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

Diff of /code/trunk/ChangeLog

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

revision 197 by ph10, Tue Jul 31 10:50:18 2007 UTC revision 586 by ph10, Wed Jan 12 17:36:47 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.3 05-Jul-07  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    4.  If pcregrep was compiled under Windows, there was a reference to the
21        function pcregrep_exit() before it was defined. I am assuming this was
22        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
23        reported by a user. I've moved the definition above the reference.
24    
25    
26    Version 8.11 10-Dec-2010
27    ------------------------
28    
29    1.  (*THEN) was not working properly if there were untried alternatives prior
30        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
31        backtracked to try for "b" instead of moving to the next alternative branch
32        at the same level (in this case, to look for "c"). The Perl documentation
33        is clear that when (*THEN) is backtracked onto, it goes to the "next
34        alternative in the innermost enclosing group".
35    
36    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
37        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
38        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
39        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
40        (*THEN).
41    
42    3.  If \s appeared in a character class, it removed the VT character from
43        the class, even if it had been included by some previous item, for example
44        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
45        of \s, but is part of the POSIX "space" class.)
46    
47    4.  A partial match never returns an empty string (because you can always
48        match an empty string at the end of the subject); however the checking for
49        an empty string was starting at the "start of match" point. This has been
50        changed to the "earliest inspected character" point, because the returned
51        data for a partial match starts at this character. This means that, for
52        example, /(?<=abc)def/ gives a partial match for the subject "abc"
53        (previously it gave "no match").
54    
55    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
56        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
57        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
58        has an implication that the given string is incomplete (because a partial
59        match is preferred over a full match). For this reason, these items now
60        give a partial match in this situation. [Aside: previously, the one case
61        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
62        match rather than a full match, which was wrong by the old rules, but is
63        now correct.]
64    
65    6.  There was a bug in the handling of #-introduced comments, recognized when
66        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
67        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
68        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
69        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
70        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
71        places in pcre_compile().
72    
73    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
74        comments when looking ahead for named forward references to subpatterns,
75        the only newline sequence it recognized was NL. It now handles newlines
76        according to the set newline convention.
77    
78    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
79        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
80        cater for a lack of strerror(). These oversights have been fixed.
81    
82    9.  Added --match-limit and --recursion-limit to pcregrep.
83    
84    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
85    
86    11. When the -o option was used, pcregrep was setting a return code of 1, even
87        when matches were found, and --line-buffered was not being honoured.
88    
89    12. Added an optional parentheses number to the -o and --only-matching options
90        of pcregrep.
91    
92    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
93        can match an empty string. The code to do it in pcretest and pcredemo
94        needed fixing:
95    
96        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
97            only one byte after an empty string match just before CRLF (this case
98            just got forgotten; "any" and "anycrlf" were OK).
99    
100        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
101            mode when an empty string match preceded an ASCII character followed by
102            a non-ASCII character. (The code for advancing by one character rather
103            than one byte was nonsense.)
104    
105        (c) The pcredemo.c sample program did not have any code at all to handle
106            the cases when CRLF is a valid newline sequence.
107    
108    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
109        as a starting offset was within the subject string. There is now a new
110        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
111        negative or greater than the length of the string. In order to test this,
112        pcretest is extended to allow the setting of negative starting offsets.
113    
114    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
115        starting offset points to the beginning of a UTF-8 character was
116        unnecessarily clumsy. I tidied it up.
117    
118    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
119        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
120    
121    17. Nobody had reported that the --include_dir option, which was added in
122        release 7.7 should have been called --include-dir (hyphen, not underscore)
123        for compatibility with GNU grep. I have changed it to --include-dir, but
124        left --include_dir as an undocumented synonym, and the same for
125        --exclude-dir, though that is not available in GNU grep, at least as of
126        release 2.5.4.
127    
128    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
129        characters from a string of bytes) have been redefined so as not to use
130        loops, in order to improve performance in some environments. At the same
131        time, I abstracted some of the common code into auxiliary macros to save
132        repetition (this should not affect the compiled code).
133    
134    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
135        compile-time error is now given if \c is not followed by an ASCII
136        character, that is, a byte less than 128. (In EBCDIC mode, the code is
137        different, and any byte value is allowed.)
138    
139    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
140        START_OPTIMIZE option, which is now allowed at compile time - but just
141        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
142        to pcregrep and other applications that have no direct access to PCRE
143        options. The new /Y option in pcretest sets this option when calling
144        pcre_compile().
145    
146    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
147        back references. Groups containing recursive back references were forced to
148        be atomic by that change, but in the case of named groups, the amount of
149        memory required was incorrectly computed, leading to "Failed: internal
150        error: code overflow". This has been fixed.
151    
152    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
153        pcretest.c, to avoid build problems in some Borland environments.
154    
155    
156    Version 8.10 25-Jun-2010
157    ------------------------
158    
159    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
160        THEN.
161    
162    2.  (*ACCEPT) was not working when inside an atomic group.
163    
164    3.  Inside a character class, \B is treated as a literal by default, but
165        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
166        causes the error). The code is unchanged, but I tidied the documentation.
167    
168    4.  Inside a character class, PCRE always treated \R and \X as literals,
169        whereas Perl faults them if its -w option is set. I have changed PCRE so
170        that it faults them when PCRE_EXTRA is set.
171    
172    5.  Added support for \N, which always matches any character other than
173        newline. (It is the same as "." when PCRE_DOTALL is not set.)
174    
175    6.  When compiling pcregrep with newer versions of gcc which may have
176        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
177        declared with attribute warn_unused_result" were given. Just casting the
178        result to (void) does not stop the warnings; a more elaborate fudge is
179        needed. I've used a macro to implement this.
180    
181    7.  Minor change to pcretest.c to avoid a compiler warning.
