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

Diff of /code/trunk/ChangeLog

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

revision 59 by nigel, Sat Feb 24 21:39:54 2007 UTC revision 626 by ph10, Wed Jul 20 17:51:54 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.13 30-Apr-2011
5    ------------------------
6    
7    1.  The Unicode data tables have been updated to Unicode 6.0.0.
8    
9    2.  Two minor typos in pcre_internal.h have been fixed.
10    
11    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
12        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
13        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
14    
15    4.  There were a number of related bugs in the code for matching backrefences
16        caselessly in UTF-8 mode when codes for the characters concerned were
17        different numbers of bytes. For example, U+023A and U+2C65 are an upper
18        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
19        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
20        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
21        2-byte code at the end of the subject (it thought there wasn't enough data
22        left).
23    
24    5.  Comprehensive information about what went wrong is now returned by
25        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
26        as the output vector has at least 2 elements. The offset of the start of
27        the failing character and a reason code are placed in the vector.
28    
29    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
30        now returned is for the first byte of the failing character, instead of the
31        last byte inspected. This is an incompatible change, but I hope it is small
32        enough not to be a problem. It makes the returned offset consistent with
33        pcre_exec() and pcre_dfa_exec().
34    
35    7.  pcretest now gives a text phrase as well as the error number when
36        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
37        failure, the offset and reason code are output.
38    
39    8.  When \R was used with a maximizing quantifier it failed to skip backwards
40        over a \r\n pair if the subsequent match failed. Instead, it just skipped
41        back over a single character (\n). This seems wrong (because it treated the
42        two characters as a single entity when going forwards), conflicts with the
43        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
44        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
45        has been changed.
46    
47    9.  Some internal refactoring has changed the processing so that the handling
48        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
49        time (the PCRE_DOTALL option was changed this way some time ago: version
50        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
51        which was always a bit of a fudge. It also means that there is one less
52        argument for the match() function, which reduces its stack requirements
53        slightly. This change also fixes an incompatibility with Perl: the pattern
54        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
55    
56    10. More internal refactoring has drastically reduced the number of recursive
57        calls to match() for possessively repeated groups such as (abc)++ when
58        using pcre_exec().
59    
60    11. While implementing 10, a number of bugs in the handling of groups were
61        discovered and fixed:
62    
63        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
64        (a|)*(?1) gave a compile-time internal error.
65        ((a|)+)+  did not notice that the outer group could match an empty string.
66        (^a|^)+   was not marked as anchored.
67        (.*a|.*)+ was not marked as matching at start or after a newline.
68    
69    12. Yet more internal refactoring has removed another argument from the match()
70        function. Special calls to this function are now indicated by setting a
71        value in a variable in the "match data" data block.
72    
73    13. Be more explicit in pcre_study() instead of relying on "default" for
74        opcodes that mean there is no starting character; this means that when new
75        ones are added and accidentally left out of pcre_study(), testing should
76        pick them up.
77    
78    14. The -s option of pcretest has been documented for ages as being an old
79        synonym of -m (show memory usage). I have changed it to mean "force study
80        for every regex", that is, assume /S for every regex. This is similar to -i
81        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
82        using it. It makes it easier to run collections of tests with and without
83        study enabled, and thereby test pcre_study() more easily. All the standard
84        tests are now run with and without -s (but some patterns can be marked as
85        "never study" - see 20 below).
86    
87    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
88        restoration of the capturing data to the outer values was not happening
89        correctly.
90    
91    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
92        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
93        pattern had matched an empty string, and so incorrectly returned a no
94        match.
95    
96    17. There was optimizing code for the last branch of non-capturing parentheses,
97        and also for the obeyed branch of a conditional subexpression, which used
98        tail recursion to cut down on stack usage. Unfortunately, not that there is
99        the possibility of (*THEN) occurring in these branches, tail recursion is
100        no longer possible because the return has to be checked for (*THEN). These
101        two optimizations have therefore been removed.
102    
103    18. If a pattern containing \R was studied, it was assumed that \R always
104        matched two bytes, thus causing the minimum subject length to be
105        incorrectly computed because \R can also match just one byte.
106    
107    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
108        was incorrectly computed.
109    
110    20. If /S is present twice on a test pattern in pcretest input, it *disables*
111        studying, thereby overriding the use of -s on the command line. This is
112        necessary for one or two tests to keep the output identical in both cases.
113    
114    21. When (*ACCEPT) was used in an assertion that matched an empty string and
115        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
116    
117    22. When an atomic group that contained a capturing parenthesis was
118        successfully matched, but the branch in which it appeared failed, the
119        capturing was not being forgotten if a higher numbered group was later
120        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
121        group 1 to "a", when in fact it should be unset. This applied to multi-
122        branched capturing and non-capturing groups, repeated or not, and also to
123        positive assertions (capturing in negative assertions is not well defined
124        in PCRE) and also to nested atomic groups.
125    
126    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
127        subject after a captured substring (to make it easier to tell which of a
128        number of identical substrings has been captured).
129    
130    24. The way atomic groups are processed by pcre_exec() has been changed so that
131        if they are repeated, backtracking one repetition now resets captured
132        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
133        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
134        "aaa". Previously, it would have been "a". As part of this code
135        refactoring, the way recursive calls are handled has also been changed.
136    
137    24. If an assertion condition captured any substrings, they were not passed
138        back unless some other capturing happened later. For example, if
139        (?(?=(a))a) was matched against "a", no capturing was returned.
140    
141    25. When studying a pattern that contained subroutine calls or assertions,
142        the code for finding the minimum length of a possible match was handling
143        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
144        group 1 called group 2 while simultaneously a separate group 2 called group
145        1). A stack overflow occurred in this case. I have fixed this by limiting
146        the recursion depth to 10.
147    
148    26. Updated RunTest.bat in the distribution to the version supplied by Tom
149        Fortmann. This supports explicit test numbers on the command line, and has
150        argument validation and error reporting.
151    
152    27. An instance of \X with an unlimited repeat could fail if at any point the
153        first character it looked at was a mark character.
154    
155    28. Some minor code refactoring concerning Unicode properties and scripts
156        should reduce the stack requirement of match() slightly.
157    
158    29. Added the '=' option to pcretest to check the setting of unused capturing
159        slots at the end of the pattern, which are documented as being -1, but are
160        not included in the return count.
161    
162    
163    Version 8.12 15-Jan-2011
164    ------------------------
165    
166    1.  Fixed some typos in the markup of the man pages, and wrote a script that
167        checks for such things as part of the documentation building process.
168    
169    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
170        --match-limit and --recursion-limit options (added for 8.11). In
171        particular, this made one of the standard tests fail. (The integer value
172        went into the wrong half of a long int.)
173    
174    3.  If the --colour option was given to pcregrep with -v (invert match), it
175        did strange things, either producing crazy output, or crashing. It should,
176        of course, ignore a request for colour when reporting lines that do not
177        match.
178    
179    4.  Another pcregrep bug caused similar problems if --colour was specified with
180        -M (multiline) and the pattern match finished with a line ending.
181    
182    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
183        matched in multiline mode, the following line was shown as part of the
184        match. This seems wrong, so I have changed it.
185    
186    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
187        the check for further matches in the same line (so they could be coloured)
188        to overrun the end of the current line. If another match was found, it was
189        incorrectly shown (and then shown again when found in the next line).
190    
191    7.  If pcregrep was compiled under Windows, there was a reference to the
192        function pcregrep_exit() before it was defined. I am assuming this was
193        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
194        reported by a user. I've moved the definition above the reference.
195    
196    
197    Version 8.11 10-Dec-2010
198    ------------------------
199    
200    1.  (*THEN) was not working properly if there were untried alternatives prior
201        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
202        backtracked to try for "b" instead of moving to the next alternative branch
203        at the same level (in this case, to look for "c"). The Perl documentation
204        is clear that when (*THEN) is backtracked onto, it goes to the "next
205        alternative in the innermost enclosing group".
206    
207    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
208        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
209        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
210        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
211        (*THEN).
212    
213    3.  If \s appeared in a character class, it removed the VT character from
214        the class, even if it had been included by some previous item, for example
215        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
216        of \s, but is part of the POSIX "space" class.)
217    
218    4.  A partial match never returns an empty string (because you can always
219        match an empty string at the end of the subject); however the checking for
220        an empty string was starting at the "start of match" point. This has been
221        changed to the "earliest inspected character" point, because the returned
222        data for a partial match starts at this character. This means that, for
223        example, /(?<=abc)def/ gives a partial match for the subject "abc"
224        (previously it gave "no match").
225    
226    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
227        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
228        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
229        has an implication that the given string is incomplete (because a partial
230        match is preferred over a full match). For this reason, these items now
231        give a partial match in this situation. [Aside: previously, the one case
232        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
233        match rather than a full match, which was wrong by the old rules, but is
234        now correct.]
235    
236    6.  There was a bug in the handling of #-introduced comments, recognized when
237        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
238        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
239        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
240        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
241        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
242        places in pcre_compile().
243    
244    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
245        comments when looking ahead for named forward references to subpatterns,
246        the only newline sequence it recognized was NL. It now handles newlines
247        according to the set newline convention.
248    
249    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
250        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
251        cater for a lack of strerror(). These oversights have been fixed.
252    
253    9.  Added --match-limit and --recursion-limit to pcregrep.
254    
255    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
256    
257    11. When the -o option was used, pcregrep was setting a return code of 1, even
258        when matches were found, and --line-buffered was not being honoured.
259    
260    12. Added an optional parentheses number to the -o and --only-matching options
261        of pcregrep.
262    
263    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
264        can match an empty string. The code to do it in pcretest and pcredemo
265        needed fixing:
266    
267        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
268            only one byte after an empty string match just before CRLF (this case
269            just got forgotten; "any" and "anycrlf" were OK).
270    
271        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
272            mode when an empty string match preceded an ASCII character followed by
273            a non-ASCII character. (The code for advancing by one character rather
274            than one byte was nonsense.)
275    
276        (c) The pcredemo.c sample program did not have any code at all to handle
277            the cases when CRLF is a valid newline sequence.
278    
279    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
280        as a starting offset was within the subject string. There is now a new
281        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
282        negative or greater than the length of the string. In order to test this,
283        pcretest is extended to allow the setting of negative starting offsets.
284    
285    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
286        starting offset points to the beginning of a UTF-8 character was
287        unnecessarily clumsy. I tidied it up.
288    
289    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
290        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
291    
292    17. Nobody had reported that the --include_dir option, which was added in
293        release 7.7 should have been called --include-dir (hyphen, not underscore)
294        for compatibility with GNU grep. I have changed it to --include-dir, but
295        left --include_dir as an undocumented synonym, and the same for
296        --exclude-dir, though that is not available in GNU grep, at least as of
297        release 2.5.4.
298    
299    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
300        characters from a string of bytes) have been redefined so as not to use
301        loops, in order to improve performance in some environments. At the same
302        time, I abstracted some of the common code into auxiliary macros to save
303        repetition (this should not affect the compiled code).
304    
305    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
306        compile-time error is now given if \c is not followed by an ASCII
307        character, that is, a byte less than 128. (In EBCDIC mode, the code is
308        different, and any byte value is allowed.)
309    
310    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
311        START_OPTIMIZE option, which is now allowed at compile time - but just
312        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
313        to pcregrep and other applications that have no direct access to PCRE
314        options. The new /Y option in pcretest sets this option when calling
315        pcre_compile().
316    
317    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
318        back references. Groups containing recursive back references were forced to
319        be atomic by that change, but in the case of named groups, the amount of
320        memory required was incorrectly computed, leading to "Failed: internal
321        error: code overflow". This has been fixed.
322    
323    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
324        pcretest.c, to avoid build problems in some Borland environments.
325    
326    
327    Version 8.10 25-Jun-2010
328    ------------------------
329    
330    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
331        THEN.
332    
333    2.  (*ACCEPT) was not working when inside an atomic group.
334    
335    3.  Inside a character class, \B is treated as a literal by default, but
336        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
337        causes the error). The code is unchanged, but I tidied the documentation.
338    
339    4.  Inside a character class, PCRE always treated \R and \X as literals,
340        whereas Perl faults them if its -w option is set. I have changed PCRE so
341        that it faults them when PCRE_EXTRA is set.
342    
343    5.  Added support for \N, which always matches any character other than
344        newline. (It is the same as "." when PCRE_DOTALL is not set.)
345    
346    6.  When compiling pcregrep with newer versions of gcc which may have
347        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
348        declared with attribute warn_unused_result" were given. Just casting the
349        result to (void) does not stop the warnings; a more elaborate fudge is
350        needed. I've used a macro to implement this.
351    
352    7.  Minor change to pcretest.c to avoid a compiler warning.
353    
354    8.  Added four artifical Unicode properties to help with an option to make
355        \s etc use properties (see next item). The new properties are: Xan
356        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
357    
358    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
359        use Unicode properties. (*UCP) at the start of a pattern can be used to set
360        this option. Modified pcretest to add /W to test this facility. Added
361        REG_UCP to make it available via the POSIX interface.
362    
363    10. Added --line-buffered to pcregrep.
364    
365    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
366        studied, and the match started with a letter with a code point greater than
367        127 whose first byte was different to the first byte of the other case of
368        the letter, the other case of this starting letter was not recognized
369        (#976).
370    
371    12. If a pattern that was studied started with a repeated Unicode property
372        test, for example, \p{Nd}+, there was the theoretical possibility of
373        setting up an incorrect bitmap of starting bytes, but fortunately it could
374        not have actually happened in practice until change 8 above was made (it
375        added property types that matched character-matching opcodes).
376    
377    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
378        possible starting bytes for non-anchored patterns.
379    
380    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
381        \R, and also a number of cases that involve Unicode properties, both
382        explicit and implicit when PCRE_UCP is set.
383    
384    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
385        input, it could crash or give wrong results if characters with values
386        greater than 0xc0 were present in the subject string. (Detail: it assumed
387        UTF-8 input when processing these items.)
388    
389    16. Added a lot of (int) casts to avoid compiler warnings in systems where
390        size_t is 64-bit (#991).
391    
392    17. Added a check for running out of memory when PCRE is compiled with
393        --disable-stack-for-recursion (#990).
394    
395    18. If the last data line in a file for pcretest does not have a newline on
396        the end, a newline was missing in the output.
397    
398    19. The default pcre_chartables.c file recognizes only ASCII characters (values
399        less than 128) in its various bitmaps. However, there is a facility for
400        generating tables according to the current locale when PCRE is compiled. It
401        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
402        characters, are recognized by isspace() and therefore were getting set in
403        these tables, and indeed these tables seem to approximate to ISO 8859. This
404        caused a problem in UTF-8 mode when pcre_study() was used to create a list
405        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
406        which of course cannot start UTF-8 characters. I have changed the code so
407        that only real ASCII characters (less than 128) and the correct starting
408        bytes for UTF-8 encodings are set for characters greater than 127 when in
409        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
410        altogether.)
411    
412    20. Added the /T option to pcretest so as to be able to run tests with non-
413        standard character tables, thus making it possible to include the tests
414        used for 19 above in the standard set of tests.
415    
416    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
417        reference to a subpattern the other side of a comment that contains an
418        opening parenthesis caused either an internal compiling error, or a
419        reference to the wrong subpattern.
420    
421    
422    Version 8.02 19-Mar-2010
423    ------------------------
424    
425    1.  The Unicode data tables have been updated to Unicode 5.2.0.
426    
427    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
428        configured.
429    
430    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
431        original author of that file, following a query about its status.
432    
433    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
434        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
435    
436    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
437        quantifier applied to a forward-referencing subroutine call, could compile
438        incorrect code or give the error "internal error: previously-checked
439        referenced subpattern not found".
