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

Diff of /code/trunk/ChangeLog

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

revision 5 by nigel, Sat Feb 24 21:38:05 2007 UTC revision 501 by ph10, Sun Mar 7 11:49:54 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.02 01-Mar-2010
5    ------------------------
6    
7    1.  The Unicode data tables have been updated to Unicode 5.2.0.
8    
9    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
10        configured.
11    
12    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
13        original author of that file, following a query about its status.
14    
15    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
16        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
17    
18    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
19        quantifier applied to a forward-referencing subroutine call, could compile
20        incorrect code or give the error "internal error: previously-checked
21        referenced subpattern not found".
22    
23    6.  Both MS Visual Studio and Symbian OS have problems with initializing
24        variables to point to external functions. For these systems, therefore,
25        pcre_malloc etc. are now initialized to local functions that call the
26        relevant global functions.
27    
28    7.  There were two entries missing in the vectors called coptable and poptable
29        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
30        I've fixed the data, and added a kludgy way of testing at compile time that
31        the lengths are correct (equal to the number of opcodes).
32    
33    8.  Following on from 7, I added a similar kludge to check the length of the
34        eint vector in pcreposix.c.
35    
36    9.  Error texts for pcre_compile() are held as one long string to avoid too
37        much relocation at load time. To find a text, the string is searched,
38        counting zeros. There was no check for running off the end of the string,
39        which could happen if a new error number was added without updating the
40        string.
41    
42    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
43    
44    11. \K was not working if it appeared in an atomic group or in a group that
45        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
46        \K is "not well defined" if used in an assertion. PCRE now accepts it if
47        the assertion is positive, but not if it is negative.
48    
49    12. Change 11 fortuitously reduced the size of the stack frame used in the
50        "match()" function of pcre_exec.c by one pointer. Forthcoming
51        implementation of support for (*MARK) will need an extra pointer on the
52        stack; I have reserved it now, so that the stack frame size does not
53        decrease.
54    
55    
56    Version 8.01 19-Jan-2010
57    ------------------------
58    
59    1.  If a pattern contained a conditional subpattern with only one branch (in
60        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
61        computed the wrong minimum data length (which is of course zero for such
62        subpatterns). This could cause incorrect "no match" results.
63    
64    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
65        the pattern is reset in the first branch, pcre_compile() failed with
66        "internal error: code overflow at offset...". This happened only when
67        the reset was to the original external option setting. (An optimization
68        abstracts leading options settings into an external setting, which was the
69        cause of this.)
70    
71    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
72        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
73        assertion pattern did not match (meaning that the assertion was true), it
74        was incorrectly treated as false if the SKIP had been reached during the
75        matching. This also applied to assertions used as conditions.
76    
77    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
78        assertion subpattern, including such a pattern used as a condition,
79        unpredictable results occurred, instead of the error return
80        PCRE_ERROR_DFA_UITEM.
81    
82    5.  The C++ GlobalReplace function was not working like Perl for the special
83        situation when an empty string is matched. It now does the fancy magic
84        stuff that is necessary.
85    
86    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
87        removed. (These were left over from very, very early versions of PCRE.)
88    
89    7.  Some cosmetic changes to the code to make life easier when compiling it
90        as part of something else:
91    
92        (a) Change DEBUG to PCRE_DEBUG.
93    
94        (b) In pcre_compile(), rename the member of the "branch_chain" structure
95            called "current" as "current_branch", to prevent a collision with the
96            Linux macro when compiled as a kernel module.
97    
98        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
99            prevent a collision with the Linux macro when compiled as a kernel
100            module.
101    
102    8.  In pcre_compile() there are some checks for integer overflows that used to
103        cast potentially large values to (double). This has been changed to that
104        when building, a check for int64_t is made, and if it is found, it is used
105        instead, thus avoiding the use of floating point arithmetic. (There is no
106        other use of FP in PCRE.) If int64_t is not found, the fallback is to
107        double.
108    
109    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
110        2005 (difference between two addresses compared to an unsigned value).
111    
112    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
113        custom one, because of the following reported problem in Windows:
114    
115          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
116              under Win32.
117          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
118              therefore missing the function definition.
119          - The compiler thus generates a "C" signature for the test function.
120          - The linker fails to find the "C" function.
121          - PCRE fails to configure if asked to do so against libbz2.
122    
123    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
124        messages were output:
125    
126          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
127          rerunning libtoolize, to keep the correct libtool macros in-tree.
128          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
129    
130        I have done both of these things.
131    
132    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
133        most of the time, it *can* run out if it is given a pattern that contains a
134        runaway infinite recursion. I updated the discussion in the pcrestack man
135        page.
136    
137    13. Now that we have gone to the x.xx style of version numbers, the minor
138        version may start with zero. Using 08 or 09 is a bad idea because users
139        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
140        interpreted as invalid octal numbers. I've updated the previous comment in
141        configure.ac, and also added a check that gives an error if 08 or 09 are
142        used.
143    
144    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
145        causing partial matching to fail when the end of the subject matched \W
146        in a UTF-8 pattern where \W was quantified with a minimum of 3.
147    
148    15. There were some discrepancies between the declarations in pcre_internal.h
149        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
150        their definitions. The declarations used "const uschar *" and the
151        definitions used USPTR. Even though USPTR is normally defined as "const
152        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
153        reported that: "This difference in casting confuses some C++ compilers, for
154        example, SunCC recognizes above declarations as different functions and
155        generates broken code for hbpcre." I have changed the declarations to use
156        USPTR.
157    
158    16. GNU libtool is named differently on some systems. The autogen.sh script now
159        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
160        (FreeBSD).
