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

Diff of /code/trunk/ChangeLog

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

revision 149 by ph10, Mon Apr 16 15:28:08 2007 UTC revision 701 by ph10, Tue Sep 20 11:30:56 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.1 12-Mar-07  Version 8.20 12-Sep-2011
5    ------------------------
6    
7    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
8        a POSIX class. After further experiments with Perl, which convinced me that
9        Perl has bugs and confusions, a closing square bracket is no longer allowed
10        in a POSIX name. This bug also affected patterns with classes that started
11        with full stops.
12    
13    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
14        captured substring, but while checking the failing first alternative,
15        substring 1 is temporarily captured. If the output vector supplied to
16        pcre_exec() was not big enough for this capture, the yield of the function
17        was still zero ("insufficient space for captured substrings"). This cannot
18        be totally fixed without adding another stack variable, which seems a lot
19        of expense for a edge case. However, I have improved the situation in cases
20        such as /(a)(b)x|abc/ matched against "abc", where the return code
21        indicates that fewer than the maximum number of slots in the ovector have
22        been set.
23    
24    3.  Related to (2) above: when there are more back references in a pattern than
25        slots in the output vector, pcre_exec() uses temporary memory during
26        matching, and copies in the captures as far as possible afterwards. It was
27        using the entire output vector, but this conflicts with the specification
28        that only 2/3 is used for passing back captured substrings. Now it uses
29        only the first 2/3, for compatibility. This is, of course, another edge
30        case.
31    
32    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
33        main code base, and can be used by building with --enable-jit. When this is
34        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
35        runtime --no-jit option is given.
36    
37    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
38        ovector, the return from the function was zero, implying that there were
39        other matches that did not fit. The correct "exactly full" value is now
40        returned.
41    
42    6.  If a subpattern that was called recursively or as a subroutine contained
43        (*PRUNE) or any other control that caused it to give a non-standard return,
44        invalid errors such as "Error -26 (nested recursion at the same subject
45        position)" or even infinite loops could occur.
46    
47    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
48        computing the minimum length on reaching *ACCEPT, and so ended up with the
49        wrong value of 1 rather than 0. Further investigation indicates that
50        computing a minimum subject length in the presence of *ACCEPT is difficult
51        (think back references, subroutine calls), and so I have changed the code
52        so that no minimum is registered for a pattern that contains *ACCEPT.
53    
54    8.  If (*THEN) was present in the first (true) branch of a conditional group,
55        it was not handled as intended.
56    
57    9.  Replaced RunTest.bat with the much improved version provided by Sheri
58        Pierce.
59    
60    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
61        the first byte in a match must be "a".
62    
63    
64    Version 8.13 16-Aug-2011
65    ------------------------
66    
67    1.  The Unicode data tables have been updated to Unicode 6.0.0.
68    
69    2.  Two minor typos in pcre_internal.h have been fixed.
70    
71    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
72        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
73        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
74    
75    4.  There were a number of related bugs in the code for matching backrefences
76        caselessly in UTF-8 mode when codes for the characters concerned were
77        different numbers of bytes. For example, U+023A and U+2C65 are an upper
78        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
79        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
80        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
81        2-byte code at the end of the subject (it thought there wasn't enough data
82        left).
83    
84    5.  Comprehensive information about what went wrong is now returned by
85        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
86        as the output vector has at least 2 elements. The offset of the start of
87        the failing character and a reason code are placed in the vector.
88    
89    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
90        now returned is for the first byte of the failing character, instead of the
91        last byte inspected. This is an incompatible change, but I hope it is small
92        enough not to be a problem. It makes the returned offset consistent with
93        pcre_exec() and pcre_dfa_exec().
94    
95    7.  pcretest now gives a text phrase as well as the error number when
96        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
97        failure, the offset and reason code are output.
98    
99    8.  When \R was used with a maximizing quantifier it failed to skip backwards
100        over a \r\n pair if the subsequent match failed. Instead, it just skipped
101        back over a single character (\n). This seems wrong (because it treated the
102        two characters as a single entity when going forwards), conflicts with the
103        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
104        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
105        has been changed.
106    
107    9.  Some internal refactoring has changed the processing so that the handling
108        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
109        time (the PCRE_DOTALL option was changed this way some time ago: version
110        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
111        which was always a bit of a fudge. It also means that there is one less
112        argument for the match() function, which reduces its stack requirements
113        slightly. This change also fixes an incompatibility with Perl: the pattern
114        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
115    
116    10. More internal refactoring has drastically reduced the number of recursive
117        calls to match() for possessively repeated groups such as (abc)++ when
118        using pcre_exec().
119    
120    11. While implementing 10, a number of bugs in the handling of groups were
121        discovered and fixed:
122    
123        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
124        (a|)*(?1) gave a compile-time internal error.
125        ((a|)+)+  did not notice that the outer group could match an empty string.
126        (^a|^)+   was not marked as anchored.
127        (.*a|.*)+ was not marked as matching at start or after a newline.
128    
129    12. Yet more internal refactoring has removed another argument from the match()
130        function. Special calls to this function are now indicated by setting a
131        value in a variable in the "match data" data block.
132    
133    13. Be more explicit in pcre_study() instead of relying on "default" for
134        opcodes that mean there is no starting character; this means that when new
135        ones are added and accidentally left out of pcre_study(), testing should
136        pick them up.
137    
138    14. The -s option of pcretest has been documented for ages as being an old
139        synonym of -m (show memory usage). I have changed it to mean "force study
140        for every regex", that is, assume /S for every regex. This is similar to -i
141        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
142        using it. It makes it easier to run collections of tests with and without
143        study enabled, and thereby test pcre_study() more easily. All the standard
144        tests are now run with and without -s (but some patterns can be marked as
145        "never study" - see 20 below).
146    
147    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
148        restoration of the capturing data to the outer values was not happening
149        correctly.
150    
151    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
152        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
153        pattern had matched an empty string, and so incorrectly returned a no
154        match.
155    
156    17. There was optimizing code for the last branch of non-capturing parentheses,
157        and also for the obeyed branch of a conditional subexpression, which used
158        tail recursion to cut down on stack usage. Unfortunately, now that there is
159        the possibility of (*THEN) occurring in these branches, tail recursion is
160        no longer possible because the return has to be checked for (*THEN). These
161        two optimizations have therefore been removed.
162    
163    18. If a pattern containing \R was studied, it was assumed that \R always
164        matched two bytes, thus causing the minimum subject length to be
165        incorrectly computed because \R can also match just one byte.
166    
167    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
168        was incorrectly computed.
169    
170    20. If /S is present twice on a test pattern in pcretest input, it now
171        *disables* studying, thereby overriding the use of -s on the command line
172        (see 14 above). This is necessary for one or two tests to keep the output
173        identical in both cases.
174    
175    21. When (*ACCEPT) was used in an assertion that matched an empty string and
176        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
177    
178    22. When an atomic group that contained a capturing parenthesis was
179        successfully matched, but the branch in which it appeared failed, the
180        capturing was not being forgotten if a higher numbered group was later
181        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
182        group 1 to "a", when in fact it should be unset. This applied to multi-
183        branched capturing and non-capturing groups, repeated or not, and also to
184        positive assertions (capturing in negative assertions does not happen
185        in PCRE) and also to nested atomic groups.
186    
187    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
188        subject after a captured substring, to make it easier to tell which of a
189        number of identical substrings has been captured.
190    
191    24. The way atomic groups are processed by pcre_exec() has been changed so that
192        if they are repeated, backtracking one repetition now resets captured
193        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
194        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
195        "aaa". Previously, it would have been "a". As part of this code
196        refactoring, the way recursive calls are handled has also been changed.
