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

Diff of /code/trunk/ChangeLog

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

revision 116 by ph10, Fri Mar 9 15:23:02 2007 UTC revision 712 by ph10, Mon Sep 26 16:31:42 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.1 05-Mar-07  Version 8.20 23-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    11. Change 17 for 8.13 increased the recursion depth for patterns like
64        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
65        pattern contains any instances of (*THEN). If it does not, the old
66        optimizations are restored. It would be nice to do this on a per-group
67        basis, but at the moment that is not feasible.
68    
69    12. In some environments, the output of pcretest -C is CRLF terminated. This
70        broke RunTest's code that checks for the link size. A single white space
71        after the value is now allowed for.
72    
73    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
74        For "fr", it uses the Windows-specific input and output files.
75    
76    
77    Version 8.13 16-Aug-2011
78    ------------------------
79    
80    1.  The Unicode data tables have been updated to Unicode 6.0.0.
81    
82    2.  Two minor typos in pcre_internal.h have been fixed.
83    
84    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
85        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
86        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
87    
88    4.  There were a number of related bugs in the code for matching backrefences
89        caselessly in UTF-8 mode when codes for the characters concerned were
90        different numbers of bytes. For example, U+023A and U+2C65 are an upper
91        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
92        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
93        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
94        2-byte code at the end of the subject (it thought there wasn't enough data
95        left).
96    
97    5.  Comprehensive information about what went wrong is now returned by
98        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
99        as the output vector has at least 2 elements. The offset of the start of
100        the failing character and a reason code are placed in the vector.
101    
102    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
103        now returned is for the first byte of the failing character, instead of the
104        last byte inspected. This is an incompatible change, but I hope it is small
105        enough not to be a problem. It makes the returned offset consistent with
106        pcre_exec() and pcre_dfa_exec().
107    
108    7.  pcretest now gives a text phrase as well as the error number when
109        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
110        failure, the offset and reason code are output.
111    
112    8.  When \R was used with a maximizing quantifier it failed to skip backwards
113        over a \r\n pair if the subsequent match failed. Instead, it just skipped
114        back over a single character (\n). This seems wrong (because it treated the
115        two characters as a single entity when going forwards), conflicts with the
116        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
117        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
118        has been changed.
119    
120    9.  Some internal refactoring has changed the processing so that the handling
121        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
122        time (the PCRE_DOTALL option was changed this way some time ago: version
123        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
124        which was always a bit of a fudge. It also means that there is one less
125        argument for the match() function, which reduces its stack requirements
126        slightly. This change also fixes an incompatibility with Perl: the pattern
127        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
128    
129    10. More internal refactoring has drastically reduced the number of recursive
130        calls to match() for possessively repeated groups such as (abc)++ when
131        using pcre_exec().
132    
133    11. While implementing 10, a number of bugs in the handling of groups were
134        discovered and fixed:
135    
136        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
137        (a|)*(?1) gave a compile-time internal error.
138        ((a|)+)+  did not notice that the outer group could match an empty string.
139        (^a|^)+   was not marked as anchored.
140        (.*a|.*)+ was not marked as matching at start or after a newline.
141    
142    12. Yet more internal refactoring has removed another argument from the match()
143        function. Special calls to this function are now indicated by setting a
144        value in a variable in the "match data" data block.
145    
146    13. Be more explicit in pcre_study() instead of relying on "default" for
147        opcodes that mean there is no starting character; this means that when new
148        ones are added and accidentally left out of pcre_study(), testing should
149        pick them up.
150    
151    14. The -s option of pcretest has been documented for ages as being an old
152        synonym of -m (show memory usage). I have changed it to mean "force study
153        for every regex", that is, assume /S for every regex. This is similar to -i
154        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
155        using it. It makes it easier to run collections of tests with and without
156        study enabled, and thereby test pcre_study() more easily. All the standard
157        tests are now run with and without -s (but some patterns can be marked as
158        "never study" - see 20 below).
159    
160    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
161        restoration of the capturing data to the outer values was not happening
162        correctly.
163    
164    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
165        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
166        pattern had matched an empty string, and so incorrectly returned a no
167        match.
168    
169    17. There was optimizing code for the last branch of non-capturing parentheses,
170        and also for the obeyed branch of a conditional subexpression, which used
171        tail recursion to cut down on stack usage. Unfortunately, now that there is
172        the possibility of (*THEN) occurring in these branches, tail recursion is
173        no longer possible because the return has to be checked for (*THEN). These
174        two optimizations have therefore been removed. [But see 8.20/11 above.]
175    
176    18. If a pattern containing \R was studied, it was assumed that \R always
177        matched two bytes, thus causing the minimum subject length to be
178        incorrectly computed because \R can also match just one byte.
179    
180    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
181        was incorrectly computed.
182    
183    20. If /S is present twice on a test pattern in pcretest input, it now
184        *disables* studying, thereby overriding the use of -s on the command line
185        (see 14 above). This is necessary for one or two tests to keep the output
186        identical in both cases.
187    
188    21. When (*ACCEPT) was used in an assertion that matched an empty string and
189        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
190    
191    22. When an atomic group that contained a capturing parenthesis was
192        successfully matched, but the branch in which it appeared failed, the
193        capturing was not being forgotten if a higher numbered group was later
194        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
195        group 1 to "a", when in fact it should be unset. This applied to multi-
196        branched capturing and non-capturing groups, repeated or not, and also to
197        positive assertions (capturing in negative assertions does not happen
198        in PCRE) and also to nested atomic groups.
199    
200    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
201        subject after a captured substring, to make it easier to tell which of a
202        number of identical substrings has been captured.
203    
204    24. The way atomic groups are processed by pcre_exec() has been changed so that
205        if they are repeated, backtracking one repetition now resets captured
206        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
207        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
208        "aaa". Previously, it would have been "a". As part of this code
209        refactoring, the way recursive calls are handled has also been changed.
210    
211    25. If an assertion condition captured any substrings, they were not passed
212        back unless some other capturing happened later. For example, if
213        (?(?=(a))a) was matched against "a", no capturing was returned.
214    
215    26. When studying a pattern that contained subroutine calls or assertions,
216        the code for finding the minimum length of a possible match was handling
217        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
218        group 1 called group 2 while simultaneously a separate group 2 called group
219        1). A stack overflow occurred in this case. I have fixed this by limiting
220        the recursion depth to 10.
221    
222    27. Updated RunTest.bat in the distribution to the version supplied by Tom
223        Fortmann. This supports explicit test numbers on the command line, and has
224        argument validation and error reporting.
225    
226    28. An instance of \X with an unlimited repeat could fail if at any point the
227        first character it looked at was a mark character.
228    
229    29. Some minor code refactoring concerning Unicode properties and scripts
230        should reduce the stack requirement of match() slightly.
231    
232    30. Added the '=' option to pcretest to check the setting of unused capturing
233        slots at the end of the pattern, which are documented as being -1, but are
234        not included in the return count.
235    
236    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
237        compiled something random. Now it gives a compile-time error (as does
238        Perl).
239    
240    32. A *MARK encountered during the processing of a positive assertion is now
241        recorded and passed back (compatible with Perl).
242    
243    33. If --only-matching or --colour was set on a pcregrep call whose pattern
244        had alternative anchored branches, the search for a second match in a line
245        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
246        matched the line "0102" twice. The same bug affected patterns that started
247        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
248        twice.
249    
250    34. Previously, PCRE did not allow quantification of assertions. However, Perl
251        does, and because of capturing effects, quantifying parenthesized
252        assertions may at times be useful. Quantifiers are now allowed for
253        parenthesized assertions.
254    
255    35. A minor code tidy in pcre_compile() when checking options for \R usage.
256    
257    36. \g was being checked for fancy things in a character class, when it should
258        just be a literal "g".
259    
260    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
261        appearance of a nested POSIX class supersedes an apparent external class.
262        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
263        unescaped square brackets may also appear as part of class names. For
264        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
265        more like Perl. (But see 8.20/1 above.)
266    
267    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
268        was because it thought it was \N{name}, which is not supported).
269    
270    39. Add minix to OS list not supporting the -S option in pcretest.