161    
162    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
163        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
164        comment: "Figure out how to create a longlong from a string: strtoll and
165        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
166        instance, but it only takes 2 args instead of 3!"
167    
168    18. A subtle bug concerned with back references has been fixed by a change of
169        specification, with a corresponding code fix. A pattern such as
170        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
171        refers, was giving matches when it shouldn't. For example, xa=xaaa would
172        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
173        same bug. Such groups have to be quantified to be useful, or contained
174        inside another quantified group. (If there's no repetition, the reference
175        can never match.) The problem arises because, having left the group and
176        moved on to the rest of the pattern, a later failure that backtracks into
177        the group uses the captured value from the final iteration of the group
178        rather than the correct earlier one. I have fixed this in PCRE by forcing
179        any group that contains a reference to itself to be an atomic group; that
180        is, there cannot be any backtracking into it once it has completed. This is
181        similar to recursive and subroutine calls.
182    
183    
184    Version 8.00 19-Oct-09
185    ----------------------
186    
187    1.  The table for translating pcre_compile() error codes into POSIX error codes
188        was out-of-date, and there was no check on the pcre_compile() error code
189        being within the table. This could lead to an OK return being given in
190        error.
191    
192    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
193        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
194        in a Windows environment.
195    
196    3.  The pcregrep --count option prints the count for each file even when it is
197        zero, as does GNU grep. However, pcregrep was also printing all files when
198        --files-with-matches was added. Now, when both options are given, it prints
199        counts only for those files that have at least one match. (GNU grep just
200        prints the file name in this circumstance, but including the count seems
201        more useful - otherwise, why use --count?) Also ensured that the
202        combination -clh just lists non-zero counts, with no names.
203    
204    4.  The long form of the pcregrep -F option was incorrectly implemented as
205        --fixed_strings instead of --fixed-strings. This is an incompatible change,
206        but it seems right to fix it, and I didn't think it was worth preserving
207        the old behaviour.
208    
209    5.  The command line items --regex=pattern and --regexp=pattern were not
210        recognized by pcregrep, which required --regex pattern or --regexp pattern
211        (with a space rather than an '='). The man page documented the '=' forms,
212        which are compatible with GNU grep; these now work.
213    
214    6.  No libpcreposix.pc file was created for pkg-config; there was just
215        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
216    
217    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
218        when UCP support is not needed, by modifying the Python script that
219        generates it from Unicode data files. This should not matter if the module
220        is correctly used as a library, but I received one complaint about 50K of
221        unwanted data. My guess is that the person linked everything into his
222        program rather than using a library. Anyway, it does no harm.
223    
224    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
225        was a minimum greater than 1 for a wide character in a possessive
226        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
227        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
228        character. Chaos in the form of incorrect output or a compiling loop could
229        result.
230    
231    9.  The restrictions on what a pattern can contain when partial matching is
232        requested for pcre_exec() have been removed. All patterns can now be
233        partially matched by this function. In addition, if there are at least two
234        slots in the offset vector, the offset of the earliest inspected character
235        for the match and the offset of the end of the subject are set in them when
236        PCRE_ERROR_PARTIAL is returned.
237    
238    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
239        synonymous with PCRE_PARTIAL, for backwards compatibility, and
240        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
241        and may be more useful for multi-segment matching.
242    
243    11. Partial matching with pcre_exec() is now more intuitive. A partial match
244        used to be given if ever the end of the subject was reached; now it is
245        given only if matching could not proceed because another character was
246        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
247        string "Z", which now yields "no match" instead of "partial match". In the
248        case of pcre_dfa_exec(), "no match" is given if every matching path for the
249        final character ended with (*FAIL).
250    
251    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
252        if the pattern had a "must contain" character that was already found in the
253        earlier partial match, unless partial matching was again requested. For
254        example, with the pattern /dog.(body)?/, the "must contain" character is
255        "g". If the first part-match was for the string "dog", restarting with
256        "sbody" failed. This bug has been fixed.
257    
258    13. The string returned by pcre_dfa_exec() after a partial match has been
259        changed so that it starts at the first inspected character rather than the
260        first character of the match. This makes a difference only if the pattern
261        starts with a lookbehind assertion or \b or \B (\K is not supported by
262        pcre_dfa_exec()). It's an incompatible change, but it makes the two
263        matching functions compatible, and I think it's the right thing to do.
264    
265    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
266        so that the demonstration program is easily available in environments where
267        PCRE has not been installed from source.
268    
269    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
270        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
271        library.
272    
273    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
274        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
275        is not the first non-POSIX option to be added. Clearly some people find
276        these options useful.
277    
278    17. If a caller to the POSIX matching function regexec() passes a non-zero
279        value for nmatch with a NULL value for pmatch, the value of
280        nmatch is forced to zero.
281    
282    18. RunGrepTest did not have a test for the availability of the -u option of
283        the diff command, as RunTest does. It now checks in the same way as
284        RunTest, and also checks for the -b option.
285    
286    19. If an odd number of negated classes containing just a single character
287        interposed, within parentheses, between a forward reference to a named
288        subpattern and the definition of the subpattern, compilation crashed with
289        an internal error, complaining that it could not find the referenced
290        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
291        [The bug was that it was starting one character too far in when skipping
292        over the character class, thus treating the ] as data rather than
293        terminating the class. This meant it could skip too much.]
294    
295    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
296        /g option in pcretest when the pattern contains \K, which makes it possible
297        to have an empty string match not at the start, even when the pattern is
298        anchored. Updated pcretest and pcredemo to use this option.
