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

Diff of /code/trunk/ChangeLog

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

revision 49 by nigel, Sat Feb 24 21:39:33 2007 UTC revision 620 by ph10, Sun Jul 17 13:53:14 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    
149    Version 8.12 15-Jan-2011
150    ------------------------
151    
152    1.  Fixed some typos in the markup of the man pages, and wrote a script that
153        checks for such things as part of the documentation building process.
154    
155    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
156        --match-limit and --recursion-limit options (added for 8.11). In
157        particular, this made one of the standard tests fail. (The integer value
158        went into the wrong half of a long int.)
159    
160    3.  If the --colour option was given to pcregrep with -v (invert match), it
161        did strange things, either producing crazy output, or crashing. It should,
162        of course, ignore a request for colour when reporting lines that do not
163        match.
164    
165    4.  Another pcregrep bug caused similar problems if --colour was specified with
166        -M (multiline) and the pattern match finished with a line ending.
167    
168    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
169        matched in multiline mode, the following line was shown as part of the
170        match. This seems wrong, so I have changed it.
171    
172    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
173        the check for further matches in the same line (so they could be coloured)
174        to overrun the end of the current line. If another match was found, it was
175        incorrectly shown (and then shown again when found in the next line).
176    
177    7.  If pcregrep was compiled under Windows, there was a reference to the
178        function pcregrep_exit() before it was defined. I am assuming this was
179        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
180        reported by a user. I've moved the definition above the reference.
181    
182    
183    Version 8.11 10-Dec-2010
184    ------------------------
185    
186    1.  (*THEN) was not working properly if there were untried alternatives prior
187        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
188        backtracked to try for "b" instead of moving to the next alternative branch
189        at the same level (in this case, to look for "c"). The Perl documentation
190        is clear that when (*THEN) is backtracked onto, it goes to the "next
191        alternative in the innermost enclosing group".
192    
193    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
194        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
195        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
196        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
197        (*THEN).
198    
199    3.  If \s appeared in a character class, it removed the VT character from
200        the class, even if it had been included by some previous item, for example
201        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
202        of \s, but is part of the POSIX "space" class.)
203    
204    4.  A partial match never returns an empty string (because you can always
205        match an empty string at the end of the subject); however the checking for
206        an empty string was starting at the "start of match" point. This has been
207        changed to the "earliest inspected character" point, because the returned
208        data for a partial match starts at this character. This means that, for
209        example, /(?<=abc)def/ gives a partial match for the subject "abc"
210        (previously it gave "no match").
211    
212    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
213        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
214        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
215        has an implication that the given string is incomplete (because a partial
216        match is preferred over a full match). For this reason, these items now
217        give a partial match in this situation. [Aside: previously, the one case
218        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
219        match rather than a full match, which was wrong by the old rules, but is
220        now correct.]
221    
222    6.  There was a bug in the handling of #-introduced comments, recognized when
223        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
224        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
225        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
226        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
227        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
228        places in pcre_compile().
229    
230    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
231        comments when looking ahead for named forward references to subpatterns,
232        the only newline sequence it recognized was NL. It now handles newlines
233        according to the set newline convention.
234    
235    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
236        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
237        cater for a lack of strerror(). These oversights have been fixed.
238    
239    9.  Added --match-limit and --recursion-limit to pcregrep.
240    
241    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
242    
243    11. When the -o option was used, pcregrep was setting a return code of 1, even
244        when matches were found, and --line-buffered was not being honoured.
245    
246    12. Added an optional parentheses number to the -o and --only-matching options
247        of pcregrep.
248    
249    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
250        can match an empty string. The code to do it in pcretest and pcredemo
251        needed fixing:
252    
253        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
254            only one byte after an empty string match just before CRLF (this case
255            just got forgotten; "any" and "anycrlf" were OK).
256    
257        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
258            mode when an empty string match preceded an ASCII character followed by
259            a non-ASCII character. (The code for advancing by one character rather
260            than one byte was nonsense.)
261    
262        (c) The pcredemo.c sample program did not have any code at all to handle
263            the cases when CRLF is a valid newline sequence.
264    
265    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
266        as a starting offset was within the subject string. There is now a new
267        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
268        negative or greater than the length of the string. In order to test this,
269        pcretest is extended to allow the setting of negative starting offsets.
270    
271    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
272        starting offset points to the beginning of a UTF-8 character was
273        unnecessarily clumsy. I tidied it up.
274    
275    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
276        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
277    
278    17. Nobody had reported that the --include_dir option, which was added in
279        release 7.7 should have been called --include-dir (hyphen, not underscore)
280        for compatibility with GNU grep. I have changed it to --include-dir, but
281        left --include_dir as an undocumented synonym, and the same for
282        --exclude-dir, though that is not available in GNU grep, at least as of
283        release 2.5.4.
284    
285    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
286        characters from a string of bytes) have been redefined so as not to use
287        loops, in order to improve performance in some environments. At the same
288        time, I abstracted some of the common code into auxiliary macros to save
289        repetition (this should not affect the compiled code).
290    
291    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
292        compile-time error is now given if \c is not followed by an ASCII
293        character, that is, a byte less than 128. (In EBCDIC mode, the code is
294        different, and any byte value is allowed.)
295    
296    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
297        START_OPTIMIZE option, which is now allowed at compile time - but just
298        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
299        to pcregrep and other applications that have no direct access to PCRE
300        options. The new /Y option in pcretest sets this option when calling
301        pcre_compile().
302    
303    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
304        back references. Groups containing recursive back references were forced to
305        be atomic by that change, but in the case of named groups, the amount of
306        memory required was incorrectly computed, leading to "Failed: internal
307        error: code overflow". This has been fixed.
308    
309    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
310        pcretest.c, to avoid build problems in some Borland environments.
311    
312    
313    Version 8.10 25-Jun-2010
314    ------------------------
315    
316    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
317        THEN.
318    
319    2.  (*ACCEPT) was not working when inside an atomic group.
320    
321    3.  Inside a character class, \B is treated as a literal by default, but
322        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
323        causes the error). The code is unchanged, but I tidied the documentation.
324    
325    4.  Inside a character class, PCRE always treated \R and \X as literals,
326        whereas Perl faults them if its -w option is set. I have changed PCRE so
327        that it faults them when PCRE_EXTRA is set.
328    
329    5.  Added support for \N, which always matches any character other than
330        newline. (It is the same as "." when PCRE_DOTALL is not set.)
331    
332    6.  When compiling pcregrep with newer versions of gcc which may have
333        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
334        declared with attribute warn_unused_result" were given. Just casting the
335        result to (void) does not stop the warnings; a more elaborate fudge is
336        needed. I've used a macro to implement this.
337    
338    7.  Minor change to pcretest.c to avoid a compiler warning.
339    
340    8.  Added four artifical Unicode properties to help with an option to make
341        \s etc use properties (see next item). The new properties are: Xan
342        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
343    
344    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
345        use Unicode properties. (*UCP) at the start of a pattern can be used to set
346        this option. Modified pcretest to add /W to test this facility. Added
347        REG_UCP to make it available via the POSIX interface.
348    
349    10. Added --line-buffered to pcregrep.
350    
351    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
352        studied, and the match started with a letter with a code point greater than
353        127 whose first byte was different to the first byte of the other case of
354        the letter, the other case of this starting letter was not recognized
355        (#976).
356    
357    12. If a pattern that was studied started with a repeated Unicode property
358        test, for example, \p{Nd}+, there was the theoretical possibility of
359        setting up an incorrect bitmap of starting bytes, but fortunately it could
360        not have actually happened in practice until change 8 above was made (it
361        added property types that matched character-matching opcodes).
362    
363    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
364        possible starting bytes for non-anchored patterns.
365    
366    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
367        \R, and also a number of cases that involve Unicode properties, both
368        explicit and implicit when PCRE_UCP is set.
369    
370    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
371        input, it could crash or give wrong results if characters with values
372        greater than 0xc0 were present in the subject string. (Detail: it assumed
373        UTF-8 input when processing these items.)
374    
375    16. Added a lot of (int) casts to avoid compiler warnings in systems where
376        size_t is 64-bit (#991).
377    
378    17. Added a check for running out of memory when PCRE is compiled with
379        --disable-stack-for-recursion (#990).
380    
381    18. If the last data line in a file for pcretest does not have a newline on
382        the end, a newline was missing in the output.
383    
384    19. The default pcre_chartables.c file recognizes only ASCII characters (values
385        less than 128) in its various bitmaps. However, there is a facility for
386        generating tables according to the current locale when PCRE is compiled. It
387        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
388        characters, are recognized by isspace() and therefore were getting set in
389        these tables, and indeed these tables seem to approximate to ISO 8859. This
390        caused a problem in UTF-8 mode when pcre_study() was used to create a list
391        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
392        which of course cannot start UTF-8 characters. I have changed the code so
393        that only real ASCII characters (less than 128) and the correct starting
394        bytes for UTF-8 encodings are set for characters greater than 127 when in
395        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
396        altogether.)
397    
398    20. Added the /T option to pcretest so as to be able to run tests with non-
399        standard character tables, thus making it possible to include the tests
400        used for 19 above in the standard set of tests.
401    
402    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
403        reference to a subpattern the other side of a comment that contains an
404        opening parenthesis caused either an internal compiling error, or a
405        reference to the wrong subpattern.
406    
407    
408    Version 8.02 19-Mar-2010
409    ------------------------
410    
411    1.  The Unicode data tables have been updated to Unicode 5.2.0.
412    
413    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
414        configured.
415    
416    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
417        original author of that file, following a query about its status.
418    
419    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
420        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
421    
422    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
423        quantifier applied to a forward-referencing subroutine call, could compile
424        incorrect code or give the error "internal error: previously-checked
425        referenced subpattern not found".
426    
427    6.  Both MS Visual Studio and Symbian OS have problems with initializing
428        variables to point to external functions. For these systems, therefore,
429        pcre_malloc etc. are now initialized to local functions that call the
430        relevant global functions.
431    
432    7.  There were two entries missing in the vectors called coptable and poptable
433        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
434        I've fixed the data, and added a kludgy way of testing at compile time that
435        the lengths are correct (equal to the number of opcodes).
436    
437    8.  Following on from 7, I added a similar kludge to check the length of the
438        eint vector in pcreposix.c.
439    
440    9.  Error texts for pcre_compile() are held as one long string to avoid too
441        much relocation at load time. To find a text, the string is searched,
442        counting zeros. There was no check for running off the end of the string,
443        which could happen if a new error number was added without updating the
444        string.
445    
446    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
447    
448    11. \K was not working if it appeared in an atomic group or in a group that
449        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
450        \K is "not well defined" if used in an assertion. PCRE now accepts it if
451        the assertion is positive, but not if it is negative.
452    
453    12. Change 11 fortuitously reduced the size of the stack frame used in the
454        "match()" function of pcre_exec.c by one pointer. Forthcoming
455        implementation of support for (*MARK) will need an extra pointer on the
456        stack; I have reserved it now, so that the stack frame size does not
457        decrease.
458    
459    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
460        item in branch that calls a recursion is a subroutine call - as in the
461        second branch in the above example - was incorrectly given the compile-
462        time error "recursive call could loop indefinitely" because pcre_compile()
463        was not correctly checking the subroutine for matching a non-empty string.
464    
465    14. The checks for overrunning compiling workspace could trigger after an
466        overrun had occurred. This is a "should never occur" error, but it can be
467        triggered by pathological patterns such as hundreds of nested parentheses.
468        The checks now trigger 100 bytes before the end of the workspace.
469    
470    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
471    
472    
473    Version 8.01 19-Jan-2010
474    ------------------------
475    
476    1.  If a pattern contained a conditional subpattern with only one branch (in
477        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
478        computed the wrong minimum data length (which is of course zero for such
479        subpatterns). This could cause incorrect "no match" results.
480    
481    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
482        the pattern is reset in the first branch, pcre_compile() failed with
483        "internal error: code overflow at offset...". This happened only when
484        the reset was to the original external option setting. (An optimization
485        abstracts leading options settings into an external setting, which was the
486        cause of this.)
487    
488    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
489        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
490        assertion pattern did not match (meaning that the assertion was true), it
491        was incorrectly treated as false if the SKIP had been reached during the
492        matching. This also applied to assertions used as conditions.
493    
494    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
495        assertion subpattern, including such a pattern used as a condition,
496        unpredictable results occurred, instead of the error return
497        PCRE_ERROR_DFA_UITEM.
498    
499    5.  The C++ GlobalReplace function was not working like Perl for the special
500        situation when an empty string is matched. It now does the fancy magic
501        stuff that is necessary.
502    
503    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
504        removed. (These were left over from very, very early versions of PCRE.)
505    
506    7.  Some cosmetic changes to the code to make life easier when compiling it
507        as part of something else:
508    
509        (a) Change DEBUG to PCRE_DEBUG.
510    
511        (b) In pcre_compile(), rename the member of the "branch_chain" structure
512            called "current" as "current_branch", to prevent a collision with the
513            Linux macro when compiled as a kernel module.
514    
515        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
516            prevent a collision with the Linux macro when compiled as a kernel
517            module.
518    
519    8.  In pcre_compile() there are some checks for integer overflows that used to
520        cast potentially large values to (double). This has been changed to that
521        when building, a check for int64_t is made, and if it is found, it is used
522        instead, thus avoiding the use of floating point arithmetic. (There is no
523        other use of FP in PCRE.) If int64_t is not found, the fallback is to
524        double.
525    
526    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
527        2005 (difference between two addresses compared to an unsigned value).
528    
529    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
530        custom one, because of the following reported problem in Windows:
531    
532          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
533              under Win32.
534          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
535              therefore missing the function definition.
536          - The compiler thus generates a "C" signature for the test function.
537          - The linker fails to find the "C" function.
538          - PCRE fails to configure if asked to do so against libbz2.
539    
540    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
541        messages were output:
542    
543          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
544          rerunning libtoolize, to keep the correct libtool macros in-tree.
545          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
546    
547        I have done both of these things.
548    
549    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
550        most of the time, it *can* run out if it is given a pattern that contains a
551        runaway infinite recursion. I updated the discussion in the pcrestack man
552        page.
553    
554    13. Now that we have gone to the x.xx style of version numbers, the minor
555        version may start with zero. Using 08 or 09 is a bad idea because users
556        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
557        interpreted as invalid octal numbers. I've updated the previous comment in
558        configure.ac, and also added a check that gives an error if 08 or 09 are
559        used.
560    
561    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
562        causing partial matching to fail when the end of the subject matched \W
563        in a UTF-8 pattern where \W was quantified with a minimum of 3.
564    
565    15. There were some discrepancies between the declarations in pcre_internal.h
566        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
567        their definitions. The declarations used "const uschar *" and the
568        definitions used USPTR. Even though USPTR is normally defined as "const
569        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
570        reported that: "This difference in casting confuses some C++ compilers, for
571        example, SunCC recognizes above declarations as different functions and
572        generates broken code for hbpcre." I have changed the declarations to use
573        USPTR.
574    
575    16. GNU libtool is named differently on some systems. The autogen.sh script now
576        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
577        (FreeBSD).
578    
579    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
580        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
581        comment: "Figure out how to create a longlong from a string: strtoll and
582        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
583        instance, but it only takes 2 args instead of 3!"