271    
272    40. PCRE tries to detect cases of infinite recursion at compile time, but it
273        cannot analyze patterns in sufficient detail to catch mutual recursions
274        such as ((?1))((?2)). There is now a runtime test that gives an error if a
275        subgroup is called recursively as a subpattern for a second time at the
276        same position in the subject string. In previous releases this might have
277        been caught by the recursion limit, or it might have run out of stack.
278    
279    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
280        happen only once. PCRE was, however incorrectly giving a compile time error
281        "recursive call could loop indefinitely" because it cannot analyze the
282        pattern in sufficient detail. The compile time test no longer happens when
283        PCRE is compiling a conditional subpattern, but actual runaway loops are
284        now caught at runtime (see 40 above).
285    
286    42. It seems that Perl allows any characters other than a closing parenthesis
287        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
288        has been changed to be the same.
289    
290    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
291        as not to get warnings when autogen.sh is called. Also changed
292        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
293    
294    44. To help people who use pcregrep to scan files containing exceedingly long
295        lines, the following changes have been made:
296    
297        (a) The default value of the buffer size parameter has been increased from
298            8K to 20K. (The actual buffer used is three times this size.)
299    
300        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
301            PCRE is built.
302    
303        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
304            to be set at run time.
305    
306        (d) Numerical values in pcregrep options can be followed by K or M, for
307            example --buffer-size=50K.
308    
309        (e) If a line being scanned overflows pcregrep's buffer, an error is now
310            given and the return code is set to 2.
311    
312    45. Add a pointer to the latest mark to the callout data block.
313    
314    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
315        partial match of an empty string instead of no match. This was specific to
316        the use of ".".
317    
318    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
319        complete match instead of a partial match. This bug was dependent on both
320        the PCRE_UTF8 and PCRE_DOTALL options being set.
321    
322    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
323        starting byte set, because \b was not being ignored.
324    
325    
326    Version 8.12 15-Jan-2011
327    ------------------------
328    
329    1.  Fixed some typos in the markup of the man pages, and wrote a script that
330        checks for such things as part of the documentation building process.
331    
332    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
333        --match-limit and --recursion-limit options (added for 8.11). In
334        particular, this made one of the standard tests fail. (The integer value
335        went into the wrong half of a long int.)
336    
337    3.  If the --colour option was given to pcregrep with -v (invert match), it
338        did strange things, either producing crazy output, or crashing. It should,
339        of course, ignore a request for colour when reporting lines that do not
340        match.
341    
342    4.  Another pcregrep bug caused similar problems if --colour was specified with
343        -M (multiline) and the pattern match finished with a line ending.
344    
345    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
346        matched in multiline mode, the following line was shown as part of the
347        match. This seems wrong, so I have changed it.
348    
349    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
350        the check for further matches in the same line (so they could be coloured)
351        to overrun the end of the current line. If another match was found, it was
352        incorrectly shown (and then shown again when found in the next line).
353    
354    7.  If pcregrep was compiled under Windows, there was a reference to the
355        function pcregrep_exit() before it was defined. I am assuming this was
356        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
357        reported by a user. I've moved the definition above the reference.
358    
359    
360    Version 8.11 10-Dec-2010
361    ------------------------
362    
363    1.  (*THEN) was not working properly if there were untried alternatives prior
364        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
365        backtracked to try for "b" instead of moving to the next alternative branch
366        at the same level (in this case, to look for "c"). The Perl documentation
367        is clear that when (*THEN) is backtracked onto, it goes to the "next
368        alternative in the innermost enclosing group".
369    
370    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
371        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
372        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
373        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
374        (*THEN).
375    
376    3.  If \s appeared in a character class, it removed the VT character from
377        the class, even if it had been included by some previous item, for example
378        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
379        of \s, but is part of the POSIX "space" class.)
380    
381    4.  A partial match never returns an empty string (because you can always
382        match an empty string at the end of the subject); however the checking for
383        an empty string was starting at the "start of match" point. This has been
384        changed to the "earliest inspected character" point, because the returned
385        data for a partial match starts at this character. This means that, for
386        example, /(?<=abc)def/ gives a partial match for the subject "abc"
387        (previously it gave "no match").
388    
389    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
390        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
391        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
392        has an implication that the given string is incomplete (because a partial
393        match is preferred over a full match). For this reason, these items now
394        give a partial match in this situation. [Aside: previously, the one case
395        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
396        match rather than a full match, which was wrong by the old rules, but is
397        now correct.]
398    
399    6.  There was a bug in the handling of #-introduced comments, recognized when
400        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
401        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
402        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
403        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
404        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
405        places in pcre_compile().
406    
407    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
408        comments when looking ahead for named forward references to subpatterns,
409        the only newline sequence it recognized was NL. It now handles newlines
410        according to the set newline convention.
411    
412    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
413        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
414        cater for a lack of strerror(). These oversights have been fixed.
415    
416    9.  Added --match-limit and --recursion-limit to pcregrep.
417    
418    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
419    
420    11. When the -o option was used, pcregrep was setting a return code of 1, even
421        when matches were found, and --line-buffered was not being honoured.
422    
423    12. Added an optional parentheses number to the -o and --only-matching options
424        of pcregrep.
425    
426    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
427        can match an empty string. The code to do it in pcretest and pcredemo
428        needed fixing:
429    
430        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
431            only one byte after an empty string match just before CRLF (this case
432            just got forgotten; "any" and "anycrlf" were OK).
433    
434        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
435            mode when an empty string match preceded an ASCII character followed by
436            a non-ASCII character. (The code for advancing by one character rather
437            than one byte was nonsense.)
438    
439        (c) The pcredemo.c sample program did not have any code at all to handle
440            the cases when CRLF is a valid newline sequence.
441    
442    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
443        as a starting offset was within the subject string. There is now a new
444        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
445        negative or greater than the length of the string. In order to test this,
446        pcretest is extended to allow the setting of negative starting offsets.
447    
448    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
449        starting offset points to the beginning of a UTF-8 character was
450        unnecessarily clumsy. I tidied it up.
451    
452    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
453        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
454    
455    17. Nobody had reported that the --include_dir option, which was added in
456        release 7.7 should have been called --include-dir (hyphen, not underscore)
457        for compatibility with GNU grep. I have changed it to --include-dir, but
458        left --include_dir as an undocumented synonym, and the same for
459        --exclude-dir, though that is not available in GNU grep, at least as of
460        release 2.5.4.
461    
462    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
463        characters from a string of bytes) have been redefined so as not to use
464        loops, in order to improve performance in some environments. At the same
465        time, I abstracted some of the common code into auxiliary macros to save
466        repetition (this should not affect the compiled code).
467    
468    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
469        compile-time error is now given if \c is not followed by an ASCII
470        character, that is, a byte less than 128. (In EBCDIC mode, the code is
471        different, and any byte value is allowed.)
472    
473    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
474        START_OPTIMIZE option, which is now allowed at compile time - but just
475        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
476        to pcregrep and other applications that have no direct access to PCRE
477        options. The new /Y option in pcretest sets this option when calling
478        pcre_compile().
479    
480    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
481        back references. Groups containing recursive back references were forced to
482        be atomic by that change, but in the case of named groups, the amount of
483        memory required was incorrectly computed, leading to "Failed: internal
484        error: code overflow". This has been fixed.
485    
486    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
487        pcretest.c, to avoid build problems in some Borland environments.
488    
489    
490    Version 8.10 25-Jun-2010
491    ------------------------
492    
493    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
494        THEN.
495    
496    2.  (*ACCEPT) was not working when inside an atomic group.
497    
498    3.  Inside a character class, \B is treated as a literal by default, but
499        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
500        causes the error). The code is unchanged, but I tidied the documentation.
501    
502    4.  Inside a character class, PCRE always treated \R and \X as literals,
503        whereas Perl faults them if its -w option is set. I have changed PCRE so
504        that it faults them when PCRE_EXTRA is set.
505    
506    5.  Added support for \N, which always matches any character other than
507        newline. (It is the same as "." when PCRE_DOTALL is not set.)
508    
509    6.  When compiling pcregrep with newer versions of gcc which may have
510        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
511        declared with attribute warn_unused_result" were given. Just casting the
512        result to (void) does not stop the warnings; a more elaborate fudge is
513        needed. I've used a macro to implement this.
514    
515    7.  Minor change to pcretest.c to avoid a compiler warning.