299    
300    21. If the maximum number of capturing subpatterns in a recursion was greater
301        than the maximum at the outer level, the higher number was returned, but
302        with unset values at the outer level. The correct (outer level) value is
303        now given.
304    
305    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
306        PCRE did not set those parentheses (unlike Perl). I have now found a way to
307        make it do so. The string so far is captured, making this feature
308        compatible with Perl.
309    
310    23. The tests have been re-organized, adding tests 11 and 12, to make it
311        possible to check the Perl 5.10 features against Perl 5.10.
312    
313    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
314        pattern matches a fixed length string. PCRE did not allow this; now it
315        does. Neither allows recursion.
316    
317    25. I finally figured out how to implement a request to provide the minimum
318        length of subject string that was needed in order to match a given pattern.
319        (It was back references and recursion that I had previously got hung up
320        on.) This code has now been added to pcre_study(); it finds a lower bound
321        to the length of subject needed. It is not necessarily the greatest lower
322        bound, but using it to avoid searching strings that are too short does give
323        some useful speed-ups. The value is available to calling programs via
324        pcre_fullinfo().
325    
326    26. While implementing 25, I discovered to my embarrassment that pcretest had
327        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
328        study optimizations had never been tested with that matching function.
329        Oops. What is worse, even when it was passed study data, there was a bug in
330        pcre_dfa_exec() that meant it never actually used it. Double oops. There
331        were also very few tests of studied patterns with pcre_dfa_exec().
332    
333    27. If (?| is used to create subpatterns with duplicate numbers, they are now
334        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
335        on the other side of the coin, they are no longer allowed to have different
336        names, because these cannot be distinguished in PCRE, and this has caused
337        confusion. (This is a difference from Perl.)
338    
339    28. When duplicate subpattern names are present (necessarily with different
340        numbers, as required by 27 above), and a test is made by name in a
341        conditional pattern, either for a subpattern having been matched, or for
342        recursion in such a pattern, all the associated numbered subpatterns are
343        tested, and the overall condition is true if the condition is true for any
344        one of them. This is the way Perl works, and is also more like the way
345        testing by number works.
346    
347    
348    Version 7.9 11-Apr-09
349    ---------------------
350    
351    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
352        (pcretest), all targets were linked against these libraries. This included
353        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
354        libraries. This caused unwanted dependencies to be created. This problem
355        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
356        pcretest is linked with readline.
357    
358    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
359        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
360        moved outside it again, because FALSE and TRUE are already defined in AIX,
361        but BOOL is not.
362    
363    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
364        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
365    
366    4.  The pcregrep documentation said spaces were inserted as well as colons (or
367        hyphens) following file names and line numbers when outputting matching
368        lines. This is not true; no spaces are inserted. I have also clarified the
369        wording for the --colour (or --color) option.
370    
371    5.  In pcregrep, when --colour was used with -o, the list of matching strings
372        was not coloured; this is different to GNU grep, so I have changed it to be
373        the same.
374    
375    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
376        each matching line was coloured. Now it goes on to look for further matches
377        of any of the test patterns, which is the same behaviour as GNU grep.
378    
379    7.  A pattern that could match an empty string could cause pcregrep to loop; it
380        doesn't make sense to accept an empty string match in pcregrep, so I have
381        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
382        seems to be how GNU grep behaves.
383    
384    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
385        start or after a newline", because the conditional assertion was not being
386        correctly handled. The rule now is that both the assertion and what follows
387        in the first alternative must satisfy the test.
388    
389    9.  If auto-callout was enabled in a pattern with a conditional group whose
390        condition was an assertion, PCRE could crash during matching, both with
391        pcre_exec() and pcre_dfa_exec().
392    
393    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
394        used for matching.
395    
396    11. Unicode property support in character classes was not working for
397        characters (bytes) greater than 127 when not in UTF-8 mode.
398    
399    12. Added the -M command line option to pcretest.
400    
401    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
402    
403    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
404    
405    16. Added comments and documentation about mis-use of no_arg in the C++
406        wrapper.
407    
408    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
409        from Martin Jerabek that uses macro names for all relevant character and
410        string constants.
411    
412    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
413        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
414        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
415        these, but not everybody uses configure.
416    
417    19. A conditional group that had only one branch was not being correctly
418        recognized as an item that could match an empty string. This meant that an
419        enclosing group might also not be so recognized, causing infinite looping
420        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
421        with the subject "ab", where knowledge that the repeated group can match
422        nothing is needed in order to break the loop.
423    
424    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
425        exec(), but without supplying a callout function, matching went wrong.
426    
427    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
428        leak if the size of the offset vector was greater than 30. When the vector
429        is smaller, the saved offsets during recursion go onto a local stack
430        vector, but for larger vectors malloc() is used. It was failing to free
431        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
432        error, in fact).
433    
434    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
435        heapframe that is used only when UTF-8 support is enabled. This caused no
436        problem, but was untidy.
437    
438    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
439        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
440        included within another project.
441    
442    24. Steven Van Ingelgem's patches to add more options to the CMake support,
443        slightly modified by me:
444    
445          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
446              not building pcregrep.
447    
448          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
449              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
450    
451    25. Forward references, both numeric and by name, in patterns that made use of
452        duplicate group numbers, could behave incorrectly or give incorrect errors,
453        because when scanning forward to find the reference group, PCRE was not
454        taking into account the duplicate group numbers. A pattern such as
455        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
456    
457    26. Changed a few more instances of "const unsigned char *" to USPTR, making
458        the feature of a custom pointer more persuasive (as requested by a user).
