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

Diff of /code/trunk/ChangeLog

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

revision 15 by nigel, Sat Feb 24 21:38:25 2007 UTC revision 619 by ph10, Sun Jul 17 13:23:14 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.13 30-Apr-2011
5    ------------------------
6    
7    1.  The Unicode data tables have been updated to Unicode 6.0.0.
8    
9    2.  Two minor typos in pcre_internal.h have been fixed.
10    
11    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
12        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
13        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
14    
15    4.  There were a number of related bugs in the code for matching backrefences
16        caselessly in UTF-8 mode when codes for the characters concerned were
17        different numbers of bytes. For example, U+023A and U+2C65 are an upper
18        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
19        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
20        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
21        2-byte code at the end of the subject (it thought there wasn't enough data
22        left).
23    
24    5.  Comprehensive information about what went wrong is now returned by
25        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
26        as the output vector has at least 2 elements. The offset of the start of
27        the failing character and a reason code are placed in the vector.
28    
29    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
30        now returned is for the first byte of the failing character, instead of the
31        last byte inspected. This is an incompatible change, but I hope it is small
32        enough not to be a problem. It makes the returned offset consistent with
33        pcre_exec() and pcre_dfa_exec().
34    
35    7.  pcretest now gives a text phrase as well as the error number when
36        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
37        failure, the offset and reason code are output.
38    
39    8.  When \R was used with a maximizing quantifier it failed to skip backwards
40        over a \r\n pair if the subsequent match failed. Instead, it just skipped
41        back over a single character (\n). This seems wrong (because it treated the
42        two characters as a single entity when going forwards), conflicts with the
43        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
44        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
45        has been changed.
46    
47    9.  Some internal refactoring has changed the processing so that the handling
48        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
49        time (the PCRE_DOTALL option was changed this way some time ago: version
50        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
51        which was always a bit of a fudge. It also means that there is one less
52        argument for the match() function, which reduces its stack requirements
53        slightly. This change also fixes an incompatibility with Perl: the pattern
54        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
55    
56    10. More internal refactoring has drastically reduced the number of recursive
57        calls to match() for possessively repeated groups such as (abc)++ when
58        using pcre_exec().
59    
60    11. While implementing 10, a number of bugs in the handling of groups were
61        discovered and fixed:
62    
63        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
64        (a|)*(?1) gave a compile-time internal error.
65        ((a|)+)+  did not notice that the outer group could match an empty string.
66        (^a|^)+   was not marked as anchored.
67        (.*a|.*)+ was not marked as matching at start or after a newline.
68    
69    12. Yet more internal refactoring has removed another argument from the match()
70        function. Special calls to this function are now indicated by setting a
71        value in a variable in the "match data" data block.
72    
73    13. Be more explicit in pcre_study() instead of relying on "default" for
74        opcodes that mean there is no starting character; this means that when new
75        ones are added and accidentally left out of pcre_study(), testing should
76        pick them up.
77    
78    14. The -s option of pcretest has been documented for ages as being an old
79        synonym of -m (show memory usage). I have changed it to mean "force study
80        for every regex", that is, assume /S for every regex. This is similar to -i
81        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
82        using it. It makes it easier to run collections of tests with and without
83        study enabled, and thereby test pcre_study() more easily. All the standard
84        tests are now run with and without -s (but some patterns can be marked as
85        "never study" - see 20 below).
86    
87    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
88        restoration of the capturing data to the outer values was not happening
89        correctly.
90    
91    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
92        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
93        pattern had matched an empty string, and so incorrectly returned a no
94        match.
95    
96    17. There was optimizing code for the last branch of non-capturing parentheses,
97        and also for the obeyed branch of a conditional subexpression, which used
98        tail recursion to cut down on stack usage. Unfortunately, not that there is
99        the possibility of (*THEN) occurring in these branches, tail recursion is
100        no longer possible because the return has to be checked for (*THEN). These
101        two optimizations have therefore been removed.
102    
103    18. If a pattern containing \R was studied, it was assumed that \R always
104        matched two bytes, thus causing the minimum subject length to be
105        incorrectly computed because \R can also match just one byte.
106    
107    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
108        was incorrectly computed.
109    
110    20. If /S is present twice on a test pattern in pcretest input, it *disables*
111        studying, thereby overriding the use of -s on the command line. This is
112        necessary for one or two tests to keep the output identical in both cases.
113    
114    21. When (*ACCEPT) was used in an assertion that matched an empty string and
115        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
116    
117    22. When an atomic group that contained a capturing parenthesis was
118        successfully matched, but the branch in which it appeared failed, the
119        capturing was not being forgotten if a higher numbered group was later
120        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
121        group 1 to "a", when in fact it should be unset. This applied to multi-
122        branched capturing and non-capturing groups, repeated or not, and also to
123        positive assertions (capturing in negative assertions is not well defined
124        in PCRE) and also to nested atomic groups.
125    
126    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
127        subject after a captured substring (to make it easier to tell which of a
128        number of identical substrings has been captured).
129    
130    24. The way atomic groups are processed by pcre_exec() has been changed so that
131        if they are repeated, backtracking one repetition now resets captured
132        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
133        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
134        "aaa". Previously, it would have been "a". As part of this code
135        refactoring, the way recursive calls are handled has also been changed.
136    
137    24. If an assertion condition captured any substrings, they were not passed
138        back unless some other capturing happened later. For example, if
139        (?(?=(a))a) was matched against "a", no capturing was returned.
140    
141    
142    Version 8.12 15-Jan-2011
143    ------------------------
144    
145    1.  Fixed some typos in the markup of the man pages, and wrote a script that
146        checks for such things as part of the documentation building process.
147    
148    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
149        --match-limit and --recursion-limit options (added for 8.11). In
150        particular, this made one of the standard tests fail. (The integer value
151        went into the wrong half of a long int.)
152    
153    3.  If the --colour option was given to pcregrep with -v (invert match), it
154        did strange things, either producing crazy output, or crashing. It should,
155        of course, ignore a request for colour when reporting lines that do not
156        match.
157    
158    4.  Another pcregrep bug caused similar problems if --colour was specified with
159        -M (multiline) and the pattern match finished with a line ending.
160    
161    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
162        matched in multiline mode, the following line was shown as part of the
163        match. This seems wrong, so I have changed it.
164    
165    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
166        the check for further matches in the same line (so they could be coloured)
167        to overrun the end of the current line. If another match was found, it was
168        incorrectly shown (and then shown again when found in the next line).
169    
170    7.  If pcregrep was compiled under Windows, there was a reference to the
171        function pcregrep_exit() before it was defined. I am assuming this was
172        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
173        reported by a user. I've moved the definition above the reference.
174    
175    
176    Version 8.11 10-Dec-2010
177    ------------------------
178    
179    1.  (*THEN) was not working properly if there were untried alternatives prior
180        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
181        backtracked to try for "b" instead of moving to the next alternative branch
182        at the same level (in this case, to look for "c"). The Perl documentation
183        is clear that when (*THEN) is backtracked onto, it goes to the "next
184        alternative in the innermost enclosing group".
185    
186    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
187        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
188        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
189        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
190        (*THEN).
191    
192    3.  If \s appeared in a character class, it removed the VT character from
193        the class, even if it had been included by some previous item, for example
194        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
195        of \s, but is part of the POSIX "space" class.)
196    
197    4.  A partial match never returns an empty string (because you can always
198        match an empty string at the end of the subject); however the checking for
199        an empty string was starting at the "start of match" point. This has been
200        changed to the "earliest inspected character" point, because the returned
201        data for a partial match starts at this character. This means that, for
202        example, /(?<=abc)def/ gives a partial match for the subject "abc"
203        (previously it gave "no match").
204    
205    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
206        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
207        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
208        has an implication that the given string is incomplete (because a partial
209        match is preferred over a full match). For this reason, these items now
210        give a partial match in this situation. [Aside: previously, the one case
211        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
212        match rather than a full match, which was wrong by the old rules, but is
213        now correct.]
214    
215    6.  There was a bug in the handling of #-introduced comments, recognized when
216        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
217        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
218        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
219        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
220        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
221        places in pcre_compile().
222    
223    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
224        comments when looking ahead for named forward references to subpatterns,
225        the only newline sequence it recognized was NL. It now handles newlines
226        according to the set newline convention.
227    
228    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
229        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
230        cater for a lack of strerror(). These oversights have been fixed.
231    
232    9.  Added --match-limit and --recursion-limit to pcregrep.
233    
234    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
235    
236    11. When the -o option was used, pcregrep was setting a return code of 1, even
237        when matches were found, and --line-buffered was not being honoured.
238    
239    12. Added an optional parentheses number to the -o and --only-matching options
240        of pcregrep.
241    
242    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
243        can match an empty string. The code to do it in pcretest and pcredemo
244        needed fixing:
245    
246        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
247            only one byte after an empty string match just before CRLF (this case
248            just got forgotten; "any" and "anycrlf" were OK).
249    
250        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
251            mode when an empty string match preceded an ASCII character followed by
252            a non-ASCII character. (The code for advancing by one character rather
253            than one byte was nonsense.)
254    
255        (c) The pcredemo.c sample program did not have any code at all to handle
256            the cases when CRLF is a valid newline sequence.
257    
258    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
259        as a starting offset was within the subject string. There is now a new
260        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
261        negative or greater than the length of the string. In order to test this,
262        pcretest is extended to allow the setting of negative starting offsets.
263    
264    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
265        starting offset points to the beginning of a UTF-8 character was
266        unnecessarily clumsy. I tidied it up.
267    
268    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
269        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
270    
271    17. Nobody had reported that the --include_dir option, which was added in
272        release 7.7 should have been called --include-dir (hyphen, not underscore)
273        for compatibility with GNU grep. I have changed it to --include-dir, but
274        left --include_dir as an undocumented synonym, and the same for
275        --exclude-dir, though that is not available in GNU grep, at least as of
276        release 2.5.4.
277    
278    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
279        characters from a string of bytes) have been redefined so as not to use
280        loops, in order to improve performance in some environments. At the same
281        time, I abstracted some of the common code into auxiliary macros to save
282        repetition (this should not affect the compiled code).
283    
284    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
285        compile-time error is now given if \c is not followed by an ASCII
286        character, that is, a byte less than 128. (In EBCDIC mode, the code is
287        different, and any byte value is allowed.)
288    
289    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
290        START_OPTIMIZE option, which is now allowed at compile time - but just
291        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
292        to pcregrep and other applications that have no direct access to PCRE
293        options. The new /Y option in pcretest sets this option when calling
294        pcre_compile().
295    
296    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
297        back references. Groups containing recursive back references were forced to
298        be atomic by that change, but in the case of named groups, the amount of
299        memory required was incorrectly computed, leading to "Failed: internal
300        error: code overflow". This has been fixed.
301    
302    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
303        pcretest.c, to avoid build problems in some Borland environments.
304    
305    
306    Version 8.10 25-Jun-2010
307    ------------------------
308    
309    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
310        THEN.
311    
312    2.  (*ACCEPT) was not working when inside an atomic group.
313    
314    3.  Inside a character class, \B is treated as a literal by default, but
315        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
316        causes the error). The code is unchanged, but I tidied the documentation.
317    
318    4.  Inside a character class, PCRE always treated \R and \X as literals,
319        whereas Perl faults them if its -w option is set. I have changed PCRE so
320        that it faults them when PCRE_EXTRA is set.
321    
322    5.  Added support for \N, which always matches any character other than
323        newline. (It is the same as "." when PCRE_DOTALL is not set.)
324    
325    6.  When compiling pcregrep with newer versions of gcc which may have
326        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
327        declared with attribute warn_unused_result" were given. Just casting the
328        result to (void) does not stop the warnings; a more elaborate fudge is
329        needed. I've used a macro to implement this.
330    
331    7.  Minor change to pcretest.c to avoid a compiler warning.
332    
333    8.  Added four artifical Unicode properties to help with an option to make
334        \s etc use properties (see next item). The new properties are: Xan
335        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
336    
337    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
338        use Unicode properties. (*UCP) at the start of a pattern can be used to set
339        this option. Modified pcretest to add /W to test this facility. Added
340        REG_UCP to make it available via the POSIX interface.
341    
342    10. Added --line-buffered to pcregrep.
343    
344    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
345        studied, and the match started with a letter with a code point greater than
346        127 whose first byte was different to the first byte of the other case of
347        the letter, the other case of this starting letter was not recognized
348        (#976).
349    
350    12. If a pattern that was studied started with a repeated Unicode property
351        test, for example, \p{Nd}+, there was the theoretical possibility of
352        setting up an incorrect bitmap of starting bytes, but fortunately it could
353        not have actually happened in practice until change 8 above was made (it
354        added property types that matched character-matching opcodes).
355    
356    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
357        possible starting bytes for non-anchored patterns.
358    
359    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
360        \R, and also a number of cases that involve Unicode properties, both
361        explicit and implicit when PCRE_UCP is set.
362    
363    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
364        input, it could crash or give wrong results if characters with values
365        greater than 0xc0 were present in the subject string. (Detail: it assumed
366        UTF-8 input when processing these items.)
367    
368    16. Added a lot of (int) casts to avoid compiler warnings in systems where
369        size_t is 64-bit (#991).
370    
371    17. Added a check for running out of memory when PCRE is compiled with
372        --disable-stack-for-recursion (#990).
373    
374    18. If the last data line in a file for pcretest does not have a newline on
375        the end, a newline was missing in the output.
376    
377    19. The default pcre_chartables.c file recognizes only ASCII characters (values
378        less than 128) in its various bitmaps. However, there is a facility for
379        generating tables according to the current locale when PCRE is compiled. It
380        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
381        characters, are recognized by isspace() and therefore were getting set in
382        these tables, and indeed these tables seem to approximate to ISO 8859. This
383        caused a problem in UTF-8 mode when pcre_study() was used to create a list
384        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
385        which of course cannot start UTF-8 characters. I have changed the code so
386        that only real ASCII characters (less than 128) and the correct starting
387        bytes for UTF-8 encodings are set for characters greater than 127 when in
388        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
389        altogether.)
390    
391    20. Added the /T option to pcretest so as to be able to run tests with non-
392        standard character tables, thus making it possible to include the tests
393        used for 19 above in the standard set of tests.
394    
395    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
396        reference to a subpattern the other side of a comment that contains an
397        opening parenthesis caused either an internal compiling error, or a
398        reference to the wrong subpattern.
399    
400    
401    Version 8.02 19-Mar-2010
402    ------------------------
403    
404    1.  The Unicode data tables have been updated to Unicode 5.2.0.
405    
406    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
407        configured.
408    
409    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
410        original author of that file, following a query about its status.
411    
412    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
413        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
414    
415    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
416        quantifier applied to a forward-referencing subroutine call, could compile
417        incorrect code or give the error "internal error: previously-checked
418        referenced subpattern not found".
419    
420    6.  Both MS Visual Studio and Symbian OS have problems with initializing
421        variables to point to external functions. For these systems, therefore,
422        pcre_malloc etc. are now initialized to local functions that call the
423        relevant global functions.
424    
425    7.  There were two entries missing in the vectors called coptable and poptable
426        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
427        I've fixed the data, and added a kludgy way of testing at compile time that
428        the lengths are correct (equal to the number of opcodes).
429    
430    8.  Following on from 7, I added a similar kludge to check the length of the
431        eint vector in pcreposix.c.
432    
433    9.  Error texts for pcre_compile() are held as one long string to avoid too
434        much relocation at load time. To find a text, the string is searched,
435        counting zeros. There was no check for running off the end of the string,
436        which could happen if a new error number was added without updating the
437        string.
438    
439    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
440    
441    11. \K was not working if it appeared in an atomic group or in a group that
442        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
443        \K is "not well defined" if used in an assertion. PCRE now accepts it if
444        the assertion is positive, but not if it is negative.
445    
446    12. Change 11 fortuitously reduced the size of the stack frame used in the
447        "match()" function of pcre_exec.c by one pointer. Forthcoming
448        implementation of support for (*MARK) will need an extra pointer on the
449        stack; I have reserved it now, so that the stack frame size does not
450        decrease.
451    
452    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
453        item in branch that calls a recursion is a subroutine call - as in the
454        second branch in the above example - was incorrectly given the compile-
455        time error "recursive call could loop indefinitely" because pcre_compile()
456        was not correctly checking the subroutine for matching a non-empty string.
457    
458    14. The checks for overrunning compiling workspace could trigger after an
459        overrun had occurred. This is a "should never occur" error, but it can be
460        triggered by pathological patterns such as hundreds of nested parentheses.
461        The checks now trigger 100 bytes before the end of the workspace.
462    
463    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
464    
465    
466    Version 8.01 19-Jan-2010
467    ------------------------
468    
469    1.  If a pattern contained a conditional subpattern with only one branch (in
470        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
471        computed the wrong minimum data length (which is of course zero for such
472        subpatterns). This could cause incorrect "no match" results.