440    
441    6.  Both MS Visual Studio and Symbian OS have problems with initializing
442        variables to point to external functions. For these systems, therefore,
443        pcre_malloc etc. are now initialized to local functions that call the
444        relevant global functions.
445    
446    7.  There were two entries missing in the vectors called coptable and poptable
447        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
448        I've fixed the data, and added a kludgy way of testing at compile time that
449        the lengths are correct (equal to the number of opcodes).
450    
451    8.  Following on from 7, I added a similar kludge to check the length of the
452        eint vector in pcreposix.c.
453    
454    9.  Error texts for pcre_compile() are held as one long string to avoid too
455        much relocation at load time. To find a text, the string is searched,
456        counting zeros. There was no check for running off the end of the string,
457        which could happen if a new error number was added without updating the
458        string.
459    
460    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
461    
462    11. \K was not working if it appeared in an atomic group or in a group that
463        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
464        \K is "not well defined" if used in an assertion. PCRE now accepts it if
465        the assertion is positive, but not if it is negative.
466    
467    12. Change 11 fortuitously reduced the size of the stack frame used in the
468        "match()" function of pcre_exec.c by one pointer. Forthcoming
469        implementation of support for (*MARK) will need an extra pointer on the
470        stack; I have reserved it now, so that the stack frame size does not
471        decrease.
472    
473    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
474        item in branch that calls a recursion is a subroutine call - as in the
475        second branch in the above example - was incorrectly given the compile-
476        time error "recursive call could loop indefinitely" because pcre_compile()
477        was not correctly checking the subroutine for matching a non-empty string.
478    
479    14. The checks for overrunning compiling workspace could trigger after an
480        overrun had occurred. This is a "should never occur" error, but it can be
481        triggered by pathological patterns such as hundreds of nested parentheses.
482        The checks now trigger 100 bytes before the end of the workspace.
483    
484    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
485    
486    
487    Version 8.01 19-Jan-2010
488    ------------------------
489    
490    1.  If a pattern contained a conditional subpattern with only one branch (in
491        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
492        computed the wrong minimum data length (which is of course zero for such
493        subpatterns). This could cause incorrect "no match" results.
494    
495    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
496        the pattern is reset in the first branch, pcre_compile() failed with
497        "internal error: code overflow at offset...". This happened only when
498        the reset was to the original external option setting. (An optimization
499        abstracts leading options settings into an external setting, which was the
500        cause of this.)
501    
502    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
503        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
504        assertion pattern did not match (meaning that the assertion was true), it
505        was incorrectly treated as false if the SKIP had been reached during the
506        matching. This also applied to assertions used as conditions.
507    
508    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
509        assertion subpattern, including such a pattern used as a condition,
510        unpredictable results occurred, instead of the error return
511        PCRE_ERROR_DFA_UITEM.
512    
513    5.  The C++ GlobalReplace function was not working like Perl for the special
514        situation when an empty string is matched. It now does the fancy magic
515        stuff that is necessary.
516    
517    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
518        removed. (These were left over from very, very early versions of PCRE.)
519    
520    7.  Some cosmetic changes to the code to make life easier when compiling it
521        as part of something else:
522    
523        (a) Change DEBUG to PCRE_DEBUG.
524    
525        (b) In pcre_compile(), rename the member of the "branch_chain" structure
526            called "current" as "current_branch", to prevent a collision with the
527            Linux macro when compiled as a kernel module.
528    
529        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
530            prevent a collision with the Linux macro when compiled as a kernel
531            module.
532    
533    8.  In pcre_compile() there are some checks for integer overflows that used to
534        cast potentially large values to (double). This has been changed to that
535        when building, a check for int64_t is made, and if it is found, it is used
536        instead, thus avoiding the use of floating point arithmetic. (There is no
537        other use of FP in PCRE.) If int64_t is not found, the fallback is to
538        double.
539    
540    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
541        2005 (difference between two addresses compared to an unsigned value).
542    
543    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
544        custom one, because of the following reported problem in Windows:
545    
546          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
547              under Win32.
548          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
549              therefore missing the function definition.
550          - The compiler thus generates a "C" signature for the test function.
551          - The linker fails to find the "C" function.
552          - PCRE fails to configure if asked to do so against libbz2.
553    
554    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
555        messages were output:
556    
557          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
558          rerunning libtoolize, to keep the correct libtool macros in-tree.
559          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
560    
561        I have done both of these things.
562    
563    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
564        most of the time, it *can* run out if it is given a pattern that contains a
565        runaway infinite recursion. I updated the discussion in the pcrestack man
566        page.
567    
568    13. Now that we have gone to the x.xx style of version numbers, the minor
569        version may start with zero. Using 08 or 09 is a bad idea because users
570        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
571        interpreted as invalid octal numbers. I've updated the previous comment in
572        configure.ac, and also added a check that gives an error if 08 or 09 are
573        used.
574    
575    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
576        causing partial matching to fail when the end of the subject matched \W
577        in a UTF-8 pattern where \W was quantified with a minimum of 3.
578    
579    15. There were some discrepancies between the declarations in pcre_internal.h
580        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
581        their definitions. The declarations used "const uschar *" and the
582        definitions used USPTR. Even though USPTR is normally defined as "const
583        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
584        reported that: "This difference in casting confuses some C++ compilers, for
585        example, SunCC recognizes above declarations as different functions and
586        generates broken code for hbpcre." I have changed the declarations to use
587        USPTR.
588    
589    16. GNU libtool is named differently on some systems. The autogen.sh script now
590        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
591        (FreeBSD).
592    
593    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
594        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
595        comment: "Figure out how to create a longlong from a string: strtoll and
596        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
597        instance, but it only takes 2 args instead of 3!"
598    
599    18. A subtle bug concerned with back references has been fixed by a change of
600        specification, with a corresponding code fix. A pattern such as
601        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
602        refers, was giving matches when it shouldn't. For example, xa=xaaa would
603        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
604        same bug. Such groups have to be quantified to be useful, or contained
605        inside another quantified group. (If there's no repetition, the reference
606        can never match.) The problem arises because, having left the group and
607        moved on to the rest of the pattern, a later failure that backtracks into
608        the group uses the captured value from the final iteration of the group
609        rather than the correct earlier one. I have fixed this in PCRE by forcing
610        any group that contains a reference to itself to be an atomic group; that
611        is, there cannot be any backtracking into it once it has completed. This is
612        similar to recursive and subroutine calls.
613    
614    
615    Version 8.00 19-Oct-09
616    ----------------------
617    
618    1.  The table for translating pcre_compile() error codes into POSIX error codes
619        was out-of-date, and there was no check on the pcre_compile() error code
620        being within the table. This could lead to an OK return being given in
621        error.
622    
623    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
624        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
625        in a Windows environment.
626    
627    3.  The pcregrep --count option prints the count for each file even when it is
628        zero, as does GNU grep. However, pcregrep was also printing all files when
629        --files-with-matches was added. Now, when both options are given, it prints
630        counts only for those files that have at least one match. (GNU grep just
631        prints the file name in this circumstance, but including the count seems
632        more useful - otherwise, why use --count?) Also ensured that the
633        combination -clh just lists non-zero counts, with no names.
634    
635    4.  The long form of the pcregrep -F option was incorrectly implemented as
636        --fixed_strings instead of --fixed-strings. This is an incompatible change,
637        but it seems right to fix it, and I didn't think it was worth preserving
638        the old behaviour.
639    
640    5.  The command line items --regex=pattern and --regexp=pattern were not
641        recognized by pcregrep, which required --regex pattern or --regexp pattern
642        (with a space rather than an '='). The man page documented the '=' forms,
643        which are compatible with GNU grep; these now work.
644    
645    6.  No libpcreposix.pc file was created for pkg-config; there was just
646        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
647    
648    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
649        when UCP support is not needed, by modifying the Python script that
650        generates it from Unicode data files. This should not matter if the module
651        is correctly used as a library, but I received one complaint about 50K of
652        unwanted data. My guess is that the person linked everything into his
653        program rather than using a library. Anyway, it does no harm.
654    
655    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
656        was a minimum greater than 1 for a wide character in a possessive
657        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
658        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
659        character. Chaos in the form of incorrect output or a compiling loop could
660        result.
661    
662    9.  The restrictions on what a pattern can contain when partial matching is
663        requested for pcre_exec() have been removed. All patterns can now be
664        partially matched by this function. In addition, if there are at least two
665        slots in the offset vector, the offset of the earliest inspected character
666        for the match and the offset of the end of the subject are set in them when
667        PCRE_ERROR_PARTIAL is returned.
668    
669    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
670        synonymous with PCRE_PARTIAL, for backwards compatibility, and
671        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
672        and may be more useful for multi-segment matching.
673    
674    11. Partial matching with pcre_exec() is now more intuitive. A partial match
675        used to be given if ever the end of the subject was reached; now it is
676        given only if matching could not proceed because another character was
677        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
678        string "Z", which now yields "no match" instead of "partial match". In the
679        case of pcre_dfa_exec(), "no match" is given if every matching path for the
680        final character ended with (*FAIL).
681    
682    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
683        if the pattern had a "must contain" character that was already found in the
684        earlier partial match, unless partial matching was again requested. For
685        example, with the pattern /dog.(body)?/, the "must contain" character is
686        "g". If the first part-match was for the string "dog", restarting with
687        "sbody" failed. This bug has been fixed.
688    
689    13. The string returned by pcre_dfa_exec() after a partial match has been
690        changed so that it starts at the first inspected character rather than the
691        first character of the match. This makes a difference only if the pattern
692        starts with a lookbehind assertion or \b or \B (\K is not supported by
693        pcre_dfa_exec()). It's an incompatible change, but it makes the two
694        matching functions compatible, and I think it's the right thing to do.
695    
696    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
697        so that the demonstration program is easily available in environments where
698        PCRE has not been installed from source.
699    
700    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
701        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
702        library.
703    
704    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
705        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
706        is not the first non-POSIX option to be added. Clearly some people find
707        these options useful.
708    
709    17. If a caller to the POSIX matching function regexec() passes a non-zero
710        value for nmatch with a NULL value for pmatch, the value of
711        nmatch is forced to zero.
712    
713    18. RunGrepTest did not have a test for the availability of the -u option of
714        the diff command, as RunTest does. It now checks in the same way as
715        RunTest, and also checks for the -b option.
716    
717    19. If an odd number of negated classes containing just a single character
718        interposed, within parentheses, between a forward reference to a named
719        subpattern and the definition of the subpattern, compilation crashed with
720        an internal error, complaining that it could not find the referenced
721        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
722        [The bug was that it was starting one character too far in when skipping
723        over the character class, thus treating the ] as data rather than
724        terminating the class. This meant it could skip too much.]
725    
726    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
727        /g option in pcretest when the pattern contains \K, which makes it possible
728        to have an empty string match not at the start, even when the pattern is
729        anchored. Updated pcretest and pcredemo to use this option.
730    
731    21. If the maximum number of capturing subpatterns in a recursion was greater
732        than the maximum at the outer level, the higher number was returned, but
733        with unset values at the outer level. The correct (outer level) value is
734        now given.
735    
736    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
737        PCRE did not set those parentheses (unlike Perl). I have now found a way to
738        make it do so. The string so far is captured, making this feature
739        compatible with Perl.
740    
741    23. The tests have been re-organized, adding tests 11 and 12, to make it
742        possible to check the Perl 5.10 features against Perl 5.10.
743    
744    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
745        pattern matches a fixed length string. PCRE did not allow this; now it
746        does. Neither allows recursion.
747    
748    25. I finally figured out how to implement a request to provide the minimum
749        length of subject string that was needed in order to match a given pattern.
750        (It was back references and recursion that I had previously got hung up
751        on.) This code has now been added to pcre_study(); it finds a lower bound
752        to the length of subject needed. It is not necessarily the greatest lower
753        bound, but using it to avoid searching strings that are too short does give
754        some useful speed-ups. The value is available to calling programs via
755        pcre_fullinfo().
756    
757    26. While implementing 25, I discovered to my embarrassment that pcretest had
758        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
759        study optimizations had never been tested with that matching function.
760        Oops. What is worse, even when it was passed study data, there was a bug in
761        pcre_dfa_exec() that meant it never actually used it. Double oops. There
762        were also very few tests of studied patterns with pcre_dfa_exec().
763    
764    27. If (?| is used to create subpatterns with duplicate numbers, they are now
765        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
766        on the other side of the coin, they are no longer allowed to have different
767        names, because these cannot be distinguished in PCRE, and this has caused
768        confusion. (This is a difference from Perl.)
769    
770    28. When duplicate subpattern names are present (necessarily with different
771        numbers, as required by 27 above), and a test is made by name in a
772        conditional pattern, either for a subpattern having been matched, or for
773        recursion in such a pattern, all the associated numbered subpatterns are
774        tested, and the overall condition is true if the condition is true for any
775        one of them. This is the way Perl works, and is also more like the way
776        testing by number works.
777    
778    
779    Version 7.9 11-Apr-09
780    ---------------------
781    
782    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
783        (pcretest), all targets were linked against these libraries. This included
784        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
785        libraries. This caused unwanted dependencies to be created. This problem
786        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
787        pcretest is linked with readline.
788    
789    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
790        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
791        moved outside it again, because FALSE and TRUE are already defined in AIX,
792        but BOOL is not.
793    
794    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
795        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
796    
797    4.  The pcregrep documentation said spaces were inserted as well as colons (or
798        hyphens) following file names and line numbers when outputting matching
799        lines. This is not true; no spaces are inserted. I have also clarified the
800        wording for the --colour (or --color) option.
801    
802    5.  In pcregrep, when --colour was used with -o, the list of matching strings
803        was not coloured; this is different to GNU grep, so I have changed it to be
804        the same.
805    
806    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
807        each matching line was coloured. Now it goes on to look for further matches
808        of any of the test patterns, which is the same behaviour as GNU grep.
809    
810    7.  A pattern that could match an empty string could cause pcregrep to loop; it
811        doesn't make sense to accept an empty string match in pcregrep, so I have
812        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
813        seems to be how GNU grep behaves.
814    
815    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
816        start or after a newline", because the conditional assertion was not being
817        correctly handled. The rule now is that both the assertion and what follows
818        in the first alternative must satisfy the test.
819    
820    9.  If auto-callout was enabled in a pattern with a conditional group whose
821        condition was an assertion, PCRE could crash during matching, both with
822        pcre_exec() and pcre_dfa_exec().
823    
824    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
825        used for matching.
826    
827    11. Unicode property support in character classes was not working for
828        characters (bytes) greater than 127 when not in UTF-8 mode.
829    
830    12. Added the -M command line option to pcretest.
831    
832    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
833    
834    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
835    
836    16. Added comments and documentation about mis-use of no_arg in the C++
837        wrapper.
838    
839    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
840        from Martin Jerabek that uses macro names for all relevant character and
841        string constants.
842    
843    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
844        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
845        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
846        these, but not everybody uses configure.
847    
848    19. A conditional group that had only one branch was not being correctly
849        recognized as an item that could match an empty string. This meant that an
850        enclosing group might also not be so recognized, causing infinite looping
851        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
852        with the subject "ab", where knowledge that the repeated group can match
853        nothing is needed in order to break the loop.
854    
855    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
856        exec(), but without supplying a callout function, matching went wrong.
857    
858    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
859        leak if the size of the offset vector was greater than 30. When the vector
860        is smaller, the saved offsets during recursion go onto a local stack
861        vector, but for larger vectors malloc() is used. It was failing to free
862        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
863        error, in fact).
864    
865    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
866        heapframe that is used only when UTF-8 support is enabled. This caused no
867        problem, but was untidy.
