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

Diff of /code/trunk/ChangeLog

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

revision 11 by nigel, Sat Feb 24 21:38:17 2007 UTC revision 613 by ph10, Sat Jul 2 16:59:52 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    
118    Version 8.12 15-Jan-2011
119    ------------------------
120    
121    1.  Fixed some typos in the markup of the man pages, and wrote a script that
122        checks for such things as part of the documentation building process.
123    
124    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
125        --match-limit and --recursion-limit options (added for 8.11). In
126        particular, this made one of the standard tests fail. (The integer value
127        went into the wrong half of a long int.)
128    
129    3.  If the --colour option was given to pcregrep with -v (invert match), it
130        did strange things, either producing crazy output, or crashing. It should,
131        of course, ignore a request for colour when reporting lines that do not
132        match.
133    
134    4.  Another pcregrep bug caused similar problems if --colour was specified with
135        -M (multiline) and the pattern match finished with a line ending.
136    
137    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
138        matched in multiline mode, the following line was shown as part of the
139        match. This seems wrong, so I have changed it.
140    
141    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
142        the check for further matches in the same line (so they could be coloured)
143        to overrun the end of the current line. If another match was found, it was
144        incorrectly shown (and then shown again when found in the next line).
145    
146    7.  If pcregrep was compiled under Windows, there was a reference to the
147        function pcregrep_exit() before it was defined. I am assuming this was
148        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
149        reported by a user. I've moved the definition above the reference.
150    
151    
152    Version 8.11 10-Dec-2010
153    ------------------------
154    
155    1.  (*THEN) was not working properly if there were untried alternatives prior
156        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
157        backtracked to try for "b" instead of moving to the next alternative branch
158        at the same level (in this case, to look for "c"). The Perl documentation
159        is clear that when (*THEN) is backtracked onto, it goes to the "next
160        alternative in the innermost enclosing group".
161    
162    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
163        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
164        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
165        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
166        (*THEN).
167    
168    3.  If \s appeared in a character class, it removed the VT character from
169        the class, even if it had been included by some previous item, for example
170        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
171        of \s, but is part of the POSIX "space" class.)
172    
173    4.  A partial match never returns an empty string (because you can always
174        match an empty string at the end of the subject); however the checking for
175        an empty string was starting at the "start of match" point. This has been
176        changed to the "earliest inspected character" point, because the returned
177        data for a partial match starts at this character. This means that, for
178        example, /(?<=abc)def/ gives a partial match for the subject "abc"
179        (previously it gave "no match").
180    
181    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
182        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
183        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
184        has an implication that the given string is incomplete (because a partial
185        match is preferred over a full match). For this reason, these items now
186        give a partial match in this situation. [Aside: previously, the one case
187        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
188        match rather than a full match, which was wrong by the old rules, but is
189        now correct.]
190    
191    6.  There was a bug in the handling of #-introduced comments, recognized when
192        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
193        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
194        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
195        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
196        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
197        places in pcre_compile().
198    
199    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
200        comments when looking ahead for named forward references to subpatterns,
201        the only newline sequence it recognized was NL. It now handles newlines
202        according to the set newline convention.
203    
204    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
205        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
206        cater for a lack of strerror(). These oversights have been fixed.
207    
208    9.  Added --match-limit and --recursion-limit to pcregrep.
209    
210    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
211    
212    11. When the -o option was used, pcregrep was setting a return code of 1, even
213        when matches were found, and --line-buffered was not being honoured.
214    
215    12. Added an optional parentheses number to the -o and --only-matching options
216        of pcregrep.
217    
218    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
219        can match an empty string. The code to do it in pcretest and pcredemo
220        needed fixing:
221    
222        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
223            only one byte after an empty string match just before CRLF (this case
224            just got forgotten; "any" and "anycrlf" were OK).
225    
226        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
227            mode when an empty string match preceded an ASCII character followed by
228            a non-ASCII character. (The code for advancing by one character rather
229            than one byte was nonsense.)
230    
231        (c) The pcredemo.c sample program did not have any code at all to handle
232            the cases when CRLF is a valid newline sequence.
233    
234    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
235        as a starting offset was within the subject string. There is now a new
236        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
237        negative or greater than the length of the string. In order to test this,
238        pcretest is extended to allow the setting of negative starting offsets.
239    
240    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
241        starting offset points to the beginning of a UTF-8 character was
242        unnecessarily clumsy. I tidied it up.
243    
244    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
245        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
246    
247    17. Nobody had reported that the --include_dir option, which was added in
248        release 7.7 should have been called --include-dir (hyphen, not underscore)
249        for compatibility with GNU grep. I have changed it to --include-dir, but
250        left --include_dir as an undocumented synonym, and the same for
251        --exclude-dir, though that is not available in GNU grep, at least as of
252        release 2.5.4.
253    
254    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
255        characters from a string of bytes) have been redefined so as not to use
256        loops, in order to improve performance in some environments. At the same
257        time, I abstracted some of the common code into auxiliary macros to save
258        repetition (this should not affect the compiled code).
259    
260    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
261        compile-time error is now given if \c is not followed by an ASCII
262        character, that is, a byte less than 128. (In EBCDIC mode, the code is
263        different, and any byte value is allowed.)
264    
265    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
266        START_OPTIMIZE option, which is now allowed at compile time - but just
267        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
268        to pcregrep and other applications that have no direct access to PCRE
269        options. The new /Y option in pcretest sets this option when calling
270        pcre_compile().
271    
272    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
273        back references. Groups containing recursive back references were forced to
274        be atomic by that change, but in the case of named groups, the amount of
275        memory required was incorrectly computed, leading to "Failed: internal
276        error: code overflow". This has been fixed.
277    
278    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
279        pcretest.c, to avoid build problems in some Borland environments.
280    
281    
282    Version 8.10 25-Jun-2010
283    ------------------------
284    
285    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
286        THEN.
287    
288    2.  (*ACCEPT) was not working when inside an atomic group.
289    
290    3.  Inside a character class, \B is treated as a literal by default, but
291        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
292        causes the error). The code is unchanged, but I tidied the documentation.
293    
294    4.  Inside a character class, PCRE always treated \R and \X as literals,
295        whereas Perl faults them if its -w option is set. I have changed PCRE so
296        that it faults them when PCRE_EXTRA is set.
297    
298    5.  Added support for \N, which always matches any character other than
299        newline. (It is the same as "." when PCRE_DOTALL is not set.)
300    
301    6.  When compiling pcregrep with newer versions of gcc which may have
302        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
303        declared with attribute warn_unused_result" were given. Just casting the
304        result to (void) does not stop the warnings; a more elaborate fudge is
305        needed. I've used a macro to implement this.
306    
307    7.  Minor change to pcretest.c to avoid a compiler warning.
308    
309    8.  Added four artifical Unicode properties to help with an option to make
310        \s etc use properties (see next item). The new properties are: Xan
311        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
312    
313    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
314        use Unicode properties. (*UCP) at the start of a pattern can be used to set
315        this option. Modified pcretest to add /W to test this facility. Added
316        REG_UCP to make it available via the POSIX interface.
317    
318    10. Added --line-buffered to pcregrep.
319    
320    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
321        studied, and the match started with a letter with a code point greater than
322        127 whose first byte was different to the first byte of the other case of
323        the letter, the other case of this starting letter was not recognized
324        (#976).
325    
326    12. If a pattern that was studied started with a repeated Unicode property
327        test, for example, \p{Nd}+, there was the theoretical possibility of
328        setting up an incorrect bitmap of starting bytes, but fortunately it could
329        not have actually happened in practice until change 8 above was made (it
330        added property types that matched character-matching opcodes).
331    
332    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
333        possible starting bytes for non-anchored patterns.
334    
335    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
336        \R, and also a number of cases that involve Unicode properties, both
337        explicit and implicit when PCRE_UCP is set.
338    
339    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
340        input, it could crash or give wrong results if characters with values
341        greater than 0xc0 were present in the subject string. (Detail: it assumed
342        UTF-8 input when processing these items.)
343    
344    16. Added a lot of (int) casts to avoid compiler warnings in systems where
345        size_t is 64-bit (#991).
346    
347    17. Added a check for running out of memory when PCRE is compiled with
348        --disable-stack-for-recursion (#990).
349    
350    18. If the last data line in a file for pcretest does not have a newline on
351        the end, a newline was missing in the output.
352    
353    19. The default pcre_chartables.c file recognizes only ASCII characters (values
354        less than 128) in its various bitmaps. However, there is a facility for
355        generating tables according to the current locale when PCRE is compiled. It
356        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
357        characters, are recognized by isspace() and therefore were getting set in
358        these tables, and indeed these tables seem to approximate to ISO 8859. This
359        caused a problem in UTF-8 mode when pcre_study() was used to create a list
360        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
361        which of course cannot start UTF-8 characters. I have changed the code so
362        that only real ASCII characters (less than 128) and the correct starting
363        bytes for UTF-8 encodings are set for characters greater than 127 when in
364        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
365        altogether.)
366    
367    20. Added the /T option to pcretest so as to be able to run tests with non-
368        standard character tables, thus making it possible to include the tests
369        used for 19 above in the standard set of tests.
370    
371    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
372        reference to a subpattern the other side of a comment that contains an
373        opening parenthesis caused either an internal compiling error, or a
374        reference to the wrong subpattern.
375    
376    
377    Version 8.02 19-Mar-2010
378    ------------------------
379    
380    1.  The Unicode data tables have been updated to Unicode 5.2.0.
381    
382    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
383        configured.
384    
385    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
386        original author of that file, following a query about its status.
387    
388    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
389        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
390    
391    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
392        quantifier applied to a forward-referencing subroutine call, could compile
393        incorrect code or give the error "internal error: previously-checked
394        referenced subpattern not found".
395    
396    6.  Both MS Visual Studio and Symbian OS have problems with initializing
397        variables to point to external functions. For these systems, therefore,
398        pcre_malloc etc. are now initialized to local functions that call the
399        relevant global functions.
400    
401    7.  There were two entries missing in the vectors called coptable and poptable
402        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
403        I've fixed the data, and added a kludgy way of testing at compile time that
404        the lengths are correct (equal to the number of opcodes).
405    
406    8.  Following on from 7, I added a similar kludge to check the length of the
407        eint vector in pcreposix.c.
408    
409    9.  Error texts for pcre_compile() are held as one long string to avoid too
410        much relocation at load time. To find a text, the string is searched,
411        counting zeros. There was no check for running off the end of the string,
412        which could happen if a new error number was added without updating the
413        string.
414    
415    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
416    
417    11. \K was not working if it appeared in an atomic group or in a group that
418        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
419        \K is "not well defined" if used in an assertion. PCRE now accepts it if
420        the assertion is positive, but not if it is negative.
421    
422    12. Change 11 fortuitously reduced the size of the stack frame used in the
423        "match()" function of pcre_exec.c by one pointer. Forthcoming
424        implementation of support for (*MARK) will need an extra pointer on the
425        stack; I have reserved it now, so that the stack frame size does not
426        decrease.
427    
428    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
429        item in branch that calls a recursion is a subroutine call - as in the
430        second branch in the above example - was incorrectly given the compile-
431        time error "recursive call could loop indefinitely" because pcre_compile()
432        was not correctly checking the subroutine for matching a non-empty string.
433    
434    14. The checks for overrunning compiling workspace could trigger after an
435        overrun had occurred. This is a "should never occur" error, but it can be
436        triggered by pathological patterns such as hundreds of nested parentheses.
437        The checks now trigger 100 bytes before the end of the workspace.
438    
439    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
440    
441    
442    Version 8.01 19-Jan-2010
443    ------------------------
444    
445    1.  If a pattern contained a conditional subpattern with only one branch (in
446        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
447        computed the wrong minimum data length (which is of course zero for such
448        subpatterns). This could cause incorrect "no match" results.
449    
450    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
451        the pattern is reset in the first branch, pcre_compile() failed with
452        "internal error: code overflow at offset...". This happened only when
453        the reset was to the original external option setting. (An optimization
454        abstracts leading options settings into an external setting, which was the
455        cause of this.)
456    
457    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
458        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
459        assertion pattern did not match (meaning that the assertion was true), it
460        was incorrectly treated as false if the SKIP had been reached during the
461        matching. This also applied to assertions used as conditions.
462    
463    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
464        assertion subpattern, including such a pattern used as a condition,
465        unpredictable results occurred, instead of the error return
466        PCRE_ERROR_DFA_UITEM.
467    
468    5.  The C++ GlobalReplace function was not working like Perl for the special
469        situation when an empty string is matched. It now does the fancy magic
470        stuff that is necessary.
471    
472    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
473        removed. (These were left over from very, very early versions of PCRE.)
474    
475    7.  Some cosmetic changes to the code to make life easier when compiling it
476        as part of something else:
477    
478        (a) Change DEBUG to PCRE_DEBUG.
479    
480        (b) In pcre_compile(), rename the member of the "branch_chain" structure
481            called "current" as "current_branch", to prevent a collision with the
482            Linux macro when compiled as a kernel module.
483    
484        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
485            prevent a collision with the Linux macro when compiled as a kernel
486            module.
487    
488    8.  In pcre_compile() there are some checks for integer overflows that used to
489        cast potentially large values to (double). This has been changed to that
490        when building, a check for int64_t is made, and if it is found, it is used
491        instead, thus avoiding the use of floating point arithmetic. (There is no
492        other use of FP in PCRE.) If int64_t is not found, the fallback is to
493        double.
494    
495    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
496        2005 (difference between two addresses compared to an unsigned value).
497    
498    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
499        custom one, because of the following reported problem in Windows:
500    
501          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
502              under Win32.
503          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
504              therefore missing the function definition.
505          - The compiler thus generates a "C" signature for the test function.
506          - The linker fails to find the "C" function.
507          - PCRE fails to configure if asked to do so against libbz2.
508    
509    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
510        messages were output:
511    
512          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
513          rerunning libtoolize, to keep the correct libtool macros in-tree.
514          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
515    
516        I have done both of these things.
517    
518    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
519        most of the time, it *can* run out if it is given a pattern that contains a
520        runaway infinite recursion. I updated the discussion in the pcrestack man
521        page.
522    
523    13. Now that we have gone to the x.xx style of version numbers, the minor
524        version may start with zero. Using 08 or 09 is a bad idea because users
525        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
526        interpreted as invalid octal numbers. I've updated the previous comment in
527        configure.ac, and also added a check that gives an error if 08 or 09 are
528        used.
529    
530    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
531        causing partial matching to fail when the end of the subject matched \W
532        in a UTF-8 pattern where \W was quantified with a minimum of 3.
533    
534    15. There were some discrepancies between the declarations in pcre_internal.h
535        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
536        their definitions. The declarations used "const uschar *" and the
537        definitions used USPTR. Even though USPTR is normally defined as "const
538        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
539        reported that: "This difference in casting confuses some C++ compilers, for
540        example, SunCC recognizes above declarations as different functions and
541        generates broken code for hbpcre." I have changed the declarations to use
542        USPTR.
543    
544    16. GNU libtool is named differently on some systems. The autogen.sh script now
545        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
546        (FreeBSD).
547    
548    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
549        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
550        comment: "Figure out how to create a longlong from a string: strtoll and
551        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
552        instance, but it only takes 2 args instead of 3!"
553    
554    18. A subtle bug concerned with back references has been fixed by a change of
555        specification, with a corresponding code fix. A pattern such as
556        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
557        refers, was giving matches when it shouldn't. For example, xa=xaaa would
558        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
559        same bug. Such groups have to be quantified to be useful, or contained
560        inside another quantified group. (If there's no repetition, the reference
561        can never match.) The problem arises because, having left the group and
562        moved on to the rest of the pattern, a later failure that backtracks into
563        the group uses the captured value from the final iteration of the group
564        rather than the correct earlier one. I have fixed this in PCRE by forcing
565        any group that contains a reference to itself to be an atomic group; that
566        is, there cannot be any backtracking into it once it has completed. This is
567        similar to recursive and subroutine calls.
568    
569    
570    Version 8.00 19-Oct-09
571    ----------------------
572    
573    1.  The table for translating pcre_compile() error codes into POSIX error codes
574        was out-of-date, and there was no check on the pcre_compile() error code
575        being within the table. This could lead to an OK return being given in
576        error.
577    
578    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
579        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
580        in a Windows environment.