473    
474    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
475        the pattern is reset in the first branch, pcre_compile() failed with
476        "internal error: code overflow at offset...". This happened only when
477        the reset was to the original external option setting. (An optimization
478        abstracts leading options settings into an external setting, which was the
479        cause of this.)
480    
481    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
482        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
483        assertion pattern did not match (meaning that the assertion was true), it
484        was incorrectly treated as false if the SKIP had been reached during the
485        matching. This also applied to assertions used as conditions.
486    
487    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
488        assertion subpattern, including such a pattern used as a condition,
489        unpredictable results occurred, instead of the error return
490        PCRE_ERROR_DFA_UITEM.
491    
492    5.  The C++ GlobalReplace function was not working like Perl for the special
493        situation when an empty string is matched. It now does the fancy magic
494        stuff that is necessary.
495    
496    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
497        removed. (These were left over from very, very early versions of PCRE.)
498    
499    7.  Some cosmetic changes to the code to make life easier when compiling it
500        as part of something else:
501    
502        (a) Change DEBUG to PCRE_DEBUG.
503    
504        (b) In pcre_compile(), rename the member of the "branch_chain" structure
505            called "current" as "current_branch", to prevent a collision with the
506            Linux macro when compiled as a kernel module.
507    
508        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
509            prevent a collision with the Linux macro when compiled as a kernel
510            module.
511    
512    8.  In pcre_compile() there are some checks for integer overflows that used to
513        cast potentially large values to (double). This has been changed to that
514        when building, a check for int64_t is made, and if it is found, it is used
515        instead, thus avoiding the use of floating point arithmetic. (There is no
516        other use of FP in PCRE.) If int64_t is not found, the fallback is to
517        double.
518    
519    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
520        2005 (difference between two addresses compared to an unsigned value).
521    
522    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
523        custom one, because of the following reported problem in Windows:
524    
525          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
526              under Win32.
527          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
528              therefore missing the function definition.
529          - The compiler thus generates a "C" signature for the test function.
530          - The linker fails to find the "C" function.
531          - PCRE fails to configure if asked to do so against libbz2.
532    
533    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
534        messages were output:
535    
536          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
537          rerunning libtoolize, to keep the correct libtool macros in-tree.
538          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
539    
540        I have done both of these things.
541    
542    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
543        most of the time, it *can* run out if it is given a pattern that contains a
544        runaway infinite recursion. I updated the discussion in the pcrestack man
545        page.
546    
547    13. Now that we have gone to the x.xx style of version numbers, the minor
548        version may start with zero. Using 08 or 09 is a bad idea because users
549        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
550        interpreted as invalid octal numbers. I've updated the previous comment in
551        configure.ac, and also added a check that gives an error if 08 or 09 are
552        used.
553    
554    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
555        causing partial matching to fail when the end of the subject matched \W
556        in a UTF-8 pattern where \W was quantified with a minimum of 3.
557    
558    15. There were some discrepancies between the declarations in pcre_internal.h
559        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
560        their definitions. The declarations used "const uschar *" and the
561        definitions used USPTR. Even though USPTR is normally defined as "const
562        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
563        reported that: "This difference in casting confuses some C++ compilers, for
564        example, SunCC recognizes above declarations as different functions and
565        generates broken code for hbpcre." I have changed the declarations to use
566        USPTR.
567    
568    16. GNU libtool is named differently on some systems. The autogen.sh script now
569        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
570        (FreeBSD).
571    
572    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
573        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
574        comment: "Figure out how to create a longlong from a string: strtoll and
575        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
576        instance, but it only takes 2 args instead of 3!"
577    
578    18. A subtle bug concerned with back references has been fixed by a change of
579        specification, with a corresponding code fix. A pattern such as
580        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
581        refers, was giving matches when it shouldn't. For example, xa=xaaa would
582        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
583        same bug. Such groups have to be quantified to be useful, or contained
584        inside another quantified group. (If there's no repetition, the reference
585        can never match.) The problem arises because, having left the group and
586        moved on to the rest of the pattern, a later failure that backtracks into
587        the group uses the captured value from the final iteration of the group
588        rather than the correct earlier one. I have fixed this in PCRE by forcing
589        any group that contains a reference to itself to be an atomic group; that
590        is, there cannot be any backtracking into it once it has completed. This is
591        similar to recursive and subroutine calls.
592    
593    
594    Version 8.00 19-Oct-09
595    ----------------------
596    
597    1.  The table for translating pcre_compile() error codes into POSIX error codes
598        was out-of-date, and there was no check on the pcre_compile() error code
599        being within the table. This could lead to an OK return being given in
600        error.
601    
602    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
603        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
604        in a Windows environment.
605    
606    3.  The pcregrep --count option prints the count for each file even when it is
607        zero, as does GNU grep. However, pcregrep was also printing all files when
608        --files-with-matches was added. Now, when both options are given, it prints
609        counts only for those files that have at least one match. (GNU grep just
610        prints the file name in this circumstance, but including the count seems
611        more useful - otherwise, why use --count?) Also ensured that the
612        combination -clh just lists non-zero counts, with no names.
613    
614    4.  The long form of the pcregrep -F option was incorrectly implemented as
615        --fixed_strings instead of --fixed-strings. This is an incompatible change,
616        but it seems right to fix it, and I didn't think it was worth preserving
617        the old behaviour.
618    
619    5.  The command line items --regex=pattern and --regexp=pattern were not
620        recognized by pcregrep, which required --regex pattern or --regexp pattern
621        (with a space rather than an '='). The man page documented the '=' forms,
622        which are compatible with GNU grep; these now work.
623    
624    6.  No libpcreposix.pc file was created for pkg-config; there was just
625        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
626    
627    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
628        when UCP support is not needed, by modifying the Python script that
629        generates it from Unicode data files. This should not matter if the module
630        is correctly used as a library, but I received one complaint about 50K of
631        unwanted data. My guess is that the person linked everything into his
632        program rather than using a library. Anyway, it does no harm.
633    
634    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
635        was a minimum greater than 1 for a wide character in a possessive
636        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
637        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
638        character. Chaos in the form of incorrect output or a compiling loop could
639        result.
640    
641    9.  The restrictions on what a pattern can contain when partial matching is
642        requested for pcre_exec() have been removed. All patterns can now be
643        partially matched by this function. In addition, if there are at least two
644        slots in the offset vector, the offset of the earliest inspected character
645        for the match and the offset of the end of the subject are set in them when
646        PCRE_ERROR_PARTIAL is returned.
647    
648    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
649        synonymous with PCRE_PARTIAL, for backwards compatibility, and
650        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
651        and may be more useful for multi-segment matching.
652    
653    11. Partial matching with pcre_exec() is now more intuitive. A partial match
654        used to be given if ever the end of the subject was reached; now it is
655        given only if matching could not proceed because another character was
656        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
657        string "Z", which now yields "no match" instead of "partial match". In the
658        case of pcre_dfa_exec(), "no match" is given if every matching path for the
659        final character ended with (*FAIL).
660    
661    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
662        if the pattern had a "must contain" character that was already found in the
663        earlier partial match, unless partial matching was again requested. For
664        example, with the pattern /dog.(body)?/, the "must contain" character is
665        "g". If the first part-match was for the string "dog", restarting with
666        "sbody" failed. This bug has been fixed.
667    
668    13. The string returned by pcre_dfa_exec() after a partial match has been
669        changed so that it starts at the first inspected character rather than the
670        first character of the match. This makes a difference only if the pattern
671        starts with a lookbehind assertion or \b or \B (\K is not supported by
672        pcre_dfa_exec()). It's an incompatible change, but it makes the two
673        matching functions compatible, and I think it's the right thing to do.
674    
675    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
676        so that the demonstration program is easily available in environments where
677        PCRE has not been installed from source.
678    
679    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
680        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
681        library.
682    
683    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
684        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
685        is not the first non-POSIX option to be added. Clearly some people find
686        these options useful.
687    
688    17. If a caller to the POSIX matching function regexec() passes a non-zero
689        value for nmatch with a NULL value for pmatch, the value of
690        nmatch is forced to zero.
691    
692    18. RunGrepTest did not have a test for the availability of the -u option of
693        the diff command, as RunTest does. It now checks in the same way as
694        RunTest, and also checks for the -b option.
695    
696    19. If an odd number of negated classes containing just a single character
697        interposed, within parentheses, between a forward reference to a named
698        subpattern and the definition of the subpattern, compilation crashed with
699        an internal error, complaining that it could not find the referenced
700        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
701        [The bug was that it was starting one character too far in when skipping
702        over the character class, thus treating the ] as data rather than
703        terminating the class. This meant it could skip too much.]
704    
705    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
706        /g option in pcretest when the pattern contains \K, which makes it possible
707        to have an empty string match not at the start, even when the pattern is
708        anchored. Updated pcretest and pcredemo to use this option.
709    
710    21. If the maximum number of capturing subpatterns in a recursion was greater
711        than the maximum at the outer level, the higher number was returned, but
712        with unset values at the outer level. The correct (outer level) value is
713        now given.
714    
715    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
716        PCRE did not set those parentheses (unlike Perl). I have now found a way to
717        make it do so. The string so far is captured, making this feature
718        compatible with Perl.
719    
720    23. The tests have been re-organized, adding tests 11 and 12, to make it
721        possible to check the Perl 5.10 features against Perl 5.10.
722    
723    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
724        pattern matches a fixed length string. PCRE did not allow this; now it
725        does. Neither allows recursion.
726    
727    25. I finally figured out how to implement a request to provide the minimum
728        length of subject string that was needed in order to match a given pattern.
729        (It was back references and recursion that I had previously got hung up
730        on.) This code has now been added to pcre_study(); it finds a lower bound
731        to the length of subject needed. It is not necessarily the greatest lower
732        bound, but using it to avoid searching strings that are too short does give
733        some useful speed-ups. The value is available to calling programs via
734        pcre_fullinfo().
735    
736    26. While implementing 25, I discovered to my embarrassment that pcretest had
737        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
738        study optimizations had never been tested with that matching function.
739        Oops. What is worse, even when it was passed study data, there was a bug in
740        pcre_dfa_exec() that meant it never actually used it. Double oops. There
741        were also very few tests of studied patterns with pcre_dfa_exec().
742    
743    27. If (?| is used to create subpatterns with duplicate numbers, they are now
744        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
745        on the other side of the coin, they are no longer allowed to have different
746        names, because these cannot be distinguished in PCRE, and this has caused
747        confusion. (This is a difference from Perl.)
748    
749    28. When duplicate subpattern names are present (necessarily with different
750        numbers, as required by 27 above), and a test is made by name in a
751        conditional pattern, either for a subpattern having been matched, or for
752        recursion in such a pattern, all the associated numbered subpatterns are
753        tested, and the overall condition is true if the condition is true for any
754        one of them. This is the way Perl works, and is also more like the way
755        testing by number works.
756    
757    
758    Version 7.9 11-Apr-09
759    ---------------------
760    
761    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
762        (pcretest), all targets were linked against these libraries. This included
763        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
764        libraries. This caused unwanted dependencies to be created. This problem
765        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
766        pcretest is linked with readline.
767    
768    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
769        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
770        moved outside it again, because FALSE and TRUE are already defined in AIX,
771        but BOOL is not.
772    
773    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
774        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
775    
776    4.  The pcregrep documentation said spaces were inserted as well as colons (or
777        hyphens) following file names and line numbers when outputting matching
778        lines. This is not true; no spaces are inserted. I have also clarified the
779        wording for the --colour (or --color) option.
780    
781    5.  In pcregrep, when --colour was used with -o, the list of matching strings
782        was not coloured; this is different to GNU grep, so I have changed it to be
783        the same.
784    
785    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
786        each matching line was coloured. Now it goes on to look for further matches
787        of any of the test patterns, which is the same behaviour as GNU grep.
788    
789    7.  A pattern that could match an empty string could cause pcregrep to loop; it
790        doesn't make sense to accept an empty string match in pcregrep, so I have
791        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
792        seems to be how GNU grep behaves.
793    
794    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
795        start or after a newline", because the conditional assertion was not being
796        correctly handled. The rule now is that both the assertion and what follows
797        in the first alternative must satisfy the test.
798    
799    9.  If auto-callout was enabled in a pattern with a conditional group whose
800        condition was an assertion, PCRE could crash during matching, both with
801        pcre_exec() and pcre_dfa_exec().
802    
803    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
804        used for matching.
805    
806    11. Unicode property support in character classes was not working for
807        characters (bytes) greater than 127 when not in UTF-8 mode.
808    
809    12. Added the -M command line option to pcretest.
810    
811    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
812    
813    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
814    
815    16. Added comments and documentation about mis-use of no_arg in the C++
816        wrapper.
817    
818    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
819        from Martin Jerabek that uses macro names for all relevant character and
820        string constants.
821    
822    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
823        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
824        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
825        these, but not everybody uses configure.
826    
827    19. A conditional group that had only one branch was not being correctly
828        recognized as an item that could match an empty string. This meant that an
829        enclosing group might also not be so recognized, causing infinite looping
830        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
831        with the subject "ab", where knowledge that the repeated group can match
832        nothing is needed in order to break the loop.
833    
834    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
835        exec(), but without supplying a callout function, matching went wrong.
836    
837    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
838        leak if the size of the offset vector was greater than 30. When the vector
839        is smaller, the saved offsets during recursion go onto a local stack
840        vector, but for larger vectors malloc() is used. It was failing to free
841        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
842        error, in fact).
843    
844    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
845        heapframe that is used only when UTF-8 support is enabled. This caused no
846        problem, but was untidy.
847    
848    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
849        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
850        included within another project.
851    
852    24. Steven Van Ingelgem's patches to add more options to the CMake support,
853        slightly modified by me:
854    
855          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
856              not building pcregrep.
857    
858          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
859              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
860    
861    25. Forward references, both numeric and by name, in patterns that made use of
862        duplicate group numbers, could behave incorrectly or give incorrect errors,
863        because when scanning forward to find the reference group, PCRE was not
864        taking into account the duplicate group numbers. A pattern such as
865        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
866    
867    26. Changed a few more instances of "const unsigned char *" to USPTR, making
868        the feature of a custom pointer more persuasive (as requested by a user).
869    
870    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
871        pcretest.c, inside #ifndefs, because it seems they are sometimes already
872        pre-defined.
873    
874    28. Added support for (*UTF8) at the start of a pattern.
875    
876    29. Arrange for flags added by the "release type" setting in CMake to be shown
877        in the configuration summary.
878    
879    
880    Version 7.8 05-Sep-08
881    ---------------------
882    
883    1.  Replaced UCP searching code with optimized version as implemented for Ad
884        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
885        stage table and inline lookup instead of a function, giving speed ups of 2
886        to 5 times on some simple patterns that I tested. Permission was given to
887        distribute the MultiStage2.py script that generates the tables (it's not in
888        the tarball, but is in the Subversion repository).
889    
890    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
891        scripts.
892    
893    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
894        a group with a zero qualifier. The result of the study could be incorrect,
895        or the function might crash, depending on the pattern.
896    
897    4.  Caseless matching was not working for non-ASCII characters in back
898        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
899        It now works when Unicode Property Support is available.
900    
901    5.  In pcretest, an escape such as \x{de} in the data was always generating
902        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
903        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
904        truncation.
905    
906    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
907    
908    7.  Added two (int) casts to pcregrep when printing the difference of two
909        pointers, in case they are 64-bit values.
910    
911    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
912        test 2 if it fails.
913    
914    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
915        and a #define of that name to empty if it is not externally set. This is to
916        allow users of MSVC to set it if necessary.
917    
918    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
919        the convenience functions in the pcre_get.c source file.
920    
921    11. An option change at the start of a pattern that had top-level alternatives
922        could cause overwriting and/or a crash. This command provoked a crash in
923        some environments:
924    
925          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
926    
927        This potential security problem was recorded as CVE-2008-2371.
928    
929    12. For a pattern where the match had to start at the beginning or immediately
930        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
931        pcre_dfa_exec() could read past the end of the passed subject if there was
932        no match. To help with detecting such bugs (e.g. with valgrind), I modified
933        pcretest so that it places the subject at the end of its malloc-ed buffer.
934    
935    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
936        exec() might read past the end of the data buffer in UTF-8 mode.
937    
938    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
939        the data contained the byte 0x85 as part of a UTF-8 character within its
940        first line. This applied both to normal and DFA matching.
941    
942    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
943        /^[^d]*?$/8 failed to match "abc".
944    
945    16. Added a missing copyright notice to pcrecpp_internal.h.
946    
947    17. Make it more clear in the documentation that values returned from
948        pcre_exec() in ovector are byte offsets, not character counts.
949    
950    18. Tidied a few places to stop certain compilers from issuing warnings.
951    
952    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
953        supplied by Stefan Weber. I made a further small update for 7.8 because
954        there is a change of source arrangements: the pcre_searchfuncs.c module is
955        replaced by pcre_ucd.c.