868    
869    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
870        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
871        included within another project.
872    
873    24. Steven Van Ingelgem's patches to add more options to the CMake support,
874        slightly modified by me:
875    
876          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
877              not building pcregrep.
878    
879          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
880              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
881    
882    25. Forward references, both numeric and by name, in patterns that made use of
883        duplicate group numbers, could behave incorrectly or give incorrect errors,
884        because when scanning forward to find the reference group, PCRE was not
885        taking into account the duplicate group numbers. A pattern such as
886        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
887    
888    26. Changed a few more instances of "const unsigned char *" to USPTR, making
889        the feature of a custom pointer more persuasive (as requested by a user).
890    
891    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
892        pcretest.c, inside #ifndefs, because it seems they are sometimes already
893        pre-defined.
894    
895    28. Added support for (*UTF8) at the start of a pattern.
896    
897    29. Arrange for flags added by the "release type" setting in CMake to be shown
898        in the configuration summary.
899    
900    
901    Version 7.8 05-Sep-08
902    ---------------------
903    
904    1.  Replaced UCP searching code with optimized version as implemented for Ad
905        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
906        stage table and inline lookup instead of a function, giving speed ups of 2
907        to 5 times on some simple patterns that I tested. Permission was given to
908        distribute the MultiStage2.py script that generates the tables (it's not in
909        the tarball, but is in the Subversion repository).
910    
911    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
912        scripts.
913    
914    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
915        a group with a zero qualifier. The result of the study could be incorrect,
916        or the function might crash, depending on the pattern.
917    
918    4.  Caseless matching was not working for non-ASCII characters in back
919        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
920        It now works when Unicode Property Support is available.
921    
922    5.  In pcretest, an escape such as \x{de} in the data was always generating
923        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
924        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
925        truncation.
926    
927    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
928    
929    7.  Added two (int) casts to pcregrep when printing the difference of two
930        pointers, in case they are 64-bit values.
931    
932    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
933        test 2 if it fails.
934    
935    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
936        and a #define of that name to empty if it is not externally set. This is to
937        allow users of MSVC to set it if necessary.
938    
939    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
940        the convenience functions in the pcre_get.c source file.
941    
942    11. An option change at the start of a pattern that had top-level alternatives
943        could cause overwriting and/or a crash. This command provoked a crash in
944        some environments:
945    
946          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
947    
948        This potential security problem was recorded as CVE-2008-2371.
949    
950    12. For a pattern where the match had to start at the beginning or immediately
951        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
952        pcre_dfa_exec() could read past the end of the passed subject if there was
953        no match. To help with detecting such bugs (e.g. with valgrind), I modified
954        pcretest so that it places the subject at the end of its malloc-ed buffer.
955    
956    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
957        exec() might read past the end of the data buffer in UTF-8 mode.
958    
959    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
960        the data contained the byte 0x85 as part of a UTF-8 character within its
961        first line. This applied both to normal and DFA matching.
962    
963    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
964        /^[^d]*?$/8 failed to match "abc".
965    
966    16. Added a missing copyright notice to pcrecpp_internal.h.
967    
968    17. Make it more clear in the documentation that values returned from
969        pcre_exec() in ovector are byte offsets, not character counts.
970    
971    18. Tidied a few places to stop certain compilers from issuing warnings.
972    
973    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
974        supplied by Stefan Weber. I made a further small update for 7.8 because
975        there is a change of source arrangements: the pcre_searchfuncs.c module is
976        replaced by pcre_ucd.c.
977    
978    
979    Version 7.7 07-May-08
980    ---------------------
981    
982    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
983        a string to a long long, pretend we don't even have a long long." This is
984        done by checking for the strtoq, strtoll, and _strtoi64 functions.
985    
986    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
987        pre-7.6 versions, which defined a global no_arg variable instead of putting
988        it in the RE class. (See also #8 below.)
989    
990    3.  Remove a line of dead code, identified by coverity and reported by Nuno
991        Lopes.
992    
993    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
994    
995        (1) The include/exclude patterns were being applied to the whole pathnames
996            of files, instead of just to the final components.
997    
998        (2) If there was more than one level of directory, the subdirectories were
999            skipped unless they satisfied the include/exclude conditions. This is
1000            inconsistent with GNU grep (and could even be seen as contrary to the
1001            pcregrep specification - which I improved to make it absolutely clear).
1002            The action now is always to scan all levels of directory, and just
1003            apply the include/exclude patterns to regular files.
1004    
1005    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1006        --exclude_dir in the tests to avoid scanning .svn directories.
1007    
1008    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1009        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1010        doesn't support NULs in patterns.
1011    
1012    7.  Added some missing "const"s to declarations of static tables in
1013        pcre_compile.c and pcre_dfa_exec.c.
1014    
1015    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1016        caused by fix #2  above. (Subsequently also a second patch to fix the
1017        first patch. And a third patch - this was a messy problem.)
1018    
1019    9.  Applied Craig's patch to remove the use of push_back().
1020    
1021    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1022        matching function regexec().
1023    
1024    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1025        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1026        references. PCRE supports relative numbers with this syntax (I don't think
1027        Oniguruma does).
1028    
1029    12. Previously, a group with a zero repeat such as (...){0} was completely
1030        omitted from the compiled regex. However, this means that if the group
1031        was called as a subroutine from elsewhere in the pattern, things went wrong
1032        (an internal error was given). Such groups are now left in the compiled
1033        pattern, with a new opcode that causes them to be skipped at execution
1034        time.
1035    
1036    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1037        to the way PCRE behaves:
1038    
1039        (a) A lone ] character is dis-allowed (Perl treats it as data).
1040    
1041        (b) A back reference to an unmatched subpattern matches an empty string
1042            (Perl fails the current match path).
1043    
1044        (c) A data ] in a character class must be notated as \] because if the
1045            first data character in a class is ], it defines an empty class. (In
1046            Perl it is not possible to have an empty class.) The empty class []
1047            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1048            The negative empty class [^] matches any one character, independently
1049            of the DOTALL setting.
1050    
1051    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1052        non-existent subpattern following a character class starting with ']' and
1053        containing () gave an internal compiling error instead of "reference to
1054        non-existent subpattern". Fortunately, when the pattern did exist, the
1055        compiled code was correct. (When scanning forwards to check for the
1056        existencd of the subpattern, it was treating the data ']' as terminating
1057        the class, so got the count wrong. When actually compiling, the reference
1058        was subsequently set up correctly.)
1059    
1060    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1061        it was being rejected as not supported by pcre_dfa_exec(), even though
1062        other assertions are supported. I have made pcre_dfa_exec() support
1063        (*FAIL).
1064    
1065    16. The implementation of 13c above involved the invention of a new opcode,
1066        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1067        cannot be changed at match time, I realized I could make a small
1068        improvement to matching performance by compiling OP_ALLANY instead of
1069        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1070        on the OP_ANY path.
1071    
1072    17. Compiling pcretest on Windows with readline support failed without the
1073        following two fixes: (1) Make the unistd.h include conditional on
1074        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1075    
1076    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1077        ncurses library to be included for pcretest when ReadLine support is
1078        requested, but also to allow for it to be overridden. This patch came from
1079        Daniel Bergström.
1080    
1081    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1082        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1083        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1084        spotting this.
1085    
1086    
1087    Version 7.6 28-Jan-08
1088    ---------------------
1089    
1090    1.  A character class containing a very large number of characters with
1091        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1092        overflow.
1093    
1094    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1095        HAVE_LONG_LONG is not defined.
1096    
1097    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1098        bring them up to date and include new features. This patch includes:
1099    
1100        - Fixed PH's badly added libz and libbz2 support.
1101        - Fixed a problem with static linking.
1102        - Added pcredemo. [But later removed - see 7 below.]
1103        - Fixed dftables problem and added an option.
1104        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1105            HAVE_LONG_LONG.
1106        - Added readline support for pcretest.
1107        - Added an listing of the option settings after cmake has run.
1108    
1109    4.  A user submitted a patch to Makefile that makes it easy to create
1110        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1111        Makefile.am that cause it to include this special target, without
1112        affecting anything else. Note that the same mingw target plus all
1113        the other distribution libraries and programs are now supported
1114        when configuring with CMake (see 6 below) instead of with
1115        Configure/Make.
1116    
1117    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1118        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1119        exported in the Windows port". It has not yet been confirmed that the patch
1120        solves the problem, but it does no harm.
1121    
1122    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1123        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1124        with CMake, and also correct the comment about stack recursion.
1125    
1126    7.  Remove the automatic building of pcredemo from the ./configure system and
1127        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1128        of a program that users should build themselves after PCRE is installed, so
1129        building it automatically is not really right. What is more, it gave
1130        trouble in some build environments.
1131    
1132    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1133    
1134    
1135    Version 7.5 10-Jan-08
1136    ---------------------
1137    
1138    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1139        values in parens when parsing an RE using the C++ wrapper."
1140    
1141    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1142        Characters greater than 255 were excluded from the class instead of being
1143        included.
1144    
1145    3.  The same bug as (2) above applied to negated POSIX classes such as
1146        [:^space:].
1147    
1148    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1149        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1150        I have changed it.
1151    
1152    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1153        first named subpattern) and a construct such as (?&a) would reference the
1154        first named subpattern whose name started with "a" (in other words, the
1155        length check was missing). Both these problems are fixed. "Subpattern name
1156        expected" is now given for (?&) (a zero-length name), and this patch also
1157        makes it give the same error for \k'' (previously it complained that that
1158        was a reference to a non-existent subpattern).
1159    
1160    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1161        this is right because (?- can be followed by option settings as well as by
1162        digits. I have, however, made the messages clearer.
1163    
1164    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1165        than the number used in the conditional) now cause a compile-time error.
1166        This is actually not compatible with Perl, which accepts such patterns, but
1167        treats the conditional as always being FALSE (as PCRE used to), but it
1168        seems to me that giving a diagnostic is better.
1169    
1170    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1171        and messages.
1172    
1173    9.  Fix two occurrences of "backslash" in comments that should have been
1174        "backspace".
1175    
1176    10. Remove two redundant lines of code that can never be obeyed (their function
1177        was moved elsewhere).
1178    
1179    11. The program that makes PCRE's Unicode character property table had a bug
1180        which caused it to generate incorrect table entries for sequences of
1181        characters that have the same character type, but are in different scripts.
1182        It amalgamated them into a single range, with the script of the first of
1183        them. In other words, some characters were in the wrong script. There were
1184        thirteen such cases, affecting characters in the following ranges:
1185    
1186          U+002b0 - U+002c1
1187          U+0060c - U+0060d
1188          U+0061e - U+00612
1189          U+0064b - U+0065e
1190          U+0074d - U+0076d
1191          U+01800 - U+01805
1192          U+01d00 - U+01d77
1193          U+01d9b - U+01dbf
1194          U+0200b - U+0200f
1195          U+030fc - U+030fe
1196          U+03260 - U+0327f
1197          U+0fb46 - U+0fbb1
1198          U+10450 - U+1049d
1199    
1200    12. The -o option (show only the matching part of a line) for pcregrep was not
1201        compatible with GNU grep in that, if there was more than one match in a
1202        line, it showed only the first of them. It now behaves in the same way as
1203        GNU grep.
1204    
1205    13. If the -o and -v options were combined for pcregrep, it printed a blank
1206        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1207        does the same. The return code can be used to tell if there were any
1208        non-matching lines.
1209    
1210    14. Added --file-offsets and --line-offsets to pcregrep.
1211    
1212    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1213        infinitely looping recursion. The bug was that positive lookaheads were not
1214        being skipped when checking for a possible empty match (negative lookaheads
1215        and both kinds of lookbehind were skipped).
1216    
1217    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1218        inclusion of <windows.h> to before rather than after the definition of
1219        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1220    
1221    17. Specifying a possessive quantifier with a specific limit for a Unicode
1222        character property caused pcre_compile() to compile bad code, which led at
1223        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1224        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1225        caused the error; without that there was no problem.
1226    
1227    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1228    
1229    19. Added --enable-pcretest-libreadline.
1230    
1231    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1232        RE::GlobalReplace(). As a result, the number of replacements returned was
1233        double what it should be. I removed one of the increments, but Craig sent a
1234        later patch that removed the other one (the right fix) and added unit tests
1235        that check the return values (which was not done before).
1236    
1237    21. Several CMake things:
1238    
1239        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1240            the names libpcre and libpcreposix, not just pcre and pcreposix.
1241    
1242        (2) The above change means that pcretest and pcregrep are now correctly
1243            linked with the newly-built libraries, not previously installed ones.
1244    
1245        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1246    
1247    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1248        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1249        UTF-8 newline character). The key issue is that the pattern starts .*;
1250        this means that the match must be either at the beginning, or after a
1251        newline. The bug was in the code for advancing after a failed match and
1252        checking that the new position followed a newline. It was not taking
1253        account of UTF-8 characters correctly.
1254    
1255    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1256        character classes. PCRE was not treating the sequence [:...:] as a
1257        character class unless the ... were all letters. Perl, however, seems to
1258        allow any characters between [: and :], though of course it rejects as
1259        unknown any "names" that contain non-letters, because all the known class
1260        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1261        for example, whereas PCRE did not - it did not recognize a POSIX character
1262        class. This seemed a bit dangerous, so the code has been changed to be
1263        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1264        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1265        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1266        Perl does, and where it didn't before.
1267    
1268    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1269        Windows environments %n is disabled by default.
1270    
1271    
1272    Version 7.4 21-Sep-07
1273    ---------------------
1274    
1275    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1276        means that a class such as [\s] counted as "explicit reference to CR or
1277        LF". That isn't really right - the whole point of the change was to try to
1278        help when there was an actual mention of one of the two characters. So now
1279        the change happens only if \r or \n (or a literal CR or LF) character is
1280        encountered.
1281    
1282    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1283        of both had grown to the point where there were only 3 bits left.
1284        Fortunately, there was spare space in the data structure, and so I have
1285        moved the internal flags into a new 16-bit field to free up more option
1286        bits.
1287    
1288    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1289        but did not set the internal JCHANGED flag - either of these is enough to
1290        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1291        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1292        start sets both bits.
1293    
1294    4.  Added options (at build time, compile time, exec time) to change \R from
1295        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1296    
1297    5.  doc/pcresyntax.html was missing from the distribution.
1298    
1299    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1300        compatibility, even though it is no longer used.
1301    
1302    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1303        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1304        windows.h file is present (where different names are used). [This was
1305        reversed later after testing - see 16 below.]
1306    
1307    8.  Changed all #include <config.h> to #include "config.h". There were also
1308        some further <pcre.h> cases that I changed to "pcre.h".
1309    
1310    9.  When pcregrep was used with the --colour option, it missed the line ending
1311        sequence off the lines that it output.
1312    
1313    10. It was pointed out to me that arrays of string pointers cause lots of
1314        relocations when a shared library is dynamically loaded. A technique of
1315        using a single long string with a table of offsets can drastically reduce
1316        these. I have refactored PCRE in four places to do this. The result is
1317        dramatic:
1318    
1319          Originally:                          290
1320          After changing UCP table:            187
1321          After changing error message table:   43
1322          After changing table of "verbs"       36
1323          After changing table of Posix names   22
1324    
1325        Thanks to the folks working on Gregex for glib for this insight.
1326    
1327    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1328        unicode-properties was also set.
1329    
1330    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1331    
1332    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1333        checked only for CRLF.
1334    
1335    14. Added casts to pcretest.c to avoid compiler warnings.
1336    
1337    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1338    
1339    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1340        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1341        entirely. This removes changes made in 7 above.
1342    
1343    17. The CMake files have been updated, and there is now more information about
1344        building with CMake in the NON-UNIX-USE document.