459    
460    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
461        pcretest.c, inside #ifndefs, because it seems they are sometimes already
462        pre-defined.
463    
464    28. Added support for (*UTF8) at the start of a pattern.
465    
466    29. Arrange for flags added by the "release type" setting in CMake to be shown
467        in the configuration summary.
468    
469    
470    Version 7.8 05-Sep-08
471    ---------------------
472    
473    1.  Replaced UCP searching code with optimized version as implemented for Ad
474        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
475        stage table and inline lookup instead of a function, giving speed ups of 2
476        to 5 times on some simple patterns that I tested. Permission was given to
477        distribute the MultiStage2.py script that generates the tables (it's not in
478        the tarball, but is in the Subversion repository).
479    
480    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
481        scripts.
482    
483    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
484        a group with a zero qualifier. The result of the study could be incorrect,
485        or the function might crash, depending on the pattern.
486    
487    4.  Caseless matching was not working for non-ASCII characters in back
488        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
489        It now works when Unicode Property Support is available.
490    
491    5.  In pcretest, an escape such as \x{de} in the data was always generating
492        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
493        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
494        truncation.
495    
496    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
497    
498    7.  Added two (int) casts to pcregrep when printing the difference of two
499        pointers, in case they are 64-bit values.
500    
501    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
502        test 2 if it fails.
503    
504    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
505        and a #define of that name to empty if it is not externally set. This is to
506        allow users of MSVC to set it if necessary.
507    
508    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
509        the convenience functions in the pcre_get.c source file.
510    
511    11. An option change at the start of a pattern that had top-level alternatives
512        could cause overwriting and/or a crash. This command provoked a crash in
513        some environments:
514    
515          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
516    
517        This potential security problem was recorded as CVE-2008-2371.
518    
519    12. For a pattern where the match had to start at the beginning or immediately
520        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
521        pcre_dfa_exec() could read past the end of the passed subject if there was
522        no match. To help with detecting such bugs (e.g. with valgrind), I modified
523        pcretest so that it places the subject at the end of its malloc-ed buffer.
524    
525    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
526        exec() might read past the end of the data buffer in UTF-8 mode.
527    
528    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
529        the data contained the byte 0x85 as part of a UTF-8 character within its
530        first line. This applied both to normal and DFA matching.
531    
532    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
533        /^[^d]*?$/8 failed to match "abc".
534    
535    16. Added a missing copyright notice to pcrecpp_internal.h.
536    
537    17. Make it more clear in the documentation that values returned from
538        pcre_exec() in ovector are byte offsets, not character counts.
539    
540    18. Tidied a few places to stop certain compilers from issuing warnings.
541    
542    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
543        supplied by Stefan Weber. I made a further small update for 7.8 because
544        there is a change of source arrangements: the pcre_searchfuncs.c module is
545        replaced by pcre_ucd.c.
546    
547    
548    Version 7.7 07-May-08
549    ---------------------
550    
551    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
552        a string to a long long, pretend we don't even have a long long." This is
553        done by checking for the strtoq, strtoll, and _strtoi64 functions.
554    
555    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
556        pre-7.6 versions, which defined a global no_arg variable instead of putting
557        it in the RE class. (See also #8 below.)
558    
559    3.  Remove a line of dead code, identified by coverity and reported by Nuno
560        Lopes.
561    
562    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
563    
564        (1) The include/exclude patterns were being applied to the whole pathnames
565            of files, instead of just to the final components.
566    
567        (2) If there was more than one level of directory, the subdirectories were
568            skipped unless they satisfied the include/exclude conditions. This is
569            inconsistent with GNU grep (and could even be seen as contrary to the
570            pcregrep specification - which I improved to make it absolutely clear).
571            The action now is always to scan all levels of directory, and just
572            apply the include/exclude patterns to regular files.
573    
574    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
575        --exclude_dir in the tests to avoid scanning .svn directories.
576    
577    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
578        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
579        doesn't support NULs in patterns.
580    
581    7.  Added some missing "const"s to declarations of static tables in
582        pcre_compile.c and pcre_dfa_exec.c.
583    
584    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
585        caused by fix #2  above. (Subsequently also a second patch to fix the
586        first patch. And a third patch - this was a messy problem.)
587    
588    9.  Applied Craig's patch to remove the use of push_back().
589    
590    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
591        matching function regexec().
592    
593    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
594        which, however, unlike Perl's \g{...}, are subroutine calls, not back
595        references. PCRE supports relative numbers with this syntax (I don't think
596        Oniguruma does).
597    
598    12. Previously, a group with a zero repeat such as (...){0} was completely
599        omitted from the compiled regex. However, this means that if the group
600        was called as a subroutine from elsewhere in the pattern, things went wrong
601        (an internal error was given). Such groups are now left in the compiled
602        pattern, with a new opcode that causes them to be skipped at execution
603        time.
604    
605    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
606        to the way PCRE behaves:
607    
608        (a) A lone ] character is dis-allowed (Perl treats it as data).
609    
610        (b) A back reference to an unmatched subpattern matches an empty string
611            (Perl fails the current match path).
612    
613        (c) A data ] in a character class must be notated as \] because if the
614            first data character in a class is ], it defines an empty class. (In
615            Perl it is not possible to have an empty class.) The empty class []
616            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
617            The negative empty class [^] matches any one character, independently
618            of the DOTALL setting.
