that might have provoked this were recursions,
412        possessively repeated groups, and atomic groups.
413    
414    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
415        and pcregrep, because some OS require shared objects to be explicitly
416        passed to ld, causing the link step to fail if they are not.
417    
418    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
419        mode, patterns that started with \h* or \R* might be incorrectly matched.
420    
421    
422    Version 8.30 04-February-2012
423    -----------------------------
424    
425    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
426        name is defined in ctype.h.
427    
428    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
429        only in quite long subpatterns.
430    
431    3.  Removed the function pcre_info(), which has been obsolete and deprecated
432        since it was replaced by pcre_fullinfo() in February 2000.
433    
434    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
435        match a (*MARK), and the match failed at the start of the subject, a
436        reference to memory before the start of the subject could occur. This bug
437        was introduced by fix 17 of release 8.21.
438    
439    5.  A reference to an unset group with zero minimum repetition was giving
440        totally wrong answers (in non-JavaScript-compatibility mode). For example,
441        /(another)?(\1?)test/ matched against "hello world test". This bug was
442        introduced in release 8.13.
443    
444    6.  Add support for 16-bit character strings (a large amount of work involving
445        many changes and refactorings).
446    
447    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
448        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
449        from a file.
450    
451    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
452        rounding is not applied in this particular case).
453    
454    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
455        if they appear, or are escaped, in patterns.
456    
457    10. Get rid of a number of -Wunused-but-set-variable warnings.
458    
459    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
460        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
461        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
462        also returns the mark "x". This bug applied to capturing parentheses,
463        non-capturing parentheses, and atomic parentheses. It also applied to some
464        assertions.
465    
466    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
467        information out of configure.ac instead of relying on pcre.h.generic, which
468        is not stored in the repository.
469    
470    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
471        -lreadline.
472    
473    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
474    
475    15. Applied Graycode's patch to put the top-level frame on the stack rather
476        than the heap when not using the stack for recursion. This gives a
477        performance improvement in many cases when recursion is not deep.
478    
479    16. Experimental code added to "pcretest -C" to output the stack frame size.
480    
481    
482    Version 8.21 12-Dec-2011
483    ------------------------
484    
485    1.  Updating the JIT compiler.
486    
487    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
488        are added as well.
489    
490    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
491        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
492        calling _pcre_jit_exec. Some extra comments are added.
493    
494    4.  (*MARK) settings inside atomic groups that do not contain any capturing
495        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
496        was introduced by change 18 for 8.20.
497    
498    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
499        ECMA-262 standard.
500    
501    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
502        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
503        This bug was probably introduced by change 9 of 8.13.
504    
505    7.  While fixing 6 above, I noticed that a number of other items were being
506        incorrectly rejected as "not fixed length". This arose partly because newer
507        opcodes had not been added to the fixed-length checking code. I have (a)
508        corrected the bug and added tests for these items, and (b) arranged for an
509        error to occur if an unknown opcode is encountered while checking for fixed
510        length instead of just assuming "not fixed length". The items that were
511        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
512        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
513        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
514    
515    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
516        being incorrectly compiled and would have given unpredicatble results.
517    
518    9.  A possessively repeated subpattern with minimum repeat count greater than
519        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
520        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
521        the first (A) could occur when it should not.
522    
523    10. Add a cast and remove a redundant test from the code.
524    
525    11. JIT should use pcre_malloc/pcre_free for allocation.
526    
527    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
528        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
529        is anything other than /usr, -L is still shown).
530    
531    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
532    
533    14. Perl does not support \N without a following name in a [] class; PCRE now
534        also gives an error.
535    
536    15. If a forward reference was repeated with an upper limit of around 2000,
537        it caused the error "internal error: overran compiling workspace". The
538        maximum number of forward references (including repeats) was limited by the
539        internal workspace, and dependent on the LINK_SIZE. The code has been
540        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
541        the default depends on LINK_SIZE. There is a new upper limit (for safety)
542        of around 200,000 forward references. While doing this, I also speeded up
543        the filling in of repeated forward references.
544    
545    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
546        incorrectly expecting the subject to contain another "a" after the start.
547    
548    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
549        in the match, the SKIP should be ignored. This was not happening; instead
550        the SKIP was being treated as NOMATCH. For patterns such as
551        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
552        tested.
553    
554    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
555        now much more compatible with Perl, in particular in cases where the result
556        is a non-match for a non-anchored pattern. For example, if
557        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
558        "m", where previously it did not return a name. A side effect of this
559        change is that for partial matches, the last encountered mark name is
560        returned, as for non matches. A number of tests that were previously not
561        Perl-compatible have been moved into the Perl-compatible test files. The
562        refactoring has had the pleasing side effect of removing one argument from
563        the match() function, thus reducing its stack requirements.
564    
565    19. If the /S+ option was used in pcretest to study a pattern using JIT,
566        subsequent uses of /S (without +) incorrectly behaved like /S+.
567    
568    21. Retrieve executable code size support for the JIT compiler and fixing
569        some warnings.
570    
571    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
572        not work when the shorter character appeared right at the end of the
573        subject string.
574    
575    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
576        systems.
577    
578    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
579        output it when the /M option is used in pcretest.
580    
581    25. The CheckMan script was not being included in the distribution. Also, added
582        an explicit "perl" to run Perl scripts from the PrepareRelease script
583        because this is reportedly needed in Windows.
584    
585    26. If study data was being save in a file and studying had not found a set of
586        "starts with" bytes for the pattern, the data written to the file (though
587        never used) was taken from uninitialized memory and so caused valgrind to
588        complain.
589    
590    27. Updated RunTest.bat as provided by Sheri Pierce.
591    
592    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
593    
594    29. Computation of memory usage for the table of capturing group names was
595        giving an unnecessarily large value.
596    
597    
598    Version 8.20 21-Oct-2011
599    ------------------------
600    
601    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
602        a POSIX class. After further experiments with Perl, which convinced me that
603        Perl has bugs and confusions, a closing square bracket is no longer allowed
604        in a POSIX name. This bug also affected patterns with classes that started
605        with full stops.
606    
607    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
608        captured substring, but while checking the failing first alternative,
609        substring 1 is temporarily captured. If the output vector supplied to
610        pcre_exec() was not big enough for this capture, the yield of the function
611        was still zero ("insufficient space for captured substrings"). This cannot
612        be totally fixed without adding another stack variable, which seems a lot
613        of expense for a edge case. However, I have improved the situation in cases
614        such as /(a)(b)x|abc/ matched against "abc", where the return code
615        indicates that fewer than the maximum number of slots in the ovector have
616        been set.
617    
618    3.  Related to (2) above: when there are more back references in a pattern than
619        slots in the output vector, pcre_exec() uses temporary memory during
620        matching, and copies in the captures as far as possible afterwards. It was
621        using the entire output vector, but this conflicts with the specification
622        that only 2/3 is used for passing back captured substrings. Now it uses
623        only the first 2/3, for compatibility. This is, of course, another edge
624        case.
625    
626    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
627        main code base, and can be used by building with --enable-jit. When this is
628        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
629        runtime --no-jit option is given.
630    
631    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
632        ovector, the return from the function was zero, implying that there were
633        other matches that did not fit. The correct "exactly full" value is now
634        returned.
635    
636    6.  If a subpattern that was called recursively or as a subroutine contained
637        (*PRUNE) or any other control that caused it to give a non-standard return,
638        invalid errors such as "Error -26 (nested recursion at the same subject
639        position)" or even infinite loops could occur.
640    
641    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
642        computing the minimum length on reaching *ACCEPT, and so ended up with the
643        wrong value of 1 rather than 0. Further investigation indicates that
644        computing a minimum subject length in the presence of *ACCEPT is difficult
645        (think back references, subroutine calls), and so I have changed the code
646        so that no minimum is registered for a pattern that contains *ACCEPT.
647    
648    8.  If (*THEN) was present in the first (true) branch of a conditional group,
649        it was not handled as intended. [But see 16 below.]
650    
651    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
652        Sheri Pierce.
653    
654    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
655        the first byte in a match must be "a".
656    
657    11. Change 17 for 8.13 increased the recursion depth for patterns like
658        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
659        pattern contains any instances of (*THEN). If it does not, the old
660        optimizations are restored. It would be nice to do this on a per-group
661        basis, but at the moment that is not feasible.
662    
663    12. In some environments, the output of pcretest -C is CRLF terminated. This
664        broke RunTest's code that checks for the link size. A single white space
665        character after the value is now allowed for.
666    
667    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
668        For "fr", it uses the Windows-specific input and output files.
669    
670    14. If (*THEN) appeared in a group that was called recursively or as a
671        subroutine, it did not work as intended. [But see next item.]
672    
673    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
674        pattern fragments (but not containing any | characters). If A and B are
675        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
676        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
677        D. In other words, Perl considers parentheses that do not contain any |
678        characters to be part of a surrounding alternative, whereas PCRE was
679        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
680        differently. PCRE now behaves in the same way as Perl, except in the case
681        of subroutine/recursion calls such as (?1) which have in any case always
682        been different (but PCRE had them first :-).
683    
684    16. Related to 15 above: Perl does not treat the | in a conditional group as
685        creating alternatives. Such a group is treated in the same way as an
686        ordinary group without any | characters when processing (*THEN). PCRE has
687        been changed to match Perl's behaviour.