197    
198    25. If an assertion condition captured any substrings, they were not passed
199        back unless some other capturing happened later. For example, if
200        (?(?=(a))a) was matched against "a", no capturing was returned.
201    
202    26. When studying a pattern that contained subroutine calls or assertions,
203        the code for finding the minimum length of a possible match was handling
204        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
205        group 1 called group 2 while simultaneously a separate group 2 called group
206        1). A stack overflow occurred in this case. I have fixed this by limiting
207        the recursion depth to 10.
208    
209    27. Updated RunTest.bat in the distribution to the version supplied by Tom
210        Fortmann. This supports explicit test numbers on the command line, and has
211        argument validation and error reporting.
212    
213    28. An instance of \X with an unlimited repeat could fail if at any point the
214        first character it looked at was a mark character.
215    
216    29. Some minor code refactoring concerning Unicode properties and scripts
217        should reduce the stack requirement of match() slightly.
218    
219    30. Added the '=' option to pcretest to check the setting of unused capturing
220        slots at the end of the pattern, which are documented as being -1, but are
221        not included in the return count.
222    
223    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
224        compiled something random. Now it gives a compile-time error (as does
225        Perl).
226    
227    32. A *MARK encountered during the processing of a positive assertion is now
228        recorded and passed back (compatible with Perl).
229    
230    33. If --only-matching or --colour was set on a pcregrep call whose pattern
231        had alternative anchored branches, the search for a second match in a line
232        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
233        matched the line "0102" twice. The same bug affected patterns that started
234        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
235        twice.
236    
237    34. Previously, PCRE did not allow quantification of assertions. However, Perl
238        does, and because of capturing effects, quantifying parenthesized
239        assertions may at times be useful. Quantifiers are now allowed for
240        parenthesized assertions.
241    
242    35. A minor code tidy in pcre_compile() when checking options for \R usage.
243    
244    36. \g was being checked for fancy things in a character class, when it should
245        just be a literal "g".
246    
247    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
248        appearance of a nested POSIX class supersedes an apparent external class.
249        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
250        unescaped square brackets may also appear as part of class names. For
251        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
252        more like Perl. (But see 8.20/1 above.)
253    
254    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
255        was because it thought it was \N{name}, which is not supported).
256    
257    39. Add minix to OS list not supporting the -S option in pcretest.
258    
259    40. PCRE tries to detect cases of infinite recursion at compile time, but it
260        cannot analyze patterns in sufficient detail to catch mutual recursions
261        such as ((?1))((?2)). There is now a runtime test that gives an error if a
262        subgroup is called recursively as a subpattern for a second time at the
263        same position in the subject string. In previous releases this might have
264        been caught by the recursion limit, or it might have run out of stack.
265    
266    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
267        happen only once. PCRE was, however incorrectly giving a compile time error
268        "recursive call could loop indefinitely" because it cannot analyze the
269        pattern in sufficient detail. The compile time test no longer happens when
270        PCRE is compiling a conditional subpattern, but actual runaway loops are
271        now caught at runtime (see 40 above).
272    
273    42. It seems that Perl allows any characters other than a closing parenthesis
274        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
275        has been changed to be the same.
276    
277    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
278        as not to get warnings when autogen.sh is called. Also changed
279        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
280    
281    44. To help people who use pcregrep to scan files containing exceedingly long
282        lines, the following changes have been made:
283    
284        (a) The default value of the buffer size parameter has been increased from
285            8K to 20K. (The actual buffer used is three times this size.)
286    
287        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
288            PCRE is built.
289    
290        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
291            to be set at run time.
292    
293        (d) Numerical values in pcregrep options can be followed by K or M, for
294            example --buffer-size=50K.
295    
296        (e) If a line being scanned overflows pcregrep's buffer, an error is now
297            given and the return code is set to 2.
298    
299    45. Add a pointer to the latest mark to the callout data block.
300    
301    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
302        partial match of an empty string instead of no match. This was specific to
303        the use of ".".
304    
305    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
306        complete match instead of a partial match. This bug was dependent on both
307        the PCRE_UTF8 and PCRE_DOTALL options being set.
308    
309    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
310        starting byte set, because \b was not being ignored.
311    
312    
313    Version 8.12 15-Jan-2011
314    ------------------------
315    
316    1.  Fixed some typos in the markup of the man pages, and wrote a script that
317        checks for such things as part of the documentation building process.
318    
319    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
320        --match-limit and --recursion-limit options (added for 8.11). In
321        particular, this made one of the standard tests fail. (The integer value
322        went into the wrong half of a long int.)
323    
324    3.  If the --colour option was given to pcregrep with -v (invert match), it
325        did strange things, either producing crazy output, or crashing. It should,
326        of course, ignore a request for colour when reporting lines that do not
327        match.
328    
329    4.  Another pcregrep bug caused similar problems if --colour was specified with
330        -M (multiline) and the pattern match finished with a line ending.
331    
332    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
333        matched in multiline mode, the following line was shown as part of the
334        match. This seems wrong, so I have changed it.
335    
336    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
337        the check for further matches in the same line (so they could be coloured)
338        to overrun the end of the current line. If another match was found, it was
339        incorrectly shown (and then shown again when found in the next line).
340    
341    7.  If pcregrep was compiled under Windows, there was a reference to the
342        function pcregrep_exit() before it was defined. I am assuming this was
343        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
344        reported by a user. I've moved the definition above the reference.
345    
346    
347    Version 8.11 10-Dec-2010
348    ------------------------
349    
350    1.  (*THEN) was not working properly if there were untried alternatives prior
351        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
352        backtracked to try for "b" instead of moving to the next alternative branch
353        at the same level (in this case, to look for "c"). The Perl documentation
354        is clear that when (*THEN) is backtracked onto, it goes to the "next
355        alternative in the innermost enclosing group".
356    
357    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
358        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
359        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
360        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
361        (*THEN).
362    
363    3.  If \s appeared in a character class, it removed the VT character from
364        the class, even if it had been included by some previous item, for example
365        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
366        of \s, but is part of the POSIX "space" class.)
367    
368    4.  A partial match never returns an empty string (because you can always
369        match an empty string at the end of the subject); however the checking for
370        an empty string was starting at the "start of match" point. This has been
371        changed to the "earliest inspected character" point, because the returned
372        data for a partial match starts at this character. This means that, for
373        example, /(?<=abc)def/ gives a partial match for the subject "abc"
374        (previously it gave "no match").
375    
376    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
377        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
378        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
379        has an implication that the given string is incomplete (because a partial
380        match is preferred over a full match). For this reason, these items now
381        give a partial match in this situation. [Aside: previously, the one case
382        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
383        match rather than a full match, which was wrong by the old rules, but is
384        now correct.]
385    
386    6.  There was a bug in the handling of #-introduced comments, recognized when
387        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
388        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
389        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
390        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
391        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
392        places in pcre_compile().
393    
394    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
395        comments when looking ahead for named forward references to subpatterns,
396        the only newline sequence it recognized was NL. It now handles newlines
397        according to the set newline convention.
398    
399    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
400        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
401        cater for a lack of strerror(). These oversights have been fixed.
402    
403    9.  Added --match-limit and --recursion-limit to pcregrep.
404    
405    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
406    
407    11. When the -o option was used, pcregrep was setting a return code of 1, even
408        when matches were found, and --line-buffered was not being honoured.
409    
410    12. Added an optional parentheses number to the -o and --only-matching options
411        of pcregrep.
412    
413    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
414        can match an empty string. The code to do it in pcretest and pcredemo
415        needed fixing:
416    
417        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
418            only one byte after an empty string match just before CRLF (this case
419            just got forgotten; "any" and "anycrlf" were OK).
420    
421        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
422            mode when an empty string match preceded an ASCII character followed by
423            a non-ASCII character. (The code for advancing by one character rather
424            than one byte was nonsense.)