956    
957    
958    Version 7.7 07-May-08
959    ---------------------
960    
961    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
962        a string to a long long, pretend we don't even have a long long." This is
963        done by checking for the strtoq, strtoll, and _strtoi64 functions.
964    
965    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
966        pre-7.6 versions, which defined a global no_arg variable instead of putting
967        it in the RE class. (See also #8 below.)
968    
969    3.  Remove a line of dead code, identified by coverity and reported by Nuno
970        Lopes.
971    
972    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
973    
974        (1) The include/exclude patterns were being applied to the whole pathnames
975            of files, instead of just to the final components.
976    
977        (2) If there was more than one level of directory, the subdirectories were
978            skipped unless they satisfied the include/exclude conditions. This is
979            inconsistent with GNU grep (and could even be seen as contrary to the
980            pcregrep specification - which I improved to make it absolutely clear).
981            The action now is always to scan all levels of directory, and just
982            apply the include/exclude patterns to regular files.
983    
984    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
985        --exclude_dir in the tests to avoid scanning .svn directories.
986    
987    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
988        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
989        doesn't support NULs in patterns.
990    
991    7.  Added some missing "const"s to declarations of static tables in
992        pcre_compile.c and pcre_dfa_exec.c.
993    
994    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
995        caused by fix #2  above. (Subsequently also a second patch to fix the
996        first patch. And a third patch - this was a messy problem.)
997    
998    9.  Applied Craig's patch to remove the use of push_back().
999    
1000    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1001        matching function regexec().
1002    
1003    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1004        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1005        references. PCRE supports relative numbers with this syntax (I don't think
1006        Oniguruma does).
1007    
1008    12. Previously, a group with a zero repeat such as (...){0} was completely
1009        omitted from the compiled regex. However, this means that if the group
1010        was called as a subroutine from elsewhere in the pattern, things went wrong
1011        (an internal error was given). Such groups are now left in the compiled
1012        pattern, with a new opcode that causes them to be skipped at execution
1013        time.
1014    
1015    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1016        to the way PCRE behaves:
1017    
1018        (a) A lone ] character is dis-allowed (Perl treats it as data).
1019    
1020        (b) A back reference to an unmatched subpattern matches an empty string
1021            (Perl fails the current match path).
1022    
1023        (c) A data ] in a character class must be notated as \] because if the
1024            first data character in a class is ], it defines an empty class. (In
1025            Perl it is not possible to have an empty class.) The empty class []
1026            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1027            The negative empty class [^] matches any one character, independently
1028            of the DOTALL setting.
1029    
1030    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1031        non-existent subpattern following a character class starting with ']' and
1032        containing () gave an internal compiling error instead of "reference to
1033        non-existent subpattern". Fortunately, when the pattern did exist, the
1034        compiled code was correct. (When scanning forwards to check for the
1035        existencd of the subpattern, it was treating the data ']' as terminating
1036        the class, so got the count wrong. When actually compiling, the reference
1037        was subsequently set up correctly.)
1038    
1039    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1040        it was being rejected as not supported by pcre_dfa_exec(), even though
1041        other assertions are supported. I have made pcre_dfa_exec() support
1042        (*FAIL).
1043    
1044    16. The implementation of 13c above involved the invention of a new opcode,
1045        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1046        cannot be changed at match time, I realized I could make a small
1047        improvement to matching performance by compiling OP_ALLANY instead of
1048        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1049        on the OP_ANY path.
1050    
1051    17. Compiling pcretest on Windows with readline support failed without the
1052        following two fixes: (1) Make the unistd.h include conditional on
1053        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1054    
1055    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1056        ncurses library to be included for pcretest when ReadLine support is
1057        requested, but also to allow for it to be overridden. This patch came from
1058        Daniel Bergström.
1059    
1060    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1061        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1062        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1063        spotting this.
1064    
1065    
1066    Version 7.6 28-Jan-08
1067    ---------------------
1068    
1069    1.  A character class containing a very large number of characters with
1070        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1071        overflow.
1072    
1073    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1074        HAVE_LONG_LONG is not defined.
1075    
1076    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1077        bring them up to date and include new features. This patch includes:
1078    
1079        - Fixed PH's badly added libz and libbz2 support.
1080        - Fixed a problem with static linking.
1081        - Added pcredemo. [But later removed - see 7 below.]
1082        - Fixed dftables problem and added an option.
1083        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1084            HAVE_LONG_LONG.
1085        - Added readline support for pcretest.
1086        - Added an listing of the option settings after cmake has run.
1087    
1088    4.  A user submitted a patch to Makefile that makes it easy to create
1089        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1090        Makefile.am that cause it to include this special target, without
1091        affecting anything else. Note that the same mingw target plus all
1092        the other distribution libraries and programs are now supported
1093        when configuring with CMake (see 6 below) instead of with
1094        Configure/Make.
1095    
1096    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1097        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1098        exported in the Windows port". It has not yet been confirmed that the patch
1099        solves the problem, but it does no harm.
1100    
1101    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1102        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1103        with CMake, and also correct the comment about stack recursion.
1104    
1105    7.  Remove the automatic building of pcredemo from the ./configure system and
1106        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1107        of a program that users should build themselves after PCRE is installed, so
1108        building it automatically is not really right. What is more, it gave
1109        trouble in some build environments.
1110    
1111    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1112    
1113    
1114    Version 7.5 10-Jan-08
1115    ---------------------
1116    
1117    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1118        values in parens when parsing an RE using the C++ wrapper."
1119    
1120    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1121        Characters greater than 255 were excluded from the class instead of being
1122        included.
1123    
1124    3.  The same bug as (2) above applied to negated POSIX classes such as
1125        [:^space:].
1126    
1127    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1128        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1129        I have changed it.
1130    
1131    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1132        first named subpattern) and a construct such as (?&a) would reference the
1133        first named subpattern whose name started with "a" (in other words, the
1134        length check was missing). Both these problems are fixed. "Subpattern name
1135        expected" is now given for (?&) (a zero-length name), and this patch also
1136        makes it give the same error for \k'' (previously it complained that that
1137        was a reference to a non-existent subpattern).
1138    
1139    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1140        this is right because (?- can be followed by option settings as well as by
1141        digits. I have, however, made the messages clearer.
1142    
1143    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1144        than the number used in the conditional) now cause a compile-time error.
1145        This is actually not compatible with Perl, which accepts such patterns, but
1146        treats the conditional as always being FALSE (as PCRE used to), but it
1147        seems to me that giving a diagnostic is better.
1148    
1149    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1150        and messages.
1151    
1152    9.  Fix two occurrences of "backslash" in comments that should have been
1153        "backspace".
1154    
1155    10. Remove two redundant lines of code that can never be obeyed (their function
1156        was moved elsewhere).
1157    
1158    11. The program that makes PCRE's Unicode character property table had a bug
1159        which caused it to generate incorrect table entries for sequences of
1160        characters that have the same character type, but are in different scripts.
1161        It amalgamated them into a single range, with the script of the first of
1162        them. In other words, some characters were in the wrong script. There were
1163        thirteen such cases, affecting characters in the following ranges:
1164    
1165          U+002b0 - U+002c1
1166          U+0060c - U+0060d
1167          U+0061e - U+00612
1168          U+0064b - U+0065e
1169          U+0074d - U+0076d
1170          U+01800 - U+01805
1171          U+01d00 - U+01d77
1172          U+01d9b - U+01dbf
1173          U+0200b - U+0200f
1174          U+030fc - U+030fe
1175          U+03260 - U+0327f
1176          U+0fb46 - U+0fbb1
1177          U+10450 - U+1049d
1178    
1179    12. The -o option (show only the matching part of a line) for pcregrep was not
1180        compatible with GNU grep in that, if there was more than one match in a
1181        line, it showed only the first of them. It now behaves in the same way as
1182        GNU grep.
1183    
1184    13. If the -o and -v options were combined for pcregrep, it printed a blank
1185        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1186        does the same. The return code can be used to tell if there were any
1187        non-matching lines.
1188    
1189    14. Added --file-offsets and --line-offsets to pcregrep.
1190    
1191    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1192        infinitely looping recursion. The bug was that positive lookaheads were not
1193        being skipped when checking for a possible empty match (negative lookaheads
1194        and both kinds of lookbehind were skipped).
1195    
1196    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1197        inclusion of <windows.h> to before rather than after the definition of
1198        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1199    
1200    17. Specifying a possessive quantifier with a specific limit for a Unicode
1201        character property caused pcre_compile() to compile bad code, which led at
1202        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1203        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1204        caused the error; without that there was no problem.
1205    
1206    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1207    
1208    19. Added --enable-pcretest-libreadline.
1209    
1210    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1211        RE::GlobalReplace(). As a result, the number of replacements returned was
1212        double what it should be. I removed one of the increments, but Craig sent a
1213        later patch that removed the other one (the right fix) and added unit tests
1214        that check the return values (which was not done before).
1215    
1216    21. Several CMake things:
1217    
1218        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1219            the names libpcre and libpcreposix, not just pcre and pcreposix.
1220    
1221        (2) The above change means that pcretest and pcregrep are now correctly
1222            linked with the newly-built libraries, not previously installed ones.
1223    
1224        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1225    
1226    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1227        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1228        UTF-8 newline character). The key issue is that the pattern starts .*;
1229        this means that the match must be either at the beginning, or after a
1230        newline. The bug was in the code for advancing after a failed match and
1231        checking that the new position followed a newline. It was not taking
1232        account of UTF-8 characters correctly.
1233    
1234    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1235        character classes. PCRE was not treating the sequence [:...:] as a
1236        character class unless the ... were all letters. Perl, however, seems to
1237        allow any characters between [: and :], though of course it rejects as
1238        unknown any "names" that contain non-letters, because all the known class
1239        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1240        for example, whereas PCRE did not - it did not recognize a POSIX character
1241        class. This seemed a bit dangerous, so the code has been changed to be
1242        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1243        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1244        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1245        Perl does, and where it didn't before.
1246    
1247    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1248        Windows environments %n is disabled by default.
1249    
1250    
1251    Version 7.4 21-Sep-07
1252    ---------------------
1253    
1254    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1255        means that a class such as [\s] counted as "explicit reference to CR or
1256        LF". That isn't really right - the whole point of the change was to try to
1257        help when there was an actual mention of one of the two characters. So now
1258        the change happens only if \r or \n (or a literal CR or LF) character is
1259        encountered.
1260    
1261    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1262        of both had grown to the point where there were only 3 bits left.
1263        Fortunately, there was spare space in the data structure, and so I have
1264        moved the internal flags into a new 16-bit field to free up more option
1265        bits.
1266    
1267    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1268        but did not set the internal JCHANGED flag - either of these is enough to
1269        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1270        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1271        start sets both bits.
1272    
1273    4.  Added options (at build time, compile time, exec time) to change \R from
1274        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1275    
1276    5.  doc/pcresyntax.html was missing from the distribution.
1277    
1278    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1279        compatibility, even though it is no longer used.
1280    
1281    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1282        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1283        windows.h file is present (where different names are used). [This was
1284        reversed later after testing - see 16 below.]
1285    
1286    8.  Changed all #include <config.h> to #include "config.h". There were also
1287        some further <pcre.h> cases that I changed to "pcre.h".
1288    
1289    9.  When pcregrep was used with the --colour option, it missed the line ending
1290        sequence off the lines that it output.
1291    
1292    10. It was pointed out to me that arrays of string pointers cause lots of
1293        relocations when a shared library is dynamically loaded. A technique of
1294        using a single long string with a table of offsets can drastically reduce
1295        these. I have refactored PCRE in four places to do this. The result is
1296        dramatic:
1297    
1298          Originally:                          290
1299          After changing UCP table:            187
1300          After changing error message table:   43
1301          After changing table of "verbs"       36
1302          After changing table of Posix names   22
1303    
1304        Thanks to the folks working on Gregex for glib for this insight.
1305    
1306    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1307        unicode-properties was also set.
1308    
1309    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1310    
1311    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1312        checked only for CRLF.
1313    
1314    14. Added casts to pcretest.c to avoid compiler warnings.
1315    
1316    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1317    
1318    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1319        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1320        entirely. This removes changes made in 7 above.
1321    
1322    17. The CMake files have been updated, and there is now more information about
1323        building with CMake in the NON-UNIX-USE document.
1324    
1325    
1326    Version 7.3 28-Aug-07
1327    ---------------------
1328    
1329     1. In the rejigging of the build system that eventually resulted in 7.1, the
1330        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1331        brackets there is not right, since it causes compilers to look for an
1332        installed pcre.h, not the version that is in the source that is being
1333        compiled (which of course may be different). I have changed it back to:
1334    
1335          #include "pcre.h"
1336    
1337        I have a vague recollection that the change was concerned with compiling in
1338        different directories, but in the new build system, that is taken care of
1339        by the VPATH setting the Makefile.
1340    
1341     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1342        when the subject happened to end in the byte 0x85 (e.g. if the last
1343        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1344        characters but of course it shouldn't be taken as a newline when it is part
1345        of another character. The bug was that, for an unlimited repeat of . in
1346        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1347        characters when looking for a newline.
1348    
1349     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1350    
1351     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1352        in debug output.
1353    
1354     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1355        long printing in the pcrecpp unittest when running under MinGW.
1356    
1357     6. ESC_K was left out of the EBCDIC table.
1358    
1359     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1360        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1361        limit also applies to "virtual nesting" when a pattern is recursive, and in
1362        this case 1000 isn't so big. I have been able to remove this limit at the
1363        expense of backing off one optimization in certain circumstances. Normally,
1364        when pcre_exec() would call its internal match() function recursively and
1365        immediately return the result unconditionally, it uses a "tail recursion"
1366        feature to save stack. However, when a subpattern that can match an empty
1367        string has an unlimited repetition quantifier, it no longer makes this
1368        optimization. That gives it a stack frame in which to save the data for
1369        checking that an empty string has been matched. Previously this was taken
1370        from the 1000-entry workspace that had been reserved. So now there is no
1371        explicit limit, but more stack is used.
1372    
1373     8. Applied Daniel's patches to solve problems with the import/export magic
1374        syntax that is required for Windows, and which was going wrong for the
1375        pcreposix and pcrecpp parts of the library. These were overlooked when this
1376        problem was solved for the main library.
1377    
1378     9. There were some crude static tests to avoid integer overflow when computing
1379        the size of patterns that contain repeated groups with explicit upper
1380        limits. As the maximum quantifier is 65535, the maximum group length was
1381        set at 30,000 so that the product of these two numbers did not overflow a
1382        32-bit integer. However, it turns out that people want to use groups that
1383        are longer than 30,000 bytes (though not repeat them that many times).
1384        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1385        made it possible to implement the integer overflow checks in a much more
1386        dynamic way, which I have now done. The artificial limitation on group
1387        length has been removed - we now have only the limit on the total length of
1388        the compiled pattern, which depends on the LINK_SIZE setting.
1389    
1390    10. Fixed a bug in the documentation for get/copy named substring when
1391        duplicate names are permitted. If none of the named substrings are set, the
1392        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1393        empty string.
1394    
1395    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1396        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1397        because the ] is interpreted as the first data character and the
1398        terminating ] is not found. PCRE has been made compatible with Perl in this
1399        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1400        cause memory overwriting.
1401    
1402    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1403        string has been matched (to stop an infinite loop). It was not recognizing
1404        a conditional subpattern that could match an empty string if that
1405        subpattern was within another subpattern. For example, it looped when
1406        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1407        condition was not nested. This bug has been fixed.
1408    
1409    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1410        past the start of the subject in the presence of bytes with the top bit
1411        set, for example "\x8aBCD".
1412    
1413    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1414        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1415    
1416    14. Optimized (?!) to (*FAIL).
1417    
1418    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1419        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1420        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1421        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1422        does: it's just the validity check that is more restrictive.
1423    
1424    16. Inserted checks for integer overflows during escape sequence (backslash)
1425        processing, and also fixed erroneous offset values for syntax errors during
1426        backslash processing.
1427    
1428    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1429        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1430    
1431    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1432        caused an overrun.
1433    
1434    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1435        something other than just ASCII characters) inside a group that had an
1436        unlimited repeat caused a loop at compile time (while checking to see
1437        whether the group could match an empty string).
1438    
1439    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1440        [\P{Any}] did so. (Error in the code for printing property names.)
1441    
1442    21. An orphan \E inside a character class could cause a crash.
1443    
1444    22. A repeated capturing bracket such as (A)? could cause a wild memory
1445        reference during compilation.
1446    
1447    23. There are several functions in pcre_compile() that scan along a compiled
1448        expression for various reasons (e.g. to see if it's fixed length for look
1449        behind). There were bugs in these functions when a repeated \p or \P was
1450        present in the pattern. These operators have additional parameters compared
1451        with \d, etc, and these were not being taken into account when moving along
1452        the compiled data. Specifically:
1453    
1454        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1455            length.
