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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.107  
changed lines
  Added in v.699

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12