1345    
1346    
1347    Version 7.3 28-Aug-07
1348    ---------------------
1349    
1350     1. In the rejigging of the build system that eventually resulted in 7.1, the
1351        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1352        brackets there is not right, since it causes compilers to look for an
1353        installed pcre.h, not the version that is in the source that is being
1354        compiled (which of course may be different). I have changed it back to:
1355    
1356          #include "pcre.h"
1357    
1358        I have a vague recollection that the change was concerned with compiling in
1359        different directories, but in the new build system, that is taken care of
1360        by the VPATH setting the Makefile.
1361    
1362     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1363        when the subject happened to end in the byte 0x85 (e.g. if the last
1364        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1365        characters but of course it shouldn't be taken as a newline when it is part
1366        of another character. The bug was that, for an unlimited repeat of . in
1367        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1368        characters when looking for a newline.
1369    
1370     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1371    
1372     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1373        in debug output.
1374    
1375     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1376        long printing in the pcrecpp unittest when running under MinGW.
1377    
1378     6. ESC_K was left out of the EBCDIC table.
1379    
1380     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1381        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1382        limit also applies to "virtual nesting" when a pattern is recursive, and in
1383        this case 1000 isn't so big. I have been able to remove this limit at the
1384        expense of backing off one optimization in certain circumstances. Normally,
1385        when pcre_exec() would call its internal match() function recursively and
1386        immediately return the result unconditionally, it uses a "tail recursion"
1387        feature to save stack. However, when a subpattern that can match an empty
1388        string has an unlimited repetition quantifier, it no longer makes this
1389        optimization. That gives it a stack frame in which to save the data for
1390        checking that an empty string has been matched. Previously this was taken
1391        from the 1000-entry workspace that had been reserved. So now there is no
1392        explicit limit, but more stack is used.
1393    
1394     8. Applied Daniel's patches to solve problems with the import/export magic
1395        syntax that is required for Windows, and which was going wrong for the
1396        pcreposix and pcrecpp parts of the library. These were overlooked when this
1397        problem was solved for the main library.
1398    
1399     9. There were some crude static tests to avoid integer overflow when computing
1400        the size of patterns that contain repeated groups with explicit upper
1401        limits. As the maximum quantifier is 65535, the maximum group length was
1402        set at 30,000 so that the product of these two numbers did not overflow a
1403        32-bit integer. However, it turns out that people want to use groups that
1404        are longer than 30,000 bytes (though not repeat them that many times).
1405        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1406        made it possible to implement the integer overflow checks in a much more
1407        dynamic way, which I have now done. The artificial limitation on group
1408        length has been removed - we now have only the limit on the total length of
1409        the compiled pattern, which depends on the LINK_SIZE setting.
1410    
1411    10. Fixed a bug in the documentation for get/copy named substring when
1412        duplicate names are permitted. If none of the named substrings are set, the
1413        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1414        empty string.
1415    
1416    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1417        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1418        because the ] is interpreted as the first data character and the
1419        terminating ] is not found. PCRE has been made compatible with Perl in this
1420        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1421        cause memory overwriting.
1422    
1423    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1424        string has been matched (to stop an infinite loop). It was not recognizing
1425        a conditional subpattern that could match an empty string if that
1426        subpattern was within another subpattern. For example, it looped when
1427        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1428        condition was not nested. This bug has been fixed.
1429    
1430    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1431        past the start of the subject in the presence of bytes with the top bit
1432        set, for example "\x8aBCD".
1433    
1434    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1435        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1436    
1437    14. Optimized (?!) to (*FAIL).
1438    
1439    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1440        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1441        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1442        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1443        does: it's just the validity check that is more restrictive.
1444    
1445    16. Inserted checks for integer overflows during escape sequence (backslash)
1446        processing, and also fixed erroneous offset values for syntax errors during
1447        backslash processing.
1448    
1449    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1450        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1451    
1452    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1453        caused an overrun.
1454    
1455    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1456        something other than just ASCII characters) inside a group that had an
1457        unlimited repeat caused a loop at compile time (while checking to see
1458        whether the group could match an empty string).
1459    
1460    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1461        [\P{Any}] did so. (Error in the code for printing property names.)
1462    
1463    21. An orphan \E inside a character class could cause a crash.
1464    
1465    22. A repeated capturing bracket such as (A)? could cause a wild memory
1466        reference during compilation.
1467    
1468    23. There are several functions in pcre_compile() that scan along a compiled
1469        expression for various reasons (e.g. to see if it's fixed length for look
1470        behind). There were bugs in these functions when a repeated \p or \P was
1471        present in the pattern. These operators have additional parameters compared
1472        with \d, etc, and these were not being taken into account when moving along
1473        the compiled data. Specifically:
1474    
1475        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1476            length.
1477    
1478        (b) An item such as \pL+ within a repeated group could cause crashes or
1479            loops.
1480    
1481        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1482            "reference to non-existent subpattern" error.
1483    
1484        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1485    
1486    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1487        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1488    
1489    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1490    
1491    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1492        character were causing crashes (broken optimization).
1493    
1494    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1495        \p or \P) caused a compile-time loop.
1496    
1497    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1498        break. For example, the unstudied pattern [\r\n]A does not match the string
1499        "\r\nA" because change 7.0/46 below moves the current point on by two
1500        characters after failing to match at the start. However, the pattern \nA
1501        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1502        the same is true. There doesn't seem any very clean way out of this, but
1503        what I have chosen to do makes the common cases work: PCRE now takes note
1504        of whether there can be an explicit match for \r or \n anywhere in the
1505        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1506        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1507        pattern has explicit CR or LF references.
1508    
1509    29. Added (*CR) etc for changing newline setting at start of pattern.
1510    
1511    
1512    Version 7.2 19-Jun-07
1513    ---------------------
1514    
1515     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1516        which is apparently normally available under Windows.
1517    
1518     2. Re-jig the pcregrep tests with different newline settings in an attempt
1519        to make them independent of the local environment's newline setting.
1520    
1521     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1522    
1523     4. Some of the "internals" tests were previously cut out when the link size
1524        was not 2, because the output contained actual offsets. The recent new
1525        "Z" feature of pcretest means that these can be cut out, making the tests
1526        usable with all link sizes.
1527    
1528     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1529        stack recursion. This gives a massive performance boost under BSD, but just
1530        a small improvement under Linux. However, it saves one field in the frame
1531        in all cases.
1532    
1533     6. Added more features from the forthcoming Perl 5.10:
1534    
1535        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1536            recursion call. It refers to the nth most recently opened parentheses.
1537    
1538        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1539            to be opened parentheses.
1540    
1541        (c) Conditions that refer to capturing parentheses can be specified
1542            relatively, for example, (?(-2)... or (?(+3)...
1543    
1544        (d) \K resets the start of the current match so that everything before
1545            is not part of it.
1546    
1547        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1548    
1549        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1550            reference syntax.
1551    
1552        (g) (?| introduces a group in which the numbering of parentheses in each
1553            alternative starts with the same number.
1554    
1555        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1556    
1557     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1558        PCRE_INFO_JCHANGED.
1559    
1560     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1561        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1562        for detecting groups that can match an empty string.
1563    
1564     9. A pattern with a very large number of alternatives (more than several
1565        hundred) was running out of internal workspace during the pre-compile
1566        phase, where pcre_compile() figures out how much memory will be needed. A
1567        bit of new cunning has reduced the workspace needed for groups with
1568        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1569        workspace instead of running out of the 4096 that are available.
1570    
1571    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1572    
1573    11. Applied patch from Google to remove an optimization that didn't quite work.
1574        The report of the bug said:
1575    
1576          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1577          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1578          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1579    
1580    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1581        it matched the wrong number of bytes.
1582    
1583    
1584    Version 7.1 24-Apr-07
1585    ---------------------
1586    
1587     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1588        that is more "standard", making use of automake and other Autotools. There
1589        is some re-arrangement of the files and adjustment of comments consequent
1590        on this.
1591    
1592     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1593        for recursive directory scanning broke on some systems because the files
1594        are not scanned in any specific order and on different systems the order
1595        was different. A call to "sort" has been inserted into RunGrepTest for the
1596        approprate test as a short-term fix. In the longer term there may be an
1597        alternative.
1598    
1599     3. I had an email from Eric Raymond about problems translating some of PCRE's
1600        man pages to HTML (despite the fact that I distribute HTML pages, some
1601        people do their own conversions for various reasons). The problems
1602        concerned the use of low-level troff macros .br and .in. I have therefore
1603        removed all such uses from the man pages (some were redundant, some could
1604        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1605        HTML has been updated to handle .nf/.fi and to complain if it encounters
1606        .br or .in.
1607    
1608     4. Updated comments in configure.ac that get placed in config.h.in and also
1609        arranged for config.h to be included in the distribution, with the name
1610        config.h.generic, for the benefit of those who have to compile without
1611        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1612    
1613     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1614        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1615        makevp.bat for the current PCRE, using the additional files
1616        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1617    
1618     6. A Windows user reported a minor discrepancy with test 2, which turned out
1619        to be caused by a trailing space on an input line that had got lost in his
1620        copy. The trailing space was an accident, so I've just removed it.
1621    
1622     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1623        that is needed.
1624    
1625     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1626        as "const" (a) because they are and (b) because it helps the PHP
1627        maintainers who have recently made a script to detect big data structures
1628        in the php code that should be moved to the .rodata section. I remembered
1629        to update Builducptable as well, so it won't revert if ucptable.h is ever
1630        re-created.
1631    
1632     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1633        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1634        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1635        support is not required. This saves 1.5-2K of code, which is important in
1636        some applications.
1637    
1638        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1639        so as not to refer to the tables, even though these functions will never be
1640        called when UTF-8 support is disabled. Otherwise there are problems with a
1641        shared library.
1642    
1643    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1644    
1645        (a) It was defining its arguments as char * instead of void *.
1646    
1647        (b) It was assuming that all moves were upwards in memory; this was true
1648            a long time ago when I wrote it, but is no longer the case.
1649    
1650        The emulated memove() is provided for those environments that have neither
1651        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1652        is clearly not the case, as these two bugs were recently reported.
1653    
1654    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1655        and Detrail to create the HTML documentation, the .txt form of the man
1656        pages, and it removes trailing spaces from listed files. It also creates
1657        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1658        case, it wraps all the #defines with #ifndefs. This script should be run
1659        before "make dist".
1660    
1661    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1662        with Unicode property support.
1663    
1664        (a) For a maximizing quantifier, if the two different cases of the
1665            character were of different lengths in their UTF-8 codings (there are
1666            some cases like this - I found 11), and the matching function had to
1667            back up over a mixture of the two cases, it incorrectly assumed they
1668            were both the same length.
1669    
1670        (b) When PCRE was configured to use the heap rather than the stack for
1671            recursion during matching, it was not correctly preserving the data for
1672            the other case of a UTF-8 character when checking ahead for a match
1673            while processing a minimizing repeat. If the check also involved
1674            matching a wide character, but failed, corruption could cause an
1675            erroneous result when trying to check for a repeat of the original
1676            character.
1677    
1678    13. Some tidying changes to the testing mechanism:
1679    
1680        (a) The RunTest script now detects the internal link size and whether there
1681            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1682            values substituted by "configure". (The RunGrepTest script already did
1683            this for UTF-8.) The configure.ac script no longer substitutes the
1684            relevant variables.
1685    
1686        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1687            with length and offset values. This means that the output is different
1688            for different internal link sizes. Test 2 is skipped for link sizes
1689            other than 2 because of this, bypassing the problem. Unfortunately,
1690            there was also a test in test 3 (the locale tests) that used /B and
1691            failed for link sizes other than 2. Rather than cut the whole test out,
1692            I have added a new /Z option to pcretest that replaces the length and
1693            offset values with spaces. This is now used to make test 3 independent
1694            of link size. (Test 2 will be tidied up later.)
1695    
1696    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1697        segmentation fault instead of returning the appropriate error message.
1698    
1699    15. In multiline mode when the newline sequence was set to "any", the pattern
1700        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1701        This doesn't seem right; it now treats the CRLF combination as the line
1702        ending, and so does not match in that case. It's only a pattern such as ^$
1703        that would hit this one: something like ^ABC$ would have failed after \r
1704        and then tried again after \r\n.
1705    
1706    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1707        in an attempt to make files that differ only in their line terminators
1708        compare equal. This works on Linux.
1709    
1710    17. Under certain error circumstances pcregrep might try to free random memory
1711        as it exited. This is now fixed, thanks to valgrind.
1712    
1713    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1714        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1715        was because its rules for how to advance for /g after matching an empty
1716        string at the end of a line did not allow for this case. They now check for
1717        it specially.
1718    
1719    20. pcretest is supposed to handle patterns and data of any length, by
1720        extending its buffers when necessary. It was getting this wrong when the
1721        buffer for a data line had to be extended.
1722    
1723    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1724        CRLF as a newline sequence.
1725    
1726    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1727        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1728        I have nevertheless tidied it up.
1729    
1730    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1731    
1732    24. Added a man page for pcre-config.
1733    
1734    
1735    Version 7.0 19-Dec-06
1736    ---------------------
1737    
1738     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1739        moving to gcc 4.1.1.
1740    
1741     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1742        sys/time.h, which is documented as needed for this function. It doesn't
1743        seem to matter on Linux, but it showed up on some releases of OS X.
1744    
1745     3. It seems that there are systems where bytes whose values are greater than
1746        127 match isprint() in the "C" locale. The "C" locale should be the
1747        default when a C program starts up. In most systems, only ASCII printing
1748        characters match isprint(). This difference caused the output from pcretest
1749        to vary, making some of the tests fail. I have changed pcretest so that:
1750    
1751        (a) When it is outputting text in the compiled version of a pattern, bytes
1752            other than 32-126 are always shown as hex escapes.
1753    
1754        (b) When it is outputting text that is a matched part of a subject string,
1755            it does the same, unless a different locale has been set for the match
1756            (using the /L modifier). In this case, it uses isprint() to decide.
1757    
1758     4. Fixed a major bug that caused incorrect computation of the amount of memory
1759        required for a compiled pattern when options that changed within the
1760        pattern affected the logic of the preliminary scan that determines the
1761        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1762        that the computed length was too small. The symptoms of this bug were
1763        either the PCRE error "internal error: code overflow" from pcre_compile(),
1764        or a glibc crash with a message such as "pcretest: free(): invalid next
1765        size (fast)". Examples of patterns that provoked this bug (shown in
1766        pcretest format) are:
1767    
1768          /(?-x: )/x
1769          /(?x)(?-x: \s*#\s*)/
1770          /((?i)[\x{c0}])/8
1771          /(?i:[\x{c0}])/8
1772    
1773        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1774        is now done differently.
1775    
1776     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1777        wrapper classes; (b) implement a new function in the C++ scanner that is
1778        more efficient than the old way of doing things because it avoids levels of
1779        recursion in the regex matching; (c) add a paragraph to the documentation
1780        for the FullMatch() function.
1781    
1782     6. The escape sequence \n was being treated as whatever was defined as
1783        "newline". Not only was this contrary to the documentation, which states
1784        that \n is character 10 (hex 0A), but it also went horribly wrong when
1785        "newline" was defined as CRLF. This has been fixed.
1786    
1787     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1788        was being set to -1 for the "end of line" case (supposedly a value that no
1789        character can have). Though this value is never used (the check for end of
1790        line is "zero bytes in current character"), it caused compiler complaints.
1791        I've changed it to 0xffffffff.
1792    
1793     8. In pcre_version.c, the version string was being built by a sequence of
1794        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1795        string (as it is for production releases) called a macro with an empty
1796        argument. The C standard says the result of this is undefined. The gcc
1797        compiler treats it as an empty string (which was what was wanted) but it is
1798        reported that Visual C gives an error. The source has been hacked around to
1799        avoid this problem.