1456    
1457        (b) An item such as \pL+ within a repeated group could cause crashes or
1458            loops.
1459    
1460        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1461            "reference to non-existent subpattern" error.
1462    
1463        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1464    
1465    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1466        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1467    
1468    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1469    
1470    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1471        character were causing crashes (broken optimization).
1472    
1473    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1474        \p or \P) caused a compile-time loop.
1475    
1476    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1477        break. For example, the unstudied pattern [\r\n]A does not match the string
1478        "\r\nA" because change 7.0/46 below moves the current point on by two
1479        characters after failing to match at the start. However, the pattern \nA
1480        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1481        the same is true. There doesn't seem any very clean way out of this, but
1482        what I have chosen to do makes the common cases work: PCRE now takes note
1483        of whether there can be an explicit match for \r or \n anywhere in the
1484        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1485        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1486        pattern has explicit CR or LF references.
1487    
1488    29. Added (*CR) etc for changing newline setting at start of pattern.
1489    
1490    
1491    Version 7.2 19-Jun-07
1492    ---------------------
1493    
1494     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1495        which is apparently normally available under Windows.
1496    
1497     2. Re-jig the pcregrep tests with different newline settings in an attempt
1498        to make them independent of the local environment's newline setting.
1499    
1500     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1501    
1502     4. Some of the "internals" tests were previously cut out when the link size
1503        was not 2, because the output contained actual offsets. The recent new
1504        "Z" feature of pcretest means that these can be cut out, making the tests
1505        usable with all link sizes.
1506    
1507     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1508        stack recursion. This gives a massive performance boost under BSD, but just
1509        a small improvement under Linux. However, it saves one field in the frame
1510        in all cases.
1511    
1512     6. Added more features from the forthcoming Perl 5.10:
1513    
1514        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1515            recursion call. It refers to the nth most recently opened parentheses.
1516    
1517        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1518            to be opened parentheses.
1519    
1520        (c) Conditions that refer to capturing parentheses can be specified
1521            relatively, for example, (?(-2)... or (?(+3)...
1522    
1523        (d) \K resets the start of the current match so that everything before
1524            is not part of it.
1525    
1526        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1527    
1528        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1529            reference syntax.
1530    
1531        (g) (?| introduces a group in which the numbering of parentheses in each
1532            alternative starts with the same number.
1533    
1534        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1535    
1536     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1537        PCRE_INFO_JCHANGED.
1538    
1539     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1540        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1541        for detecting groups that can match an empty string.
1542    
1543     9. A pattern with a very large number of alternatives (more than several
1544        hundred) was running out of internal workspace during the pre-compile
1545        phase, where pcre_compile() figures out how much memory will be needed. A
1546        bit of new cunning has reduced the workspace needed for groups with
1547        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1548        workspace instead of running out of the 4096 that are available.
1549    
1550    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1551    
1552    11. Applied patch from Google to remove an optimization that didn't quite work.
1553        The report of the bug said:
1554    
1555          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1556          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1557          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1558    
1559    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1560        it matched the wrong number of bytes.
1561    
1562    
1563    Version 7.1 24-Apr-07
1564    ---------------------
1565    
1566     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1567        that is more "standard", making use of automake and other Autotools. There
1568        is some re-arrangement of the files and adjustment of comments consequent
1569        on this.
1570    
1571     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1572        for recursive directory scanning broke on some systems because the files
1573        are not scanned in any specific order and on different systems the order
1574        was different. A call to "sort" has been inserted into RunGrepTest for the
1575        approprate test as a short-term fix. In the longer term there may be an
1576        alternative.
1577    
1578     3. I had an email from Eric Raymond about problems translating some of PCRE's
1579        man pages to HTML (despite the fact that I distribute HTML pages, some
1580        people do their own conversions for various reasons). The problems
1581        concerned the use of low-level troff macros .br and .in. I have therefore
1582        removed all such uses from the man pages (some were redundant, some could
1583        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1584        HTML has been updated to handle .nf/.fi and to complain if it encounters
1585        .br or .in.
1586    
1587     4. Updated comments in configure.ac that get placed in config.h.in and also
1588        arranged for config.h to be included in the distribution, with the name
1589        config.h.generic, for the benefit of those who have to compile without
1590        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1591    
1592     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1593        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1594        makevp.bat for the current PCRE, using the additional files
1595        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1596    
1597     6. A Windows user reported a minor discrepancy with test 2, which turned out
1598        to be caused by a trailing space on an input line that had got lost in his
1599        copy. The trailing space was an accident, so I've just removed it.
1600    
1601     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1602        that is needed.
1603    
1604     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1605        as "const" (a) because they are and (b) because it helps the PHP
1606        maintainers who have recently made a script to detect big data structures
1607        in the php code that should be moved to the .rodata section. I remembered
1608        to update Builducptable as well, so it won't revert if ucptable.h is ever
1609        re-created.
1610    
1611     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1612        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1613        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1614        support is not required. This saves 1.5-2K of code, which is important in
1615        some applications.
1616    
1617        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1618        so as not to refer to the tables, even though these functions will never be
1619        called when UTF-8 support is disabled. Otherwise there are problems with a
1620        shared library.
1621    
1622    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1623    
1624        (a) It was defining its arguments as char * instead of void *.
1625    
1626        (b) It was assuming that all moves were upwards in memory; this was true
1627            a long time ago when I wrote it, but is no longer the case.
1628    
1629        The emulated memove() is provided for those environments that have neither
1630        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1631        is clearly not the case, as these two bugs were recently reported.
1632    
1633    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1634        and Detrail to create the HTML documentation, the .txt form of the man
1635        pages, and it removes trailing spaces from listed files. It also creates
1636        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1637        case, it wraps all the #defines with #ifndefs. This script should be run
1638        before "make dist".
1639    
1640    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1641        with Unicode property support.
1642    
1643        (a) For a maximizing quantifier, if the two different cases of the
1644            character were of different lengths in their UTF-8 codings (there are
1645            some cases like this - I found 11), and the matching function had to
1646            back up over a mixture of the two cases, it incorrectly assumed they
1647            were both the same length.
1648    
1649        (b) When PCRE was configured to use the heap rather than the stack for
1650            recursion during matching, it was not correctly preserving the data for
1651            the other case of a UTF-8 character when checking ahead for a match
1652            while processing a minimizing repeat. If the check also involved
1653            matching a wide character, but failed, corruption could cause an
1654            erroneous result when trying to check for a repeat of the original
1655            character.
1656    
1657    13. Some tidying changes to the testing mechanism:
1658    
1659        (a) The RunTest script now detects the internal link size and whether there
1660            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1661            values substituted by "configure". (The RunGrepTest script already did
1662            this for UTF-8.) The configure.ac script no longer substitutes the
1663            relevant variables.
1664    
1665        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1666            with length and offset values. This means that the output is different
1667            for different internal link sizes. Test 2 is skipped for link sizes
1668            other than 2 because of this, bypassing the problem. Unfortunately,
1669            there was also a test in test 3 (the locale tests) that used /B and
1670            failed for link sizes other than 2. Rather than cut the whole test out,
1671            I have added a new /Z option to pcretest that replaces the length and
1672            offset values with spaces. This is now used to make test 3 independent
1673            of link size. (Test 2 will be tidied up later.)
1674    
1675    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1676        segmentation fault instead of returning the appropriate error message.
1677    
1678    15. In multiline mode when the newline sequence was set to "any", the pattern
1679        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1680        This doesn't seem right; it now treats the CRLF combination as the line
1681        ending, and so does not match in that case. It's only a pattern such as ^$
1682        that would hit this one: something like ^ABC$ would have failed after \r
1683        and then tried again after \r\n.
1684    
1685    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1686        in an attempt to make files that differ only in their line terminators
1687        compare equal. This works on Linux.
1688    
1689    17. Under certain error circumstances pcregrep might try to free random memory
1690        as it exited. This is now fixed, thanks to valgrind.
1691    
1692    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1693        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1694        was because its rules for how to advance for /g after matching an empty
1695        string at the end of a line did not allow for this case. They now check for
1696        it specially.
1697    
1698    20. pcretest is supposed to handle patterns and data of any length, by
1699        extending its buffers when necessary. It was getting this wrong when the
1700        buffer for a data line had to be extended.
1701    
1702    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1703        CRLF as a newline sequence.
1704    
1705    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1706        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1707        I have nevertheless tidied it up.
1708    
1709    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1710    
1711    24. Added a man page for pcre-config.
1712    
1713    
1714    Version 7.0 19-Dec-06
1715    ---------------------
1716    
1717     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1718        moving to gcc 4.1.1.
1719    
1720     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1721        sys/time.h, which is documented as needed for this function. It doesn't
1722        seem to matter on Linux, but it showed up on some releases of OS X.
1723    
1724     3. It seems that there are systems where bytes whose values are greater than
1725        127 match isprint() in the "C" locale. The "C" locale should be the
1726        default when a C program starts up. In most systems, only ASCII printing
1727        characters match isprint(). This difference caused the output from pcretest
1728        to vary, making some of the tests fail. I have changed pcretest so that:
1729    
1730        (a) When it is outputting text in the compiled version of a pattern, bytes
1731            other than 32-126 are always shown as hex escapes.
1732    
1733        (b) When it is outputting text that is a matched part of a subject string,
1734            it does the same, unless a different locale has been set for the match
1735            (using the /L modifier). In this case, it uses isprint() to decide.
1736    
1737     4. Fixed a major bug that caused incorrect computation of the amount of memory
1738        required for a compiled pattern when options that changed within the
1739        pattern affected the logic of the preliminary scan that determines the
1740        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1741        that the computed length was too small. The symptoms of this bug were
1742        either the PCRE error "internal error: code overflow" from pcre_compile(),
1743        or a glibc crash with a message such as "pcretest: free(): invalid next
1744        size (fast)". Examples of patterns that provoked this bug (shown in
1745        pcretest format) are:
1746    
1747          /(?-x: )/x
1748          /(?x)(?-x: \s*#\s*)/
1749          /((?i)[\x{c0}])/8
1750          /(?i:[\x{c0}])/8
1751    
1752        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1753        is now done differently.
1754    
1755     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1756        wrapper classes; (b) implement a new function in the C++ scanner that is
1757        more efficient than the old way of doing things because it avoids levels of
1758        recursion in the regex matching; (c) add a paragraph to the documentation
1759        for the FullMatch() function.
1760    
1761     6. The escape sequence \n was being treated as whatever was defined as
1762        "newline". Not only was this contrary to the documentation, which states
1763        that \n is character 10 (hex 0A), but it also went horribly wrong when
1764        "newline" was defined as CRLF. This has been fixed.
1765    
1766     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1767        was being set to -1 for the "end of line" case (supposedly a value that no
1768        character can have). Though this value is never used (the check for end of
1769        line is "zero bytes in current character"), it caused compiler complaints.
1770        I've changed it to 0xffffffff.
1771    
1772     8. In pcre_version.c, the version string was being built by a sequence of
1773        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1774        string (as it is for production releases) called a macro with an empty
1775        argument. The C standard says the result of this is undefined. The gcc
1776        compiler treats it as an empty string (which was what was wanted) but it is
1777        reported that Visual C gives an error. The source has been hacked around to
1778        avoid this problem.
1779    
1780     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1781        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1782        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1783        of them did).
1784    
1785    10. Originally, pcretest opened its input and output without "b"; then I was
1786        told that "b" was needed in some environments, so it was added for release
1787        5.0 to both the input and output. (It makes no difference on Unix-like
1788        systems.) Later I was told that it is wrong for the input on Windows. I've
1789        now abstracted the modes into two macros, to make it easier to fiddle with
1790        them, and removed "b" from the input mode under Windows.
1791    
1792    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1793    
1794    12. Added -help and --help to pcretest as an official way of being reminded
1795        of the options.
1796    
1797    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1798        and pcrecpp.cc because they annoy compilers at high warning levels.
1799    
1800    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1801    
1802    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1803        scripts are not C programs :-) and which was not noticed because it works
1804        on Linux.
1805    
1806    16. pcretest is supposed to handle any length of pattern and data line (as one
1807        line or as a continued sequence of lines) by extending its input buffer if
1808        necessary. This feature was broken for very long pattern lines, leading to
1809        a string of junk being passed to pcre_compile() if the pattern was longer
1810        than about 50K.
1811    
1812    17. I have done a major re-factoring of the way pcre_compile() computes the
1813        amount of memory needed for a compiled pattern. Previously, there was code
1814        that made a preliminary scan of the pattern in order to do this. That was
1815        OK when PCRE was new, but as the facilities have expanded, it has become
1816        harder and harder to keep it in step with the real compile phase, and there
1817        have been a number of bugs (see for example, 4 above). I have now found a
1818        cunning way of running the real compile function in a "fake" mode that
1819        enables it to compute how much memory it would need, while actually only
1820        ever using a few hundred bytes of working memory and without too many
1821        tests of the mode. This should make future maintenance and development
1822        easier. A side effect of this work is that the limit of 200 on the nesting
1823        depth of parentheses has been removed (though this was never a serious
1824        limitation, I suspect). However, there is a downside: pcre_compile() now
1825        runs more slowly than before (30% or more, depending on the pattern). I
1826        hope this isn't a big issue. There is no effect on runtime performance.
1827    
1828    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1829        newline (only possible for the last line of a file) and it was a
1830        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1831    
1832    19. Added additional timing features to pcretest. (1) The -tm option now times
1833        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1834        separate command line item, by a number that specifies the number of
1835        repeats to use when timing. The default is 50000; this gives better
1836        precision, but takes uncomfortably long for very large patterns.
1837    
1838    20. Extended pcre_study() to be more clever in cases where a branch of a
1839        subpattern has no definite first character. For example, (a*|b*)[cd] would
1840        previously give no result from pcre_study(). Now it recognizes that the
1841        first character must be a, b, c, or d.
1842    
1843    21. There was an incorrect error "recursive call could loop indefinitely" if
1844        a subpattern (or the entire pattern) that was being tested for matching an
1845        empty string contained only one non-empty item after a nested subpattern.
1846        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1847        incorrectly, because the \d was being skipped in the check.
1848    
1849    22. The pcretest program now has a new pattern option /B and a command line
1850        option -b, which is equivalent to adding /B to every pattern. This causes
1851        it to show the compiled bytecode, without the additional information that
1852        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1853        is the same as /B/I).
1854    
1855    23. A new optimization is now able automatically to treat some sequences such
1856        as a*b as a*+b. More specifically, if something simple (such as a character
1857        or a simple class like \d) has an unlimited quantifier, and is followed by
1858        something that cannot possibly match the quantified thing, the quantifier
1859        is automatically "possessified".
1860    
1861    24. A recursive reference to a subpattern whose number was greater than 39
1862        went wrong under certain circumstances in UTF-8 mode. This bug could also
1863        have affected the operation of pcre_study().
1864    
1865    25. Realized that a little bit of performance could be had by replacing
1866        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1867    
1868    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1869    
1870    27. Possessive quantifiers such as a++ were previously implemented by turning
1871        them into atomic groups such as ($>a+). Now they have their own opcodes,
1872        which improves performance. This includes the automatically created ones
1873        from 23 above.
1874    
1875    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1876        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1877        the first matched character to be a colon. This applied both to named and
1878        numbered groups.
1879    
1880    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1881    
1882    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1883        building PCRE on Windows easier, so I have included it in the distribution.
1884    
1885    31. There is now a check in pcretest against a ridiculously large number being
1886        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1887        loop, the loop is abandoned.
1888    
1889    32. Forward references to subpatterns in conditions such as (?(2)...) where
1890        subpattern 2 is defined later cause pcre_compile() to search forwards in
1891        the pattern for the relevant set of parentheses. This search went wrong
1892        when there were unescaped parentheses in a character class, parentheses
1893        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1894    
1895    33. "Subroutine" calls and backreferences were previously restricted to
1896        referencing subpatterns earlier in the regex. This restriction has now
1897        been removed.
1898    
1899    34. Added a number of extra features that are going to be in Perl 5.10. On the
1900        whole, these are just syntactic alternatives for features that PCRE had
1901        previously implemented using the Python syntax or my own invention. The
1902        other formats are all retained for compatibility.
1903    
1904        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1905            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1906            also .NET compatible.
1907    
1908        (b) A recursion or subroutine call to a named group can now be defined as
1909            (?&name) as well as (?P>name).
1910    
1911        (c) A backreference to a named group can now be defined as \k<name> or
1912            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1913            5.10, are also .NET compatible.
1914    
1915        (d) A conditional reference to a named group can now use the syntax
1916            (?(<name>) or (?('name') as well as (?(name).
1917    
1918        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1919            groups (named and numbered) that are never evaluated inline, but can be
1920            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1921            is always false. There may be only one alternative in such a group.
1922    
1923        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1924            as the simple (?(R). The condition is true only if the most recent
1925            recursion is that of the given number or name. It does not search out
1926            through the entire recursion stack.
