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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.53  
changed lines
  Added in v.611

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12