581    
582    3.  The pcregrep --count option prints the count for each file even when it is
583        zero, as does GNU grep. However, pcregrep was also printing all files when
584        --files-with-matches was added. Now, when both options are given, it prints
585        counts only for those files that have at least one match. (GNU grep just
586        prints the file name in this circumstance, but including the count seems
587        more useful - otherwise, why use --count?) Also ensured that the
588        combination -clh just lists non-zero counts, with no names.
589    
590    4.  The long form of the pcregrep -F option was incorrectly implemented as
591        --fixed_strings instead of --fixed-strings. This is an incompatible change,
592        but it seems right to fix it, and I didn't think it was worth preserving
593        the old behaviour.
594    
595    5.  The command line items --regex=pattern and --regexp=pattern were not
596        recognized by pcregrep, which required --regex pattern or --regexp pattern
597        (with a space rather than an '='). The man page documented the '=' forms,
598        which are compatible with GNU grep; these now work.
599    
600    6.  No libpcreposix.pc file was created for pkg-config; there was just
601        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
602    
603    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
604        when UCP support is not needed, by modifying the Python script that
605        generates it from Unicode data files. This should not matter if the module
606        is correctly used as a library, but I received one complaint about 50K of
607        unwanted data. My guess is that the person linked everything into his
608        program rather than using a library. Anyway, it does no harm.
609    
610    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
611        was a minimum greater than 1 for a wide character in a possessive
612        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
613        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
614        character. Chaos in the form of incorrect output or a compiling loop could
615        result.
616    
617    9.  The restrictions on what a pattern can contain when partial matching is
618        requested for pcre_exec() have been removed. All patterns can now be
619        partially matched by this function. In addition, if there are at least two
620        slots in the offset vector, the offset of the earliest inspected character
621        for the match and the offset of the end of the subject are set in them when
622        PCRE_ERROR_PARTIAL is returned.
623    
624    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
625        synonymous with PCRE_PARTIAL, for backwards compatibility, and
626        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
627        and may be more useful for multi-segment matching.
628    
629    11. Partial matching with pcre_exec() is now more intuitive. A partial match
630        used to be given if ever the end of the subject was reached; now it is
631        given only if matching could not proceed because another character was
632        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
633        string "Z", which now yields "no match" instead of "partial match". In the
634        case of pcre_dfa_exec(), "no match" is given if every matching path for the
635        final character ended with (*FAIL).
636    
637    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
638        if the pattern had a "must contain" character that was already found in the
639        earlier partial match, unless partial matching was again requested. For
640        example, with the pattern /dog.(body)?/, the "must contain" character is
641        "g". If the first part-match was for the string "dog", restarting with
642        "sbody" failed. This bug has been fixed.
643    
644    13. The string returned by pcre_dfa_exec() after a partial match has been
645        changed so that it starts at the first inspected character rather than the
646        first character of the match. This makes a difference only if the pattern
647        starts with a lookbehind assertion or \b or \B (\K is not supported by
648        pcre_dfa_exec()). It's an incompatible change, but it makes the two
649        matching functions compatible, and I think it's the right thing to do.
650    
651    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
652        so that the demonstration program is easily available in environments where
653        PCRE has not been installed from source.
654    
655    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
656        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
657        library.
658    
659    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
660        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
661        is not the first non-POSIX option to be added. Clearly some people find
662        these options useful.
663    
664    17. If a caller to the POSIX matching function regexec() passes a non-zero
665        value for nmatch with a NULL value for pmatch, the value of
666        nmatch is forced to zero.
667    
668    18. RunGrepTest did not have a test for the availability of the -u option of
669        the diff command, as RunTest does. It now checks in the same way as
670        RunTest, and also checks for the -b option.
671    
672    19. If an odd number of negated classes containing just a single character
673        interposed, within parentheses, between a forward reference to a named
674        subpattern and the definition of the subpattern, compilation crashed with
675        an internal error, complaining that it could not find the referenced
676        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
677        [The bug was that it was starting one character too far in when skipping
678        over the character class, thus treating the ] as data rather than
679        terminating the class. This meant it could skip too much.]
680    
681    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
682        /g option in pcretest when the pattern contains \K, which makes it possible
683        to have an empty string match not at the start, even when the pattern is
684        anchored. Updated pcretest and pcredemo to use this option.
685    
686    21. If the maximum number of capturing subpatterns in a recursion was greater
687        than the maximum at the outer level, the higher number was returned, but
688        with unset values at the outer level. The correct (outer level) value is
689        now given.
690    
691    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
692        PCRE did not set those parentheses (unlike Perl). I have now found a way to
693        make it do so. The string so far is captured, making this feature
694        compatible with Perl.
695    
696    23. The tests have been re-organized, adding tests 11 and 12, to make it
697        possible to check the Perl 5.10 features against Perl 5.10.
698    
699    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
700        pattern matches a fixed length string. PCRE did not allow this; now it
701        does. Neither allows recursion.
702    
703    25. I finally figured out how to implement a request to provide the minimum
704        length of subject string that was needed in order to match a given pattern.
705        (It was back references and recursion that I had previously got hung up
706        on.) This code has now been added to pcre_study(); it finds a lower bound
707        to the length of subject needed. It is not necessarily the greatest lower
708        bound, but using it to avoid searching strings that are too short does give
709        some useful speed-ups. The value is available to calling programs via
710        pcre_fullinfo().
711    
712    26. While implementing 25, I discovered to my embarrassment that pcretest had
713        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
714        study optimizations had never been tested with that matching function.
715        Oops. What is worse, even when it was passed study data, there was a bug in
716        pcre_dfa_exec() that meant it never actually used it. Double oops. There
717        were also very few tests of studied patterns with pcre_dfa_exec().
718    
719    27. If (?| is used to create subpatterns with duplicate numbers, they are now
720        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
721        on the other side of the coin, they are no longer allowed to have different
722        names, because these cannot be distinguished in PCRE, and this has caused
723        confusion. (This is a difference from Perl.)
724    
725    28. When duplicate subpattern names are present (necessarily with different
726        numbers, as required by 27 above), and a test is made by name in a
727        conditional pattern, either for a subpattern having been matched, or for
728        recursion in such a pattern, all the associated numbered subpatterns are
729        tested, and the overall condition is true if the condition is true for any
730        one of them. This is the way Perl works, and is also more like the way
731        testing by number works.
732    
733    
734    Version 7.9 11-Apr-09
735    ---------------------
736    
737    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
738        (pcretest), all targets were linked against these libraries. This included
739        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
740        libraries. This caused unwanted dependencies to be created. This problem
741        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
742        pcretest is linked with readline.
743    
744    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
745        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
746        moved outside it again, because FALSE and TRUE are already defined in AIX,
747        but BOOL is not.
748    
749    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
750        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
751    
752    4.  The pcregrep documentation said spaces were inserted as well as colons (or
753        hyphens) following file names and line numbers when outputting matching
754        lines. This is not true; no spaces are inserted. I have also clarified the
755        wording for the --colour (or --color) option.
756    
757    5.  In pcregrep, when --colour was used with -o, the list of matching strings
758        was not coloured; this is different to GNU grep, so I have changed it to be
759        the same.
760    
761    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
762        each matching line was coloured. Now it goes on to look for further matches
763        of any of the test patterns, which is the same behaviour as GNU grep.
764    
765    7.  A pattern that could match an empty string could cause pcregrep to loop; it
766        doesn't make sense to accept an empty string match in pcregrep, so I have
767        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
768        seems to be how GNU grep behaves.
769    
770    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
771        start or after a newline", because the conditional assertion was not being
772        correctly handled. The rule now is that both the assertion and what follows
773        in the first alternative must satisfy the test.
774    
775    9.  If auto-callout was enabled in a pattern with a conditional group whose
776        condition was an assertion, PCRE could crash during matching, both with
777        pcre_exec() and pcre_dfa_exec().
778    
779    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
780        used for matching.
781    
782    11. Unicode property support in character classes was not working for
783        characters (bytes) greater than 127 when not in UTF-8 mode.
784    
785    12. Added the -M command line option to pcretest.
786    
787    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
788    
789    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
790    
791    16. Added comments and documentation about mis-use of no_arg in the C++
792        wrapper.
793    
794    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
795        from Martin Jerabek that uses macro names for all relevant character and
796        string constants.
797    
798    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
799        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
800        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
801        these, but not everybody uses configure.
802    
803    19. A conditional group that had only one branch was not being correctly
804        recognized as an item that could match an empty string. This meant that an
805        enclosing group might also not be so recognized, causing infinite looping
806        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
807        with the subject "ab", where knowledge that the repeated group can match
808        nothing is needed in order to break the loop.
809    
810    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
811        exec(), but without supplying a callout function, matching went wrong.
812    
813    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
814        leak if the size of the offset vector was greater than 30. When the vector
815        is smaller, the saved offsets during recursion go onto a local stack
816        vector, but for larger vectors malloc() is used. It was failing to free
817        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
818        error, in fact).
819    
820    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
821        heapframe that is used only when UTF-8 support is enabled. This caused no
822        problem, but was untidy.
823    
824    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
825        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
826        included within another project.
827    
828    24. Steven Van Ingelgem's patches to add more options to the CMake support,
829        slightly modified by me:
830    
831          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
832              not building pcregrep.
833    
834          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
835              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
836    
837    25. Forward references, both numeric and by name, in patterns that made use of
838        duplicate group numbers, could behave incorrectly or give incorrect errors,
839        because when scanning forward to find the reference group, PCRE was not
840        taking into account the duplicate group numbers. A pattern such as
841        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
842    
843    26. Changed a few more instances of "const unsigned char *" to USPTR, making
844        the feature of a custom pointer more persuasive (as requested by a user).
845    
846    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
847        pcretest.c, inside #ifndefs, because it seems they are sometimes already
848        pre-defined.
849    
850    28. Added support for (*UTF8) at the start of a pattern.
851    
852    29. Arrange for flags added by the "release type" setting in CMake to be shown
853        in the configuration summary.
854    
855    
856    Version 7.8 05-Sep-08
857    ---------------------
858    
859    1.  Replaced UCP searching code with optimized version as implemented for Ad
860        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
861        stage table and inline lookup instead of a function, giving speed ups of 2
862        to 5 times on some simple patterns that I tested. Permission was given to
863        distribute the MultiStage2.py script that generates the tables (it's not in
864        the tarball, but is in the Subversion repository).
865    
866    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
867        scripts.
868    
869    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
870        a group with a zero qualifier. The result of the study could be incorrect,
871        or the function might crash, depending on the pattern.
872    
873    4.  Caseless matching was not working for non-ASCII characters in back
874        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
875        It now works when Unicode Property Support is available.
876    
877    5.  In pcretest, an escape such as \x{de} in the data was always generating
878        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
879        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
880        truncation.
881    
882    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
883    
884    7.  Added two (int) casts to pcregrep when printing the difference of two
885        pointers, in case they are 64-bit values.
886    
887    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
888        test 2 if it fails.
889    
890    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
891        and a #define of that name to empty if it is not externally set. This is to
892        allow users of MSVC to set it if necessary.
893    
894    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
895        the convenience functions in the pcre_get.c source file.
896    
897    11. An option change at the start of a pattern that had top-level alternatives
898        could cause overwriting and/or a crash. This command provoked a crash in
899        some environments:
900    
901          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
902    
903        This potential security problem was recorded as CVE-2008-2371.
904    
905    12. For a pattern where the match had to start at the beginning or immediately
906        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
907        pcre_dfa_exec() could read past the end of the passed subject if there was
908        no match. To help with detecting such bugs (e.g. with valgrind), I modified
909        pcretest so that it places the subject at the end of its malloc-ed buffer.
910    
911    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
912        exec() might read past the end of the data buffer in UTF-8 mode.
913    
914    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
915        the data contained the byte 0x85 as part of a UTF-8 character within its
916        first line. This applied both to normal and DFA matching.
917    
918    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
919        /^[^d]*?$/8 failed to match "abc".
920    
921    16. Added a missing copyright notice to pcrecpp_internal.h.
922    
923    17. Make it more clear in the documentation that values returned from
924        pcre_exec() in ovector are byte offsets, not character counts.
925    
926    18. Tidied a few places to stop certain compilers from issuing warnings.
927    
928    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
929        supplied by Stefan Weber. I made a further small update for 7.8 because
930        there is a change of source arrangements: the pcre_searchfuncs.c module is
931        replaced by pcre_ucd.c.
932    
933    
934    Version 7.7 07-May-08
935    ---------------------
936    
937    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
938        a string to a long long, pretend we don't even have a long long." This is
939        done by checking for the strtoq, strtoll, and _strtoi64 functions.
940    
941    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
942        pre-7.6 versions, which defined a global no_arg variable instead of putting
943        it in the RE class. (See also #8 below.)
944    
945    3.  Remove a line of dead code, identified by coverity and reported by Nuno
946        Lopes.
947    
948    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
949    
950        (1) The include/exclude patterns were being applied to the whole pathnames
951            of files, instead of just to the final components.
952    
953        (2) If there was more than one level of directory, the subdirectories were
954            skipped unless they satisfied the include/exclude conditions. This is
955            inconsistent with GNU grep (and could even be seen as contrary to the
956            pcregrep specification - which I improved to make it absolutely clear).
957            The action now is always to scan all levels of directory, and just
958            apply the include/exclude patterns to regular files.
959    
960    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
961        --exclude_dir in the tests to avoid scanning .svn directories.
962    
963    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
964        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
965        doesn't support NULs in patterns.
966    
967    7.  Added some missing "const"s to declarations of static tables in
968        pcre_compile.c and pcre_dfa_exec.c.
969    
970    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
971        caused by fix #2  above. (Subsequently also a second patch to fix the
972        first patch. And a third patch - this was a messy problem.)
973    
974    9.  Applied Craig's patch to remove the use of push_back().
975    
976    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
977        matching function regexec().
978    
979    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
980        which, however, unlike Perl's \g{...}, are subroutine calls, not back
981        references. PCRE supports relative numbers with this syntax (I don't think
982        Oniguruma does).
983    
984    12. Previously, a group with a zero repeat such as (...){0} was completely
985        omitted from the compiled regex. However, this means that if the group
986        was called as a subroutine from elsewhere in the pattern, things went wrong
987        (an internal error was given). Such groups are now left in the compiled
988        pattern, with a new opcode that causes them to be skipped at execution
989        time.
990    
991    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
992        to the way PCRE behaves:
993    
994        (a) A lone ] character is dis-allowed (Perl treats it as data).
995    
996        (b) A back reference to an unmatched subpattern matches an empty string
997            (Perl fails the current match path).
998    
999        (c) A data ] in a character class must be notated as \] because if the
1000            first data character in a class is ], it defines an empty class. (In
1001            Perl it is not possible to have an empty class.) The empty class []
1002            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1003            The negative empty class [^] matches any one character, independently
1004            of the DOTALL setting.
1005    
1006    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1007        non-existent subpattern following a character class starting with ']' and
1008        containing () gave an internal compiling error instead of "reference to
1009        non-existent subpattern". Fortunately, when the pattern did exist, the
1010        compiled code was correct. (When scanning forwards to check for the
1011        existencd of the subpattern, it was treating the data ']' as terminating
1012        the class, so got the count wrong. When actually compiling, the reference
1013        was subsequently set up correctly.)
1014    
1015    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1016        it was being rejected as not supported by pcre_dfa_exec(), even though
1017        other assertions are supported. I have made pcre_dfa_exec() support
1018        (*FAIL).
1019    
1020    16. The implementation of 13c above involved the invention of a new opcode,
1021        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1022        cannot be changed at match time, I realized I could make a small
1023        improvement to matching performance by compiling OP_ALLANY instead of
1024        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1025        on the OP_ANY path.
1026    
1027    17. Compiling pcretest on Windows with readline support failed without the
1028        following two fixes: (1) Make the unistd.h include conditional on
1029        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1030    
1031    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1032        ncurses library to be included for pcretest when ReadLine support is
1033        requested, but also to allow for it to be overridden. This patch came from
1034        Daniel Bergström.
1035    
1036    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1037        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1038        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1039        spotting this.
1040    
1041    
1042    Version 7.6 28-Jan-08
1043    ---------------------
1044    
1045    1.  A character class containing a very large number of characters with
1046        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1047        overflow.
1048    
1049    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1050        HAVE_LONG_LONG is not defined.
1051    
1052    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1053        bring them up to date and include new features. This patch includes:
1054    
1055        - Fixed PH's badly added libz and libbz2 support.