1927    
1928        (g) The escape \gN or \g{N} has been added, where N is a positive or
1929            negative number, specifying an absolute or relative reference.
1930    
1931    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1932        some "unreachable code" warnings.
1933    
1934    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1935        things, this adds five new scripts.
1936    
1937    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1938        There were also incompatibilities regarding the handling of \Q..\E inside
1939        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1940        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1941    
1942    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1943        matches an empty string, and forcibly breaks the loop. There were bugs in
1944        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1945        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1946        separate and independent bugs (that affected different cases) have been
1947        fixed.
1948    
1949    39. Refactored the code to abolish the use of different opcodes for small
1950        capturing bracket numbers. This is a tidy that I avoided doing when I
1951        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1952        The new approach is not only tidier, it makes it possible to reduce the
1953        memory needed to fix the previous bug (38).
1954    
1955    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1956        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1957        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1958        mode.
1959    
1960    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1961        report.
1962    
1963    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1964        copy construction and assignment in the C++ wrapper.
1965    
1966    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1967        couple of bugs that could have given wrong results in the "--newline=crlf"
1968        case.
1969    
1970    44. Added a number of casts and did some reorganization of signed/unsigned int
1971        variables following suggestions from Dair Grant. Also renamed the variable
1972        "this" as "item" because it is a C++ keyword.
1973    
1974    45. Arranged for dftables to add
1975    
1976          #include "pcre_internal.h"
1977    
1978        to pcre_chartables.c because without it, gcc 4.x may remove the array
1979        definition from the final binary if PCRE is built into a static library and
1980        dead code stripping is activated.
1981    
1982    46. For an unanchored pattern, if a match attempt fails at the start of a
1983        newline sequence, and the newline setting is CRLF or ANY, and the next two
1984        characters are CRLF, advance by two characters instead of one.
1985    
1986    
1987    Version 6.7 04-Jul-06
1988    ---------------------
1989    
1990     1. In order to handle tests when input lines are enormously long, pcretest has
1991        been re-factored so that it automatically extends its buffers when
1992        necessary. The code is crude, but this _is_ just a test program. The
1993        default size has been increased from 32K to 50K.
1994    
1995     2. The code in pcre_study() was using the value of the re argument before
1996        testing it for NULL. (Of course, in any sensible call of the function, it
1997        won't be NULL.)
1998    
1999     3. The memmove() emulation function in pcre_internal.h, which is used on
2000        systems that lack both memmove() and bcopy() - that is, hardly ever -
2001        was missing a "static" storage class specifier.
2002    
2003     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2004        containing an extended class (one that cannot be represented by a bitmap
2005        because it contains high-valued characters or Unicode property items, e.g.
2006        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2007        pattern, but PCRE should not loop if you do not (it no longer does).
2008        [Detail: two cases were found: (a) a repeated subpattern containing an
2009        extended class; (b) a recursive reference to a subpattern that followed a
2010        previous extended class. It wasn't skipping over the extended class
2011        correctly when UTF-8 mode was not set.]
2012    
2013     5. A negated single-character class was not being recognized as fixed-length
2014        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2015        compile error "lookbehind assertion is not fixed length".
2016    
2017     6. The RunPerlTest auxiliary script was showing an unexpected difference
2018        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2019        write a Perl script that can interpret lines of an input file either as
2020        byte characters or as UTF-8, which is what "perltest" was being required to
2021        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2022        can't do is switch easily at run time between having the "use utf8;" pragma
2023        or not. In the end, I fudged it by using the RunPerlTest script to insert
2024        "use utf8;" explicitly for the UTF-8 tests.
2025    
2026     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2027        the end of the subject string, contrary to the documentation and to what
2028        Perl does. This was true of both matching functions. Now it matches only at
2029        the start of the subject and immediately after *internal* newlines.
2030    
2031     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2032        a pointer to an int instead of a pointer to an unsigned long int. This
2033        caused problems on 64-bit systems.
2034    
2035     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2036        instance of the 'standard' template library not being so standard".
2037    
2038    10. There was no check on the number of named subpatterns nor the maximum
2039        length of a subpattern name. The product of these values is used to compute
2040        the size of the memory block for a compiled pattern. By supplying a very
2041        long subpattern name and a large number of named subpatterns, the size
2042        computation could be caused to overflow. This is now prevented by limiting
2043        the length of names to 32 characters, and the number of named subpatterns
2044        to 10,000.
2045    
2046    11. Subpatterns that are repeated with specific counts have to be replicated in
2047        the compiled pattern. The size of memory for this was computed from the
2048        length of the subpattern and the repeat count. The latter is limited to
2049        65535, but there was no limit on the former, meaning that integer overflow
2050        could in principle occur. The compiled length of a repeated subpattern is
2051        now limited to 30,000 bytes in order to prevent this.
2052    
2053    12. Added the optional facility to have named substrings with the same name.
2054    
2055    13. Added the ability to use a named substring as a condition, using the
2056        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2057        are numbers (not recommended). Forward references are permitted.
2058    
2059    14. Added forward references in named backreferences (if you see what I mean).
2060    
2061    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2062        pattern could run off the end of the subject. For example, the pattern
2063        "(?s)(.{1,5})"8 did this with the subject "ab".
2064    
2065    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2066        PCRE_CASELESS was set when matching characters that were quantified with ?
2067        or *.
2068    
2069    17. A character class other than a single negated character that had a minimum
2070        but no maximum quantifier - for example [ab]{6,} - was not handled
2071        correctly by pce_dfa_exec(). It would match only one character.
2072    
2073    18. A valid (though odd) pattern that looked like a POSIX character
2074        class but used an invalid character after [ (for example [[,abc,]]) caused
2075        pcre_compile() to give the error "Failed: internal error: code overflow" or
2076        in some cases to crash with a glibc free() error. This could even happen if
2077        the pattern terminated after [[ but there just happened to be a sequence of
2078        letters, a binary zero, and a closing ] in the memory that followed.
2079    
2080    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2081        over the years. Originally (before any Unicode support), just the bottom 8
2082        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2083        output from "man perlunicode" includes this:
2084    
2085          The regular expression compiler produces polymorphic opcodes.  That
2086          is, the pattern adapts to the data and automatically switches to
2087          the Unicode character scheme when presented with Unicode data--or
2088          instead uses a traditional byte scheme when presented with byte
2089          data.
2090    
2091        Sadly, a wide octal escape does not cause a switch, and in a string with
2092        no other multibyte characters, these octal escapes are treated as before.
2093        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2094        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2095        Unicode string.
2096    
2097        I have not perpetrated such confusion in PCRE. Up till now, it took just
2098        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2099        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2100        translate to the appropriate multibyte character.
2101    
2102    29. Applied some refactoring to reduce the number of warnings from Microsoft
2103        and Borland compilers. This has included removing the fudge introduced
2104        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2105        a warning about an unused variable.
2106    
2107    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2108        characters since release 4.0, because Perl (from release 5.004) does not.
2109        [Or at least, is documented not to: some releases seem to be in conflict
2110        with the documentation.] However, when a pattern was studied with
2111        pcre_study() and all its branches started with \s, PCRE still included VT
2112        as a possible starting character. Of course, this did no harm; it just
2113        caused an unnecessary match attempt.
2114    
2115    22. Removed a now-redundant internal flag bit that recorded the fact that case
2116        dependency changed within the pattern. This was once needed for "required
2117        byte" processing, but is no longer used. This recovers a now-scarce options
2118        bit. Also moved the least significant internal flag bit to the most-
2119        significant bit of the word, which was not previously used (hangover from
2120        the days when it was an int rather than a uint) to free up another bit for
2121        the future.
2122    
2123    23. Added support for CRLF line endings as well as CR and LF. As well as the
2124        default being selectable at build time, it can now be changed at runtime
2125        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2126        specify that it is scanning data with non-default line endings.
2127    
2128    24. Changed the definition of CXXLINK to make it agree with the definition of
2129        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2130    
2131    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2132        recursions. This makes a big different to stack usage for some patterns.
2133    
2134    26. If a subpattern containing a named recursion or subroutine reference such
2135        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2136        the space required for the compiled pattern went wrong and gave too small a
2137        value. Depending on the environment, this could lead to "Failed: internal
2138        error: code overflow at offset 49" or "glibc detected double free or
2139        corruption" errors.
2140    
2141    27. Applied patches from Google (a) to support the new newline modes and (b) to
2142        advance over multibyte UTF-8 characters in GlobalReplace.
2143    
2144    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2145        difference for some implementation of PCRE in some Windows version.
2146    
2147    29. Added some extra testing facilities to pcretest:
2148    
2149        \q<number>   in a data line sets the "match limit" value
2150        \Q<number>   in a data line sets the "match recursion limt" value
2151        -S <number>  sets the stack size, where <number> is in megabytes
2152    
2153        The -S option isn't available for Windows.
2154    
2155    
2156    Version 6.6 06-Feb-06
2157    ---------------------
2158    
2159     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2160        in pcreposix.h. I have copied the definition from pcre.h.
2161    
2162     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2163        because pcre.h is no longer a built file.
2164    
2165     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2166        not normally included in the compiled code.
2167    
2168    
2169    Version 6.5 01-Feb-06
2170    ---------------------
2171    
2172     1. When using the partial match feature with pcre_dfa_exec(), it was not
2173        anchoring the second and subsequent partial matches at the new starting
2174        point. This could lead to incorrect results. For example, with the pattern
2175        /1234/, partially matching against "123" and then "a4" gave a match.
2176    
2177     2. Changes to pcregrep:
2178    
2179        (a) All non-match returns from pcre_exec() were being treated as failures
2180            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2181            error message is output. Some extra information is given for the
2182            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2183            probably the only errors that are likely to be caused by users (by
2184            specifying a regex that has nested indefinite repeats, for instance).
2185            If there are more than 20 of these errors, pcregrep is abandoned.
2186    
2187        (b) A binary zero was treated as data while matching, but terminated the
2188            output line if it was written out. This has been fixed: binary zeroes
2189            are now no different to any other data bytes.
2190    
2191        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2192            used to set a locale for matching. The --locale=xxxx long option has
2193            been added (no short equivalent) to specify a locale explicitly on the
2194            pcregrep command, overriding the environment variables.
2195    
2196        (d) When -B was used with -n, some line numbers in the output were one less
2197            than they should have been.
2198    
2199        (e) Added the -o (--only-matching) option.
2200    
2201        (f) If -A or -C was used with -c (count only), some lines of context were
2202            accidentally printed for the final match.
2203    
2204        (g) Added the -H (--with-filename) option.
2205    
2206        (h) The combination of options -rh failed to suppress file names for files
2207            that were found from directory arguments.
2208    
2209        (i) Added the -D (--devices) and -d (--directories) options.
2210    
2211        (j) Added the -F (--fixed-strings) option.
2212    
2213        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2214    
2215        (l) Added the --colo(u)r option.
2216    
2217        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2218            is not present by default.
2219    
2220     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2221        items such as (?R) or (?1), when the recursion could match a number of
2222        alternatives. If it matched one of the alternatives, but subsequently,
2223        outside the recursion, there was a failure, the code tried to back up into
2224        the recursion. However, because of the way PCRE is implemented, this is not
2225        possible, and the result was an incorrect result from the match.
2226    
2227        In order to prevent this happening, the specification of recursion has
2228        been changed so that all such subpatterns are automatically treated as
2229        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2230    
2231     4. I had overlooked the fact that, in some locales, there are characters for
2232        which isalpha() is true but neither isupper() nor islower() are true. In
2233        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2234        and ordfeminine) are like this. This affected the treatment of \w and \W
2235        when they appeared in character classes, but not when they appeared outside
2236        a character class. The bit map for "word" characters is now created
2237        separately from the results of isalnum() instead of just taking it from the
2238        upper, lower, and digit maps. (Plus the underscore character, of course.)
2239    
2240     5. The above bug also affected the handling of POSIX character classes such as
2241        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2242        permanent tables. Instead, the bit maps for such a class were previously
2243        created as the appropriate unions of the upper, lower, and digit bitmaps.
2244        Now they are created by subtraction from the [[:word:]] class, which has
2245        its own bitmap.
2246    
2247     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2248        It is created by subtracting the vertical space characters (\x09, \x0a,
2249        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2250        subtraction was done in the overall bitmap for a character class, meaning
2251        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2252        be recognized. This bug has been fixed.
2253    
2254     7. Patches from the folks at Google:
2255    
2256          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2257          real life, but is still worth protecting against".
2258    
2259          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2260          regular expressions".
2261    
2262          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2263          have it.
2264    
2265          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2266          "configure" and the latter not, in order to fix a problem somebody had
2267          with compiling the Arg class on HP-UX.
2268    
2269          (e) Improve the error-handling of the C++ wrapper a little bit.
2270    
2271          (f) New tests for checking recursion limiting.
2272    
2273     8. The pcre_memmove() function, which is used only if the environment does not
2274        have a standard memmove() function (and is therefore rarely compiled),
2275        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2276        returning a result (though PCRE never actually uses the result).
2277    
2278     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2279        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2280        returned instead of calling malloc() with an overflowing number that would
2281        most likely cause subsequent chaos.
2282    
2283    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2284    
2285    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2286        with this option is matched, the nmatch and pmatch options of regexec() are
2287        ignored.
2288    
2289    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2290        provided in case anyone wants to the the POSIX interface with UTF-8
2291        strings.
2292    
2293    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2294        C++ linking (needed for some HP-UX environments).
2295    
2296    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2297        (unused parameter) and in the pcre_printint() function (omitted "default"
2298        switch label when the default is to do nothing).
2299    
2300    15. Added some code to make it possible, when PCRE is compiled as a C++
2301        library, to replace subject pointers for pcre_exec() with a smart pointer
2302        class, thus making it possible to process discontinuous strings.
2303    
2304    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2305        much the same function. They were added by different people who were trying
2306        to make PCRE easy to compile on non-Unix systems. It has been suggested
2307        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2308        for compiling on Windows systems. I have therefore replaced it with
2309        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2310        defaults to "extern" for C or "extern C" for C++, which works fine on
2311        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2312        SCOPE with something explicit in config.h. In addition:
2313    
2314        (a) pcreposix.h still had just "extern" instead of either of these macros;
2315            I have replaced it with PCRE_DATA_SCOPE.
2316    
2317        (b) Functions such as _pcre_xclass(), which are internal to the library,
2318            but external in the C sense, all had PCRE_EXPORT in their definitions.
2319            This is apparently wrong for the Windows case, so I have removed it.
2320            (It makes no difference on Unix-like systems.)
2321    
2322    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2323        of recursive calls to match(). This is different to MATCH_LIMIT because
2324        that limits the total number of calls to match(), not all of which increase
2325        the depth of recursion. Limiting the recursion depth limits the amount of
2326        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2327        when PCRE is compiled, and changed at run time. A patch from Google adds
2328        this functionality to the C++ interface.
2329    
2330    18. Changes to the handling of Unicode character properties:
2331    
2332        (a) Updated the table to Unicode 4.1.0.
2333    
2334        (b) Recognize characters that are not in the table as "Cn" (undefined).
2335    
2336        (c) I revised the way the table is implemented to a much improved format
2337            which includes recognition of ranges. It now supports the ranges that
2338            are defined in UnicodeData.txt, and it also amalgamates other
2339            characters into ranges. This has reduced the number of entries in the
2340            table from around 16,000 to around 3,000, thus reducing its size
2341            considerably. I realized I did not need to use a tree structure after
2342            all - a binary chop search is just as efficient. Having reduced the
2343            number of entries, I extended their size from 6 bytes to 8 bytes to
2344            allow for more data.
2345    
2346        (d) Added support for Unicode script names via properties such as \p{Han}.
2347    
2348    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2349        matching that character.
2350    
2351    20. When matching a repeated Unicode property with a minimum greater than zero,
2352        (for example \pL{2,}), PCRE could look past the end of the subject if it
2353        reached it while seeking the minimum number of characters. This could
2354        happen only if some of the characters were more than one byte long, because
2355        there is a check for at least the minimum number of bytes.
2356    
2357    21. Refactored the implementation of \p and \P so as to be more general, to
2358        allow for more different types of property in future. This has changed the
2359        compiled form incompatibly. Anybody with saved compiled patterns that use
2360        \p or \P will have to recompile them.
2361    
2362    22. Added "Any" and "L&" to the supported property types.
2363    
2364    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2365        but give a compile time error if the value is greater than 0xff.
2366    
2367    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2368        accidentally not being installed or uninstalled.
2369    
2370    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2371        made were to insert the current release number. This seemed silly, because
2372        it made things harder for people building PCRE on systems that don't run
2373        "configure". I have turned pcre.h into a distributed file, no longer built
2374        by "configure", with the version identification directly included. There is
2375        no longer a pcre.h.in file.
2376    
2377        However, this change necessitated a change to the pcre-config script as
2378        well. It is built from pcre-config.in, and one of the substitutions was the
2379        release number. I have updated configure.ac so that ./configure now finds
2380        the release number by grepping pcre.h.