516    
517    8.  Added four artifical Unicode properties to help with an option to make
518        \s etc use properties (see next item). The new properties are: Xan
519        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
520    
521    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
522        use Unicode properties. (*UCP) at the start of a pattern can be used to set
523        this option. Modified pcretest to add /W to test this facility. Added
524        REG_UCP to make it available via the POSIX interface.
525    
526    10. Added --line-buffered to pcregrep.
527    
528    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
529        studied, and the match started with a letter with a code point greater than
530        127 whose first byte was different to the first byte of the other case of
531        the letter, the other case of this starting letter was not recognized
532        (#976).
533    
534    12. If a pattern that was studied started with a repeated Unicode property
535        test, for example, \p{Nd}+, there was the theoretical possibility of
536        setting up an incorrect bitmap of starting bytes, but fortunately it could
537        not have actually happened in practice until change 8 above was made (it
538        added property types that matched character-matching opcodes).
539    
540    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
541        possible starting bytes for non-anchored patterns.
542    
543    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
544        \R, and also a number of cases that involve Unicode properties, both
545        explicit and implicit when PCRE_UCP is set.
546    
547    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
548        input, it could crash or give wrong results if characters with values
549        greater than 0xc0 were present in the subject string. (Detail: it assumed
550        UTF-8 input when processing these items.)
551    
552    16. Added a lot of (int) casts to avoid compiler warnings in systems where
553        size_t is 64-bit (#991).
554    
555    17. Added a check for running out of memory when PCRE is compiled with
556        --disable-stack-for-recursion (#990).
557    
558    18. If the last data line in a file for pcretest does not have a newline on
559        the end, a newline was missing in the output.
560    
561    19. The default pcre_chartables.c file recognizes only ASCII characters (values
562        less than 128) in its various bitmaps. However, there is a facility for
563        generating tables according to the current locale when PCRE is compiled. It
564        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
565        characters, are recognized by isspace() and therefore were getting set in
566        these tables, and indeed these tables seem to approximate to ISO 8859. This
567        caused a problem in UTF-8 mode when pcre_study() was used to create a list
568        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
569        which of course cannot start UTF-8 characters. I have changed the code so
570        that only real ASCII characters (less than 128) and the correct starting
571        bytes for UTF-8 encodings are set for characters greater than 127 when in
572        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
573        altogether.)
574    
575    20. Added the /T option to pcretest so as to be able to run tests with non-
576        standard character tables, thus making it possible to include the tests
577        used for 19 above in the standard set of tests.
578    
579    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
580        reference to a subpattern the other side of a comment that contains an
581        opening parenthesis caused either an internal compiling error, or a
582        reference to the wrong subpattern.
583    
584    
585    Version 8.02 19-Mar-2010
586    ------------------------
587    
588    1.  The Unicode data tables have been updated to Unicode 5.2.0.
589    
590    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
591        configured.
592    
593    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
594        original author of that file, following a query about its status.
595    
596    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
597        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
598    
599    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
600        quantifier applied to a forward-referencing subroutine call, could compile
601        incorrect code or give the error "internal error: previously-checked
602        referenced subpattern not found".
603    
604    6.  Both MS Visual Studio and Symbian OS have problems with initializing
605        variables to point to external functions. For these systems, therefore,
606        pcre_malloc etc. are now initialized to local functions that call the
607        relevant global functions.
608    
609    7.  There were two entries missing in the vectors called coptable and poptable
610        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
611        I've fixed the data, and added a kludgy way of testing at compile time that
612        the lengths are correct (equal to the number of opcodes).
613    
614    8.  Following on from 7, I added a similar kludge to check the length of the
615        eint vector in pcreposix.c.
616    
617    9.  Error texts for pcre_compile() are held as one long string to avoid too
618        much relocation at load time. To find a text, the string is searched,
619        counting zeros. There was no check for running off the end of the string,
620        which could happen if a new error number was added without updating the
621        string.
622    
623    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
624    
625    11. \K was not working if it appeared in an atomic group or in a group that
626        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
627        \K is "not well defined" if used in an assertion. PCRE now accepts it if
628        the assertion is positive, but not if it is negative.
629    
630    12. Change 11 fortuitously reduced the size of the stack frame used in the
631        "match()" function of pcre_exec.c by one pointer. Forthcoming
632        implementation of support for (*MARK) will need an extra pointer on the
633        stack; I have reserved it now, so that the stack frame size does not
634        decrease.
635    
636    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
637        item in branch that calls a recursion is a subroutine call - as in the
638        second branch in the above example - was incorrectly given the compile-
639        time error "recursive call could loop indefinitely" because pcre_compile()
640        was not correctly checking the subroutine for matching a non-empty string.
641    
642    14. The checks for overrunning compiling workspace could trigger after an
643        overrun had occurred. This is a "should never occur" error, but it can be
644        triggered by pathological patterns such as hundreds of nested parentheses.
645        The checks now trigger 100 bytes before the end of the workspace.
646    
647    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
648    
649    
650    Version 8.01 19-Jan-2010
651    ------------------------
652    
653    1.  If a pattern contained a conditional subpattern with only one branch (in
654        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
655        computed the wrong minimum data length (which is of course zero for such
656        subpatterns). This could cause incorrect "no match" results.
657    
658    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
659        the pattern is reset in the first branch, pcre_compile() failed with
660        "internal error: code overflow at offset...". This happened only when
661        the reset was to the original external option setting. (An optimization
662        abstracts leading options settings into an external setting, which was the
663        cause of this.)
664    
665    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
666        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
667        assertion pattern did not match (meaning that the assertion was true), it
668        was incorrectly treated as false if the SKIP had been reached during the
669        matching. This also applied to assertions used as conditions.
670    
671    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
672        assertion subpattern, including such a pattern used as a condition,
673        unpredictable results occurred, instead of the error return
674        PCRE_ERROR_DFA_UITEM.
675    
676    5.  The C++ GlobalReplace function was not working like Perl for the special
677        situation when an empty string is matched. It now does the fancy magic
678        stuff that is necessary.
679    
680    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
681        removed. (These were left over from very, very early versions of PCRE.)
682    
683    7.  Some cosmetic changes to the code to make life easier when compiling it
684        as part of something else:
685    
686        (a) Change DEBUG to PCRE_DEBUG.
687    
688        (b) In pcre_compile(), rename the member of the "branch_chain" structure
689            called "current" as "current_branch", to prevent a collision with the
690            Linux macro when compiled as a kernel module.
691    
692        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
693            prevent a collision with the Linux macro when compiled as a kernel
694            module.
695    
696    8.  In pcre_compile() there are some checks for integer overflows that used to
697        cast potentially large values to (double). This has been changed to that
698        when building, a check for int64_t is made, and if it is found, it is used
699        instead, thus avoiding the use of floating point arithmetic. (There is no
700        other use of FP in PCRE.) If int64_t is not found, the fallback is to
701        double.
702    
703    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
704        2005 (difference between two addresses compared to an unsigned value).
705    
706    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
707        custom one, because of the following reported problem in Windows:
708    
709          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
710              under Win32.
711          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
712              therefore missing the function definition.
713          - The compiler thus generates a "C" signature for the test function.
714          - The linker fails to find the "C" function.
715          - PCRE fails to configure if asked to do so against libbz2.
716    
717    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
718        messages were output:
719    
720          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
721          rerunning libtoolize, to keep the correct libtool macros in-tree.
722          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
723    
724        I have done both of these things.
725    
726    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
727        most of the time, it *can* run out if it is given a pattern that contains a
728        runaway infinite recursion. I updated the discussion in the pcrestack man
729        page.
730    
731    13. Now that we have gone to the x.xx style of version numbers, the minor
732        version may start with zero. Using 08 or 09 is a bad idea because users
733        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
734        interpreted as invalid octal numbers. I've updated the previous comment in
735        configure.ac, and also added a check that gives an error if 08 or 09 are
736        used.
737    
738    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
739        causing partial matching to fail when the end of the subject matched \W
740        in a UTF-8 pattern where \W was quantified with a minimum of 3.