182    
183    8.  Added four artifical Unicode properties to help with an option to make
184        \s etc use properties (see next item). The new properties are: Xan
185        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
186    
187    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
188        use Unicode properties. (*UCP) at the start of a pattern can be used to set
189        this option. Modified pcretest to add /W to test this facility. Added
190        REG_UCP to make it available via the POSIX interface.
191    
192    10. Added --line-buffered to pcregrep.
193    
194    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
195        studied, and the match started with a letter with a code point greater than
196        127 whose first byte was different to the first byte of the other case of
197        the letter, the other case of this starting letter was not recognized
198        (#976).
199    
200    12. If a pattern that was studied started with a repeated Unicode property
201        test, for example, \p{Nd}+, there was the theoretical possibility of
202        setting up an incorrect bitmap of starting bytes, but fortunately it could
203        not have actually happened in practice until change 8 above was made (it
204        added property types that matched character-matching opcodes).
205    
206    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
207        possible starting bytes for non-anchored patterns.
208    
209    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
210        \R, and also a number of cases that involve Unicode properties, both
211        explicit and implicit when PCRE_UCP is set.
212    
213    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
214        input, it could crash or give wrong results if characters with values
215        greater than 0xc0 were present in the subject string. (Detail: it assumed
216        UTF-8 input when processing these items.)
217    
218    16. Added a lot of (int) casts to avoid compiler warnings in systems where
219        size_t is 64-bit (#991).
220    
221    17. Added a check for running out of memory when PCRE is compiled with
222        --disable-stack-for-recursion (#990).
223    
224    18. If the last data line in a file for pcretest does not have a newline on
225        the end, a newline was missing in the output.
226    
227    19. The default pcre_chartables.c file recognizes only ASCII characters (values
228        less than 128) in its various bitmaps. However, there is a facility for
229        generating tables according to the current locale when PCRE is compiled. It
230        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
231        characters, are recognized by isspace() and therefore were getting set in
232        these tables, and indeed these tables seem to approximate to ISO 8859. This
233        caused a problem in UTF-8 mode when pcre_study() was used to create a list
234        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
235        which of course cannot start UTF-8 characters. I have changed the code so
236        that only real ASCII characters (less than 128) and the correct starting
237        bytes for UTF-8 encodings are set for characters greater than 127 when in
238        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
239        altogether.)
240    
241    20. Added the /T option to pcretest so as to be able to run tests with non-
242        standard character tables, thus making it possible to include the tests
243        used for 19 above in the standard set of tests.
244    
245    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
246        reference to a subpattern the other side of a comment that contains an
247        opening parenthesis caused either an internal compiling error, or a
248        reference to the wrong subpattern.
249    
250    
251    Version 8.02 19-Mar-2010
252    ------------------------
253    
254    1.  The Unicode data tables have been updated to Unicode 5.2.0.
255    
256    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
257        configured.
258    
259    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
260        original author of that file, following a query about its status.
261    
262    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
263        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
264    
265    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
266        quantifier applied to a forward-referencing subroutine call, could compile
267        incorrect code or give the error "internal error: previously-checked
268        referenced subpattern not found".
269    
270    6.  Both MS Visual Studio and Symbian OS have problems with initializing
271        variables to point to external functions. For these systems, therefore,
272        pcre_malloc etc. are now initialized to local functions that call the
273        relevant global functions.
274    
275    7.  There were two entries missing in the vectors called coptable and poptable
276        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
277        I've fixed the data, and added a kludgy way of testing at compile time that
278        the lengths are correct (equal to the number of opcodes).
279    
280    8.  Following on from 7, I added a similar kludge to check the length of the
281        eint vector in pcreposix.c.
282    
283    9.  Error texts for pcre_compile() are held as one long string to avoid too
284        much relocation at load time. To find a text, the string is searched,
285        counting zeros. There was no check for running off the end of the string,
286        which could happen if a new error number was added without updating the
287        string.
288    
289    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
290    
291    11. \K was not working if it appeared in an atomic group or in a group that
292        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
293        \K is "not well defined" if used in an assertion. PCRE now accepts it if
294        the assertion is positive, but not if it is negative.
295    
296    12. Change 11 fortuitously reduced the size of the stack frame used in the
297        "match()" function of pcre_exec.c by one pointer. Forthcoming
298        implementation of support for (*MARK) will need an extra pointer on the
299        stack; I have reserved it now, so that the stack frame size does not
300        decrease.
301    
302    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
303        item in branch that calls a recursion is a subroutine call - as in the
304        second branch in the above example - was incorrectly given the compile-
305        time error "recursive call could loop indefinitely" because pcre_compile()
306        was not correctly checking the subroutine for matching a non-empty string.
307    
308    14. The checks for overrunning compiling workspace could trigger after an
309        overrun had occurred. This is a "should never occur" error, but it can be
310        triggered by pathological patterns such as hundreds of nested parentheses.
311        The checks now trigger 100 bytes before the end of the workspace.
312    
313    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
314    
315    
316    Version 8.01 19-Jan-2010
317    ------------------------
318    
319    1.  If a pattern contained a conditional subpattern with only one branch (in
320        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
321        computed the wrong minimum data length (which is of course zero for such
322        subpatterns). This could cause incorrect "no match" results.
323    
324    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
325        the pattern is reset in the first branch, pcre_compile() failed with
326        "internal error: code overflow at offset...". This happened only when
327        the reset was to the original external option setting. (An optimization
328        abstracts leading options settings into an external setting, which was the
329        cause of this.)
330    
331    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
332        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
333        assertion pattern did not match (meaning that the assertion was true), it
334        was incorrectly treated as false if the SKIP had been reached during the
335        matching. This also applied to assertions used as conditions.
