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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.312  
changed lines
  Added in v.873

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12