425    
426        (c) The pcredemo.c sample program did not have any code at all to handle
427            the cases when CRLF is a valid newline sequence.
428    
429    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
430        as a starting offset was within the subject string. There is now a new
431        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
432        negative or greater than the length of the string. In order to test this,
433        pcretest is extended to allow the setting of negative starting offsets.
434    
435    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
436        starting offset points to the beginning of a UTF-8 character was
437        unnecessarily clumsy. I tidied it up.
438    
439    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
440        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
441    
442    17. Nobody had reported that the --include_dir option, which was added in
443        release 7.7 should have been called --include-dir (hyphen, not underscore)
444        for compatibility with GNU grep. I have changed it to --include-dir, but
445        left --include_dir as an undocumented synonym, and the same for
446        --exclude-dir, though that is not available in GNU grep, at least as of
447        release 2.5.4.
448    
449    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
450        characters from a string of bytes) have been redefined so as not to use
451        loops, in order to improve performance in some environments. At the same
452        time, I abstracted some of the common code into auxiliary macros to save
453        repetition (this should not affect the compiled code).
454    
455    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
456        compile-time error is now given if \c is not followed by an ASCII
457        character, that is, a byte less than 128. (In EBCDIC mode, the code is
458        different, and any byte value is allowed.)
459    
460    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
461        START_OPTIMIZE option, which is now allowed at compile time - but just
462        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
463        to pcregrep and other applications that have no direct access to PCRE
464        options. The new /Y option in pcretest sets this option when calling
465        pcre_compile().
466    
467    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
468        back references. Groups containing recursive back references were forced to
469        be atomic by that change, but in the case of named groups, the amount of
470        memory required was incorrectly computed, leading to "Failed: internal
471        error: code overflow". This has been fixed.
472    
473    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
474        pcretest.c, to avoid build problems in some Borland environments.
475    
476    
477    Version 8.10 25-Jun-2010
478    ------------------------
479    
480    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
481        THEN.
482    
483    2.  (*ACCEPT) was not working when inside an atomic group.
484    
485    3.  Inside a character class, \B is treated as a literal by default, but
486        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
487        causes the error). The code is unchanged, but I tidied the documentation.
488    
489    4.  Inside a character class, PCRE always treated \R and \X as literals,
490        whereas Perl faults them if its -w option is set. I have changed PCRE so
491        that it faults them when PCRE_EXTRA is set.
492    
493    5.  Added support for \N, which always matches any character other than
494        newline. (It is the same as "." when PCRE_DOTALL is not set.)
495    
496    6.  When compiling pcregrep with newer versions of gcc which may have
497        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
498        declared with attribute warn_unused_result" were given. Just casting the
499        result to (void) does not stop the warnings; a more elaborate fudge is
500        needed. I've used a macro to implement this.
501    
502    7.  Minor change to pcretest.c to avoid a compiler warning.
503    
504    8.  Added four artifical Unicode properties to help with an option to make
505        \s etc use properties (see next item). The new properties are: Xan
506        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
507    
508    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
509        use Unicode properties. (*UCP) at the start of a pattern can be used to set
510        this option. Modified pcretest to add /W to test this facility. Added
511        REG_UCP to make it available via the POSIX interface.
512    
513    10. Added --line-buffered to pcregrep.
514    
515    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
516        studied, and the match started with a letter with a code point greater than
517        127 whose first byte was different to the first byte of the other case of
518        the letter, the other case of this starting letter was not recognized
519        (#976).
520    
521    12. If a pattern that was studied started with a repeated Unicode property
522        test, for example, \p{Nd}+, there was the theoretical possibility of
523        setting up an incorrect bitmap of starting bytes, but fortunately it could
524        not have actually happened in practice until change 8 above was made (it
525        added property types that matched character-matching opcodes).
526    
527    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
528        possible starting bytes for non-anchored patterns.
529    
530    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
531        \R, and also a number of cases that involve Unicode properties, both
532        explicit and implicit when PCRE_UCP is set.
533    
534    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
535        input, it could crash or give wrong results if characters with values
536        greater than 0xc0 were present in the subject string. (Detail: it assumed
537        UTF-8 input when processing these items.)
538    
539    16. Added a lot of (int) casts to avoid compiler warnings in systems where
540        size_t is 64-bit (#991).
541    
542    17. Added a check for running out of memory when PCRE is compiled with
543        --disable-stack-for-recursion (#990).
544    
545    18. If the last data line in a file for pcretest does not have a newline on
546        the end, a newline was missing in the output.
547    
548    19. The default pcre_chartables.c file recognizes only ASCII characters (values
549        less than 128) in its various bitmaps. However, there is a facility for
550        generating tables according to the current locale when PCRE is compiled. It
551        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
552        characters, are recognized by isspace() and therefore were getting set in
553        these tables, and indeed these tables seem to approximate to ISO 8859. This
554        caused a problem in UTF-8 mode when pcre_study() was used to create a list
555        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
556        which of course cannot start UTF-8 characters. I have changed the code so
557        that only real ASCII characters (less than 128) and the correct starting
558        bytes for UTF-8 encodings are set for characters greater than 127 when in
559        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
560        altogether.)
561    
562    20. Added the /T option to pcretest so as to be able to run tests with non-
563        standard character tables, thus making it possible to include the tests
564        used for 19 above in the standard set of tests.
565    
566    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
567        reference to a subpattern the other side of a comment that contains an
568        opening parenthesis caused either an internal compiling error, or a
569        reference to the wrong subpattern.
570    
571    
572    Version 8.02 19-Mar-2010
573    ------------------------
574    
575    1.  The Unicode data tables have been updated to Unicode 5.2.0.
576    
577    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
578        configured.
579    
580    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
581        original author of that file, following a query about its status.
582    
583    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
584        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
585    
586    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
587        quantifier applied to a forward-referencing subroutine call, could compile
588        incorrect code or give the error "internal error: previously-checked
589        referenced subpattern not found".
590    
591    6.  Both MS Visual Studio and Symbian OS have problems with initializing
592        variables to point to external functions. For these systems, therefore,
593        pcre_malloc etc. are now initialized to local functions that call the
594        relevant global functions.
595    
596    7.  There were two entries missing in the vectors called coptable and poptable
597        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
598        I've fixed the data, and added a kludgy way of testing at compile time that
599        the lengths are correct (equal to the number of opcodes).
600    
601    8.  Following on from 7, I added a similar kludge to check the length of the
602        eint vector in pcreposix.c.
603    
604    9.  Error texts for pcre_compile() are held as one long string to avoid too
605        much relocation at load time. To find a text, the string is searched,
606        counting zeros. There was no check for running off the end of the string,
607        which could happen if a new error number was added without updating the
608        string.
609    
610    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
611    
612    11. \K was not working if it appeared in an atomic group or in a group that
613        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
614        \K is "not well defined" if used in an assertion. PCRE now accepts it if
615        the assertion is positive, but not if it is negative.
616    
617    12. Change 11 fortuitously reduced the size of the stack frame used in the
618        "match()" function of pcre_exec.c by one pointer. Forthcoming
619        implementation of support for (*MARK) will need an extra pointer on the
620        stack; I have reserved it now, so that the stack frame size does not
621        decrease.
622    
623    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
624        item in branch that calls a recursion is a subroutine call - as in the
625        second branch in the above example - was incorrectly given the compile-
626        time error "recursive call could loop indefinitely" because pcre_compile()
627        was not correctly checking the subroutine for matching a non-empty string.
628    
629    14. The checks for overrunning compiling workspace could trigger after an
630        overrun had occurred. This is a "should never occur" error, but it can be
631        triggered by pathological patterns such as hundreds of nested parentheses.