584    
585    18. A subtle bug concerned with back references has been fixed by a change of
586        specification, with a corresponding code fix. A pattern such as
587        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
588        refers, was giving matches when it shouldn't. For example, xa=xaaa would
589        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
590        same bug. Such groups have to be quantified to be useful, or contained
591        inside another quantified group. (If there's no repetition, the reference
592        can never match.) The problem arises because, having left the group and
593        moved on to the rest of the pattern, a later failure that backtracks into
594        the group uses the captured value from the final iteration of the group
595        rather than the correct earlier one. I have fixed this in PCRE by forcing
596        any group that contains a reference to itself to be an atomic group; that
597        is, there cannot be any backtracking into it once it has completed. This is
598        similar to recursive and subroutine calls.
599    
600    
601    Version 8.00 19-Oct-09
602    ----------------------
603    
604    1.  The table for translating pcre_compile() error codes into POSIX error codes
605        was out-of-date, and there was no check on the pcre_compile() error code
606        being within the table. This could lead to an OK return being given in
607        error.
608    
609    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
610        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
611        in a Windows environment.
612    
613    3.  The pcregrep --count option prints the count for each file even when it is
614        zero, as does GNU grep. However, pcregrep was also printing all files when
615        --files-with-matches was added. Now, when both options are given, it prints
616        counts only for those files that have at least one match. (GNU grep just
617        prints the file name in this circumstance, but including the count seems
618        more useful - otherwise, why use --count?) Also ensured that the
619        combination -clh just lists non-zero counts, with no names.
620    
621    4.  The long form of the pcregrep -F option was incorrectly implemented as
622        --fixed_strings instead of --fixed-strings. This is an incompatible change,
623        but it seems right to fix it, and I didn't think it was worth preserving
624        the old behaviour.
625    
626    5.  The command line items --regex=pattern and --regexp=pattern were not
627        recognized by pcregrep, which required --regex pattern or --regexp pattern
628        (with a space rather than an '='). The man page documented the '=' forms,
629        which are compatible with GNU grep; these now work.
630    
631    6.  No libpcreposix.pc file was created for pkg-config; there was just
632        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
633    
634    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
635        when UCP support is not needed, by modifying the Python script that
636        generates it from Unicode data files. This should not matter if the module
637        is correctly used as a library, but I received one complaint about 50K of
638        unwanted data. My guess is that the person linked everything into his
639        program rather than using a library. Anyway, it does no harm.
640    
641    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
642        was a minimum greater than 1 for a wide character in a possessive
643        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
644        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
645        character. Chaos in the form of incorrect output or a compiling loop could
646        result.
647    
648    9.  The restrictions on what a pattern can contain when partial matching is
649        requested for pcre_exec() have been removed. All patterns can now be
650        partially matched by this function. In addition, if there are at least two
651        slots in the offset vector, the offset of the earliest inspected character
652        for the match and the offset of the end of the subject are set in them when
653        PCRE_ERROR_PARTIAL is returned.
654    
655    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
656        synonymous with PCRE_PARTIAL, for backwards compatibility, and
657        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
658        and may be more useful for multi-segment matching.
659    
660    11. Partial matching with pcre_exec() is now more intuitive. A partial match
661        used to be given if ever the end of the subject was reached; now it is
662        given only if matching could not proceed because another character was
663        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
664        string "Z", which now yields "no match" instead of "partial match". In the
665        case of pcre_dfa_exec(), "no match" is given if every matching path for the
666        final character ended with (*FAIL).
667    
668    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
669        if the pattern had a "must contain" character that was already found in the
670        earlier partial match, unless partial matching was again requested. For
671        example, with the pattern /dog.(body)?/, the "must contain" character is
672        "g". If the first part-match was for the string "dog", restarting with
673        "sbody" failed. This bug has been fixed.
674    
675    13. The string returned by pcre_dfa_exec() after a partial match has been
676        changed so that it starts at the first inspected character rather than the
677        first character of the match. This makes a difference only if the pattern
678        starts with a lookbehind assertion or \b or \B (\K is not supported by
679        pcre_dfa_exec()). It's an incompatible change, but it makes the two
680        matching functions compatible, and I think it's the right thing to do.
681    
682    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
683        so that the demonstration program is easily available in environments where
684        PCRE has not been installed from source.
685    
686    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
687        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
688        library.
689    
690    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
691        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
692        is not the first non-POSIX option to be added. Clearly some people find
693        these options useful.
694    
695    17. If a caller to the POSIX matching function regexec() passes a non-zero
696        value for nmatch with a NULL value for pmatch, the value of
697        nmatch is forced to zero.
698    
699    18. RunGrepTest did not have a test for the availability of the -u option of
700        the diff command, as RunTest does. It now checks in the same way as
701        RunTest, and also checks for the -b option.
702    
703    19. If an odd number of negated classes containing just a single character
704        interposed, within parentheses, between a forward reference to a named
705        subpattern and the definition of the subpattern, compilation crashed with
706        an internal error, complaining that it could not find the referenced
707        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
708        [The bug was that it was starting one character too far in when skipping
709        over the character class, thus treating the ] as data rather than
710        terminating the class. This meant it could skip too much.]
711    
712    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
713        /g option in pcretest when the pattern contains \K, which makes it possible
714        to have an empty string match not at the start, even when the pattern is
715        anchored. Updated pcretest and pcredemo to use this option.
716    
717    21. If the maximum number of capturing subpatterns in a recursion was greater
718        than the maximum at the outer level, the higher number was returned, but
719        with unset values at the outer level. The correct (outer level) value is
720        now given.
721    
722    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
723        PCRE did not set those parentheses (unlike Perl). I have now found a way to
724        make it do so. The string so far is captured, making this feature
725        compatible with Perl.
726    
727    23. The tests have been re-organized, adding tests 11 and 12, to make it
728        possible to check the Perl 5.10 features against Perl 5.10.
729    
730    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
731        pattern matches a fixed length string. PCRE did not allow this; now it
732        does. Neither allows recursion.
733    
734    25. I finally figured out how to implement a request to provide the minimum
735        length of subject string that was needed in order to match a given pattern.
736        (It was back references and recursion that I had previously got hung up
737        on.) This code has now been added to pcre_study(); it finds a lower bound
738        to the length of subject needed. It is not necessarily the greatest lower
739        bound, but using it to avoid searching strings that are too short does give
740        some useful speed-ups. The value is available to calling programs via
741        pcre_fullinfo().
742    
743    26. While implementing 25, I discovered to my embarrassment that pcretest had
744        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
745        study optimizations had never been tested with that matching function.
746        Oops. What is worse, even when it was passed study data, there was a bug in
747        pcre_dfa_exec() that meant it never actually used it. Double oops. There
748        were also very few tests of studied patterns with pcre_dfa_exec().
749    
750    27. If (?| is used to create subpatterns with duplicate numbers, they are now
751        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
752        on the other side of the coin, they are no longer allowed to have different
753        names, because these cannot be distinguished in PCRE, and this has caused
754        confusion. (This is a difference from Perl.)
755    
756    28. When duplicate subpattern names are present (necessarily with different
757        numbers, as required by 27 above), and a test is made by name in a
758        conditional pattern, either for a subpattern having been matched, or for
759        recursion in such a pattern, all the associated numbered subpatterns are
760        tested, and the overall condition is true if the condition is true for any
761        one of them. This is the way Perl works, and is also more like the way
762        testing by number works.
763    
764    
765    Version 7.9 11-Apr-09
766    ---------------------
767    
768    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
769        (pcretest), all targets were linked against these libraries. This included
770        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
771        libraries. This caused unwanted dependencies to be created. This problem
772        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
773        pcretest is linked with readline.
774    
775    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
776        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
777        moved outside it again, because FALSE and TRUE are already defined in AIX,
778        but BOOL is not.
779    
780    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
781        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
782    
783    4.  The pcregrep documentation said spaces were inserted as well as colons (or
784        hyphens) following file names and line numbers when outputting matching
785        lines. This is not true; no spaces are inserted. I have also clarified the
786        wording for the --colour (or --color) option.
787    
788    5.  In pcregrep, when --colour was used with -o, the list of matching strings
789        was not coloured; this is different to GNU grep, so I have changed it to be
790        the same.
791    
792    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
793        each matching line was coloured. Now it goes on to look for further matches
794        of any of the test patterns, which is the same behaviour as GNU grep.
795    
796    7.  A pattern that could match an empty string could cause pcregrep to loop; it
797        doesn't make sense to accept an empty string match in pcregrep, so I have
798        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
799        seems to be how GNU grep behaves.
800    
801    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
802        start or after a newline", because the conditional assertion was not being
803        correctly handled. The rule now is that both the assertion and what follows
804        in the first alternative must satisfy the test.
805    
806    9.  If auto-callout was enabled in a pattern with a conditional group whose
807        condition was an assertion, PCRE could crash during matching, both with
808        pcre_exec() and pcre_dfa_exec().
809    
810    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
811        used for matching.
812    
813    11. Unicode property support in character classes was not working for
814        characters (bytes) greater than 127 when not in UTF-8 mode.
815    
816    12. Added the -M command line option to pcretest.
817    
818    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
819    
820    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
821    
822    16. Added comments and documentation about mis-use of no_arg in the C++
823        wrapper.
824    
825    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
826        from Martin Jerabek that uses macro names for all relevant character and
827        string constants.
828    
829    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
830        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
831        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
832        these, but not everybody uses configure.
833    
834    19. A conditional group that had only one branch was not being correctly
835        recognized as an item that could match an empty string. This meant that an
836        enclosing group might also not be so recognized, causing infinite looping
837        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
838        with the subject "ab", where knowledge that the repeated group can match
839        nothing is needed in order to break the loop.
840    
841    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
842        exec(), but without supplying a callout function, matching went wrong.
843    
844    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
845        leak if the size of the offset vector was greater than 30. When the vector
846        is smaller, the saved offsets during recursion go onto a local stack
847        vector, but for larger vectors malloc() is used. It was failing to free
848        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
849        error, in fact).
850    
851    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
852        heapframe that is used only when UTF-8 support is enabled. This caused no
853        problem, but was untidy.
854    
855    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
856        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
857        included within another project.
858    
859    24. Steven Van Ingelgem's patches to add more options to the CMake support,
860        slightly modified by me:
861    
862          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
863              not building pcregrep.
864    
865          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
866              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
867    
868    25. Forward references, both numeric and by name, in patterns that made use of
869        duplicate group numbers, could behave incorrectly or give incorrect errors,
870        because when scanning forward to find the reference group, PCRE was not
871        taking into account the duplicate group numbers. A pattern such as
872        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
873    
874    26. Changed a few more instances of "const unsigned char *" to USPTR, making
875        the feature of a custom pointer more persuasive (as requested by a user).
876    
877    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
878        pcretest.c, inside #ifndefs, because it seems they are sometimes already
879        pre-defined.
880    
881    28. Added support for (*UTF8) at the start of a pattern.
882    
883    29. Arrange for flags added by the "release type" setting in CMake to be shown
884        in the configuration summary.
885    
886    
887    Version 7.8 05-Sep-08
888    ---------------------
889    
890    1.  Replaced UCP searching code with optimized version as implemented for Ad
891        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
892        stage table and inline lookup instead of a function, giving speed ups of 2
893        to 5 times on some simple patterns that I tested. Permission was given to
894        distribute the MultiStage2.py script that generates the tables (it's not in
895        the tarball, but is in the Subversion repository).
896    
897    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
898        scripts.
899    
900    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
901        a group with a zero qualifier. The result of the study could be incorrect,
902        or the function might crash, depending on the pattern.
903    
904    4.  Caseless matching was not working for non-ASCII characters in back
905        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
906        It now works when Unicode Property Support is available.
907    
908    5.  In pcretest, an escape such as \x{de} in the data was always generating
909        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
910        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
911        truncation.
912    
913    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
914    
915    7.  Added two (int) casts to pcregrep when printing the difference of two
916        pointers, in case they are 64-bit values.
917    
918    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
919        test 2 if it fails.
920    
921    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
922        and a #define of that name to empty if it is not externally set. This is to
923        allow users of MSVC to set it if necessary.
924    
925    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
926        the convenience functions in the pcre_get.c source file.
927    
928    11. An option change at the start of a pattern that had top-level alternatives
929        could cause overwriting and/or a crash. This command provoked a crash in
930        some environments:
931    
932          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
933    
934        This potential security problem was recorded as CVE-2008-2371.
935    
936    12. For a pattern where the match had to start at the beginning or immediately
937        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
938        pcre_dfa_exec() could read past the end of the passed subject if there was
939        no match. To help with detecting such bugs (e.g. with valgrind), I modified
940        pcretest so that it places the subject at the end of its malloc-ed buffer.
941    
942    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
943        exec() might read past the end of the data buffer in UTF-8 mode.
944    
945    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
946        the data contained the byte 0x85 as part of a UTF-8 character within its
947        first line. This applied both to normal and DFA matching.
948    
949    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
950        /^[^d]*?$/8 failed to match "abc".
951    
952    16. Added a missing copyright notice to pcrecpp_internal.h.
953    
954    17. Make it more clear in the documentation that values returned from
955        pcre_exec() in ovector are byte offsets, not character counts.
956    
957    18. Tidied a few places to stop certain compilers from issuing warnings.
958    
959    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
960        supplied by Stefan Weber. I made a further small update for 7.8 because
961        there is a change of source arrangements: the pcre_searchfuncs.c module is
962        replaced by pcre_ucd.c.
963    
964    
965    Version 7.7 07-May-08
966    ---------------------
967    
968    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
969        a string to a long long, pretend we don't even have a long long." This is
970        done by checking for the strtoq, strtoll, and _strtoi64 functions.
971    
972    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
973        pre-7.6 versions, which defined a global no_arg variable instead of putting
974        it in the RE class. (See also #8 below.)
975    
976    3.  Remove a line of dead code, identified by coverity and reported by Nuno
977        Lopes.
978    
979    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
980    
981        (1) The include/exclude patterns were being applied to the whole pathnames
982            of files, instead of just to the final components.
983    
984        (2) If there was more than one level of directory, the subdirectories were
985            skipped unless they satisfied the include/exclude conditions. This is
986            inconsistent with GNU grep (and could even be seen as contrary to the
987            pcregrep specification - which I improved to make it absolutely clear).
988            The action now is always to scan all levels of directory, and just
989            apply the include/exclude patterns to regular files.
990    
991    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
992        --exclude_dir in the tests to avoid scanning .svn directories.
993    
994    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
995        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
996        doesn't support NULs in patterns.
997    
998    7.  Added some missing "const"s to declarations of static tables in
999        pcre_compile.c and pcre_dfa_exec.c.
1000    
1001    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1002        caused by fix #2  above. (Subsequently also a second patch to fix the
1003        first patch. And a third patch - this was a messy problem.)
1004    
1005    9.  Applied Craig's patch to remove the use of push_back().
1006    
1007    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1008        matching function regexec().
1009    
1010    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1011        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1012        references. PCRE supports relative numbers with this syntax (I don't think
1013        Oniguruma does).
1014    
1015    12. Previously, a group with a zero repeat such as (...){0} was completely
1016        omitted from the compiled regex. However, this means that if the group
1017        was called as a subroutine from elsewhere in the pattern, things went wrong
1018        (an internal error was given). Such groups are now left in the compiled
1019        pattern, with a new opcode that causes them to be skipped at execution
1020        time.
1021    
1022    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1023        to the way PCRE behaves:
1024    
1025        (a) A lone ] character is dis-allowed (Perl treats it as data).
1026    
1027        (b) A back reference to an unmatched subpattern matches an empty string
1028            (Perl fails the current match path).