688    
689    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
690        RunGrepTest script failed.
691    
692    18. Change 22 for version 13 caused atomic groups to use more stack. This is
693        inevitable for groups that contain captures, but it can lead to a lot of
694        stack use in large patterns. The old behaviour has been restored for atomic
695        groups that do not contain any capturing parentheses.
696    
697    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
698        suppress the check for a minimum subject length at run time. (If it was
699        given to pcre_exec() or pcre_dfa_exec() it did work.)
700    
701    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
702        fail to work when decoding hex characters in data strings in EBCDIC
703        environments.
704    
705    21. It appears that in at least one Mac OS environment, the isxdigit() function
706        is implemented as a macro that evaluates to its argument more than once,
707        contravening the C 90 Standard (I haven't checked a later standard). There
708        was an instance in pcretest which caused it to go wrong when processing
709        \x{...} escapes in subject strings. The has been rewritten to avoid using
710        things like p++ in the argument of isxdigit().
711    
712    
713    Version 8.13 16-Aug-2011
714    ------------------------
715    
716    1.  The Unicode data tables have been updated to Unicode 6.0.0.
717    
718    2.  Two minor typos in pcre_internal.h have been fixed.
719    
720    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
721        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
722        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
723    
724    4.  There were a number of related bugs in the code for matching backrefences
725        caselessly in UTF-8 mode when codes for the characters concerned were
726        different numbers of bytes. For example, U+023A and U+2C65 are an upper
727        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
728        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
729        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
730        2-byte code at the end of the subject (it thought there wasn't enough data
731        left).
732    
733    5.  Comprehensive information about what went wrong is now returned by
734        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
735        as the output vector has at least 2 elements. The offset of the start of
736        the failing character and a reason code are placed in the vector.
737    
738    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
739        now returned is for the first byte of the failing character, instead of the
740        last byte inspected. This is an incompatible change, but I hope it is small
741        enough not to be a problem. It makes the returned offset consistent with
742        pcre_exec() and pcre_dfa_exec().
743    
744    7.  pcretest now gives a text phrase as well as the error number when
745        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
746        failure, the offset and reason code are output.
747    
748    8.  When \R was used with a maximizing quantifier it failed to skip backwards
749        over a \r\n pair if the subsequent match failed. Instead, it just skipped
750        back over a single character (\n). This seems wrong (because it treated the
751        two characters as a single entity when going forwards), conflicts with the
752        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
753        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
754        has been changed.
755    
756    9.  Some internal refactoring has changed the processing so that the handling
757        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
758        time (the PCRE_DOTALL option was changed this way some time ago: version
759        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
760        which was always a bit of a fudge. It also means that there is one less
761        argument for the match() function, which reduces its stack requirements
762        slightly. This change also fixes an incompatibility with Perl: the pattern
763        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
764    
765    10. More internal refactoring has drastically reduced the number of recursive
766        calls to match() for possessively repeated groups such as (abc)++ when
767        using pcre_exec().
768    
769    11. While implementing 10, a number of bugs in the handling of groups were
770        discovered and fixed:
771    
772        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
773        (a|)*(?1) gave a compile-time internal error.
774        ((a|)+)+  did not notice that the outer group could match an empty string.
775        (^a|^)+   was not marked as anchored.
776        (.*a|.*)+ was not marked as matching at start or after a newline.
777    
778    12. Yet more internal refactoring has removed another argument from the match()
779        function. Special calls to this function are now indicated by setting a
780        value in a variable in the "match data" data block.
781    
782    13. Be more explicit in pcre_study() instead of relying on "default" for
783        opcodes that mean there is no starting character; this means that when new
784        ones are added and accidentally left out of pcre_study(), testing should
785        pick them up.
786    
787    14. The -s option of pcretest has been documented for ages as being an old
788        synonym of -m (show memory usage). I have changed it to mean "force study
789        for every regex", that is, assume /S for every regex. This is similar to -i
790        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
791        using it. It makes it easier to run collections of tests with and without
792        study enabled, and thereby test pcre_study() more easily. All the standard
793        tests are now run with and without -s (but some patterns can be marked as
794        "never study" - see 20 below).
795    
796    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
797        restoration of the capturing data to the outer values was not happening
798        correctly.
799    
800    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
801        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
802        pattern had matched an empty string, and so incorrectly returned a no
803        match.
804    
805    17. There was optimizing code for the last branch of non-capturing parentheses,
806        and also for the obeyed branch of a conditional subexpression, which used
807        tail recursion to cut down on stack usage. Unfortunately, now that there is
808        the possibility of (*THEN) occurring in these branches, tail recursion is
809        no longer possible because the return has to be checked for (*THEN). These
810        two optimizations have therefore been removed. [But see 8.20/11 above.]
811    
812    18. If a pattern containing \R was studied, it was assumed that \R always
813        matched two bytes, thus causing the minimum subject length to be
814        incorrectly computed because \R can also match just one byte.
815    
816    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
817        was incorrectly computed.
818    
819    20. If /S is present twice on a test pattern in pcretest input, it now
820        *disables* studying, thereby overriding the use of -s on the command line
821        (see 14 above). This is necessary for one or two tests to keep the output
822        identical in both cases.
823    
824    21. When (*ACCEPT) was used in an assertion that matched an empty string and
825        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
826    
827    22. When an atomic group that contained a capturing parenthesis was
828        successfully matched, but the branch in which it appeared failed, the
829        capturing was not being forgotten if a higher numbered group was later
830        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
831        group 1 to "a", when in fact it should be unset. This applied to multi-
832        branched capturing and non-capturing groups, repeated or not, and also to
833        positive assertions (capturing in negative assertions does not happen
834        in PCRE) and also to nested atomic groups.
835    
836    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
837        subject after a captured substring, to make it easier to tell which of a
838        number of identical substrings has been captured.
839    
840    24. The way atomic groups are processed by pcre_exec() has been changed so that
841        if they are repeated, backtracking one repetition now resets captured
842        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
843        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
844        "aaa". Previously, it would have been "a". As part of this code
845        refactoring, the way recursive calls are handled has also been changed.
846    
847    25. If an assertion condition captured any substrings, they were not passed
848        back unless some other capturing happened later. For example, if
849        (?(?=(a))a) was matched against "a", no capturing was returned.
850    
851    26. When studying a pattern that contained subroutine calls or assertions,
852        the code for finding the minimum length of a possible match was handling
853        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
854        group 1 called group 2 while simultaneously a separate group 2 called group
855        1). A stack overflow occurred in this case. I have fixed this by limiting
856        the recursion depth to 10.
857    
858    27. Updated RunTest.bat in the distribution to the version supplied by Tom
859        Fortmann. This supports explicit test numbers on the command line, and has
860        argument validation and error reporting.
861    
862    28. An instance of \X with an unlimited repeat could fail if at any point the
863        first character it looked at was a mark character.
864    
865    29. Some minor code refactoring concerning Unicode properties and scripts
866        should reduce the stack requirement of match() slightly.
867    
868    30. Added the '=' option to pcretest to check the setting of unused capturing
869        slots at the end of the pattern, which are documented as being -1, but are
870        not included in the return count.
871    
872    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
873        compiled something random. Now it gives a compile-time error (as does
874        Perl).
875    
876    32. A *MARK encountered during the processing of a positive assertion is now
877        recorded and passed back (compatible with Perl).
878    
879    33. If --only-matching or --colour was set on a pcregrep call whose pattern
880        had alternative anchored branches, the search for a second match in a line
881        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
882        matched the line "0102" twice. The same bug affected patterns that started
883        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
884        twice.
885    
886    34. Previously, PCRE did not allow quantification of assertions. However, Perl
887        does, and because of capturing effects, quantifying parenthesized
888        assertions may at times be useful. Quantifiers are now allowed for
889        parenthesized assertions.
890    
891    35. A minor code tidy in pcre_compile() when checking options for \R usage.
892    
893    36. \g was being checked for fancy things in a character class, when it should
894        just be a literal "g".
895    
896    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
897        appearance of a nested POSIX class supersedes an apparent external class.
898        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
899        unescaped square brackets may also appear as part of class names. For
900        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
901        more like Perl. (But see 8.20/1 above.)
902    
903    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
904        was because it thought it was \N{name}, which is not supported).
905    
906    39. Add minix to OS list not supporting the -S option in pcretest.
907    
908    40. PCRE tries to detect cases of infinite recursion at compile time, but it
909        cannot analyze patterns in sufficient detail to catch mutual recursions
910        such as ((?1))((?2)). There is now a runtime test that gives an error if a
911        subgroup is called recursively as a subpattern for a second time at the
912        same position in the subject string. In previous releases this might have
913        been caught by the recursion limit, or it might have run out of stack.
914    
915    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
916        happen only once. PCRE was, however incorrectly giving a compile time error
917        "recursive call could loop indefinitely" because it cannot analyze the
918        pattern in sufficient detail. The compile time test no longer happens when
919        PCRE is compiling a conditional subpattern, but actual runaway loops are
920        now caught at runtime (see 40 above).
921    
922    42. It seems that Perl allows any characters other than a closing parenthesis
923        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
924        has been changed to be the same.
