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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.261  
changed lines
  Added in v.919

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12