1029    
1030        (c) A data ] in a character class must be notated as \] because if the
1031            first data character in a class is ], it defines an empty class. (In
1032            Perl it is not possible to have an empty class.) The empty class []
1033            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1034            The negative empty class [^] matches any one character, independently
1035            of the DOTALL setting.
1036    
1037    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1038        non-existent subpattern following a character class starting with ']' and
1039        containing () gave an internal compiling error instead of "reference to
1040        non-existent subpattern". Fortunately, when the pattern did exist, the
1041        compiled code was correct. (When scanning forwards to check for the
1042        existencd of the subpattern, it was treating the data ']' as terminating
1043        the class, so got the count wrong. When actually compiling, the reference
1044        was subsequently set up correctly.)
1045    
1046    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1047        it was being rejected as not supported by pcre_dfa_exec(), even though
1048        other assertions are supported. I have made pcre_dfa_exec() support
1049        (*FAIL).
1050    
1051    16. The implementation of 13c above involved the invention of a new opcode,
1052        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1053        cannot be changed at match time, I realized I could make a small
1054        improvement to matching performance by compiling OP_ALLANY instead of
1055        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1056        on the OP_ANY path.
1057    
1058    17. Compiling pcretest on Windows with readline support failed without the
1059        following two fixes: (1) Make the unistd.h include conditional on
1060        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1061    
1062    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1063        ncurses library to be included for pcretest when ReadLine support is
1064        requested, but also to allow for it to be overridden. This patch came from
1065        Daniel Bergström.
1066    
1067    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1068        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1069        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1070        spotting this.
1071    
1072    
1073    Version 7.6 28-Jan-08
1074    ---------------------
1075    
1076    1.  A character class containing a very large number of characters with
1077        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1078        overflow.
1079    
1080    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1081        HAVE_LONG_LONG is not defined.
1082    
1083    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1084        bring them up to date and include new features. This patch includes:
1085    
1086        - Fixed PH's badly added libz and libbz2 support.
1087        - Fixed a problem with static linking.
1088        - Added pcredemo. [But later removed - see 7 below.]
1089        - Fixed dftables problem and added an option.
1090        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1091            HAVE_LONG_LONG.
1092        - Added readline support for pcretest.
1093        - Added an listing of the option settings after cmake has run.
1094    
1095    4.  A user submitted a patch to Makefile that makes it easy to create
1096        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1097        Makefile.am that cause it to include this special target, without
1098        affecting anything else. Note that the same mingw target plus all
1099        the other distribution libraries and programs are now supported
1100        when configuring with CMake (see 6 below) instead of with
1101        Configure/Make.
1102    
1103    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1104        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1105        exported in the Windows port". It has not yet been confirmed that the patch
1106        solves the problem, but it does no harm.
1107    
1108    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1109        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1110        with CMake, and also correct the comment about stack recursion.
1111    
1112    7.  Remove the automatic building of pcredemo from the ./configure system and
1113        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1114        of a program that users should build themselves after PCRE is installed, so
1115        building it automatically is not really right. What is more, it gave
1116        trouble in some build environments.
1117    
1118    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1119    
1120    
1121    Version 7.5 10-Jan-08
1122    ---------------------
1123    
1124    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1125        values in parens when parsing an RE using the C++ wrapper."
1126    
1127    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1128        Characters greater than 255 were excluded from the class instead of being
1129        included.
1130    
1131    3.  The same bug as (2) above applied to negated POSIX classes such as
1132        [:^space:].
1133    
1134    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1135        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1136        I have changed it.
1137    
1138    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1139        first named subpattern) and a construct such as (?&a) would reference the
1140        first named subpattern whose name started with "a" (in other words, the
1141        length check was missing). Both these problems are fixed. "Subpattern name
1142        expected" is now given for (?&) (a zero-length name), and this patch also
1143        makes it give the same error for \k'' (previously it complained that that
1144        was a reference to a non-existent subpattern).
1145    
1146    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1147        this is right because (?- can be followed by option settings as well as by
1148        digits. I have, however, made the messages clearer.
1149    
1150    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1151        than the number used in the conditional) now cause a compile-time error.
1152        This is actually not compatible with Perl, which accepts such patterns, but
1153        treats the conditional as always being FALSE (as PCRE used to), but it
1154        seems to me that giving a diagnostic is better.
1155    
1156    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1157        and messages.
1158    
1159    9.  Fix two occurrences of "backslash" in comments that should have been
1160        "backspace".
1161    
1162    10. Remove two redundant lines of code that can never be obeyed (their function
1163        was moved elsewhere).
1164    
1165    11. The program that makes PCRE's Unicode character property table had a bug
1166        which caused it to generate incorrect table entries for sequences of
1167        characters that have the same character type, but are in different scripts.
1168        It amalgamated them into a single range, with the script of the first of
1169        them. In other words, some characters were in the wrong script. There were
1170        thirteen such cases, affecting characters in the following ranges:
1171    
1172          U+002b0 - U+002c1
1173          U+0060c - U+0060d
1174          U+0061e - U+00612
1175          U+0064b - U+0065e
1176          U+0074d - U+0076d
1177          U+01800 - U+01805
1178          U+01d00 - U+01d77
1179          U+01d9b - U+01dbf
1180          U+0200b - U+0200f
1181          U+030fc - U+030fe
1182          U+03260 - U+0327f
1183          U+0fb46 - U+0fbb1
1184          U+10450 - U+1049d
1185    
1186    12. The -o option (show only the matching part of a line) for pcregrep was not
1187        compatible with GNU grep in that, if there was more than one match in a
1188        line, it showed only the first of them. It now behaves in the same way as
1189        GNU grep.
1190    
1191    13. If the -o and -v options were combined for pcregrep, it printed a blank
1192        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1193        does the same. The return code can be used to tell if there were any
1194        non-matching lines.
1195    
1196    14. Added --file-offsets and --line-offsets to pcregrep.
1197    
1198    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1199        infinitely looping recursion. The bug was that positive lookaheads were not
1200        being skipped when checking for a possible empty match (negative lookaheads
1201        and both kinds of lookbehind were skipped).
1202    
1203    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1204        inclusion of <windows.h> to before rather than after the definition of
1205        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1206    
1207    17. Specifying a possessive quantifier with a specific limit for a Unicode
1208        character property caused pcre_compile() to compile bad code, which led at
1209        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1210        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1211        caused the error; without that there was no problem.
1212    
1213    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1214    
1215    19. Added --enable-pcretest-libreadline.
1216    
1217    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1218        RE::GlobalReplace(). As a result, the number of replacements returned was
1219        double what it should be. I removed one of the increments, but Craig sent a
1220        later patch that removed the other one (the right fix) and added unit tests
1221        that check the return values (which was not done before).
1222    
1223    21. Several CMake things:
1224    
1225        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1226            the names libpcre and libpcreposix, not just pcre and pcreposix.
1227    
1228        (2) The above change means that pcretest and pcregrep are now correctly
1229            linked with the newly-built libraries, not previously installed ones.
1230    
1231        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1232    
1233    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1234        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1235        UTF-8 newline character). The key issue is that the pattern starts .*;
1236        this means that the match must be either at the beginning, or after a
1237        newline. The bug was in the code for advancing after a failed match and
1238        checking that the new position followed a newline. It was not taking
1239        account of UTF-8 characters correctly.
1240    
1241    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1242        character classes. PCRE was not treating the sequence [:...:] as a
1243        character class unless the ... were all letters. Perl, however, seems to
1244        allow any characters between [: and :], though of course it rejects as
1245        unknown any "names" that contain non-letters, because all the known class
1246        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1247        for example, whereas PCRE did not - it did not recognize a POSIX character
1248        class. This seemed a bit dangerous, so the code has been changed to be
1249        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1250        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1251        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1252        Perl does, and where it didn't before.
1253    
1254    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1255        Windows environments %n is disabled by default.
1256    
1257    
1258    Version 7.4 21-Sep-07
1259    ---------------------
1260    
1261    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1262        means that a class such as [\s] counted as "explicit reference to CR or
1263        LF". That isn't really right - the whole point of the change was to try to
1264        help when there was an actual mention of one of the two characters. So now
1265        the change happens only if \r or \n (or a literal CR or LF) character is
1266        encountered.
1267    
1268    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1269        of both had grown to the point where there were only 3 bits left.
1270        Fortunately, there was spare space in the data structure, and so I have
1271        moved the internal flags into a new 16-bit field to free up more option
1272        bits.
1273    
1274    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1275        but did not set the internal JCHANGED flag - either of these is enough to
1276        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1277        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1278        start sets both bits.
1279    
1280    4.  Added options (at build time, compile time, exec time) to change \R from
1281        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1282    
1283    5.  doc/pcresyntax.html was missing from the distribution.
1284    
1285    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1286        compatibility, even though it is no longer used.
1287    
1288    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1289        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1290        windows.h file is present (where different names are used). [This was
1291        reversed later after testing - see 16 below.]
1292    
1293    8.  Changed all #include <config.h> to #include "config.h". There were also
1294        some further <pcre.h> cases that I changed to "pcre.h".
1295    
1296    9.  When pcregrep was used with the --colour option, it missed the line ending
1297        sequence off the lines that it output.
1298    
1299    10. It was pointed out to me that arrays of string pointers cause lots of
1300        relocations when a shared library is dynamically loaded. A technique of
1301        using a single long string with a table of offsets can drastically reduce
1302        these. I have refactored PCRE in four places to do this. The result is
1303        dramatic:
1304    
1305          Originally:                          290
1306          After changing UCP table:            187
1307          After changing error message table:   43
1308          After changing table of "verbs"       36
1309          After changing table of Posix names   22
1310    
1311        Thanks to the folks working on Gregex for glib for this insight.
1312    
1313    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1314        unicode-properties was also set.
1315    
1316    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1317    
1318    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1319        checked only for CRLF.
1320    
1321    14. Added casts to pcretest.c to avoid compiler warnings.
1322    
1323    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1324    
1325    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1326        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1327        entirely. This removes changes made in 7 above.
1328    
1329    17. The CMake files have been updated, and there is now more information about
1330        building with CMake in the NON-UNIX-USE document.
1331    
1332    
1333    Version 7.3 28-Aug-07
1334    ---------------------
1335    
1336     1. In the rejigging of the build system that eventually resulted in 7.1, the
1337        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1338        brackets there is not right, since it causes compilers to look for an
1339        installed pcre.h, not the version that is in the source that is being
1340        compiled (which of course may be different). I have changed it back to:
1341    
1342          #include "pcre.h"
1343    
1344        I have a vague recollection that the change was concerned with compiling in
1345        different directories, but in the new build system, that is taken care of
1346        by the VPATH setting the Makefile.
1347    
1348     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1349        when the subject happened to end in the byte 0x85 (e.g. if the last
1350        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1351        characters but of course it shouldn't be taken as a newline when it is part
1352        of another character. The bug was that, for an unlimited repeat of . in
1353        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1354        characters when looking for a newline.
1355    
1356     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1357    
1358     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1359        in debug output.
1360    
1361     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1362        long printing in the pcrecpp unittest when running under MinGW.
1363    
1364     6. ESC_K was left out of the EBCDIC table.
1365    
1366     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1367        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1368        limit also applies to "virtual nesting" when a pattern is recursive, and in
1369        this case 1000 isn't so big. I have been able to remove this limit at the
1370        expense of backing off one optimization in certain circumstances. Normally,
1371        when pcre_exec() would call its internal match() function recursively and
1372        immediately return the result unconditionally, it uses a "tail recursion"
1373        feature to save stack. However, when a subpattern that can match an empty
1374        string has an unlimited repetition quantifier, it no longer makes this
1375        optimization. That gives it a stack frame in which to save the data for
1376        checking that an empty string has been matched. Previously this was taken
1377        from the 1000-entry workspace that had been reserved. So now there is no
1378        explicit limit, but more stack is used.
1379    
1380     8. Applied Daniel's patches to solve problems with the import/export magic
1381        syntax that is required for Windows, and which was going wrong for the
1382        pcreposix and pcrecpp parts of the library. These were overlooked when this
1383        problem was solved for the main library.
1384    
1385     9. There were some crude static tests to avoid integer overflow when computing
1386        the size of patterns that contain repeated groups with explicit upper
1387        limits. As the maximum quantifier is 65535, the maximum group length was
1388        set at 30,000 so that the product of these two numbers did not overflow a
1389        32-bit integer. However, it turns out that people want to use groups that
1390        are longer than 30,000 bytes (though not repeat them that many times).
1391        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1392        made it possible to implement the integer overflow checks in a much more
1393        dynamic way, which I have now done. The artificial limitation on group
1394        length has been removed - we now have only the limit on the total length of
1395        the compiled pattern, which depends on the LINK_SIZE setting.
1396    
1397    10. Fixed a bug in the documentation for get/copy named substring when
1398        duplicate names are permitted. If none of the named substrings are set, the
1399        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1400        empty string.
1401    
1402    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1403        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1404        because the ] is interpreted as the first data character and the
1405        terminating ] is not found. PCRE has been made compatible with Perl in this
1406        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1407        cause memory overwriting.
1408    
1409    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1410        string has been matched (to stop an infinite loop). It was not recognizing
1411        a conditional subpattern that could match an empty string if that
1412        subpattern was within another subpattern. For example, it looped when
1413        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1414        condition was not nested. This bug has been fixed.
1415    
1416    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1417        past the start of the subject in the presence of bytes with the top bit
1418        set, for example "\x8aBCD".
1419    
1420    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1421        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1422    
1423    14. Optimized (?!) to (*FAIL).
1424    
1425    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1426        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1427        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1428        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1429        does: it's just the validity check that is more restrictive.
1430    
1431    16. Inserted checks for integer overflows during escape sequence (backslash)
1432        processing, and also fixed erroneous offset values for syntax errors during
1433        backslash processing.
1434    
1435    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1436        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1437    
1438    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1439        caused an overrun.
1440    
1441    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1442        something other than just ASCII characters) inside a group that had an
1443        unlimited repeat caused a loop at compile time (while checking to see
1444        whether the group could match an empty string).
1445    
1446    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1447        [\P{Any}] did so. (Error in the code for printing property names.)
1448    
1449    21. An orphan \E inside a character class could cause a crash.
1450    
1451    22. A repeated capturing bracket such as (A)? could cause a wild memory
1452        reference during compilation.
1453    
1454    23. There are several functions in pcre_compile() that scan along a compiled
1455        expression for various reasons (e.g. to see if it's fixed length for look
1456        behind). There were bugs in these functions when a repeated \p or \P was
1457        present in the pattern. These operators have additional parameters compared
1458        with \d, etc, and these were not being taken into account when moving along
1459        the compiled data. Specifically:
1460    
1461        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1462            length.
1463    
1464        (b) An item such as \pL+ within a repeated group could cause crashes or
1465            loops.
1466    
1467        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1468            "reference to non-existent subpattern" error.
1469    
1470        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1471    
1472    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1473        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1474    
1475    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1476    
1477    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1478        character were causing crashes (broken optimization).
1479    
1480    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1481        \p or \P) caused a compile-time loop.
1482    
1483    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1484        break. For example, the unstudied pattern [\r\n]A does not match the string
1485        "\r\nA" because change 7.0/46 below moves the current point on by two
1486        characters after failing to match at the start. However, the pattern \nA
1487        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1488        the same is true. There doesn't seem any very clean way out of this, but
1489        what I have chosen to do makes the common cases work: PCRE now takes note
1490        of whether there can be an explicit match for \r or \n anywhere in the
1491        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1492        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1493        pattern has explicit CR or LF references.