925    
926    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
927        as not to get warnings when autogen.sh is called. Also changed
928        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
929    
930    44. To help people who use pcregrep to scan files containing exceedingly long
931        lines, the following changes have been made:
932    
933        (a) The default value of the buffer size parameter has been increased from
934            8K to 20K. (The actual buffer used is three times this size.)
935    
936        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
937            PCRE is built.
938    
939        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
940            to be set at run time.
941    
942        (d) Numerical values in pcregrep options can be followed by K or M, for
943            example --buffer-size=50K.
944    
945        (e) If a line being scanned overflows pcregrep's buffer, an error is now
946            given and the return code is set to 2.
947    
948    45. Add a pointer to the latest mark to the callout data block.
949    
950    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
951        partial match of an empty string instead of no match. This was specific to
952        the use of ".".
953    
954    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
955        complete match instead of a partial match. This bug was dependent on both
956        the PCRE_UTF8 and PCRE_DOTALL options being set.
957    
958    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
959        starting byte set, because \b was not being ignored.
960    
961    
962    Version 8.12 15-Jan-2011
963    ------------------------
964    
965    1.  Fixed some typos in the markup of the man pages, and wrote a script that
966        checks for such things as part of the documentation building process.
967    
968    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
969        --match-limit and --recursion-limit options (added for 8.11). In
970        particular, this made one of the standard tests fail. (The integer value
971        went into the wrong half of a long int.)
972    
973    3.  If the --colour option was given to pcregrep with -v (invert match), it
974        did strange things, either producing crazy output, or crashing. It should,
975        of course, ignore a request for colour when reporting lines that do not
976        match.
977    
978    4.  Another pcregrep bug caused similar problems if --colour was specified with
979        -M (multiline) and the pattern match finished with a line ending.
980    
981    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
982        matched in multiline mode, the following line was shown as part of the
983        match. This seems wrong, so I have changed it.
984    
985    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
986        the check for further matches in the same line (so they could be coloured)
987        to overrun the end of the current line. If another match was found, it was
988        incorrectly shown (and then shown again when found in the next line).
989    
990    7.  If pcregrep was compiled under Windows, there was a reference to the
991        function pcregrep_exit() before it was defined. I am assuming this was
992        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
993        reported by a user. I've moved the definition above the reference.
994    
995    
996    Version 8.11 10-Dec-2010
997    ------------------------
998    
999    1.  (*THEN) was not working properly if there were untried alternatives prior
1000        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1001        backtracked to try for "b" instead of moving to the next alternative branch
1002        at the same level (in this case, to look for "c"). The Perl documentation
1003        is clear that when (*THEN) is backtracked onto, it goes to the "next
1004        alternative in the innermost enclosing group".
1005    
1006    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1007        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1008        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1009        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1010        (*THEN).
1011    
1012    3.  If \s appeared in a character class, it removed the VT character from
1013        the class, even if it had been included by some previous item, for example
1014        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1015        of \s, but is part of the POSIX "space" class.)
1016    
1017    4.  A partial match never returns an empty string (because you can always
1018        match an empty string at the end of the subject); however the checking for
1019        an empty string was starting at the "start of match" point. This has been
1020        changed to the "earliest inspected character" point, because the returned
1021        data for a partial match starts at this character. This means that, for
1022        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1023        (previously it gave "no match").
1024    
1025    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1026        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1027        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1028        has an implication that the given string is incomplete (because a partial
1029        match is preferred over a full match). For this reason, these items now
1030        give a partial match in this situation. [Aside: previously, the one case
1031        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1032        match rather than a full match, which was wrong by the old rules, but is
1033        now correct.]
1034    
1035    6.  There was a bug in the handling of #-introduced comments, recognized when
1036        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1037        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1038        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1039        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1040        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1041        places in pcre_compile().
1042    
1043    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1044        comments when looking ahead for named forward references to subpatterns,
1045        the only newline sequence it recognized was NL. It now handles newlines
1046        according to the set newline convention.
1047    
1048    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1049        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1050        cater for a lack of strerror(). These oversights have been fixed.
1051    
1052    9.  Added --match-limit and --recursion-limit to pcregrep.
1053    
1054    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1055    
1056    11. When the -o option was used, pcregrep was setting a return code of 1, even
1057        when matches were found, and --line-buffered was not being honoured.
1058    
1059    12. Added an optional parentheses number to the -o and --only-matching options
1060        of pcregrep.
1061    
1062    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1063        can match an empty string. The code to do it in pcretest and pcredemo
1064        needed fixing:
1065    
1066        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1067            only one byte after an empty string match just before CRLF (this case
1068            just got forgotten; "any" and "anycrlf" were OK).
1069    
1070        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1071            mode when an empty string match preceded an ASCII character followed by
1072            a non-ASCII character. (The code for advancing by one character rather
1073            than one byte was nonsense.)
1074    
1075        (c) The pcredemo.c sample program did not have any code at all to handle
1076            the cases when CRLF is a valid newline sequence.
1077    
1078    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1079        as a starting offset was within the subject string. There is now a new
1080        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1081        negative or greater than the length of the string. In order to test this,
1082        pcretest is extended to allow the setting of negative starting offsets.
1083    
1084    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1085        starting offset points to the beginning of a UTF-8 character was
1086        unnecessarily clumsy. I tidied it up.
1087    
1088    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1089        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1090    
1091    17. Nobody had reported that the --include_dir option, which was added in
1092        release 7.7 should have been called --include-dir (hyphen, not underscore)
1093        for compatibility with GNU grep. I have changed it to --include-dir, but
1094        left --include_dir as an undocumented synonym, and the same for
1095        --exclude-dir, though that is not available in GNU grep, at least as of
1096        release 2.5.4.
1097    
1098    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1099        characters from a string of bytes) have been redefined so as not to use
1100        loops, in order to improve performance in some environments. At the same
1101        time, I abstracted some of the common code into auxiliary macros to save
1102        repetition (this should not affect the compiled code).
1103    
1104    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1105        compile-time error is now given if \c is not followed by an ASCII
1106        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1107        different, and any byte value is allowed.)
1108    
1109    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1110        START_OPTIMIZE option, which is now allowed at compile time - but just
1111        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1112        to pcregrep and other applications that have no direct access to PCRE
1113        options. The new /Y option in pcretest sets this option when calling
1114        pcre_compile().
1115    
1116    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1117        back references. Groups containing recursive back references were forced to
1118        be atomic by that change, but in the case of named groups, the amount of
1119        memory required was incorrectly computed, leading to "Failed: internal
1120        error: code overflow". This has been fixed.
1121    
1122    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1123        pcretest.c, to avoid build problems in some Borland environments.
1124    
1125    
1126    Version 8.10 25-Jun-2010
1127    ------------------------
1128    
1129    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1130        THEN.
1131    
1132    2.  (*ACCEPT) was not working when inside an atomic group.
1133    
1134    3.  Inside a character class, \B is treated as a literal by default, but
1135        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1136        causes the error). The code is unchanged, but I tidied the documentation.
1137    
1138    4.  Inside a character class, PCRE always treated \R and \X as literals,
1139        whereas Perl faults them if its -w option is set. I have changed PCRE so
1140        that it faults them when PCRE_EXTRA is set.
1141    
1142    5.  Added support for \N, which always matches any character other than
1143        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1144    
1145    6.  When compiling pcregrep with newer versions of gcc which may have
1146        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1147        declared with attribute warn_unused_result" were given. Just casting the
1148        result to (void) does not stop the warnings; a more elaborate fudge is
1149        needed. I've used a macro to implement this.
1150    
1151    7.  Minor change to pcretest.c to avoid a compiler warning.
1152    
1153    8.  Added four artifical Unicode properties to help with an option to make
1154        \s etc use properties (see next item). The new properties are: Xan
1155        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1156    
1157    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1158        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1159        this option. Modified pcretest to add /W to test this facility. Added
1160        REG_UCP to make it available via the POSIX interface.
1161    
1162    10. Added --line-buffered to pcregrep.
1163    
1164    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1165        studied, and the match started with a letter with a code point greater than
1166        127 whose first byte was different to the first byte of the other case of
1167        the letter, the other case of this starting letter was not recognized
1168        (#976).
1169    
1170    12. If a pattern that was studied started with a repeated Unicode property
1171        test, for example, \p{Nd}+, there was the theoretical possibility of
1172        setting up an incorrect bitmap of starting bytes, but fortunately it could
1173        not have actually happened in practice until change 8 above was made (it
1174        added property types that matched character-matching opcodes).
1175    
1176    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1177        possible starting bytes for non-anchored patterns.
1178    
1179    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1180        \R, and also a number of cases that involve Unicode properties, both
1181        explicit and implicit when PCRE_UCP is set.
1182    
1183    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1184        input, it could crash or give wrong results if characters with values
1185        greater than 0xc0 were present in the subject string. (Detail: it assumed
1186        UTF-8 input when processing these items.)
1187    
1188    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1189        size_t is 64-bit (#991).
1190    
1191    17. Added a check for running out of memory when PCRE is compiled with
1192        --disable-stack-for-recursion (#990).
1193    
1194    18. If the last data line in a file for pcretest does not have a newline on
1195        the end, a newline was missing in the output.