1056        - Fixed a problem with static linking.
1057        - Added pcredemo. [But later removed - see 7 below.]
1058        - Fixed dftables problem and added an option.
1059        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1060            HAVE_LONG_LONG.
1061        - Added readline support for pcretest.
1062        - Added an listing of the option settings after cmake has run.
1063    
1064    4.  A user submitted a patch to Makefile that makes it easy to create
1065        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1066        Makefile.am that cause it to include this special target, without
1067        affecting anything else. Note that the same mingw target plus all
1068        the other distribution libraries and programs are now supported
1069        when configuring with CMake (see 6 below) instead of with
1070        Configure/Make.
1071    
1072    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1073        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1074        exported in the Windows port". It has not yet been confirmed that the patch
1075        solves the problem, but it does no harm.
1076    
1077    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1078        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1079        with CMake, and also correct the comment about stack recursion.
1080    
1081    7.  Remove the automatic building of pcredemo from the ./configure system and
1082        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1083        of a program that users should build themselves after PCRE is installed, so
1084        building it automatically is not really right. What is more, it gave
1085        trouble in some build environments.
1086    
1087    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1088    
1089    
1090    Version 7.5 10-Jan-08
1091    ---------------------
1092    
1093    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1094        values in parens when parsing an RE using the C++ wrapper."
1095    
1096    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1097        Characters greater than 255 were excluded from the class instead of being
1098        included.
1099    
1100    3.  The same bug as (2) above applied to negated POSIX classes such as
1101        [:^space:].
1102    
1103    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1104        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1105        I have changed it.
1106    
1107    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1108        first named subpattern) and a construct such as (?&a) would reference the
1109        first named subpattern whose name started with "a" (in other words, the
1110        length check was missing). Both these problems are fixed. "Subpattern name
1111        expected" is now given for (?&) (a zero-length name), and this patch also
1112        makes it give the same error for \k'' (previously it complained that that
1113        was a reference to a non-existent subpattern).
1114    
1115    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1116        this is right because (?- can be followed by option settings as well as by
1117        digits. I have, however, made the messages clearer.
1118    
1119    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1120        than the number used in the conditional) now cause a compile-time error.
1121        This is actually not compatible with Perl, which accepts such patterns, but
1122        treats the conditional as always being FALSE (as PCRE used to), but it
1123        seems to me that giving a diagnostic is better.
1124    
1125    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1126        and messages.
1127    
1128    9.  Fix two occurrences of "backslash" in comments that should have been
1129        "backspace".
1130    
1131    10. Remove two redundant lines of code that can never be obeyed (their function
1132        was moved elsewhere).
1133    
1134    11. The program that makes PCRE's Unicode character property table had a bug
1135        which caused it to generate incorrect table entries for sequences of
1136        characters that have the same character type, but are in different scripts.
1137        It amalgamated them into a single range, with the script of the first of
1138        them. In other words, some characters were in the wrong script. There were
1139        thirteen such cases, affecting characters in the following ranges:
1140    
1141          U+002b0 - U+002c1
1142          U+0060c - U+0060d
1143          U+0061e - U+00612
1144          U+0064b - U+0065e
1145          U+0074d - U+0076d
1146          U+01800 - U+01805
1147          U+01d00 - U+01d77
1148          U+01d9b - U+01dbf
1149          U+0200b - U+0200f
1150          U+030fc - U+030fe
1151          U+03260 - U+0327f
1152          U+0fb46 - U+0fbb1
1153          U+10450 - U+1049d
1154    
1155    12. The -o option (show only the matching part of a line) for pcregrep was not
1156        compatible with GNU grep in that, if there was more than one match in a
1157        line, it showed only the first of them. It now behaves in the same way as
1158        GNU grep.
1159    
1160    13. If the -o and -v options were combined for pcregrep, it printed a blank
1161        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1162        does the same. The return code can be used to tell if there were any
1163        non-matching lines.
1164    
1165    14. Added --file-offsets and --line-offsets to pcregrep.
1166    
1167    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1168        infinitely looping recursion. The bug was that positive lookaheads were not
1169        being skipped when checking for a possible empty match (negative lookaheads
1170        and both kinds of lookbehind were skipped).
1171    
1172    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1173        inclusion of <windows.h> to before rather than after the definition of
1174        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1175    
1176    17. Specifying a possessive quantifier with a specific limit for a Unicode
1177        character property caused pcre_compile() to compile bad code, which led at
1178        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1179        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1180        caused the error; without that there was no problem.
1181    
1182    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1183    
1184    19. Added --enable-pcretest-libreadline.
1185    
1186    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1187        RE::GlobalReplace(). As a result, the number of replacements returned was
1188        double what it should be. I removed one of the increments, but Craig sent a
1189        later patch that removed the other one (the right fix) and added unit tests
1190        that check the return values (which was not done before).
1191    
1192    21. Several CMake things:
1193    
1194        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1195            the names libpcre and libpcreposix, not just pcre and pcreposix.
1196    
1197        (2) The above change means that pcretest and pcregrep are now correctly
1198            linked with the newly-built libraries, not previously installed ones.
1199    
1200        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1201    
1202    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1203        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1204        UTF-8 newline character). The key issue is that the pattern starts .*;
1205        this means that the match must be either at the beginning, or after a
1206        newline. The bug was in the code for advancing after a failed match and
1207        checking that the new position followed a newline. It was not taking
1208        account of UTF-8 characters correctly.
1209    
1210    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1211        character classes. PCRE was not treating the sequence [:...:] as a
1212        character class unless the ... were all letters. Perl, however, seems to
1213        allow any characters between [: and :], though of course it rejects as
1214        unknown any "names" that contain non-letters, because all the known class
1215        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1216        for example, whereas PCRE did not - it did not recognize a POSIX character
1217        class. This seemed a bit dangerous, so the code has been changed to be
1218        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1219        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1220        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1221        Perl does, and where it didn't before.
1222    
1223    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1224        Windows environments %n is disabled by default.
1225    
1226    
1227    Version 7.4 21-Sep-07
1228    ---------------------
1229    
1230    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1231        means that a class such as [\s] counted as "explicit reference to CR or
1232        LF". That isn't really right - the whole point of the change was to try to
1233        help when there was an actual mention of one of the two characters. So now
1234        the change happens only if \r or \n (or a literal CR or LF) character is
1235        encountered.
1236    
1237    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1238        of both had grown to the point where there were only 3 bits left.
1239        Fortunately, there was spare space in the data structure, and so I have
1240        moved the internal flags into a new 16-bit field to free up more option
1241        bits.
1242    
1243    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1244        but did not set the internal JCHANGED flag - either of these is enough to
1245        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1246        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1247        start sets both bits.
1248    
1249    4.  Added options (at build time, compile time, exec time) to change \R from
1250        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1251    
1252    5.  doc/pcresyntax.html was missing from the distribution.
1253    
1254    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1255        compatibility, even though it is no longer used.
1256    
1257    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1258        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1259        windows.h file is present (where different names are used). [This was
1260        reversed later after testing - see 16 below.]
1261    
1262    8.  Changed all #include <config.h> to #include "config.h". There were also
1263        some further <pcre.h> cases that I changed to "pcre.h".
1264    
1265    9.  When pcregrep was used with the --colour option, it missed the line ending
1266        sequence off the lines that it output.
1267    
1268    10. It was pointed out to me that arrays of string pointers cause lots of
1269        relocations when a shared library is dynamically loaded. A technique of
1270        using a single long string with a table of offsets can drastically reduce
1271        these. I have refactored PCRE in four places to do this. The result is
1272        dramatic:
1273    
1274          Originally:                          290
1275          After changing UCP table:            187
1276          After changing error message table:   43
1277          After changing table of "verbs"       36
1278          After changing table of Posix names   22
1279    
1280        Thanks to the folks working on Gregex for glib for this insight.
1281    
1282    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1283        unicode-properties was also set.
1284    
1285    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1286    
1287    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1288        checked only for CRLF.
1289    
1290    14. Added casts to pcretest.c to avoid compiler warnings.
1291    
1292    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1293    
1294    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1295        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1296        entirely. This removes changes made in 7 above.
1297    
1298    17. The CMake files have been updated, and there is now more information about
1299        building with CMake in the NON-UNIX-USE document.
1300    
1301    
1302    Version 7.3 28-Aug-07
1303    ---------------------
1304    
1305     1. In the rejigging of the build system that eventually resulted in 7.1, the
1306        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1307        brackets there is not right, since it causes compilers to look for an
1308        installed pcre.h, not the version that is in the source that is being
1309        compiled (which of course may be different). I have changed it back to:
1310    
1311          #include "pcre.h"
1312    
1313        I have a vague recollection that the change was concerned with compiling in
1314        different directories, but in the new build system, that is taken care of
1315        by the VPATH setting the Makefile.
1316    
1317     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1318        when the subject happened to end in the byte 0x85 (e.g. if the last
1319        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1320        characters but of course it shouldn't be taken as a newline when it is part
1321        of another character. The bug was that, for an unlimited repeat of . in
1322        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1323        characters when looking for a newline.
1324    
1325     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1326    
1327     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1328        in debug output.
1329    
1330     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1331        long printing in the pcrecpp unittest when running under MinGW.
1332    
1333     6. ESC_K was left out of the EBCDIC table.
1334    
1335     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1336        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1337        limit also applies to "virtual nesting" when a pattern is recursive, and in
1338        this case 1000 isn't so big. I have been able to remove this limit at the
1339        expense of backing off one optimization in certain circumstances. Normally,
1340        when pcre_exec() would call its internal match() function recursively and
1341        immediately return the result unconditionally, it uses a "tail recursion"
1342        feature to save stack. However, when a subpattern that can match an empty
1343        string has an unlimited repetition quantifier, it no longer makes this
1344        optimization. That gives it a stack frame in which to save the data for
1345        checking that an empty string has been matched. Previously this was taken
1346        from the 1000-entry workspace that had been reserved. So now there is no
1347        explicit limit, but more stack is used.
1348    
1349     8. Applied Daniel's patches to solve problems with the import/export magic
1350        syntax that is required for Windows, and which was going wrong for the
1351        pcreposix and pcrecpp parts of the library. These were overlooked when this
1352        problem was solved for the main library.
1353    
1354     9. There were some crude static tests to avoid integer overflow when computing
1355        the size of patterns that contain repeated groups with explicit upper
1356        limits. As the maximum quantifier is 65535, the maximum group length was
1357        set at 30,000 so that the product of these two numbers did not overflow a
1358        32-bit integer. However, it turns out that people want to use groups that
1359        are longer than 30,000 bytes (though not repeat them that many times).
1360        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1361        made it possible to implement the integer overflow checks in a much more
1362        dynamic way, which I have now done. The artificial limitation on group
1363        length has been removed - we now have only the limit on the total length of
1364        the compiled pattern, which depends on the LINK_SIZE setting.
1365    
1366    10. Fixed a bug in the documentation for get/copy named substring when
1367        duplicate names are permitted. If none of the named substrings are set, the
1368        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1369        empty string.
1370    
1371    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1372        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1373        because the ] is interpreted as the first data character and the
1374        terminating ] is not found. PCRE has been made compatible with Perl in this
1375        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1376        cause memory overwriting.
1377    
1378    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1379        string has been matched (to stop an infinite loop). It was not recognizing
1380        a conditional subpattern that could match an empty string if that
1381        subpattern was within another subpattern. For example, it looped when
1382        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1383        condition was not nested. This bug has been fixed.
1384    
1385    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1386        past the start of the subject in the presence of bytes with the top bit
1387        set, for example "\x8aBCD".
1388    
1389    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1390        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1391    
1392    14. Optimized (?!) to (*FAIL).
1393    
1394    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1395        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1396        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1397        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1398        does: it's just the validity check that is more restrictive.
1399    
1400    16. Inserted checks for integer overflows during escape sequence (backslash)
1401        processing, and also fixed erroneous offset values for syntax errors during
1402        backslash processing.
1403    
1404    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1405        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1406    
1407    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1408        caused an overrun.
1409    
1410    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1411        something other than just ASCII characters) inside a group that had an
1412        unlimited repeat caused a loop at compile time (while checking to see
1413        whether the group could match an empty string).
1414    
1415    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1416        [\P{Any}] did so. (Error in the code for printing property names.)
1417    
1418    21. An orphan \E inside a character class could cause a crash.
1419    
1420    22. A repeated capturing bracket such as (A)? could cause a wild memory
1421        reference during compilation.
1422    
1423    23. There are several functions in pcre_compile() that scan along a compiled
1424        expression for various reasons (e.g. to see if it's fixed length for look
1425        behind). There were bugs in these functions when a repeated \p or \P was
1426        present in the pattern. These operators have additional parameters compared
1427        with \d, etc, and these were not being taken into account when moving along
1428        the compiled data. Specifically:
1429    
1430        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1431            length.
1432    
1433        (b) An item such as \pL+ within a repeated group could cause crashes or
1434            loops.
1435    
1436        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1437            "reference to non-existent subpattern" error.
1438    
1439        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1440    
1441    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1442        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1443    
1444    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1445    
1446    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1447        character were causing crashes (broken optimization).
1448    
1449    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1450        \p or \P) caused a compile-time loop.
1451    
1452    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1453        break. For example, the unstudied pattern [\r\n]A does not match the string
1454        "\r\nA" because change 7.0/46 below moves the current point on by two
1455        characters after failing to match at the start. However, the pattern \nA
1456        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1457        the same is true. There doesn't seem any very clean way out of this, but
1458        what I have chosen to do makes the common cases work: PCRE now takes note
1459        of whether there can be an explicit match for \r or \n anywhere in the
1460        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1461        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1462        pattern has explicit CR or LF references.
1463    
1464    29. Added (*CR) etc for changing newline setting at start of pattern.
1465    
1466    
1467    Version 7.2 19-Jun-07
1468    ---------------------
1469    
1470     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1471        which is apparently normally available under Windows.
1472    
1473     2. Re-jig the pcregrep tests with different newline settings in an attempt
1474        to make them independent of the local environment's newline setting.
1475    
1476     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1477    
1478     4. Some of the "internals" tests were previously cut out when the link size
1479        was not 2, because the output contained actual offsets. The recent new
1480        "Z" feature of pcretest means that these can be cut out, making the tests
1481        usable with all link sizes.
1482    
1483     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1484        stack recursion. This gives a massive performance boost under BSD, but just
1485        a small improvement under Linux. However, it saves one field in the frame
1486        in all cases.
1487    
1488     6. Added more features from the forthcoming Perl 5.10:
1489    
1490        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1491            recursion call. It refers to the nth most recently opened parentheses.
1492    
1493        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1494            to be opened parentheses.
1495    
1496        (c) Conditions that refer to capturing parentheses can be specified
1497            relatively, for example, (?(-2)... or (?(+3)...
1498    
1499        (d) \K resets the start of the current match so that everything before
1500            is not part of it.
1501    
1502        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1503    
1504        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1505            reference syntax.
1506    
1507        (g) (?| introduces a group in which the numbering of parentheses in each
1508            alternative starts with the same number.
1509    
1510        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1511    
1512     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1513        PCRE_INFO_JCHANGED.
1514    
1515     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1516        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1517        for detecting groups that can match an empty string.
1518    
1519     9. A pattern with a very large number of alternatives (more than several
1520        hundred) was running out of internal workspace during the pre-compile
1521        phase, where pcre_compile() figures out how much memory will be needed. A
1522        bit of new cunning has reduced the workspace needed for groups with
1523        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1524        workspace instead of running out of the 4096 that are available.
1525    
1526    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1527    
1528    11. Applied patch from Google to remove an optimization that didn't quite work.
1529        The report of the bug said:
1530    
1531          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1532          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1533          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1534    
1535    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1536        it matched the wrong number of bytes.
1537    
1538    
1539    Version 7.1 24-Apr-07
1540    ---------------------
1541    
1542     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1543        that is more "standard", making use of automake and other Autotools. There
1544        is some re-arrangement of the files and adjustment of comments consequent
1545        on this.
1546    
1547     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1548        for recursive directory scanning broke on some systems because the files
1549        are not scanned in any specific order and on different systems the order
1550        was different. A call to "sort" has been inserted into RunGrepTest for the
1551        approprate test as a short-term fix. In the longer term there may be an
1552        alternative.