741    
742    15. There were some discrepancies between the declarations in pcre_internal.h
743        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
744        their definitions. The declarations used "const uschar *" and the
745        definitions used USPTR. Even though USPTR is normally defined as "const
746        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
747        reported that: "This difference in casting confuses some C++ compilers, for
748        example, SunCC recognizes above declarations as different functions and
749        generates broken code for hbpcre." I have changed the declarations to use
750        USPTR.
751    
752    16. GNU libtool is named differently on some systems. The autogen.sh script now
753        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
754        (FreeBSD).
755    
756    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
757        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
758        comment: "Figure out how to create a longlong from a string: strtoll and
759        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
760        instance, but it only takes 2 args instead of 3!"
761    
762    18. A subtle bug concerned with back references has been fixed by a change of
763        specification, with a corresponding code fix. A pattern such as
764        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
765        refers, was giving matches when it shouldn't. For example, xa=xaaa would
766        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
767        same bug. Such groups have to be quantified to be useful, or contained
768        inside another quantified group. (If there's no repetition, the reference
769        can never match.) The problem arises because, having left the group and
770        moved on to the rest of the pattern, a later failure that backtracks into
771        the group uses the captured value from the final iteration of the group
772        rather than the correct earlier one. I have fixed this in PCRE by forcing
773        any group that contains a reference to itself to be an atomic group; that
774        is, there cannot be any backtracking into it once it has completed. This is
775        similar to recursive and subroutine calls.
776    
777    
778    Version 8.00 19-Oct-09
779    ----------------------
780    
781    1.  The table for translating pcre_compile() error codes into POSIX error codes
782        was out-of-date, and there was no check on the pcre_compile() error code
783        being within the table. This could lead to an OK return being given in
784        error.
785    
786    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
787        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
788        in a Windows environment.
789    
790    3.  The pcregrep --count option prints the count for each file even when it is
791        zero, as does GNU grep. However, pcregrep was also printing all files when
792        --files-with-matches was added. Now, when both options are given, it prints
793        counts only for those files that have at least one match. (GNU grep just
794        prints the file name in this circumstance, but including the count seems
795        more useful - otherwise, why use --count?) Also ensured that the
796        combination -clh just lists non-zero counts, with no names.
797    
798    4.  The long form of the pcregrep -F option was incorrectly implemented as
799        --fixed_strings instead of --fixed-strings. This is an incompatible change,
800        but it seems right to fix it, and I didn't think it was worth preserving
801        the old behaviour.
802    
803    5.  The command line items --regex=pattern and --regexp=pattern were not
804        recognized by pcregrep, which required --regex pattern or --regexp pattern
805        (with a space rather than an '='). The man page documented the '=' forms,
806        which are compatible with GNU grep; these now work.
807    
808    6.  No libpcreposix.pc file was created for pkg-config; there was just
809        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
810    
811    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
812        when UCP support is not needed, by modifying the Python script that
813        generates it from Unicode data files. This should not matter if the module
814        is correctly used as a library, but I received one complaint about 50K of
815        unwanted data. My guess is that the person linked everything into his
816        program rather than using a library. Anyway, it does no harm.
817    
818    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
819        was a minimum greater than 1 for a wide character in a possessive
820        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
821        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
822        character. Chaos in the form of incorrect output or a compiling loop could
823        result.
824    
825    9.  The restrictions on what a pattern can contain when partial matching is
826        requested for pcre_exec() have been removed. All patterns can now be
827        partially matched by this function. In addition, if there are at least two
828        slots in the offset vector, the offset of the earliest inspected character
829        for the match and the offset of the end of the subject are set in them when
830        PCRE_ERROR_PARTIAL is returned.
831    
832    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
833        synonymous with PCRE_PARTIAL, for backwards compatibility, and
834        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
835        and may be more useful for multi-segment matching.
836    
837    11. Partial matching with pcre_exec() is now more intuitive. A partial match
838        used to be given if ever the end of the subject was reached; now it is
839        given only if matching could not proceed because another character was
840        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
841        string "Z", which now yields "no match" instead of "partial match". In the
842        case of pcre_dfa_exec(), "no match" is given if every matching path for the
843        final character ended with (*FAIL).
844    
845    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
846        if the pattern had a "must contain" character that was already found in the
847        earlier partial match, unless partial matching was again requested. For
848        example, with the pattern /dog.(body)?/, the "must contain" character is
849        "g". If the first part-match was for the string "dog", restarting with
850        "sbody" failed. This bug has been fixed.
851    
852    13. The string returned by pcre_dfa_exec() after a partial match has been
853        changed so that it starts at the first inspected character rather than the
854        first character of the match. This makes a difference only if the pattern
855        starts with a lookbehind assertion or \b or \B (\K is not supported by
856        pcre_dfa_exec()). It's an incompatible change, but it makes the two
857        matching functions compatible, and I think it's the right thing to do.
858    
859    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
860        so that the demonstration program is easily available in environments where
861        PCRE has not been installed from source.
862    
863    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
864        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
865        library.
866    
867    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
868        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
869        is not the first non-POSIX option to be added. Clearly some people find
870        these options useful.
871    
872    17. If a caller to the POSIX matching function regexec() passes a non-zero
873        value for nmatch with a NULL value for pmatch, the value of
874        nmatch is forced to zero.
875    
876    18. RunGrepTest did not have a test for the availability of the -u option of
877        the diff command, as RunTest does. It now checks in the same way as
878        RunTest, and also checks for the -b option.
879    
880    19. If an odd number of negated classes containing just a single character
881        interposed, within parentheses, between a forward reference to a named
882        subpattern and the definition of the subpattern, compilation crashed with
883        an internal error, complaining that it could not find the referenced
884        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
885        [The bug was that it was starting one character too far in when skipping
886        over the character class, thus treating the ] as data rather than
887        terminating the class. This meant it could skip too much.]
888    
889    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
890        /g option in pcretest when the pattern contains \K, which makes it possible
891        to have an empty string match not at the start, even when the pattern is
892        anchored. Updated pcretest and pcredemo to use this option.
893    
894    21. If the maximum number of capturing subpatterns in a recursion was greater
895        than the maximum at the outer level, the higher number was returned, but
896        with unset values at the outer level. The correct (outer level) value is
897        now given.
898    
899    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
900        PCRE did not set those parentheses (unlike Perl). I have now found a way to
901        make it do so. The string so far is captured, making this feature
902        compatible with Perl.
903    
904    23. The tests have been re-organized, adding tests 11 and 12, to make it
905        possible to check the Perl 5.10 features against Perl 5.10.
906    
907    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
908        pattern matches a fixed length string. PCRE did not allow this; now it
909        does. Neither allows recursion.
910    
911    25. I finally figured out how to implement a request to provide the minimum
912        length of subject string that was needed in order to match a given pattern.
913        (It was back references and recursion that I had previously got hung up
914        on.) This code has now been added to pcre_study(); it finds a lower bound
915        to the length of subject needed. It is not necessarily the greatest lower
916        bound, but using it to avoid searching strings that are too short does give
917        some useful speed-ups. The value is available to calling programs via
918        pcre_fullinfo().
919    
920    26. While implementing 25, I discovered to my embarrassment that pcretest had
921        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
922        study optimizations had never been tested with that matching function.
923        Oops. What is worse, even when it was passed study data, there was a bug in
924        pcre_dfa_exec() that meant it never actually used it. Double oops. There
925        were also very few tests of studied patterns with pcre_dfa_exec().
926    
927    27. If (?| is used to create subpatterns with duplicate numbers, they are now
928        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
929        on the other side of the coin, they are no longer allowed to have different
930        names, because these cannot be distinguished in PCRE, and this has caused
931        confusion. (This is a difference from Perl.)
932    
933    28. When duplicate subpattern names are present (necessarily with different
934        numbers, as required by 27 above), and a test is made by name in a
935        conditional pattern, either for a subpattern having been matched, or for
936        recursion in such a pattern, all the associated numbered subpatterns are
937        tested, and the overall condition is true if the condition is true for any
938        one of them. This is the way Perl works, and is also more like the way
939        testing by number works.
940    
941    
942    Version 7.9 11-Apr-09
943    ---------------------
944    
945    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
946        (pcretest), all targets were linked against these libraries. This included
947        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
948        libraries. This caused unwanted dependencies to be created. This problem
949        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
950        pcretest is linked with readline.
951    
952    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
953        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
954        moved outside it again, because FALSE and TRUE are already defined in AIX,
955        but BOOL is not.