1196    
1197    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1198        less than 128) in its various bitmaps. However, there is a facility for
1199        generating tables according to the current locale when PCRE is compiled. It
1200        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1201        characters, are recognized by isspace() and therefore were getting set in
1202        these tables, and indeed these tables seem to approximate to ISO 8859. This
1203        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1204        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1205        which of course cannot start UTF-8 characters. I have changed the code so
1206        that only real ASCII characters (less than 128) and the correct starting
1207        bytes for UTF-8 encodings are set for characters greater than 127 when in
1208        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1209        altogether.)
1210    
1211    20. Added the /T option to pcretest so as to be able to run tests with non-
1212        standard character tables, thus making it possible to include the tests
1213        used for 19 above in the standard set of tests.
1214    
1215    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1216        reference to a subpattern the other side of a comment that contains an
1217        opening parenthesis caused either an internal compiling error, or a
1218        reference to the wrong subpattern.
1219    
1220    
1221    Version 8.02 19-Mar-2010
1222    ------------------------
1223    
1224    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1225    
1226    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1227        configured.
1228    
1229    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1230        original author of that file, following a query about its status.
1231    
1232    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1233        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1234    
1235    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1236        quantifier applied to a forward-referencing subroutine call, could compile
1237        incorrect code or give the error "internal error: previously-checked
1238        referenced subpattern not found".
1239    
1240    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1241        variables to point to external functions. For these systems, therefore,
1242        pcre_malloc etc. are now initialized to local functions that call the
1243        relevant global functions.
1244    
1245    7.  There were two entries missing in the vectors called coptable and poptable
1246        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1247        I've fixed the data, and added a kludgy way of testing at compile time that
1248        the lengths are correct (equal to the number of opcodes).
1249    
1250    8.  Following on from 7, I added a similar kludge to check the length of the
1251        eint vector in pcreposix.c.
1252    
1253    9.  Error texts for pcre_compile() are held as one long string to avoid too
1254        much relocation at load time. To find a text, the string is searched,
1255        counting zeros. There was no check for running off the end of the string,
1256        which could happen if a new error number was added without updating the
1257        string.
1258    
1259    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1260    
1261    11. \K was not working if it appeared in an atomic group or in a group that
1262        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1263        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1264        the assertion is positive, but not if it is negative.
1265    
1266    12. Change 11 fortuitously reduced the size of the stack frame used in the
1267        "match()" function of pcre_exec.c by one pointer. Forthcoming
1268        implementation of support for (*MARK) will need an extra pointer on the
1269        stack; I have reserved it now, so that the stack frame size does not
1270        decrease.
1271    
1272    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1273        item in branch that calls a recursion is a subroutine call - as in the
1274        second branch in the above example - was incorrectly given the compile-
1275        time error "recursive call could loop indefinitely" because pcre_compile()
1276        was not correctly checking the subroutine for matching a non-empty string.
1277    
1278    14. The checks for overrunning compiling workspace could trigger after an
1279        overrun had occurred. This is a "should never occur" error, but it can be
1280        triggered by pathological patterns such as hundreds of nested parentheses.
1281        The checks now trigger 100 bytes before the end of the workspace.
1282    
1283    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1284    
1285    
1286    Version 8.01 19-Jan-2010
1287    ------------------------
1288    
1289    1.  If a pattern contained a conditional subpattern with only one branch (in
1290        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1291        computed the wrong minimum data length (which is of course zero for such
1292        subpatterns). This could cause incorrect "no match" results.
1293    
1294    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1295        the pattern is reset in the first branch, pcre_compile() failed with
1296        "internal error: code overflow at offset...". This happened only when
1297        the reset was to the original external option setting. (An optimization
1298        abstracts leading options settings into an external setting, which was the
1299        cause of this.)
1300    
1301    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1302        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1303        assertion pattern did not match (meaning that the assertion was true), it
1304        was incorrectly treated as false if the SKIP had been reached during the
1305        matching. This also applied to assertions used as conditions.
1306    
1307    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1308        assertion subpattern, including such a pattern used as a condition,
1309        unpredictable results occurred, instead of the error return
1310        PCRE_ERROR_DFA_UITEM.
1311    
1312    5.  The C++ GlobalReplace function was not working like Perl for the special
1313        situation when an empty string is matched. It now does the fancy magic
1314        stuff that is necessary.
1315    
1316    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1317        removed. (These were left over from very, very early versions of PCRE.)
1318    
1319    7.  Some cosmetic changes to the code to make life easier when compiling it
1320        as part of something else:
1321    
1322        (a) Change DEBUG to PCRE_DEBUG.
1323    
1324        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1325            called "current" as "current_branch", to prevent a collision with the
1326            Linux macro when compiled as a kernel module.
1327    
1328        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1329            prevent a collision with the Linux macro when compiled as a kernel
1330            module.
1331    
1332    8.  In pcre_compile() there are some checks for integer overflows that used to
1333        cast potentially large values to (double). This has been changed to that
1334        when building, a check for int64_t is made, and if it is found, it is used
1335        instead, thus avoiding the use of floating point arithmetic. (There is no
1336        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1337        double.
1338    
1339    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1340        2005 (difference between two addresses compared to an unsigned value).
1341    
1342    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1343        custom one, because of the following reported problem in Windows:
1344    
1345          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1346              under Win32.
1347          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1348              therefore missing the function definition.
1349          - The compiler thus generates a "C" signature for the test function.
1350          - The linker fails to find the "C" function.
1351          - PCRE fails to configure if asked to do so against libbz2.
1352    
1353    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1354        messages were output:
1355    
1356          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1357          rerunning libtoolize, to keep the correct libtool macros in-tree.
1358          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1359    
1360        I have done both of these things.
1361    
1362    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1363        most of the time, it *can* run out if it is given a pattern that contains a
1364        runaway infinite recursion. I updated the discussion in the pcrestack man
1365        page.
1366    
1367    13. Now that we have gone to the x.xx style of version numbers, the minor
1368        version may start with zero. Using 08 or 09 is a bad idea because users
1369        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1370        interpreted as invalid octal numbers. I've updated the previous comment in
1371        configure.ac, and also added a check that gives an error if 08 or 09 are
1372        used.
1373    
1374    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1375        causing partial matching to fail when the end of the subject matched \W
1376        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1377    
1378    15. There were some discrepancies between the declarations in pcre_internal.h
1379        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1380        their definitions. The declarations used "const uschar *" and the
1381        definitions used USPTR. Even though USPTR is normally defined as "const
1382        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1383        reported that: "This difference in casting confuses some C++ compilers, for
1384        example, SunCC recognizes above declarations as different functions and
1385        generates broken code for hbpcre." I have changed the declarations to use
1386        USPTR.
1387    
1388    16. GNU libtool is named differently on some systems. The autogen.sh script now
1389        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1390        (FreeBSD).
1391    
1392    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1393        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1394        comment: "Figure out how to create a longlong from a string: strtoll and
1395        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1396        instance, but it only takes 2 args instead of 3!"
1397    
1398    18. A subtle bug concerned with back references has been fixed by a change of
1399        specification, with a corresponding code fix. A pattern such as
1400        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1401        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1402        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1403        same bug. Such groups have to be quantified to be useful, or contained
1404        inside another quantified group. (If there's no repetition, the reference
1405        can never match.) The problem arises because, having left the group and
1406        moved on to the rest of the pattern, a later failure that backtracks into
1407        the group uses the captured value from the final iteration of the group
1408        rather than the correct earlier one. I have fixed this in PCRE by forcing
1409        any group that contains a reference to itself to be an atomic group; that
1410        is, there cannot be any backtracking into it once it has completed. This is
1411        similar to recursive and subroutine calls.
1412    
1413    
1414    Version 8.00 19-Oct-09
1415    ----------------------
1416    
1417    1.  The table for translating pcre_compile() error codes into POSIX error codes
1418        was out-of-date, and there was no check on the pcre_compile() error code
1419        being within the table. This could lead to an OK return being given in
1420        error.
1421    
1422    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1423        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1424        in a Windows environment.
1425    
1426    3.  The pcregrep --count option prints the count for each file even when it is
1427        zero, as does GNU grep. However, pcregrep was also printing all files when
1428        --files-with-matches was added. Now, when both options are given, it prints
1429        counts only for those files that have at least one match. (GNU grep just
1430        prints the file name in this circumstance, but including the count seems
1431        more useful - otherwise, why use --count?) Also ensured that the
1432        combination -clh just lists non-zero counts, with no names.
1433    
1434    4.  The long form of the pcregrep -F option was incorrectly implemented as
1435        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1436        but it seems right to fix it, and I didn't think it was worth preserving
1437        the old behaviour.
1438    
1439    5.  The command line items --regex=pattern and --regexp=pattern were not
1440        recognized by pcregrep, which required --regex pattern or --regexp pattern
1441        (with a space rather than an '='). The man page documented the '=' forms,
1442        which are compatible with GNU grep; these now work.
1443    
1444    6.  No libpcreposix.pc file was created for pkg-config; there was just
1445        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1446    
1447    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1448        when UCP support is not needed, by modifying the Python script that
1449        generates it from Unicode data files. This should not matter if the module
1450        is correctly used as a library, but I received one complaint about 50K of
1451        unwanted data. My guess is that the person linked everything into his
1452        program rather than using a library. Anyway, it does no harm.