1553    
1554     3. I had an email from Eric Raymond about problems translating some of PCRE's
1555        man pages to HTML (despite the fact that I distribute HTML pages, some
1556        people do their own conversions for various reasons). The problems
1557        concerned the use of low-level troff macros .br and .in. I have therefore
1558        removed all such uses from the man pages (some were redundant, some could
1559        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1560        HTML has been updated to handle .nf/.fi and to complain if it encounters
1561        .br or .in.
1562    
1563     4. Updated comments in configure.ac that get placed in config.h.in and also
1564        arranged for config.h to be included in the distribution, with the name
1565        config.h.generic, for the benefit of those who have to compile without
1566        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1567    
1568     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1569        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1570        makevp.bat for the current PCRE, using the additional files
1571        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1572    
1573     6. A Windows user reported a minor discrepancy with test 2, which turned out
1574        to be caused by a trailing space on an input line that had got lost in his
1575        copy. The trailing space was an accident, so I've just removed it.
1576    
1577     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1578        that is needed.
1579    
1580     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1581        as "const" (a) because they are and (b) because it helps the PHP
1582        maintainers who have recently made a script to detect big data structures
1583        in the php code that should be moved to the .rodata section. I remembered
1584        to update Builducptable as well, so it won't revert if ucptable.h is ever
1585        re-created.
1586    
1587     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1588        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1589        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1590        support is not required. This saves 1.5-2K of code, which is important in
1591        some applications.
1592    
1593        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1594        so as not to refer to the tables, even though these functions will never be
1595        called when UTF-8 support is disabled. Otherwise there are problems with a
1596        shared library.
1597    
1598    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1599    
1600        (a) It was defining its arguments as char * instead of void *.
1601    
1602        (b) It was assuming that all moves were upwards in memory; this was true
1603            a long time ago when I wrote it, but is no longer the case.
1604    
1605        The emulated memove() is provided for those environments that have neither
1606        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1607        is clearly not the case, as these two bugs were recently reported.
1608    
1609    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1610        and Detrail to create the HTML documentation, the .txt form of the man
1611        pages, and it removes trailing spaces from listed files. It also creates
1612        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1613        case, it wraps all the #defines with #ifndefs. This script should be run
1614        before "make dist".
1615    
1616    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1617        with Unicode property support.
1618    
1619        (a) For a maximizing quantifier, if the two different cases of the
1620            character were of different lengths in their UTF-8 codings (there are
1621            some cases like this - I found 11), and the matching function had to
1622            back up over a mixture of the two cases, it incorrectly assumed they
1623            were both the same length.
1624    
1625        (b) When PCRE was configured to use the heap rather than the stack for
1626            recursion during matching, it was not correctly preserving the data for
1627            the other case of a UTF-8 character when checking ahead for a match
1628            while processing a minimizing repeat. If the check also involved
1629            matching a wide character, but failed, corruption could cause an
1630            erroneous result when trying to check for a repeat of the original
1631            character.
1632    
1633    13. Some tidying changes to the testing mechanism:
1634    
1635        (a) The RunTest script now detects the internal link size and whether there
1636            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1637            values substituted by "configure". (The RunGrepTest script already did
1638            this for UTF-8.) The configure.ac script no longer substitutes the
1639            relevant variables.
1640    
1641        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1642            with length and offset values. This means that the output is different
1643            for different internal link sizes. Test 2 is skipped for link sizes
1644            other than 2 because of this, bypassing the problem. Unfortunately,
1645            there was also a test in test 3 (the locale tests) that used /B and
1646            failed for link sizes other than 2. Rather than cut the whole test out,
1647            I have added a new /Z option to pcretest that replaces the length and
1648            offset values with spaces. This is now used to make test 3 independent
1649            of link size. (Test 2 will be tidied up later.)
1650    
1651    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1652        segmentation fault instead of returning the appropriate error message.
1653    
1654    15. In multiline mode when the newline sequence was set to "any", the pattern
1655        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1656        This doesn't seem right; it now treats the CRLF combination as the line
1657        ending, and so does not match in that case. It's only a pattern such as ^$
1658        that would hit this one: something like ^ABC$ would have failed after \r
1659        and then tried again after \r\n.
1660    
1661    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1662        in an attempt to make files that differ only in their line terminators
1663        compare equal. This works on Linux.
1664    
1665    17. Under certain error circumstances pcregrep might try to free random memory
1666        as it exited. This is now fixed, thanks to valgrind.
1667    
1668    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1669        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1670        was because its rules for how to advance for /g after matching an empty
1671        string at the end of a line did not allow for this case. They now check for
1672        it specially.
1673    
1674    20. pcretest is supposed to handle patterns and data of any length, by
1675        extending its buffers when necessary. It was getting this wrong when the
1676        buffer for a data line had to be extended.
1677    
1678    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1679        CRLF as a newline sequence.
1680    
1681    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1682        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1683        I have nevertheless tidied it up.
1684    
1685    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1686    
1687    24. Added a man page for pcre-config.
1688    
1689    
1690    Version 7.0 19-Dec-06
1691    ---------------------
1692    
1693     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1694        moving to gcc 4.1.1.
1695    
1696     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1697        sys/time.h, which is documented as needed for this function. It doesn't
1698        seem to matter on Linux, but it showed up on some releases of OS X.
1699    
1700     3. It seems that there are systems where bytes whose values are greater than
1701        127 match isprint() in the "C" locale. The "C" locale should be the
1702        default when a C program starts up. In most systems, only ASCII printing
1703        characters match isprint(). This difference caused the output from pcretest
1704        to vary, making some of the tests fail. I have changed pcretest so that:
1705    
1706        (a) When it is outputting text in the compiled version of a pattern, bytes
1707            other than 32-126 are always shown as hex escapes.
1708    
1709        (b) When it is outputting text that is a matched part of a subject string,
1710            it does the same, unless a different locale has been set for the match
1711            (using the /L modifier). In this case, it uses isprint() to decide.
1712    
1713     4. Fixed a major bug that caused incorrect computation of the amount of memory
1714        required for a compiled pattern when options that changed within the
1715        pattern affected the logic of the preliminary scan that determines the
1716        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1717        that the computed length was too small. The symptoms of this bug were
1718        either the PCRE error "internal error: code overflow" from pcre_compile(),
1719        or a glibc crash with a message such as "pcretest: free(): invalid next
1720        size (fast)". Examples of patterns that provoked this bug (shown in
1721        pcretest format) are:
1722    
1723          /(?-x: )/x
1724          /(?x)(?-x: \s*#\s*)/
1725          /((?i)[\x{c0}])/8
1726          /(?i:[\x{c0}])/8
1727    
1728        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1729        is now done differently.
1730    
1731     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1732        wrapper classes; (b) implement a new function in the C++ scanner that is
1733        more efficient than the old way of doing things because it avoids levels of
1734        recursion in the regex matching; (c) add a paragraph to the documentation
1735        for the FullMatch() function.
1736    
1737     6. The escape sequence \n was being treated as whatever was defined as
1738        "newline". Not only was this contrary to the documentation, which states
1739        that \n is character 10 (hex 0A), but it also went horribly wrong when
1740        "newline" was defined as CRLF. This has been fixed.
1741    
1742     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1743        was being set to -1 for the "end of line" case (supposedly a value that no
1744        character can have). Though this value is never used (the check for end of
1745        line is "zero bytes in current character"), it caused compiler complaints.
1746        I've changed it to 0xffffffff.
1747    
1748     8. In pcre_version.c, the version string was being built by a sequence of
1749        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1750        string (as it is for production releases) called a macro with an empty
1751        argument. The C standard says the result of this is undefined. The gcc
1752        compiler treats it as an empty string (which was what was wanted) but it is
1753        reported that Visual C gives an error. The source has been hacked around to
1754        avoid this problem.
1755    
1756     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1757        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1758        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1759        of them did).
1760    
1761    10. Originally, pcretest opened its input and output without "b"; then I was
1762        told that "b" was needed in some environments, so it was added for release
1763        5.0 to both the input and output. (It makes no difference on Unix-like
1764        systems.) Later I was told that it is wrong for the input on Windows. I've
1765        now abstracted the modes into two macros, to make it easier to fiddle with
1766        them, and removed "b" from the input mode under Windows.
1767    
1768    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1769    
1770    12. Added -help and --help to pcretest as an official way of being reminded
1771        of the options.
1772    
1773    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1774        and pcrecpp.cc because they annoy compilers at high warning levels.
1775    
1776    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1777    
1778    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1779        scripts are not C programs :-) and which was not noticed because it works
1780        on Linux.
1781    
1782    16. pcretest is supposed to handle any length of pattern and data line (as one
1783        line or as a continued sequence of lines) by extending its input buffer if
1784        necessary. This feature was broken for very long pattern lines, leading to
1785        a string of junk being passed to pcre_compile() if the pattern was longer
1786        than about 50K.
1787    
1788    17. I have done a major re-factoring of the way pcre_compile() computes the
1789        amount of memory needed for a compiled pattern. Previously, there was code
1790        that made a preliminary scan of the pattern in order to do this. That was
1791        OK when PCRE was new, but as the facilities have expanded, it has become
1792        harder and harder to keep it in step with the real compile phase, and there
1793        have been a number of bugs (see for example, 4 above). I have now found a
1794        cunning way of running the real compile function in a "fake" mode that
1795        enables it to compute how much memory it would need, while actually only
1796        ever using a few hundred bytes of working memory and without too many
1797        tests of the mode. This should make future maintenance and development
1798        easier. A side effect of this work is that the limit of 200 on the nesting
1799        depth of parentheses has been removed (though this was never a serious
1800        limitation, I suspect). However, there is a downside: pcre_compile() now
1801        runs more slowly than before (30% or more, depending on the pattern). I
1802        hope this isn't a big issue. There is no effect on runtime performance.
1803    
1804    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1805        newline (only possible for the last line of a file) and it was a
1806        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1807    
1808    19. Added additional timing features to pcretest. (1) The -tm option now times
1809        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1810        separate command line item, by a number that specifies the number of
1811        repeats to use when timing. The default is 50000; this gives better
1812        precision, but takes uncomfortably long for very large patterns.
1813    
1814    20. Extended pcre_study() to be more clever in cases where a branch of a
1815        subpattern has no definite first character. For example, (a*|b*)[cd] would
1816        previously give no result from pcre_study(). Now it recognizes that the
1817        first character must be a, b, c, or d.
1818    
1819    21. There was an incorrect error "recursive call could loop indefinitely" if
1820        a subpattern (or the entire pattern) that was being tested for matching an
1821        empty string contained only one non-empty item after a nested subpattern.
1822        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1823        incorrectly, because the \d was being skipped in the check.
1824    
1825    22. The pcretest program now has a new pattern option /B and a command line
1826        option -b, which is equivalent to adding /B to every pattern. This causes
1827        it to show the compiled bytecode, without the additional information that
1828        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1829        is the same as /B/I).
1830    
1831    23. A new optimization is now able automatically to treat some sequences such
1832        as a*b as a*+b. More specifically, if something simple (such as a character
1833        or a simple class like \d) has an unlimited quantifier, and is followed by
1834        something that cannot possibly match the quantified thing, the quantifier
1835        is automatically "possessified".
1836    
1837    24. A recursive reference to a subpattern whose number was greater than 39
1838        went wrong under certain circumstances in UTF-8 mode. This bug could also
1839        have affected the operation of pcre_study().
1840    
1841    25. Realized that a little bit of performance could be had by replacing
1842        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1843    
1844    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1845    
1846    27. Possessive quantifiers such as a++ were previously implemented by turning
1847        them into atomic groups such as ($>a+). Now they have their own opcodes,
1848        which improves performance. This includes the automatically created ones
1849        from 23 above.
1850    
1851    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1852        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1853        the first matched character to be a colon. This applied both to named and
1854        numbered groups.
1855    
1856    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1857    
1858    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1859        building PCRE on Windows easier, so I have included it in the distribution.
1860    
1861    31. There is now a check in pcretest against a ridiculously large number being
1862        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1863        loop, the loop is abandoned.
1864    
1865    32. Forward references to subpatterns in conditions such as (?(2)...) where
1866        subpattern 2 is defined later cause pcre_compile() to search forwards in
1867        the pattern for the relevant set of parentheses. This search went wrong
1868        when there were unescaped parentheses in a character class, parentheses
1869        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1870    
1871    33. "Subroutine" calls and backreferences were previously restricted to
1872        referencing subpatterns earlier in the regex. This restriction has now
1873        been removed.
1874    
1875    34. Added a number of extra features that are going to be in Perl 5.10. On the
1876        whole, these are just syntactic alternatives for features that PCRE had
1877        previously implemented using the Python syntax or my own invention. The
1878        other formats are all retained for compatibility.
1879    
1880        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1881            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1882            also .NET compatible.
1883    
1884        (b) A recursion or subroutine call to a named group can now be defined as
1885            (?&name) as well as (?P>name).
1886    
1887        (c) A backreference to a named group can now be defined as \k<name> or
1888            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1889            5.10, are also .NET compatible.
1890    
1891        (d) A conditional reference to a named group can now use the syntax
1892            (?(<name>) or (?('name') as well as (?(name).
1893    
1894        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1895            groups (named and numbered) that are never evaluated inline, but can be
1896            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1897            is always false. There may be only one alternative in such a group.
1898    
1899        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1900            as the simple (?(R). The condition is true only if the most recent
1901            recursion is that of the given number or name. It does not search out
1902            through the entire recursion stack.
1903    
1904        (g) The escape \gN or \g{N} has been added, where N is a positive or
1905            negative number, specifying an absolute or relative reference.
1906    
1907    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1908        some "unreachable code" warnings.
1909    
1910    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1911        things, this adds five new scripts.
1912    
1913    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1914        There were also incompatibilities regarding the handling of \Q..\E inside
1915        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1916        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1917    
1918    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1919        matches an empty string, and forcibly breaks the loop. There were bugs in
1920        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1921        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1922        separate and independent bugs (that affected different cases) have been
1923        fixed.
1924    
1925    39. Refactored the code to abolish the use of different opcodes for small
1926        capturing bracket numbers. This is a tidy that I avoided doing when I
1927        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1928        The new approach is not only tidier, it makes it possible to reduce the
1929        memory needed to fix the previous bug (38).
1930    
1931    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1932        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1933        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1934        mode.
1935    
1936    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1937        report.
1938    
1939    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1940        copy construction and assignment in the C++ wrapper.
1941    
1942    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1943        couple of bugs that could have given wrong results in the "--newline=crlf"
1944        case.
1945    
1946    44. Added a number of casts and did some reorganization of signed/unsigned int
1947        variables following suggestions from Dair Grant. Also renamed the variable
1948        "this" as "item" because it is a C++ keyword.
1949    
1950    45. Arranged for dftables to add
1951    
1952          #include "pcre_internal.h"
1953    
1954        to pcre_chartables.c because without it, gcc 4.x may remove the array
1955        definition from the final binary if PCRE is built into a static library and
1956        dead code stripping is activated.
1957    
1958    46. For an unanchored pattern, if a match attempt fails at the start of a
1959        newline sequence, and the newline setting is CRLF or ANY, and the next two
1960        characters are CRLF, advance by two characters instead of one.
1961    
1962    
1963    Version 6.7 04-Jul-06
1964    ---------------------
1965    
1966     1. In order to handle tests when input lines are enormously long, pcretest has
1967        been re-factored so that it automatically extends its buffers when
1968        necessary. The code is crude, but this _is_ just a test program. The
1969        default size has been increased from 32K to 50K.
1970    
1971     2. The code in pcre_study() was using the value of the re argument before
1972        testing it for NULL. (Of course, in any sensible call of the function, it
1973        won't be NULL.)
1974    
1975     3. The memmove() emulation function in pcre_internal.h, which is used on
1976        systems that lack both memmove() and bcopy() - that is, hardly ever -
1977        was missing a "static" storage class specifier.
1978    
1979     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1980        containing an extended class (one that cannot be represented by a bitmap
1981        because it contains high-valued characters or Unicode property items, e.g.
1982        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1983        pattern, but PCRE should not loop if you do not (it no longer does).
1984        [Detail: two cases were found: (a) a repeated subpattern containing an
1985        extended class; (b) a recursive reference to a subpattern that followed a
1986        previous extended class. It wasn't skipping over the extended class
1987        correctly when UTF-8 mode was not set.]
1988    
1989     5. A negated single-character class was not being recognized as fixed-length
1990        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1991        compile error "lookbehind assertion is not fixed length".