2381    
2382    26. Added the ability to run the tests under valgrind.
2383    
2384    
2385    Version 6.4 05-Sep-05
2386    ---------------------
2387    
2388     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2389        "--" to be printed when multiple files were scanned, even when none of the
2390        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2391        consider it to be a bug, and have restored the previous behaviour.
2392    
2393     2. A couple of code tidies to get rid of compiler warnings.
2394    
2395     3. The pcretest program used to cheat by referring to symbols in the library
2396        whose names begin with _pcre_. These are internal symbols that are not
2397        really supposed to be visible externally, and in some environments it is
2398        possible to suppress them. The cheating is now confined to including
2399        certain files from the library's source, which is a bit cleaner.
2400    
2401     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2402        file's purpose clearer.
2403    
2404     5. Reorganized pcre_ucp_findchar().
2405    
2406    
2407    Version 6.3 15-Aug-05
2408    ---------------------
2409    
2410     1. The file libpcre.pc.in did not have general read permission in the tarball.
2411    
2412     2. There were some problems when building without C++ support:
2413    
2414        (a) If C++ support was not built, "make install" and "make test" still
2415            tried to test it.
2416    
2417        (b) There were problems when the value of CXX was explicitly set. Some
2418            changes have been made to try to fix these, and ...
2419    
2420        (c) --disable-cpp can now be used to explicitly disable C++ support.
2421    
2422        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2423            backslash in a target when C++ was disabled. This confuses some
2424            versions of "make", apparently. Using an intermediate variable solves
2425            this. (Same for CPP_LOBJ.)
2426    
2427     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2428        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2429        necessary on certain architectures.
2430    
2431     4. Added a setting of -export-symbols-regex to the link command to remove
2432        those symbols that are exported in the C sense, but actually are local
2433        within the library, and not documented. Their names all begin with
2434        "_pcre_". This is not a perfect job, because (a) we have to except some
2435        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2436        available (and never for static libraries). I have made a note to try to
2437        find a way round (a) in the future.
2438    
2439    
2440    Version 6.2 01-Aug-05
2441    ---------------------
2442    
2443     1. There was no test for integer overflow of quantifier values. A construction
2444        such as {1111111111111111} would give undefined results. What is worse, if
2445        a minimum quantifier for a parenthesized subpattern overflowed and became
2446        negative, the calculation of the memory size went wrong. This could have
2447        led to memory overwriting.
2448    
2449     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2450    
2451     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2452        operating environments where this matters.
2453    
2454     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2455        PCRE options from within the C++ wrapper.
2456    
2457     5. Named capturing subpatterns were not being correctly counted when a pattern
2458        was compiled. This caused two problems: (a) If there were more than 100
2459        such subpatterns, the calculation of the memory needed for the whole
2460        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2461        back references of the form \12, where the number was greater than 9, were
2462        not recognized as back references, even though there were sufficient
2463        previous subpatterns.
2464    
2465     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2466        versions of gcc, e.g. 2.95.4.
2467    
2468    
2469    Version 6.1 21-Jun-05
2470    ---------------------
2471    
2472     1. There was one reference to the variable "posix" in pcretest.c that was not
2473        surrounded by "#if !defined NOPOSIX".
2474    
2475     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2476        the cross-check on the old pcre_info() function, for the benefit of the
2477        cut-down version of PCRE that is currently imported into Exim.
2478    
2479     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2480        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2481        patterns that start (?i) but I don't think that matters. The use of (?i) is
2482        just an example; this all applies to the other options as well.
2483    
2484     4. Since libtool seems to echo the compile commands it is issuing, the output
2485        from "make" can be reduced a bit by putting "@" in front of each libtool
2486        compile command.
2487    
2488     5. Patch from the folks at Google for configure.in to be a bit more thorough
2489        in checking for a suitable C++ installation before trying to compile the
2490        C++ stuff. This should fix a reported problem when a compiler was present,
2491        but no suitable headers.
2492    
2493     6. The man pages all had just "PCRE" as their title. I have changed them to
2494        be the relevant file name. I have also arranged that these names are
2495        retained in the file doc/pcre.txt, which is a concatenation in text format
2496        of all the man pages except the little individual ones for each function.
2497    
2498     7. The NON-UNIX-USE file had not been updated for the different set of source
2499        files that come with release 6. I also added a few comments about the C++
2500        wrapper.
2501    
2502    
2503    Version 6.0 07-Jun-05
2504    ---------------------
2505    
2506     1. Some minor internal re-organization to help with my DFA experiments.
2507    
2508     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2509        didn't matter for the library itself when fully configured, but did matter
2510        when compiling without UCP support, or within Exim, where the ucp files are
2511        not imported.
2512    
2513     3. Refactoring of the library code to split up the various functions into
2514        different source modules. The addition of the new DFA matching code (see
2515        below) to a single monolithic source would have made it really too
2516        unwieldy, quite apart from causing all the code to be include in a
2517        statically linked application, when only some functions are used. This is
2518        relevant even without the DFA addition now that patterns can be compiled in
2519        one application and matched in another.
2520    
2521        The downside of splitting up is that there have to be some external
2522        functions and data tables that are used internally in different modules of
2523        the library but which are not part of the API. These have all had their
2524        names changed to start with "_pcre_" so that they are unlikely to clash
2525        with other external names.
2526    
2527     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2528        a different (DFA) algorithm. Although it is slower than the original
2529        function, it does have some advantages for certain types of matching
2530        problem.
2531    
2532     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2533        including restarting after a partial match.
2534    
2535     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2536        defined when compiling for Windows was sent to me. I have put it into the
2537        code, though I have no means of testing or verifying it.
2538    
2539     7. Added the pcre_refcount() auxiliary function.
2540    
2541     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2542        match before or at the first newline in the subject string. In pcretest,
2543        the /f option on a pattern can be used to set this.
2544    
2545     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2546        would behave wrongly. This has been present in PCRE since release 4.0.
2547    
2548    10. A number of changes to the pcregrep command:
2549    
2550        (a) Refactored how -x works; insert ^(...)$ instead of setting
2551            PCRE_ANCHORED and checking the length, in preparation for adding
2552            something similar for -w.
2553    
2554        (b) Added the -w (match as a word) option.
2555    
2556        (c) Refactored the way lines are read and buffered so as to have more
2557            than one at a time available.
2558    
2559        (d) Implemented a pcregrep test script.
2560    
2561        (e) Added the -M (multiline match) option. This allows patterns to match
2562            over several lines of the subject. The buffering ensures that at least
2563            8K, or the rest of the document (whichever is the shorter) is available
2564            for matching (and similarly the previous 8K for lookbehind assertions).
2565    
2566        (f) Changed the --help output so that it now says
2567    
2568              -w, --word-regex(p)
2569    
2570            instead of two lines, one with "regex" and the other with "regexp"
2571            because that confused at least one person since the short forms are the
2572            same. (This required a bit of code, as the output is generated
2573            automatically from a table. It wasn't just a text change.)
2574    
2575        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2576            option but starts with a hyphen. Could be a pattern or a path name
2577            starting with a hyphen, for instance.
2578    
2579        (h) "-" can be given as a file name to represent stdin.
2580    
2581        (i) When file names are being printed, "(standard input)" is used for
2582            the standard input, for compatibility with GNU grep. Previously
2583            "<stdin>" was used.
2584    
2585        (j) The option --label=xxx can be used to supply a name to be used for
2586            stdin when file names are being printed. There is no short form.
2587    
2588        (k) Re-factored the options decoding logic because we are going to add
2589            two more options that take data. Such options can now be given in four
2590            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2591    
2592        (l) Added the -A, -B, and -C options for requesting that lines of context
2593            around matches be printed.
2594    
2595        (m) Added the -L option to print the names of files that do not contain
2596            any matching lines, that is, the complement of -l.
2597    
2598        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2599            continue to scan other files.
2600    
2601        (o) The -s option was incorrectly implemented. For compatibility with other
2602            greps, it now suppresses the error message for a non-existent or non-
2603            accessible file (but not the return code). There is a new option called
2604            -q that suppresses the output of matching lines, which was what -s was
2605            previously doing.
2606    
2607        (p) Added --include and --exclude options to specify files for inclusion
2608            and exclusion when recursing.
2609    
2610    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2611        Hopefully, it now does.
2612    
2613    12. Missing cast in pcre_study().
2614    
2615    13. Added an "uninstall" target to the makefile.
2616    
2617    14. Replaced "extern" in the function prototypes in Makefile.in with
2618        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2619        world, but is set differently for Windows.
2620    
2621    15. Added a second compiling function called pcre_compile2(). The only
2622        difference is that it has an extra argument, which is a pointer to an
2623        integer error code. When there is a compile-time failure, this is set
2624        non-zero, in addition to the error test pointer being set to point to an
2625        error message. The new argument may be NULL if no error number is required
2626        (but then you may as well call pcre_compile(), which is now just a
2627        wrapper). This facility is provided because some applications need a
2628        numeric error indication, but it has also enabled me to tidy up the way
2629        compile-time errors are handled in the POSIX wrapper.
2630    
2631    16. Added VPATH=.libs to the makefile; this should help when building with one
2632        prefix path and installing with another. (Or so I'm told by someone who
2633        knows more about this stuff than I do.)
2634    
2635    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2636        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2637        match everything, including newlines. This is not POSIX-compatible, but
2638        somebody wanted the feature. From pcretest it can be activated by using
2639        both the P and the s flags.
2640    
2641    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2642    
2643    19. libpcre.pc was being incorrectly installed as executable.
2644    
2645    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2646        it now also looks for '\r' so that it will work unmodified on Windows.
2647    
2648    21. Added Google's contributed C++ wrapper to the distribution.
2649    
2650    22. Added some untidy missing memory free() calls in pcretest, to keep
2651        Electric Fence happy when testing.
2652    
2653    
2654    
2655    Version 5.0 13-Sep-04
2656    ---------------------
2657    
2658     1. Internal change: literal characters are no longer packed up into items
2659        containing multiple characters in a single byte-string. Each character
2660        is now matched using a separate opcode. However, there may be more than one
2661        byte in the character in UTF-8 mode.
2662    
2663     2. The pcre_callout_block structure has two new fields: pattern_position and
2664        next_item_length. These contain the offset in the pattern to the next match
2665        item, and its length, respectively.
2666    
2667     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2668        insertion of callouts before each pattern item. Added the /C option to
2669        pcretest to make use of this.
2670    
2671     4. On the advice of a Windows user, the lines
2672    
2673          #if defined(_WIN32) || defined(WIN32)
2674          _setmode( _fileno( stdout ), 0x8000 );
2675          #endif  /* defined(_WIN32) || defined(WIN32) */
2676    
2677        have been added to the source of pcretest. This apparently does useful
2678        magic in relation to line terminators.
2679    
2680     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2681        for the benefit of those environments where the "b" makes a difference.
2682    
2683     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2684        to know about it. I have put a hack into configure.in that adds in code
2685        to set GCC=yes if CC=icc. This seems to end up at a point in the
2686        generated configure script that is early enough to affect the setting of
2687        compiler options, which is what is needed, but I have no means of testing
2688        whether it really works. (The user who reported this had patched the
2689        generated configure script, which of course I cannot do.)
2690    
2691        LATER: After change 22 below (new libtool files), the configure script
2692        seems to know about icc (and also ecc). Therefore, I have commented out
2693        this hack in configure.in.
2694    
2695     7. Added support for pkg-config (2 patches were sent in).
2696    
2697     8. Negated POSIX character classes that used a combination of internal tables
2698        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2699        [[:^ascii]]. Typically, they would match almost any characters. The other
2700        POSIX classes were not broken in this way.
2701    
2702     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2703        to find the match, as PCRE was deluded into thinking that the match had to
2704        start at the start point or following a newline. The same bug applied to
2705        patterns with negative forward assertions or any backward assertions
2706        preceding ".*" at the start, unless the pattern required a fixed first
2707        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2708    
2709    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2710        starting at the last subject character, bytes beyond the end of the subject
2711        string were read.
2712    
2713    11. Renamed the variable "class" as "classbits" to make life easier for C++
2714        users. (Previously there was a macro definition, but it apparently wasn't
2715        enough.)
2716    
2717    12. Added the new field "tables" to the extra data so that tables can be passed
2718        in at exec time, or the internal tables can be re-selected. This allows
2719        a compiled regex to be saved and re-used at a later time by a different
2720        program that might have everything at different addresses.
2721    
2722    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2723        -R library as well as a -L library.
2724    
2725    14. The debugging options of pcretest (-d on the command line or D on a
2726        pattern) showed incorrect output for anything following an extended class
2727        that contained multibyte characters and which was followed by a quantifier.
2728    
2729    15. Added optional support for general category Unicode character properties
2730        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2731        support. It adds about 90K to the size of the library. The meanings of the
2732        inbuilt class escapes such as \d and \s have NOT been changed.
2733    
2734    16. Updated pcredemo.c to include calls to free() to release the memory for the
2735        compiled pattern.
2736    
2737    17. The generated file chartables.c was being created in the source directory
2738        instead of in the building directory. This caused the build to fail if the
2739        source directory was different from the building directory, and was
2740        read-only.
2741    
2742    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2743        file. No doubt somebody will tell me if they don't make sense... Also added
2744        Dan Mooney's comments about building on OpenVMS.
2745    
2746    19. Added support for partial matching via the PCRE_PARTIAL option for
2747        pcre_exec() and the \P data escape in pcretest.
2748    
2749    20. Extended pcretest with 3 new pattern features:
2750    
2751        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2752              write the compiled pattern to the file whose name is "rest-of-line".
2753              This is a straight binary dump of the data, with the saved pointer to
2754              the character tables forced to be NULL. The study data, if any, is
2755              written too. After writing, pcretest reads a new pattern.
2756    
2757        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2758              compiled pattern from the given file. There must not be any
2759              occurrences of "<" in the file name (pretty unlikely); if there are,
2760              pcretest will instead treat the initial "<" as a pattern delimiter.
2761              After reading in the pattern, pcretest goes on to read data lines as
2762              usual.
2763    
2764        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2765              and 16-bit fields in a compiled pattern, to simulate a pattern that
2766              was compiled on a host of opposite endianness.
2767    
2768    21. The pcre-exec() function can now cope with patterns that were compiled on
2769        hosts of opposite endianness, with this restriction:
2770    
2771          As for any compiled expression that is saved and used later, the tables
2772          pointer field cannot be preserved; the extra_data field in the arguments
2773          to pcre_exec() should be used to pass in a tables address if a value
2774          other than the default internal tables were used at compile time.
2775    
2776    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2777        now diagnosed as an error. Previously, most of the time, a negative number
2778        would have been treated as zero, but if in addition "ovector" was passed as
2779        NULL, a crash could occur.
2780    
2781    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2782        new versions from the libtool 1.5 distribution (the last one is a copy of
2783        a file called libtool.m4). This seems to have fixed the need to patch
2784        "configure" to support Darwin 1.3 (which I used to do). However, I still
2785        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2786        workstation).
2787    
2788    24. Changed the PCRE licence to be the more standard "BSD" licence.
2789    
2790    
2791    Version 4.5 01-Dec-03
2792    ---------------------
2793    
2794     1. There has been some re-arrangement of the code for the match() function so
2795        that it can be compiled in a version that does not call itself recursively.
2796        Instead, it keeps those local variables that need separate instances for
2797        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2798        needs to "recurse". Keeping track of where control must go is done by means
2799        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2800        hide most of the details from the main code, and operates only if
2801        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2802        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2803        operating.
2804    
2805        To make it easier for callers to provide specially tailored get/free
2806        functions for this usage, two new functions, pcre_stack_malloc, and
2807        pcre_stack_free, are used. They are always called in strict stacking order,
2808        and the size of block requested is always the same.
2809    
2810        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2811        PCRE has been compiled to use the stack or the heap for recursion. The
2812        -C option of pcretest uses this to show which version is compiled.
2813    
2814        A new data escape \S, is added to pcretest; it causes the amounts of store
2815        obtained and freed by both kinds of malloc/free at match time to be added
2816        to the output.
2817    
2818     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2819        what's available on my current Linux desktop machine.
2820    
2821     3. When matching a UTF-8 string, the test for a valid string at the start has
2822        been extended. If start_offset is not zero, PCRE now checks that it points
2823        to a byte that is the start of a UTF-8 character. If not, it returns
2824        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2825        this is necessary because there may be backward assertions in the pattern.
2826        When matching the same subject several times, it may save resources to use
2827        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2828    
2829     4. The code for checking the validity of UTF-8 strings has been tightened so
2830        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2831        containing "overlong sequences".
2832    
2833     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2834        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2835        should have been "&", but it just so happened that all the cases this let
2836        through by mistake were picked up later in the function.
2837    
2838     6. I had used a variable called "isblank" - this is a C99 function, causing
2839        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2840    
2841     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2842        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2843        so the version is shown; (c) stop "make test" echoing "./RunTest".