1800    
1801     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1802        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1803        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1804        of them did).
1805    
1806    10. Originally, pcretest opened its input and output without "b"; then I was
1807        told that "b" was needed in some environments, so it was added for release
1808        5.0 to both the input and output. (It makes no difference on Unix-like
1809        systems.) Later I was told that it is wrong for the input on Windows. I've
1810        now abstracted the modes into two macros, to make it easier to fiddle with
1811        them, and removed "b" from the input mode under Windows.
1812    
1813    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1814    
1815    12. Added -help and --help to pcretest as an official way of being reminded
1816        of the options.
1817    
1818    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1819        and pcrecpp.cc because they annoy compilers at high warning levels.
1820    
1821    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1822    
1823    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1824        scripts are not C programs :-) and which was not noticed because it works
1825        on Linux.
1826    
1827    16. pcretest is supposed to handle any length of pattern and data line (as one
1828        line or as a continued sequence of lines) by extending its input buffer if
1829        necessary. This feature was broken for very long pattern lines, leading to
1830        a string of junk being passed to pcre_compile() if the pattern was longer
1831        than about 50K.
1832    
1833    17. I have done a major re-factoring of the way pcre_compile() computes the
1834        amount of memory needed for a compiled pattern. Previously, there was code
1835        that made a preliminary scan of the pattern in order to do this. That was
1836        OK when PCRE was new, but as the facilities have expanded, it has become
1837        harder and harder to keep it in step with the real compile phase, and there
1838        have been a number of bugs (see for example, 4 above). I have now found a
1839        cunning way of running the real compile function in a "fake" mode that
1840        enables it to compute how much memory it would need, while actually only
1841        ever using a few hundred bytes of working memory and without too many
1842        tests of the mode. This should make future maintenance and development
1843        easier. A side effect of this work is that the limit of 200 on the nesting
1844        depth of parentheses has been removed (though this was never a serious
1845        limitation, I suspect). However, there is a downside: pcre_compile() now
1846        runs more slowly than before (30% or more, depending on the pattern). I
1847        hope this isn't a big issue. There is no effect on runtime performance.
1848    
1849    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1850        newline (only possible for the last line of a file) and it was a
1851        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1852    
1853    19. Added additional timing features to pcretest. (1) The -tm option now times
1854        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1855        separate command line item, by a number that specifies the number of
1856        repeats to use when timing. The default is 50000; this gives better
1857        precision, but takes uncomfortably long for very large patterns.
1858    
1859    20. Extended pcre_study() to be more clever in cases where a branch of a
1860        subpattern has no definite first character. For example, (a*|b*)[cd] would
1861        previously give no result from pcre_study(). Now it recognizes that the
1862        first character must be a, b, c, or d.
1863    
1864    21. There was an incorrect error "recursive call could loop indefinitely" if
1865        a subpattern (or the entire pattern) that was being tested for matching an
1866        empty string contained only one non-empty item after a nested subpattern.
1867        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1868        incorrectly, because the \d was being skipped in the check.
1869    
1870    22. The pcretest program now has a new pattern option /B and a command line
1871        option -b, which is equivalent to adding /B to every pattern. This causes
1872        it to show the compiled bytecode, without the additional information that
1873        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1874        is the same as /B/I).
1875    
1876    23. A new optimization is now able automatically to treat some sequences such
1877        as a*b as a*+b. More specifically, if something simple (such as a character
1878        or a simple class like \d) has an unlimited quantifier, and is followed by
1879        something that cannot possibly match the quantified thing, the quantifier
1880        is automatically "possessified".
1881    
1882    24. A recursive reference to a subpattern whose number was greater than 39
1883        went wrong under certain circumstances in UTF-8 mode. This bug could also
1884        have affected the operation of pcre_study().
1885    
1886    25. Realized that a little bit of performance could be had by replacing
1887        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1888    
1889    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1890    
1891    27. Possessive quantifiers such as a++ were previously implemented by turning
1892        them into atomic groups such as ($>a+). Now they have their own opcodes,
1893        which improves performance. This includes the automatically created ones
1894        from 23 above.
1895    
1896    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1897        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1898        the first matched character to be a colon. This applied both to named and
1899        numbered groups.
1900    
1901    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1902    
1903    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1904        building PCRE on Windows easier, so I have included it in the distribution.
1905    
1906    31. There is now a check in pcretest against a ridiculously large number being
1907        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1908        loop, the loop is abandoned.
1909    
1910    32. Forward references to subpatterns in conditions such as (?(2)...) where
1911        subpattern 2 is defined later cause pcre_compile() to search forwards in
1912        the pattern for the relevant set of parentheses. This search went wrong
1913        when there were unescaped parentheses in a character class, parentheses
1914        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1915    
1916    33. "Subroutine" calls and backreferences were previously restricted to
1917        referencing subpatterns earlier in the regex. This restriction has now
1918        been removed.
1919    
1920    34. Added a number of extra features that are going to be in Perl 5.10. On the
1921        whole, these are just syntactic alternatives for features that PCRE had
1922        previously implemented using the Python syntax or my own invention. The
1923        other formats are all retained for compatibility.
1924    
1925        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1926            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1927            also .NET compatible.
1928    
1929        (b) A recursion or subroutine call to a named group can now be defined as
1930            (?&name) as well as (?P>name).
1931    
1932        (c) A backreference to a named group can now be defined as \k<name> or
1933            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1934            5.10, are also .NET compatible.
1935    
1936        (d) A conditional reference to a named group can now use the syntax
1937            (?(<name>) or (?('name') as well as (?(name).
1938    
1939        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1940            groups (named and numbered) that are never evaluated inline, but can be
1941            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1942            is always false. There may be only one alternative in such a group.
1943    
1944        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1945            as the simple (?(R). The condition is true only if the most recent
1946            recursion is that of the given number or name. It does not search out
1947            through the entire recursion stack.
1948    
1949        (g) The escape \gN or \g{N} has been added, where N is a positive or
1950            negative number, specifying an absolute or relative reference.
1951    
1952    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1953        some "unreachable code" warnings.
1954    
1955    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1956        things, this adds five new scripts.
1957    
1958    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1959        There were also incompatibilities regarding the handling of \Q..\E inside
1960        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1961        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1962    
1963    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1964        matches an empty string, and forcibly breaks the loop. There were bugs in
1965        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1966        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1967        separate and independent bugs (that affected different cases) have been
1968        fixed.
1969    
1970    39. Refactored the code to abolish the use of different opcodes for small
1971        capturing bracket numbers. This is a tidy that I avoided doing when I
1972        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1973        The new approach is not only tidier, it makes it possible to reduce the
1974        memory needed to fix the previous bug (38).
1975    
1976    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1977        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1978        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1979        mode.
1980    
1981    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1982        report.
1983    
1984    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1985        copy construction and assignment in the C++ wrapper.
1986    
1987    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1988        couple of bugs that could have given wrong results in the "--newline=crlf"
1989        case.
1990    
1991    44. Added a number of casts and did some reorganization of signed/unsigned int
1992        variables following suggestions from Dair Grant. Also renamed the variable
1993        "this" as "item" because it is a C++ keyword.
1994    
1995    45. Arranged for dftables to add
1996    
1997          #include "pcre_internal.h"
1998    
1999        to pcre_chartables.c because without it, gcc 4.x may remove the array
2000        definition from the final binary if PCRE is built into a static library and
2001        dead code stripping is activated.
2002    
2003    46. For an unanchored pattern, if a match attempt fails at the start of a
2004        newline sequence, and the newline setting is CRLF or ANY, and the next two
2005        characters are CRLF, advance by two characters instead of one.
2006    
2007    
2008    Version 6.7 04-Jul-06
2009    ---------------------
2010    
2011     1. In order to handle tests when input lines are enormously long, pcretest has
2012        been re-factored so that it automatically extends its buffers when
2013        necessary. The code is crude, but this _is_ just a test program. The
2014        default size has been increased from 32K to 50K.
2015    
2016     2. The code in pcre_study() was using the value of the re argument before
2017        testing it for NULL. (Of course, in any sensible call of the function, it
2018        won't be NULL.)
2019    
2020     3. The memmove() emulation function in pcre_internal.h, which is used on
2021        systems that lack both memmove() and bcopy() - that is, hardly ever -
2022        was missing a "static" storage class specifier.
2023    
2024     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2025        containing an extended class (one that cannot be represented by a bitmap
2026        because it contains high-valued characters or Unicode property items, e.g.
2027        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2028        pattern, but PCRE should not loop if you do not (it no longer does).
2029        [Detail: two cases were found: (a) a repeated subpattern containing an
2030        extended class; (b) a recursive reference to a subpattern that followed a
2031        previous extended class. It wasn't skipping over the extended class
2032        correctly when UTF-8 mode was not set.]
2033    
2034     5. A negated single-character class was not being recognized as fixed-length
2035        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2036        compile error "lookbehind assertion is not fixed length".
2037    
2038     6. The RunPerlTest auxiliary script was showing an unexpected difference
2039        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2040        write a Perl script that can interpret lines of an input file either as
2041        byte characters or as UTF-8, which is what "perltest" was being required to
2042        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2043        can't do is switch easily at run time between having the "use utf8;" pragma
2044        or not. In the end, I fudged it by using the RunPerlTest script to insert
2045        "use utf8;" explicitly for the UTF-8 tests.
2046    
2047     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2048        the end of the subject string, contrary to the documentation and to what
2049        Perl does. This was true of both matching functions. Now it matches only at
2050        the start of the subject and immediately after *internal* newlines.
2051    
2052     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2053        a pointer to an int instead of a pointer to an unsigned long int. This
2054        caused problems on 64-bit systems.
2055    
2056     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2057        instance of the 'standard' template library not being so standard".
2058    
2059    10. There was no check on the number of named subpatterns nor the maximum
2060        length of a subpattern name. The product of these values is used to compute
2061        the size of the memory block for a compiled pattern. By supplying a very
2062        long subpattern name and a large number of named subpatterns, the size
2063        computation could be caused to overflow. This is now prevented by limiting
2064        the length of names to 32 characters, and the number of named subpatterns
2065        to 10,000.
2066    
2067    11. Subpatterns that are repeated with specific counts have to be replicated in
2068        the compiled pattern. The size of memory for this was computed from the
2069        length of the subpattern and the repeat count. The latter is limited to
2070        65535, but there was no limit on the former, meaning that integer overflow
2071        could in principle occur. The compiled length of a repeated subpattern is
2072        now limited to 30,000 bytes in order to prevent this.
2073    
2074    12. Added the optional facility to have named substrings with the same name.
2075    
2076    13. Added the ability to use a named substring as a condition, using the
2077        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2078        are numbers (not recommended). Forward references are permitted.
2079    
2080    14. Added forward references in named backreferences (if you see what I mean).
2081    
2082    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2083        pattern could run off the end of the subject. For example, the pattern
2084        "(?s)(.{1,5})"8 did this with the subject "ab".
2085    
2086    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2087        PCRE_CASELESS was set when matching characters that were quantified with ?
2088        or *.
2089    
2090    17. A character class other than a single negated character that had a minimum
2091        but no maximum quantifier - for example [ab]{6,} - was not handled
2092        correctly by pce_dfa_exec(). It would match only one character.
2093    
2094    18. A valid (though odd) pattern that looked like a POSIX character
2095        class but used an invalid character after [ (for example [[,abc,]]) caused
2096        pcre_compile() to give the error "Failed: internal error: code overflow" or
2097        in some cases to crash with a glibc free() error. This could even happen if
2098        the pattern terminated after [[ but there just happened to be a sequence of
2099        letters, a binary zero, and a closing ] in the memory that followed.
2100    
2101    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2102        over the years. Originally (before any Unicode support), just the bottom 8
2103        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2104        output from "man perlunicode" includes this:
2105    
2106          The regular expression compiler produces polymorphic opcodes.  That
2107          is, the pattern adapts to the data and automatically switches to
2108          the Unicode character scheme when presented with Unicode data--or
2109          instead uses a traditional byte scheme when presented with byte
2110          data.
2111    
2112        Sadly, a wide octal escape does not cause a switch, and in a string with
2113        no other multibyte characters, these octal escapes are treated as before.
2114        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2115        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2116        Unicode string.
2117    
2118        I have not perpetrated such confusion in PCRE. Up till now, it took just
2119        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2120        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2121        translate to the appropriate multibyte character.
2122    
2123    29. Applied some refactoring to reduce the number of warnings from Microsoft
2124        and Borland compilers. This has included removing the fudge introduced
2125        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2126        a warning about an unused variable.
2127    
2128    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2129        characters since release 4.0, because Perl (from release 5.004) does not.
2130        [Or at least, is documented not to: some releases seem to be in conflict
2131        with the documentation.] However, when a pattern was studied with
2132        pcre_study() and all its branches started with \s, PCRE still included VT
2133        as a possible starting character. Of course, this did no harm; it just
2134        caused an unnecessary match attempt.
2135    
2136    22. Removed a now-redundant internal flag bit that recorded the fact that case
2137        dependency changed within the pattern. This was once needed for "required
2138        byte" processing, but is no longer used. This recovers a now-scarce options
2139        bit. Also moved the least significant internal flag bit to the most-
2140        significant bit of the word, which was not previously used (hangover from
2141        the days when it was an int rather than a uint) to free up another bit for
2142        the future.
2143    
2144    23. Added support for CRLF line endings as well as CR and LF. As well as the
2145        default being selectable at build time, it can now be changed at runtime
2146        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2147        specify that it is scanning data with non-default line endings.
2148    
2149    24. Changed the definition of CXXLINK to make it agree with the definition of
2150        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2151    
2152    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2153        recursions. This makes a big different to stack usage for some patterns.
2154    
2155    26. If a subpattern containing a named recursion or subroutine reference such
2156        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2157        the space required for the compiled pattern went wrong and gave too small a
2158        value. Depending on the environment, this could lead to "Failed: internal
2159        error: code overflow at offset 49" or "glibc detected double free or
2160        corruption" errors.
2161    
2162    27. Applied patches from Google (a) to support the new newline modes and (b) to
2163        advance over multibyte UTF-8 characters in GlobalReplace.
2164    
2165    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2166        difference for some implementation of PCRE in some Windows version.
2167    
2168    29. Added some extra testing facilities to pcretest:
2169    
2170        \q<number>   in a data line sets the "match limit" value
2171        \Q<number>   in a data line sets the "match recursion limt" value
2172        -S <number>  sets the stack size, where <number> is in megabytes
2173    
2174        The -S option isn't available for Windows.
2175    
2176    
2177    Version 6.6 06-Feb-06
2178    ---------------------
2179    
2180     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2181        in pcreposix.h. I have copied the definition from pcre.h.
2182    
2183     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2184        because pcre.h is no longer a built file.
2185    
2186     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2187        not normally included in the compiled code.
2188    
2189    
2190    Version 6.5 01-Feb-06
2191    ---------------------
2192    
2193     1. When using the partial match feature with pcre_dfa_exec(), it was not
2194        anchoring the second and subsequent partial matches at the new starting
2195        point. This could lead to incorrect results. For example, with the pattern
2196        /1234/, partially matching against "123" and then "a4" gave a match.
2197    
2198     2. Changes to pcregrep:
2199    
2200        (a) All non-match returns from pcre_exec() were being treated as failures
2201            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2202            error message is output. Some extra information is given for the
2203            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2204            probably the only errors that are likely to be caused by users (by
2205            specifying a regex that has nested indefinite repeats, for instance).