1992    
1993     6. The RunPerlTest auxiliary script was showing an unexpected difference
1994        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1995        write a Perl script that can interpret lines of an input file either as
1996        byte characters or as UTF-8, which is what "perltest" was being required to
1997        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1998        can't do is switch easily at run time between having the "use utf8;" pragma
1999        or not. In the end, I fudged it by using the RunPerlTest script to insert
2000        "use utf8;" explicitly for the UTF-8 tests.
2001    
2002     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2003        the end of the subject string, contrary to the documentation and to what
2004        Perl does. This was true of both matching functions. Now it matches only at
2005        the start of the subject and immediately after *internal* newlines.
2006    
2007     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2008        a pointer to an int instead of a pointer to an unsigned long int. This
2009        caused problems on 64-bit systems.
2010    
2011     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2012        instance of the 'standard' template library not being so standard".
2013    
2014    10. There was no check on the number of named subpatterns nor the maximum
2015        length of a subpattern name. The product of these values is used to compute
2016        the size of the memory block for a compiled pattern. By supplying a very
2017        long subpattern name and a large number of named subpatterns, the size
2018        computation could be caused to overflow. This is now prevented by limiting
2019        the length of names to 32 characters, and the number of named subpatterns
2020        to 10,000.
2021    
2022    11. Subpatterns that are repeated with specific counts have to be replicated in
2023        the compiled pattern. The size of memory for this was computed from the
2024        length of the subpattern and the repeat count. The latter is limited to
2025        65535, but there was no limit on the former, meaning that integer overflow
2026        could in principle occur. The compiled length of a repeated subpattern is
2027        now limited to 30,000 bytes in order to prevent this.
2028    
2029    12. Added the optional facility to have named substrings with the same name.
2030    
2031    13. Added the ability to use a named substring as a condition, using the
2032        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2033        are numbers (not recommended). Forward references are permitted.
2034    
2035    14. Added forward references in named backreferences (if you see what I mean).
2036    
2037    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2038        pattern could run off the end of the subject. For example, the pattern
2039        "(?s)(.{1,5})"8 did this with the subject "ab".
2040    
2041    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2042        PCRE_CASELESS was set when matching characters that were quantified with ?
2043        or *.
2044    
2045    17. A character class other than a single negated character that had a minimum
2046        but no maximum quantifier - for example [ab]{6,} - was not handled
2047        correctly by pce_dfa_exec(). It would match only one character.
2048    
2049    18. A valid (though odd) pattern that looked like a POSIX character
2050        class but used an invalid character after [ (for example [[,abc,]]) caused
2051        pcre_compile() to give the error "Failed: internal error: code overflow" or
2052        in some cases to crash with a glibc free() error. This could even happen if
2053        the pattern terminated after [[ but there just happened to be a sequence of
2054        letters, a binary zero, and a closing ] in the memory that followed.
2055    
2056    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2057        over the years. Originally (before any Unicode support), just the bottom 8
2058        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2059        output from "man perlunicode" includes this:
2060    
2061          The regular expression compiler produces polymorphic opcodes.  That
2062          is, the pattern adapts to the data and automatically switches to
2063          the Unicode character scheme when presented with Unicode data--or
2064          instead uses a traditional byte scheme when presented with byte
2065          data.
2066    
2067        Sadly, a wide octal escape does not cause a switch, and in a string with
2068        no other multibyte characters, these octal escapes are treated as before.
2069        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2070        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2071        Unicode string.
2072    
2073        I have not perpetrated such confusion in PCRE. Up till now, it took just
2074        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2075        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2076        translate to the appropriate multibyte character.
2077    
2078    29. Applied some refactoring to reduce the number of warnings from Microsoft
2079        and Borland compilers. This has included removing the fudge introduced
2080        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2081        a warning about an unused variable.
2082    
2083    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2084        characters since release 4.0, because Perl (from release 5.004) does not.
2085        [Or at least, is documented not to: some releases seem to be in conflict
2086        with the documentation.] However, when a pattern was studied with
2087        pcre_study() and all its branches started with \s, PCRE still included VT
2088        as a possible starting character. Of course, this did no harm; it just
2089        caused an unnecessary match attempt.
2090    
2091    22. Removed a now-redundant internal flag bit that recorded the fact that case
2092        dependency changed within the pattern. This was once needed for "required
2093        byte" processing, but is no longer used. This recovers a now-scarce options
2094        bit. Also moved the least significant internal flag bit to the most-
2095        significant bit of the word, which was not previously used (hangover from
2096        the days when it was an int rather than a uint) to free up another bit for
2097        the future.
2098    
2099    23. Added support for CRLF line endings as well as CR and LF. As well as the
2100        default being selectable at build time, it can now be changed at runtime
2101        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2102        specify that it is scanning data with non-default line endings.
2103    
2104    24. Changed the definition of CXXLINK to make it agree with the definition of
2105        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2106    
2107    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2108        recursions. This makes a big different to stack usage for some patterns.
2109    
2110    26. If a subpattern containing a named recursion or subroutine reference such
2111        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2112        the space required for the compiled pattern went wrong and gave too small a
2113        value. Depending on the environment, this could lead to "Failed: internal
2114        error: code overflow at offset 49" or "glibc detected double free or
2115        corruption" errors.
2116    
2117    27. Applied patches from Google (a) to support the new newline modes and (b) to
2118        advance over multibyte UTF-8 characters in GlobalReplace.
2119    
2120    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2121        difference for some implementation of PCRE in some Windows version.
2122    
2123    29. Added some extra testing facilities to pcretest:
2124    
2125        \q<number>   in a data line sets the "match limit" value
2126        \Q<number>   in a data line sets the "match recursion limt" value
2127        -S <number>  sets the stack size, where <number> is in megabytes
2128    
2129        The -S option isn't available for Windows.
2130    
2131    
2132    Version 6.6 06-Feb-06
2133    ---------------------
2134    
2135     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2136        in pcreposix.h. I have copied the definition from pcre.h.
2137    
2138     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2139        because pcre.h is no longer a built file.
2140    
2141     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2142        not normally included in the compiled code.
2143    
2144    
2145    Version 6.5 01-Feb-06
2146    ---------------------
2147    
2148     1. When using the partial match feature with pcre_dfa_exec(), it was not
2149        anchoring the second and subsequent partial matches at the new starting
2150        point. This could lead to incorrect results. For example, with the pattern
2151        /1234/, partially matching against "123" and then "a4" gave a match.
2152    
2153     2. Changes to pcregrep:
2154    
2155        (a) All non-match returns from pcre_exec() were being treated as failures
2156            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2157            error message is output. Some extra information is given for the
2158            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2159            probably the only errors that are likely to be caused by users (by
2160            specifying a regex that has nested indefinite repeats, for instance).
2161            If there are more than 20 of these errors, pcregrep is abandoned.
2162    
2163        (b) A binary zero was treated as data while matching, but terminated the
2164            output line if it was written out. This has been fixed: binary zeroes
2165            are now no different to any other data bytes.
2166    
2167        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2168            used to set a locale for matching. The --locale=xxxx long option has
2169            been added (no short equivalent) to specify a locale explicitly on the
2170            pcregrep command, overriding the environment variables.
2171    
2172        (d) When -B was used with -n, some line numbers in the output were one less
2173            than they should have been.
2174    
2175        (e) Added the -o (--only-matching) option.
2176    
2177        (f) If -A or -C was used with -c (count only), some lines of context were
2178            accidentally printed for the final match.
2179    
2180        (g) Added the -H (--with-filename) option.
2181    
2182        (h) The combination of options -rh failed to suppress file names for files
2183            that were found from directory arguments.
2184    
2185        (i) Added the -D (--devices) and -d (--directories) options.
2186    
2187        (j) Added the -F (--fixed-strings) option.
2188    
2189        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2190    
2191        (l) Added the --colo(u)r option.
2192    
2193        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2194            is not present by default.
2195    
2196     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2197        items such as (?R) or (?1), when the recursion could match a number of
2198        alternatives. If it matched one of the alternatives, but subsequently,
2199        outside the recursion, there was a failure, the code tried to back up into
2200        the recursion. However, because of the way PCRE is implemented, this is not
2201        possible, and the result was an incorrect result from the match.
2202    
2203        In order to prevent this happening, the specification of recursion has
2204        been changed so that all such subpatterns are automatically treated as
2205        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2206    
2207     4. I had overlooked the fact that, in some locales, there are characters for
2208        which isalpha() is true but neither isupper() nor islower() are true. In
2209        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2210        and ordfeminine) are like this. This affected the treatment of \w and \W
2211        when they appeared in character classes, but not when they appeared outside
2212        a character class. The bit map for "word" characters is now created
2213        separately from the results of isalnum() instead of just taking it from the
2214        upper, lower, and digit maps. (Plus the underscore character, of course.)
2215    
2216     5. The above bug also affected the handling of POSIX character classes such as
2217        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2218        permanent tables. Instead, the bit maps for such a class were previously
2219        created as the appropriate unions of the upper, lower, and digit bitmaps.
2220        Now they are created by subtraction from the [[:word:]] class, which has
2221        its own bitmap.
2222    
2223     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2224        It is created by subtracting the vertical space characters (\x09, \x0a,
2225        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2226        subtraction was done in the overall bitmap for a character class, meaning
2227        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2228        be recognized. This bug has been fixed.
2229    
2230     7. Patches from the folks at Google:
2231    
2232          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2233          real life, but is still worth protecting against".
2234    
2235          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2236          regular expressions".
2237    
2238          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2239          have it.
2240    
2241          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2242          "configure" and the latter not, in order to fix a problem somebody had
2243          with compiling the Arg class on HP-UX.
2244    
2245          (e) Improve the error-handling of the C++ wrapper a little bit.
2246    
2247          (f) New tests for checking recursion limiting.
2248    
2249     8. The pcre_memmove() function, which is used only if the environment does not
2250        have a standard memmove() function (and is therefore rarely compiled),
2251        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2252        returning a result (though PCRE never actually uses the result).
2253    
2254     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2255        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2256        returned instead of calling malloc() with an overflowing number that would
2257        most likely cause subsequent chaos.
2258    
2259    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2260    
2261    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2262        with this option is matched, the nmatch and pmatch options of regexec() are
2263        ignored.
2264    
2265    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2266        provided in case anyone wants to the the POSIX interface with UTF-8
2267        strings.
2268    
2269    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2270        C++ linking (needed for some HP-UX environments).
2271    
2272    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2273        (unused parameter) and in the pcre_printint() function (omitted "default"
2274        switch label when the default is to do nothing).
2275    
2276    15. Added some code to make it possible, when PCRE is compiled as a C++
2277        library, to replace subject pointers for pcre_exec() with a smart pointer
2278        class, thus making it possible to process discontinuous strings.
2279    
2280    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2281        much the same function. They were added by different people who were trying
2282        to make PCRE easy to compile on non-Unix systems. It has been suggested
2283        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2284        for compiling on Windows systems. I have therefore replaced it with
2285        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2286        defaults to "extern" for C or "extern C" for C++, which works fine on
2287        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2288        SCOPE with something explicit in config.h. In addition:
2289    
2290        (a) pcreposix.h still had just "extern" instead of either of these macros;
2291            I have replaced it with PCRE_DATA_SCOPE.
2292    
2293        (b) Functions such as _pcre_xclass(), which are internal to the library,
2294            but external in the C sense, all had PCRE_EXPORT in their definitions.
2295            This is apparently wrong for the Windows case, so I have removed it.
2296            (It makes no difference on Unix-like systems.)
2297    
2298    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2299        of recursive calls to match(). This is different to MATCH_LIMIT because
2300        that limits the total number of calls to match(), not all of which increase
2301        the depth of recursion. Limiting the recursion depth limits the amount of
2302        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2303        when PCRE is compiled, and changed at run time. A patch from Google adds
2304        this functionality to the C++ interface.
2305    
2306    18. Changes to the handling of Unicode character properties:
2307    
2308        (a) Updated the table to Unicode 4.1.0.
2309    
2310        (b) Recognize characters that are not in the table as "Cn" (undefined).
2311    
2312        (c) I revised the way the table is implemented to a much improved format
2313            which includes recognition of ranges. It now supports the ranges that
2314            are defined in UnicodeData.txt, and it also amalgamates other
2315            characters into ranges. This has reduced the number of entries in the
2316            table from around 16,000 to around 3,000, thus reducing its size
2317            considerably. I realized I did not need to use a tree structure after
2318            all - a binary chop search is just as efficient. Having reduced the
2319            number of entries, I extended their size from 6 bytes to 8 bytes to
2320            allow for more data.
2321    
2322        (d) Added support for Unicode script names via properties such as \p{Han}.
2323    
2324    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2325        matching that character.
2326    
2327    20. When matching a repeated Unicode property with a minimum greater than zero,
2328        (for example \pL{2,}), PCRE could look past the end of the subject if it
2329        reached it while seeking the minimum number of characters. This could
2330        happen only if some of the characters were more than one byte long, because
2331        there is a check for at least the minimum number of bytes.
2332    
2333    21. Refactored the implementation of \p and \P so as to be more general, to
2334        allow for more different types of property in future. This has changed the
2335        compiled form incompatibly. Anybody with saved compiled patterns that use
2336        \p or \P will have to recompile them.
2337    
2338    22. Added "Any" and "L&" to the supported property types.
2339    
2340    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2341        but give a compile time error if the value is greater than 0xff.
2342    
2343    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2344        accidentally not being installed or uninstalled.
2345    
2346    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2347        made were to insert the current release number. This seemed silly, because
2348        it made things harder for people building PCRE on systems that don't run
2349        "configure". I have turned pcre.h into a distributed file, no longer built
2350        by "configure", with the version identification directly included. There is
2351        no longer a pcre.h.in file.
2352    
2353        However, this change necessitated a change to the pcre-config script as
2354        well. It is built from pcre-config.in, and one of the substitutions was the
2355        release number. I have updated configure.ac so that ./configure now finds
2356        the release number by grepping pcre.h.
2357    
2358    26. Added the ability to run the tests under valgrind.
2359    
2360    
2361    Version 6.4 05-Sep-05
2362    ---------------------
2363    
2364     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2365        "--" to be printed when multiple files were scanned, even when none of the
2366        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2367        consider it to be a bug, and have restored the previous behaviour.
2368    
2369     2. A couple of code tidies to get rid of compiler warnings.
2370    
2371     3. The pcretest program used to cheat by referring to symbols in the library
2372        whose names begin with _pcre_. These are internal symbols that are not
2373        really supposed to be visible externally, and in some environments it is
2374        possible to suppress them. The cheating is now confined to including
2375        certain files from the library's source, which is a bit cleaner.
2376    
2377     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2378        file's purpose clearer.
2379    
2380     5. Reorganized pcre_ucp_findchar().
2381    
2382    
2383    Version 6.3 15-Aug-05
2384    ---------------------
2385    
2386     1. The file libpcre.pc.in did not have general read permission in the tarball.
2387    
2388     2. There were some problems when building without C++ support:
2389    
2390        (a) If C++ support was not built, "make install" and "make test" still
2391            tried to test it.
2392    
2393        (b) There were problems when the value of CXX was explicitly set. Some
2394            changes have been made to try to fix these, and ...
2395    
2396        (c) --disable-cpp can now be used to explicitly disable C++ support.
2397    
2398        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2399            backslash in a target when C++ was disabled. This confuses some
2400            versions of "make", apparently. Using an intermediate variable solves
2401            this. (Same for CPP_LOBJ.)
2402    
2403     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2404        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2405        necessary on certain architectures.
2406    
2407     4. Added a setting of -export-symbols-regex to the link command to remove
2408        those symbols that are exported in the C sense, but actually are local
2409        within the library, and not documented. Their names all begin with
2410        "_pcre_". This is not a perfect job, because (a) we have to except some
2411        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2412        available (and never for static libraries). I have made a note to try to
2413        find a way round (a) in the future.
2414    
2415    
2416    Version 6.2 01-Aug-05
2417    ---------------------
2418    
2419     1. There was no test for integer overflow of quantifier values. A construction
2420        such as {1111111111111111} would give undefined results. What is worse, if
2421        a minimum quantifier for a parenthesized subpattern overflowed and became
2422        negative, the calculation of the memory size went wrong. This could have
2423        led to memory overwriting.
2424    
2425     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2426    
2427     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2428        operating environments where this matters.
2429    
2430     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2431        PCRE options from within the C++ wrapper.