632        The checks now trigger 100 bytes before the end of the workspace.
633    
634    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
635    
636    
637    Version 8.01 19-Jan-2010
638    ------------------------
639    
640    1.  If a pattern contained a conditional subpattern with only one branch (in
641        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
642        computed the wrong minimum data length (which is of course zero for such
643        subpatterns). This could cause incorrect "no match" results.
644    
645    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
646        the pattern is reset in the first branch, pcre_compile() failed with
647        "internal error: code overflow at offset...". This happened only when
648        the reset was to the original external option setting. (An optimization
649        abstracts leading options settings into an external setting, which was the
650        cause of this.)
651    
652    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
653        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
654        assertion pattern did not match (meaning that the assertion was true), it
655        was incorrectly treated as false if the SKIP had been reached during the
656        matching. This also applied to assertions used as conditions.
657    
658    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
659        assertion subpattern, including such a pattern used as a condition,
660        unpredictable results occurred, instead of the error return
661        PCRE_ERROR_DFA_UITEM.
662    
663    5.  The C++ GlobalReplace function was not working like Perl for the special
664        situation when an empty string is matched. It now does the fancy magic
665        stuff that is necessary.
666    
667    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
668        removed. (These were left over from very, very early versions of PCRE.)
669    
670    7.  Some cosmetic changes to the code to make life easier when compiling it
671        as part of something else:
672    
673        (a) Change DEBUG to PCRE_DEBUG.
674    
675        (b) In pcre_compile(), rename the member of the "branch_chain" structure
676            called "current" as "current_branch", to prevent a collision with the
677            Linux macro when compiled as a kernel module.
678    
679        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
680            prevent a collision with the Linux macro when compiled as a kernel
681            module.
682    
683    8.  In pcre_compile() there are some checks for integer overflows that used to
684        cast potentially large values to (double). This has been changed to that
685        when building, a check for int64_t is made, and if it is found, it is used
686        instead, thus avoiding the use of floating point arithmetic. (There is no
687        other use of FP in PCRE.) If int64_t is not found, the fallback is to
688        double.
689    
690    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
691        2005 (difference between two addresses compared to an unsigned value).
692    
693    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
694        custom one, because of the following reported problem in Windows:
695    
696          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
697              under Win32.
698          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
699              therefore missing the function definition.
700          - The compiler thus generates a "C" signature for the test function.
701          - The linker fails to find the "C" function.
702          - PCRE fails to configure if asked to do so against libbz2.
703    
704    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
705        messages were output:
706    
707          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
708          rerunning libtoolize, to keep the correct libtool macros in-tree.
709          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
710    
711        I have done both of these things.
712    
713    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
714        most of the time, it *can* run out if it is given a pattern that contains a
715        runaway infinite recursion. I updated the discussion in the pcrestack man
716        page.
717    
718    13. Now that we have gone to the x.xx style of version numbers, the minor
719        version may start with zero. Using 08 or 09 is a bad idea because users
720        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
721        interpreted as invalid octal numbers. I've updated the previous comment in
722        configure.ac, and also added a check that gives an error if 08 or 09 are
723        used.
724    
725    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
726        causing partial matching to fail when the end of the subject matched \W
727        in a UTF-8 pattern where \W was quantified with a minimum of 3.
728    
729    15. There were some discrepancies between the declarations in pcre_internal.h
730        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
731        their definitions. The declarations used "const uschar *" and the
732        definitions used USPTR. Even though USPTR is normally defined as "const
733        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
734        reported that: "This difference in casting confuses some C++ compilers, for
735        example, SunCC recognizes above declarations as different functions and
736        generates broken code for hbpcre." I have changed the declarations to use
737        USPTR.
738    
739    16. GNU libtool is named differently on some systems. The autogen.sh script now
740        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
741        (FreeBSD).
742    
743    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
744        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
745        comment: "Figure out how to create a longlong from a string: strtoll and
746        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
747        instance, but it only takes 2 args instead of 3!"
748    
749    18. A subtle bug concerned with back references has been fixed by a change of
750        specification, with a corresponding code fix. A pattern such as
751        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
752        refers, was giving matches when it shouldn't. For example, xa=xaaa would
753        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
754        same bug. Such groups have to be quantified to be useful, or contained
755        inside another quantified group. (If there's no repetition, the reference
756        can never match.) The problem arises because, having left the group and
757        moved on to the rest of the pattern, a later failure that backtracks into
758        the group uses the captured value from the final iteration of the group
759        rather than the correct earlier one. I have fixed this in PCRE by forcing
760        any group that contains a reference to itself to be an atomic group; that
761        is, there cannot be any backtracking into it once it has completed. This is
762        similar to recursive and subroutine calls.
763    
764    
765    Version 8.00 19-Oct-09
766    ----------------------
767    
768    1.  The table for translating pcre_compile() error codes into POSIX error codes
769        was out-of-date, and there was no check on the pcre_compile() error code
770        being within the table. This could lead to an OK return being given in
771        error.
772    
773    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
774        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
775        in a Windows environment.
776    
777    3.  The pcregrep --count option prints the count for each file even when it is
778        zero, as does GNU grep. However, pcregrep was also printing all files when
779        --files-with-matches was added. Now, when both options are given, it prints
780        counts only for those files that have at least one match. (GNU grep just
781        prints the file name in this circumstance, but including the count seems
782        more useful - otherwise, why use --count?) Also ensured that the
783        combination -clh just lists non-zero counts, with no names.
784    
785    4.  The long form of the pcregrep -F option was incorrectly implemented as
786        --fixed_strings instead of --fixed-strings. This is an incompatible change,
787        but it seems right to fix it, and I didn't think it was worth preserving
788        the old behaviour.
789    
790    5.  The command line items --regex=pattern and --regexp=pattern were not
791        recognized by pcregrep, which required --regex pattern or --regexp pattern
792        (with a space rather than an '='). The man page documented the '=' forms,
793        which are compatible with GNU grep; these now work.
794    
795    6.  No libpcreposix.pc file was created for pkg-config; there was just
796        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
797    
798    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
799        when UCP support is not needed, by modifying the Python script that
800        generates it from Unicode data files. This should not matter if the module
801        is correctly used as a library, but I received one complaint about 50K of
802        unwanted data. My guess is that the person linked everything into his
803        program rather than using a library. Anyway, it does no harm.
804    
805    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
806        was a minimum greater than 1 for a wide character in a possessive
807        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
808        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
809        character. Chaos in the form of incorrect output or a compiling loop could
810        result.
811    
812    9.  The restrictions on what a pattern can contain when partial matching is
813        requested for pcre_exec() have been removed. All patterns can now be
814        partially matched by this function. In addition, if there are at least two
815        slots in the offset vector, the offset of the earliest inspected character
816        for the match and the offset of the end of the subject are set in them when
817        PCRE_ERROR_PARTIAL is returned.
818    
819    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
820        synonymous with PCRE_PARTIAL, for backwards compatibility, and
821        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
822        and may be more useful for multi-segment matching.
823    
824    11. Partial matching with pcre_exec() is now more intuitive. A partial match
825        used to be given if ever the end of the subject was reached; now it is
826        given only if matching could not proceed because another character was
827        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
828        string "Z", which now yields "no match" instead of "partial match". In the
829        case of pcre_dfa_exec(), "no match" is given if every matching path for the
830        final character ended with (*FAIL).
831    
832    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
833        if the pattern had a "must contain" character that was already found in the
834        earlier partial match, unless partial matching was again requested. For
835        example, with the pattern /dog.(body)?/, the "must contain" character is
836        "g". If the first part-match was for the string "dog", restarting with
837        "sbody" failed. This bug has been fixed.