1494    
1495    29. Added (*CR) etc for changing newline setting at start of pattern.
1496    
1497    
1498    Version 7.2 19-Jun-07
1499    ---------------------
1500    
1501     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1502        which is apparently normally available under Windows.
1503    
1504     2. Re-jig the pcregrep tests with different newline settings in an attempt
1505        to make them independent of the local environment's newline setting.
1506    
1507     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1508    
1509     4. Some of the "internals" tests were previously cut out when the link size
1510        was not 2, because the output contained actual offsets. The recent new
1511        "Z" feature of pcretest means that these can be cut out, making the tests
1512        usable with all link sizes.
1513    
1514     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1515        stack recursion. This gives a massive performance boost under BSD, but just
1516        a small improvement under Linux. However, it saves one field in the frame
1517        in all cases.
1518    
1519     6. Added more features from the forthcoming Perl 5.10:
1520    
1521        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1522            recursion call. It refers to the nth most recently opened parentheses.
1523    
1524        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1525            to be opened parentheses.
1526    
1527        (c) Conditions that refer to capturing parentheses can be specified
1528            relatively, for example, (?(-2)... or (?(+3)...
1529    
1530        (d) \K resets the start of the current match so that everything before
1531            is not part of it.
1532    
1533        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1534    
1535        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1536            reference syntax.
1537    
1538        (g) (?| introduces a group in which the numbering of parentheses in each
1539            alternative starts with the same number.
1540    
1541        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1542    
1543     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1544        PCRE_INFO_JCHANGED.
1545    
1546     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1547        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1548        for detecting groups that can match an empty string.
1549    
1550     9. A pattern with a very large number of alternatives (more than several
1551        hundred) was running out of internal workspace during the pre-compile
1552        phase, where pcre_compile() figures out how much memory will be needed. A
1553        bit of new cunning has reduced the workspace needed for groups with
1554        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1555        workspace instead of running out of the 4096 that are available.
1556    
1557    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1558    
1559    11. Applied patch from Google to remove an optimization that didn't quite work.
1560        The report of the bug said:
1561    
1562          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1563          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1564          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1565    
1566    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1567        it matched the wrong number of bytes.
1568    
1569    
1570    Version 7.1 24-Apr-07
1571    ---------------------
1572    
1573     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1574        that is more "standard", making use of automake and other Autotools. There
1575        is some re-arrangement of the files and adjustment of comments consequent
1576        on this.
1577    
1578     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1579        for recursive directory scanning broke on some systems because the files
1580        are not scanned in any specific order and on different systems the order
1581        was different. A call to "sort" has been inserted into RunGrepTest for the
1582        approprate test as a short-term fix. In the longer term there may be an
1583        alternative.
1584    
1585     3. I had an email from Eric Raymond about problems translating some of PCRE's
1586        man pages to HTML (despite the fact that I distribute HTML pages, some
1587        people do their own conversions for various reasons). The problems
1588        concerned the use of low-level troff macros .br and .in. I have therefore
1589        removed all such uses from the man pages (some were redundant, some could
1590        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1591        HTML has been updated to handle .nf/.fi and to complain if it encounters
1592        .br or .in.
1593    
1594     4. Updated comments in configure.ac that get placed in config.h.in and also
1595        arranged for config.h to be included in the distribution, with the name
1596        config.h.generic, for the benefit of those who have to compile without
1597        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1598    
1599     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1600        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1601        makevp.bat for the current PCRE, using the additional files
1602        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1603    
1604     6. A Windows user reported a minor discrepancy with test 2, which turned out
1605        to be caused by a trailing space on an input line that had got lost in his
1606        copy. The trailing space was an accident, so I've just removed it.
1607    
1608     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1609        that is needed.
1610    
1611     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1612        as "const" (a) because they are and (b) because it helps the PHP
1613        maintainers who have recently made a script to detect big data structures
1614        in the php code that should be moved to the .rodata section. I remembered
1615        to update Builducptable as well, so it won't revert if ucptable.h is ever
1616        re-created.
1617    
1618     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1619        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1620        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1621        support is not required. This saves 1.5-2K of code, which is important in
1622        some applications.
1623    
1624        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1625        so as not to refer to the tables, even though these functions will never be
1626        called when UTF-8 support is disabled. Otherwise there are problems with a
1627        shared library.
1628    
1629    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1630    
1631        (a) It was defining its arguments as char * instead of void *.
1632    
1633        (b) It was assuming that all moves were upwards in memory; this was true
1634            a long time ago when I wrote it, but is no longer the case.
1635    
1636        The emulated memove() is provided for those environments that have neither
1637        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1638        is clearly not the case, as these two bugs were recently reported.
1639    
1640    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1641        and Detrail to create the HTML documentation, the .txt form of the man
1642        pages, and it removes trailing spaces from listed files. It also creates
1643        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1644        case, it wraps all the #defines with #ifndefs. This script should be run
1645        before "make dist".
1646    
1647    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1648        with Unicode property support.
1649    
1650        (a) For a maximizing quantifier, if the two different cases of the
1651            character were of different lengths in their UTF-8 codings (there are
1652            some cases like this - I found 11), and the matching function had to
1653            back up over a mixture of the two cases, it incorrectly assumed they
1654            were both the same length.
1655    
1656        (b) When PCRE was configured to use the heap rather than the stack for
1657            recursion during matching, it was not correctly preserving the data for
1658            the other case of a UTF-8 character when checking ahead for a match
1659            while processing a minimizing repeat. If the check also involved
1660            matching a wide character, but failed, corruption could cause an
1661            erroneous result when trying to check for a repeat of the original
1662            character.
1663    
1664    13. Some tidying changes to the testing mechanism:
1665    
1666        (a) The RunTest script now detects the internal link size and whether there
1667            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1668            values substituted by "configure". (The RunGrepTest script already did
1669            this for UTF-8.) The configure.ac script no longer substitutes the
1670            relevant variables.
1671    
1672        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1673            with length and offset values. This means that the output is different
1674            for different internal link sizes. Test 2 is skipped for link sizes
1675            other than 2 because of this, bypassing the problem. Unfortunately,
1676            there was also a test in test 3 (the locale tests) that used /B and
1677            failed for link sizes other than 2. Rather than cut the whole test out,
1678            I have added a new /Z option to pcretest that replaces the length and
1679            offset values with spaces. This is now used to make test 3 independent
1680            of link size. (Test 2 will be tidied up later.)
1681    
1682    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1683        segmentation fault instead of returning the appropriate error message.
1684    
1685    15. In multiline mode when the newline sequence was set to "any", the pattern
1686        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1687        This doesn't seem right; it now treats the CRLF combination as the line
1688        ending, and so does not match in that case. It's only a pattern such as ^$
1689        that would hit this one: something like ^ABC$ would have failed after \r
1690        and then tried again after \r\n.
1691    
1692    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1693        in an attempt to make files that differ only in their line terminators
1694        compare equal. This works on Linux.
1695    
1696    17. Under certain error circumstances pcregrep might try to free random memory
1697        as it exited. This is now fixed, thanks to valgrind.
1698    
1699    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1700        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1701        was because its rules for how to advance for /g after matching an empty
1702        string at the end of a line did not allow for this case. They now check for
1703        it specially.
1704    
1705    20. pcretest is supposed to handle patterns and data of any length, by
1706        extending its buffers when necessary. It was getting this wrong when the
1707        buffer for a data line had to be extended.
1708    
1709    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1710        CRLF as a newline sequence.
1711    
1712    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1713        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1714        I have nevertheless tidied it up.
1715    
1716    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1717    
1718    24. Added a man page for pcre-config.
1719    
1720    
1721    Version 7.0 19-Dec-06
1722    ---------------------
1723    
1724     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1725        moving to gcc 4.1.1.
1726    
1727     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1728        sys/time.h, which is documented as needed for this function. It doesn't
1729        seem to matter on Linux, but it showed up on some releases of OS X.
1730    
1731     3. It seems that there are systems where bytes whose values are greater than
1732        127 match isprint() in the "C" locale. The "C" locale should be the
1733        default when a C program starts up. In most systems, only ASCII printing
1734        characters match isprint(). This difference caused the output from pcretest
1735        to vary, making some of the tests fail. I have changed pcretest so that:
1736    
1737        (a) When it is outputting text in the compiled version of a pattern, bytes
1738            other than 32-126 are always shown as hex escapes.
1739    
1740        (b) When it is outputting text that is a matched part of a subject string,
1741            it does the same, unless a different locale has been set for the match
1742            (using the /L modifier). In this case, it uses isprint() to decide.
1743    
1744     4. Fixed a major bug that caused incorrect computation of the amount of memory
1745        required for a compiled pattern when options that changed within the
1746        pattern affected the logic of the preliminary scan that determines the
1747        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1748        that the computed length was too small. The symptoms of this bug were
1749        either the PCRE error "internal error: code overflow" from pcre_compile(),
1750        or a glibc crash with a message such as "pcretest: free(): invalid next
1751        size (fast)". Examples of patterns that provoked this bug (shown in
1752        pcretest format) are:
1753    
1754          /(?-x: )/x
1755          /(?x)(?-x: \s*#\s*)/
1756          /((?i)[\x{c0}])/8
1757          /(?i:[\x{c0}])/8
1758    
1759        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1760        is now done differently.
1761    
1762     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1763        wrapper classes; (b) implement a new function in the C++ scanner that is
1764        more efficient than the old way of doing things because it avoids levels of
1765        recursion in the regex matching; (c) add a paragraph to the documentation
1766        for the FullMatch() function.
1767    
1768     6. The escape sequence \n was being treated as whatever was defined as
1769        "newline". Not only was this contrary to the documentation, which states
1770        that \n is character 10 (hex 0A), but it also went horribly wrong when
1771        "newline" was defined as CRLF. This has been fixed.
1772    
1773     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1774        was being set to -1 for the "end of line" case (supposedly a value that no
1775        character can have). Though this value is never used (the check for end of
1776        line is "zero bytes in current character"), it caused compiler complaints.
1777        I've changed it to 0xffffffff.
1778    
1779     8. In pcre_version.c, the version string was being built by a sequence of
1780        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1781        string (as it is for production releases) called a macro with an empty
1782        argument. The C standard says the result of this is undefined. The gcc
1783        compiler treats it as an empty string (which was what was wanted) but it is
1784        reported that Visual C gives an error. The source has been hacked around to
1785        avoid this problem.
1786    
1787     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1788        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1789        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1790        of them did).
1791    
1792    10. Originally, pcretest opened its input and output without "b"; then I was
1793        told that "b" was needed in some environments, so it was added for release
1794        5.0 to both the input and output. (It makes no difference on Unix-like
1795        systems.) Later I was told that it is wrong for the input on Windows. I've
1796        now abstracted the modes into two macros, to make it easier to fiddle with
1797        them, and removed "b" from the input mode under Windows.
1798    
1799    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1800    
1801    12. Added -help and --help to pcretest as an official way of being reminded
1802        of the options.
1803    
1804    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1805        and pcrecpp.cc because they annoy compilers at high warning levels.
1806    
1807    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1808    
1809    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1810        scripts are not C programs :-) and which was not noticed because it works
1811        on Linux.
1812    
1813    16. pcretest is supposed to handle any length of pattern and data line (as one
1814        line or as a continued sequence of lines) by extending its input buffer if
1815        necessary. This feature was broken for very long pattern lines, leading to
1816        a string of junk being passed to pcre_compile() if the pattern was longer
1817        than about 50K.
1818    
1819    17. I have done a major re-factoring of the way pcre_compile() computes the
1820        amount of memory needed for a compiled pattern. Previously, there was code
1821        that made a preliminary scan of the pattern in order to do this. That was
1822        OK when PCRE was new, but as the facilities have expanded, it has become
1823        harder and harder to keep it in step with the real compile phase, and there
1824        have been a number of bugs (see for example, 4 above). I have now found a
1825        cunning way of running the real compile function in a "fake" mode that
1826        enables it to compute how much memory it would need, while actually only
1827        ever using a few hundred bytes of working memory and without too many
1828        tests of the mode. This should make future maintenance and development
1829        easier. A side effect of this work is that the limit of 200 on the nesting
1830        depth of parentheses has been removed (though this was never a serious
1831        limitation, I suspect). However, there is a downside: pcre_compile() now
1832        runs more slowly than before (30% or more, depending on the pattern). I
1833        hope this isn't a big issue. There is no effect on runtime performance.
1834    
1835    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1836        newline (only possible for the last line of a file) and it was a
1837        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1838    
1839    19. Added additional timing features to pcretest. (1) The -tm option now times
1840        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1841        separate command line item, by a number that specifies the number of
1842        repeats to use when timing. The default is 50000; this gives better
1843        precision, but takes uncomfortably long for very large patterns.
1844    
1845    20. Extended pcre_study() to be more clever in cases where a branch of a
1846        subpattern has no definite first character. For example, (a*|b*)[cd] would
1847        previously give no result from pcre_study(). Now it recognizes that the
1848        first character must be a, b, c, or d.
1849    
1850    21. There was an incorrect error "recursive call could loop indefinitely" if
1851        a subpattern (or the entire pattern) that was being tested for matching an
1852        empty string contained only one non-empty item after a nested subpattern.
1853        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1854        incorrectly, because the \d was being skipped in the check.
1855    
1856    22. The pcretest program now has a new pattern option /B and a command line
1857        option -b, which is equivalent to adding /B to every pattern. This causes
1858        it to show the compiled bytecode, without the additional information that
1859        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1860        is the same as /B/I).
1861    
1862    23. A new optimization is now able automatically to treat some sequences such
1863        as a*b as a*+b. More specifically, if something simple (such as a character
1864        or a simple class like \d) has an unlimited quantifier, and is followed by
1865        something that cannot possibly match the quantified thing, the quantifier
1866        is automatically "possessified".
1867    
1868    24. A recursive reference to a subpattern whose number was greater than 39
1869        went wrong under certain circumstances in UTF-8 mode. This bug could also
1870        have affected the operation of pcre_study().
1871    
1872    25. Realized that a little bit of performance could be had by replacing
1873        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1874    
1875    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1876    
1877    27. Possessive quantifiers such as a++ were previously implemented by turning
1878        them into atomic groups such as ($>a+). Now they have their own opcodes,
1879        which improves performance. This includes the automatically created ones
1880        from 23 above.
1881    
1882    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1883        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1884        the first matched character to be a colon. This applied both to named and
1885        numbered groups.
1886    
1887    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1888    
1889    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1890        building PCRE on Windows easier, so I have included it in the distribution.
1891    
1892    31. There is now a check in pcretest against a ridiculously large number being
1893        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1894        loop, the loop is abandoned.
1895    
1896    32. Forward references to subpatterns in conditions such as (?(2)...) where
1897        subpattern 2 is defined later cause pcre_compile() to search forwards in
1898        the pattern for the relevant set of parentheses. This search went wrong
1899        when there were unescaped parentheses in a character class, parentheses
1900        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1901    
1902    33. "Subroutine" calls and backreferences were previously restricted to
1903        referencing subpatterns earlier in the regex. This restriction has now
1904        been removed.
1905    
1906    34. Added a number of extra features that are going to be in Perl 5.10. On the
1907        whole, these are just syntactic alternatives for features that PCRE had
1908        previously implemented using the Python syntax or my own invention. The
1909        other formats are all retained for compatibility.
1910    
1911        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1912            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1913            also .NET compatible.
1914    
1915        (b) A recursion or subroutine call to a named group can now be defined as
1916            (?&name) as well as (?P>name).