336    
337    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
338        assertion subpattern, including such a pattern used as a condition,
339        unpredictable results occurred, instead of the error return
340        PCRE_ERROR_DFA_UITEM.
341    
342    5.  The C++ GlobalReplace function was not working like Perl for the special
343        situation when an empty string is matched. It now does the fancy magic
344        stuff that is necessary.
345    
346    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
347        removed. (These were left over from very, very early versions of PCRE.)
348    
349    7.  Some cosmetic changes to the code to make life easier when compiling it
350        as part of something else:
351    
352        (a) Change DEBUG to PCRE_DEBUG.
353    
354        (b) In pcre_compile(), rename the member of the "branch_chain" structure
355            called "current" as "current_branch", to prevent a collision with the
356            Linux macro when compiled as a kernel module.
357    
358        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
359            prevent a collision with the Linux macro when compiled as a kernel
360            module.
361    
362    8.  In pcre_compile() there are some checks for integer overflows that used to
363        cast potentially large values to (double). This has been changed to that
364        when building, a check for int64_t is made, and if it is found, it is used
365        instead, thus avoiding the use of floating point arithmetic. (There is no
366        other use of FP in PCRE.) If int64_t is not found, the fallback is to
367        double.
368    
369    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
370        2005 (difference between two addresses compared to an unsigned value).
371    
372    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
373        custom one, because of the following reported problem in Windows:
374    
375          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
376              under Win32.
377          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
378              therefore missing the function definition.
379          - The compiler thus generates a "C" signature for the test function.
380          - The linker fails to find the "C" function.
381          - PCRE fails to configure if asked to do so against libbz2.
382    
383    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
384        messages were output:
385    
386          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
387          rerunning libtoolize, to keep the correct libtool macros in-tree.
388          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
389    
390        I have done both of these things.
391    
392    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
393        most of the time, it *can* run out if it is given a pattern that contains a
394        runaway infinite recursion. I updated the discussion in the pcrestack man
395        page.
396    
397    13. Now that we have gone to the x.xx style of version numbers, the minor
398        version may start with zero. Using 08 or 09 is a bad idea because users
399        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
400        interpreted as invalid octal numbers. I've updated the previous comment in
401        configure.ac, and also added a check that gives an error if 08 or 09 are
402        used.
403    
404    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
405        causing partial matching to fail when the end of the subject matched \W
406        in a UTF-8 pattern where \W was quantified with a minimum of 3.
407    
408    15. There were some discrepancies between the declarations in pcre_internal.h
409        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
410        their definitions. The declarations used "const uschar *" and the
411        definitions used USPTR. Even though USPTR is normally defined as "const
412        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
413        reported that: "This difference in casting confuses some C++ compilers, for
414        example, SunCC recognizes above declarations as different functions and
415        generates broken code for hbpcre." I have changed the declarations to use
416        USPTR.
417    
418    16. GNU libtool is named differently on some systems. The autogen.sh script now
419        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
420        (FreeBSD).
421    
422    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
423        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
424        comment: "Figure out how to create a longlong from a string: strtoll and
425        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
426        instance, but it only takes 2 args instead of 3!"
427    
428    18. A subtle bug concerned with back references has been fixed by a change of
429        specification, with a corresponding code fix. A pattern such as
430        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
431        refers, was giving matches when it shouldn't. For example, xa=xaaa would
432        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
433        same bug. Such groups have to be quantified to be useful, or contained
434        inside another quantified group. (If there's no repetition, the reference
435        can never match.) The problem arises because, having left the group and
436        moved on to the rest of the pattern, a later failure that backtracks into
437        the group uses the captured value from the final iteration of the group
438        rather than the correct earlier one. I have fixed this in PCRE by forcing
439        any group that contains a reference to itself to be an atomic group; that
440        is, there cannot be any backtracking into it once it has completed. This is
441        similar to recursive and subroutine calls.
442    
443    
444    Version 8.00 19-Oct-09
445    ----------------------
446    
447    1.  The table for translating pcre_compile() error codes into POSIX error codes
448        was out-of-date, and there was no check on the pcre_compile() error code
449        being within the table. This could lead to an OK return being given in
450        error.
451    
452    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
453        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
454        in a Windows environment.
455    
456    3.  The pcregrep --count option prints the count for each file even when it is
457        zero, as does GNU grep. However, pcregrep was also printing all files when
458        --files-with-matches was added. Now, when both options are given, it prints
459        counts only for those files that have at least one match. (GNU grep just
460        prints the file name in this circumstance, but including the count seems
461        more useful - otherwise, why use --count?) Also ensured that the
462        combination -clh just lists non-zero counts, with no names.
463    
464    4.  The long form of the pcregrep -F option was incorrectly implemented as
465        --fixed_strings instead of --fixed-strings. This is an incompatible change,
466        but it seems right to fix it, and I didn't think it was worth preserving
467        the old behaviour.
468    
469    5.  The command line items --regex=pattern and --regexp=pattern were not
470        recognized by pcregrep, which required --regex pattern or --regexp pattern
471        (with a space rather than an '='). The man page documented the '=' forms,
472        which are compatible with GNU grep; these now work.
473    
474    6.  No libpcreposix.pc file was created for pkg-config; there was just
475        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
476    
477    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
478        when UCP support is not needed, by modifying the Python script that
479        generates it from Unicode data files. This should not matter if the module
480        is correctly used as a library, but I received one complaint about 50K of
481        unwanted data. My guess is that the person linked everything into his
482        program rather than using a library. Anyway, it does no harm.
483    
484    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
485        was a minimum greater than 1 for a wide character in a possessive
486        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
487        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
488        character. Chaos in the form of incorrect output or a compiling loop could
489        result.