838    
839    13. The string returned by pcre_dfa_exec() after a partial match has been
840        changed so that it starts at the first inspected character rather than the
841        first character of the match. This makes a difference only if the pattern
842        starts with a lookbehind assertion or \b or \B (\K is not supported by
843        pcre_dfa_exec()). It's an incompatible change, but it makes the two
844        matching functions compatible, and I think it's the right thing to do.
845    
846    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
847        so that the demonstration program is easily available in environments where
848        PCRE has not been installed from source.
849    
850    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
851        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
852        library.
853    
854    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
855        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
856        is not the first non-POSIX option to be added. Clearly some people find
857        these options useful.
858    
859    17. If a caller to the POSIX matching function regexec() passes a non-zero
860        value for nmatch with a NULL value for pmatch, the value of
861        nmatch is forced to zero.
862    
863    18. RunGrepTest did not have a test for the availability of the -u option of
864        the diff command, as RunTest does. It now checks in the same way as
865        RunTest, and also checks for the -b option.
866    
867    19. If an odd number of negated classes containing just a single character
868        interposed, within parentheses, between a forward reference to a named
869        subpattern and the definition of the subpattern, compilation crashed with
870        an internal error, complaining that it could not find the referenced
871        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
872        [The bug was that it was starting one character too far in when skipping
873        over the character class, thus treating the ] as data rather than
874        terminating the class. This meant it could skip too much.]
875    
876    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
877        /g option in pcretest when the pattern contains \K, which makes it possible
878        to have an empty string match not at the start, even when the pattern is
879        anchored. Updated pcretest and pcredemo to use this option.
880    
881    21. If the maximum number of capturing subpatterns in a recursion was greater
882        than the maximum at the outer level, the higher number was returned, but
883        with unset values at the outer level. The correct (outer level) value is
884        now given.
885    
886    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
887        PCRE did not set those parentheses (unlike Perl). I have now found a way to
888        make it do so. The string so far is captured, making this feature
889        compatible with Perl.
890    
891    23. The tests have been re-organized, adding tests 11 and 12, to make it
892        possible to check the Perl 5.10 features against Perl 5.10.
893    
894    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
895        pattern matches a fixed length string. PCRE did not allow this; now it
896        does. Neither allows recursion.
897    
898    25. I finally figured out how to implement a request to provide the minimum
899        length of subject string that was needed in order to match a given pattern.
900        (It was back references and recursion that I had previously got hung up
901        on.) This code has now been added to pcre_study(); it finds a lower bound
902        to the length of subject needed. It is not necessarily the greatest lower
903        bound, but using it to avoid searching strings that are too short does give
904        some useful speed-ups. The value is available to calling programs via
905        pcre_fullinfo().
906    
907    26. While implementing 25, I discovered to my embarrassment that pcretest had
908        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
909        study optimizations had never been tested with that matching function.
910        Oops. What is worse, even when it was passed study data, there was a bug in
911        pcre_dfa_exec() that meant it never actually used it. Double oops. There
912        were also very few tests of studied patterns with pcre_dfa_exec().
913    
914    27. If (?| is used to create subpatterns with duplicate numbers, they are now
915        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
916        on the other side of the coin, they are no longer allowed to have different
917        names, because these cannot be distinguished in PCRE, and this has caused
918        confusion. (This is a difference from Perl.)
919    
920    28. When duplicate subpattern names are present (necessarily with different
921        numbers, as required by 27 above), and a test is made by name in a
922        conditional pattern, either for a subpattern having been matched, or for
923        recursion in such a pattern, all the associated numbered subpatterns are
924        tested, and the overall condition is true if the condition is true for any
925        one of them. This is the way Perl works, and is also more like the way
926        testing by number works.
927    
928    
929    Version 7.9 11-Apr-09
930    ---------------------
931    
932    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
933        (pcretest), all targets were linked against these libraries. This included
934        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
935        libraries. This caused unwanted dependencies to be created. This problem
936        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
937        pcretest is linked with readline.
938    
939    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
940        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
941        moved outside it again, because FALSE and TRUE are already defined in AIX,
942        but BOOL is not.
943    
944    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
945        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
946    
947    4.  The pcregrep documentation said spaces were inserted as well as colons (or
948        hyphens) following file names and line numbers when outputting matching
949        lines. This is not true; no spaces are inserted. I have also clarified the
950        wording for the --colour (or --color) option.
951    
952    5.  In pcregrep, when --colour was used with -o, the list of matching strings
953        was not coloured; this is different to GNU grep, so I have changed it to be
954        the same.
955    
956    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
957        each matching line was coloured. Now it goes on to look for further matches
958        of any of the test patterns, which is the same behaviour as GNU grep.
959    
960    7.  A pattern that could match an empty string could cause pcregrep to loop; it
961        doesn't make sense to accept an empty string match in pcregrep, so I have
962        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
963        seems to be how GNU grep behaves.
964    
965    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
966        start or after a newline", because the conditional assertion was not being
967        correctly handled. The rule now is that both the assertion and what follows
968        in the first alternative must satisfy the test.
969    
970    9.  If auto-callout was enabled in a pattern with a conditional group whose
971        condition was an assertion, PCRE could crash during matching, both with
972        pcre_exec() and pcre_dfa_exec().
973    
974    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
975        used for matching.
976    
977    11. Unicode property support in character classes was not working for
978        characters (bytes) greater than 127 when not in UTF-8 mode.
979    
980    12. Added the -M command line option to pcretest.
981    
982    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
983    
984    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
985    
986    16. Added comments and documentation about mis-use of no_arg in the C++
987        wrapper.
988    
989    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
990        from Martin Jerabek that uses macro names for all relevant character and
991        string constants.
992    
993    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
994        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
995        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
996        these, but not everybody uses configure.
997    
998    19. A conditional group that had only one branch was not being correctly
999        recognized as an item that could match an empty string. This meant that an
1000        enclosing group might also not be so recognized, causing infinite looping
1001        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1002        with the subject "ab", where knowledge that the repeated group can match
1003        nothing is needed in order to break the loop.
1004    
1005    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1006        exec(), but without supplying a callout function, matching went wrong.
1007    
1008    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1009        leak if the size of the offset vector was greater than 30. When the vector
1010        is smaller, the saved offsets during recursion go onto a local stack
1011        vector, but for larger vectors malloc() is used. It was failing to free
1012        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1013        error, in fact).
1014    
1015    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1016        heapframe that is used only when UTF-8 support is enabled. This caused no
1017        problem, but was untidy.
1018    
1019    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1020        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1021        included within another project.
1022    
1023    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1024        slightly modified by me:
1025    
1026          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1027              not building pcregrep.
1028    
1029          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1030              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1031    
1032    25. Forward references, both numeric and by name, in patterns that made use of
1033        duplicate group numbers, could behave incorrectly or give incorrect errors,
1034        because when scanning forward to find the reference group, PCRE was not
1035        taking into account the duplicate group numbers. A pattern such as
1036        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1037    
1038    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1039        the feature of a custom pointer more persuasive (as requested by a user).
1040    
1041    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1042        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1043        pre-defined.
1044    
1045    28. Added support for (*UTF8) at the start of a pattern.
1046    
1047    29. Arrange for flags added by the "release type" setting in CMake to be shown
1048        in the configuration summary.
1049    
1050    
1051    Version 7.8 05-Sep-08
1052    ---------------------
1053    
1054    1.  Replaced UCP searching code with optimized version as implemented for Ad
1055        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1056        stage table and inline lookup instead of a function, giving speed ups of 2
1057        to 5 times on some simple patterns that I tested. Permission was given to
1058        distribute the MultiStage2.py script that generates the tables (it's not in
1059        the tarball, but is in the Subversion repository).
1060    
1061    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1062        scripts.
1063    
1064    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1065        a group with a zero qualifier. The result of the study could be incorrect,
1066        or the function might crash, depending on the pattern.