2206            If there are more than 20 of these errors, pcregrep is abandoned.
2207    
2208        (b) A binary zero was treated as data while matching, but terminated the
2209            output line if it was written out. This has been fixed: binary zeroes
2210            are now no different to any other data bytes.
2211    
2212        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2213            used to set a locale for matching. The --locale=xxxx long option has
2214            been added (no short equivalent) to specify a locale explicitly on the
2215            pcregrep command, overriding the environment variables.
2216    
2217        (d) When -B was used with -n, some line numbers in the output were one less
2218            than they should have been.
2219    
2220        (e) Added the -o (--only-matching) option.
2221    
2222        (f) If -A or -C was used with -c (count only), some lines of context were
2223            accidentally printed for the final match.
2224    
2225        (g) Added the -H (--with-filename) option.
2226    
2227        (h) The combination of options -rh failed to suppress file names for files
2228            that were found from directory arguments.
2229    
2230        (i) Added the -D (--devices) and -d (--directories) options.
2231    
2232        (j) Added the -F (--fixed-strings) option.
2233    
2234        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2235    
2236        (l) Added the --colo(u)r option.
2237    
2238        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2239            is not present by default.
2240    
2241     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2242        items such as (?R) or (?1), when the recursion could match a number of
2243        alternatives. If it matched one of the alternatives, but subsequently,
2244        outside the recursion, there was a failure, the code tried to back up into
2245        the recursion. However, because of the way PCRE is implemented, this is not
2246        possible, and the result was an incorrect result from the match.
2247    
2248        In order to prevent this happening, the specification of recursion has
2249        been changed so that all such subpatterns are automatically treated as
2250        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2251    
2252     4. I had overlooked the fact that, in some locales, there are characters for
2253        which isalpha() is true but neither isupper() nor islower() are true. In
2254        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2255        and ordfeminine) are like this. This affected the treatment of \w and \W
2256        when they appeared in character classes, but not when they appeared outside
2257        a character class. The bit map for "word" characters is now created
2258        separately from the results of isalnum() instead of just taking it from the
2259        upper, lower, and digit maps. (Plus the underscore character, of course.)
2260    
2261     5. The above bug also affected the handling of POSIX character classes such as
2262        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2263        permanent tables. Instead, the bit maps for such a class were previously
2264        created as the appropriate unions of the upper, lower, and digit bitmaps.
2265        Now they are created by subtraction from the [[:word:]] class, which has
2266        its own bitmap.
2267    
2268     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2269        It is created by subtracting the vertical space characters (\x09, \x0a,
2270        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2271        subtraction was done in the overall bitmap for a character class, meaning
2272        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2273        be recognized. This bug has been fixed.
2274    
2275     7. Patches from the folks at Google:
2276    
2277          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2278          real life, but is still worth protecting against".
2279    
2280          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2281          regular expressions".
2282    
2283          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2284          have it.
2285    
2286          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2287          "configure" and the latter not, in order to fix a problem somebody had
2288          with compiling the Arg class on HP-UX.
2289    
2290          (e) Improve the error-handling of the C++ wrapper a little bit.
2291    
2292          (f) New tests for checking recursion limiting.
2293    
2294     8. The pcre_memmove() function, which is used only if the environment does not
2295        have a standard memmove() function (and is therefore rarely compiled),
2296        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2297        returning a result (though PCRE never actually uses the result).
2298    
2299     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2300        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2301        returned instead of calling malloc() with an overflowing number that would
2302        most likely cause subsequent chaos.
2303    
2304    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2305    
2306    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2307        with this option is matched, the nmatch and pmatch options of regexec() are
2308        ignored.
2309    
2310    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2311        provided in case anyone wants to the the POSIX interface with UTF-8
2312        strings.
2313    
2314    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2315        C++ linking (needed for some HP-UX environments).
2316    
2317    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2318        (unused parameter) and in the pcre_printint() function (omitted "default"
2319        switch label when the default is to do nothing).
2320    
2321    15. Added some code to make it possible, when PCRE is compiled as a C++
2322        library, to replace subject pointers for pcre_exec() with a smart pointer
2323        class, thus making it possible to process discontinuous strings.
2324    
2325    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2326        much the same function. They were added by different people who were trying
2327        to make PCRE easy to compile on non-Unix systems. It has been suggested
2328        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2329        for compiling on Windows systems. I have therefore replaced it with
2330        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2331        defaults to "extern" for C or "extern C" for C++, which works fine on
2332        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2333        SCOPE with something explicit in config.h. In addition:
2334    
2335        (a) pcreposix.h still had just "extern" instead of either of these macros;
2336            I have replaced it with PCRE_DATA_SCOPE.
2337    
2338        (b) Functions such as _pcre_xclass(), which are internal to the library,
2339            but external in the C sense, all had PCRE_EXPORT in their definitions.
2340            This is apparently wrong for the Windows case, so I have removed it.
2341            (It makes no difference on Unix-like systems.)
2342    
2343    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2344        of recursive calls to match(). This is different to MATCH_LIMIT because
2345        that limits the total number of calls to match(), not all of which increase
2346        the depth of recursion. Limiting the recursion depth limits the amount of
2347        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2348        when PCRE is compiled, and changed at run time. A patch from Google adds
2349        this functionality to the C++ interface.
2350    
2351    18. Changes to the handling of Unicode character properties:
2352    
2353        (a) Updated the table to Unicode 4.1.0.
2354    
2355        (b) Recognize characters that are not in the table as "Cn" (undefined).
2356    
2357        (c) I revised the way the table is implemented to a much improved format
2358            which includes recognition of ranges. It now supports the ranges that
2359            are defined in UnicodeData.txt, and it also amalgamates other
2360            characters into ranges. This has reduced the number of entries in the
2361            table from around 16,000 to around 3,000, thus reducing its size
2362            considerably. I realized I did not need to use a tree structure after
2363            all - a binary chop search is just as efficient. Having reduced the
2364            number of entries, I extended their size from 6 bytes to 8 bytes to
2365            allow for more data.
2366    
2367        (d) Added support for Unicode script names via properties such as \p{Han}.
2368    
2369    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2370        matching that character.
2371    
2372    20. When matching a repeated Unicode property with a minimum greater than zero,
2373        (for example \pL{2,}), PCRE could look past the end of the subject if it
2374        reached it while seeking the minimum number of characters. This could
2375        happen only if some of the characters were more than one byte long, because
2376        there is a check for at least the minimum number of bytes.
2377    
2378    21. Refactored the implementation of \p and \P so as to be more general, to
2379        allow for more different types of property in future. This has changed the
2380        compiled form incompatibly. Anybody with saved compiled patterns that use
2381        \p or \P will have to recompile them.
2382    
2383    22. Added "Any" and "L&" to the supported property types.
2384    
2385    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2386        but give a compile time error if the value is greater than 0xff.
2387    
2388    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2389        accidentally not being installed or uninstalled.
2390    
2391    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2392        made were to insert the current release number. This seemed silly, because
2393        it made things harder for people building PCRE on systems that don't run
2394        "configure". I have turned pcre.h into a distributed file, no longer built
2395        by "configure", with the version identification directly included. There is
2396        no longer a pcre.h.in file.
2397    
2398        However, this change necessitated a change to the pcre-config script as
2399        well. It is built from pcre-config.in, and one of the substitutions was the
2400        release number. I have updated configure.ac so that ./configure now finds
2401        the release number by grepping pcre.h.
2402    
2403    26. Added the ability to run the tests under valgrind.
2404    
2405    
2406    Version 6.4 05-Sep-05
2407    ---------------------
2408    
2409     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2410        "--" to be printed when multiple files were scanned, even when none of the
2411        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2412        consider it to be a bug, and have restored the previous behaviour.
2413    
2414     2. A couple of code tidies to get rid of compiler warnings.
2415    
2416     3. The pcretest program used to cheat by referring to symbols in the library
2417        whose names begin with _pcre_. These are internal symbols that are not
2418        really supposed to be visible externally, and in some environments it is
2419        possible to suppress them. The cheating is now confined to including
2420        certain files from the library's source, which is a bit cleaner.
2421    
2422     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2423        file's purpose clearer.
2424    
2425     5. Reorganized pcre_ucp_findchar().
2426    
2427    
2428    Version 6.3 15-Aug-05
2429    ---------------------
2430    
2431     1. The file libpcre.pc.in did not have general read permission in the tarball.
2432    
2433     2. There were some problems when building without C++ support:
2434    
2435        (a) If C++ support was not built, "make install" and "make test" still
2436            tried to test it.
2437    
2438        (b) There were problems when the value of CXX was explicitly set. Some
2439            changes have been made to try to fix these, and ...
2440    
2441        (c) --disable-cpp can now be used to explicitly disable C++ support.
2442    
2443        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2444            backslash in a target when C++ was disabled. This confuses some
2445            versions of "make", apparently. Using an intermediate variable solves
2446            this. (Same for CPP_LOBJ.)
2447    
2448     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2449        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2450        necessary on certain architectures.
2451    
2452     4. Added a setting of -export-symbols-regex to the link command to remove
2453        those symbols that are exported in the C sense, but actually are local
2454        within the library, and not documented. Their names all begin with
2455        "_pcre_". This is not a perfect job, because (a) we have to except some
2456        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2457        available (and never for static libraries). I have made a note to try to
2458        find a way round (a) in the future.
2459    
2460    
2461    Version 6.2 01-Aug-05
2462    ---------------------
2463    
2464     1. There was no test for integer overflow of quantifier values. A construction
2465        such as {1111111111111111} would give undefined results. What is worse, if
2466        a minimum quantifier for a parenthesized subpattern overflowed and became
2467        negative, the calculation of the memory size went wrong. This could have
2468        led to memory overwriting.
2469    
2470     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2471    
2472     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2473        operating environments where this matters.
2474    
2475     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2476        PCRE options from within the C++ wrapper.
2477    
2478     5. Named capturing subpatterns were not being correctly counted when a pattern
2479        was compiled. This caused two problems: (a) If there were more than 100
2480        such subpatterns, the calculation of the memory needed for the whole
2481        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2482        back references of the form \12, where the number was greater than 9, were
2483        not recognized as back references, even though there were sufficient
2484        previous subpatterns.
2485    
2486     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2487        versions of gcc, e.g. 2.95.4.
2488    
2489    
2490    Version 6.1 21-Jun-05
2491    ---------------------
2492    
2493     1. There was one reference to the variable "posix" in pcretest.c that was not
2494        surrounded by "#if !defined NOPOSIX".
2495    
2496     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2497        the cross-check on the old pcre_info() function, for the benefit of the
2498        cut-down version of PCRE that is currently imported into Exim.
2499    
2500     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2501        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2502        patterns that start (?i) but I don't think that matters. The use of (?i) is
2503        just an example; this all applies to the other options as well.
2504    
2505     4. Since libtool seems to echo the compile commands it is issuing, the output
2506        from "make" can be reduced a bit by putting "@" in front of each libtool
2507        compile command.
2508    
2509     5. Patch from the folks at Google for configure.in to be a bit more thorough
2510        in checking for a suitable C++ installation before trying to compile the
2511        C++ stuff. This should fix a reported problem when a compiler was present,
2512        but no suitable headers.
2513    
2514     6. The man pages all had just "PCRE" as their title. I have changed them to
2515        be the relevant file name. I have also arranged that these names are
2516        retained in the file doc/pcre.txt, which is a concatenation in text format
2517        of all the man pages except the little individual ones for each function.
2518    
2519     7. The NON-UNIX-USE file had not been updated for the different set of source
2520        files that come with release 6. I also added a few comments about the C++
2521        wrapper.
2522    
2523    
2524    Version 6.0 07-Jun-05
2525    ---------------------
2526    
2527     1. Some minor internal re-organization to help with my DFA experiments.
2528    
2529     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2530        didn't matter for the library itself when fully configured, but did matter
2531        when compiling without UCP support, or within Exim, where the ucp files are
2532        not imported.
2533    
2534     3. Refactoring of the library code to split up the various functions into
2535        different source modules. The addition of the new DFA matching code (see
2536        below) to a single monolithic source would have made it really too
2537        unwieldy, quite apart from causing all the code to be include in a
2538        statically linked application, when only some functions are used. This is
2539        relevant even without the DFA addition now that patterns can be compiled in
2540        one application and matched in another.
2541    
2542        The downside of splitting up is that there have to be some external
2543        functions and data tables that are used internally in different modules of
2544        the library but which are not part of the API. These have all had their
2545        names changed to start with "_pcre_" so that they are unlikely to clash
2546        with other external names.
2547    
2548     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2549        a different (DFA) algorithm. Although it is slower than the original
2550        function, it does have some advantages for certain types of matching
2551        problem.
2552    
2553     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2554        including restarting after a partial match.
2555    
2556     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2557        defined when compiling for Windows was sent to me. I have put it into the
2558        code, though I have no means of testing or verifying it.
2559    
2560     7. Added the pcre_refcount() auxiliary function.
2561    
2562     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2563        match before or at the first newline in the subject string. In pcretest,
2564        the /f option on a pattern can be used to set this.
2565    
2566     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2567        would behave wrongly. This has been present in PCRE since release 4.0.
2568    
2569    10. A number of changes to the pcregrep command:
2570    
2571        (a) Refactored how -x works; insert ^(...)$ instead of setting
2572            PCRE_ANCHORED and checking the length, in preparation for adding
2573            something similar for -w.
2574    
2575        (b) Added the -w (match as a word) option.
2576    
2577        (c) Refactored the way lines are read and buffered so as to have more
2578            than one at a time available.
2579    
2580        (d) Implemented a pcregrep test script.
2581    
2582        (e) Added the -M (multiline match) option. This allows patterns to match
2583            over several lines of the subject. The buffering ensures that at least
2584            8K, or the rest of the document (whichever is the shorter) is available
2585            for matching (and similarly the previous 8K for lookbehind assertions).
2586    
2587        (f) Changed the --help output so that it now says
2588    
2589              -w, --word-regex(p)
2590    
2591            instead of two lines, one with "regex" and the other with "regexp"
2592            because that confused at least one person since the short forms are the
2593            same. (This required a bit of code, as the output is generated
2594            automatically from a table. It wasn't just a text change.)
2595    
2596        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2597            option but starts with a hyphen. Could be a pattern or a path name
2598            starting with a hyphen, for instance.
2599    
2600        (h) "-" can be given as a file name to represent stdin.
2601    
2602        (i) When file names are being printed, "(standard input)" is used for
2603            the standard input, for compatibility with GNU grep. Previously
2604            "<stdin>" was used.
2605    
2606        (j) The option --label=xxx can be used to supply a name to be used for
2607            stdin when file names are being printed. There is no short form.
2608    
2609        (k) Re-factored the options decoding logic because we are going to add
2610            two more options that take data. Such options can now be given in four
2611            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2612    
2613        (l) Added the -A, -B, and -C options for requesting that lines of context
2614            around matches be printed.
2615    
2616        (m) Added the -L option to print the names of files that do not contain
2617            any matching lines, that is, the complement of -l.
2618    
2619        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2620            continue to scan other files.
2621    
2622        (o) The -s option was incorrectly implemented. For compatibility with other
2623            greps, it now suppresses the error message for a non-existent or non-
2624            accessible file (but not the return code). There is a new option called
2625            -q that suppresses the output of matching lines, which was what -s was
2626            previously doing.
2627    
2628        (p) Added --include and --exclude options to specify files for inclusion
2629            and exclusion when recursing.
2630    
2631    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2632        Hopefully, it now does.
2633    
2634    12. Missing cast in pcre_study().
2635    
2636    13. Added an "uninstall" target to the makefile.
2637    
2638    14. Replaced "extern" in the function prototypes in Makefile.in with
2639        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2640        world, but is set differently for Windows.