490    
491    9.  The restrictions on what a pattern can contain when partial matching is
492        requested for pcre_exec() have been removed. All patterns can now be
493        partially matched by this function. In addition, if there are at least two
494        slots in the offset vector, the offset of the earliest inspected character
495        for the match and the offset of the end of the subject are set in them when
496        PCRE_ERROR_PARTIAL is returned.
497    
498    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
499        synonymous with PCRE_PARTIAL, for backwards compatibility, and
500        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
501        and may be more useful for multi-segment matching.
502    
503    11. Partial matching with pcre_exec() is now more intuitive. A partial match
504        used to be given if ever the end of the subject was reached; now it is
505        given only if matching could not proceed because another character was
506        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
507        string "Z", which now yields "no match" instead of "partial match". In the
508        case of pcre_dfa_exec(), "no match" is given if every matching path for the
509        final character ended with (*FAIL).
510    
511    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
512        if the pattern had a "must contain" character that was already found in the
513        earlier partial match, unless partial matching was again requested. For
514        example, with the pattern /dog.(body)?/, the "must contain" character is
515        "g". If the first part-match was for the string "dog", restarting with
516        "sbody" failed. This bug has been fixed.
517    
518    13. The string returned by pcre_dfa_exec() after a partial match has been
519        changed so that it starts at the first inspected character rather than the
520        first character of the match. This makes a difference only if the pattern
521        starts with a lookbehind assertion or \b or \B (\K is not supported by
522        pcre_dfa_exec()). It's an incompatible change, but it makes the two
523        matching functions compatible, and I think it's the right thing to do.
524    
525    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
526        so that the demonstration program is easily available in environments where
527        PCRE has not been installed from source.
528    
529    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
530        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
531        library.
532    
533    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
534        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
535        is not the first non-POSIX option to be added. Clearly some people find
536        these options useful.
537    
538    17. If a caller to the POSIX matching function regexec() passes a non-zero
539        value for nmatch with a NULL value for pmatch, the value of
540        nmatch is forced to zero.
541    
542    18. RunGrepTest did not have a test for the availability of the -u option of
543        the diff command, as RunTest does. It now checks in the same way as
544        RunTest, and also checks for the -b option.
545    
546    19. If an odd number of negated classes containing just a single character
547        interposed, within parentheses, between a forward reference to a named
548        subpattern and the definition of the subpattern, compilation crashed with
549        an internal error, complaining that it could not find the referenced
550        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
551        [The bug was that it was starting one character too far in when skipping
552        over the character class, thus treating the ] as data rather than
553        terminating the class. This meant it could skip too much.]
554    
555    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
556        /g option in pcretest when the pattern contains \K, which makes it possible
557        to have an empty string match not at the start, even when the pattern is
558        anchored. Updated pcretest and pcredemo to use this option.
559    
560    21. If the maximum number of capturing subpatterns in a recursion was greater
561        than the maximum at the outer level, the higher number was returned, but
562        with unset values at the outer level. The correct (outer level) value is
563        now given.
564    
565    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
566        PCRE did not set those parentheses (unlike Perl). I have now found a way to
567        make it do so. The string so far is captured, making this feature
568        compatible with Perl.
569    
570    23. The tests have been re-organized, adding tests 11 and 12, to make it
571        possible to check the Perl 5.10 features against Perl 5.10.
572    
573    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
574        pattern matches a fixed length string. PCRE did not allow this; now it
575        does. Neither allows recursion.
576    
577    25. I finally figured out how to implement a request to provide the minimum
578        length of subject string that was needed in order to match a given pattern.
579        (It was back references and recursion that I had previously got hung up
580        on.) This code has now been added to pcre_study(); it finds a lower bound
581        to the length of subject needed. It is not necessarily the greatest lower
582        bound, but using it to avoid searching strings that are too short does give
583        some useful speed-ups. The value is available to calling programs via
584        pcre_fullinfo().
585    
586    26. While implementing 25, I discovered to my embarrassment that pcretest had
587        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
588        study optimizations had never been tested with that matching function.
589        Oops. What is worse, even when it was passed study data, there was a bug in
590        pcre_dfa_exec() that meant it never actually used it. Double oops. There
591        were also very few tests of studied patterns with pcre_dfa_exec().
592    
593    27. If (?| is used to create subpatterns with duplicate numbers, they are now
594        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
595        on the other side of the coin, they are no longer allowed to have different
596        names, because these cannot be distinguished in PCRE, and this has caused
597        confusion. (This is a difference from Perl.)
598    
599    28. When duplicate subpattern names are present (necessarily with different
600        numbers, as required by 27 above), and a test is made by name in a
601        conditional pattern, either for a subpattern having been matched, or for
602        recursion in such a pattern, all the associated numbered subpatterns are
603        tested, and the overall condition is true if the condition is true for any
604        one of them. This is the way Perl works, and is also more like the way
605        testing by number works.
606    
607    
608    Version 7.9 11-Apr-09
609    ---------------------
610    
611    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
612        (pcretest), all targets were linked against these libraries. This included
613        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
614        libraries. This caused unwanted dependencies to be created. This problem
615        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
616        pcretest is linked with readline.
617    
618    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
619        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
620        moved outside it again, because FALSE and TRUE are already defined in AIX,
621        but BOOL is not.
622    
623    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
624        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
625    
626    4.  The pcregrep documentation said spaces were inserted as well as colons (or
627        hyphens) following file names and line numbers when outputting matching
628        lines. This is not true; no spaces are inserted. I have also clarified the
629        wording for the --colour (or --color) option.
630    
631    5.  In pcregrep, when --colour was used with -o, the list of matching strings
632        was not coloured; this is different to GNU grep, so I have changed it to be
633        the same.
634    
635    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
636        each matching line was coloured. Now it goes on to look for further matches
637        of any of the test patterns, which is the same behaviour as GNU grep.