2844    
2845     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2846    
2847     9. The prototype for memmove() for systems that don't have it was using
2848        size_t, but the inclusion of the header that defines size_t was later. I've
2849        moved the #includes for the C headers earlier to avoid this.
2850    
2851    10. Added some adjustments to the code to make it easier to compiler on certain
2852        special systems:
2853    
2854          (a) Some "const" qualifiers were missing.
2855          (b) Added the macro EXPORT before all exported functions; by default this
2856              is defined to be empty.
2857          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2858              that it reads its output file name as an argument instead of writing
2859              to the standard output and assuming this can be redirected.
2860    
2861    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2862        class containing characters with values greater than 255, PCRE compilation
2863        went into a loop.
2864    
2865    12. A recursive reference to a subpattern that was within another subpattern
2866        that had a minimum quantifier of zero caused PCRE to crash. For example,
2867        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2868        recursion. If the recursively-called subpattern itself had a zero repeat,
2869        that was OK.
2870    
2871    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2872        buffer into which it was copied (for escape processing) was still set at
2873        1024, so long lines caused crashes.
2874    
2875    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2876        "internal error: code overflow...". This applied to any character class
2877        that was followed by a possessive quantifier.
2878    
2879    15. Modified the Makefile to add libpcre.la as a prerequisite for
2880        libpcreposix.la because I was told this is needed for a parallel build to
2881        work.
2882    
2883    16. If a pattern that contained .* following optional items at the start was
2884        studied, the wrong optimizing data was generated, leading to matching
2885        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2886        matching string must start with a or b or c. The correct conclusion for
2887        this pattern is that a match can start with any character.
2888    
2889    
2890    Version 4.4 13-Aug-03
2891    ---------------------
2892    
2893     1. In UTF-8 mode, a character class containing characters with values between
2894        127 and 255 was not handled correctly if the compiled pattern was studied.
2895        In fixing this, I have also improved the studying algorithm for such
2896        classes (slightly).
2897    
2898     2. Three internal functions had redundant arguments passed to them. Removal
2899        might give a very teeny performance improvement.
2900    
2901     3. Documentation bug: the value of the capture_top field in a callout is *one
2902        more than* the number of the hightest numbered captured substring.
2903    
2904     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2905        in incorrectly linking with a previously installed version. They now link
2906        explicitly with libpcre.la.
2907    
2908     5. configure.in no longer needs to recognize Cygwin specially.
2909    
2910     6. A problem in pcre.in for Windows platforms is fixed.
2911    
2912     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2913        pcretest, it used to include the size of the study block as part of its
2914        output. Unfortunately, the structure contains a field that has a different
2915        size on different hardware architectures. This meant that the tests that
2916        showed this size failed. As the block is currently always of a fixed size,
2917        this information isn't actually particularly useful in pcretest output, so
2918        I have just removed it.
2919    
2920     8. Three pre-processor statements accidentally did not start in column 1.
2921        Sadly, there are *still* compilers around that complain, even though
2922        standard C has not required this for well over a decade. Sigh.
2923    
2924     9. In pcretest, the code for checking callouts passed small integers in the
2925        callout_data field, which is a void * field. However, some picky compilers
2926        complained about the casts involved for this on 64-bit systems. Now
2927        pcretest passes the address of the small integer instead, which should get
2928        rid of the warnings.
2929    
2930    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2931        both compile and run time, and gives an error if an invalid UTF-8 sequence
2932        is found. There is a option for disabling this check in cases where the
2933        string is known to be correct and/or the maximum performance is wanted.
2934    
2935    11. In response to a bug report, I changed one line in Makefile.in from
2936    
2937            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2938        to
2939            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2940    
2941        to look similar to other lines, but I have no way of telling whether this
2942        is the right thing to do, as I do not use Windows. No doubt I'll get told
2943        if it's wrong...
2944    
2945    
2946    Version 4.3 21-May-03
2947    ---------------------
2948    
2949    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2950       Makefile.
2951    
2952    2. Some refactoring to improve the quality of the code:
2953    
2954       (i)   The utf8_table... variables are now declared "const".
2955    
2956       (ii)  The code for \cx, which used the "case flipping" table to upper case
2957             lower case letters, now just substracts 32. This is ASCII-specific,
2958             but the whole concept of \cx is ASCII-specific, so it seems
2959             reasonable.
2960    
2961       (iii) PCRE was using its character types table to recognize decimal and
2962             hexadecimal digits in the pattern. This is silly, because it handles
2963             only 0-9, a-f, and A-F, but the character types table is locale-
2964             specific, which means strange things might happen. A private
2965             table is now used for this - though it costs 256 bytes, a table is
2966             much faster than multiple explicit tests. Of course, the standard
2967             character types table is still used for matching digits in subject
2968             strings against \d.
2969    
2970       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2971             ending in _t are). So I've renamed it as ESC_tee.
2972    
2973    3. The first argument for regexec() in the POSIX wrapper should have been
2974       defined as "const".
2975    
2976    4. Changed pcretest to use malloc() for its buffers so that they can be
2977       Electric Fenced for debugging.
2978    
2979    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2980       to read one or more bytes before the start of the subject string. Often this
2981       had no effect on PCRE's behaviour, but in some circumstances it could
2982       provoke a segmentation fault.
2983    
2984    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2985       to try to read one or more bytes before the start of the subject string.
2986    
2987    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2988       UTF-8 support could misbehave in various ways if the subject string
2989       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2990       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2991       back over UTF-8 characters.)
2992    
2993    
2994    Version 4.2 14-Apr-03
2995    ---------------------
2996    
2997    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2998    
2999    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3000         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3001         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3002         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3003         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3004           and BUILD_EXEEXT
3005         Note: automatic setting of the BUILD variables is not yet working
3006         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3007           compile-time but not at link-time
3008         [LINK]: use for linking executables only
3009         make different versions for Windows and non-Windows
3010         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3011           libraries
3012         [LINK_FOR_BUILD]: new variable
3013         [OBJEXT]: use throughout
3014         [EXEEXT]: use throughout
3015         <winshared>: new target
3016         <wininstall>: new target
3017         <dftables.o>: use native compiler
3018         <dftables>: use native linker
3019         <install>: handle Windows platform correctly
3020         <clean>: ditto
3021         <check>: ditto
3022         copy DLL to top builddir before testing
3023    
3024       As part of these changes, -no-undefined was removed again. This was reported
3025       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3026       in any case.
3027    
3028    3. Some tidies to get rid of compiler warnings:
3029    
3030       . In the match_data structure, match_limit was an unsigned long int, whereas
3031         match_call_count was an int. I've made them both unsigned long ints.
3032    
3033       . In pcretest the fact that a const uschar * doesn't automatically cast to
3034         a void * provoked a warning.
3035    
3036       . Turning on some more compiler warnings threw up some "shadow" variables
3037         and a few more missing casts.
3038    
3039    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3040       option, a class that contained a single character with a value between 128
3041       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3042    
3043    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3044       option, a class that contained several characters, but with at least one
3045       whose value was between 128 and 255 caused PCRE to crash.
3046    
3047    
3048    Version 4.1 12-Mar-03
3049    ---------------------
3050    
3051    1. Compiling with gcc -pedantic found a couple of places where casts were
3052    needed, and a string in dftables.c that was longer than standard compilers are
3053    required to support.
3054    
3055    2. Compiling with Sun's compiler found a few more places where the code could
3056    be tidied up in order to avoid warnings.
3057    
3058    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3059    first of these names is deprecated in the latest Autoconf in favour of the name
3060    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3061    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3062    analogy I have changed it to CFLAGS_FOR_BUILD.
3063    
3064    4. Added -no-undefined to the linking command in the Makefile, because this is
3065    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3066    linking step for the pcreposix library.
3067    
3068    5. PCRE was failing to diagnose the case of two named groups with the same
3069    name.
3070    
3071    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3072    literal character that is needed in the subject for a match, and scans along to
3073    ensure that it is present before embarking on the full matching process. This
3074    saves time in cases of nested unlimited repeats that are never going to match.
3075    Problem: the scan can take a lot of time if the subject is very long (e.g.
3076    megabytes), thus penalizing straightforward matches. It is now done only if the
3077    amount of subject to be scanned is less than 1000 bytes.
3078    
3079    7. A lesser problem with the same optimization is that it was recording the
3080    first character of an anchored pattern as "needed", thus provoking a search
3081    right along the subject, even when the first match of the pattern was going to
3082    fail. The "needed" character is now not set for anchored patterns, unless it
3083    follows something in the pattern that is of non-fixed length. Thus, it still
3084    fulfils its original purpose of finding quick non-matches in cases of nested
3085    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3086    
3087    
3088    Version 4.0 17-Feb-03
3089    ---------------------
3090    
3091    1. If a comment in an extended regex that started immediately after a meta-item
3092    extended to the end of string, PCRE compiled incorrect data. This could lead to
3093    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3094    
3095    2. Moved to autoconf 2.53 and libtool 1.4.2.
3096    
3097    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3098    the special perltest8 script is no longer needed - all the tests can be run
3099    from a single perltest script.
3100    
3101    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3102    by \s. It has now been removed in PCRE. This means it isn't recognized as
3103    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3104    class [:space:] *does* include VT, thereby creating a mess.
3105    
3106    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3107    space and tab.
3108    
3109    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3110    its new features into the main test script, reducing the number of scripts.
3111    
3112    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3113    were backward compatible, and made the (?i) apply to the whole pattern, as if
3114    /i were given. Now it behaves more logically, and applies the option setting
3115    only to what follows. PCRE has been changed to follow suit. However, if it
3116    finds options settings right at the start of the pattern, it extracts them into
3117    the global options, as before. Thus, they show up in the info data.
3118    
3119    8. Added support for the \Q...\E escape sequence. Characters in between are
3120    treated as literals. This is slightly different from Perl in that $ and @ are
3121    also handled as literals inside the quotes. In Perl, they will cause variable
3122    interpolation. Note the following examples:
3123    
3124        Pattern            PCRE matches      Perl matches
3125    
3126        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3127        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3128        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3129    
3130    For compatibility with Perl, \Q...\E sequences are recognized inside character
3131    classes as well as outside them.
3132    
3133    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3134    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3135    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3136    signed/unsigned warnings.
3137    
3138    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3139    option for pcretest, so I've replaced it by a simple function that does just
3140    that job.
3141    
3142    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3143    "pcregrep -".
3144    
3145    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3146    Java package. This provides some syntactic sugar for simple cases of what my
3147    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3148    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3149    item, you can use this simplified notation. Note that only makes sense with
3150    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3151    greediness, whatever the setting of the PCRE_UNGREEDY option.
3152    
3153    13. A change of greediness default within a pattern was not taking effect at
3154    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3155    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3156    was abstracted outside.
3157    
3158    14. PCRE now supports the \G assertion. It is true when the current matching
3159    position is at the start point of the match. This differs from \A when the
3160    starting offset is non-zero. Used with the /g option of pcretest (or similar
3161    code), it works in the same way as it does for Perl's /g option. If all
3162    alternatives of a regex begin with \G, the expression is anchored to the start
3163    match position, and the "anchored" flag is set in the compiled expression.
3164    
3165    15. Some bugs concerning the handling of certain option changes within patterns
3166    have been fixed. These applied to options other than (?ims). For example,
3167    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3168    been the other way round. Some of this was related to change 7 above.
3169    
3170    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3171    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3172    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3173    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3174    
3175    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3176    mode. Unlike ".", it always matches newline, whatever the setting of
3177    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3178    assertions. Perl allows it, but it doesn't (in general) work because it can't
3179    calculate the length of the lookbehind. At least, that's the case for Perl
3180    5.8.0 - I've been told they are going to document that it doesn't work in
3181    future.
3182    
3183    18. Added an error diagnosis for escapes that PCRE does not support: these are
3184    \L, \l, \N, \P, \p, \U, \u, and \X.
3185    
3186    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3187    reading past the end of the pattern in cases such as /[abcd/.
3188    
3189    20. PCRE was getting more memory than necessary for patterns with classes that
3190    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3191    
3192    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3193    compiling PCRE for use with Virtual Pascal.
3194    
3195    22. Small fix to the Makefile to make it work properly if the build is done
3196    outside the source tree.
3197    
3198    23. Added a new extension: a condition to go with recursion. If a conditional
3199    subpattern starts with (?(R) the "true" branch is used if recursion has
3200    happened, whereas the "false" branch is used only at the top level.
3201    
3202    24. When there was a very long string of literal characters (over 255 bytes
3203    without UTF support, over 250 bytes with UTF support), the computation of how
3204    much memory was required could be incorrect, leading to segfaults or other
3205    strange effects.
3206    
3207    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3208    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3209    there was a subsequent back reference to those brackets. This meant that, for
3210    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3211    possible to check for precisely this case. All we can do is abandon the
3212    optimization if .* occurs inside capturing brackets when there are any back
3213    references whatsoever. (See below for a better fix that came later.)
3214    
3215    26. The handling of the optimization for finding the first character of a
3216    non-anchored pattern, and for finding a character that is required later in the
3217    match were failing in some cases. This didn't break the matching; it just
3218    failed to optimize when it could. The way this is done has been re-implemented.
3219    
3220    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3221    
3222    28. Added a new feature that provides some of the functionality that Perl
3223    provides with (?{...}). The facility is termed a "callout". The way it is done
3224    in PCRE is for the caller to provide an optional function, by setting
3225    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3226    global variable. By default it is unset, which disables all calling out. To get
3227    the function called, the regex must include (?C) at appropriate points. This
3228    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3229    This provides a means of identifying different callout points. When PCRE
3230    reaches such a point in the regex, if pcre_callout has been set, the external
3231    function is called. It is provided with data in a structure called
3232    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3233    matching continues; if it returns a non-zero value, the match at the current
3234    point fails. However, backtracking will occur if possible. [This was changed
3235    later and other features added - see item 49 below.]
3236    
3237    29. pcretest is upgraded to test the callout functionality. It provides a
3238    callout function that displays information. By default, it shows the start of
3239    the match and the current position in the text. There are some new data escapes
3240    to vary what happens:
3241    
3242        \C+         in addition, show current contents of captured substrings
3243        \C-         do not supply a callout function
3244        \C!n        return 1 when callout number n is reached
3245        \C!n!m      return 1 when callout number n is reached for the mth time
3246    
3247    30. If pcregrep was called with the -l option and just a single file name, it
3248    output "<stdin>" if a match was found, instead of the file name.
3249    
3250    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3251    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3252    pcre_exec(). This saves a malloc/free per call. The default value of
3253    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3254    when configuring.
3255    
3256    32. The default maximum size of a compiled pattern is 64K. There have been a
3257    few cases of people hitting this limit. The code now uses macros to handle the
3258    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3259    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3260    configuring. Tests 2 and 5 work only with 2-byte links because they output
3261    debugging information about compiled patterns.
3262    
3263    33. Internal code re-arrangements:
3264    
3265    (a) Moved the debugging function for printing out a compiled regex into
3266        its own source file (printint.c) and used #include to pull it into
3267        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3268        separate copies.
3269    
3270    (b) Defined the list of op-code names for debugging as a macro in
3271        internal.h so that it is next to the definition of the opcodes.
3272    
3273    (c) Defined a table of op-code lengths for simpler skipping along compiled
3274        code. This is again a macro in internal.h so that it is next to the
3275        definition of the opcodes.
3276    
3277    34. Added support for recursive calls to individual subpatterns, along the
3278    lines of Robin Houston's patch (but implemented somewhat differently).
3279    
3280    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3281    allow it to read and process whole directories in Win32. This code was
3282    contributed by Lionel Fourquaux; it has not been tested by me.
3283    
3284    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3285    used to name a group. Names consist of alphanumerics and underscores, and must
3286    be unique. Back references use the syntax (?P=name) and recursive calls use
3287    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3288    numbers. The function pcre_fullinfo() can be used after compilation to extract
3289    a name/number map. There are three relevant calls:
3290    
3291      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3292      PCRE_INFO_NAMECOUNT            yields the number of entries
3293      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3294    
3295    The map is a vector of fixed-size entries. The size of each entry depends on
3296    the length of the longest name used. The first two bytes of each entry are the
3297    group number, most significant byte first. There follows the corresponding
3298    name, zero terminated. The names are in alphabetical order.
3299    
3300    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3301    case instead of 255. Making it the same both with and without UTF-8 support
3302    means that the same test output works with both.
3303    
3304    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3305    calling malloc() with a zero argument.
3306    
3307    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3308    optimization. I've improved things by keeping a bitmap of backreferences with
3309    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3310    fact referenced, the optimization can be applied. It is unlikely that a
3311    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3312    the match to follow \n) will appear inside brackets with a number greater than
3313    31, but if it does, any back reference > 31 suppresses the optimization.