2432    
2433     5. Named capturing subpatterns were not being correctly counted when a pattern
2434        was compiled. This caused two problems: (a) If there were more than 100
2435        such subpatterns, the calculation of the memory needed for the whole
2436        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2437        back references of the form \12, where the number was greater than 9, were
2438        not recognized as back references, even though there were sufficient
2439        previous subpatterns.
2440    
2441     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2442        versions of gcc, e.g. 2.95.4.
2443    
2444    
2445    Version 6.1 21-Jun-05
2446    ---------------------
2447    
2448     1. There was one reference to the variable "posix" in pcretest.c that was not
2449        surrounded by "#if !defined NOPOSIX".
2450    
2451     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2452        the cross-check on the old pcre_info() function, for the benefit of the
2453        cut-down version of PCRE that is currently imported into Exim.
2454    
2455     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2456        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2457        patterns that start (?i) but I don't think that matters. The use of (?i) is
2458        just an example; this all applies to the other options as well.
2459    
2460     4. Since libtool seems to echo the compile commands it is issuing, the output
2461        from "make" can be reduced a bit by putting "@" in front of each libtool
2462        compile command.
2463    
2464     5. Patch from the folks at Google for configure.in to be a bit more thorough
2465        in checking for a suitable C++ installation before trying to compile the
2466        C++ stuff. This should fix a reported problem when a compiler was present,
2467        but no suitable headers.
2468    
2469     6. The man pages all had just "PCRE" as their title. I have changed them to
2470        be the relevant file name. I have also arranged that these names are
2471        retained in the file doc/pcre.txt, which is a concatenation in text format
2472        of all the man pages except the little individual ones for each function.
2473    
2474     7. The NON-UNIX-USE file had not been updated for the different set of source
2475        files that come with release 6. I also added a few comments about the C++
2476        wrapper.
2477    
2478    
2479    Version 6.0 07-Jun-05
2480    ---------------------
2481    
2482     1. Some minor internal re-organization to help with my DFA experiments.
2483    
2484     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2485        didn't matter for the library itself when fully configured, but did matter
2486        when compiling without UCP support, or within Exim, where the ucp files are
2487        not imported.
2488    
2489     3. Refactoring of the library code to split up the various functions into
2490        different source modules. The addition of the new DFA matching code (see
2491        below) to a single monolithic source would have made it really too
2492        unwieldy, quite apart from causing all the code to be include in a
2493        statically linked application, when only some functions are used. This is
2494        relevant even without the DFA addition now that patterns can be compiled in
2495        one application and matched in another.
2496    
2497        The downside of splitting up is that there have to be some external
2498        functions and data tables that are used internally in different modules of
2499        the library but which are not part of the API. These have all had their
2500        names changed to start with "_pcre_" so that they are unlikely to clash
2501        with other external names.
2502    
2503     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2504        a different (DFA) algorithm. Although it is slower than the original
2505        function, it does have some advantages for certain types of matching
2506        problem.
2507    
2508     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2509        including restarting after a partial match.
2510    
2511     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2512        defined when compiling for Windows was sent to me. I have put it into the
2513        code, though I have no means of testing or verifying it.
2514    
2515     7. Added the pcre_refcount() auxiliary function.
2516    
2517     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2518        match before or at the first newline in the subject string. In pcretest,
2519        the /f option on a pattern can be used to set this.
2520    
2521     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2522        would behave wrongly. This has been present in PCRE since release 4.0.
2523    
2524    10. A number of changes to the pcregrep command:
2525    
2526        (a) Refactored how -x works; insert ^(...)$ instead of setting
2527            PCRE_ANCHORED and checking the length, in preparation for adding
2528            something similar for -w.
2529    
2530        (b) Added the -w (match as a word) option.
2531    
2532        (c) Refactored the way lines are read and buffered so as to have more
2533            than one at a time available.
2534    
2535        (d) Implemented a pcregrep test script.
2536    
2537        (e) Added the -M (multiline match) option. This allows patterns to match
2538            over several lines of the subject. The buffering ensures that at least
2539            8K, or the rest of the document (whichever is the shorter) is available
2540            for matching (and similarly the previous 8K for lookbehind assertions).
2541    
2542        (f) Changed the --help output so that it now says
2543    
2544              -w, --word-regex(p)
2545    
2546            instead of two lines, one with "regex" and the other with "regexp"
2547            because that confused at least one person since the short forms are the
2548            same. (This required a bit of code, as the output is generated
2549            automatically from a table. It wasn't just a text change.)
2550    
2551        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2552            option but starts with a hyphen. Could be a pattern or a path name
2553            starting with a hyphen, for instance.
2554    
2555        (h) "-" can be given as a file name to represent stdin.
2556    
2557        (i) When file names are being printed, "(standard input)" is used for
2558            the standard input, for compatibility with GNU grep. Previously
2559            "<stdin>" was used.
2560    
2561        (j) The option --label=xxx can be used to supply a name to be used for
2562            stdin when file names are being printed. There is no short form.
2563    
2564        (k) Re-factored the options decoding logic because we are going to add
2565            two more options that take data. Such options can now be given in four
2566            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2567    
2568        (l) Added the -A, -B, and -C options for requesting that lines of context
2569            around matches be printed.
2570    
2571        (m) Added the -L option to print the names of files that do not contain
2572            any matching lines, that is, the complement of -l.
2573    
2574        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2575            continue to scan other files.
2576    
2577        (o) The -s option was incorrectly implemented. For compatibility with other
2578            greps, it now suppresses the error message for a non-existent or non-
2579            accessible file (but not the return code). There is a new option called
2580            -q that suppresses the output of matching lines, which was what -s was
2581            previously doing.
2582    
2583        (p) Added --include and --exclude options to specify files for inclusion
2584            and exclusion when recursing.
2585    
2586    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2587        Hopefully, it now does.
2588    
2589    12. Missing cast in pcre_study().
2590    
2591    13. Added an "uninstall" target to the makefile.
2592    
2593    14. Replaced "extern" in the function prototypes in Makefile.in with
2594        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2595        world, but is set differently for Windows.
2596    
2597    15. Added a second compiling function called pcre_compile2(). The only
2598        difference is that it has an extra argument, which is a pointer to an
2599        integer error code. When there is a compile-time failure, this is set
2600        non-zero, in addition to the error test pointer being set to point to an
2601        error message. The new argument may be NULL if no error number is required
2602        (but then you may as well call pcre_compile(), which is now just a
2603        wrapper). This facility is provided because some applications need a
2604        numeric error indication, but it has also enabled me to tidy up the way
2605        compile-time errors are handled in the POSIX wrapper.
2606    
2607    16. Added VPATH=.libs to the makefile; this should help when building with one
2608        prefix path and installing with another. (Or so I'm told by someone who
2609        knows more about this stuff than I do.)
2610    
2611    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2612        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2613        match everything, including newlines. This is not POSIX-compatible, but
2614        somebody wanted the feature. From pcretest it can be activated by using
2615        both the P and the s flags.
2616    
2617    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2618    
2619    19. libpcre.pc was being incorrectly installed as executable.
2620    
2621    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2622        it now also looks for '\r' so that it will work unmodified on Windows.
2623    
2624    21. Added Google's contributed C++ wrapper to the distribution.
2625    
2626    22. Added some untidy missing memory free() calls in pcretest, to keep
2627        Electric Fence happy when testing.
2628    
2629    
2630    
2631    Version 5.0 13-Sep-04
2632    ---------------------
2633    
2634     1. Internal change: literal characters are no longer packed up into items
2635        containing multiple characters in a single byte-string. Each character
2636        is now matched using a separate opcode. However, there may be more than one
2637        byte in the character in UTF-8 mode.
2638    
2639     2. The pcre_callout_block structure has two new fields: pattern_position and
2640        next_item_length. These contain the offset in the pattern to the next match
2641        item, and its length, respectively.
2642    
2643     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2644        insertion of callouts before each pattern item. Added the /C option to
2645        pcretest to make use of this.
2646    
2647     4. On the advice of a Windows user, the lines
2648    
2649          #if defined(_WIN32) || defined(WIN32)
2650          _setmode( _fileno( stdout ), 0x8000 );
2651          #endif  /* defined(_WIN32) || defined(WIN32) */
2652    
2653        have been added to the source of pcretest. This apparently does useful
2654        magic in relation to line terminators.
2655    
2656     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2657        for the benefit of those environments where the "b" makes a difference.
2658    
2659     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2660        to know about it. I have put a hack into configure.in that adds in code
2661        to set GCC=yes if CC=icc. This seems to end up at a point in the
2662        generated configure script that is early enough to affect the setting of
2663        compiler options, which is what is needed, but I have no means of testing
2664        whether it really works. (The user who reported this had patched the
2665        generated configure script, which of course I cannot do.)
2666    
2667        LATER: After change 22 below (new libtool files), the configure script
2668        seems to know about icc (and also ecc). Therefore, I have commented out
2669        this hack in configure.in.
2670    
2671     7. Added support for pkg-config (2 patches were sent in).
2672    
2673     8. Negated POSIX character classes that used a combination of internal tables
2674        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2675        [[:^ascii]]. Typically, they would match almost any characters. The other
2676        POSIX classes were not broken in this way.
2677    
2678     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2679        to find the match, as PCRE was deluded into thinking that the match had to
2680        start at the start point or following a newline. The same bug applied to
2681        patterns with negative forward assertions or any backward assertions
2682        preceding ".*" at the start, unless the pattern required a fixed first
2683        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2684    
2685    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2686        starting at the last subject character, bytes beyond the end of the subject
2687        string were read.
2688    
2689    11. Renamed the variable "class" as "classbits" to make life easier for C++
2690        users. (Previously there was a macro definition, but it apparently wasn't
2691        enough.)
2692    
2693    12. Added the new field "tables" to the extra data so that tables can be passed
2694        in at exec time, or the internal tables can be re-selected. This allows
2695        a compiled regex to be saved and re-used at a later time by a different
2696        program that might have everything at different addresses.
2697    
2698    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2699        -R library as well as a -L library.
2700    
2701    14. The debugging options of pcretest (-d on the command line or D on a
2702        pattern) showed incorrect output for anything following an extended class
2703        that contained multibyte characters and which was followed by a quantifier.
2704    
2705    15. Added optional support for general category Unicode character properties
2706        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2707        support. It adds about 90K to the size of the library. The meanings of the
2708        inbuilt class escapes such as \d and \s have NOT been changed.
2709    
2710    16. Updated pcredemo.c to include calls to free() to release the memory for the
2711        compiled pattern.
2712    
2713    17. The generated file chartables.c was being created in the source directory
2714        instead of in the building directory. This caused the build to fail if the
2715        source directory was different from the building directory, and was
2716        read-only.
2717    
2718    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2719        file. No doubt somebody will tell me if they don't make sense... Also added
2720        Dan Mooney's comments about building on OpenVMS.
2721    
2722    19. Added support for partial matching via the PCRE_PARTIAL option for
2723        pcre_exec() and the \P data escape in pcretest.
2724    
2725    20. Extended pcretest with 3 new pattern features:
2726    
2727        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2728              write the compiled pattern to the file whose name is "rest-of-line".
2729              This is a straight binary dump of the data, with the saved pointer to
2730              the character tables forced to be NULL. The study data, if any, is
2731              written too. After writing, pcretest reads a new pattern.
2732    
2733        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2734              compiled pattern from the given file. There must not be any
2735              occurrences of "<" in the file name (pretty unlikely); if there are,
2736              pcretest will instead treat the initial "<" as a pattern delimiter.
2737              After reading in the pattern, pcretest goes on to read data lines as
2738              usual.
2739    
2740        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2741              and 16-bit fields in a compiled pattern, to simulate a pattern that
2742              was compiled on a host of opposite endianness.
2743    
2744    21. The pcre-exec() function can now cope with patterns that were compiled on
2745        hosts of opposite endianness, with this restriction:
2746    
2747          As for any compiled expression that is saved and used later, the tables
2748          pointer field cannot be preserved; the extra_data field in the arguments
2749          to pcre_exec() should be used to pass in a tables address if a value
2750          other than the default internal tables were used at compile time.
2751    
2752    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2753        now diagnosed as an error. Previously, most of the time, a negative number
2754        would have been treated as zero, but if in addition "ovector" was passed as
2755        NULL, a crash could occur.
2756    
2757    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2758        new versions from the libtool 1.5 distribution (the last one is a copy of
2759        a file called libtool.m4). This seems to have fixed the need to patch
2760        "configure" to support Darwin 1.3 (which I used to do). However, I still
2761        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2762        workstation).
2763    
2764    24. Changed the PCRE licence to be the more standard "BSD" licence.
2765    
2766    
2767    Version 4.5 01-Dec-03
2768    ---------------------
2769    
2770     1. There has been some re-arrangement of the code for the match() function so
2771        that it can be compiled in a version that does not call itself recursively.
2772        Instead, it keeps those local variables that need separate instances for
2773        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2774        needs to "recurse". Keeping track of where control must go is done by means
2775        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2776        hide most of the details from the main code, and operates only if
2777        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2778        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2779        operating.
2780    
2781        To make it easier for callers to provide specially tailored get/free
2782        functions for this usage, two new functions, pcre_stack_malloc, and
2783        pcre_stack_free, are used. They are always called in strict stacking order,
2784        and the size of block requested is always the same.
2785    
2786        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2787        PCRE has been compiled to use the stack or the heap for recursion. The
2788        -C option of pcretest uses this to show which version is compiled.
2789    
2790        A new data escape \S, is added to pcretest; it causes the amounts of store
2791        obtained and freed by both kinds of malloc/free at match time to be added
2792        to the output.
2793    
2794     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2795        what's available on my current Linux desktop machine.
2796    
2797     3. When matching a UTF-8 string, the test for a valid string at the start has
2798        been extended. If start_offset is not zero, PCRE now checks that it points
2799        to a byte that is the start of a UTF-8 character. If not, it returns
2800        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2801        this is necessary because there may be backward assertions in the pattern.
2802        When matching the same subject several times, it may save resources to use
2803        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2804    
2805     4. The code for checking the validity of UTF-8 strings has been tightened so
2806        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2807        containing "overlong sequences".
2808    
2809     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2810        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2811        should have been "&", but it just so happened that all the cases this let
2812        through by mistake were picked up later in the function.
2813    
2814     6. I had used a variable called "isblank" - this is a C99 function, causing
2815        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2816    
2817     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2818        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2819        so the version is shown; (c) stop "make test" echoing "./RunTest".
2820    
2821     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2822    
2823     9. The prototype for memmove() for systems that don't have it was using
2824        size_t, but the inclusion of the header that defines size_t was later. I've
2825        moved the #includes for the C headers earlier to avoid this.
2826    
2827    10. Added some adjustments to the code to make it easier to compiler on certain
2828        special systems:
2829    
2830          (a) Some "const" qualifiers were missing.
2831          (b) Added the macro EXPORT before all exported functions; by default this
2832              is defined to be empty.
2833          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2834              that it reads its output file name as an argument instead of writing
2835              to the standard output and assuming this can be redirected.
2836    
2837    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2838        class containing characters with values greater than 255, PCRE compilation
2839        went into a loop.
2840    
2841    12. A recursive reference to a subpattern that was within another subpattern
2842        that had a minimum quantifier of zero caused PCRE to crash. For example,
2843        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2844        recursion. If the recursively-called subpattern itself had a zero repeat,
2845        that was OK.
2846    
2847    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2848        buffer into which it was copied (for escape processing) was still set at
2849        1024, so long lines caused crashes.
2850    
2851    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2852        "internal error: code overflow...". This applied to any character class
2853        that was followed by a possessive quantifier.
2854    
2855    15. Modified the Makefile to add libpcre.la as a prerequisite for
2856        libpcreposix.la because I was told this is needed for a parallel build to
2857        work.
2858    
2859    16. If a pattern that contained .* following optional items at the start was
2860        studied, the wrong optimizing data was generated, leading to matching
2861        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2862        matching string must start with a or b or c. The correct conclusion for
2863        this pattern is that a match can start with any character.
2864    
2865    
2866    Version 4.4 13-Aug-03
2867    ---------------------
2868    
2869     1. In UTF-8 mode, a character class containing characters with values between
2870        127 and 255 was not handled correctly if the compiled pattern was studied.
2871        In fixing this, I have also improved the studying algorithm for such
2872        classes (slightly).