1067    
1068    4.  Caseless matching was not working for non-ASCII characters in back
1069        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1070        It now works when Unicode Property Support is available.
1071    
1072    5.  In pcretest, an escape such as \x{de} in the data was always generating
1073        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1074        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1075        truncation.
1076    
1077    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1078    
1079    7.  Added two (int) casts to pcregrep when printing the difference of two
1080        pointers, in case they are 64-bit values.
1081    
1082    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1083        test 2 if it fails.
1084    
1085    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1086        and a #define of that name to empty if it is not externally set. This is to
1087        allow users of MSVC to set it if necessary.
1088    
1089    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1090        the convenience functions in the pcre_get.c source file.
1091    
1092    11. An option change at the start of a pattern that had top-level alternatives
1093        could cause overwriting and/or a crash. This command provoked a crash in
1094        some environments:
1095    
1096          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1097    
1098        This potential security problem was recorded as CVE-2008-2371.
1099    
1100    12. For a pattern where the match had to start at the beginning or immediately
1101        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1102        pcre_dfa_exec() could read past the end of the passed subject if there was
1103        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1104        pcretest so that it places the subject at the end of its malloc-ed buffer.
1105    
1106    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1107        exec() might read past the end of the data buffer in UTF-8 mode.
1108    
1109    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1110        the data contained the byte 0x85 as part of a UTF-8 character within its
1111        first line. This applied both to normal and DFA matching.
1112    
1113    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1114        /^[^d]*?$/8 failed to match "abc".
1115    
1116    16. Added a missing copyright notice to pcrecpp_internal.h.
1117    
1118    17. Make it more clear in the documentation that values returned from
1119        pcre_exec() in ovector are byte offsets, not character counts.
1120    
1121    18. Tidied a few places to stop certain compilers from issuing warnings.
1122    
1123    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1124        supplied by Stefan Weber. I made a further small update for 7.8 because
1125        there is a change of source arrangements: the pcre_searchfuncs.c module is
1126        replaced by pcre_ucd.c.
1127    
1128    
1129    Version 7.7 07-May-08
1130    ---------------------
1131    
1132    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1133        a string to a long long, pretend we don't even have a long long." This is
1134        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1135    
1136    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1137        pre-7.6 versions, which defined a global no_arg variable instead of putting
1138        it in the RE class. (See also #8 below.)
1139    
1140    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1141        Lopes.
1142    
1143    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1144    
1145        (1) The include/exclude patterns were being applied to the whole pathnames
1146            of files, instead of just to the final components.
1147    
1148        (2) If there was more than one level of directory, the subdirectories were
1149            skipped unless they satisfied the include/exclude conditions. This is
1150            inconsistent with GNU grep (and could even be seen as contrary to the
1151            pcregrep specification - which I improved to make it absolutely clear).
1152            The action now is always to scan all levels of directory, and just
1153            apply the include/exclude patterns to regular files.
1154    
1155    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1156        --exclude_dir in the tests to avoid scanning .svn directories.
1157    
1158    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1159        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1160        doesn't support NULs in patterns.
1161    
1162    7.  Added some missing "const"s to declarations of static tables in
1163        pcre_compile.c and pcre_dfa_exec.c.
1164    
1165    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1166        caused by fix #2  above. (Subsequently also a second patch to fix the
1167        first patch. And a third patch - this was a messy problem.)
1168    
1169    9.  Applied Craig's patch to remove the use of push_back().
1170    
1171    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1172        matching function regexec().
1173    
1174    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1175        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1176        references. PCRE supports relative numbers with this syntax (I don't think
1177        Oniguruma does).
1178    
1179    12. Previously, a group with a zero repeat such as (...){0} was completely
1180        omitted from the compiled regex. However, this means that if the group
1181        was called as a subroutine from elsewhere in the pattern, things went wrong
1182        (an internal error was given). Such groups are now left in the compiled
1183        pattern, with a new opcode that causes them to be skipped at execution
1184        time.
1185    
1186    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1187        to the way PCRE behaves:
1188    
1189        (a) A lone ] character is dis-allowed (Perl treats it as data).
1190    
1191        (b) A back reference to an unmatched subpattern matches an empty string
1192            (Perl fails the current match path).
1193    
1194        (c) A data ] in a character class must be notated as \] because if the
1195            first data character in a class is ], it defines an empty class. (In
1196            Perl it is not possible to have an empty class.) The empty class []
1197            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1198            The negative empty class [^] matches any one character, independently
1199            of the DOTALL setting.
1200    
1201    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1202        non-existent subpattern following a character class starting with ']' and
1203        containing () gave an internal compiling error instead of "reference to
1204        non-existent subpattern". Fortunately, when the pattern did exist, the
1205        compiled code was correct. (When scanning forwards to check for the
1206        existencd of the subpattern, it was treating the data ']' as terminating
1207        the class, so got the count wrong. When actually compiling, the reference
1208        was subsequently set up correctly.)
1209    
1210    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1211        it was being rejected as not supported by pcre_dfa_exec(), even though
1212        other assertions are supported. I have made pcre_dfa_exec() support
1213        (*FAIL).
1214    
1215    16. The implementation of 13c above involved the invention of a new opcode,
1216        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1217        cannot be changed at match time, I realized I could make a small
1218        improvement to matching performance by compiling OP_ALLANY instead of
1219        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1220        on the OP_ANY path.
1221    
1222    17. Compiling pcretest on Windows with readline support failed without the
1223        following two fixes: (1) Make the unistd.h include conditional on
1224        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1225    
1226    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1227        ncurses library to be included for pcretest when ReadLine support is
1228        requested, but also to allow for it to be overridden. This patch came from
1229        Daniel Bergström.
1230    
1231    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1232        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1233        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1234        spotting this.
1235    
1236    
1237    Version 7.6 28-Jan-08
1238    ---------------------
1239    
1240    1.  A character class containing a very large number of characters with
1241        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1242        overflow.
1243    
1244    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1245        HAVE_LONG_LONG is not defined.
1246    
1247    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1248        bring them up to date and include new features. This patch includes:
1249    
1250        - Fixed PH's badly added libz and libbz2 support.
1251        - Fixed a problem with static linking.
1252        - Added pcredemo. [But later removed - see 7 below.]
1253        - Fixed dftables problem and added an option.
1254        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1255            HAVE_LONG_LONG.
1256        - Added readline support for pcretest.
1257        - Added an listing of the option settings after cmake has run.
1258    
1259    4.  A user submitted a patch to Makefile that makes it easy to create
1260        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1261        Makefile.am that cause it to include this special target, without
1262        affecting anything else. Note that the same mingw target plus all
1263        the other distribution libraries and programs are now supported
1264        when configuring with CMake (see 6 below) instead of with
1265        Configure/Make.
1266    
1267    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1268        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1269        exported in the Windows port". It has not yet been confirmed that the patch
1270        solves the problem, but it does no harm.
1271    
1272    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1273        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1274        with CMake, and also correct the comment about stack recursion.
1275    
1276    7.  Remove the automatic building of pcredemo from the ./configure system and
1277        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1278        of a program that users should build themselves after PCRE is installed, so
1279        building it automatically is not really right. What is more, it gave
1280        trouble in some build environments.
1281    
1282    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1283    
1284    
1285    Version 7.5 10-Jan-08
1286    ---------------------
1287    
1288    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1289        values in parens when parsing an RE using the C++ wrapper."
1290    
1291    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1292        Characters greater than 255 were excluded from the class instead of being
1293        included.
1294    
1295    3.  The same bug as (2) above applied to negated POSIX classes such as
1296        [:^space:].
1297    
1298    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1299        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1300        I have changed it.