1917    
1918        (c) A backreference to a named group can now be defined as \k<name> or
1919            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1920            5.10, are also .NET compatible.
1921    
1922        (d) A conditional reference to a named group can now use the syntax
1923            (?(<name>) or (?('name') as well as (?(name).
1924    
1925        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1926            groups (named and numbered) that are never evaluated inline, but can be
1927            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1928            is always false. There may be only one alternative in such a group.
1929    
1930        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1931            as the simple (?(R). The condition is true only if the most recent
1932            recursion is that of the given number or name. It does not search out
1933            through the entire recursion stack.
1934    
1935        (g) The escape \gN or \g{N} has been added, where N is a positive or
1936            negative number, specifying an absolute or relative reference.
1937    
1938    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1939        some "unreachable code" warnings.
1940    
1941    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1942        things, this adds five new scripts.
1943    
1944    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1945        There were also incompatibilities regarding the handling of \Q..\E inside
1946        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1947        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1948    
1949    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1950        matches an empty string, and forcibly breaks the loop. There were bugs in
1951        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1952        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1953        separate and independent bugs (that affected different cases) have been
1954        fixed.
1955    
1956    39. Refactored the code to abolish the use of different opcodes for small
1957        capturing bracket numbers. This is a tidy that I avoided doing when I
1958        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1959        The new approach is not only tidier, it makes it possible to reduce the
1960        memory needed to fix the previous bug (38).
1961    
1962    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1963        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1964        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1965        mode.
1966    
1967    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1968        report.
1969    
1970    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1971        copy construction and assignment in the C++ wrapper.
1972    
1973    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1974        couple of bugs that could have given wrong results in the "--newline=crlf"
1975        case.
1976    
1977    44. Added a number of casts and did some reorganization of signed/unsigned int
1978        variables following suggestions from Dair Grant. Also renamed the variable
1979        "this" as "item" because it is a C++ keyword.
1980    
1981    45. Arranged for dftables to add
1982    
1983          #include "pcre_internal.h"
1984    
1985        to pcre_chartables.c because without it, gcc 4.x may remove the array
1986        definition from the final binary if PCRE is built into a static library and
1987        dead code stripping is activated.
1988    
1989    46. For an unanchored pattern, if a match attempt fails at the start of a
1990        newline sequence, and the newline setting is CRLF or ANY, and the next two
1991        characters are CRLF, advance by two characters instead of one.
1992    
1993    
1994    Version 6.7 04-Jul-06
1995    ---------------------
1996    
1997     1. In order to handle tests when input lines are enormously long, pcretest has
1998        been re-factored so that it automatically extends its buffers when
1999        necessary. The code is crude, but this _is_ just a test program. The
2000        default size has been increased from 32K to 50K.
2001    
2002     2. The code in pcre_study() was using the value of the re argument before
2003        testing it for NULL. (Of course, in any sensible call of the function, it
2004        won't be NULL.)
2005    
2006     3. The memmove() emulation function in pcre_internal.h, which is used on
2007        systems that lack both memmove() and bcopy() - that is, hardly ever -
2008        was missing a "static" storage class specifier.
2009    
2010     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2011        containing an extended class (one that cannot be represented by a bitmap
2012        because it contains high-valued characters or Unicode property items, e.g.
2013        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2014        pattern, but PCRE should not loop if you do not (it no longer does).
2015        [Detail: two cases were found: (a) a repeated subpattern containing an
2016        extended class; (b) a recursive reference to a subpattern that followed a
2017        previous extended class. It wasn't skipping over the extended class
2018        correctly when UTF-8 mode was not set.]
2019    
2020     5. A negated single-character class was not being recognized as fixed-length
2021        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2022        compile error "lookbehind assertion is not fixed length".
2023    
2024     6. The RunPerlTest auxiliary script was showing an unexpected difference
2025        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2026        write a Perl script that can interpret lines of an input file either as
2027        byte characters or as UTF-8, which is what "perltest" was being required to
2028        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2029        can't do is switch easily at run time between having the "use utf8;" pragma
2030        or not. In the end, I fudged it by using the RunPerlTest script to insert
2031        "use utf8;" explicitly for the UTF-8 tests.
2032    
2033     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2034        the end of the subject string, contrary to the documentation and to what
2035        Perl does. This was true of both matching functions. Now it matches only at
2036        the start of the subject and immediately after *internal* newlines.
2037    
2038     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2039        a pointer to an int instead of a pointer to an unsigned long int. This
2040        caused problems on 64-bit systems.
2041    
2042     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2043        instance of the 'standard' template library not being so standard".
2044    
2045    10. There was no check on the number of named subpatterns nor the maximum
2046        length of a subpattern name. The product of these values is used to compute
2047        the size of the memory block for a compiled pattern. By supplying a very
2048        long subpattern name and a large number of named subpatterns, the size
2049        computation could be caused to overflow. This is now prevented by limiting
2050        the length of names to 32 characters, and the number of named subpatterns
2051        to 10,000.
2052    
2053    11. Subpatterns that are repeated with specific counts have to be replicated in
2054        the compiled pattern. The size of memory for this was computed from the
2055        length of the subpattern and the repeat count. The latter is limited to
2056        65535, but there was no limit on the former, meaning that integer overflow
2057        could in principle occur. The compiled length of a repeated subpattern is
2058        now limited to 30,000 bytes in order to prevent this.
2059    
2060    12. Added the optional facility to have named substrings with the same name.
2061    
2062    13. Added the ability to use a named substring as a condition, using the
2063        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2064        are numbers (not recommended). Forward references are permitted.
2065    
2066    14. Added forward references in named backreferences (if you see what I mean).
2067    
2068    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2069        pattern could run off the end of the subject. For example, the pattern
2070        "(?s)(.{1,5})"8 did this with the subject "ab".
2071    
2072    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2073        PCRE_CASELESS was set when matching characters that were quantified with ?
2074        or *.
2075    
2076    17. A character class other than a single negated character that had a minimum
2077        but no maximum quantifier - for example [ab]{6,} - was not handled
2078        correctly by pce_dfa_exec(). It would match only one character.
2079    
2080    18. A valid (though odd) pattern that looked like a POSIX character
2081        class but used an invalid character after [ (for example [[,abc,]]) caused
2082        pcre_compile() to give the error "Failed: internal error: code overflow" or
2083        in some cases to crash with a glibc free() error. This could even happen if
2084        the pattern terminated after [[ but there just happened to be a sequence of
2085        letters, a binary zero, and a closing ] in the memory that followed.
2086    
2087    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2088        over the years. Originally (before any Unicode support), just the bottom 8
2089        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2090        output from "man perlunicode" includes this:
2091    
2092          The regular expression compiler produces polymorphic opcodes.  That
2093          is, the pattern adapts to the data and automatically switches to
2094          the Unicode character scheme when presented with Unicode data--or
2095          instead uses a traditional byte scheme when presented with byte
2096          data.
2097    
2098        Sadly, a wide octal escape does not cause a switch, and in a string with
2099        no other multibyte characters, these octal escapes are treated as before.
2100        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2101        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2102        Unicode string.
2103    
2104        I have not perpetrated such confusion in PCRE. Up till now, it took just
2105        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2106        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2107        translate to the appropriate multibyte character.
2108    
2109    29. Applied some refactoring to reduce the number of warnings from Microsoft
2110        and Borland compilers. This has included removing the fudge introduced
2111        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2112        a warning about an unused variable.
2113    
2114    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2115        characters since release 4.0, because Perl (from release 5.004) does not.
2116        [Or at least, is documented not to: some releases seem to be in conflict
2117        with the documentation.] However, when a pattern was studied with
2118        pcre_study() and all its branches started with \s, PCRE still included VT
2119        as a possible starting character. Of course, this did no harm; it just
2120        caused an unnecessary match attempt.
2121    
2122    22. Removed a now-redundant internal flag bit that recorded the fact that case
2123        dependency changed within the pattern. This was once needed for "required
2124        byte" processing, but is no longer used. This recovers a now-scarce options
2125        bit. Also moved the least significant internal flag bit to the most-
2126        significant bit of the word, which was not previously used (hangover from
2127        the days when it was an int rather than a uint) to free up another bit for
2128        the future.
2129    
2130    23. Added support for CRLF line endings as well as CR and LF. As well as the
2131        default being selectable at build time, it can now be changed at runtime
2132        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2133        specify that it is scanning data with non-default line endings.
2134    
2135    24. Changed the definition of CXXLINK to make it agree with the definition of
2136        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2137    
2138    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2139        recursions. This makes a big different to stack usage for some patterns.
2140    
2141    26. If a subpattern containing a named recursion or subroutine reference such
2142        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2143        the space required for the compiled pattern went wrong and gave too small a
2144        value. Depending on the environment, this could lead to "Failed: internal
2145        error: code overflow at offset 49" or "glibc detected double free or
2146        corruption" errors.
2147    
2148    27. Applied patches from Google (a) to support the new newline modes and (b) to
2149        advance over multibyte UTF-8 characters in GlobalReplace.
2150    
2151    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2152        difference for some implementation of PCRE in some Windows version.
2153    
2154    29. Added some extra testing facilities to pcretest:
2155    
2156        \q<number>   in a data line sets the "match limit" value
2157        \Q<number>   in a data line sets the "match recursion limt" value
2158        -S <number>  sets the stack size, where <number> is in megabytes
2159    
2160        The -S option isn't available for Windows.
2161    
2162    
2163    Version 6.6 06-Feb-06
2164    ---------------------
2165    
2166     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2167        in pcreposix.h. I have copied the definition from pcre.h.
2168    
2169     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2170        because pcre.h is no longer a built file.
2171    
2172     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2173        not normally included in the compiled code.
2174    
2175    
2176    Version 6.5 01-Feb-06
2177    ---------------------
2178    
2179     1. When using the partial match feature with pcre_dfa_exec(), it was not
2180        anchoring the second and subsequent partial matches at the new starting
2181        point. This could lead to incorrect results. For example, with the pattern
2182        /1234/, partially matching against "123" and then "a4" gave a match.
2183    
2184     2. Changes to pcregrep:
2185    
2186        (a) All non-match returns from pcre_exec() were being treated as failures
2187            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2188            error message is output. Some extra information is given for the
2189            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2190            probably the only errors that are likely to be caused by users (by
2191            specifying a regex that has nested indefinite repeats, for instance).
2192            If there are more than 20 of these errors, pcregrep is abandoned.
2193    
2194        (b) A binary zero was treated as data while matching, but terminated the
2195            output line if it was written out. This has been fixed: binary zeroes
2196            are now no different to any other data bytes.
2197    
2198        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2199            used to set a locale for matching. The --locale=xxxx long option has
2200            been added (no short equivalent) to specify a locale explicitly on the
2201            pcregrep command, overriding the environment variables.
2202    
2203        (d) When -B was used with -n, some line numbers in the output were one less
2204            than they should have been.
2205    
2206        (e) Added the -o (--only-matching) option.
2207    
2208        (f) If -A or -C was used with -c (count only), some lines of context were
2209            accidentally printed for the final match.
2210    
2211        (g) Added the -H (--with-filename) option.
2212    
2213        (h) The combination of options -rh failed to suppress file names for files
2214            that were found from directory arguments.
2215    
2216        (i) Added the -D (--devices) and -d (--directories) options.
2217    
2218        (j) Added the -F (--fixed-strings) option.
2219    
2220        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2221    
2222        (l) Added the --colo(u)r option.
2223    
2224        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2225            is not present by default.
2226    
2227     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2228        items such as (?R) or (?1), when the recursion could match a number of
2229        alternatives. If it matched one of the alternatives, but subsequently,
2230        outside the recursion, there was a failure, the code tried to back up into
2231        the recursion. However, because of the way PCRE is implemented, this is not
2232        possible, and the result was an incorrect result from the match.
2233    
2234        In order to prevent this happening, the specification of recursion has
2235        been changed so that all such subpatterns are automatically treated as
2236        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2237    
2238     4. I had overlooked the fact that, in some locales, there are characters for
2239        which isalpha() is true but neither isupper() nor islower() are true. In
2240        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2241        and ordfeminine) are like this. This affected the treatment of \w and \W
2242        when they appeared in character classes, but not when they appeared outside
2243        a character class. The bit map for "word" characters is now created
2244        separately from the results of isalnum() instead of just taking it from the
2245        upper, lower, and digit maps. (Plus the underscore character, of course.)
2246    
2247     5. The above bug also affected the handling of POSIX character classes such as
2248        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2249        permanent tables. Instead, the bit maps for such a class were previously
2250        created as the appropriate unions of the upper, lower, and digit bitmaps.
2251        Now they are created by subtraction from the [[:word:]] class, which has
2252        its own bitmap.
2253    
2254     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2255        It is created by subtracting the vertical space characters (\x09, \x0a,
2256        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2257        subtraction was done in the overall bitmap for a character class, meaning
2258        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2259        be recognized. This bug has been fixed.
2260    
2261     7. Patches from the folks at Google:
2262    
2263          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2264          real life, but is still worth protecting against".
2265    
2266          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2267          regular expressions".
2268    
2269          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2270          have it.
2271    
2272          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2273          "configure" and the latter not, in order to fix a problem somebody had
2274          with compiling the Arg class on HP-UX.
2275    
2276          (e) Improve the error-handling of the C++ wrapper a little bit.
2277    
2278          (f) New tests for checking recursion limiting.
2279    
2280     8. The pcre_memmove() function, which is used only if the environment does not
2281        have a standard memmove() function (and is therefore rarely compiled),
2282        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2283        returning a result (though PCRE never actually uses the result).
2284    
2285     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2286        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2287        returned instead of calling malloc() with an overflowing number that would
2288        most likely cause subsequent chaos.
2289    
2290    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2291    
2292    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2293        with this option is matched, the nmatch and pmatch options of regexec() are
2294        ignored.
2295    
2296    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2297        provided in case anyone wants to the the POSIX interface with UTF-8
2298        strings.
2299    
2300    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2301        C++ linking (needed for some HP-UX environments).
2302    
2303    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2304        (unused parameter) and in the pcre_printint() function (omitted "default"
2305        switch label when the default is to do nothing).
2306    
2307    15. Added some code to make it possible, when PCRE is compiled as a C++
2308        library, to replace subject pointers for pcre_exec() with a smart pointer
2309        class, thus making it possible to process discontinuous strings.
2310    
2311    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2312        much the same function. They were added by different people who were trying
2313        to make PCRE easy to compile on non-Unix systems. It has been suggested
2314        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2315        for compiling on Windows systems. I have therefore replaced it with
2316        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2317        defaults to "extern" for C or "extern C" for C++, which works fine on
2318        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2319        SCOPE with something explicit in config.h. In addition:
2320    
2321        (a) pcreposix.h still had just "extern" instead of either of these macros;
2322            I have replaced it with PCRE_DATA_SCOPE.
2323    
2324        (b) Functions such as _pcre_xclass(), which are internal to the library,
2325            but external in the C sense, all had PCRE_EXPORT in their definitions.
2326            This is apparently wrong for the Windows case, so I have removed it.
2327            (It makes no difference on Unix-like systems.)
2328    
2329    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2330        of recursive calls to match(). This is different to MATCH_LIMIT because
2331        that limits the total number of calls to match(), not all of which increase
2332        the depth of recursion. Limiting the recursion depth limits the amount of
2333        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2334        when PCRE is compiled, and changed at run time. A patch from Google adds
2335        this functionality to the C++ interface.
2336    
2337    18. Changes to the handling of Unicode character properties:
2338    
2339        (a) Updated the table to Unicode 4.1.0.