619    
620    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
621        non-existent subpattern following a character class starting with ']' and
622        containing () gave an internal compiling error instead of "reference to
623        non-existent subpattern". Fortunately, when the pattern did exist, the
624        compiled code was correct. (When scanning forwards to check for the
625        existencd of the subpattern, it was treating the data ']' as terminating
626        the class, so got the count wrong. When actually compiling, the reference
627        was subsequently set up correctly.)
628    
629    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
630        it was being rejected as not supported by pcre_dfa_exec(), even though
631        other assertions are supported. I have made pcre_dfa_exec() support
632        (*FAIL).
633    
634    16. The implementation of 13c above involved the invention of a new opcode,
635        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
636        cannot be changed at match time, I realized I could make a small
637        improvement to matching performance by compiling OP_ALLANY instead of
638        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
639        on the OP_ANY path.
640    
641    17. Compiling pcretest on Windows with readline support failed without the
642        following two fixes: (1) Make the unistd.h include conditional on
643        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
644    
645    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
646        ncurses library to be included for pcretest when ReadLine support is
647        requested, but also to allow for it to be overridden. This patch came from
648        Daniel Bergström.
649    
650    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
651        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
652        any errors with the current Unicode tables. Thanks to Peter Kankowski for
653        spotting this.
654    
655    
656    Version 7.6 28-Jan-08
657    ---------------------
658    
659    1.  A character class containing a very large number of characters with
660        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
661        overflow.
662    
663    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
664        HAVE_LONG_LONG is not defined.
665    
666    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
667        bring them up to date and include new features. This patch includes:
668    
669        - Fixed PH's badly added libz and libbz2 support.
670        - Fixed a problem with static linking.
671        - Added pcredemo. [But later removed - see 7 below.]
672        - Fixed dftables problem and added an option.
673        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
674            HAVE_LONG_LONG.
675        - Added readline support for pcretest.
676        - Added an listing of the option settings after cmake has run.
677    
678    4.  A user submitted a patch to Makefile that makes it easy to create
679        "pcre.dll" under mingw when using Configure/Make. I added stuff to
680        Makefile.am that cause it to include this special target, without
681        affecting anything else. Note that the same mingw target plus all
682        the other distribution libraries and programs are now supported
683        when configuring with CMake (see 6 below) instead of with
684        Configure/Make.
685    
686    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
687        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
688        exported in the Windows port". It has not yet been confirmed that the patch
689        solves the problem, but it does no harm.
690    
691    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
692        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
693        with CMake, and also correct the comment about stack recursion.
694    
695    7.  Remove the automatic building of pcredemo from the ./configure system and
696        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
697        of a program that users should build themselves after PCRE is installed, so
698        building it automatically is not really right. What is more, it gave
699        trouble in some build environments.
700    
701    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
702    
703    
704    Version 7.5 10-Jan-08
705    ---------------------
706    
707    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
708        values in parens when parsing an RE using the C++ wrapper."
709    
710    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
711        Characters greater than 255 were excluded from the class instead of being
712        included.
713    
714    3.  The same bug as (2) above applied to negated POSIX classes such as
715        [:^space:].
716    
717    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
718        defined or documented. It seems to have been a typo for PCRE_STATIC, so
719        I have changed it.
720    
721    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
722        first named subpattern) and a construct such as (?&a) would reference the
723        first named subpattern whose name started with "a" (in other words, the
724        length check was missing). Both these problems are fixed. "Subpattern name
725        expected" is now given for (?&) (a zero-length name), and this patch also
726        makes it give the same error for \k'' (previously it complained that that
727        was a reference to a non-existent subpattern).
728    
729    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
730        this is right because (?- can be followed by option settings as well as by
731        digits. I have, however, made the messages clearer.
732    
733    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
734        than the number used in the conditional) now cause a compile-time error.
735        This is actually not compatible with Perl, which accepts such patterns, but
736        treats the conditional as always being FALSE (as PCRE used to), but it
737        seems to me that giving a diagnostic is better.
738    
739    8.  Change "alphameric" to the more common word "alphanumeric" in comments
740        and messages.
741    
742    9.  Fix two occurrences of "backslash" in comments that should have been
743        "backspace".
744    
745    10. Remove two redundant lines of code that can never be obeyed (their function
746        was moved elsewhere).
747    
748    11. The program that makes PCRE's Unicode character property table had a bug
749        which caused it to generate incorrect table entries for sequences of
750        characters that have the same character type, but are in different scripts.
751        It amalgamated them into a single range, with the script of the first of
752        them. In other words, some characters were in the wrong script. There were
753        thirteen such cases, affecting characters in the following ranges:
754    
755          U+002b0 - U+002c1
756          U+0060c - U+0060d
757          U+0061e - U+00612
758          U+0064b - U+0065e
759          U+0074d - U+0076d
760          U+01800 - U+01805
761          U+01d00 - U+01d77
762          U+01d9b - U+01dbf
763          U+0200b - U+0200f
764          U+030fc - U+030fe
765          U+03260 - U+0327f
766          U+0fb46 - U+0fbb1
767          U+10450 - U+1049d
768    
769    12. The -o option (show only the matching part of a line) for pcregrep was not
770        compatible with GNU grep in that, if there was more than one match in a
771        line, it showed only the first of them. It now behaves in the same way as
772        GNU grep.
773    
774    13. If the -o and -v options were combined for pcregrep, it printed a blank
775        line for every non-matching line. GNU grep prints nothing, and pcregrep now
776        does the same. The return code can be used to tell if there were any
777        non-matching lines.
778    
779    14. Added --file-offsets and --line-offsets to pcregrep.