638    
639    7.  A pattern that could match an empty string could cause pcregrep to loop; it
640        doesn't make sense to accept an empty string match in pcregrep, so I have
641        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
642        seems to be how GNU grep behaves.
643    
644    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
645        start or after a newline", because the conditional assertion was not being
646        correctly handled. The rule now is that both the assertion and what follows
647        in the first alternative must satisfy the test.
648    
649    9.  If auto-callout was enabled in a pattern with a conditional group whose
650        condition was an assertion, PCRE could crash during matching, both with
651        pcre_exec() and pcre_dfa_exec().
652    
653    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
654        used for matching.
655    
656    11. Unicode property support in character classes was not working for
657        characters (bytes) greater than 127 when not in UTF-8 mode.
658    
659    12. Added the -M command line option to pcretest.
660    
661    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
662    
663    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
664    
665    16. Added comments and documentation about mis-use of no_arg in the C++
666        wrapper.
667    
668    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
669        from Martin Jerabek that uses macro names for all relevant character and
670        string constants.
671    
672    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
673        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
674        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
675        these, but not everybody uses configure.
676    
677    19. A conditional group that had only one branch was not being correctly
678        recognized as an item that could match an empty string. This meant that an
679        enclosing group might also not be so recognized, causing infinite looping
680        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
681        with the subject "ab", where knowledge that the repeated group can match
682        nothing is needed in order to break the loop.
683    
684    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
685        exec(), but without supplying a callout function, matching went wrong.
686    
687    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
688        leak if the size of the offset vector was greater than 30. When the vector
689        is smaller, the saved offsets during recursion go onto a local stack
690        vector, but for larger vectors malloc() is used. It was failing to free
691        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
692        error, in fact).
693    
694    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
695        heapframe that is used only when UTF-8 support is enabled. This caused no
696        problem, but was untidy.
697    
698    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
699        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
700        included within another project.
701    
702    24. Steven Van Ingelgem's patches to add more options to the CMake support,
703        slightly modified by me:
704    
705          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
706              not building pcregrep.
707    
708          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
709              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
710    
711    25. Forward references, both numeric and by name, in patterns that made use of
712        duplicate group numbers, could behave incorrectly or give incorrect errors,
713        because when scanning forward to find the reference group, PCRE was not
714        taking into account the duplicate group numbers. A pattern such as
715        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
716    
717    26. Changed a few more instances of "const unsigned char *" to USPTR, making
718        the feature of a custom pointer more persuasive (as requested by a user).
719    
720    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
721        pcretest.c, inside #ifndefs, because it seems they are sometimes already
722        pre-defined.
723    
724    28. Added support for (*UTF8) at the start of a pattern.
725    
726    29. Arrange for flags added by the "release type" setting in CMake to be shown
727        in the configuration summary.
728    
729    
730    Version 7.8 05-Sep-08
731    ---------------------
732    
733    1.  Replaced UCP searching code with optimized version as implemented for Ad
734        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
735        stage table and inline lookup instead of a function, giving speed ups of 2
736        to 5 times on some simple patterns that I tested. Permission was given to
737        distribute the MultiStage2.py script that generates the tables (it's not in
738        the tarball, but is in the Subversion repository).
739    
740    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
741        scripts.
742    
743    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
744        a group with a zero qualifier. The result of the study could be incorrect,
745        or the function might crash, depending on the pattern.
746    
747    4.  Caseless matching was not working for non-ASCII characters in back
748        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
749        It now works when Unicode Property Support is available.
750    
751    5.  In pcretest, an escape such as \x{de} in the data was always generating
752        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
753        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
754        truncation.
755    
756    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
757    
758    7.  Added two (int) casts to pcregrep when printing the difference of two
759        pointers, in case they are 64-bit values.
760    
761    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
762        test 2 if it fails.
763    
764    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
765        and a #define of that name to empty if it is not externally set. This is to
766        allow users of MSVC to set it if necessary.
767    
768    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
769        the convenience functions in the pcre_get.c source file.
770    
771    11. An option change at the start of a pattern that had top-level alternatives
772        could cause overwriting and/or a crash. This command provoked a crash in
773        some environments:
774    
775          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
776    
777        This potential security problem was recorded as CVE-2008-2371.
778    
779    12. For a pattern where the match had to start at the beginning or immediately
780        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
781        pcre_dfa_exec() could read past the end of the passed subject if there was
782        no match. To help with detecting such bugs (e.g. with valgrind), I modified
783        pcretest so that it places the subject at the end of its malloc-ed buffer.
784    
785    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
786        exec() might read past the end of the data buffer in UTF-8 mode.
787    
788    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
789        the data contained the byte 0x85 as part of a UTF-8 character within its
790        first line. This applied both to normal and DFA matching.
791    
792    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
793        /^[^d]*?$/8 failed to match "abc".
794    
795    16. Added a missing copyright notice to pcrecpp_internal.h.
796    
797    17. Make it more clear in the documentation that values returned from
798        pcre_exec() in ovector are byte offsets, not character counts.
799    
800    18. Tidied a few places to stop certain compilers from issuing warnings.
801    
802    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
803        supplied by Stefan Weber. I made a further small update for 7.8 because
804        there is a change of source arrangements: the pcre_searchfuncs.c module is
805        replaced by pcre_ucd.c.
806    
807    
808    Version 7.7 07-May-08
809    ---------------------
810    
811    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
812        a string to a long long, pretend we don't even have a long long." This is
813        done by checking for the strtoq, strtoll, and _strtoi64 functions.
814    
815    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
816        pre-7.6 versions, which defined a global no_arg variable instead of putting
817        it in the RE class. (See also #8 below.)