1453    
1454    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1455        was a minimum greater than 1 for a wide character in a possessive
1456        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1457        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1458        character. Chaos in the form of incorrect output or a compiling loop could
1459        result.
1460    
1461    9.  The restrictions on what a pattern can contain when partial matching is
1462        requested for pcre_exec() have been removed. All patterns can now be
1463        partially matched by this function. In addition, if there are at least two
1464        slots in the offset vector, the offset of the earliest inspected character
1465        for the match and the offset of the end of the subject are set in them when
1466        PCRE_ERROR_PARTIAL is returned.
1467    
1468    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1469        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1470        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1471        and may be more useful for multi-segment matching.
1472    
1473    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1474        used to be given if ever the end of the subject was reached; now it is
1475        given only if matching could not proceed because another character was
1476        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1477        string "Z", which now yields "no match" instead of "partial match". In the
1478        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1479        final character ended with (*FAIL).
1480    
1481    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1482        if the pattern had a "must contain" character that was already found in the
1483        earlier partial match, unless partial matching was again requested. For
1484        example, with the pattern /dog.(body)?/, the "must contain" character is
1485        "g". If the first part-match was for the string "dog", restarting with
1486        "sbody" failed. This bug has been fixed.
1487    
1488    13. The string returned by pcre_dfa_exec() after a partial match has been
1489        changed so that it starts at the first inspected character rather than the
1490        first character of the match. This makes a difference only if the pattern
1491        starts with a lookbehind assertion or \b or \B (\K is not supported by
1492        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1493        matching functions compatible, and I think it's the right thing to do.
1494    
1495    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1496        so that the demonstration program is easily available in environments where
1497        PCRE has not been installed from source.
1498    
1499    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1500        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1501        library.
1502    
1503    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1504        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1505        is not the first non-POSIX option to be added. Clearly some people find
1506        these options useful.
1507    
1508    17. If a caller to the POSIX matching function regexec() passes a non-zero
1509        value for nmatch with a NULL value for pmatch, the value of
1510        nmatch is forced to zero.
1511    
1512    18. RunGrepTest did not have a test for the availability of the -u option of
1513        the diff command, as RunTest does. It now checks in the same way as
1514        RunTest, and also checks for the -b option.
1515    
1516    19. If an odd number of negated classes containing just a single character
1517        interposed, within parentheses, between a forward reference to a named
1518        subpattern and the definition of the subpattern, compilation crashed with
1519        an internal error, complaining that it could not find the referenced
1520        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1521        [The bug was that it was starting one character too far in when skipping
1522        over the character class, thus treating the ] as data rather than
1523        terminating the class. This meant it could skip too much.]
1524    
1525    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1526        /g option in pcretest when the pattern contains \K, which makes it possible
1527        to have an empty string match not at the start, even when the pattern is
1528        anchored. Updated pcretest and pcredemo to use this option.
1529    
1530    21. If the maximum number of capturing subpatterns in a recursion was greater
1531        than the maximum at the outer level, the higher number was returned, but
1532        with unset values at the outer level. The correct (outer level) value is
1533        now given.
1534    
1535    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1536        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1537        make it do so. The string so far is captured, making this feature
1538        compatible with Perl.
1539    
1540    23. The tests have been re-organized, adding tests 11 and 12, to make it
1541        possible to check the Perl 5.10 features against Perl 5.10.
1542    
1543    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1544        pattern matches a fixed length string. PCRE did not allow this; now it
1545        does. Neither allows recursion.
1546    
1547    25. I finally figured out how to implement a request to provide the minimum
1548        length of subject string that was needed in order to match a given pattern.
1549        (It was back references and recursion that I had previously got hung up
1550        on.) This code has now been added to pcre_study(); it finds a lower bound
1551        to the length of subject needed. It is not necessarily the greatest lower
1552        bound, but using it to avoid searching strings that are too short does give
1553        some useful speed-ups. The value is available to calling programs via
1554        pcre_fullinfo().
1555    
1556    26. While implementing 25, I discovered to my embarrassment that pcretest had
1557        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1558        study optimizations had never been tested with that matching function.
1559        Oops. What is worse, even when it was passed study data, there was a bug in
1560        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1561        were also very few tests of studied patterns with pcre_dfa_exec().
1562    
1563    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1564        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1565        on the other side of the coin, they are no longer allowed to have different
1566        names, because these cannot be distinguished in PCRE, and this has caused
1567        confusion. (This is a difference from Perl.)
1568    
1569    28. When duplicate subpattern names are present (necessarily with different
1570        numbers, as required by 27 above), and a test is made by name in a
1571        conditional pattern, either for a subpattern having been matched, or for
1572        recursion in such a pattern, all the associated numbered subpatterns are
1573        tested, and the overall condition is true if the condition is true for any
1574        one of them. This is the way Perl works, and is also more like the way
1575        testing by number works.
1576    
1577    
1578    Version 7.9 11-Apr-09
1579    ---------------------
1580    
1581    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1582        (pcretest), all targets were linked against these libraries. This included
1583        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1584        libraries. This caused unwanted dependencies to be created. This problem
1585        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1586        pcretest is linked with readline.
1587    
1588    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1589        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1590        moved outside it again, because FALSE and TRUE are already defined in AIX,
1591        but BOOL is not.
1592    
1593    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1594        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1595    
1596    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1597        hyphens) following file names and line numbers when outputting matching
1598        lines. This is not true; no spaces are inserted. I have also clarified the
1599        wording for the --colour (or --color) option.
1600    
1601    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1602        was not coloured; this is different to GNU grep, so I have changed it to be
1603        the same.
1604    
1605    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1606        each matching line was coloured. Now it goes on to look for further matches
1607        of any of the test patterns, which is the same behaviour as GNU grep.
1608    
1609    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1610        doesn't make sense to accept an empty string match in pcregrep, so I have
1611        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1612        seems to be how GNU grep behaves.
1613    
1614    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1615        start or after a newline", because the conditional assertion was not being
1616        correctly handled. The rule now is that both the assertion and what follows
1617        in the first alternative must satisfy the test.
1618    
1619    9.  If auto-callout was enabled in a pattern with a conditional group whose
1620        condition was an assertion, PCRE could crash during matching, both with
1621        pcre_exec() and pcre_dfa_exec().
1622    
1623    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1624        used for matching.
1625    
1626    11. Unicode property support in character classes was not working for
1627        characters (bytes) greater than 127 when not in UTF-8 mode.
1628    
1629    12. Added the -M command line option to pcretest.
1630    
1631    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1632    
1633    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1634    
1635    16. Added comments and documentation about mis-use of no_arg in the C++
1636        wrapper.
1637    
1638    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1639        from Martin Jerabek that uses macro names for all relevant character and
1640        string constants.
1641    
1642    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1643        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1644        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1645        these, but not everybody uses configure.
1646    
1647    19. A conditional group that had only one branch was not being correctly
1648        recognized as an item that could match an empty string. This meant that an
1649        enclosing group might also not be so recognized, causing infinite looping
1650        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1651        with the subject "ab", where knowledge that the repeated group can match
1652        nothing is needed in order to break the loop.
1653    
1654    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1655        exec(), but without supplying a callout function, matching went wrong.
1656    
1657    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1658        leak if the size of the offset vector was greater than 30. When the vector
1659        is smaller, the saved offsets during recursion go onto a local stack
1660        vector, but for larger vectors malloc() is used. It was failing to free
1661        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1662        error, in fact).
1663    
1664    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1665        heapframe that is used only when UTF-8 support is enabled. This caused no
1666        problem, but was untidy.
1667    
1668    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1669        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1670        included within another project.
1671    
1672    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1673        slightly modified by me:
1674    
1675          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1676              not building pcregrep.
1677    
1678          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1679              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1680    
1681    25. Forward references, both numeric and by name, in patterns that made use of
1682        duplicate group numbers, could behave incorrectly or give incorrect errors,
1683        because when scanning forward to find the reference group, PCRE was not
1684        taking into account the duplicate group numbers. A pattern such as
1685        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1686    
1687    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1688        the feature of a custom pointer more persuasive (as requested by a user).
1689    
1690    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1691        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1692        pre-defined.
1693    
1694    28. Added support for (*UTF8) at the start of a pattern.
1695    
1696    29. Arrange for flags added by the "release type" setting in CMake to be shown
1697        in the configuration summary.
1698    
1699    
1700    Version 7.8 05-Sep-08
1701    ---------------------
1702    
1703    1.  Replaced UCP searching code with optimized version as implemented for Ad
1704        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1705        stage table and inline lookup instead of a function, giving speed ups of 2
1706        to 5 times on some simple patterns that I tested. Permission was given to
1707        distribute the MultiStage2.py script that generates the tables (it's not in
1708        the tarball, but is in the Subversion repository).
1709    
1710    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1711        scripts.
1712    
1713    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1714        a group with a zero qualifier. The result of the study could be incorrect,
1715        or the function might crash, depending on the pattern.
1716    
1717    4.  Caseless matching was not working for non-ASCII characters in back
1718        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1719        It now works when Unicode Property Support is available.