780    
781    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
782        infinitely looping recursion. The bug was that positive lookaheads were not
783        being skipped when checking for a possible empty match (negative lookaheads
784        and both kinds of lookbehind were skipped).
785    
786    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
787        inclusion of <windows.h> to before rather than after the definition of
788        INVALID_FILE_ATTRIBUTES (patch from David Byron).
789    
790    17. Specifying a possessive quantifier with a specific limit for a Unicode
791        character property caused pcre_compile() to compile bad code, which led at
792        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
793        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
794        caused the error; without that there was no problem.
795    
796    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
797    
798    19. Added --enable-pcretest-libreadline.
799    
800    20. In pcrecpp.cc, the variable 'count' was incremented twice in
801        RE::GlobalReplace(). As a result, the number of replacements returned was
802        double what it should be. I removed one of the increments, but Craig sent a
803        later patch that removed the other one (the right fix) and added unit tests
804        that check the return values (which was not done before).
805    
806    21. Several CMake things:
807    
808        (1) Arranged that, when cmake is used on Unix, the libraries end up with
809            the names libpcre and libpcreposix, not just pcre and pcreposix.
810    
811        (2) The above change means that pcretest and pcregrep are now correctly
812            linked with the newly-built libraries, not previously installed ones.
813    
814        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
815    
816    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
817        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
818        UTF-8 newline character). The key issue is that the pattern starts .*;
819        this means that the match must be either at the beginning, or after a
820        newline. The bug was in the code for advancing after a failed match and
821        checking that the new position followed a newline. It was not taking
822        account of UTF-8 characters correctly.
823    
824    23. PCRE was behaving differently from Perl in the way it recognized POSIX
825        character classes. PCRE was not treating the sequence [:...:] as a
826        character class unless the ... were all letters. Perl, however, seems to
827        allow any characters between [: and :], though of course it rejects as
828        unknown any "names" that contain non-letters, because all the known class
829        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
830        for example, whereas PCRE did not - it did not recognize a POSIX character
831        class. This seemed a bit dangerous, so the code has been changed to be
832        closer to Perl. The behaviour is not identical to Perl, because PCRE will
833        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
834        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
835        Perl does, and where it didn't before.
836    
837    24. Rewrite so as to remove the single use of %n from pcregrep because in some
838        Windows environments %n is disabled by default.
839    
840    
841    Version 7.4 21-Sep-07
842    ---------------------
843    
844    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
845        means that a class such as [\s] counted as "explicit reference to CR or
846        LF". That isn't really right - the whole point of the change was to try to
847        help when there was an actual mention of one of the two characters. So now
848        the change happens only if \r or \n (or a literal CR or LF) character is
849        encountered.
850    
851    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
852        of both had grown to the point where there were only 3 bits left.
853        Fortunately, there was spare space in the data structure, and so I have
854        moved the internal flags into a new 16-bit field to free up more option
855        bits.
856    
857    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
858        but did not set the internal JCHANGED flag - either of these is enough to
859        control the way the "get" function works - but the PCRE_INFO_JCHANGED
860        facility is supposed to tell if (?J) was ever used, so now (?J) at the
861        start sets both bits.
862    
863    4.  Added options (at build time, compile time, exec time) to change \R from
864        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
865    
866    5.  doc/pcresyntax.html was missing from the distribution.
867    
868    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
869        compatibility, even though it is no longer used.
870    
871    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
872        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
873        windows.h file is present (where different names are used). [This was
874        reversed later after testing - see 16 below.]
875    
876    8.  Changed all #include <config.h> to #include "config.h". There were also
877        some further <pcre.h> cases that I changed to "pcre.h".
878    
879    9.  When pcregrep was used with the --colour option, it missed the line ending
880        sequence off the lines that it output.
881    
882    10. It was pointed out to me that arrays of string pointers cause lots of
883        relocations when a shared library is dynamically loaded. A technique of
884        using a single long string with a table of offsets can drastically reduce
885        these. I have refactored PCRE in four places to do this. The result is
886        dramatic:
887    
888          Originally:                          290
889          After changing UCP table:            187
890          After changing error message table:   43
891          After changing table of "verbs"       36
892          After changing table of Posix names   22
893    
894        Thanks to the folks working on Gregex for glib for this insight.
895    
896    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
897        unicode-properties was also set.
898    
899    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
900    
901    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
902        checked only for CRLF.
903    
904    14. Added casts to pcretest.c to avoid compiler warnings.
905    
906    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
907    
908    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
909        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
910        entirely. This removes changes made in 7 above.
911    
912    17. The CMake files have been updated, and there is now more information about
913        building with CMake in the NON-UNIX-USE document.
914    
915    
916    Version 7.3 28-Aug-07
917    ---------------------
918    
919     1. In the rejigging of the build system that eventually resulted in 7.1, the
920        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
921        brackets there is not right, since it causes compilers to look for an
922        installed pcre.h, not the version that is in the source that is being
923        compiled (which of course may be different). I have changed it back to:
924    
925          #include "pcre.h"
926    
927        I have a vague recollection that the change was concerned with compiling in
928        different directories, but in the new build system, that is taken care of
929        by the VPATH setting the Makefile.
930    
931     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
932        when the subject happened to end in the byte 0x85 (e.g. if the last
933        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
934        characters but of course it shouldn't be taken as a newline when it is part
935        of another character. The bug was that, for an unlimited repeat of . in
936        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
937        characters when looking for a newline.
938    
939     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
940    
941     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
942        in debug output.
943    
944     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
945        long printing in the pcrecpp unittest when running under MinGW.
946    
947     6. ESC_K was left out of the EBCDIC table.