818    
819    3.  Remove a line of dead code, identified by coverity and reported by Nuno
820        Lopes.
821    
822    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
823    
824        (1) The include/exclude patterns were being applied to the whole pathnames
825            of files, instead of just to the final components.
826    
827        (2) If there was more than one level of directory, the subdirectories were
828            skipped unless they satisfied the include/exclude conditions. This is
829            inconsistent with GNU grep (and could even be seen as contrary to the
830            pcregrep specification - which I improved to make it absolutely clear).
831            The action now is always to scan all levels of directory, and just
832            apply the include/exclude patterns to regular files.
833    
834    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
835        --exclude_dir in the tests to avoid scanning .svn directories.
836    
837    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
838        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
839        doesn't support NULs in patterns.
840    
841    7.  Added some missing "const"s to declarations of static tables in
842        pcre_compile.c and pcre_dfa_exec.c.
843    
844    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
845        caused by fix #2  above. (Subsequently also a second patch to fix the
846        first patch. And a third patch - this was a messy problem.)
847    
848    9.  Applied Craig's patch to remove the use of push_back().
849    
850    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
851        matching function regexec().
852    
853    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
854        which, however, unlike Perl's \g{...}, are subroutine calls, not back
855        references. PCRE supports relative numbers with this syntax (I don't think
856        Oniguruma does).
857    
858    12. Previously, a group with a zero repeat such as (...){0} was completely
859        omitted from the compiled regex. However, this means that if the group
860        was called as a subroutine from elsewhere in the pattern, things went wrong
861        (an internal error was given). Such groups are now left in the compiled
862        pattern, with a new opcode that causes them to be skipped at execution
863        time.
864    
865    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
866        to the way PCRE behaves:
867    
868        (a) A lone ] character is dis-allowed (Perl treats it as data).
869    
870        (b) A back reference to an unmatched subpattern matches an empty string
871            (Perl fails the current match path).
872    
873        (c) A data ] in a character class must be notated as \] because if the
874            first data character in a class is ], it defines an empty class. (In
875            Perl it is not possible to have an empty class.) The empty class []
876            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
877            The negative empty class [^] matches any one character, independently
878            of the DOTALL setting.
879    
880    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
881        non-existent subpattern following a character class starting with ']' and
882        containing () gave an internal compiling error instead of "reference to
883        non-existent subpattern". Fortunately, when the pattern did exist, the
884        compiled code was correct. (When scanning forwards to check for the
885        existencd of the subpattern, it was treating the data ']' as terminating
886        the class, so got the count wrong. When actually compiling, the reference
887        was subsequently set up correctly.)
888    
889    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
890        it was being rejected as not supported by pcre_dfa_exec(), even though
891        other assertions are supported. I have made pcre_dfa_exec() support
892        (*FAIL).
893    
894    16. The implementation of 13c above involved the invention of a new opcode,
895        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
896        cannot be changed at match time, I realized I could make a small
897        improvement to matching performance by compiling OP_ALLANY instead of
898        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
899        on the OP_ANY path.
900    
901    17. Compiling pcretest on Windows with readline support failed without the
902        following two fixes: (1) Make the unistd.h include conditional on
903        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
904    
905    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
906        ncurses library to be included for pcretest when ReadLine support is
907        requested, but also to allow for it to be overridden. This patch came from
908        Daniel Bergström.
909    
910    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
911        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
912        any errors with the current Unicode tables. Thanks to Peter Kankowski for
913        spotting this.
914    
915    
916    Version 7.6 28-Jan-08
917    ---------------------
918    
919    1.  A character class containing a very large number of characters with
920        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
921        overflow.
922    
923    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
924        HAVE_LONG_LONG is not defined.
925    
926    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
927        bring them up to date and include new features. This patch includes:
928    
929        - Fixed PH's badly added libz and libbz2 support.
930        - Fixed a problem with static linking.
931        - Added pcredemo. [But later removed - see 7 below.]
932        - Fixed dftables problem and added an option.
933        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
934            HAVE_LONG_LONG.
935        - Added readline support for pcretest.
936        - Added an listing of the option settings after cmake has run.
937    
938    4.  A user submitted a patch to Makefile that makes it easy to create
939        "pcre.dll" under mingw when using Configure/Make. I added stuff to
940        Makefile.am that cause it to include this special target, without
941        affecting anything else. Note that the same mingw target plus all
942        the other distribution libraries and programs are now supported
943        when configuring with CMake (see 6 below) instead of with
944        Configure/Make.
945    
946    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
947        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
948        exported in the Windows port". It has not yet been confirmed that the patch
949        solves the problem, but it does no harm.
950    
951    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
952        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
953        with CMake, and also correct the comment about stack recursion.
954    
955    7.  Remove the automatic building of pcredemo from the ./configure system and
956        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
957        of a program that users should build themselves after PCRE is installed, so
958        building it automatically is not really right. What is more, it gave
959        trouble in some build environments.
960    
961    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
962    
963    
964    Version 7.5 10-Jan-08
965    ---------------------
966    
967    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
968        values in parens when parsing an RE using the C++ wrapper."
969    
970    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
971        Characters greater than 255 were excluded from the class instead of being
972        included.
973    
974    3.  The same bug as (2) above applied to negated POSIX classes such as
975        [:^space:].
976    
977    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
978        defined or documented. It seems to have been a typo for PCRE_STATIC, so
979        I have changed it.
980    
981    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
982        first named subpattern) and a construct such as (?&a) would reference the
983        first named subpattern whose name started with "a" (in other words, the
984        length check was missing). Both these problems are fixed. "Subpattern name
985        expected" is now given for (?&) (a zero-length name), and this patch also
986        makes it give the same error for \k'' (previously it complained that that
987        was a reference to a non-existent subpattern).
988    
989    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
990        this is right because (?- can be followed by option settings as well as by
991        digits. I have, however, made the messages clearer.