1720    
1721    5.  In pcretest, an escape such as \x{de} in the data was always generating
1722        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1723        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1724        truncation.
1725    
1726    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1727    
1728    7.  Added two (int) casts to pcregrep when printing the difference of two
1729        pointers, in case they are 64-bit values.
1730    
1731    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1732        test 2 if it fails.
1733    
1734    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1735        and a #define of that name to empty if it is not externally set. This is to
1736        allow users of MSVC to set it if necessary.
1737    
1738    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1739        the convenience functions in the pcre_get.c source file.
1740    
1741    11. An option change at the start of a pattern that had top-level alternatives
1742        could cause overwriting and/or a crash. This command provoked a crash in
1743        some environments:
1744    
1745          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1746    
1747        This potential security problem was recorded as CVE-2008-2371.
1748    
1749    12. For a pattern where the match had to start at the beginning or immediately
1750        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1751        pcre_dfa_exec() could read past the end of the passed subject if there was
1752        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1753        pcretest so that it places the subject at the end of its malloc-ed buffer.
1754    
1755    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1756        exec() might read past the end of the data buffer in UTF-8 mode.
1757    
1758    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1759        the data contained the byte 0x85 as part of a UTF-8 character within its
1760        first line. This applied both to normal and DFA matching.
1761    
1762    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1763        /^[^d]*?$/8 failed to match "abc".
1764    
1765    16. Added a missing copyright notice to pcrecpp_internal.h.
1766    
1767    17. Make it more clear in the documentation that values returned from
1768        pcre_exec() in ovector are byte offsets, not character counts.
1769    
1770    18. Tidied a few places to stop certain compilers from issuing warnings.
1771    
1772    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1773        supplied by Stefan Weber. I made a further small update for 7.8 because
1774        there is a change of source arrangements: the pcre_searchfuncs.c module is
1775        replaced by pcre_ucd.c.
1776    
1777    
1778    Version 7.7 07-May-08
1779    ---------------------
1780    
1781    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1782        a string to a long long, pretend we don't even have a long long." This is
1783        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1784    
1785    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1786        pre-7.6 versions, which defined a global no_arg variable instead of putting
1787        it in the RE class. (See also #8 below.)
1788    
1789    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1790        Lopes.
1791    
1792    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1793    
1794        (1) The include/exclude patterns were being applied to the whole pathnames
1795            of files, instead of just to the final components.
1796    
1797        (2) If there was more than one level of directory, the subdirectories were
1798            skipped unless they satisfied the include/exclude conditions. This is
1799            inconsistent with GNU grep (and could even be seen as contrary to the
1800            pcregrep specification - which I improved to make it absolutely clear).
1801            The action now is always to scan all levels of directory, and just
1802            apply the include/exclude patterns to regular files.
1803    
1804    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1805        --exclude_dir in the tests to avoid scanning .svn directories.
1806    
1807    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1808        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1809        doesn't support NULs in patterns.
1810    
1811    7.  Added some missing "const"s to declarations of static tables in
1812        pcre_compile.c and pcre_dfa_exec.c.
1813    
1814    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1815        caused by fix #2  above. (Subsequently also a second patch to fix the
1816        first patch. And a third patch - this was a messy problem.)
1817    
1818    9.  Applied Craig's patch to remove the use of push_back().
1819    
1820    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1821        matching function regexec().
1822    
1823    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1824        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1825        references. PCRE supports relative numbers with this syntax (I don't think
1826        Oniguruma does).
1827    
1828    12. Previously, a group with a zero repeat such as (...){0} was completely
1829        omitted from the compiled regex. However, this means that if the group
1830        was called as a subroutine from elsewhere in the pattern, things went wrong
1831        (an internal error was given). Such groups are now left in the compiled
1832        pattern, with a new opcode that causes them to be skipped at execution
1833        time.
1834    
1835    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1836        to the way PCRE behaves:
1837    
1838        (a) A lone ] character is dis-allowed (Perl treats it as data).
1839    
1840        (b) A back reference to an unmatched subpattern matches an empty string
1841            (Perl fails the current match path).
1842    
1843        (c) A data ] in a character class must be notated as \] because if the
1844            first data character in a class is ], it defines an empty class. (In
1845            Perl it is not possible to have an empty class.) The empty class []
1846            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1847            The negative empty class [^] matches any one character, independently
1848            of the DOTALL setting.
1849    
1850    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1851        non-existent subpattern following a character class starting with ']' and
1852        containing () gave an internal compiling error instead of "reference to
1853        non-existent subpattern". Fortunately, when the pattern did exist, the
1854        compiled code was correct. (When scanning forwards to check for the
1855        existencd of the subpattern, it was treating the data ']' as terminating
1856        the class, so got the count wrong. When actually compiling, the reference
1857        was subsequently set up correctly.)
1858    
1859    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1860        it was being rejected as not supported by pcre_dfa_exec(), even though
1861        other assertions are supported. I have made pcre_dfa_exec() support
1862        (*FAIL).
1863    
1864    16. The implementation of 13c above involved the invention of a new opcode,
1865        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1866        cannot be changed at match time, I realized I could make a small
1867        improvement to matching performance by compiling OP_ALLANY instead of
1868        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1869        on the OP_ANY path.
1870    
1871    17. Compiling pcretest on Windows with readline support failed without the
1872        following two fixes: (1) Make the unistd.h include conditional on
1873        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1874    
1875    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1876        ncurses library to be included for pcretest when ReadLine support is
1877        requested, but also to allow for it to be overridden. This patch came from
1878        Daniel Bergström.
1879    
1880    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1881        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1882        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1883        spotting this.
1884    
1885    
1886    Version 7.6 28-Jan-08
1887    ---------------------
1888    
1889    1.  A character class containing a very large number of characters with
1890        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1891        overflow.
1892    
1893    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1894        HAVE_LONG_LONG is not defined.
1895    
1896    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1897        bring them up to date and include new features. This patch includes:
1898    
1899        - Fixed PH's badly added libz and libbz2 support.
1900        - Fixed a problem with static linking.
1901        - Added pcredemo. [But later removed - see 7 below.]
1902        - Fixed dftables problem and added an option.
1903        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1904            HAVE_LONG_LONG.
1905        - Added readline support for pcretest.
1906        - Added an listing of the option settings after cmake has run.
1907    
1908    4.  A user submitted a patch to Makefile that makes it easy to create
1909        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1910        Makefile.am that cause it to include this special target, without
1911        affecting anything else. Note that the same mingw target plus all
1912        the other distribution libraries and programs are now supported
1913        when configuring with CMake (see 6 below) instead of with
1914        Configure/Make.
1915    
1916    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1917        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1918        exported in the Windows port". It has not yet been confirmed that the patch
1919        solves the problem, but it does no harm.
1920    
1921    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1922        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1923        with CMake, and also correct the comment about stack recursion.
1924    
1925    7.  Remove the automatic building of pcredemo from the ./configure system and
1926        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1927        of a program that users should build themselves after PCRE is installed, so
1928        building it automatically is not really right. What is more, it gave
1929        trouble in some build environments.
1930    
1931    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1932    
1933    
1934    Version 7.5 10-Jan-08
1935    ---------------------
1936    
1937    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1938        values in parens when parsing an RE using the C++ wrapper."
1939    
1940    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1941        Characters greater than 255 were excluded from the class instead of being
1942        included.
1943    
1944    3.  The same bug as (2) above applied to negated POSIX classes such as
1945        [:^space:].
1946    
1947    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1948        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1949        I have changed it.
1950    
1951    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1952        first named subpattern) and a construct such as (?&a) would reference the
1953        first named subpattern whose name started with "a" (in other words, the
1954        length check was missing). Both these problems are fixed. "Subpattern name
1955        expected" is now given for (?&) (a zero-length name), and this patch also
1956        makes it give the same error for \k'' (previously it complained that that
1957        was a reference to a non-existent subpattern).
1958    
1959    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1960        this is right because (?- can be followed by option settings as well as by
1961        digits. I have, however, made the messages clearer.
1962    
1963    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1964        than the number used in the conditional) now cause a compile-time error.
1965        This is actually not compatible with Perl, which accepts such patterns, but
1966        treats the conditional as always being FALSE (as PCRE used to), but it
1967        seems to me that giving a diagnostic is better.
1968    
1969    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1970        and messages.
1971    
1972    9.  Fix two occurrences of "backslash" in comments that should have been
1973        "backspace".
1974    
1975    10. Remove two redundant lines of code that can never be obeyed (their function
1976        was moved elsewhere).
1977    
1978    11. The program that makes PCRE's Unicode character property table had a bug
1979        which caused it to generate incorrect table entries for sequences of
1980        characters that have the same character type, but are in different scripts.