948    
949     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
950        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
951        limit also applies to "virtual nesting" when a pattern is recursive, and in
952        this case 1000 isn't so big. I have been able to remove this limit at the
953        expense of backing off one optimization in certain circumstances. Normally,
954        when pcre_exec() would call its internal match() function recursively and
955        immediately return the result unconditionally, it uses a "tail recursion"
956        feature to save stack. However, when a subpattern that can match an empty
957        string has an unlimited repetition quantifier, it no longer makes this
958        optimization. That gives it a stack frame in which to save the data for
959        checking that an empty string has been matched. Previously this was taken
960        from the 1000-entry workspace that had been reserved. So now there is no
961        explicit limit, but more stack is used.
962    
963     8. Applied Daniel's patches to solve problems with the import/export magic
964        syntax that is required for Windows, and which was going wrong for the
965        pcreposix and pcrecpp parts of the library. These were overlooked when this
966        problem was solved for the main library.
967    
968     9. There were some crude static tests to avoid integer overflow when computing
969        the size of patterns that contain repeated groups with explicit upper
970        limits. As the maximum quantifier is 65535, the maximum group length was
971        set at 30,000 so that the product of these two numbers did not overflow a
972        32-bit integer. However, it turns out that people want to use groups that
973        are longer than 30,000 bytes (though not repeat them that many times).
974        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
975        made it possible to implement the integer overflow checks in a much more
976        dynamic way, which I have now done. The artificial limitation on group
977        length has been removed - we now have only the limit on the total length of
978        the compiled pattern, which depends on the LINK_SIZE setting.
979    
980    10. Fixed a bug in the documentation for get/copy named substring when
981        duplicate names are permitted. If none of the named substrings are set, the
982        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
983        empty string.
984    
985    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
986        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
987        because the ] is interpreted as the first data character and the
988        terminating ] is not found. PCRE has been made compatible with Perl in this
989        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
990        cause memory overwriting.
991    
992    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
993        string has been matched (to stop an infinite loop). It was not recognizing
994        a conditional subpattern that could match an empty string if that
995        subpattern was within another subpattern. For example, it looped when
996        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
997        condition was not nested. This bug has been fixed.
998    
999    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1000        past the start of the subject in the presence of bytes with the top bit
1001        set, for example "\x8aBCD".
1002    
1003    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1004        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1005    
1006    14. Optimized (?!) to (*FAIL).
1007    
1008    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1009        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1010        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1011        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1012        does: it's just the validity check that is more restrictive.
1013    
1014    16. Inserted checks for integer overflows during escape sequence (backslash)
1015        processing, and also fixed erroneous offset values for syntax errors during
1016        backslash processing.
1017    
1018    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1019        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1020    
1021    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1022        caused an overrun.
1023    
1024    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1025        something other than just ASCII characters) inside a group that had an
1026        unlimited repeat caused a loop at compile time (while checking to see
1027        whether the group could match an empty string).
1028    
1029    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1030        [\P{Any}] did so. (Error in the code for printing property names.)
1031    
1032    21. An orphan \E inside a character class could cause a crash.
1033    
1034    22. A repeated capturing bracket such as (A)? could cause a wild memory
1035        reference during compilation.
1036    
1037    23. There are several functions in pcre_compile() that scan along a compiled
1038        expression for various reasons (e.g. to see if it's fixed length for look
1039        behind). There were bugs in these functions when a repeated \p or \P was
1040        present in the pattern. These operators have additional parameters compared
1041        with \d, etc, and these were not being taken into account when moving along
1042        the compiled data. Specifically:
1043    
1044        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1045            length.
1046    
1047        (b) An item such as \pL+ within a repeated group could cause crashes or
1048            loops.
1049    
1050        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1051            "reference to non-existent subpattern" error.
1052    
1053        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1054    
1055    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1056        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1057    
1058    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1059    
1060    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1061        character were causing crashes (broken optimization).
1062    
1063    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1064        \p or \P) caused a compile-time loop.
1065    
1066    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1067        break. For example, the unstudied pattern [\r\n]A does not match the string
1068        "\r\nA" because change 7.0/46 below moves the current point on by two
1069        characters after failing to match at the start. However, the pattern \nA
1070        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1071        the same is true. There doesn't seem any very clean way out of this, but
1072        what I have chosen to do makes the common cases work: PCRE now takes note
1073        of whether there can be an explicit match for \r or \n anywhere in the
1074        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1075        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1076        pattern has explicit CR or LF references.
1077    
1078    29. Added (*CR) etc for changing newline setting at start of pattern.
1079    
1080    
1081    Version 7.2 19-Jun-07
1082    ---------------------
1083    
1084     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1085        which is apparently normally available under Windows.
1086    
1087     2. Re-jig the pcregrep tests with different newline settings in an attempt
1088        to make them independent of the local environment's newline setting.
1089    
1090     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1091    
1092     4. Some of the "internals" tests were previously cut out when the link size
1093        was not 2, because the output contained actual offsets. The recent new
1094        "Z" feature of pcretest means that these can be cut out, making the tests
1095        usable with all link sizes.
1096    
1097     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1098        stack recursion. This gives a massive performance boost under BSD, but just
1099        a small improvement under Linux. However, it saves one field in the frame
1100        in all cases.
1101    
1102     6. Added more features from the forthcoming Perl 5.10:
1103    
1104        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1105            recursion call. It refers to the nth most recently opened parentheses.
1106    
1107        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1108            to be opened parentheses.
1109    
1110        (c) Conditions that refer to capturing parentheses can be specified
1111            relatively, for example, (?(-2)... or (?(+3)...