1301    
1302    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1303        first named subpattern) and a construct such as (?&a) would reference the
1304        first named subpattern whose name started with "a" (in other words, the
1305        length check was missing). Both these problems are fixed. "Subpattern name
1306        expected" is now given for (?&) (a zero-length name), and this patch also
1307        makes it give the same error for \k'' (previously it complained that that
1308        was a reference to a non-existent subpattern).
1309    
1310    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1311        this is right because (?- can be followed by option settings as well as by
1312        digits. I have, however, made the messages clearer.
1313    
1314    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1315        than the number used in the conditional) now cause a compile-time error.
1316        This is actually not compatible with Perl, which accepts such patterns, but
1317        treats the conditional as always being FALSE (as PCRE used to), but it
1318        seems to me that giving a diagnostic is better.
1319    
1320    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1321        and messages.
1322    
1323    9.  Fix two occurrences of "backslash" in comments that should have been
1324        "backspace".
1325    
1326    10. Remove two redundant lines of code that can never be obeyed (their function
1327        was moved elsewhere).
1328    
1329    11. The program that makes PCRE's Unicode character property table had a bug
1330        which caused it to generate incorrect table entries for sequences of
1331        characters that have the same character type, but are in different scripts.
1332        It amalgamated them into a single range, with the script of the first of
1333        them. In other words, some characters were in the wrong script. There were
1334        thirteen such cases, affecting characters in the following ranges:
1335    
1336          U+002b0 - U+002c1
1337          U+0060c - U+0060d
1338          U+0061e - U+00612
1339          U+0064b - U+0065e
1340          U+0074d - U+0076d
1341          U+01800 - U+01805
1342          U+01d00 - U+01d77
1343          U+01d9b - U+01dbf
1344          U+0200b - U+0200f
1345          U+030fc - U+030fe
1346          U+03260 - U+0327f
1347          U+0fb46 - U+0fbb1
1348          U+10450 - U+1049d
1349    
1350    12. The -o option (show only the matching part of a line) for pcregrep was not
1351        compatible with GNU grep in that, if there was more than one match in a
1352        line, it showed only the first of them. It now behaves in the same way as
1353        GNU grep.
1354    
1355    13. If the -o and -v options were combined for pcregrep, it printed a blank
1356        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1357        does the same. The return code can be used to tell if there were any
1358        non-matching lines.
1359    
1360    14. Added --file-offsets and --line-offsets to pcregrep.
1361    
1362    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1363        infinitely looping recursion. The bug was that positive lookaheads were not
1364        being skipped when checking for a possible empty match (negative lookaheads
1365        and both kinds of lookbehind were skipped).
1366    
1367    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1368        inclusion of <windows.h> to before rather than after the definition of
1369        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1370    
1371    17. Specifying a possessive quantifier with a specific limit for a Unicode
1372        character property caused pcre_compile() to compile bad code, which led at
1373        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1374        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1375        caused the error; without that there was no problem.
1376    
1377    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1378    
1379    19. Added --enable-pcretest-libreadline.
1380    
1381    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1382        RE::GlobalReplace(). As a result, the number of replacements returned was
1383        double what it should be. I removed one of the increments, but Craig sent a
1384        later patch that removed the other one (the right fix) and added unit tests
1385        that check the return values (which was not done before).
1386    
1387    21. Several CMake things:
1388    
1389        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1390            the names libpcre and libpcreposix, not just pcre and pcreposix.
1391    
1392        (2) The above change means that pcretest and pcregrep are now correctly
1393            linked with the newly-built libraries, not previously installed ones.
1394    
1395        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1396    
1397    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1398        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1399        UTF-8 newline character). The key issue is that the pattern starts .*;
1400        this means that the match must be either at the beginning, or after a
1401        newline. The bug was in the code for advancing after a failed match and
1402        checking that the new position followed a newline. It was not taking
1403        account of UTF-8 characters correctly.
1404    
1405    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1406        character classes. PCRE was not treating the sequence [:...:] as a
1407        character class unless the ... were all letters. Perl, however, seems to
1408        allow any characters between [: and :], though of course it rejects as
1409        unknown any "names" that contain non-letters, because all the known class
1410        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1411        for example, whereas PCRE did not - it did not recognize a POSIX character
1412        class. This seemed a bit dangerous, so the code has been changed to be
1413        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1414        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1415        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1416        Perl does, and where it didn't before.
1417    
1418    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1419        Windows environments %n is disabled by default.
1420    
1421    
1422    Version 7.4 21-Sep-07
1423    ---------------------
1424    
1425    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1426        means that a class such as [\s] counted as "explicit reference to CR or
1427        LF". That isn't really right - the whole point of the change was to try to
1428        help when there was an actual mention of one of the two characters. So now
1429        the change happens only if \r or \n (or a literal CR or LF) character is
1430        encountered.
1431    
1432    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1433        of both had grown to the point where there were only 3 bits left.
1434        Fortunately, there was spare space in the data structure, and so I have
1435        moved the internal flags into a new 16-bit field to free up more option
1436        bits.
1437    
1438    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1439        but did not set the internal JCHANGED flag - either of these is enough to
1440        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1441        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1442        start sets both bits.
1443    
1444    4.  Added options (at build time, compile time, exec time) to change \R from
1445        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1446    
1447    5.  doc/pcresyntax.html was missing from the distribution.
1448    
1449    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1450        compatibility, even though it is no longer used.
1451    
1452    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1453        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1454        windows.h file is present (where different names are used). [This was
1455        reversed later after testing - see 16 below.]
1456    
1457    8.  Changed all #include <config.h> to #include "config.h". There were also
1458        some further <pcre.h> cases that I changed to "pcre.h".
1459    
1460    9.  When pcregrep was used with the --colour option, it missed the line ending
1461        sequence off the lines that it output.
1462    
1463    10. It was pointed out to me that arrays of string pointers cause lots of
1464        relocations when a shared library is dynamically loaded. A technique of
1465        using a single long string with a table of offsets can drastically reduce
1466        these. I have refactored PCRE in four places to do this. The result is
1467        dramatic:
1468    
1469          Originally:                          290
1470          After changing UCP table:            187
1471          After changing error message table:   43
1472          After changing table of "verbs"       36
1473          After changing table of Posix names   22
1474    
1475        Thanks to the folks working on Gregex for glib for this insight.
1476    
1477    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1478        unicode-properties was also set.
1479    
1480    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1481    
1482    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1483        checked only for CRLF.
1484    
1485    14. Added casts to pcretest.c to avoid compiler warnings.
1486    
1487    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1488    
1489    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1490        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1491        entirely. This removes changes made in 7 above.
1492    
1493    17. The CMake files have been updated, and there is now more information about
1494        building with CMake in the NON-UNIX-USE document.
1495    
1496    
1497    Version 7.3 28-Aug-07
1498    ---------------------
1499    
1500     1. In the rejigging of the build system that eventually resulted in 7.1, the
1501        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1502        brackets there is not right, since it causes compilers to look for an
1503        installed pcre.h, not the version that is in the source that is being
1504        compiled (which of course may be different). I have changed it back to:
1505    
1506          #include "pcre.h"
1507    
1508        I have a vague recollection that the change was concerned with compiling in
1509        different directories, but in the new build system, that is taken care of
1510        by the VPATH setting the Makefile.
1511    
1512     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1513        when the subject happened to end in the byte 0x85 (e.g. if the last
1514        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1515        characters but of course it shouldn't be taken as a newline when it is part
1516        of another character. The bug was that, for an unlimited repeat of . in
1517        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1518        characters when looking for a newline.
1519    
1520     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1521    
1522     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1523        in debug output.
1524    
1525     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1526        long printing in the pcrecpp unittest when running under MinGW.
1527    
1528     6. ESC_K was left out of the EBCDIC table.