2873    
2874     2. Three internal functions had redundant arguments passed to them. Removal
2875        might give a very teeny performance improvement.
2876    
2877     3. Documentation bug: the value of the capture_top field in a callout is *one
2878        more than* the number of the hightest numbered captured substring.
2879    
2880     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2881        in incorrectly linking with a previously installed version. They now link
2882        explicitly with libpcre.la.
2883    
2884     5. configure.in no longer needs to recognize Cygwin specially.
2885    
2886     6. A problem in pcre.in for Windows platforms is fixed.
2887    
2888     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2889        pcretest, it used to include the size of the study block as part of its
2890        output. Unfortunately, the structure contains a field that has a different
2891        size on different hardware architectures. This meant that the tests that
2892        showed this size failed. As the block is currently always of a fixed size,
2893        this information isn't actually particularly useful in pcretest output, so
2894        I have just removed it.
2895    
2896     8. Three pre-processor statements accidentally did not start in column 1.
2897        Sadly, there are *still* compilers around that complain, even though
2898        standard C has not required this for well over a decade. Sigh.
2899    
2900     9. In pcretest, the code for checking callouts passed small integers in the
2901        callout_data field, which is a void * field. However, some picky compilers
2902        complained about the casts involved for this on 64-bit systems. Now
2903        pcretest passes the address of the small integer instead, which should get
2904        rid of the warnings.
2905    
2906    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2907        both compile and run time, and gives an error if an invalid UTF-8 sequence
2908        is found. There is a option for disabling this check in cases where the
2909        string is known to be correct and/or the maximum performance is wanted.
2910    
2911    11. In response to a bug report, I changed one line in Makefile.in from
2912    
2913            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2914        to
2915            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2916    
2917        to look similar to other lines, but I have no way of telling whether this
2918        is the right thing to do, as I do not use Windows. No doubt I'll get told
2919        if it's wrong...
2920    
2921    
2922    Version 4.3 21-May-03
2923    ---------------------
2924    
2925    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2926       Makefile.
2927    
2928    2. Some refactoring to improve the quality of the code:
2929    
2930       (i)   The utf8_table... variables are now declared "const".
2931    
2932       (ii)  The code for \cx, which used the "case flipping" table to upper case
2933             lower case letters, now just substracts 32. This is ASCII-specific,
2934             but the whole concept of \cx is ASCII-specific, so it seems
2935             reasonable.
2936    
2937       (iii) PCRE was using its character types table to recognize decimal and
2938             hexadecimal digits in the pattern. This is silly, because it handles
2939             only 0-9, a-f, and A-F, but the character types table is locale-
2940             specific, which means strange things might happen. A private
2941             table is now used for this - though it costs 256 bytes, a table is
2942             much faster than multiple explicit tests. Of course, the standard
2943             character types table is still used for matching digits in subject
2944             strings against \d.
2945    
2946       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2947             ending in _t are). So I've renamed it as ESC_tee.
2948    
2949    3. The first argument for regexec() in the POSIX wrapper should have been
2950       defined as "const".
2951    
2952    4. Changed pcretest to use malloc() for its buffers so that they can be
2953       Electric Fenced for debugging.
2954    
2955    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2956       to read one or more bytes before the start of the subject string. Often this
2957       had no effect on PCRE's behaviour, but in some circumstances it could
2958       provoke a segmentation fault.
2959    
2960    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2961       to try to read one or more bytes before the start of the subject string.
2962    
2963    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2964       UTF-8 support could misbehave in various ways if the subject string
2965       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2966       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2967       back over UTF-8 characters.)
2968    
2969    
2970    Version 4.2 14-Apr-03
2971    ---------------------
2972    
2973    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2974    
2975    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2976         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2977         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2978         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2979         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2980           and BUILD_EXEEXT
2981         Note: automatic setting of the BUILD variables is not yet working
2982         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2983           compile-time but not at link-time
2984         [LINK]: use for linking executables only
2985         make different versions for Windows and non-Windows
2986         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2987           libraries
2988         [LINK_FOR_BUILD]: new variable
2989         [OBJEXT]: use throughout
2990         [EXEEXT]: use throughout
2991         <winshared>: new target
2992         <wininstall>: new target
2993         <dftables.o>: use native compiler
2994         <dftables>: use native linker
2995         <install>: handle Windows platform correctly
2996         <clean>: ditto
2997         <check>: ditto
2998         copy DLL to top builddir before testing
2999    
3000       As part of these changes, -no-undefined was removed again. This was reported
3001       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3002       in any case.
3003    
3004    3. Some tidies to get rid of compiler warnings:
3005    
3006       . In the match_data structure, match_limit was an unsigned long int, whereas
3007         match_call_count was an int. I've made them both unsigned long ints.
3008    
3009       . In pcretest the fact that a const uschar * doesn't automatically cast to
3010         a void * provoked a warning.
3011    
3012       . Turning on some more compiler warnings threw up some "shadow" variables
3013         and a few more missing casts.
3014    
3015    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3016       option, a class that contained a single character with a value between 128
3017       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3018    
3019    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3020       option, a class that contained several characters, but with at least one
3021       whose value was between 128 and 255 caused PCRE to crash.
3022    
3023    
3024    Version 4.1 12-Mar-03
3025    ---------------------
3026    
3027    1. Compiling with gcc -pedantic found a couple of places where casts were
3028    needed, and a string in dftables.c that was longer than standard compilers are
3029    required to support.
3030    
3031    2. Compiling with Sun's compiler found a few more places where the code could
3032    be tidied up in order to avoid warnings.
3033    
3034    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3035    first of these names is deprecated in the latest Autoconf in favour of the name
3036    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3037    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3038    analogy I have changed it to CFLAGS_FOR_BUILD.
3039    
3040    4. Added -no-undefined to the linking command in the Makefile, because this is
3041    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3042    linking step for the pcreposix library.
3043    
3044    5. PCRE was failing to diagnose the case of two named groups with the same
3045    name.
3046    
3047    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3048    literal character that is needed in the subject for a match, and scans along to
3049    ensure that it is present before embarking on the full matching process. This
3050    saves time in cases of nested unlimited repeats that are never going to match.
3051    Problem: the scan can take a lot of time if the subject is very long (e.g.
3052    megabytes), thus penalizing straightforward matches. It is now done only if the
3053    amount of subject to be scanned is less than 1000 bytes.
3054    
3055    7. A lesser problem with the same optimization is that it was recording the
3056    first character of an anchored pattern as "needed", thus provoking a search
3057    right along the subject, even when the first match of the pattern was going to
3058    fail. The "needed" character is now not set for anchored patterns, unless it
3059    follows something in the pattern that is of non-fixed length. Thus, it still
3060    fulfils its original purpose of finding quick non-matches in cases of nested
3061    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3062    
3063    
3064    Version 4.0 17-Feb-03
3065    ---------------------
3066    
3067    1. If a comment in an extended regex that started immediately after a meta-item
3068    extended to the end of string, PCRE compiled incorrect data. This could lead to
3069    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3070    
3071    2. Moved to autoconf 2.53 and libtool 1.4.2.
3072    
3073    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3074    the special perltest8 script is no longer needed - all the tests can be run
3075    from a single perltest script.
3076    
3077    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3078    by \s. It has now been removed in PCRE. This means it isn't recognized as
3079    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3080    class [:space:] *does* include VT, thereby creating a mess.
3081    
3082    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3083    space and tab.
3084    
3085    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3086    its new features into the main test script, reducing the number of scripts.
3087    
3088    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3089    were backward compatible, and made the (?i) apply to the whole pattern, as if
3090    /i were given. Now it behaves more logically, and applies the option setting
3091    only to what follows. PCRE has been changed to follow suit. However, if it
3092    finds options settings right at the start of the pattern, it extracts them into
3093    the global options, as before. Thus, they show up in the info data.
3094    
3095    8. Added support for the \Q...\E escape sequence. Characters in between are
3096    treated as literals. This is slightly different from Perl in that $ and @ are
3097    also handled as literals inside the quotes. In Perl, they will cause variable
3098    interpolation. Note the following examples:
3099    
3100        Pattern            PCRE matches      Perl matches
3101    
3102        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3103        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3104        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3105    
3106    For compatibility with Perl, \Q...\E sequences are recognized inside character
3107    classes as well as outside them.
3108    
3109    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3110    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3111    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3112    signed/unsigned warnings.
3113    
3114    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3115    option for pcretest, so I've replaced it by a simple function that does just
3116    that job.
3117    
3118    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3119    "pcregrep -".
3120    
3121    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3122    Java package. This provides some syntactic sugar for simple cases of what my
3123    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3124    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3125    item, you can use this simplified notation. Note that only makes sense with
3126    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3127    greediness, whatever the setting of the PCRE_UNGREEDY option.
3128    
3129    13. A change of greediness default within a pattern was not taking effect at
3130    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3131    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3132    was abstracted outside.
3133    
3134    14. PCRE now supports the \G assertion. It is true when the current matching
3135    position is at the start point of the match. This differs from \A when the
3136    starting offset is non-zero. Used with the /g option of pcretest (or similar
3137    code), it works in the same way as it does for Perl's /g option. If all
3138    alternatives of a regex begin with \G, the expression is anchored to the start
3139    match position, and the "anchored" flag is set in the compiled expression.
3140    
3141    15. Some bugs concerning the handling of certain option changes within patterns
3142    have been fixed. These applied to options other than (?ims). For example,
3143    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3144    been the other way round. Some of this was related to change 7 above.
3145    
3146    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3147    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3148    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3149    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3150    
3151    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3152    mode. Unlike ".", it always matches newline, whatever the setting of
3153    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3154    assertions. Perl allows it, but it doesn't (in general) work because it can't
3155    calculate the length of the lookbehind. At least, that's the case for Perl
3156    5.8.0 - I've been told they are going to document that it doesn't work in
3157    future.
3158    
3159    18. Added an error diagnosis for escapes that PCRE does not support: these are
3160    \L, \l, \N, \P, \p, \U, \u, and \X.
3161    
3162    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3163    reading past the end of the pattern in cases such as /[abcd/.
3164    
3165    20. PCRE was getting more memory than necessary for patterns with classes that
3166    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3167    
3168    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3169    compiling PCRE for use with Virtual Pascal.
3170    
3171    22. Small fix to the Makefile to make it work properly if the build is done
3172    outside the source tree.
3173    
3174    23. Added a new extension: a condition to go with recursion. If a conditional
3175    subpattern starts with (?(R) the "true" branch is used if recursion has
3176    happened, whereas the "false" branch is used only at the top level.
3177    
3178    24. When there was a very long string of literal characters (over 255 bytes
3179    without UTF support, over 250 bytes with UTF support), the computation of how
3180    much memory was required could be incorrect, leading to segfaults or other
3181    strange effects.
3182    
3183    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3184    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3185    there was a subsequent back reference to those brackets. This meant that, for
3186    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3187    possible to check for precisely this case. All we can do is abandon the
3188    optimization if .* occurs inside capturing brackets when there are any back
3189    references whatsoever. (See below for a better fix that came later.)
3190    
3191    26. The handling of the optimization for finding the first character of a
3192    non-anchored pattern, and for finding a character that is required later in the
3193    match were failing in some cases. This didn't break the matching; it just
3194    failed to optimize when it could. The way this is done has been re-implemented.
3195    
3196    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3197    
3198    28. Added a new feature that provides some of the functionality that Perl
3199    provides with (?{...}). The facility is termed a "callout". The way it is done
3200    in PCRE is for the caller to provide an optional function, by setting
3201    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3202    global variable. By default it is unset, which disables all calling out. To get
3203    the function called, the regex must include (?C) at appropriate points. This
3204    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3205    This provides a means of identifying different callout points. When PCRE
3206    reaches such a point in the regex, if pcre_callout has been set, the external
3207    function is called. It is provided with data in a structure called
3208    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3209    matching continues; if it returns a non-zero value, the match at the current
3210    point fails. However, backtracking will occur if possible. [This was changed
3211    later and other features added - see item 49 below.]
3212    
3213    29. pcretest is upgraded to test the callout functionality. It provides a
3214    callout function that displays information. By default, it shows the start of
3215    the match and the current position in the text. There are some new data escapes
3216    to vary what happens:
3217    
3218        \C+         in addition, show current contents of captured substrings
3219        \C-         do not supply a callout function
3220        \C!n        return 1 when callout number n is reached
3221        \C!n!m      return 1 when callout number n is reached for the mth time
3222    
3223    30. If pcregrep was called with the -l option and just a single file name, it
3224    output "<stdin>" if a match was found, instead of the file name.
3225    
3226    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3227    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3228    pcre_exec(). This saves a malloc/free per call. The default value of
3229    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3230    when configuring.
3231    
3232    32. The default maximum size of a compiled pattern is 64K. There have been a
3233    few cases of people hitting this limit. The code now uses macros to handle the
3234    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3235    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3236    configuring. Tests 2 and 5 work only with 2-byte links because they output
3237    debugging information about compiled patterns.
3238    
3239    33. Internal code re-arrangements:
3240    
3241    (a) Moved the debugging function for printing out a compiled regex into
3242        its own source file (printint.c) and used #include to pull it into
3243        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3244        separate copies.
3245    
3246    (b) Defined the list of op-code names for debugging as a macro in
3247        internal.h so that it is next to the definition of the opcodes.
3248    
3249    (c) Defined a table of op-code lengths for simpler skipping along compiled
3250        code. This is again a macro in internal.h so that it is next to the
3251        definition of the opcodes.
3252    
3253    34. Added support for recursive calls to individual subpatterns, along the
3254    lines of Robin Houston's patch (but implemented somewhat differently).
3255    
3256    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3257    allow it to read and process whole directories in Win32. This code was
3258    contributed by Lionel Fourquaux; it has not been tested by me.
3259    
3260    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3261    used to name a group. Names consist of alphanumerics and underscores, and must
3262    be unique. Back references use the syntax (?P=name) and recursive calls use
3263    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3264    numbers. The function pcre_fullinfo() can be used after compilation to extract
3265    a name/number map. There are three relevant calls:
3266    
3267      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3268      PCRE_INFO_NAMECOUNT            yields the number of entries
3269      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3270    
3271    The map is a vector of fixed-size entries. The size of each entry depends on
3272    the length of the longest name used. The first two bytes of each entry are the
3273    group number, most significant byte first. There follows the corresponding
3274    name, zero terminated. The names are in alphabetical order.
3275    
3276    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3277    case instead of 255. Making it the same both with and without UTF-8 support
3278    means that the same test output works with both.
3279    
3280    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3281    calling malloc() with a zero argument.
3282    
3283    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3284    optimization. I've improved things by keeping a bitmap of backreferences with
3285    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3286    fact referenced, the optimization can be applied. It is unlikely that a
3287    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3288    the match to follow \n) will appear inside brackets with a number greater than
3289    31, but if it does, any back reference > 31 suppresses the optimization.
3290    
3291    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3292    of disabling numbered capturing parentheses. Any opening parenthesis that is
3293    not followed by ? behaves as if it were followed by ?: but named parentheses
3294    can still be used for capturing (and they will acquire numbers in the usual
3295    way).
3296    
3297    41. Redesigned the return codes from the match() function into yes/no/error so
3298    that errors can be passed back from deep inside the nested calls. A malloc
3299    failure while inside a recursive subpattern call now causes the
3300    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3301    
3302    42. It is now possible to set a limit on the number of times the match()
3303    function is called in a call to pcre_exec(). This facility makes it possible to
3304    limit the amount of recursion and backtracking, though not in a directly
3305    obvious way, because the match() function is used in a number of different
3306    circumstances. The count starts from zero for each position in the subject
3307    string (for non-anchored patterns). The default limit is, for compatibility, a
3308    large number, namely 10 000 000. You can change this in two ways:
3309    
3310    (a) When configuring PCRE before making, you can use --with-match-limit=n
3311        to set a default value for the compiled library.
3312    
3313    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3314        a different value is set. See 45 below.
3315    
3316    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3317    
3318    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3319    of things that can be changed at compile time. The first argument specifies
3320    what is wanted and the second points to where the information is to be placed.
3321    The current list of available information is:
3322    
3323      PCRE_CONFIG_UTF8
3324    
3325    The output is an integer that is set to one if UTF-8 support is available;
3326    otherwise it is set to zero.
3327    
3328      PCRE_CONFIG_NEWLINE
3329    
3330    The output is an integer that it set to the value of the code that is used for
3331    newline. It is either LF (10) or CR (13).