2340    
2341        (b) Recognize characters that are not in the table as "Cn" (undefined).
2342    
2343        (c) I revised the way the table is implemented to a much improved format
2344            which includes recognition of ranges. It now supports the ranges that
2345            are defined in UnicodeData.txt, and it also amalgamates other
2346            characters into ranges. This has reduced the number of entries in the
2347            table from around 16,000 to around 3,000, thus reducing its size
2348            considerably. I realized I did not need to use a tree structure after
2349            all - a binary chop search is just as efficient. Having reduced the
2350            number of entries, I extended their size from 6 bytes to 8 bytes to
2351            allow for more data.
2352    
2353        (d) Added support for Unicode script names via properties such as \p{Han}.
2354    
2355    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2356        matching that character.
2357    
2358    20. When matching a repeated Unicode property with a minimum greater than zero,
2359        (for example \pL{2,}), PCRE could look past the end of the subject if it
2360        reached it while seeking the minimum number of characters. This could
2361        happen only if some of the characters were more than one byte long, because
2362        there is a check for at least the minimum number of bytes.
2363    
2364    21. Refactored the implementation of \p and \P so as to be more general, to
2365        allow for more different types of property in future. This has changed the
2366        compiled form incompatibly. Anybody with saved compiled patterns that use
2367        \p or \P will have to recompile them.
2368    
2369    22. Added "Any" and "L&" to the supported property types.
2370    
2371    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2372        but give a compile time error if the value is greater than 0xff.
2373    
2374    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2375        accidentally not being installed or uninstalled.
2376    
2377    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2378        made were to insert the current release number. This seemed silly, because
2379        it made things harder for people building PCRE on systems that don't run
2380        "configure". I have turned pcre.h into a distributed file, no longer built
2381        by "configure", with the version identification directly included. There is
2382        no longer a pcre.h.in file.
2383    
2384        However, this change necessitated a change to the pcre-config script as
2385        well. It is built from pcre-config.in, and one of the substitutions was the
2386        release number. I have updated configure.ac so that ./configure now finds
2387        the release number by grepping pcre.h.
2388    
2389    26. Added the ability to run the tests under valgrind.
2390    
2391    
2392    Version 6.4 05-Sep-05
2393    ---------------------
2394    
2395     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2396        "--" to be printed when multiple files were scanned, even when none of the
2397        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2398        consider it to be a bug, and have restored the previous behaviour.
2399    
2400     2. A couple of code tidies to get rid of compiler warnings.
2401    
2402     3. The pcretest program used to cheat by referring to symbols in the library
2403        whose names begin with _pcre_. These are internal symbols that are not
2404        really supposed to be visible externally, and in some environments it is
2405        possible to suppress them. The cheating is now confined to including
2406        certain files from the library's source, which is a bit cleaner.
2407    
2408     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2409        file's purpose clearer.
2410    
2411     5. Reorganized pcre_ucp_findchar().
2412    
2413    
2414    Version 6.3 15-Aug-05
2415    ---------------------
2416    
2417     1. The file libpcre.pc.in did not have general read permission in the tarball.
2418    
2419     2. There were some problems when building without C++ support:
2420    
2421        (a) If C++ support was not built, "make install" and "make test" still
2422            tried to test it.
2423    
2424        (b) There were problems when the value of CXX was explicitly set. Some
2425            changes have been made to try to fix these, and ...
2426    
2427        (c) --disable-cpp can now be used to explicitly disable C++ support.
2428    
2429        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2430            backslash in a target when C++ was disabled. This confuses some
2431            versions of "make", apparently. Using an intermediate variable solves
2432            this. (Same for CPP_LOBJ.)
2433    
2434     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2435        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2436        necessary on certain architectures.
2437    
2438     4. Added a setting of -export-symbols-regex to the link command to remove
2439        those symbols that are exported in the C sense, but actually are local
2440        within the library, and not documented. Their names all begin with
2441        "_pcre_". This is not a perfect job, because (a) we have to except some
2442        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2443        available (and never for static libraries). I have made a note to try to
2444        find a way round (a) in the future.
2445    
2446    
2447    Version 6.2 01-Aug-05
2448    ---------------------
2449    
2450     1. There was no test for integer overflow of quantifier values. A construction
2451        such as {1111111111111111} would give undefined results. What is worse, if
2452        a minimum quantifier for a parenthesized subpattern overflowed and became
2453        negative, the calculation of the memory size went wrong. This could have
2454        led to memory overwriting.
2455    
2456     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2457    
2458     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2459        operating environments where this matters.
2460    
2461     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2462        PCRE options from within the C++ wrapper.
2463    
2464     5. Named capturing subpatterns were not being correctly counted when a pattern
2465        was compiled. This caused two problems: (a) If there were more than 100
2466        such subpatterns, the calculation of the memory needed for the whole
2467        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2468        back references of the form \12, where the number was greater than 9, were
2469        not recognized as back references, even though there were sufficient
2470        previous subpatterns.
2471    
2472     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2473        versions of gcc, e.g. 2.95.4.
2474    
2475    
2476    Version 6.1 21-Jun-05
2477    ---------------------
2478    
2479     1. There was one reference to the variable "posix" in pcretest.c that was not
2480        surrounded by "#if !defined NOPOSIX".
2481    
2482     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2483        the cross-check on the old pcre_info() function, for the benefit of the
2484        cut-down version of PCRE that is currently imported into Exim.
2485    
2486     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2487        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2488        patterns that start (?i) but I don't think that matters. The use of (?i) is
2489        just an example; this all applies to the other options as well.
2490    
2491     4. Since libtool seems to echo the compile commands it is issuing, the output
2492        from "make" can be reduced a bit by putting "@" in front of each libtool
2493        compile command.
2494    
2495     5. Patch from the folks at Google for configure.in to be a bit more thorough
2496        in checking for a suitable C++ installation before trying to compile the
2497        C++ stuff. This should fix a reported problem when a compiler was present,
2498        but no suitable headers.
2499    
2500     6. The man pages all had just "PCRE" as their title. I have changed them to
2501        be the relevant file name. I have also arranged that these names are
2502        retained in the file doc/pcre.txt, which is a concatenation in text format
2503        of all the man pages except the little individual ones for each function.
2504    
2505     7. The NON-UNIX-USE file had not been updated for the different set of source
2506        files that come with release 6. I also added a few comments about the C++
2507        wrapper.
2508    
2509    
2510    Version 6.0 07-Jun-05
2511    ---------------------
2512    
2513     1. Some minor internal re-organization to help with my DFA experiments.
2514    
2515     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2516        didn't matter for the library itself when fully configured, but did matter
2517        when compiling without UCP support, or within Exim, where the ucp files are
2518        not imported.
2519    
2520     3. Refactoring of the library code to split up the various functions into
2521        different source modules. The addition of the new DFA matching code (see
2522        below) to a single monolithic source would have made it really too
2523        unwieldy, quite apart from causing all the code to be include in a
2524        statically linked application, when only some functions are used. This is
2525        relevant even without the DFA addition now that patterns can be compiled in
2526        one application and matched in another.
2527    
2528        The downside of splitting up is that there have to be some external
2529        functions and data tables that are used internally in different modules of
2530        the library but which are not part of the API. These have all had their
2531        names changed to start with "_pcre_" so that they are unlikely to clash
2532        with other external names.
2533    
2534     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2535        a different (DFA) algorithm. Although it is slower than the original
2536        function, it does have some advantages for certain types of matching
2537        problem.
2538    
2539     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2540        including restarting after a partial match.
2541    
2542     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2543        defined when compiling for Windows was sent to me. I have put it into the
2544        code, though I have no means of testing or verifying it.
2545    
2546     7. Added the pcre_refcount() auxiliary function.
2547    
2548     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2549        match before or at the first newline in the subject string. In pcretest,
2550        the /f option on a pattern can be used to set this.
2551    
2552     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2553        would behave wrongly. This has been present in PCRE since release 4.0.
2554    
2555    10. A number of changes to the pcregrep command:
2556    
2557        (a) Refactored how -x works; insert ^(...)$ instead of setting
2558            PCRE_ANCHORED and checking the length, in preparation for adding
2559            something similar for -w.
2560    
2561        (b) Added the -w (match as a word) option.
2562    
2563        (c) Refactored the way lines are read and buffered so as to have more
2564            than one at a time available.
2565    
2566        (d) Implemented a pcregrep test script.
2567    
2568        (e) Added the -M (multiline match) option. This allows patterns to match
2569            over several lines of the subject. The buffering ensures that at least
2570            8K, or the rest of the document (whichever is the shorter) is available
2571            for matching (and similarly the previous 8K for lookbehind assertions).
2572    
2573        (f) Changed the --help output so that it now says
2574    
2575              -w, --word-regex(p)
2576    
2577            instead of two lines, one with "regex" and the other with "regexp"
2578            because that confused at least one person since the short forms are the
2579            same. (This required a bit of code, as the output is generated
2580            automatically from a table. It wasn't just a text change.)
2581    
2582        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2583            option but starts with a hyphen. Could be a pattern or a path name
2584            starting with a hyphen, for instance.
2585    
2586        (h) "-" can be given as a file name to represent stdin.
2587    
2588        (i) When file names are being printed, "(standard input)" is used for
2589            the standard input, for compatibility with GNU grep. Previously
2590            "<stdin>" was used.
2591    
2592        (j) The option --label=xxx can be used to supply a name to be used for
2593            stdin when file names are being printed. There is no short form.
2594    
2595        (k) Re-factored the options decoding logic because we are going to add
2596            two more options that take data. Such options can now be given in four
2597            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2598    
2599        (l) Added the -A, -B, and -C options for requesting that lines of context
2600            around matches be printed.
2601    
2602        (m) Added the -L option to print the names of files that do not contain
2603            any matching lines, that is, the complement of -l.
2604    
2605        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2606            continue to scan other files.
2607    
2608        (o) The -s option was incorrectly implemented. For compatibility with other
2609            greps, it now suppresses the error message for a non-existent or non-
2610            accessible file (but not the return code). There is a new option called
2611            -q that suppresses the output of matching lines, which was what -s was
2612            previously doing.
2613    
2614        (p) Added --include and --exclude options to specify files for inclusion
2615            and exclusion when recursing.
2616    
2617    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2618        Hopefully, it now does.
2619    
2620    12. Missing cast in pcre_study().
2621    
2622    13. Added an "uninstall" target to the makefile.
2623    
2624    14. Replaced "extern" in the function prototypes in Makefile.in with
2625        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2626        world, but is set differently for Windows.
2627    
2628    15. Added a second compiling function called pcre_compile2(). The only
2629        difference is that it has an extra argument, which is a pointer to an
2630        integer error code. When there is a compile-time failure, this is set
2631        non-zero, in addition to the error test pointer being set to point to an
2632        error message. The new argument may be NULL if no error number is required
2633        (but then you may as well call pcre_compile(), which is now just a
2634        wrapper). This facility is provided because some applications need a
2635        numeric error indication, but it has also enabled me to tidy up the way
2636        compile-time errors are handled in the POSIX wrapper.
2637    
2638    16. Added VPATH=.libs to the makefile; this should help when building with one
2639        prefix path and installing with another. (Or so I'm told by someone who
2640        knows more about this stuff than I do.)
2641    
2642    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2643        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2644        match everything, including newlines. This is not POSIX-compatible, but
2645        somebody wanted the feature. From pcretest it can be activated by using
2646        both the P and the s flags.
2647    
2648    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2649    
2650    19. libpcre.pc was being incorrectly installed as executable.
2651    
2652    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2653        it now also looks for '\r' so that it will work unmodified on Windows.
2654    
2655    21. Added Google's contributed C++ wrapper to the distribution.
2656    
2657    22. Added some untidy missing memory free() calls in pcretest, to keep
2658        Electric Fence happy when testing.
2659    
2660    
2661    
2662    Version 5.0 13-Sep-04
2663    ---------------------
2664    
2665     1. Internal change: literal characters are no longer packed up into items
2666        containing multiple characters in a single byte-string. Each character
2667        is now matched using a separate opcode. However, there may be more than one
2668        byte in the character in UTF-8 mode.
2669    
2670     2. The pcre_callout_block structure has two new fields: pattern_position and
2671        next_item_length. These contain the offset in the pattern to the next match
2672        item, and its length, respectively.
2673    
2674     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2675        insertion of callouts before each pattern item. Added the /C option to
2676        pcretest to make use of this.
2677    
2678     4. On the advice of a Windows user, the lines
2679    
2680          #if defined(_WIN32) || defined(WIN32)
2681          _setmode( _fileno( stdout ), 0x8000 );
2682          #endif  /* defined(_WIN32) || defined(WIN32) */
2683    
2684        have been added to the source of pcretest. This apparently does useful
2685        magic in relation to line terminators.
2686    
2687     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2688        for the benefit of those environments where the "b" makes a difference.
2689    
2690     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2691        to know about it. I have put a hack into configure.in that adds in code
2692        to set GCC=yes if CC=icc. This seems to end up at a point in the
2693        generated configure script that is early enough to affect the setting of
2694        compiler options, which is what is needed, but I have no means of testing
2695        whether it really works. (The user who reported this had patched the
2696        generated configure script, which of course I cannot do.)
2697    
2698        LATER: After change 22 below (new libtool files), the configure script
2699        seems to know about icc (and also ecc). Therefore, I have commented out
2700        this hack in configure.in.
2701    
2702     7. Added support for pkg-config (2 patches were sent in).
2703    
2704     8. Negated POSIX character classes that used a combination of internal tables
2705        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2706        [[:^ascii]]. Typically, they would match almost any characters. The other
2707        POSIX classes were not broken in this way.
2708    
2709     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2710        to find the match, as PCRE was deluded into thinking that the match had to
2711        start at the start point or following a newline. The same bug applied to
2712        patterns with negative forward assertions or any backward assertions
2713        preceding ".*" at the start, unless the pattern required a fixed first
2714        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2715    
2716    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2717        starting at the last subject character, bytes beyond the end of the subject
2718        string were read.
2719    
2720    11. Renamed the variable "class" as "classbits" to make life easier for C++
2721        users. (Previously there was a macro definition, but it apparently wasn't
2722        enough.)
2723    
2724    12. Added the new field "tables" to the extra data so that tables can be passed
2725        in at exec time, or the internal tables can be re-selected. This allows
2726        a compiled regex to be saved and re-used at a later time by a different
2727        program that might have everything at different addresses.
2728    
2729    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2730        -R library as well as a -L library.
2731    
2732    14. The debugging options of pcretest (-d on the command line or D on a
2733        pattern) showed incorrect output for anything following an extended class
2734        that contained multibyte characters and which was followed by a quantifier.
2735    
2736    15. Added optional support for general category Unicode character properties
2737        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2738        support. It adds about 90K to the size of the library. The meanings of the
2739        inbuilt class escapes such as \d and \s have NOT been changed.
2740    
2741    16. Updated pcredemo.c to include calls to free() to release the memory for the
2742        compiled pattern.
2743    
2744    17. The generated file chartables.c was being created in the source directory
2745        instead of in the building directory. This caused the build to fail if the
2746        source directory was different from the building directory, and was
2747        read-only.
2748    
2749    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2750        file. No doubt somebody will tell me if they don't make sense... Also added
2751        Dan Mooney's comments about building on OpenVMS.
2752    
2753    19. Added support for partial matching via the PCRE_PARTIAL option for
2754        pcre_exec() and the \P data escape in pcretest.