1981        It amalgamated them into a single range, with the script of the first of
1982        them. In other words, some characters were in the wrong script. There were
1983        thirteen such cases, affecting characters in the following ranges:
1984    
1985          U+002b0 - U+002c1
1986          U+0060c - U+0060d
1987          U+0061e - U+00612
1988          U+0064b - U+0065e
1989          U+0074d - U+0076d
1990          U+01800 - U+01805
1991          U+01d00 - U+01d77
1992          U+01d9b - U+01dbf
1993          U+0200b - U+0200f
1994          U+030fc - U+030fe
1995          U+03260 - U+0327f
1996          U+0fb46 - U+0fbb1
1997          U+10450 - U+1049d
1998    
1999    12. The -o option (show only the matching part of a line) for pcregrep was not
2000        compatible with GNU grep in that, if there was more than one match in a
2001        line, it showed only the first of them. It now behaves in the same way as
2002        GNU grep.
2003    
2004    13. If the -o and -v options were combined for pcregrep, it printed a blank
2005        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2006        does the same. The return code can be used to tell if there were any
2007        non-matching lines.
2008    
2009    14. Added --file-offsets and --line-offsets to pcregrep.
2010    
2011    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2012        infinitely looping recursion. The bug was that positive lookaheads were not
2013        being skipped when checking for a possible empty match (negative lookaheads
2014        and both kinds of lookbehind were skipped).
2015    
2016    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2017        inclusion of <windows.h> to before rather than after the definition of
2018        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2019    
2020    17. Specifying a possessive quantifier with a specific limit for a Unicode
2021        character property caused pcre_compile() to compile bad code, which led at
2022        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2023        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2024        caused the error; without that there was no problem.
2025    
2026    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2027    
2028    19. Added --enable-pcretest-libreadline.
2029    
2030    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2031        RE::GlobalReplace(). As a result, the number of replacements returned was
2032        double what it should be. I removed one of the increments, but Craig sent a
2033        later patch that removed the other one (the right fix) and added unit tests
2034        that check the return values (which was not done before).
2035    
2036    21. Several CMake things:
2037    
2038        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2039            the names libpcre and libpcreposix, not just pcre and pcreposix.
2040    
2041        (2) The above change means that pcretest and pcregrep are now correctly
2042            linked with the newly-built libraries, not previously installed ones.
2043    
2044        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2045    
2046    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2047        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2048        UTF-8 newline character). The key issue is that the pattern starts .*;
2049        this means that the match must be either at the beginning, or after a
2050        newline. The bug was in the code for advancing after a failed match and
2051        checking that the new position followed a newline. It was not taking
2052        account of UTF-8 characters correctly.
2053    
2054    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2055        character classes. PCRE was not treating the sequence [:...:] as a
2056        character class unless the ... were all letters. Perl, however, seems to
2057        allow any characters between [: and :], though of course it rejects as
2058        unknown any "names" that contain non-letters, because all the known class
2059        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2060        for example, whereas PCRE did not - it did not recognize a POSIX character
2061        class. This seemed a bit dangerous, so the code has been changed to be
2062        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2063        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2064        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2065        Perl does, and where it didn't before.
2066    
2067    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2068        Windows environments %n is disabled by default.
2069    
2070    
2071    Version 7.4 21-Sep-07
2072    ---------------------
2073    
2074    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2075        means that a class such as [\s] counted as "explicit reference to CR or
2076        LF". That isn't really right - the whole point of the change was to try to
2077        help when there was an actual mention of one of the two characters. So now
2078        the change happens only if \r or \n (or a literal CR or LF) character is
2079        encountered.
2080    
2081    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2082        of both had grown to the point where there were only 3 bits left.
2083        Fortunately, there was spare space in the data structure, and so I have
2084        moved the internal flags into a new 16-bit field to free up more option
2085        bits.
2086    
2087    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2088        but did not set the internal JCHANGED flag - either of these is enough to
2089        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2090        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2091        start sets both bits.
2092    
2093    4.  Added options (at build time, compile time, exec time) to change \R from
2094        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2095    
2096    5.  doc/pcresyntax.html was missing from the distribution.
2097    
2098    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2099        compatibility, even though it is no longer used.
2100    
2101    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2102        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2103        windows.h file is present (where different names are used). [This was
2104        reversed later after testing - see 16 below.]
2105    
2106    8.  Changed all #include <config.h> to #include "config.h". There were also
2107        some further <pcre.h> cases that I changed to "pcre.h".
2108    
2109    9.  When pcregrep was used with the --colour option, it missed the line ending
2110        sequence off the lines that it output.
2111    
2112    10. It was pointed out to me that arrays of string pointers cause lots of
2113        relocations when a shared library is dynamically loaded. A technique of
2114        using a single long string with a table of offsets can drastically reduce
2115        these. I have refactored PCRE in four places to do this. The result is
2116        dramatic:
2117    
2118          Originally:                          290
2119          After changing UCP table:            187
2120          After changing error message table:   43
2121          After changing table of "verbs"       36
2122          After changing table of Posix names   22
2123    
2124        Thanks to the folks working on Gregex for glib for this insight.
2125    
2126    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2127        unicode-properties was also set.
2128    
2129    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2130    
2131    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2132        checked only for CRLF.
2133    
2134    14. Added casts to pcretest.c to avoid compiler warnings.
2135    
2136    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2137    
2138    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2139        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2140        entirely. This removes changes made in 7 above.
2141    
2142    17. The CMake files have been updated, and there is now more information about
2143        building with CMake in the NON-UNIX-USE document.
2144    
2145    
2146    Version 7.3 28-Aug-07
2147    ---------------------
2148    
2149     1. In the rejigging of the build system that eventually resulted in 7.1, the
2150        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2151        brackets there is not right, since it causes compilers to look for an
2152        installed pcre.h, not the version that is in the source that is being
2153        compiled (which of course may be different). I have changed it back to:
2154    
2155          #include "pcre.h"
2156    
2157        I have a vague recollection that the change was concerned with compiling in
2158        different directories, but in the new build system, that is taken care of
2159        by the VPATH setting the Makefile.
2160    
2161     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2162        when the subject happened to end in the byte 0x85 (e.g. if the last
2163        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2164        characters but of course it shouldn't be taken as a newline when it is part
2165        of another character. The bug was that, for an unlimited repeat of . in
2166        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2167        characters when looking for a newline.
2168    
2169     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2170    
2171     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2172        in debug output.
2173    
2174     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2175        long printing in the pcrecpp unittest when running under MinGW.
2176    
2177     6. ESC_K was left out of the EBCDIC table.
2178    
2179     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2180        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2181        limit also applies to "virtual nesting" when a pattern is recursive, and in
2182        this case 1000 isn't so big. I have been able to remove this limit at the
2183        expense of backing off one optimization in certain circumstances. Normally,
2184        when pcre_exec() would call its internal match() function recursively and
2185        immediately return the result unconditionally, it uses a "tail recursion"
2186        feature to save stack. However, when a subpattern that can match an empty
2187        string has an unlimited repetition quantifier, it no longer makes this
2188        optimization. That gives it a stack frame in which to save the data for
2189        checking that an empty string has been matched. Previously this was taken
2190        from the 1000-entry workspace that had been reserved. So now there is no
2191        explicit limit, but more stack is used.
2192    
2193     8. Applied Daniel's patches to solve problems with the import/export magic
2194        syntax that is required for Windows, and which was going wrong for the
2195        pcreposix and pcrecpp parts of the library. These were overlooked when this
2196        problem was solved for the main library.
2197    
2198     9. There were some crude static tests to avoid integer overflow when computing
2199        the size of patterns that contain repeated groups with explicit upper
2200        limits. As the maximum quantifier is 65535, the maximum group length was
2201        set at 30,000 so that the product of these two numbers did not overflow a
2202        32-bit integer. However, it turns out that people want to use groups that
2203        are longer than 30,000 bytes (though not repeat them that many times).
2204        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2205        made it possible to implement the integer overflow checks in a much more
2206        dynamic way, which I have now done. The artificial limitation on group
2207        length has been removed - we now have only the limit on the total length of
2208        the compiled pattern, which depends on the LINK_SIZE setting.
2209    
2210    10. Fixed a bug in the documentation for get/copy named substring when
2211        duplicate names are permitted. If none of the named substrings are set, the
2212        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2213        empty string.
2214    
2215    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2216        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2217        because the ] is interpreted as the first data character and the
2218        terminating ] is not found. PCRE has been made compatible with Perl in this
2219        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2220        cause memory overwriting.
2221    
2222    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2223        string has been matched (to stop an infinite loop). It was not recognizing
2224        a conditional subpattern that could match an empty string if that
2225        subpattern was within another subpattern. For example, it looped when
2226        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2227        condition was not nested. This bug has been fixed.
2228    
2229    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2230        past the start of the subject in the presence of bytes with the top bit
2231        set, for example "\x8aBCD".
2232    
2233    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2234        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2235    
2236    14. Optimized (?!) to (*FAIL).
2237    
2238    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2239        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2240        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2241        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2242        does: it's just the validity check that is more restrictive.
2243    
2244    16. Inserted checks for integer overflows during escape sequence (backslash)
2245        processing, and also fixed erroneous offset values for syntax errors during
2246        backslash processing.
2247    
2248    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2249        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2250    
2251    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2252        caused an overrun.
2253    
2254    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2255        something other than just ASCII characters) inside a group that had an
2256        unlimited repeat caused a loop at compile time (while checking to see
2257        whether the group could match an empty string).
2258    
2259    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2260        [\P{Any}] did so. (Error in the code for printing property names.)
2261    
2262    21. An orphan \E inside a character class could cause a crash.