3332    
3333      PCRE_CONFIG_LINK_SIZE
3334    
3335    The output is an integer that contains the number of bytes used for internal
3336    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3337    
3338      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3339    
3340    The output is an integer that contains the threshold above which the POSIX
3341    interface uses malloc() for output vectors. See item 31 above.
3342    
3343      PCRE_CONFIG_MATCH_LIMIT
3344    
3345    The output is an unsigned integer that contains the default limit of the number
3346    of match() calls in a pcre_exec() execution. See 42 above.
3347    
3348    44. pcretest has been upgraded by the addition of the -C option. This causes it
3349    to extract all the available output from the new pcre_config() function, and to
3350    output it. The program then exits immediately.
3351    
3352    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3353    order to support additional features. One way would have been to define
3354    pcre_exec2() (for example) with extra arguments, but this would not have been
3355    extensible, and would also have required all calls to the original function to
3356    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3357    is used for passing in "extra" data from pcre_study().
3358    
3359    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3360    contains the following fields:
3361    
3362      flags         a bitmap indicating which of the following fields are set
3363      study_data    opaque data from pcre_study()
3364      match_limit   a way of specifying a limit on match() calls for a specific
3365                      call to pcre_exec()
3366      callout_data  data for callouts (see 49 below)
3367    
3368    The flag bits are also defined in pcre.h, and are
3369    
3370      PCRE_EXTRA_STUDY_DATA
3371      PCRE_EXTRA_MATCH_LIMIT
3372      PCRE_EXTRA_CALLOUT_DATA
3373    
3374    The pcre_study() function now returns one of these new pcre_extra blocks, with
3375    the actual study data pointed to by the study_data field, and the
3376    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3377    before. That is, this change is entirely upwards-compatible and requires no
3378    change to existing code.
3379    
3380    If you want to pass in additional data to pcre_exec(), you can either place it
3381    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3382    block.
3383    
3384    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3385    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3386    times with different match limits, until it finds the minimum value needed for
3387    pcre_exec() to complete. The value is then output. This can be instructive; for
3388    most simple matches the number is quite small, but for pathological cases it
3389    gets very large very quickly.
3390    
3391    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3392    returns the size of the data block pointed to by the study_data field in a
3393    pcre_extra block, that is, the value that was passed as the argument to
3394    pcre_malloc() when PCRE was getting memory in which to place the information
3395    created by pcre_study(). The fourth argument should point to a size_t variable.
3396    pcretest has been extended so that this information is shown after a successful
3397    pcre_study() call when information about the compiled regex is being displayed.
3398    
3399    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3400    because what follows is always an absolute path. (Later: it turns out that this
3401    is more than cosmetic for MinGW, because it doesn't like empty path
3402    components.)
3403    
3404    49. Some changes have been made to the callout feature (see 28 above):
3405    
3406    (i)  A callout function now has three choices for what it returns:
3407    
3408           0  =>  success, carry on matching
3409         > 0  =>  failure at this point, but backtrack if possible
3410         < 0  =>  serious error, return this value from pcre_exec()
3411    
3412         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3413         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3414         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3415         use by callout functions. It will never be used by PCRE itself.
3416    
3417    (ii) The pcre_extra structure (see 45 above) has a void * field called
3418         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3419         pcre_callout_block structure has a field of the same name. The contents of
3420         the field passed in the pcre_extra structure are passed to the callout
3421         function in the corresponding field in the callout block. This makes it
3422         easier to use the same callout-containing regex from multiple threads. For
3423         testing, the pcretest program has a new data escape
3424    
3425           \C*n        pass the number n (may be negative) as callout_data
3426    
3427         If the callout function in pcretest receives a non-zero value as
3428         callout_data, it returns that value.
3429    
3430    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3431    there were some redundant $(CFLAGS) in commands that are now specified as
3432    $(LINK), which already includes $(CFLAGS).
3433    
3434    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3435    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3436    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3437    one-byte characters throughout. Note that case-insensitive matching applies
3438    only to characters whose values are less than 256. PCRE doesn't support the
3439    notion of cases for higher-valued characters.
3440    
3441    (i)   A character class whose characters are all within 0-255 is handled as
3442          a bit map, and the map is inverted for negative classes. Previously, a
3443          character > 255 always failed to match such a class; however it should
3444          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3445    
3446    (ii)  A negated character class with a single character < 255 is coded as
3447          "not this character" (OP_NOT). This wasn't working properly when the test
3448          character was multibyte, either singly or repeated.
3449    
3450    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3451          mode, for example: \x{100}{2,3}.
3452    
3453    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3454          singly or repeated) now correctly test multibyte characters. However,
3455          PCRE doesn't recognize any characters with values greater than 255 as
3456          digits, spaces, or word characters. Such characters always match \D, \S,
3457          and \W, and never match \d, \s, or \w.
3458    
3459    (v)   Classes may now contain characters and character ranges with values
3460          greater than 255. For example: [ab\x{100}-\x{400}].
3461    
3462    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3463          PCRE in UTF-8 mode.
3464    
3465    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3466    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3467    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3468    value.)
3469    
3470    53. The single man page has become too large. I have therefore split it up into
3471    a number of separate man pages. These also give rise to individual HTML pages;
3472    these are now put in a separate directory, and there is an index.html page that
3473    lists them all. Some hyperlinking between the pages has been installed.
3474    
3475    54. Added convenience functions for handling named capturing parentheses.
3476    
3477    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3478    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3479    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3480    are faulted.
3481    
3482    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3483    calling configure. These values are used when compiling the dftables.c program
3484    which is run to generate the source of the default character tables. They
3485    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3486    you will need to set these values.
3487    
3488    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3489    
3490    
3491    Version 3.9 02-Jan-02
3492    ---------------------
3493    
3494    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3495    
3496    2. If --disable-static was given, the building process failed when trying to
3497    build pcretest and pcregrep. (For some reason it was using libtool to compile
3498    them, which is not right, as they aren't part of the library.)
3499    
3500    
3501    Version 3.8 18-Dec-01
3502    ---------------------
3503    
3504    1. The experimental UTF-8 code was completely screwed up. It was packing the
3505    bytes in the wrong order. How dumb can you get?
3506    
3507    
3508    Version 3.7 29-Oct-01
3509    ---------------------
3510    
3511    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3512    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3513    this didn't happen under Solaris 8, where I normally test things.
3514    
3515    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3516    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3517    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3518    this fix an hour or so after the initial 3.7 release.)
3519    
3520    
3521    Version 3.6 23-Oct-01
3522    ---------------------
3523    
3524    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3525    offsets passed as NULL with zero offset count.
3526    
3527    2. The config.guess and config.sub files had not been updated when I moved to
3528    the latest autoconf.
3529    
3530    
3531    Version 3.5 15-Aug-01
3532    ---------------------
3533    
3534    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3535    had been forgotten.
3536    
3537    2. By using declared but undefined structures, we can avoid using "void"
3538    definitions in pcre.h while keeping the internal definitions of the structures
3539    private.
3540    
3541    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3542    user point of view, this means that both static and shared libraries are built
3543    by default, but this can be individually controlled. More of the work of
3544    handling this static/shared cases is now inside libtool instead of PCRE's make
3545    file.
3546    
3547    4. The pcretest utility is now installed along with pcregrep because it is
3548    useful for users (to test regexs) and by doing this, it automatically gets
3549    relinked by libtool. The documentation has been turned into a man page, so
3550    there are now .1, .txt, and .html versions in /doc.
3551    
3552    5. Upgrades to pcregrep:
3553       (i)   Added long-form option names like gnu grep.
3554       (ii)  Added --help to list all options with an explanatory phrase.
3555       (iii) Added -r, --recursive to recurse into sub-directories.
3556       (iv)  Added -f, --file to read patterns from a file.
3557    
3558    6. pcre_exec() was referring to its "code" argument before testing that
3559    argument for NULL (and giving an error if it was NULL).
3560    
3561    7. Upgraded Makefile.in to allow for compiling in a different directory from
3562    the source directory.
3563    
3564    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3565    options bits, the pointer it was passed was to an int instead of to an unsigned
3566    long int. This mattered only on 64-bit systems.
3567    
3568    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3569    generated) instead of pcre.in, which it its source. Also made the same change
3570    in several of the .c files.
3571    
3572    10. A new release of gcc defines printf() as a macro, which broke pcretest
3573    because it had an ifdef in the middle of a string argument for printf(). Fixed
3574    by using separate calls to printf().
3575    
3576    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3577    script, to force use of CR or LF instead of \n in the source. On non-Unix
3578    systems, the value can be set in config.h.
3579    
3580    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3581    absolute limit. Changed the text of the error message to make this clear, and
3582    likewise updated the man page.
3583    
3584    13. The limit of 99 on the number of capturing subpatterns has been removed.
3585    The new limit is 65535, which I hope will not be a "real" limit.
3586    
3587    
3588    Version 3.4 22-Aug-00
3589    ---------------------
3590    
3591    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3592    
3593    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3594    
3595    
3596    Version 3.3 01-Aug-00
3597    ---------------------
3598    
3599    1. If an octal character was given, but the value was greater than \377, it
3600    was not getting masked to the least significant bits, as documented. This could
3601    lead to crashes in some systems.
3602    
3603    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3604    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3605    
3606    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3607    These just pass their arguments on to (pcre_free)(), but they are provided
3608    because some uses of PCRE bind it to non-C systems that can call its functions,
3609    but cannot call free() or pcre_free() directly.
3610    
3611    4. Add "make test" as a synonym for "make check". Corrected some comments in
3612    the Makefile.
3613    
3614    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3615    Makefile.
3616    
3617    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3618    command called pgrep for grepping around the active processes.
3619    
3620    7. Added the beginnings of support for UTF-8 character strings.
3621    
3622    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3623    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3624    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3625    out for the ar command.)
3626    
3627    
3628    Version 3.2 12-May-00
3629    ---------------------
3630    
3631    This is purely a bug fixing release.
3632    
3633    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3634    of ZA. This was just one example of several cases that could provoke this bug,
3635    which was introduced by change 9 of version 2.00. The code for breaking
3636    infinite loops after an iteration that matches an empty string was't working
3637    correctly.
3638    
3639    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3640    when matched against abbab (for example). After matching an empty string, it
3641    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3642    caused it to match further down the string than it should.
3643    
3644    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3645    was there because it doesn't seem to be needed, and it causes trouble on some
3646    systems, as it is not a Standard C header. It has been removed.
3647    
3648    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3649    were reported on the Macintosh. The changes were from
3650    
3651      while ((c = *(++ptr)) != 0 && c != '\n');
3652    to
3653      while ((c = *(++ptr)) != 0 && c != '\n') ;
3654    
3655    Totally extraordinary, but if that's what it takes...
3656    
3657    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3658    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3659    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3660    assumes the way PCRE uses memmove() (always moving upwards).
3661    
3662    6. PCRE is being used in one environment where strchr() is not available. There
3663    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3664    faster code anyway.
3665    
3666    
3667    Version 3.1 09-Feb-00
3668    ---------------------
3669    
3670    The only change in this release is the fixing of some bugs in Makefile.in for
3671    the "install" target:
3672    
3673    (1) It was failing to install pcreposix.h.
3674    
3675    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3676    
3677    
3678    Version 3.0 01-Feb-00
3679    ---------------------
3680    
3681    1. Add support for the /+ modifier to perltest (to output $` like it does in
3682    pcretest).
3683    
3684    2. Add support for the /g modifier to perltest.
3685    
3686    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3687    matches null strings.
3688    
3689    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3690    pattern. Perl treats empty patterns specially - it reuses the most recent
3691    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3692    effect.
3693    
3694    5. The POSIX interface was broken in that it was just handing over the POSIX
3695    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3696    required a bigger vector, with some working space on the end. This means that
3697    the POSIX wrapper now has to get and free some memory, and copy the results.
3698    
3699    6. Added some simple autoconf support, placing the test data and the
3700    documentation in separate directories, re-organizing some of the
3701    information files, and making it build pcre-config (a GNU standard). Also added
3702    libtool support for building PCRE as a shared library, which is now the
3703    default.
3704    
3705    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3706    09 are not valid octal constants. Single digits will be used for minor values
3707    less than 10.
3708    
3709    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3710    existing programs that set these in the POSIX interface can use PCRE without
3711    modification.
3712    
3713    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3714    return all that pcre_info() returns, plus additional data. The pcre_info()
3715    function is retained for compatibility, but is considered to be obsolete.
3716    
3717    10. Added experimental recursion feature (?R) to handle one common case that
3718    Perl 5.6 will be able to do with (?p{...}).
3719    
3720    11. Added support for POSIX character classes like [:alpha:], which Perl is
3721    adopting.
3722    
3723    
3724    Version 2.08 31-Aug-99
3725    ----------------------
3726    
3727    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3728    trying to match at the startoffset position, but instead was moving forward to
3729    the next newline as if a previous match had failed.
3730    
3731    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3732    and could get into a loop if a null string was matched other than at the start
3733    of the subject.
3734    
3735    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3736    be distinguished at compile time, and for completeness also added PCRE_DATE.
3737    
3738    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3739    in GnuWin32 environments.
3740    
3741    
3742    Version 2.07 29-Jul-99
3743    ----------------------
3744    
3745    1. The documentation is now supplied in plain text form and HTML as well as in
3746    the form of man page sources.
3747    
3748    2. C++ compilers don't like assigning (void *) values to other pointer types.
3749    In particular this affects malloc(). Although there is no problem in Standard
3750    C, I've put in casts to keep C++ compilers happy.
3751    
3752    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3753    should be (const char *).
3754    
3755    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3756    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3757    However, I haven't made this a standard facility. The documentation doesn't
3758    mention it, and the Makefile doesn't support it.
3759    
3760    5. The Makefile now contains an "install" target, with editable destinations at
3761    the top of the file. The pcretest program is not installed.
3762    
3763    6. pgrep -V now gives the PCRE version number and date.
3764    
3765    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3766    causing the entire string to be ignored, instead of just the last character.
3767    
3768    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3769    non-matching string, it can take a very, very long time, even for strings of
3770    quite modest length, because of the nested recursion. PCRE now does better in
3771    some of these cases. It does this by remembering the last required literal
3772    character in the pattern, and pre-searching the subject to ensure it is present
3773    before running the real match. In other words, it applies a heuristic to detect
3774    some types of certain failure quickly, and in the above example, if presented
3775    with a string that has no trailing " it gives "no match" very quickly.
3776    
3777    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3778    other alternatives are tried instead.
3779    
3780    
3781    Version 2.06 09-Jun-99
3782    ----------------------
3783    
3784    1. Change pcretest's output for amount of store used to show just the code
3785    space, because the remainder (the data block) varies in size between 32-bit and
3786    64-bit systems.
3787    
3788    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3789    start matching at. This allows lookbehinds to work when searching for multiple
3790    occurrences in a string.
3791    
3792    3. Added additional options to pcretest for testing multiple occurrences:
3793    
3794       /+   outputs the rest of the string that follows a match
3795       /g   loops for multiple occurrences, using the new startoffset argument
3796       /G   loops for multiple occurrences by passing an incremented pointer
3797    
3798    4. PCRE wasn't doing the "first character" optimization for patterns starting
3799    with \b or \B, though it was doing it for other lookbehind assertions. That is,
3800    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3801    the letter 'x'. On long subject strings, this gives a significant speed-up.
3802    
3803    
3804    Version 2.05 21-Apr-99
3805    ----------------------
3806    
3807    1. Changed the type of magic_number from int to long int so that it works
3808    properly on 16-bit systems.
3809    
3810    2. Fixed a bug which caused patterns starting with .* not to work correctly
3811    when the subject string contained newline characters. PCRE was assuming
3812    anchoring for such patterns in all cases, which is not correct because .* will
3813    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3814    DOTALL is set at top level; otherwise it knows that patterns starting with .*
3815    must be retried after every newline in the subject.
3816    
3817    
3818    Version 2.04 18-Feb-99
3819    ----------------------
3820    
3821    1. For parenthesized subpatterns with repeats whose minimum was zero, the
3822    computation of the store needed to hold the pattern was incorrect (too large).
3823    If such patterns were nested a few deep, this could multiply and become a real
3824    problem.
3825    
3826    2. Added /M option to pcretest to show the memory requirement of a specific
3827    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3828    
3829    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3830    compiled in such a way that the backtracking after subsequent failure was
3831