2641    
2642    15. Added a second compiling function called pcre_compile2(). The only
2643        difference is that it has an extra argument, which is a pointer to an
2644        integer error code. When there is a compile-time failure, this is set
2645        non-zero, in addition to the error test pointer being set to point to an
2646        error message. The new argument may be NULL if no error number is required
2647        (but then you may as well call pcre_compile(), which is now just a
2648        wrapper). This facility is provided because some applications need a
2649        numeric error indication, but it has also enabled me to tidy up the way
2650        compile-time errors are handled in the POSIX wrapper.
2651    
2652    16. Added VPATH=.libs to the makefile; this should help when building with one
2653        prefix path and installing with another. (Or so I'm told by someone who
2654        knows more about this stuff than I do.)
2655    
2656    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2657        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2658        match everything, including newlines. This is not POSIX-compatible, but
2659        somebody wanted the feature. From pcretest it can be activated by using
2660        both the P and the s flags.
2661    
2662    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2663    
2664    19. libpcre.pc was being incorrectly installed as executable.
2665    
2666    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2667        it now also looks for '\r' so that it will work unmodified on Windows.
2668    
2669    21. Added Google's contributed C++ wrapper to the distribution.
2670    
2671    22. Added some untidy missing memory free() calls in pcretest, to keep
2672        Electric Fence happy when testing.
2673    
2674    
2675    
2676    Version 5.0 13-Sep-04
2677    ---------------------
2678    
2679     1. Internal change: literal characters are no longer packed up into items
2680        containing multiple characters in a single byte-string. Each character
2681        is now matched using a separate opcode. However, there may be more than one
2682        byte in the character in UTF-8 mode.
2683    
2684     2. The pcre_callout_block structure has two new fields: pattern_position and
2685        next_item_length. These contain the offset in the pattern to the next match
2686        item, and its length, respectively.
2687    
2688     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2689        insertion of callouts before each pattern item. Added the /C option to
2690        pcretest to make use of this.
2691    
2692     4. On the advice of a Windows user, the lines
2693    
2694          #if defined(_WIN32) || defined(WIN32)
2695          _setmode( _fileno( stdout ), 0x8000 );
2696          #endif  /* defined(_WIN32) || defined(WIN32) */
2697    
2698        have been added to the source of pcretest. This apparently does useful
2699        magic in relation to line terminators.
2700    
2701     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2702        for the benefit of those environments where the "b" makes a difference.
2703    
2704     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2705        to know about it. I have put a hack into configure.in that adds in code
2706        to set GCC=yes if CC=icc. This seems to end up at a point in the
2707        generated configure script that is early enough to affect the setting of
2708        compiler options, which is what is needed, but I have no means of testing
2709        whether it really works. (The user who reported this had patched the
2710        generated configure script, which of course I cannot do.)
2711    
2712        LATER: After change 22 below (new libtool files), the configure script
2713        seems to know about icc (and also ecc). Therefore, I have commented out
2714        this hack in configure.in.
2715    
2716     7. Added support for pkg-config (2 patches were sent in).
2717    
2718     8. Negated POSIX character classes that used a combination of internal tables
2719        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2720        [[:^ascii]]. Typically, they would match almost any characters. The other
2721        POSIX classes were not broken in this way.
2722    
2723     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2724        to find the match, as PCRE was deluded into thinking that the match had to
2725        start at the start point or following a newline. The same bug applied to
2726        patterns with negative forward assertions or any backward assertions
2727        preceding ".*" at the start, unless the pattern required a fixed first
2728        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2729    
2730    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2731        starting at the last subject character, bytes beyond the end of the subject
2732        string were read.
2733    
2734    11. Renamed the variable "class" as "classbits" to make life easier for C++
2735        users. (Previously there was a macro definition, but it apparently wasn't
2736        enough.)
2737    
2738    12. Added the new field "tables" to the extra data so that tables can be passed
2739        in at exec time, or the internal tables can be re-selected. This allows
2740        a compiled regex to be saved and re-used at a later time by a different
2741        program that might have everything at different addresses.
2742    
2743    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2744        -R library as well as a -L library.
2745    
2746    14. The debugging options of pcretest (-d on the command line or D on a
2747        pattern) showed incorrect output for anything following an extended class
2748        that contained multibyte characters and which was followed by a quantifier.
2749    
2750    15. Added optional support for general category Unicode character properties
2751        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2752        support. It adds about 90K to the size of the library. The meanings of the
2753        inbuilt class escapes such as \d and \s have NOT been changed.
2754    
2755    16. Updated pcredemo.c to include calls to free() to release the memory for the
2756        compiled pattern.
2757    
2758    17. The generated file chartables.c was being created in the source directory
2759        instead of in the building directory. This caused the build to fail if the
2760        source directory was different from the building directory, and was
2761        read-only.
2762    
2763    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2764        file. No doubt somebody will tell me if they don't make sense... Also added
2765        Dan Mooney's comments about building on OpenVMS.
2766    
2767    19. Added support for partial matching via the PCRE_PARTIAL option for
2768        pcre_exec() and the \P data escape in pcretest.
2769    
2770    20. Extended pcretest with 3 new pattern features:
2771    
2772        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2773              write the compiled pattern to the file whose name is "rest-of-line".
2774              This is a straight binary dump of the data, with the saved pointer to
2775              the character tables forced to be NULL. The study data, if any, is
2776              written too. After writing, pcretest reads a new pattern.
2777    
2778        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2779              compiled pattern from the given file. There must not be any
2780              occurrences of "<" in the file name (pretty unlikely); if there are,
2781              pcretest will instead treat the initial "<" as a pattern delimiter.
2782              After reading in the pattern, pcretest goes on to read data lines as
2783              usual.
2784    
2785        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2786              and 16-bit fields in a compiled pattern, to simulate a pattern that
2787              was compiled on a host of opposite endianness.
2788    
2789    21. The pcre-exec() function can now cope with patterns that were compiled on
2790        hosts of opposite endianness, with this restriction:
2791    
2792          As for any compiled expression that is saved and used later, the tables
2793          pointer field cannot be preserved; the extra_data field in the arguments
2794          to pcre_exec() should be used to pass in a tables address if a value
2795          other than the default internal tables were used at compile time.
2796    
2797    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2798        now diagnosed as an error. Previously, most of the time, a negative number
2799        would have been treated as zero, but if in addition "ovector" was passed as
2800        NULL, a crash could occur.
2801    
2802    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2803        new versions from the libtool 1.5 distribution (the last one is a copy of
2804        a file called libtool.m4). This seems to have fixed the need to patch
2805        "configure" to support Darwin 1.3 (which I used to do). However, I still
2806        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2807        workstation).
2808    
2809    24. Changed the PCRE licence to be the more standard "BSD" licence.
2810    
2811    
2812    Version 4.5 01-Dec-03
2813    ---------------------
2814    
2815     1. There has been some re-arrangement of the code for the match() function so
2816        that it can be compiled in a version that does not call itself recursively.
2817        Instead, it keeps those local variables that need separate instances for
2818        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2819        needs to "recurse". Keeping track of where control must go is done by means
2820        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2821        hide most of the details from the main code, and operates only if
2822        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2823        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2824        operating.
2825    
2826        To make it easier for callers to provide specially tailored get/free
2827        functions for this usage, two new functions, pcre_stack_malloc, and
2828        pcre_stack_free, are used. They are always called in strict stacking order,
2829        and the size of block requested is always the same.
2830    
2831        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2832        PCRE has been compiled to use the stack or the heap for recursion. The
2833        -C option of pcretest uses this to show which version is compiled.
2834    
2835        A new data escape \S, is added to pcretest; it causes the amounts of store
2836        obtained and freed by both kinds of malloc/free at match time to be added
2837        to the output.
2838    
2839     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2840        what's available on my current Linux desktop machine.
2841    
2842     3. When matching a UTF-8 string, the test for a valid string at the start has
2843        been extended. If start_offset is not zero, PCRE now checks that it points
2844        to a byte that is the start of a UTF-8 character. If not, it returns
2845        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2846        this is necessary because there may be backward assertions in the pattern.
2847        When matching the same subject several times, it may save resources to use
2848        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2849    
2850     4. The code for checking the validity of UTF-8 strings has been tightened so
2851        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2852        containing "overlong sequences".
2853    
2854     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2855        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2856        should have been "&", but it just so happened that all the cases this let
2857        through by mistake were picked up later in the function.
2858    
2859     6. I had used a variable called "isblank" - this is a C99 function, causing
2860        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2861    
2862     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2863        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2864        so the version is shown; (c) stop "make test" echoing "./RunTest".
2865    
2866     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2867    
2868     9. The prototype for memmove() for systems that don't have it was using
2869        size_t, but the inclusion of the header that defines size_t was later. I've
2870        moved the #includes for the C headers earlier to avoid this.
2871    
2872    10. Added some adjustments to the code to make it easier to compiler on certain
2873        special systems:
2874    
2875          (a) Some "const" qualifiers were missing.
2876          (b) Added the macro EXPORT before all exported functions; by default this
2877              is defined to be empty.
2878          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2879              that it reads its output file name as an argument instead of writing
2880              to the standard output and assuming this can be redirected.
2881    
2882    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2883        class containing characters with values greater than 255, PCRE compilation
2884        went into a loop.
2885    
2886    12. A recursive reference to a subpattern that was within another subpattern
2887        that had a minimum quantifier of zero caused PCRE to crash. For example,
2888        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2889        recursion. If the recursively-called subpattern itself had a zero repeat,
2890        that was OK.
2891    
2892    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2893        buffer into which it was copied (for escape processing) was still set at
2894        1024, so long lines caused crashes.
2895    
2896    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2897        "internal error: code overflow...". This applied to any character class
2898        that was followed by a possessive quantifier.
2899    
2900    15. Modified the Makefile to add libpcre.la as a prerequisite for
2901        libpcreposix.la because I was told this is needed for a parallel build to
2902        work.
2903    
2904    16. If a pattern that contained .* following optional items at the start was
2905        studied, the wrong optimizing data was generated, leading to matching
2906        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2907        matching string must start with a or b or c. The correct conclusion for
2908        this pattern is that a match can start with any character.
2909    
2910    
2911    Version 4.4 13-Aug-03
2912    ---------------------
2913    
2914     1. In UTF-8 mode, a character class containing characters with values between
2915        127 and 255 was not handled correctly if the compiled pattern was studied.
2916        In fixing this, I have also improved the studying algorithm for such
2917        classes (slightly).
2918    
2919     2. Three internal functions had redundant arguments passed to them. Removal
2920        might give a very teeny performance improvement.
2921    
2922     3. Documentation bug: the value of the capture_top field in a callout is *one
2923        more than* the number of the hightest numbered captured substring.
2924    
2925     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2926        in incorrectly linking with a previously installed version. They now link
2927        explicitly with libpcre.la.
2928    
2929     5. configure.in no longer needs to recognize Cygwin specially.
2930    
2931     6. A problem in pcre.in for Windows platforms is fixed.
2932    
2933     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2934        pcretest, it used to include the size of the study block as part of its
2935        output. Unfortunately, the structure contains a field that has a different
2936        size on different hardware architectures. This meant that the tests that
2937        showed this size failed. As the block is currently always of a fixed size,
2938        this information isn't actually particularly useful in pcretest output, so
2939        I have just removed it.
2940    
2941     8. Three pre-processor statements accidentally did not start in column 1.
2942        Sadly, there are *still* compilers around that complain, even though
2943        standard C has not required this for well over a decade. Sigh.
2944    
2945     9. In pcretest, the code for checking callouts passed small integers in the
2946        callout_data field, which is a void * field. However, some picky compilers
2947        complained about the casts involved for this on 64-bit systems. Now
2948        pcretest passes the address of the small integer instead, which should get
2949        rid of the warnings.
2950    
2951    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2952        both compile and run time, and gives an error if an invalid UTF-8 sequence
2953        is found. There is a option for disabling this check in cases where the
2954        string is known to be correct and/or the maximum performance is wanted.
2955    
2956    11. In response to a bug report, I changed one line in Makefile.in from
2957    
2958            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2959        to
2960            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2961    
2962        to look similar to other lines, but I have no way of telling whether this
2963        is the right thing to do, as I do not use Windows. No doubt I'll get told
2964        if it's wrong...
2965    
2966    
2967    Version 4.3 21-May-03
2968    ---------------------
2969    
2970    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2971       Makefile.
2972    
2973    2. Some refactoring to improve the quality of the code:
2974    
2975       (i)   The utf8_table... variables are now declared "const".
2976    
2977       (ii)  The code for \cx, which used the "case flipping" table to upper case
2978             lower case letters, now just substracts 32. This is ASCII-specific,
2979             but the whole concept of \cx is ASCII-specific, so it seems
2980             reasonable.
2981    
2982       (iii) PCRE was using its character types table to recognize decimal and
2983             hexadecimal digits in the pattern. This is silly, because it handles
2984             only 0-9, a-f, and A-F, but the character types table is locale-
2985             specific, which means strange things might happen. A private
2986             table is now used for this - though it costs 256 bytes, a table is
2987             much faster than multiple explicit tests. Of course, the standard
2988             character types table is still used for matching digits in subject
2989             strings against \d.
2990    
2991       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2992             ending in _t are). So I've renamed it as ESC_tee.
2993    
2994    3. The first argument for regexec() in the POSIX wrapper should have been
2995       defined as "const".
2996    
2997    4. Changed pcretest to use malloc() for its buffers so that they can be
2998       Electric Fenced for debugging.
2999    
3000    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3001       to read one or more bytes before the start of the subject string. Often this
3002       had no effect on PCRE's behaviour, but in some circumstances it could
3003       provoke a segmentation fault.
3004    
3005    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3006       to try to read one or more bytes before the start of the subject string.
3007    
3008    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3009       UTF-8 support could misbehave in various ways if the subject string
3010       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3011       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3012       back over UTF-8 characters.)
3013    
3014    
3015    Version 4.2 14-Apr-03
3016    ---------------------
3017    
3018    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3019    
3020    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3021         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3022         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3023         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3024         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3025           and BUILD_EXEEXT
3026         Note: automatic setting of the BUILD variables is not yet working
3027         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3028           compile-time but not at link-time
3029         [LINK]: use for linking executables only
3030         make different versions for Windows and non-Windows
3031         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3032           libraries
3033         [LINK_FOR_BUILD]: new variable
3034         [OBJEXT]: use throughout
3035         [EXEEXT]: use throughout
3036         <winshared>: new target
3037         <wininstall>: new target
3038         <dftables.o>: use native compiler
3039         <dftables>: use native linker
3040         <install>: handle Windows platform correctly
3041         <clean>: ditto
3042         <check>: ditto
3043         copy DLL to top builddir before testing
3044    
3045       As part of these changes, -no-undefined was removed again. This was reported
3046       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3047       in any case.
3048    
3049    3. Some tidies to get rid of compiler warnings:
3050    
3051       . In the match_data structure, match_limit was an unsigned long int, whereas
3052         match_call_count was an int. I've made them both unsigned long ints.
3053    
3054       . In pcretest the fact that a const uschar * doesn't automatically cast to
3055         a void * provoked a warning.
3056    
3057       . Turning on some more compiler warnings threw up some "shadow" variables
3058         and a few more missing casts.
3059    
3060    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3061       option, a class that contained a single character with a value between 128
3062       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3063    
3064    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3065       option, a class that contained several characters, but with at least one
3066       whose value was between 128 and 255 caused PCRE to crash.
3067    
3068    
3069    Version 4.1 12-Mar-03
3070    ---------------------
3071    
3072    1. Compiling with gcc -pedantic found a couple of places where casts were
3073    needed, and a string in dftables.c that was longer than standard compilers are
3074    required to support.