956    
957    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
958        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
959    
960    4.  The pcregrep documentation said spaces were inserted as well as colons (or
961        hyphens) following file names and line numbers when outputting matching
962        lines. This is not true; no spaces are inserted. I have also clarified the
963        wording for the --colour (or --color) option.
964    
965    5.  In pcregrep, when --colour was used with -o, the list of matching strings
966        was not coloured; this is different to GNU grep, so I have changed it to be
967        the same.
968    
969    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
970        each matching line was coloured. Now it goes on to look for further matches
971        of any of the test patterns, which is the same behaviour as GNU grep.
972    
973    7.  A pattern that could match an empty string could cause pcregrep to loop; it
974        doesn't make sense to accept an empty string match in pcregrep, so I have
975        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
976        seems to be how GNU grep behaves.
977    
978    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
979        start or after a newline", because the conditional assertion was not being
980        correctly handled. The rule now is that both the assertion and what follows
981        in the first alternative must satisfy the test.
982    
983    9.  If auto-callout was enabled in a pattern with a conditional group whose
984        condition was an assertion, PCRE could crash during matching, both with
985        pcre_exec() and pcre_dfa_exec().
986    
987    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
988        used for matching.
989    
990    11. Unicode property support in character classes was not working for
991        characters (bytes) greater than 127 when not in UTF-8 mode.
992    
993    12. Added the -M command line option to pcretest.
994    
995    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
996    
997    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
998    
999    16. Added comments and documentation about mis-use of no_arg in the C++
1000        wrapper.
1001    
1002    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1003        from Martin Jerabek that uses macro names for all relevant character and
1004        string constants.
1005    
1006    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1007        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1008        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1009        these, but not everybody uses configure.
1010    
1011    19. A conditional group that had only one branch was not being correctly
1012        recognized as an item that could match an empty string. This meant that an
1013        enclosing group might also not be so recognized, causing infinite looping
1014        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1015        with the subject "ab", where knowledge that the repeated group can match
1016        nothing is needed in order to break the loop.
1017    
1018    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1019        exec(), but without supplying a callout function, matching went wrong.
1020    
1021    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1022        leak if the size of the offset vector was greater than 30. When the vector
1023        is smaller, the saved offsets during recursion go onto a local stack
1024        vector, but for larger vectors malloc() is used. It was failing to free
1025        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1026        error, in fact).
1027    
1028    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1029        heapframe that is used only when UTF-8 support is enabled. This caused no
1030        problem, but was untidy.
1031    
1032    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1033        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1034        included within another project.
1035    
1036    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1037        slightly modified by me:
1038    
1039          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1040              not building pcregrep.
1041    
1042          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1043              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1044    
1045    25. Forward references, both numeric and by name, in patterns that made use of
1046        duplicate group numbers, could behave incorrectly or give incorrect errors,
1047        because when scanning forward to find the reference group, PCRE was not
1048        taking into account the duplicate group numbers. A pattern such as
1049        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1050    
1051    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1052        the feature of a custom pointer more persuasive (as requested by a user).
1053    
1054    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1055        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1056        pre-defined.
1057    
1058    28. Added support for (*UTF8) at the start of a pattern.
1059    
1060    29. Arrange for flags added by the "release type" setting in CMake to be shown
1061        in the configuration summary.
1062    
1063    
1064    Version 7.8 05-Sep-08
1065    ---------------------
1066    
1067    1.  Replaced UCP searching code with optimized version as implemented for Ad
1068        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1069        stage table and inline lookup instead of a function, giving speed ups of 2
1070        to 5 times on some simple patterns that I tested. Permission was given to
1071        distribute the MultiStage2.py script that generates the tables (it's not in
1072        the tarball, but is in the Subversion repository).
1073    
1074    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1075        scripts.
1076    
1077    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1078        a group with a zero qualifier. The result of the study could be incorrect,
1079        or the function might crash, depending on the pattern.
1080    
1081    4.  Caseless matching was not working for non-ASCII characters in back
1082        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1083        It now works when Unicode Property Support is available.
1084    
1085    5.  In pcretest, an escape such as \x{de} in the data was always generating
1086        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1087        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1088        truncation.
1089    
1090    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1091    
1092    7.  Added two (int) casts to pcregrep when printing the difference of two
1093        pointers, in case they are 64-bit values.
1094    
1095    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1096        test 2 if it fails.
1097    
1098    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1099        and a #define of that name to empty if it is not externally set. This is to
1100        allow users of MSVC to set it if necessary.
1101    
1102    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1103        the convenience functions in the pcre_get.c source file.
1104    
1105    11. An option change at the start of a pattern that had top-level alternatives
1106        could cause overwriting and/or a crash. This command provoked a crash in
1107        some environments:
1108    
1109          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1110    
1111        This potential security problem was recorded as CVE-2008-2371.
1112    
1113    12. For a pattern where the match had to start at the beginning or immediately
1114        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1115        pcre_dfa_exec() could read past the end of the passed subject if there was
1116        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1117        pcretest so that it places the subject at the end of its malloc-ed buffer.
1118    
1119    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1120        exec() might read past the end of the data buffer in UTF-8 mode.
1121    
1122    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1123        the data contained the byte 0x85 as part of a UTF-8 character within its
1124        first line. This applied both to normal and DFA matching.
1125    
1126    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1127        /^[^d]*?$/8 failed to match "abc".
1128    
1129    16. Added a missing copyright notice to pcrecpp_internal.h.
1130    
1131    17. Make it more clear in the documentation that values returned from
1132        pcre_exec() in ovector are byte offsets, not character counts.
1133    
1134    18. Tidied a few places to stop certain compilers from issuing warnings.
1135    
1136    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1137        supplied by Stefan Weber. I made a further small update for 7.8 because
1138        there is a change of source arrangements: the pcre_searchfuncs.c module is
1139        replaced by pcre_ucd.c.
1140    
1141    
1142    Version 7.7 07-May-08
1143    ---------------------
1144    
1145    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1146        a string to a long long, pretend we don't even have a long long." This is
1147        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1148    
1149    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1150        pre-7.6 versions, which defined a global no_arg variable instead of putting
1151        it in the RE class. (See also #8 below.)
1152    
1153    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1154        Lopes.
1155    
1156    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1157    
1158        (1) The include/exclude patterns were being applied to the whole pathnames
1159            of files, instead of just to the final components.
1160    
1161        (2) If there was more than one level of directory, the subdirectories were
1162            skipped unless they satisfied the include/exclude conditions. This is
1163            inconsistent with GNU grep (and could even be seen as contrary to the
1164            pcregrep specification - which I improved to make it absolutely clear).
1165            The action now is always to scan all levels of directory, and just
1166            apply the include/exclude patterns to regular files.
1167    
1168    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1169        --exclude_dir in the tests to avoid scanning .svn directories.
1170    
1171    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1172        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1173        doesn't support NULs in patterns.
1174    
1175    7.  Added some missing "const"s to declarations of static tables in
1176        pcre_compile.c and pcre_dfa_exec.c.
1177    
1178    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1179        caused by fix #2  above. (Subsequently also a second patch to fix the
1180        first patch. And a third patch - this was a messy problem.)
1181    
1182    9.  Applied Craig's patch to remove the use of push_back().
1183    
1184    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1185        matching function regexec().
1186    
1187    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1188        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1189        references. PCRE supports relative numbers with this syntax (I don't think
1190        Oniguruma does).
1191    
1192    12. Previously, a group with a zero repeat such as (...){0} was completely
1193        omitted from the compiled regex. However, this means that if the group
1194        was called as a subroutine from elsewhere in the pattern, things went wrong
1195        (an internal error was given). Such groups are now left in the compiled
1196        pattern, with a new opcode that causes them to be skipped at execution
1197        time.
1198    
1199    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1200        to the way PCRE behaves:
1201    
1202        (a) A lone ] character is dis-allowed (Perl treats it as data).
1203    
1204        (b) A back reference to an unmatched subpattern matches an empty string
1205            (Perl fails the current match path).
1206    
1207        (c) A data ] in a character class must be notated as \] because if the
1208            first data character in a class is ], it defines an empty class. (In
1209            Perl it is not possible to have an empty class.) The empty class []
1210            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1211            The negative empty class [^] matches any one character, independently
1212            of the DOTALL setting.