2755    
2756    20. Extended pcretest with 3 new pattern features:
2757    
2758        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2759              write the compiled pattern to the file whose name is "rest-of-line".
2760              This is a straight binary dump of the data, with the saved pointer to
2761              the character tables forced to be NULL. The study data, if any, is
2762              written too. After writing, pcretest reads a new pattern.
2763    
2764        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2765              compiled pattern from the given file. There must not be any
2766              occurrences of "<" in the file name (pretty unlikely); if there are,
2767              pcretest will instead treat the initial "<" as a pattern delimiter.
2768              After reading in the pattern, pcretest goes on to read data lines as
2769              usual.
2770    
2771        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2772              and 16-bit fields in a compiled pattern, to simulate a pattern that
2773              was compiled on a host of opposite endianness.
2774    
2775    21. The pcre-exec() function can now cope with patterns that were compiled on
2776        hosts of opposite endianness, with this restriction:
2777    
2778          As for any compiled expression that is saved and used later, the tables
2779          pointer field cannot be preserved; the extra_data field in the arguments
2780          to pcre_exec() should be used to pass in a tables address if a value
2781          other than the default internal tables were used at compile time.
2782    
2783    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2784        now diagnosed as an error. Previously, most of the time, a negative number
2785        would have been treated as zero, but if in addition "ovector" was passed as
2786        NULL, a crash could occur.
2787    
2788    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2789        new versions from the libtool 1.5 distribution (the last one is a copy of
2790        a file called libtool.m4). This seems to have fixed the need to patch
2791        "configure" to support Darwin 1.3 (which I used to do). However, I still
2792        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2793        workstation).
2794    
2795    24. Changed the PCRE licence to be the more standard "BSD" licence.
2796    
2797    
2798    Version 4.5 01-Dec-03
2799    ---------------------
2800    
2801     1. There has been some re-arrangement of the code for the match() function so
2802        that it can be compiled in a version that does not call itself recursively.
2803        Instead, it keeps those local variables that need separate instances for
2804        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2805        needs to "recurse". Keeping track of where control must go is done by means
2806        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2807        hide most of the details from the main code, and operates only if
2808        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2809        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2810        operating.
2811    
2812        To make it easier for callers to provide specially tailored get/free
2813        functions for this usage, two new functions, pcre_stack_malloc, and
2814        pcre_stack_free, are used. They are always called in strict stacking order,
2815        and the size of block requested is always the same.
2816    
2817        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2818        PCRE has been compiled to use the stack or the heap for recursion. The
2819        -C option of pcretest uses this to show which version is compiled.
2820    
2821        A new data escape \S, is added to pcretest; it causes the amounts of store
2822        obtained and freed by both kinds of malloc/free at match time to be added
2823        to the output.
2824    
2825     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2826        what's available on my current Linux desktop machine.
2827    
2828     3. When matching a UTF-8 string, the test for a valid string at the start has
2829        been extended. If start_offset is not zero, PCRE now checks that it points
2830        to a byte that is the start of a UTF-8 character. If not, it returns
2831        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2832        this is necessary because there may be backward assertions in the pattern.
2833        When matching the same subject several times, it may save resources to use
2834        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2835    
2836     4. The code for checking the validity of UTF-8 strings has been tightened so
2837        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2838        containing "overlong sequences".
2839    
2840     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2841        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2842        should have been "&", but it just so happened that all the cases this let
2843        through by mistake were picked up later in the function.
2844    
2845     6. I had used a variable called "isblank" - this is a C99 function, causing
2846        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2847    
2848     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2849        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2850        so the version is shown; (c) stop "make test" echoing "./RunTest".
2851    
2852     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2853    
2854     9. The prototype for memmove() for systems that don't have it was using
2855        size_t, but the inclusion of the header that defines size_t was later. I've
2856        moved the #includes for the C headers earlier to avoid this.
2857    
2858    10. Added some adjustments to the code to make it easier to compiler on certain
2859        special systems:
2860    
2861          (a) Some "const" qualifiers were missing.
2862          (b) Added the macro EXPORT before all exported functions; by default this
2863              is defined to be empty.
2864          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2865              that it reads its output file name as an argument instead of writing
2866              to the standard output and assuming this can be redirected.
2867    
2868    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2869        class containing characters with values greater than 255, PCRE compilation
2870        went into a loop.
2871    
2872    12. A recursive reference to a subpattern that was within another subpattern
2873        that had a minimum quantifier of zero caused PCRE to crash. For example,
2874        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2875        recursion. If the recursively-called subpattern itself had a zero repeat,
2876        that was OK.
2877    
2878    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2879        buffer into which it was copied (for escape processing) was still set at
2880        1024, so long lines caused crashes.
2881    
2882    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2883        "internal error: code overflow...". This applied to any character class
2884        that was followed by a possessive quantifier.
2885    
2886    15. Modified the Makefile to add libpcre.la as a prerequisite for
2887        libpcreposix.la because I was told this is needed for a parallel build to
2888        work.
2889    
2890    16. If a pattern that contained .* following optional items at the start was
2891        studied, the wrong optimizing data was generated, leading to matching
2892        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2893        matching string must start with a or b or c. The correct conclusion for
2894        this pattern is that a match can start with any character.
2895    
2896    
2897    Version 4.4 13-Aug-03
2898    ---------------------
2899    
2900     1. In UTF-8 mode, a character class containing characters with values between
2901        127 and 255 was not handled correctly if the compiled pattern was studied.
2902        In fixing this, I have also improved the studying algorithm for such
2903        classes (slightly).
2904    
2905     2. Three internal functions had redundant arguments passed to them. Removal
2906        might give a very teeny performance improvement.
2907    
2908     3. Documentation bug: the value of the capture_top field in a callout is *one
2909        more than* the number of the hightest numbered captured substring.
2910    
2911     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2912        in incorrectly linking with a previously installed version. They now link
2913        explicitly with libpcre.la.
2914    
2915     5. configure.in no longer needs to recognize Cygwin specially.
2916    
2917     6. A problem in pcre.in for Windows platforms is fixed.
2918    
2919     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2920        pcretest, it used to include the size of the study block as part of its
2921        output. Unfortunately, the structure contains a field that has a different
2922        size on different hardware architectures. This meant that the tests that
2923        showed this size failed. As the block is currently always of a fixed size,
2924        this information isn't actually particularly useful in pcretest output, so
2925        I have just removed it.
2926    
2927     8. Three pre-processor statements accidentally did not start in column 1.
2928        Sadly, there are *still* compilers around that complain, even though
2929        standard C has not required this for well over a decade. Sigh.
2930    
2931     9. In pcretest, the code for checking callouts passed small integers in the
2932        callout_data field, which is a void * field. However, some picky compilers
2933        complained about the casts involved for this on 64-bit systems. Now
2934        pcretest passes the address of the small integer instead, which should get
2935        rid of the warnings.
2936    
2937    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2938        both compile and run time, and gives an error if an invalid UTF-8 sequence
2939        is found. There is a option for disabling this check in cases where the
2940        string is known to be correct and/or the maximum performance is wanted.
2941    
2942    11. In response to a bug report, I changed one line in Makefile.in from
2943    
2944            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2945        to
2946            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2947    
2948        to look similar to other lines, but I have no way of telling whether this
2949        is the right thing to do, as I do not use Windows. No doubt I'll get told
2950        if it's wrong...
2951    
2952    
2953    Version 4.3 21-May-03
2954    ---------------------
2955    
2956    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2957       Makefile.
2958    
2959    2. Some refactoring to improve the quality of the code:
2960    
2961       (i)   The utf8_table... variables are now declared "const".
2962    
2963       (ii)  The code for \cx, which used the "case flipping" table to upper case
2964             lower case letters, now just substracts 32. This is ASCII-specific,
2965             but the whole concept of \cx is ASCII-specific, so it seems
2966             reasonable.
2967    
2968       (iii) PCRE was using its character types table to recognize decimal and
2969             hexadecimal digits in the pattern. This is silly, because it handles
2970             only 0-9, a-f, and A-F, but the character types table is locale-
2971             specific, which means strange things might happen. A private
2972             table is now used for this - though it costs 256 bytes, a table is
2973             much faster than multiple explicit tests. Of course, the standard
2974             character types table is still used for matching digits in subject
2975             strings against \d.
2976    
2977       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2978             ending in _t are). So I've renamed it as ESC_tee.
2979    
2980    3. The first argument for regexec() in the POSIX wrapper should have been
2981       defined as "const".
2982    
2983    4. Changed pcretest to use malloc() for its buffers so that they can be
2984       Electric Fenced for debugging.
2985    
2986    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2987       to read one or more bytes before the start of the subject string. Often this
2988       had no effect on PCRE's behaviour, but in some circumstances it could
2989       provoke a segmentation fault.
2990    
2991    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2992       to try to read one or more bytes before the start of the subject string.
2993    
2994    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2995       UTF-8 support could misbehave in various ways if the subject string
2996       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2997       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2998       back over UTF-8 characters.)
2999    
3000    
3001    Version 4.2 14-Apr-03
3002    ---------------------
3003    
3004    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3005    
3006    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3007         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3008         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3009         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3010         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3011           and BUILD_EXEEXT
3012         Note: automatic setting of the BUILD variables is not yet working
3013         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3014           compile-time but not at link-time
3015         [LINK]: use for linking executables only
3016         make different versions for Windows and non-Windows
3017         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3018           libraries
3019         [LINK_FOR_BUILD]: new variable
3020         [OBJEXT]: use throughout
3021         [EXEEXT]: use throughout
3022         <winshared>: new target
3023         <wininstall>: new target
3024         <dftables.o>: use native compiler
3025         <dftables>: use native linker
3026         <install>: handle Windows platform correctly
3027         <clean>: ditto
3028         <check>: ditto
3029         copy DLL to top builddir before testing
3030    
3031       As part of these changes, -no-undefined was removed again. This was reported
3032       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3033       in any case.
3034    
3035    3. Some tidies to get rid of compiler warnings:
3036    
3037       . In the match_data structure, match_limit was an unsigned long int, whereas
3038         match_call_count was an int. I've made them both unsigned long ints.
3039    
3040       . In pcretest the fact that a const uschar * doesn't automatically cast to
3041         a void * provoked a warning.
3042    
3043       . Turning on some more compiler warnings threw up some "shadow" variables
3044         and a few more missing casts.
3045    
3046    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3047       option, a class that contained a single character with a value between 128
3048       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3049    
3050    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3051       option, a class that contained several characters, but with at least one
3052       whose value was between 128 and 255 caused PCRE to crash.
3053    
3054    
3055    Version 4.1 12-Mar-03
3056    ---------------------
3057    
3058    1. Compiling with gcc -pedantic found a couple of places where casts were
3059    needed, and a string in dftables.c that was longer than standard compilers are
3060    required to support.
3061    
3062    2. Compiling with Sun's compiler found a few more places where the code could
3063    be tidied up in order to avoid warnings.
3064    
3065    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3066    first of these names is deprecated in the latest Autoconf in favour of the name
3067    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3068    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3069    analogy I have changed it to CFLAGS_FOR_BUILD.
3070    
3071    4. Added -no-undefined to the linking command in the Makefile, because this is
3072    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3073    linking step for the pcreposix library.
3074    
3075    5. PCRE was failing to diagnose the case of two named groups with the same
3076    name.
3077    
3078    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3079    literal character that is needed in the subject for a match, and scans along to
3080    ensure that it is present before embarking on the full matching process. This
3081    saves time in cases of nested unlimited repeats that are never going to match.
3082    Problem: the scan can take a lot of time if the subject is very long (e.g.
3083    megabytes), thus penalizing straightforward matches. It is now done only if the
3084    amount of subject to be scanned is less than 1000 bytes.
3085    
3086    7. A lesser problem with the same optimization is that it was recording the
3087    first character of an anchored pattern as "needed", thus provoking a search
3088    right along the subject, even when the first match of the pattern was going to
3089    fail. The "needed" character is now not set for anchored patterns, unless it
3090    follows something in the pattern that is of non-fixed length. Thus, it still
3091    fulfils its original purpose of finding quick non-matches in cases of nested
3092    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3093    
3094    
3095    Version 4.0 17-Feb-03
3096    ---------------------
3097    
3098    1. If a comment in an extended regex that started immediately after a meta-item
3099    extended to the end of string, PCRE compiled incorrect data. This could lead to
3100    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3101    
3102    2. Moved to autoconf 2.53 and libtool 1.4.2.
3103    
3104    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3105    the special perltest8 script is no longer needed - all the tests can be run
3106    from a single perltest script.
3107    
3108    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3109    by \s. It has now been removed in PCRE. This means it isn't recognized as
3110    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3111    class [:space:] *does* include VT, thereby creating a mess.
3112    
3113    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3114    space and tab.
3115    
3116    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3117    its new features into the main test script, reducing the number of scripts.
3118    
3119    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3120    were backward compatible, and made the (?i) apply to the whole pattern, as if
3121    /i were given. Now it behaves more logically, and applies the option setting
3122    only to what follows. PCRE has been changed to follow suit. However, if it
3123    finds options settings right at the start of the pattern, it extracts them into
3124    the global options, as before. Thus, they show up in the info data.
3125    
3126    8. Added support for the \Q...\E escape sequence. Characters in between are
3127    treated as literals. This is slightly different from Perl in that $ and @ are
3128    also handled as literals inside the quotes. In Perl, they will cause variable
3129    interpolation. Note the following examples:
3130    
3131        Pattern            PCRE matches      Perl matches
3132    
3133        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3134        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3135        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3136    
3137    For compatibility with Perl, \Q...\E sequences are recognized inside character
3138    classes as well as outside them.
3139    
3140    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3141    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3142    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3143    signed/unsigned warnings.
3144    
3145    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3146    option for pcretest, so I've replaced it by a simple function that does just
3147    that job.
3148    
3149    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3150    "pcregrep -".
3151    
3152    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3153    Java package. This provides some syntactic sugar for simple cases of what my
3154    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3155    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3156    item, you can use this simplified notation. Note that only makes sense with
3157    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3158    greediness, whatever the setting of the PCRE_UNGREEDY option.
3159    
3160    13. A change of greediness default within a pattern was not taking effect at
3161    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3162    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3163    was abstracted outside.
3164    
3165    14. PCRE now supports the \G assertion. It is true when the current matching
3166    position is at the start point of the match. This differs from \A when the
3167    starting offset is non-zero. Used with the /g option of pcretest (or similar
3168    code), it works in the same way as it does for Perl's /g option. If all
3169    alternatives of a regex begin with \G, the expression is anchored to the start
3170    match position, and the "anchored" flag is set in the compiled expression.
3171    
3172    15. Some bugs concerning the handling of certain option changes within patterns
3173    have been fixed. These applied to options other than (?ims). For example,
3174    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3175    been the other way round. Some of this was related to change 7 above.
3176    
3177    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3178    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3179    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3180    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3181    
3182    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3183    mode. Unlike ".", it always matches newline, whatever the setting of
3184    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3185    assertions. Perl allows it, but it doesn't (in general) work because it can't
3186    calculate the length of the lookbehind. At least, that's the case for Perl
3187    5.8.0 - I've been told they are going to document that it doesn't work in
3188    future.
3189    
3190    18. Added an error diagnosis for escapes that PCRE does not support: these are
3191    \L, \l, \N, \P, \p, \U, \u, and \X.
3192    
3193    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3194    reading past the end of the pattern in cases such as /[abcd/.
3195    
3196    20. PCRE was getting more memory than necessary for patterns with classes that
3197    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3198    
3199    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3200    compiling PCRE for use with Virtual Pascal.