3075    
3076    2. Compiling with Sun's compiler found a few more places where the code could
3077    be tidied up in order to avoid warnings.
3078    
3079    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3080    first of these names is deprecated in the latest Autoconf in favour of the name
3081    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3082    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3083    analogy I have changed it to CFLAGS_FOR_BUILD.
3084    
3085    4. Added -no-undefined to the linking command in the Makefile, because this is
3086    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3087    linking step for the pcreposix library.
3088    
3089    5. PCRE was failing to diagnose the case of two named groups with the same
3090    name.
3091    
3092    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3093    literal character that is needed in the subject for a match, and scans along to
3094    ensure that it is present before embarking on the full matching process. This
3095    saves time in cases of nested unlimited repeats that are never going to match.
3096    Problem: the scan can take a lot of time if the subject is very long (e.g.
3097    megabytes), thus penalizing straightforward matches. It is now done only if the
3098    amount of subject to be scanned is less than 1000 bytes.
3099    
3100    7. A lesser problem with the same optimization is that it was recording the
3101    first character of an anchored pattern as "needed", thus provoking a search
3102    right along the subject, even when the first match of the pattern was going to
3103    fail. The "needed" character is now not set for anchored patterns, unless it
3104    follows something in the pattern that is of non-fixed length. Thus, it still
3105    fulfils its original purpose of finding quick non-matches in cases of nested
3106    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3107    
3108    
3109    Version 4.0 17-Feb-03
3110    ---------------------
3111    
3112    1. If a comment in an extended regex that started immediately after a meta-item
3113    extended to the end of string, PCRE compiled incorrect data. This could lead to
3114    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3115    
3116    2. Moved to autoconf 2.53 and libtool 1.4.2.
3117    
3118    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3119    the special perltest8 script is no longer needed - all the tests can be run
3120    from a single perltest script.
3121    
3122    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3123    by \s. It has now been removed in PCRE. This means it isn't recognized as
3124    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3125    class [:space:] *does* include VT, thereby creating a mess.
3126    
3127    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3128    space and tab.
3129    
3130    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3131    its new features into the main test script, reducing the number of scripts.
3132    
3133    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3134    were backward compatible, and made the (?i) apply to the whole pattern, as if
3135    /i were given. Now it behaves more logically, and applies the option setting
3136    only to what follows. PCRE has been changed to follow suit. However, if it
3137    finds options settings right at the start of the pattern, it extracts them into
3138    the global options, as before. Thus, they show up in the info data.
3139    
3140    8. Added support for the \Q...\E escape sequence. Characters in between are
3141    treated as literals. This is slightly different from Perl in that $ and @ are
3142    also handled as literals inside the quotes. In Perl, they will cause variable
3143    interpolation. Note the following examples:
3144    
3145        Pattern            PCRE matches      Perl matches
3146    
3147        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3148        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3149        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3150    
3151    For compatibility with Perl, \Q...\E sequences are recognized inside character
3152    classes as well as outside them.
3153    
3154    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3155    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3156    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3157    signed/unsigned warnings.
3158    
3159    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3160    option for pcretest, so I've replaced it by a simple function that does just
3161    that job.
3162    
3163    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3164    "pcregrep -".
3165    
3166    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3167    Java package. This provides some syntactic sugar for simple cases of what my
3168    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3169    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3170    item, you can use this simplified notation. Note that only makes sense with
3171    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3172    greediness, whatever the setting of the PCRE_UNGREEDY option.
3173    
3174    13. A change of greediness default within a pattern was not taking effect at
3175    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3176    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3177    was abstracted outside.
3178    
3179    14. PCRE now supports the \G assertion. It is true when the current matching
3180    position is at the start point of the match. This differs from \A when the
3181    starting offset is non-zero. Used with the /g option of pcretest (or similar
3182    code), it works in the same way as it does for Perl's /g option. If all
3183    alternatives of a regex begin with \G, the expression is anchored to the start
3184    match position, and the "anchored" flag is set in the compiled expression.
3185    
3186    15. Some bugs concerning the handling of certain option changes within patterns
3187    have been fixed. These applied to options other than (?ims). For example,
3188    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3189    been the other way round. Some of this was related to change 7 above.
3190    
3191    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3192    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3193    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3194    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3195    
3196    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3197    mode. Unlike ".", it always matches newline, whatever the setting of
3198    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3199    assertions. Perl allows it, but it doesn't (in general) work because it can't
3200    calculate the length of the lookbehind. At least, that's the case for Perl
3201    5.8.0 - I've been told they are going to document that it doesn't work in
3202    future.
3203    
3204    18. Added an error diagnosis for escapes that PCRE does not support: these are
3205    \L, \l, \N, \P, \p, \U, \u, and \X.
3206    
3207    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3208    reading past the end of the pattern in cases such as /[abcd/.
3209    
3210    20. PCRE was getting more memory than necessary for patterns with classes that
3211    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3212    
3213    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3214    compiling PCRE for use with Virtual Pascal.
3215    
3216    22. Small fix to the Makefile to make it work properly if the build is done
3217    outside the source tree.
3218    
3219    23. Added a new extension: a condition to go with recursion. If a conditional
3220    subpattern starts with (?(R) the "true" branch is used if recursion has
3221    happened, whereas the "false" branch is used only at the top level.
3222    
3223    24. When there was a very long string of literal characters (over 255 bytes
3224    without UTF support, over 250 bytes with UTF support), the computation of how
3225    much memory was required could be incorrect, leading to segfaults or other
3226    strange effects.
3227    
3228    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3229    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3230    there was a subsequent back reference to those brackets. This meant that, for
3231    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3232    possible to check for precisely this case. All we can do is abandon the
3233    optimization if .* occurs inside capturing brackets when there are any back
3234    references whatsoever. (See below for a better fix that came later.)
3235    
3236    26. The handling of the optimization for finding the first character of a
3237    non-anchored pattern, and for finding a character that is required later in the
3238    match were failing in some cases. This didn't break the matching; it just
3239    failed to optimize when it could. The way this is done has been re-implemented.
3240    
3241    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3242    
3243    28. Added a new feature that provides some of the functionality that Perl
3244    provides with (?{...}). The facility is termed a "callout". The way it is done
3245    in PCRE is for the caller to provide an optional function, by setting
3246    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3247    global variable. By default it is unset, which disables all calling out. To get
3248    the function called, the regex must include (?C) at appropriate points. This
3249    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3250    This provides a means of identifying different callout points. When PCRE
3251    reaches such a point in the regex, if pcre_callout has been set, the external
3252    function is called. It is provided with data in a structure called
3253    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3254    matching continues; if it returns a non-zero value, the match at the current
3255    point fails. However, backtracking will occur if possible. [This was changed
3256    later and other features added - see item 49 below.]
3257    
3258    29. pcretest is upgraded to test the callout functionality. It provides a
3259    callout function that displays information. By default, it shows the start of
3260    the match and the current position in the text. There are some new data escapes
3261    to vary what happens:
3262    
3263        \C+         in addition, show current contents of captured substrings
3264        \C-         do not supply a callout function
3265        \C!n        return 1 when callout number n is reached
3266        \C!n!m      return 1 when callout number n is reached for the mth time
3267    
3268    30. If pcregrep was called with the -l option and just a single file name, it
3269    output "<stdin>" if a match was found, instead of the file name.
3270    
3271    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3272    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3273    pcre_exec(). This saves a malloc/free per call. The default value of
3274    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3275    when configuring.
3276    
3277    32. The default maximum size of a compiled pattern is 64K. There have been a
3278    few cases of people hitting this limit. The code now uses macros to handle the
3279    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3280    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3281    configuring. Tests 2 and 5 work only with 2-byte links because they output
3282    debugging information about compiled patterns.
3283    
3284    33. Internal code re-arrangements:
3285    
3286    (a) Moved the debugging function for printing out a compiled regex into
3287        its own source file (printint.c) and used #include to pull it into
3288        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3289        separate copies.
3290    
3291    (b) Defined the list of op-code names for debugging as a macro in
3292        internal.h so that it is next to the definition of the opcodes.
3293    
3294    (c) Defined a table of op-code lengths for simpler skipping along compiled
3295        code. This is again a macro in internal.h so that it is next to the
3296        definition of the opcodes.
3297    
3298    34. Added support for recursive calls to individual subpatterns, along the
3299    lines of Robin Houston's patch (but implemented somewhat differently).
3300    
3301    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3302    allow it to read and process whole directories in Win32. This code was
3303    contributed by Lionel Fourquaux; it has not been tested by me.
3304    
3305    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3306    used to name a group. Names consist of alphanumerics and underscores, and must
3307    be unique. Back references use the syntax (?P=name) and recursive calls use
3308    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3309    numbers. The function pcre_fullinfo() can be used after compilation to extract
3310    a name/number map. There are three relevant calls:
3311    
3312      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3313      PCRE_INFO_NAMECOUNT            yields the number of entries
3314      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3315    
3316    The map is a vector of fixed-size entries. The size of each entry depends on
3317    the length of the longest name used. The first two bytes of each entry are the
3318    group number, most significant byte first. There follows the corresponding
3319    name, zero terminated. The names are in alphabetical order.
3320    
3321    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3322    case instead of 255. Making it the same both with and without UTF-8 support
3323    means that the same test output works with both.
3324    
3325    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3326    calling malloc() with a zero argument.
3327    
3328    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3329    optimization. I've improved things by keeping a bitmap of backreferences with
3330    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3331    fact referenced, the optimization can be applied. It is unlikely that a
3332    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3333    the match to follow \n) will appear inside brackets with a number greater than
3334    31, but if it does, any back reference > 31 suppresses the optimization.
3335    
3336    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3337    of disabling numbered capturing parentheses. Any opening parenthesis that is
3338    not followed by ? behaves as if it were followed by ?: but named parentheses
3339    can still be used for capturing (and they will acquire numbers in the usual
3340    way).
3341    
3342    41. Redesigned the return codes from the match() function into yes/no/error so
3343    that errors can be passed back from deep inside the nested calls. A malloc
3344    failure while inside a recursive subpattern call now causes the
3345    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3346    
3347    42. It is now possible to set a limit on the number of times the match()
3348    function is called in a call to pcre_exec(). This facility makes it possible to
3349    limit the amount of recursion and backtracking, though not in a directly
3350    obvious way, because the match() function is used in a number of different
3351    circumstances. The count starts from zero for each position in the subject
3352    string (for non-anchored patterns). The default limit is, for compatibility, a
3353    large number, namely 10 000 000. You can change this in two ways:
3354    
3355    (a) When configuring PCRE before making, you can use --with-match-limit=n
3356        to set a default value for the compiled library.
3357    
3358    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3359        a different value is set. See 45 below.
3360    
3361    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3362    
3363    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3364    of things that can be changed at compile time. The first argument specifies
3365    what is wanted and the second points to where the information is to be placed.
3366    The current list of available information is:
3367    
3368      PCRE_CONFIG_UTF8
3369    
3370    The output is an integer that is set to one if UTF-8 support is available;
3371    otherwise it is set to zero.
3372    
3373      PCRE_CONFIG_NEWLINE
3374    
3375    The output is an integer that it set to the value of the code that is used for
3376    newline. It is either LF (10) or CR (13).
3377    
3378      PCRE_CONFIG_LINK_SIZE
3379    
3380    The output is an integer that contains the number of bytes used for internal
3381    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3382    
3383      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3384    
3385    The output is an integer that contains the threshold above which the POSIX
3386    interface uses malloc() for output vectors. See item 31 above.
3387    
3388      PCRE_CONFIG_MATCH_LIMIT
3389    
3390    The output is an unsigned integer that contains the default limit of the number
3391    of match() calls in a pcre_exec() execution. See 42 above.
3392    
3393    44. pcretest has been upgraded by the addition of the -C option. This causes it
3394    to extract all the available output from the new pcre_config() function, and to
3395    output it. The program then exits immediately.
3396    
3397    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3398    order to support additional features. One way would have been to define
3399    pcre_exec2() (for example) with extra arguments, but this would not have been
3400    extensible, and would also have required all calls to the original function to
3401    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3402    is used for passing in "extra" data from pcre_study().
3403    
3404    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3405    contains the following fields:
3406    
3407      flags         a bitmap indicating which of the following fields are set
3408      study_data    opaque data from pcre_study()
3409      match_limit   a way of specifying a limit on match() calls for a specific
3410                      call to pcre_exec()
3411      callout_data  data for callouts (see 49 below)
3412    
3413    The flag bits are also defined in pcre.h, and are
3414    
3415      PCRE_EXTRA_STUDY_DATA
3416      PCRE_EXTRA_MATCH_LIMIT
3417      PCRE_EXTRA_CALLOUT_DATA
3418    
3419    The pcre_study() function now returns one of these new pcre_extra blocks, with
3420    the actual study data pointed to by the study_data field, and the
3421    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3422    before. That is, this change is entirely upwards-compatible and requires no
3423    change to existing code.
3424    
3425    If you want to pass in additional data to pcre_exec(), you can either place it
3426    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3427    block.
3428    
3429    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3430    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3431    times with different match limits, until it finds the minimum value needed for
3432    pcre_exec() to complete. The value is then output. This can be instructive; for
3433    most simple matches the number is quite small, but for pathological cases it
3434    gets very large very quickly.
3435    
3436    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3437    returns the size of the data block pointed to by the study_data field in a
3438    pcre_extra block, that is, the value that was passed as the argument to
3439    pcre_malloc() when PCRE was getting memory in which to place the information
3440    created by pcre_study(). The fourth argument should point to a size_t variable.
3441    pcretest has been extended so that this information is shown after a successful
3442    pcre_study() call when information about the compiled regex is being displayed.
3443    
3444    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3445    because what follows is always an absolute path. (Later: it turns out that this
3446    is more than cosmetic for MinGW, because it doesn't like empty path
3447    components.)
3448    
3449    49. Some changes have been made to the callout feature (see 28 above):
3450    
3451    (i)  A callout function now has three choices for what it returns:
3452    
3453           0  =>  success, carry on matching
3454         > 0  =>  failure at this point, but backtrack if possible
3455         < 0  =>  serious error, return this value from pcre_exec()
3456    
3457         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3458         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3459         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3460         use by callout functions. It will never be used by PCRE itself.
3461    
3462    (ii) The pcre_extra structure (see 45 above) has a void * field called
3463         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3464         pcre_callout_block structure has a field of the same name. The contents of
3465         the field passed in the pcre_extra structure are passed to the callout
3466         function in the corresponding field in the callout block. This makes it
3467         easier to use the same callout-containing regex from multiple threads. For
3468         testing, the pcretest program has a new data escape
3469    
3470           \C*n        pass the number n (may be negative) as callout_data
3471    
3472         If the callout function in pcretest receives a non-zero value as
3473         callout_data, it returns that value.
3474    
3475    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3476    there were some redundant $(CFLAGS) in commands that are now specified as
3477    $(LINK), which already includes $(CFLAGS).
3478    
3479    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3480    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3481    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3482    one-byte characters throughout. Note that case-insensitive matching applies
3483    only to characters whose values are less than 256. PCRE doesn't support the
3484    notion of cases for higher-valued characters.
3485    
3486    (i)   A character class whose characters are all within 0-255 is handled as
3487          a bit map, and the map is inverted for negative classes. Previously, a
3488          character > 255 always failed to match such a class; however it should
3489          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3490    
3491    (ii)  A negated character class with a single character < 255 is coded as
3492          "not this character" (OP_NOT). This wasn't working properly when the test
3493          character was multibyte, either singly or repeated.
3494    
3495    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3496          mode, for example: \x{100}{2,3}.
3497    
3498    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either