1213    
1214    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1215        non-existent subpattern following a character class starting with ']' and
1216        containing () gave an internal compiling error instead of "reference to
1217        non-existent subpattern". Fortunately, when the pattern did exist, the
1218        compiled code was correct. (When scanning forwards to check for the
1219        existencd of the subpattern, it was treating the data ']' as terminating
1220        the class, so got the count wrong. When actually compiling, the reference
1221        was subsequently set up correctly.)
1222    
1223    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1224        it was being rejected as not supported by pcre_dfa_exec(), even though
1225        other assertions are supported. I have made pcre_dfa_exec() support
1226        (*FAIL).
1227    
1228    16. The implementation of 13c above involved the invention of a new opcode,
1229        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1230        cannot be changed at match time, I realized I could make a small
1231        improvement to matching performance by compiling OP_ALLANY instead of
1232        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1233        on the OP_ANY path.
1234    
1235    17. Compiling pcretest on Windows with readline support failed without the
1236        following two fixes: (1) Make the unistd.h include conditional on
1237        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1238    
1239    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1240        ncurses library to be included for pcretest when ReadLine support is
1241        requested, but also to allow for it to be overridden. This patch came from
1242        Daniel Bergström.
1243    
1244    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1245        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1246        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1247        spotting this.
1248    
1249    
1250    Version 7.6 28-Jan-08
1251    ---------------------
1252    
1253    1.  A character class containing a very large number of characters with
1254        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1255        overflow.
1256    
1257    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1258        HAVE_LONG_LONG is not defined.
1259    
1260    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1261        bring them up to date and include new features. This patch includes:
1262    
1263        - Fixed PH's badly added libz and libbz2 support.
1264        - Fixed a problem with static linking.
1265        - Added pcredemo. [But later removed - see 7 below.]
1266        - Fixed dftables problem and added an option.
1267        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1268            HAVE_LONG_LONG.
1269        - Added readline support for pcretest.
1270        - Added an listing of the option settings after cmake has run.
1271    
1272    4.  A user submitted a patch to Makefile that makes it easy to create
1273        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1274        Makefile.am that cause it to include this special target, without
1275        affecting anything else. Note that the same mingw target plus all
1276        the other distribution libraries and programs are now supported
1277        when configuring with CMake (see 6 below) instead of with
1278        Configure/Make.
1279    
1280    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1281        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1282        exported in the Windows port". It has not yet been confirmed that the patch
1283        solves the problem, but it does no harm.
1284    
1285    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1286        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1287        with CMake, and also correct the comment about stack recursion.
1288    
1289    7.  Remove the automatic building of pcredemo from the ./configure system and
1290        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1291        of a program that users should build themselves after PCRE is installed, so
1292        building it automatically is not really right. What is more, it gave
1293        trouble in some build environments.
1294    
1295    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1296    
1297    
1298    Version 7.5 10-Jan-08
1299    ---------------------
1300    
1301    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1302        values in parens when parsing an RE using the C++ wrapper."
1303    
1304    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1305        Characters greater than 255 were excluded from the class instead of being
1306        included.
1307    
1308    3.  The same bug as (2) above applied to negated POSIX classes such as
1309        [:^space:].
1310    
1311    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1312        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1313        I have changed it.
1314    
1315    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1316        first named subpattern) and a construct such as (?&a) would reference the
1317        first named subpattern whose name started with "a" (in other words, the
1318        length check was missing). Both these problems are fixed. "Subpattern name
1319        expected" is now given for (?&) (a zero-length name), and this patch also
1320        makes it give the same error for \k'' (previously it complained that that
1321        was a reference to a non-existent subpattern).
1322    
1323    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1324        this is right because (?- can be followed by option settings as well as by
1325        digits. I have, however, made the messages clearer.
1326    
1327    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1328        than the number used in the conditional) now cause a compile-time error.
1329        This is actually not compatible with Perl, which accepts such patterns, but
1330        treats the conditional as always being FALSE (as PCRE used to), but it
1331        seems to me that giving a diagnostic is better.
1332    
1333    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1334        and messages.
1335    
1336    9.  Fix two occurrences of "backslash" in comments that should have been
1337        "backspace".
1338    
1339    10. Remove two redundant lines of code that can never be obeyed (their function
1340        was moved elsewhere).
1341    
1342    11. The program that makes PCRE's Unicode character property table had a bug
1343        which caused it to generate incorrect table entries for sequences of
1344        characters that have the same character type, but are in different scripts.
1345        It amalgamated them into a single range, with the script of the first of
1346        them. In other words, some characters were in the wrong script. There were
1347        thirteen such cases, affecting characters in the following ranges:
1348    
1349          U+002b0 - U+002c1
1350          U+0060c - U+0060d
1351          U+0061e - U+00612
1352          U+0064b - U+0065e
1353          U+0074d - U+0076d
1354          U+01800 - U+01805
1355          U+01d00 - U+01d77
1356          U+01d9b - U+01dbf
1357          U+0200b - U+0200f
1358          U+030fc - U+030fe
1359          U+03260 - U+0327f
1360          U+0fb46 - U+0fbb1
1361          U+10450 - U+1049d
1362    
1363    12. The -o option (show only the matching part of a line) for pcregrep was not
1364        compatible with GNU grep in that, if there was more than one match in a
1365        line, it showed only the first of them. It now behaves in the same way as
1366        GNU grep.
1367    
1368    13. If the -o and -v options were combined for pcregrep, it printed a blank
1369        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1370        does the same. The return code can be used to tell if there were any
1371        non-matching lines.
1372    
1373    14. Added --file-offsets and --line-offsets to pcregrep.
1374    
1375    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1376        infinitely looping recursion. The bug was that positive lookaheads were not
1377        being skipped when checking for a possible empty match (negative lookaheads
1378        and both kinds of lookbehind were skipped).
1379    
1380    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1381        inclusion of <windows.h> to before rather than after the definition of
1382        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1383    
1384    17. Specifying a possessive quantifier with a specific limit for a Unicode
1385        character property caused pcre_compile() to compile bad code, which led at
1386        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1387        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1388        caused the error; without that there was no problem.
1389    
1390    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1391    
1392    19. Added --enable-pcretest-libreadline.
1393    
1394    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1395        RE::GlobalReplace(). As a result, the number of replacements returned was
1396        double what it should be. I removed one of the increments, but Craig sent a
1397        later patch that removed the other one (the right fix) and added unit tests
1398        that check the return values (which was not done before).
1399    
1400    21. Several CMake things:
1401    
1402        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1403            the names libpcre and libpcreposix, not just pcre and pcreposix.
1404    
1405        (2) The above change means that pcretest and pcregrep are now correctly
1406            linked with the newly-built libraries, not previously installed ones.
1407    
1408        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1409    
1410    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1411        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1412        UTF-8 newline character). The key issue is that the pattern starts .*;
1413        this means that the match must be either at the beginning, or after a
1414        newline. The bug was in the code for advancing after a failed match and
1415        checking that the new position followed a newline. It was not taking
1416        account of UTF-8 characters correctly.
1417    
1418    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1419        character classes. PCRE was not treating the sequence [:...:] as a
1420        character class unless the ... were all letters. Perl, however, seems to
1421        allow any characters between [: and :], though of course it rejects as
1422        unknown any "names" that contain non-letters, because all the known class
1423        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1424        for example, whereas PCRE did not - it did not recognize a POSIX character
1425        class. This seemed a bit dangerous, so the code has been changed to be
1426        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1427        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1428        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1429        Perl does, and where it didn't before.
1430    
1431    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1432        Windows environments %n is disabled by default.
1433    
1434    
1435    Version 7.4 21-Sep-07
1436    ---------------------
1437    
1438    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1439        means that a class such as [\s] counted as "explicit reference to CR or
1440        LF". That isn't really right - the whole point of the change was to try to
1441        help when there was an actual mention of one of the two characters. So now
1442        the change happens only if \r or \n (or a literal CR or LF) character is
1443        encountered.
1444    
1445    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1446        of both had grown to the point where there were only 3 bits left.
1447        Fortunately, there was spare space in the data structure, and so I have
1448        moved the internal flags into a new 16-bit field to free up more option
1449        bits.
1450    
1451    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1452        but did not set the internal JCHANGED flag - either of these is enough to
1453        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1454        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1455        start sets both bits.
