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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.65  
changed lines
  Added in v.585

  ViewVC Help
Powered by ViewVC 1.1.5