1529    
1530     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1531        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1532        limit also applies to "virtual nesting" when a pattern is recursive, and in
1533        this case 1000 isn't so big. I have been able to remove this limit at the
1534        expense of backing off one optimization in certain circumstances. Normally,
1535        when pcre_exec() would call its internal match() function recursively and
1536        immediately return the result unconditionally, it uses a "tail recursion"
1537        feature to save stack. However, when a subpattern that can match an empty
1538        string has an unlimited repetition quantifier, it no longer makes this
1539        optimization. That gives it a stack frame in which to save the data for
1540        checking that an empty string has been matched. Previously this was taken
1541        from the 1000-entry workspace that had been reserved. So now there is no
1542        explicit limit, but more stack is used.
1543    
1544     8. Applied Daniel's patches to solve problems with the import/export magic
1545        syntax that is required for Windows, and which was going wrong for the
1546        pcreposix and pcrecpp parts of the library. These were overlooked when this
1547        problem was solved for the main library.
1548    
1549     9. There were some crude static tests to avoid integer overflow when computing
1550        the size of patterns that contain repeated groups with explicit upper
1551        limits. As the maximum quantifier is 65535, the maximum group length was
1552        set at 30,000 so that the product of these two numbers did not overflow a
1553        32-bit integer. However, it turns out that people want to use groups that
1554        are longer than 30,000 bytes (though not repeat them that many times).
1555        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1556        made it possible to implement the integer overflow checks in a much more
1557        dynamic way, which I have now done. The artificial limitation on group
1558        length has been removed - we now have only the limit on the total length of
1559        the compiled pattern, which depends on the LINK_SIZE setting.
1560    
1561    10. Fixed a bug in the documentation for get/copy named substring when
1562        duplicate names are permitted. If none of the named substrings are set, the
1563        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1564        empty string.
1565    
1566    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1567        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1568        because the ] is interpreted as the first data character and the
1569        terminating ] is not found. PCRE has been made compatible with Perl in this
1570        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1571        cause memory overwriting.
1572    
1573    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1574        string has been matched (to stop an infinite loop). It was not recognizing
1575        a conditional subpattern that could match an empty string if that
1576        subpattern was within another subpattern. For example, it looped when
1577        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1578        condition was not nested. This bug has been fixed.
1579    
1580    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1581        past the start of the subject in the presence of bytes with the top bit
1582        set, for example "\x8aBCD".
1583    
1584    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1585        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1586    
1587    14. Optimized (?!) to (*FAIL).
1588    
1589    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1590        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1591        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1592        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1593        does: it's just the validity check that is more restrictive.
1594    
1595    16. Inserted checks for integer overflows during escape sequence (backslash)
1596        processing, and also fixed erroneous offset values for syntax errors during
1597        backslash processing.
1598    
1599    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1600        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1601    
1602    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1603        caused an overrun.
1604    
1605    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1606        something other than just ASCII characters) inside a group that had an
1607        unlimited repeat caused a loop at compile time (while checking to see
1608        whether the group could match an empty string).
1609    
1610    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1611        [\P{Any}] did so. (Error in the code for printing property names.)
1612    
1613    21. An orphan \E inside a character class could cause a crash.
1614    
1615    22. A repeated capturing bracket such as (A)? could cause a wild memory
1616        reference during compilation.
1617    
1618    23. There are several functions in pcre_compile() that scan along a compiled
1619        expression for various reasons (e.g. to see if it's fixed length for look
1620        behind). There were bugs in these functions when a repeated \p or \P was
1621        present in the pattern. These operators have additional parameters compared
1622        with \d, etc, and these were not being taken into account when moving along
1623        the compiled data. Specifically:
1624    
1625        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1626            length.
1627    
1628        (b) An item such as \pL+ within a repeated group could cause crashes or
1629            loops.
1630    
1631        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1632            "reference to non-existent subpattern" error.
1633    
1634        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1635    
1636    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1637        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1638    
1639    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1640    
1641    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1642        character were causing crashes (broken optimization).
1643    
1644    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1645        \p or \P) caused a compile-time loop.
1646    
1647    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1648        break. For example, the unstudied pattern [\r\n]A does not match the string
1649        "\r\nA" because change 7.0/46 below moves the current point on by two
1650        characters after failing to match at the start. However, the pattern \nA
1651        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1652        the same is true. There doesn't seem any very clean way out of this, but
1653        what I have chosen to do makes the common cases work: PCRE now takes note
1654        of whether there can be an explicit match for \r or \n anywhere in the
1655        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1656        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1657        pattern has explicit CR or LF references.
1658    
1659    29. Added (*CR) etc for changing newline setting at start of pattern.
1660    
1661    
1662    Version 7.2 19-Jun-07
1663    ---------------------
1664    
1665     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1666        which is apparently normally available under Windows.
1667    
1668     2. Re-jig the pcregrep tests with different newline settings in an attempt
1669        to make them independent of the local environment's newline setting.
1670    
1671     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1672    
1673     4. Some of the "internals" tests were previously cut out when the link size
1674        was not 2, because the output contained actual offsets. The recent new
1675        "Z" feature of pcretest means that these can be cut out, making the tests
1676        usable with all link sizes.
1677    
1678     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1679        stack recursion. This gives a massive performance boost under BSD, but just
1680        a small improvement under Linux. However, it saves one field in the frame
1681        in all cases.
1682    
1683     6. Added more features from the forthcoming Perl 5.10:
1684    
1685        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1686            recursion call. It refers to the nth most recently opened parentheses.
1687    
1688        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1689            to be opened parentheses.
1690    
1691        (c) Conditions that refer to capturing parentheses can be specified
1692            relatively, for example, (?(-2)... or (?(+3)...
1693    
1694        (d) \K resets the start of the current match so that everything before
1695            is not part of it.
1696    
1697        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1698    
1699        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1700            reference syntax.
1701    
1702        (g) (?| introduces a group in which the numbering of parentheses in each
1703            alternative starts with the same number.
1704    
1705        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1706    
1707     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1708        PCRE_INFO_JCHANGED.
1709    
1710     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1711        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1712        for detecting groups that can match an empty string.
1713    
1714     9. A pattern with a very large number of alternatives (more than several
1715        hundred) was running out of internal workspace during the pre-compile
1716        phase, where pcre_compile() figures out how much memory will be needed. A
1717        bit of new cunning has reduced the workspace needed for groups with
1718        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1719        workspace instead of running out of the 4096 that are available.
1720    
1721    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1722    
1723    11. Applied patch from Google to remove an optimization that didn't quite work.
1724        The report of the bug said:
1725    
1726          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1727          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1728          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1729    
1730    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1731        it matched the wrong number of bytes.
1732    
1733    
1734    Version 7.1 24-Apr-07
1735  ---------------------  ---------------------
1736    
1737   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
# Line 135  Version 7.1 12-Mar-07 Line 1865  Version 7.1 12-Mar-07
1865      was because its rules for how to advance for /g after matching an empty      was because its rules for how to advance for /g after matching an empty
1866      string at the end of a line did not allow for this case. They now check for      string at the end of a line did not allow for this case. They now check for
1867      it specially.      it specially.
1868    
1869  20. pcretest is supposed to handle patterns and data of any length, by  20. pcretest is supposed to handle patterns and data of any length, by
1870      extending its buffers when necessary. It was getting this wrong when the      extending its buffers when necessary. It was getting this wrong when the
1871      buffer for a data line had to be extended.      buffer for a data line had to be extended.
1872    
1873  21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or  21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1874      CRLF as a newline sequence.      CRLF as a newline sequence.
1875    
1876    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1877        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1878        I have nevertheless tidied it up.
1879    
1880    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1881    
1882    24. Added a man page for pcre-config.
1883    
1884    
1885  Version 7.0 19-Dec-06  Version 7.0 19-Dec-06

Legend:
Removed from v.149  
changed lines
  Added in v.701

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12