2263    
2264    22. A repeated capturing bracket such as (A)? could cause a wild memory
2265        reference during compilation.
2266    
2267    23. There are several functions in pcre_compile() that scan along a compiled
2268        expression for various reasons (e.g. to see if it's fixed length for look
2269        behind). There were bugs in these functions when a repeated \p or \P was
2270        present in the pattern. These operators have additional parameters compared
2271        with \d, etc, and these were not being taken into account when moving along
2272        the compiled data. Specifically:
2273    
2274        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2275            length.
2276    
2277        (b) An item such as \pL+ within a repeated group could cause crashes or
2278            loops.
2279    
2280        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2281            "reference to non-existent subpattern" error.
2282    
2283        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2284    
2285    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2286        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2287    
2288    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2289    
2290    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2291        character were causing crashes (broken optimization).
2292    
2293    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2294        \p or \P) caused a compile-time loop.
2295    
2296    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2297        break. For example, the unstudied pattern [\r\n]A does not match the string
2298        "\r\nA" because change 7.0/46 below moves the current point on by two
2299        characters after failing to match at the start. However, the pattern \nA
2300        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2301        the same is true. There doesn't seem any very clean way out of this, but
2302        what I have chosen to do makes the common cases work: PCRE now takes note
2303        of whether there can be an explicit match for \r or \n anywhere in the
2304        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2305        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2306        pattern has explicit CR or LF references.
2307    
2308    29. Added (*CR) etc for changing newline setting at start of pattern.
2309    
2310    
2311    Version 7.2 19-Jun-07
2312    ---------------------
2313    
2314     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2315        which is apparently normally available under Windows.
2316    
2317     2. Re-jig the pcregrep tests with different newline settings in an attempt
2318        to make them independent of the local environment's newline setting.
2319    
2320     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2321    
2322     4. Some of the "internals" tests were previously cut out when the link size
2323        was not 2, because the output contained actual offsets. The recent new
2324        "Z" feature of pcretest means that these can be cut out, making the tests
2325        usable with all link sizes.
2326    
2327     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2328        stack recursion. This gives a massive performance boost under BSD, but just
2329        a small improvement under Linux. However, it saves one field in the frame
2330        in all cases.
2331    
2332     6. Added more features from the forthcoming Perl 5.10:
2333    
2334        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2335            recursion call. It refers to the nth most recently opened parentheses.
2336    
2337        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2338            to be opened parentheses.
2339    
2340        (c) Conditions that refer to capturing parentheses can be specified
2341            relatively, for example, (?(-2)... or (?(+3)...
2342    
2343        (d) \K resets the start of the current match so that everything before
2344            is not part of it.
2345    
2346        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2347    
2348        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2349            reference syntax.
2350    
2351        (g) (?| introduces a group in which the numbering of parentheses in each
2352            alternative starts with the same number.
2353    
2354        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2355    
2356     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2357        PCRE_INFO_JCHANGED.
2358    
2359     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2360        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2361        for detecting groups that can match an empty string.
2362    
2363     9. A pattern with a very large number of alternatives (more than several
2364        hundred) was running out of internal workspace during the pre-compile
2365        phase, where pcre_compile() figures out how much memory will be needed. A
2366        bit of new cunning has reduced the workspace needed for groups with
2367        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2368        workspace instead of running out of the 4096 that are available.
2369    
2370    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2371    
2372    11. Applied patch from Google to remove an optimization that didn't quite work.
2373        The report of the bug said:
2374    
2375          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2376          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2377          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2378    
2379    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2380        it matched the wrong number of bytes.
2381    
2382    
2383    Version 7.1 24-Apr-07
2384  ---------------------  ---------------------
2385    
2386   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2387      that is more "standard", making use of automake and other autotools.      that is more "standard", making use of automake and other Autotools. There
2388        is some re-arrangement of the files and adjustment of comments consequent
2389        on this.
2390    
2391     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2392        for recursive directory scanning broke on some systems because the files
2393        are not scanned in any specific order and on different systems the order
2394        was different. A call to "sort" has been inserted into RunGrepTest for the
2395        approprate test as a short-term fix. In the longer term there may be an
2396        alternative.
2397    
2398     3. I had an email from Eric Raymond about problems translating some of PCRE's
2399        man pages to HTML (despite the fact that I distribute HTML pages, some
2400        people do their own conversions for various reasons). The problems
2401        concerned the use of low-level troff macros .br and .in. I have therefore
2402        removed all such uses from the man pages (some were redundant, some could
2403        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2404        HTML has been updated to handle .nf/.fi and to complain if it encounters
2405        .br or .in.
2406    
2407     4. Updated comments in configure.ac that get placed in config.h.in and also
2408        arranged for config.h to be included in the distribution, with the name
2409        config.h.generic, for the benefit of those who have to compile without
2410        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2411    
2412     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2413        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2414        makevp.bat for the current PCRE, using the additional files
2415        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2416    
2417     6. A Windows user reported a minor discrepancy with test 2, which turned out
2418        to be caused by a trailing space on an input line that had got lost in his
2419        copy. The trailing space was an accident, so I've just removed it.
2420    
2421     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2422        that is needed.
2423    
2424     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2425        as "const" (a) because they are and (b) because it helps the PHP
2426        maintainers who have recently made a script to detect big data structures
2427        in the php code that should be moved to the .rodata section. I remembered
2428        to update Builducptable as well, so it won't revert if ucptable.h is ever
2429        re-created.
2430    
2431     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2432        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2433        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2434        support is not required. This saves 1.5-2K of code, which is important in
2435        some applications.
2436    
2437        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2438        so as not to refer to the tables, even though these functions will never be
2439        called when UTF-8 support is disabled. Otherwise there are problems with a
2440        shared library.
2441    
2442    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2443    
2444        (a) It was defining its arguments as char * instead of void *.
2445    
2446        (b) It was assuming that all moves were upwards in memory; this was true
2447            a long time ago when I wrote it, but is no longer the case.
2448    
2449        The emulated memove() is provided for those environments that have neither
2450        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2451        is clearly not the case, as these two bugs were recently reported.
2452    
2453    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2454        and Detrail to create the HTML documentation, the .txt form of the man
2455        pages, and it removes trailing spaces from listed files. It also creates
2456        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2457        case, it wraps all the #defines with #ifndefs. This script should be run
2458        before "make dist".
2459    
2460    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2461        with Unicode property support.
2462    
2463        (a) For a maximizing quantifier, if the two different cases of the
2464            character were of different lengths in their UTF-8 codings (there are
2465            some cases like this - I found 11), and the matching function had to
2466            back up over a mixture of the two cases, it incorrectly assumed they
2467            were both the same length.
2468    
2469        (b) When PCRE was configured to use the heap rather than the stack for
2470            recursion during matching, it was not correctly preserving the data for
2471            the other case of a UTF-8 character when checking ahead for a match
2472            while processing a minimizing repeat. If the check also involved
2473            matching a wide character, but failed, corruption could cause an
2474            erroneous result when trying to check for a repeat of the original
2475            character.
2476    
2477    13. Some tidying changes to the testing mechanism:
2478    
2479        (a) The RunTest script now detects the internal link size and whether there
2480            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2481            values substituted by "configure". (The RunGrepTest script already did
2482            this for UTF-8.) The configure.ac script no longer substitutes the
2483            relevant variables.
2484    
2485        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2486            with length and offset values. This means that the output is different
2487            for different internal link sizes. Test 2 is skipped for link sizes
2488            other than 2 because of this, bypassing the problem. Unfortunately,
2489            there was also a test in test 3 (the locale tests) that used /B and
2490            failed for link sizes other than 2. Rather than cut the whole test out,
2491            I have added a new /Z option to pcretest that replaces the length and
2492            offset values with spaces. This is now used to make test 3 independent
2493            of link size. (Test 2 will be tidied up later.)
2494    
2495    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2496        segmentation fault instead of returning the appropriate error message.
2497    
2498    15. In multiline mode when the newline sequence was set to "any", the pattern
2499        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2500        This doesn't seem right; it now treats the CRLF combination as the line
2501        ending, and so does not match in that case. It's only a pattern such as ^$
2502        that would hit this one: something like ^ABC$ would have failed after \r
2503        and then tried again after \r\n.
2504    
2505    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2506        in an attempt to make files that differ only in their line terminators
2507        compare equal. This works on Linux.
2508    
2509    17. Under certain error circumstances pcregrep might try to free random memory
2510        as it exited. This is now fixed, thanks to valgrind.
2511    
2512    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2513        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2514        was because its rules for how to advance for /g after matching an empty
2515        string at the end of a line did not allow for this case. They now check for
2516        it specially.
2517    
2518    20. pcretest is supposed to handle patterns and data of any length, by
2519        extending its buffers when necessary. It was getting this wrong when the
2520        buffer for a data line had to be extended.
2521    
2522    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2523        CRLF as a newline sequence.
2524    
2525    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2526        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2527        I have nevertheless tidied it up.
2528    
2529    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2530    
2531    24. Added a man page for pcre-config.
2532    
2533    
2534  Version 7.0 19-Dec-06  Version 7.0 19-Dec-06

Legend:
Removed from v.98  
changed lines
  Added in v.1273

  ViewVC Help
Powered by ViewVC 1.1.5