992    
993    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
994        than the number used in the conditional) now cause a compile-time error.
995        This is actually not compatible with Perl, which accepts such patterns, but
996        treats the conditional as always being FALSE (as PCRE used to), but it
997        seems to me that giving a diagnostic is better.
998    
999    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1000        and messages.
1001    
1002    9.  Fix two occurrences of "backslash" in comments that should have been
1003        "backspace".
1004    
1005    10. Remove two redundant lines of code that can never be obeyed (their function
1006        was moved elsewhere).
1007    
1008    11. The program that makes PCRE's Unicode character property table had a bug
1009        which caused it to generate incorrect table entries for sequences of
1010        characters that have the same character type, but are in different scripts.
1011        It amalgamated them into a single range, with the script of the first of
1012        them. In other words, some characters were in the wrong script. There were
1013        thirteen such cases, affecting characters in the following ranges:
1014    
1015          U+002b0 - U+002c1
1016          U+0060c - U+0060d
1017          U+0061e - U+00612
1018          U+0064b - U+0065e
1019          U+0074d - U+0076d
1020          U+01800 - U+01805
1021          U+01d00 - U+01d77
1022          U+01d9b - U+01dbf
1023          U+0200b - U+0200f
1024          U+030fc - U+030fe
1025          U+03260 - U+0327f
1026          U+0fb46 - U+0fbb1
1027          U+10450 - U+1049d
1028    
1029    12. The -o option (show only the matching part of a line) for pcregrep was not
1030        compatible with GNU grep in that, if there was more than one match in a
1031        line, it showed only the first of them. It now behaves in the same way as
1032        GNU grep.
1033    
1034    13. If the -o and -v options were combined for pcregrep, it printed a blank
1035        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1036        does the same. The return code can be used to tell if there were any
1037        non-matching lines.
1038    
1039    14. Added --file-offsets and --line-offsets to pcregrep.
1040    
1041    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1042        infinitely looping recursion. The bug was that positive lookaheads were not
1043        being skipped when checking for a possible empty match (negative lookaheads
1044        and both kinds of lookbehind were skipped).
1045    
1046    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1047        inclusion of <windows.h> to before rather than after the definition of
1048        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1049    
1050    17. Specifying a possessive quantifier with a specific limit for a Unicode
1051        character property caused pcre_compile() to compile bad code, which led at
1052        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1053        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1054        caused the error; without that there was no problem.
1055    
1056    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1057    
1058    19. Added --enable-pcretest-libreadline.
1059    
1060    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1061        RE::GlobalReplace(). As a result, the number of replacements returned was
1062        double what it should be. I removed one of the increments, but Craig sent a
1063        later patch that removed the other one (the right fix) and added unit tests
1064        that check the return values (which was not done before).
1065    
1066    21. Several CMake things:
1067    
1068        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1069            the names libpcre and libpcreposix, not just pcre and pcreposix.
1070    
1071        (2) The above change means that pcretest and pcregrep are now correctly
1072            linked with the newly-built libraries, not previously installed ones.
1073    
1074        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1075    
1076    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1077        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1078        UTF-8 newline character). The key issue is that the pattern starts .*;
1079        this means that the match must be either at the beginning, or after a
1080        newline. The bug was in the code for advancing after a failed match and
1081        checking that the new position followed a newline. It was not taking
1082        account of UTF-8 characters correctly.
1083    
1084    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1085        character classes. PCRE was not treating the sequence [:...:] as a
1086        character class unless the ... were all letters. Perl, however, seems to
1087        allow any characters between [: and :], though of course it rejects as
1088        unknown any "names" that contain non-letters, because all the known class
1089        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1090        for example, whereas PCRE did not - it did not recognize a POSIX character
1091        class. This seemed a bit dangerous, so the code has been changed to be
1092        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1093        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1094        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1095        Perl does, and where it didn't before.
1096    
1097    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1098        Windows environments %n is disabled by default.
1099    
1100    
1101    Version 7.4 21-Sep-07
1102    ---------------------
1103    
1104    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1105        means that a class such as [\s] counted as "explicit reference to CR or
1106        LF". That isn't really right - the whole point of the change was to try to
1107        help when there was an actual mention of one of the two characters. So now
1108        the change happens only if \r or \n (or a literal CR or LF) character is
1109        encountered.
1110    
1111    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1112        of both had grown to the point where there were only 3 bits left.
1113        Fortunately, there was spare space in the data structure, and so I have
1114        moved the internal flags into a new 16-bit field to free up more option
1115        bits.
1116    
1117    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1118        but did not set the internal JCHANGED flag - either of these is enough to
1119        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1120        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1121        start sets both bits.
1122    
1123    4.  Added options (at build time, compile time, exec time) to change \R from
1124        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1125    
1126    5.  doc/pcresyntax.html was missing from the distribution.
1127    
1128    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1129        compatibility, even though it is no longer used.
1130    
1131    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1132        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1133        windows.h file is present (where different names are used). [This was
1134        reversed later after testing - see 16 below.]
1135    
1136    8.  Changed all #include <config.h> to #include "config.h". There were also
1137        some further <pcre.h> cases that I changed to "pcre.h".
1138    
1139    9.  When pcregrep was used with the --colour option, it missed the line ending
1140        sequence off the lines that it output.
1141    
1142    10. It was pointed out to me that arrays of string pointers cause lots of
1143        relocations when a shared library is dynamically loaded. A technique of
1144        using a single long string with a table of offsets can drastically reduce
1145        these. I have refactored PCRE in four places to do this. The result is
1146        dramatic:
1147    
1148          Originally:                          290
1149          After changing UCP table:            187
1150          After changing error message table:   43
1151          After changing table of "verbs"       36
1152          After changing table of Posix names   22
1153    
1154        Thanks to the folks working on Gregex for glib for this insight.
1155    
1156    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1157        unicode-properties was also set.