3314    
3315    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3316    of disabling numbered capturing parentheses. Any opening parenthesis that is
3317    not followed by ? behaves as if it were followed by ?: but named parentheses
3318    can still be used for capturing (and they will acquire numbers in the usual
3319    way).
3320    
3321    41. Redesigned the return codes from the match() function into yes/no/error so
3322    that errors can be passed back from deep inside the nested calls. A malloc
3323    failure while inside a recursive subpattern call now causes the
3324    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3325    
3326    42. It is now possible to set a limit on the number of times the match()
3327    function is called in a call to pcre_exec(). This facility makes it possible to
3328    limit the amount of recursion and backtracking, though not in a directly
3329    obvious way, because the match() function is used in a number of different
3330    circumstances. The count starts from zero for each position in the subject
3331    string (for non-anchored patterns). The default limit is, for compatibility, a
3332    large number, namely 10 000 000. You can change this in two ways:
3333    
3334    (a) When configuring PCRE before making, you can use --with-match-limit=n
3335        to set a default value for the compiled library.
3336    
3337    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3338        a different value is set. See 45 below.
3339    
3340    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3341    
3342    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3343    of things that can be changed at compile time. The first argument specifies
3344    what is wanted and the second points to where the information is to be placed.
3345    The current list of available information is:
3346    
3347      PCRE_CONFIG_UTF8
3348    
3349    The output is an integer that is set to one if UTF-8 support is available;
3350    otherwise it is set to zero.
3351    
3352      PCRE_CONFIG_NEWLINE
3353    
3354    The output is an integer that it set to the value of the code that is used for
3355    newline. It is either LF (10) or CR (13).
3356    
3357      PCRE_CONFIG_LINK_SIZE
3358    
3359    The output is an integer that contains the number of bytes used for internal
3360    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3361    
3362      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3363    
3364    The output is an integer that contains the threshold above which the POSIX
3365    interface uses malloc() for output vectors. See item 31 above.
3366    
3367      PCRE_CONFIG_MATCH_LIMIT
3368    
3369    The output is an unsigned integer that contains the default limit of the number
3370    of match() calls in a pcre_exec() execution. See 42 above.
3371    
3372    44. pcretest has been upgraded by the addition of the -C option. This causes it
3373    to extract all the available output from the new pcre_config() function, and to
3374    output it. The program then exits immediately.
3375    
3376    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3377    order to support additional features. One way would have been to define
3378    pcre_exec2() (for example) with extra arguments, but this would not have been
3379    extensible, and would also have required all calls to the original function to
3380    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3381    is used for passing in "extra" data from pcre_study().
3382    
3383    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3384    contains the following fields:
3385    
3386      flags         a bitmap indicating which of the following fields are set
3387      study_data    opaque data from pcre_study()
3388      match_limit   a way of specifying a limit on match() calls for a specific
3389                      call to pcre_exec()
3390      callout_data  data for callouts (see 49 below)
3391    
3392    The flag bits are also defined in pcre.h, and are
3393    
3394      PCRE_EXTRA_STUDY_DATA
3395      PCRE_EXTRA_MATCH_LIMIT
3396      PCRE_EXTRA_CALLOUT_DATA
3397    
3398    The pcre_study() function now returns one of these new pcre_extra blocks, with
3399    the actual study data pointed to by the study_data field, and the
3400    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3401    before. That is, this change is entirely upwards-compatible and requires no
3402    change to existing code.
3403    
3404    If you want to pass in additional data to pcre_exec(), you can either place it
3405    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3406    block.
3407    
3408    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3409    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3410    times with different match limits, until it finds the minimum value needed for
3411    pcre_exec() to complete. The value is then output. This can be instructive; for
3412    most simple matches the number is quite small, but for pathological cases it
3413    gets very large very quickly.
3414    
3415    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3416    returns the size of the data block pointed to by the study_data field in a
3417    pcre_extra block, that is, the value that was passed as the argument to
3418    pcre_malloc() when PCRE was getting memory in which to place the information
3419    created by pcre_study(). The fourth argument should point to a size_t variable.
3420    pcretest has been extended so that this information is shown after a successful
3421    pcre_study() call when information about the compiled regex is being displayed.
3422    
3423    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3424    because what follows is always an absolute path. (Later: it turns out that this
3425    is more than cosmetic for MinGW, because it doesn't like empty path
3426    components.)
3427    
3428    49. Some changes have been made to the callout feature (see 28 above):
3429    
3430    (i)  A callout function now has three choices for what it returns:
3431    
3432           0  =>  success, carry on matching
3433         > 0  =>  failure at this point, but backtrack if possible
3434         < 0  =>  serious error, return this value from pcre_exec()
3435    
3436         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3437         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3438         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3439         use by callout functions. It will never be used by PCRE itself.
3440    
3441    (ii) The pcre_extra structure (see 45 above) has a void * field called
3442         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3443         pcre_callout_block structure has a field of the same name. The contents of
3444         the field passed in the pcre_extra structure are passed to the callout
3445         function in the corresponding field in the callout block. This makes it
3446         easier to use the same callout-containing regex from multiple threads. For
3447         testing, the pcretest program has a new data escape
3448    
3449           \C*n        pass the number n (may be negative) as callout_data
3450    
3451         If the callout function in pcretest receives a non-zero value as
3452         callout_data, it returns that value.
3453    
3454    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3455    there were some redundant $(CFLAGS) in commands that are now specified as
3456    $(LINK), which already includes $(CFLAGS).
3457    
3458    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3459    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3460    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3461    one-byte characters throughout. Note that case-insensitive matching applies
3462    only to characters whose values are less than 256. PCRE doesn't support the
3463    notion of cases for higher-valued characters.
3464    
3465    (i)   A character class whose characters are all within 0-255 is handled as
3466          a bit map, and the map is inverted for negative classes. Previously, a
3467          character > 255 always failed to match such a class; however it should
3468          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3469    
3470    (ii)  A negated character class with a single character < 255 is coded as
3471          "not this character" (OP_NOT). This wasn't working properly when the test
3472          character was multibyte, either singly or repeated.
3473    
3474    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3475          mode, for example: \x{100}{2,3}.
3476    
3477    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3478          singly or repeated) now correctly test multibyte characters. However,
3479          PCRE doesn't recognize any characters with values greater than 255 as
3480          digits, spaces, or word characters. Such characters always match \D, \S,
3481          and \W, and never match \d, \s, or \w.
3482    
3483    (v)   Classes may now contain characters and character ranges with values
3484          greater than 255. For example: [ab\x{100}-\x{400}].
3485    
3486    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3487          PCRE in UTF-8 mode.
3488    
3489    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3490    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3491    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3492    value.)
3493    
3494    53. The single man page has become too large. I have therefore split it up into
3495    a number of separate man pages. These also give rise to individual HTML pages;
3496    these are now put in a separate directory, and there is an index.html page that
3497    lists them all. Some hyperlinking between the pages has been installed.
3498    
3499    54. Added convenience functions for handling named capturing parentheses.
3500    
3501    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3502    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3503    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3504    are faulted.
3505    
3506    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3507    calling configure. These values are used when compiling the dftables.c program
3508    which is run to generate the source of the default character tables. They
3509    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3510    you will need to set these values.
3511    
3512    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3513    
3514    
3515    Version 3.9 02-Jan-02
3516    ---------------------
3517    
3518    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3519    
3520    2. If --disable-static was given, the building process failed when trying to
3521    build pcretest and pcregrep. (For some reason it was using libtool to compile
3522    them, which is not right, as they aren't part of the library.)
3523    
3524    
3525    Version 3.8 18-Dec-01
3526    ---------------------
3527    
3528    1. The experimental UTF-8 code was completely screwed up. It was packing the
3529    bytes in the wrong order. How dumb can you get?
3530    
3531    
3532    Version 3.7 29-Oct-01
3533    ---------------------
3534    
3535    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3536    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3537    this didn't happen under Solaris 8, where I normally test things.
3538    
3539    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3540    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3541    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3542    this fix an hour or so after the initial 3.7 release.)
3543    
3544    
3545    Version 3.6 23-Oct-01
3546    ---------------------
3547    
3548    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3549    offsets passed as NULL with zero offset count.
3550    
3551    2. The config.guess and config.sub files had not been updated when I moved to
3552    the latest autoconf.
3553    
3554    
3555    Version 3.5 15-Aug-01
3556    ---------------------
3557    
3558    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3559    had been forgotten.
3560    
3561    2. By using declared but undefined structures, we can avoid using "void"
3562    definitions in pcre.h while keeping the internal definitions of the structures
3563    private.
3564    
3565    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3566    user point of view, this means that both static and shared libraries are built
3567    by default, but this can be individually controlled. More of the work of
3568    handling this static/shared cases is now inside libtool instead of PCRE's make
3569    file.
3570    
3571    4. The pcretest utility is now installed along with pcregrep because it is
3572    useful for users (to test regexs) and by doing this, it automatically gets
3573    relinked by libtool. The documentation has been turned into a man page, so
3574    there are now .1, .txt, and .html versions in /doc.
3575    
3576    5. Upgrades to pcregrep:
3577       (i)   Added long-form option names like gnu grep.
3578       (ii)  Added --help to list all options with an explanatory phrase.
3579       (iii) Added -r, --recursive to recurse into sub-directories.
3580       (iv)  Added -f, --file to read patterns from a file.
3581    
3582    6. pcre_exec() was referring to its "code" argument before testing that
3583    argument for NULL (and giving an error if it was NULL).
3584    
3585    7. Upgraded Makefile.in to allow for compiling in a different directory from
3586    the source directory.
3587    
3588    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3589    options bits, the pointer it was passed was to an int instead of to an unsigned
3590    long int. This mattered only on 64-bit systems.
3591    
3592    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3593    generated) instead of pcre.in, which it its source. Also made the same change
3594    in several of the .c files.
3595    
3596    10. A new release of gcc defines printf() as a macro, which broke pcretest
3597    because it had an ifdef in the middle of a string argument for printf(). Fixed
3598    by using separate calls to printf().
3599    
3600    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3601    script, to force use of CR or LF instead of \n in the source. On non-Unix
3602    systems, the value can be set in config.h.
3603    
3604    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3605    absolute limit. Changed the text of the error message to make this clear, and
3606    likewise updated the man page.
3607    
3608    13. The limit of 99 on the number of capturing subpatterns has been removed.
3609    The new limit is 65535, which I hope will not be a "real" limit.
3610    
3611    
3612    Version 3.4 22-Aug-00
3613    ---------------------
3614    
3615    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3616    
3617    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3618    
3619    
3620    Version 3.3 01-Aug-00
3621    ---------------------
3622    
3623    1. If an octal character was given, but the value was greater than \377, it
3624    was not getting masked to the least significant bits, as documented. This could
3625    lead to crashes in some systems.
3626    
3627    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3628    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3629    
3630    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3631    These just pass their arguments on to (pcre_free)(), but they are provided
3632    because some uses of PCRE bind it to non-C systems that can call its functions,
3633    but cannot call free() or pcre_free() directly.
3634    
3635    4. Add "make test" as a synonym for "make check". Corrected some comments in
3636    the Makefile.
3637    
3638    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3639    Makefile.
3640    
3641    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3642    command called pgrep for grepping around the active processes.
3643    
3644    7. Added the beginnings of support for UTF-8 character strings.
3645    
3646    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3647    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3648    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3649    out for the ar command.)
3650    
3651    
3652    Version 3.2 12-May-00
3653    ---------------------
3654    
3655    This is purely a bug fixing release.
3656    
3657    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3658    of ZA. This was just one example of several cases that could provoke this bug,
3659    which was introduced by change 9 of version 2.00. The code for breaking
3660    infinite loops after an iteration that matches an empty string was't working
3661    correctly.
3662    
3663    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3664    when matched against abbab (for example). After matching an empty string, it
3665    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3666    caused it to match further down the string than it should.
3667    
3668    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3669    was there because it doesn't seem to be needed, and it causes trouble on some
3670    systems, as it is not a Standard C header. It has been removed.
3671    
3672    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3673    were reported on the Macintosh. The changes were from
3674    
3675      while ((c = *(++ptr)) != 0 && c != '\n');
3676    to
3677      while ((c = *(++ptr)) != 0 && c != '\n') ;
3678    
3679    Totally extraordinary, but if that's what it takes...
3680    
3681    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3682    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3683    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3684    assumes the way PCRE uses memmove() (always moving upwards).
3685    
3686    6. PCRE is being used in one environment where strchr() is not available. There
3687    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3688    faster code anyway.
3689    
3690    
3691    Version 3.1 09-Feb-00
3692    ---------------------
3693    
3694    The only change in this release is the fixing of some bugs in Makefile.in for
3695    the "install" target:
3696    
3697    (1) It was failing to install pcreposix.h.
3698    
3699    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3700    
3701    
3702    Version 3.0 01-Feb-00
3703    ---------------------
3704    
3705    1. Add support for the /+ modifier to perltest (to output $` like it does in
3706    pcretest).
3707    
3708    2. Add support for the /g modifier to perltest.
3709    
3710    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3711    matches null strings.
3712    
3713    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3714    pattern. Perl treats empty patterns specially - it reuses the most recent
3715    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3716    effect.
3717    
3718    5. The POSIX interface was broken in that it was just handing over the POSIX
3719    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3720    required a bigger vector, with some working space on the end. This means that
3721    the POSIX wrapper now has to get and free some memory, and copy the results.
3722    
3723    6. Added some simple autoconf support, placing the test data and the
3724    documentation in separate directories, re-organizing some of the
3725    information files, and making it build pcre-config (a GNU standard). Also added
3726    libtool support for building PCRE as a shared library, which is now the
3727    default.
3728    
3729    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3730    09 are not valid octal constants. Single digits will be used for minor values
3731    less than 10.
3732    
3733    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3734    existing programs that set these in the POSIX interface can use PCRE without
3735    modification.
3736    
3737    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3738    return all that pcre_info() returns, plus additional data. The pcre_info()
3739    function is retained for compatibility, but is considered to be obsolete.
3740    
3741    10. Added experimental recursion feature (?R) to handle one common case that
3742    Perl 5.6 will be able to do with (?p{...}).
3743    
3744    11. Added support for POSIX character classes like [:alpha:], which Perl is
3745    adopting.
3746    
3747    
3748    Version 2.08 31-Aug-99
3749    ----------------------
3750    
3751    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3752    trying to match at the startoffset position, but instead was moving forward to
3753    the next newline as if a previous match had failed.
3754    
3755    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3756    and could get into a loop if a null string was matched other than at the start
3757    of the subject.
3758    
3759    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3760    be distinguished at compile time, and for completeness also added PCRE_DATE.
3761    
3762    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3763    in GnuWin32 environments.
3764    
3765    
3766    Version 2.07 29-Jul-99
3767    ----------------------
3768    
3769    1. The documentation is now supplied in plain text form and HTML as well as in
3770    the form of man page sources.
3771    
3772    2. C++ compilers don't like assigning (void *) values to other pointer types.
3773    In particular this affects malloc(). Although there is no problem in Standard
3774    C, I've put in casts to keep C++ compilers happy.
3775    
3776    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3777    should be (const char *).
3778    
3779    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3780    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3781    However, I haven't made this a standard facility. The documentation doesn't
3782    mention it, and the Makefile doesn't support it.
3783    
3784    5. The Makefile now contains an "install" target, with editable destinations at
3785    the top of the file. The pcretest program is not installed.
3786    
3787    6. pgrep -V now gives the PCRE version number and date.
3788    
3789    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3790    causing the entire string to be ignored, instead of just the last character.
3791    
3792    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3793    non-matching string, it can take a very, very long time, even for strings of
3794    quite modest length, because of the nested recursion. PCRE now does better in
3795    some of these cases. It does this by remembering the last required literal
3796    character in the pattern, and pre-searching the subject to ensure it is present
3797    before running the real match. In other words, it applies a heuristic to detect
3798    some types of certain failure quickly, and in the above example, if presented
3799    with a string that has no trailing " it gives "no match" very quickly.
3800    
3801    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3802    other alternatives are tried instead.
3803    
3804    
3805    Version 2.06 09-Jun-99
3806    ----------------------
3807    
3808    1. Change pcretest's output for amount of store used to show just the code
3809    space, because the remainder (the data block) varies in size between 32-bit and
3810    64-bit systems.
3811    
3812    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3813    start matching at. This allows lookbehinds to work when searching for multiple
3814    occurrences in a string.
3815    
3816    3. Added additional options to pcretest for testing multiple occurrences:
3817    
3818       /+   outputs the rest of the string that follows a match
3819       /g   loops for multiple occurrences, using the new startoffset argument
3820       /G   loops for multiple occurrences by passing an incremented pointer
3821    
3822    4. PCRE wasn't doing the "first character" optimization for patterns starting
3823    with \b or \B, though it was doing it for other lookbehind assertions. That is,
3824    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3825    the letter 'x'. On long subject strings, this gives a significant speed-up.
3826    
3827    
3828    Version 2.05 21-Apr-99