1456    
1457    4.  Added options (at build time, compile time, exec time) to change \R from
1458        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1459    
1460    5.  doc/pcresyntax.html was missing from the distribution.
1461    
1462    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1463        compatibility, even though it is no longer used.
1464    
1465    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1466        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1467        windows.h file is present (where different names are used). [This was
1468        reversed later after testing - see 16 below.]
1469    
1470    8.  Changed all #include <config.h> to #include "config.h". There were also
1471        some further <pcre.h> cases that I changed to "pcre.h".
1472    
1473    9.  When pcregrep was used with the --colour option, it missed the line ending
1474        sequence off the lines that it output.
1475    
1476    10. It was pointed out to me that arrays of string pointers cause lots of
1477        relocations when a shared library is dynamically loaded. A technique of
1478        using a single long string with a table of offsets can drastically reduce
1479        these. I have refactored PCRE in four places to do this. The result is
1480        dramatic:
1481    
1482          Originally:                          290
1483          After changing UCP table:            187
1484          After changing error message table:   43
1485          After changing table of "verbs"       36
1486          After changing table of Posix names   22
1487    
1488        Thanks to the folks working on Gregex for glib for this insight.
1489    
1490    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1491        unicode-properties was also set.
1492    
1493    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1494    
1495    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1496        checked only for CRLF.
1497    
1498    14. Added casts to pcretest.c to avoid compiler warnings.
1499    
1500    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1501    
1502    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1503        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1504        entirely. This removes changes made in 7 above.
1505    
1506    17. The CMake files have been updated, and there is now more information about
1507        building with CMake in the NON-UNIX-USE document.
1508    
1509    
1510    Version 7.3 28-Aug-07
1511    ---------------------
1512    
1513     1. In the rejigging of the build system that eventually resulted in 7.1, the
1514        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1515        brackets there is not right, since it causes compilers to look for an
1516        installed pcre.h, not the version that is in the source that is being
1517        compiled (which of course may be different). I have changed it back to:
1518    
1519          #include "pcre.h"
1520    
1521        I have a vague recollection that the change was concerned with compiling in
1522        different directories, but in the new build system, that is taken care of
1523        by the VPATH setting the Makefile.
1524    
1525     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1526        when the subject happened to end in the byte 0x85 (e.g. if the last
1527        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1528        characters but of course it shouldn't be taken as a newline when it is part
1529        of another character. The bug was that, for an unlimited repeat of . in
1530        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1531        characters when looking for a newline.
1532    
1533     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1534    
1535     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1536        in debug output.
1537    
1538     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1539        long printing in the pcrecpp unittest when running under MinGW.
1540    
1541     6. ESC_K was left out of the EBCDIC table.
1542    
1543     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1544        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1545        limit also applies to "virtual nesting" when a pattern is recursive, and in
1546        this case 1000 isn't so big. I have been able to remove this limit at the
1547        expense of backing off one optimization in certain circumstances. Normally,
1548        when pcre_exec() would call its internal match() function recursively and
1549        immediately return the result unconditionally, it uses a "tail recursion"
1550        feature to save stack. However, when a subpattern that can match an empty
1551        string has an unlimited repetition quantifier, it no longer makes this
1552        optimization. That gives it a stack frame in which to save the data for
1553        checking that an empty string has been matched. Previously this was taken
1554        from the 1000-entry workspace that had been reserved. So now there is no
1555        explicit limit, but more stack is used.
1556    
1557     8. Applied Daniel's patches to solve problems with the import/export magic
1558        syntax that is required for Windows, and which was going wrong for the
1559        pcreposix and pcrecpp parts of the library. These were overlooked when this
1560        problem was solved for the main library.
1561    
1562     9. There were some crude static tests to avoid integer overflow when computing
1563        the size of patterns that contain repeated groups with explicit upper
1564        limits. As the maximum quantifier is 65535, the maximum group length was
1565        set at 30,000 so that the product of these two numbers did not overflow a
1566        32-bit integer. However, it turns out that people want to use groups that
1567        are longer than 30,000 bytes (though not repeat them that many times).
1568        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1569        made it possible to implement the integer overflow checks in a much more
1570        dynamic way, which I have now done. The artificial limitation on group
1571        length has been removed - we now have only the limit on the total length of
1572        the compiled pattern, which depends on the LINK_SIZE setting.
1573    
1574    10. Fixed a bug in the documentation for get/copy named substring when
1575        duplicate names are permitted. If none of the named substrings are set, the
1576        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1577        empty string.
1578    
1579    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1580        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1581        because the ] is interpreted as the first data character and the
1582        terminating ] is not found. PCRE has been made compatible with Perl in this
1583        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1584        cause memory overwriting.
1585    
1586    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1587        string has been matched (to stop an infinite loop). It was not recognizing
1588        a conditional subpattern that could match an empty string if that
1589        subpattern was within another subpattern. For example, it looped when
1590        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1591        condition was not nested. This bug has been fixed.
1592    
1593    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1594        past the start of the subject in the presence of bytes with the top bit
1595        set, for example "\x8aBCD".
1596    
1597    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1598        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1599    
1600    14. Optimized (?!) to (*FAIL).
1601    
1602    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1603        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1604        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1605        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1606        does: it's just the validity check that is more restrictive.
1607    
1608    16. Inserted checks for integer overflows during escape sequence (backslash)
1609        processing, and also fixed erroneous offset values for syntax errors during
1610        backslash processing.
1611    
1612    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1613        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1614    
1615    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1616        caused an overrun.
1617    
1618    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1619        something other than just ASCII characters) inside a group that had an
1620        unlimited repeat caused a loop at compile time (while checking to see
1621        whether the group could match an empty string).
1622    
1623    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1624        [\P{Any}] did so. (Error in the code for printing property names.)
1625    
1626    21. An orphan \E inside a character class could cause a crash.
1627    
1628    22. A repeated capturing bracket such as (A)? could cause a wild memory
1629        reference during compilation.
1630    
1631    23. There are several functions in pcre_compile() that scan along a compiled
1632        expression for various reasons (e.g. to see if it's fixed length for look
1633        behind). There were bugs in these functions when a repeated \p or \P was
1634        present in the pattern. These operators have additional parameters compared
1635        with \d, etc, and these were not being taken into account when moving along
1636        the compiled data. Specifically:
1637    
1638        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1639            length.
1640    
1641        (b) An item such as \pL+ within a repeated group could cause crashes or
1642            loops.
1643    
1644        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1645            "reference to non-existent subpattern" error.
1646    
1647        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1648    
1649    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1650        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1651    
1652    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1653    
1654    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1655        character were causing crashes (broken optimization).
1656    
1657    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1658        \p or \P) caused a compile-time loop.
1659    
1660    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1661        break. For example, the unstudied pattern [\r\n]A does not match the string
1662        "\r\nA" because change 7.0/46 below moves the current point on by two
1663        characters after failing to match at the start. However, the pattern \nA
1664        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1665        the same is true. There doesn't seem any very clean way out of this, but
1666        what I have chosen to do makes the common cases work: PCRE now takes note
1667        of whether there can be an explicit match for \r or \n anywhere in the
1668        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1669        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1670        pattern has explicit CR or LF references.
1671    
1672    29. Added (*CR) etc for changing newline setting at start of pattern.
1673    
1674    
1675    Version 7.2 19-Jun-07
1676    ---------------------
1677    
1678     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1679        which is apparently normally available under Windows.
1680    
1681     2. Re-jig the pcregrep tests with different newline settings in an attempt
1682        to make them independent of the local environment's newline setting.
1683    
1684     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1685    
1686     4. Some of the "internals" tests were previously cut out when the link size
1687        was not 2, because the output contained actual offsets. The recent new
1688        "Z" feature of pcretest means that these can be cut out, making the tests
1689        usable with all link sizes.
1690    
1691     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1692        stack recursion. This gives a massive performance boost under BSD, but just
1693        a small improvement under Linux. However, it saves one field in the frame
1694        in all cases.
1695    
1696     6. Added more features from the forthcoming Perl 5.10:
1697    
1698        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1699            recursion call. It refers to the nth most recently opened parentheses.
1700    
1701        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1702            to be opened parentheses.
1703    
1704        (c) Conditions that refer to capturing parentheses can be specified
1705            relatively, for example, (?(-2)... or (?(+3)...