3201    
3202    22. Small fix to the Makefile to make it work properly if the build is done
3203    outside the source tree.
3204    
3205    23. Added a new extension: a condition to go with recursion. If a conditional
3206    subpattern starts with (?(R) the "true" branch is used if recursion has
3207    happened, whereas the "false" branch is used only at the top level.
3208    
3209    24. When there was a very long string of literal characters (over 255 bytes
3210    without UTF support, over 250 bytes with UTF support), the computation of how
3211    much memory was required could be incorrect, leading to segfaults or other
3212    strange effects.
3213    
3214    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3215    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3216    there was a subsequent back reference to those brackets. This meant that, for
3217    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3218    possible to check for precisely this case. All we can do is abandon the
3219    optimization if .* occurs inside capturing brackets when there are any back
3220    references whatsoever. (See below for a better fix that came later.)
3221    
3222    26. The handling of the optimization for finding the first character of a
3223    non-anchored pattern, and for finding a character that is required later in the
3224    match were failing in some cases. This didn't break the matching; it just
3225    failed to optimize when it could. The way this is done has been re-implemented.
3226    
3227    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3228    
3229    28. Added a new feature that provides some of the functionality that Perl
3230    provides with (?{...}). The facility is termed a "callout". The way it is done
3231    in PCRE is for the caller to provide an optional function, by setting
3232    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3233    global variable. By default it is unset, which disables all calling out. To get
3234    the function called, the regex must include (?C) at appropriate points. This
3235    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3236    This provides a means of identifying different callout points. When PCRE
3237    reaches such a point in the regex, if pcre_callout has been set, the external
3238    function is called. It is provided with data in a structure called
3239    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3240    matching continues; if it returns a non-zero value, the match at the current
3241    point fails. However, backtracking will occur if possible. [This was changed
3242    later and other features added - see item 49 below.]
3243    
3244    29. pcretest is upgraded to test the callout functionality. It provides a
3245    callout function that displays information. By default, it shows the start of
3246    the match and the current position in the text. There are some new data escapes
3247    to vary what happens:
3248    
3249        \C+         in addition, show current contents of captured substrings
3250        \C-         do not supply a callout function
3251        \C!n        return 1 when callout number n is reached
3252        \C!n!m      return 1 when callout number n is reached for the mth time
3253    
3254    30. If pcregrep was called with the -l option and just a single file name, it
3255    output "<stdin>" if a match was found, instead of the file name.
3256    
3257    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3258    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3259    pcre_exec(). This saves a malloc/free per call. The default value of
3260    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3261    when configuring.
3262    
3263    32. The default maximum size of a compiled pattern is 64K. There have been a
3264    few cases of people hitting this limit. The code now uses macros to handle the
3265    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3266    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3267    configuring. Tests 2 and 5 work only with 2-byte links because they output
3268    debugging information about compiled patterns.
3269    
3270    33. Internal code re-arrangements:
3271    
3272    (a) Moved the debugging function for printing out a compiled regex into
3273        its own source file (printint.c) and used #include to pull it into
3274        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3275        separate copies.
3276    
3277    (b) Defined the list of op-code names for debugging as a macro in
3278        internal.h so that it is next to the definition of the opcodes.
3279    
3280    (c) Defined a table of op-code lengths for simpler skipping along compiled
3281        code. This is again a macro in internal.h so that it is next to the
3282        definition of the opcodes.
3283    
3284    34. Added support for recursive calls to individual subpatterns, along the
3285    lines of Robin Houston's patch (but implemented somewhat differently).
3286    
3287    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3288    allow it to read and process whole directories in Win32. This code was
3289    contributed by Lionel Fourquaux; it has not been tested by me.
3290    
3291    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3292    used to name a group. Names consist of alphanumerics and underscores, and must
3293    be unique. Back references use the syntax (?P=name) and recursive calls use
3294    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3295    numbers. The function pcre_fullinfo() can be used after compilation to extract
3296    a name/number map. There are three relevant calls:
3297    
3298      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3299      PCRE_INFO_NAMECOUNT            yields the number of entries
3300      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3301    
3302    The map is a vector of fixed-size entries. The size of each entry depends on
3303    the length of the longest name used. The first two bytes of each entry are the
3304    group number, most significant byte first. There follows the corresponding
3305    name, zero terminated. The names are in alphabetical order.
3306    
3307    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3308    case instead of 255. Making it the same both with and without UTF-8 support
3309    means that the same test output works with both.
3310    
3311    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3312    calling malloc() with a zero argument.
3313    
3314    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3315    optimization. I've improved things by keeping a bitmap of backreferences with
3316    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3317    fact referenced, the optimization can be applied. It is unlikely that a
3318    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3319    the match to follow \n) will appear inside brackets with a number greater than
3320    31, but if it does, any back reference > 31 suppresses the optimization.
3321    
3322    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3323    of disabling numbered capturing parentheses. Any opening parenthesis that is
3324    not followed by ? behaves as if it were followed by ?: but named parentheses
3325    can still be used for capturing (and they will acquire numbers in the usual
3326    way).
3327    
3328    41. Redesigned the return codes from the match() function into yes/no/error so
3329    that errors can be passed back from deep inside the nested calls. A malloc
3330    failure while inside a recursive subpattern call now causes the
3331    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3332    
3333    42. It is now possible to set a limit on the number of times the match()
3334    function is called in a call to pcre_exec(). This facility makes it possible to
3335    limit the amount of recursion and backtracking, though not in a directly
3336    obvious way, because the match() function is used in a number of different
3337    circumstances. The count starts from zero for each position in the subject
3338    string (for non-anchored patterns). The default limit is, for compatibility, a
3339    large number, namely 10 000 000. You can change this in two ways:
3340    
3341    (a) When configuring PCRE before making, you can use --with-match-limit=n
3342        to set a default value for the compiled library.
3343    
3344    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3345        a different value is set. See 45 below.
3346    
3347    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3348    
3349    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3350    of things that can be changed at compile time. The first argument specifies
3351    what is wanted and the second points to where the information is to be placed.
3352    The current list of available information is:
3353    
3354      PCRE_CONFIG_UTF8
3355    
3356    The output is an integer that is set to one if UTF-8 support is available;
3357    otherwise it is set to zero.
3358    
3359      PCRE_CONFIG_NEWLINE
3360    
3361    The output is an integer that it set to the value of the code that is used for
3362    newline. It is either LF (10) or CR (13).
3363    
3364      PCRE_CONFIG_LINK_SIZE
3365    
3366    The output is an integer that contains the number of bytes used for internal
3367    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3368    
3369      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3370    
3371    The output is an integer that contains the threshold above which the POSIX
3372    interface uses malloc() for output vectors. See item 31 above.
3373    
3374      PCRE_CONFIG_MATCH_LIMIT
3375    
3376    The output is an unsigned integer that contains the default limit of the number
3377    of match() calls in a pcre_exec() execution. See 42 above.
3378    
3379    44. pcretest has been upgraded by the addition of the -C option. This causes it
3380    to extract all the available output from the new pcre_config() function, and to
3381    output it. The program then exits immediately.
3382    
3383    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3384    order to support additional features. One way would have been to define
3385    pcre_exec2() (for example) with extra arguments, but this would not have been
3386    extensible, and would also have required all calls to the original function to
3387    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3388    is used for passing in "extra" data from pcre_study().
3389    
3390    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3391    contains the following fields:
3392    
3393      flags         a bitmap indicating which of the following fields are set
3394      study_data    opaque data from pcre_study()
3395      match_limit   a way of specifying a limit on match() calls for a specific
3396                      call to pcre_exec()
3397      callout_data  data for callouts (see 49 below)
3398    
3399    The flag bits are also defined in pcre.h, and are
3400    
3401      PCRE_EXTRA_STUDY_DATA
3402      PCRE_EXTRA_MATCH_LIMIT
3403      PCRE_EXTRA_CALLOUT_DATA
3404    
3405    The pcre_study() function now returns one of these new pcre_extra blocks, with
3406    the actual study data pointed to by the study_data field, and the
3407    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3408    before. That is, this change is entirely upwards-compatible and requires no
3409    change to existing code.
3410    
3411    If you want to pass in additional data to pcre_exec(), you can either place it
3412    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3413    block.
3414    
3415    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3416    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3417    times with different match limits, until it finds the minimum value needed for
3418    pcre_exec() to complete. The value is then output. This can be instructive; for
3419    most simple matches the number is quite small, but for pathological cases it
3420    gets very large very quickly.
3421    
3422    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3423    returns the size of the data block pointed to by the study_data field in a
3424    pcre_extra block, that is, the value that was passed as the argument to
3425    pcre_malloc() when PCRE was getting memory in which to place the information
3426    created by pcre_study(). The fourth argument should point to a size_t variable.
3427    pcretest has been extended so that this information is shown after a successful
3428    pcre_study() call when information about the compiled regex is being displayed.
3429    
3430    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3431    because what follows is always an absolute path. (Later: it turns out that this
3432    is more than cosmetic for MinGW, because it doesn't like empty path
3433    components.)
3434    
3435    49. Some changes have been made to the callout feature (see 28 above):
3436    
3437    (i)  A callout function now has three choices for what it returns:
3438    
3439           0  =>  success, carry on matching
3440         > 0  =>  failure at this point, but backtrack if possible
3441         < 0  =>  serious error, return this value from pcre_exec()
3442    
3443         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3444         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3445         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3446         use by callout functions. It will never be used by PCRE itself.
3447    
3448    (ii) The pcre_extra structure (see 45 above) has a void * field called
3449         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3450         pcre_callout_block structure has a field of the same name. The contents of
3451         the field passed in the pcre_extra structure are passed to the callout
3452         function in the corresponding field in the callout block. This makes it
3453         easier to use the same callout-containing regex from multiple threads. For
3454         testing, the pcretest program has a new data escape
3455    
3456           \C*n        pass the number n (may be negative) as callout_data
3457    
3458         If the callout function in pcretest receives a non-zero value as
3459         callout_data, it returns that value.
3460    
3461    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3462    there were some redundant $(CFLAGS) in commands that are now specified as
3463    $(LINK), which already includes $(CFLAGS).
3464    
3465    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3466    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3467    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3468    one-byte characters throughout. Note that case-insensitive matching applies
3469    only to characters whose values are less than 256. PCRE doesn't support the
3470    notion of cases for higher-valued characters.
3471    
3472    (i)   A character class whose characters are all within 0-255 is handled as
3473          a bit map, and the map is inverted for negative classes. Previously, a
3474          character > 255 always failed to match such a class; however it should
3475          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3476    
3477    (ii)  A negated character class with a single character < 255 is coded as
3478          "not this character" (OP_NOT). This wasn't working properly when the test
3479          character was multibyte, either singly or repeated.
3480    
3481    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3482          mode, for example: \x{100}{2,3}.
3483    
3484    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3485          singly or repeated) now correctly test multibyte characters. However,
3486          PCRE doesn't recognize any characters with values greater than 255 as
3487          digits, spaces, or word characters. Such characters always match \D, \S,
3488          and \W, and never match \d, \s, or \w.
3489    
3490    (v)   Classes may now contain characters and character ranges with values
3491          greater than 255. For example: [ab\x{100}-\x{400}].
3492    
3493    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3494          PCRE in UTF-8 mode.
3495    
3496    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3497    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3498    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3499    value.)
3500    
3501    53. The single man page has become too large. I have therefore split it up into
3502    a number of separate man pages. These also give rise to individual HTML pages;
3503    these are now put in a separate directory, and there is an index.html page that
3504    lists them all. Some hyperlinking between the pages has been installed.
3505    
3506    54. Added convenience functions for handling named capturing parentheses.
3507    
3508    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3509    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3510    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3511    are faulted.
3512    
3513    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3514    calling configure. These values are used when compiling the dftables.c program
3515    which is run to generate the source of the default character tables. They
3516    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3517    you will need to set these values.
3518    
3519    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3520    
3521    
3522    Version 3.9 02-Jan-02
3523    ---------------------
3524    
3525    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3526    
3527    2. If --disable-static was given, the building process failed when trying to
3528    build pcretest and pcregrep. (For some reason it was using libtool to compile
3529    them, which is not right, as they aren't part of the library.)
3530    
3531    
3532    Version 3.8 18-Dec-01
3533    ---------------------
3534    
3535    1. The experimental UTF-8 code was completely screwed up. It was packing the
3536    bytes in the wrong order. How dumb can you get?
3537    
3538    
3539    Version 3.7 29-Oct-01
3540    ---------------------
3541    
3542    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3543    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3544    this didn't happen under Solaris 8, where I normally test things.
3545    
3546    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3547    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3548    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3549    this fix an hour or so after the initial 3.7 release.)
3550    
3551    
3552    Version 3.6 23-Oct-01
3553    ---------------------
3554    
3555    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3556    offsets passed as NULL with zero offset count.
3557    
3558    2. The config.guess and config.sub files had not been updated when I moved to
3559    the latest autoconf.
3560    
3561    
3562    Version 3.5 15-Aug-01
3563    ---------------------
3564    
3565    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3566    had been forgotten.
3567    
3568    2. By using declared but undefined structures, we can avoid using "void"
3569    definitions in pcre.h while keeping the internal definitions of the structures
3570    private.
3571    
3572    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3573    user point of view, this means that both static and shared libraries are built
3574    by default, but this can be individually controlled. More of the work of
3575    handling this static/shared cases is now inside libtool instead of PCRE's make
3576    file.
3577    
3578    4. The pcretest utility is now installed along with pcregrep because it is
3579    useful for users (to test regexs) and by doing this, it automatically gets
3580    relinked by libtool. The documentation has been turned into a man page, so
3581    there are now .1, .txt, and .html versions in /doc.
3582    
3583    5. Upgrades to pcregrep:
3584       (i)   Added long-form option names like gnu grep.
3585       (ii)  Added --help to list all options with an explanatory phrase.
3586       (iii) Added -r, --recursive to recurse into sub-directories.
3587       (iv)  Added -f, --file to read patterns from a file.
3588    
3589    6. pcre_exec() was referring to its "code" argument before testing that
3590    argument for NULL (and giving an error if it was NULL).
3591    
3592    7. Upgraded Makefile.in to allow for compiling in a different directory from
3593    the source directory.
3594    
3595    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3596    options bits, the pointer it was passed was to an int instead of to an unsigned
3597    long int. This mattered only on 64-bit systems.
3598    
3599    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3600    generated) instead of pcre.in, which it its source. Also made the same change
3601    in several of the .c files.
3602    
3603    10. A new release of gcc defines printf() as a macro, which broke pcretest
3604    because it had an ifdef in the middle of a string argument for printf(). Fixed
3605    by using separate calls to printf().
3606    
3607    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3608    script, to force use of CR or LF instead of \n in the source. On non-Unix
3609    systems, the value can be set in config.h.
3610    
3611    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3612    absolute limit. Changed the text of the error message to make this clear, and
3613    likewise updated the man page.
3614    
3615    13. The limit of 99 on the number of capturing subpatterns has been removed.
3616    The new limit is 65535, which I hope will not be a "real" limit.
3617    
3618    
3619    Version 3.4 22-Aug-00
3620    ---------------------
3621    
3622    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3623    
3624    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3625    
3626    
3627  Version 3.3 01-Aug-00  Version 3.3 01-Aug-00
3628  ---------------------  ---------------------

Legend:
Removed from v.49  
changed lines
  Added in v.620

  ViewVC Help
Powered by ViewVC 1.1.5