1158    
1159    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1160    
1161    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1162        checked only for CRLF.
1163    
1164    14. Added casts to pcretest.c to avoid compiler warnings.
1165    
1166    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1167    
1168    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1169        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1170        entirely. This removes changes made in 7 above.
1171    
1172    17. The CMake files have been updated, and there is now more information about
1173        building with CMake in the NON-UNIX-USE document.
1174    
1175    
1176    Version 7.3 28-Aug-07
1177  ---------------------  ---------------------
1178    
1179   1. In the rejigging of the build system that eventually resulted in 7.1, the   1. In the rejigging of the build system that eventually resulted in 7.1, the
# Line 29  Version 7.3 05-Jul-07 Line 1201  Version 7.3 05-Jul-07
1201   4. Debugging: adjusted the names of opcodes for different kinds of parentheses   4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1202      in debug output.      in debug output.
1203    
1204   5. Arrange to use "%I64d" instead of "%lld" for long printing in the pcrecpp   5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1205      unittest when running under MinGW.      long printing in the pcrecpp unittest when running under MinGW.
1206    
1207   6. ESC_K was left out of the EBCDIC table.   6. ESC_K was left out of the EBCDIC table.
1208    
# Line 48  Version 7.3 05-Jul-07 Line 1220  Version 7.3 05-Jul-07
1220      from the 1000-entry workspace that had been reserved. So now there is no      from the 1000-entry workspace that had been reserved. So now there is no
1221      explicit limit, but more stack is used.      explicit limit, but more stack is used.
1222    
1223     8. Applied Daniel's patches to solve problems with the import/export magic
1224        syntax that is required for Windows, and which was going wrong for the
1225        pcreposix and pcrecpp parts of the library. These were overlooked when this
1226        problem was solved for the main library.
1227    
1228     9. There were some crude static tests to avoid integer overflow when computing
1229        the size of patterns that contain repeated groups with explicit upper
1230        limits. As the maximum quantifier is 65535, the maximum group length was
1231        set at 30,000 so that the product of these two numbers did not overflow a
1232        32-bit integer. However, it turns out that people want to use groups that
1233        are longer than 30,000 bytes (though not repeat them that many times).
1234        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1235        made it possible to implement the integer overflow checks in a much more
1236        dynamic way, which I have now done. The artificial limitation on group
1237        length has been removed - we now have only the limit on the total length of
1238        the compiled pattern, which depends on the LINK_SIZE setting.
1239    
1240    10. Fixed a bug in the documentation for get/copy named substring when
1241        duplicate names are permitted. If none of the named substrings are set, the
1242        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1243        empty string.
1244    
1245    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1246        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1247        because the ] is interpreted as the first data character and the
1248        terminating ] is not found. PCRE has been made compatible with Perl in this
1249        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1250        cause memory overwriting.
1251    
1252    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1253        string has been matched (to stop an infinite loop). It was not recognizing
1254        a conditional subpattern that could match an empty string if that
1255        subpattern was within another subpattern. For example, it looped when
1256        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1257        condition was not nested. This bug has been fixed.
1258    
1259    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1260        past the start of the subject in the presence of bytes with the top bit
1261        set, for example "\x8aBCD".
1262    
1263    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1264        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1265    
1266    14. Optimized (?!) to (*FAIL).
1267    
1268    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1269        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1270        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1271        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1272        does: it's just the validity check that is more restrictive.
1273    
1274    16. Inserted checks for integer overflows during escape sequence (backslash)
1275        processing, and also fixed erroneous offset values for syntax errors during
1276        backslash processing.
1277    
1278    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1279        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1280    
1281    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1282        caused an overrun.
1283    
1284    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1285        something other than just ASCII characters) inside a group that had an
1286        unlimited repeat caused a loop at compile time (while checking to see
1287        whether the group could match an empty string).
1288    
1289    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1290        [\P{Any}] did so. (Error in the code for printing property names.)
1291    
1292    21. An orphan \E inside a character class could cause a crash.
1293    
1294    22. A repeated capturing bracket such as (A)? could cause a wild memory
1295        reference during compilation.
1296    
1297    23. There are several functions in pcre_compile() that scan along a compiled
1298        expression for various reasons (e.g. to see if it's fixed length for look
1299        behind). There were bugs in these functions when a repeated \p or \P was
1300        present in the pattern. These operators have additional parameters compared
1301        with \d, etc, and these were not being taken into account when moving along
1302        the compiled data. Specifically:
1303    
1304        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1305            length.
1306    
1307        (b) An item such as \pL+ within a repeated group could cause crashes or
1308            loops.
1309    
1310        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1311            "reference to non-existent subpattern" error.
1312    
1313        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1314    
1315    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1316        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1317    
1318    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1319    
1320    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1321        character were causing crashes (broken optimization).
1322    
1323    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1324        \p or \P) caused a compile-time loop.
1325    
1326    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1327        break. For example, the unstudied pattern [\r\n]A does not match the string
1328        "\r\nA" because change 7.0/46 below moves the current point on by two
1329        characters after failing to match at the start. However, the pattern \nA
1330        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1331        the same is true. There doesn't seem any very clean way out of this, but
1332        what I have chosen to do makes the common cases work: PCRE now takes note
1333        of whether there can be an explicit match for \r or \n anywhere in the
1334        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1335        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1336        pattern has explicit CR or LF references.
1337    
1338    29. Added (*CR) etc for changing newline setting at start of pattern.
1339    
1340    
1341  Version 7.2 19-Jun-07  Version 7.2 19-Jun-07
1342  ---------------------  ---------------------

Legend:
Removed from v.197  
changed lines
  Added in v.586

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12