1706    
1707        (d) \K resets the start of the current match so that everything before
1708            is not part of it.
1709    
1710        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1711    
1712        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1713            reference syntax.
1714    
1715        (g) (?| introduces a group in which the numbering of parentheses in each
1716            alternative starts with the same number.
1717    
1718        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1719    
1720     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1721        PCRE_INFO_JCHANGED.
1722    
1723     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1724        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1725        for detecting groups that can match an empty string.
1726    
1727     9. A pattern with a very large number of alternatives (more than several
1728        hundred) was running out of internal workspace during the pre-compile
1729        phase, where pcre_compile() figures out how much memory will be needed. A
1730        bit of new cunning has reduced the workspace needed for groups with
1731        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1732        workspace instead of running out of the 4096 that are available.
1733    
1734    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1735    
1736    11. Applied patch from Google to remove an optimization that didn't quite work.
1737        The report of the bug said:
1738    
1739          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1740          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1741          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1742    
1743    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1744        it matched the wrong number of bytes.
1745    
1746    
1747    Version 7.1 24-Apr-07
1748  ---------------------  ---------------------
1749    
1750   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
1751      that is more "standard", making use of automake and other autotools. There      that is more "standard", making use of automake and other Autotools. There
1752      is some re-arrangement of the files and adjustment of comments consequent      is some re-arrangement of the files and adjustment of comments consequent
1753      on this.      on this.
1754    
# Line 26  Version 7.1 05-Mar-07 Line 1769  Version 7.1 05-Mar-07
1769      .br or .in.      .br or .in.
1770    
1771   4. Updated comments in configure.ac that get placed in config.h.in and also   4. Updated comments in configure.ac that get placed in config.h.in and also
1772      arranged for config.h to be included in the distribution, with the name      arranged for config.h to be included in the distribution, with the name
1773      config.h.generic, for the benefit of those who have to compile without      config.h.generic, for the benefit of those who have to compile without
1774      Autotools (compare pcre.h, which is now distributed as pcre.h.generic).      Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1775    
1776   5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan   5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1777      Weber: (1) pcre_internal.h was missing some function renames; (2) updated      Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1778      makevp.bat for the current PCRE, using the additional files !compile.txt,      makevp.bat for the current PCRE, using the additional files
1779      !linklib.txt, and pcregexp.pas.      makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1780    
1781   6. A Windows user reported a minor discrepancy with test 2, which turned out   6. A Windows user reported a minor discrepancy with test 2, which turned out
1782      to be caused by a trailing space on an input line that had got lost in his      to be caused by a trailing space on an input line that had got lost in his
# Line 60  Version 7.1 05-Mar-07 Line 1803  Version 7.1 05-Mar-07
1803      called when UTF-8 support is disabled. Otherwise there are problems with a      called when UTF-8 support is disabled. Otherwise there are problems with a
1804      shared library.      shared library.
1805    
1806  10. The emulated memmove() function in pcre_internal.h (provided for those  10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1807      environments that have neither memmove() nor bcopy()) was defining its  
1808      arguments as char * instead of void *.      (a) It was defining its arguments as char * instead of void *.
1809    
1810        (b) It was assuming that all moves were upwards in memory; this was true
1811            a long time ago when I wrote it, but is no longer the case.
1812    
1813        The emulated memove() is provided for those environments that have neither
1814        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1815        is clearly not the case, as these two bugs were recently reported.
1816    
1817  11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,  11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1818      and Detrail to create the HTML documentation, the .txt form of the man      and Detrail to create the HTML documentation, the .txt form of the man
1819      pages, and it removes trailing spaces from listed files. It also creates      pages, and it removes trailing spaces from listed files. It also creates
1820      pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter      pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1821      case, it wraps all the #defines with #ifndefs. This script should be run      case, it wraps all the #defines with #ifndefs. This script should be run
1822      before "make dist".      before "make dist".
1823    
1824  12. Fixed two fairly obscure bugs concerned with quantified caseless matching  12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1825      with Unicode property support.      with Unicode property support.
1826    
1827      (a) For a maximizing quantifier, if the two different cases of the      (a) For a maximizing quantifier, if the two different cases of the
1828          character were of different lengths in their UTF-8 codings (there are          character were of different lengths in their UTF-8 codings (there are
1829          some cases like this - I found 11), and the matching function had to          some cases like this - I found 11), and the matching function had to
1830          back up over a mixture of the two cases, it incorrectly assumed they          back up over a mixture of the two cases, it incorrectly assumed they
1831          were both the same length.          were both the same length.
1832    
1833      (b) When PCRE was configured to use the heap rather than the stack for      (b) When PCRE was configured to use the heap rather than the stack for
1834          recursion during matching, it was not correctly preserving the data for          recursion during matching, it was not correctly preserving the data for
1835          the other case of a UTF-8 character when checking ahead for a match          the other case of a UTF-8 character when checking ahead for a match
1836          while processing a minimizing repeat. If the check also involved          while processing a minimizing repeat. If the check also involved
1837          matching a wide character, but failed, corruption could cause an          matching a wide character, but failed, corruption could cause an
1838          erroneous result when trying to check for a repeat of the original          erroneous result when trying to check for a repeat of the original
1839          character.          character.
1840    
1841  13. Some tidying changes to the testing mechanism:  13. Some tidying changes to the testing mechanism:
1842    
1843      (a) The RunTest script now detects the internal link size and whether there      (a) The RunTest script now detects the internal link size and whether there
1844          is UTF-8 and UCP support by running ./pcretest -C instead of relying on          is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1845          values substituted by "configure". (The RunGrepTest script already did          values substituted by "configure". (The RunGrepTest script already did
1846          this for UTF-8.) The configure.ac script no longer substitutes the          this for UTF-8.) The configure.ac script no longer substitutes the
1847          relevant variables.          relevant variables.
1848    
1849      (b) The debugging options /B and /D in pcretest show the compiled bytecode      (b) The debugging options /B and /D in pcretest show the compiled bytecode
1850          with length and offset values. This means that the output is different          with length and offset values. This means that the output is different
1851          for different internal link sizes. Test 2 is skipped for link sizes          for different internal link sizes. Test 2 is skipped for link sizes
1852          other than 2 because of this, bypassing the problem. Unfortunately,          other than 2 because of this, bypassing the problem. Unfortunately,
1853          there was also a test in test 3 (the locale tests) that used /B and          there was also a test in test 3 (the locale tests) that used /B and
1854          failed for link sizes other than 2. Rather than cut the whole test out,          failed for link sizes other than 2. Rather than cut the whole test out,
1855          I have added a new /Z option to pcretest that replaces the length and          I have added a new /Z option to pcretest that replaces the length and
1856          offset values with spaces. This is now used to make test 3 independent          offset values with spaces. This is now used to make test 3 independent
1857          of link size.          of link size. (Test 2 will be tidied up later.)
1858    
1859    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1860        segmentation fault instead of returning the appropriate error message.
1861    
1862    15. In multiline mode when the newline sequence was set to "any", the pattern
1863        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1864        This doesn't seem right; it now treats the CRLF combination as the line
1865        ending, and so does not match in that case. It's only a pattern such as ^$
1866        that would hit this one: something like ^ABC$ would have failed after \r
1867        and then tried again after \r\n.
1868    
1869    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1870        in an attempt to make files that differ only in their line terminators
1871        compare equal. This works on Linux.
1872    
1873    17. Under certain error circumstances pcregrep might try to free random memory
1874        as it exited. This is now fixed, thanks to valgrind.
1875    
1876    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1877        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1878        was because its rules for how to advance for /g after matching an empty
1879        string at the end of a line did not allow for this case. They now check for
1880        it specially.
1881    
1882    20. pcretest is supposed to handle patterns and data of any length, by
1883        extending its buffers when necessary. It was getting this wrong when the
1884        buffer for a data line had to be extended.
1885    
1886    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1887        CRLF as a newline sequence.
1888    
1889    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1890        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1891        I have nevertheless tidied it up.
1892    
1893    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1894    
1895    24. Added a man page for pcre-config.
1896    
1897    
1898  Version 7.0 19-Dec-06  Version 7.0 19-Dec-06

Legend:
Removed from v.116  
changed lines
  Added in v.712

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12