1112    
1113        (d) \K resets the start of the current match so that everything before
1114            is not part of it.
1115    
1116        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1117    
1118        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1119            reference syntax.
1120    
1121        (g) (?| introduces a group in which the numbering of parentheses in each
1122            alternative starts with the same number.
1123    
1124        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1125    
1126     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1127        PCRE_INFO_JCHANGED.
1128    
1129     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1130        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1131        for detecting groups that can match an empty string.
1132    
1133     9. A pattern with a very large number of alternatives (more than several
1134        hundred) was running out of internal workspace during the pre-compile
1135        phase, where pcre_compile() figures out how much memory will be needed. A
1136        bit of new cunning has reduced the workspace needed for groups with
1137        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1138        workspace instead of running out of the 4096 that are available.
1139    
1140    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1141    
1142    11. Applied patch from Google to remove an optimization that didn't quite work.
1143        The report of the bug said:
1144    
1145          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1146          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1147          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1148    
1149    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1150        it matched the wrong number of bytes.
1151    
1152    
1153    Version 7.1 24-Apr-07
1154    ---------------------
1155    
1156     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1157        that is more "standard", making use of automake and other Autotools. There
1158        is some re-arrangement of the files and adjustment of comments consequent
1159        on this.
1160    
1161     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1162        for recursive directory scanning broke on some systems because the files
1163        are not scanned in any specific order and on different systems the order
1164        was different. A call to "sort" has been inserted into RunGrepTest for the
1165        approprate test as a short-term fix. In the longer term there may be an
1166        alternative.
1167    
1168     3. I had an email from Eric Raymond about problems translating some of PCRE's
1169        man pages to HTML (despite the fact that I distribute HTML pages, some
1170        people do their own conversions for various reasons). The problems
1171        concerned the use of low-level troff macros .br and .in. I have therefore
1172        removed all such uses from the man pages (some were redundant, some could
1173        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1174        HTML has been updated to handle .nf/.fi and to complain if it encounters
1175        .br or .in.
1176    
1177     4. Updated comments in configure.ac that get placed in config.h.in and also
1178        arranged for config.h to be included in the distribution, with the name
1179        config.h.generic, for the benefit of those who have to compile without
1180        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1181    
1182     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1183        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1184        makevp.bat for the current PCRE, using the additional files
1185        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1186    
1187     6. A Windows user reported a minor discrepancy with test 2, which turned out
1188        to be caused by a trailing space on an input line that had got lost in his
1189        copy. The trailing space was an accident, so I've just removed it.
1190    
1191     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1192        that is needed.
1193    
1194     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1195        as "const" (a) because they are and (b) because it helps the PHP
1196        maintainers who have recently made a script to detect big data structures
1197        in the php code that should be moved to the .rodata section. I remembered
1198        to update Builducptable as well, so it won't revert if ucptable.h is ever
1199        re-created.
1200    
1201     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1202        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1203        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1204        support is not required. This saves 1.5-2K of code, which is important in
1205        some applications.
1206    
1207        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1208        so as not to refer to the tables, even though these functions will never be
1209        called when UTF-8 support is disabled. Otherwise there are problems with a
1210        shared library.
1211    
1212    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1213    
1214        (a) It was defining its arguments as char * instead of void *.
1215    
1216        (b) It was assuming that all moves were upwards in memory; this was true
1217            a long time ago when I wrote it, but is no longer the case.
1218    
1219        The emulated memove() is provided for those environments that have neither
1220        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1221        is clearly not the case, as these two bugs were recently reported.
1222    
1223    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1224        and Detrail to create the HTML documentation, the .txt form of the man
1225        pages, and it removes trailing spaces from listed files. It also creates
1226        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1227        case, it wraps all the #defines with #ifndefs. This script should be run
1228        before "make dist".
1229    
1230    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1231        with Unicode property support.
1232    
1233        (a) For a maximizing quantifier, if the two different cases of the
1234            character were of different lengths in their UTF-8 codings (there are
1235            some cases like this - I found 11), and the matching function had to
1236            back up over a mixture of the two cases, it incorrectly assumed they
1237            were both the same length.
1238    
1239        (b) When PCRE was configured to use the heap rather than the stack for
1240            recursion during matching, it was not correctly preserving the data for
1241            the other case of a UTF-8 character when checking ahead for a match
1242            while processing a minimizing repeat. If the check also involved
1243            matching a wide character, but failed, corruption could cause an
1244            erroneous result when trying to check for a repeat of the original
1245            character.
1246    
1247    13. Some tidying changes to the testing mechanism:
1248    
1249        (a) The RunTest script now detects the internal link size and whether there
1250            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1251            values substituted by "configure". (The RunGrepTest script already did
1252            this for UTF-8.) The configure.ac script no longer substitutes the
1253            relevant variables.
1254    
1255        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1256            with length and offset values. This means that the output is different
1257            for different internal link sizes. Test 2 is skipped for link sizes
1258            other than 2 because of this, bypassing the problem. Unfortunately,
1259            there was also a test in test 3 (the locale tests) that used /B and
1260            failed for link sizes other than 2. Rather than cut the whole test out,
1261            I have added a new /Z option to pcretest that replaces the length and
1262            offset values with spaces. This is now used to make test 3 independent
1263            of link size. (Test 2 will be tidied up later.)
1264    
1265    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1266        segmentation fault instead of returning the appropriate error message.
1267    
1268    15. In multiline mode when the newline sequence was set to "any", the pattern
1269        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1270        This doesn't seem right; it now treats the CRLF combination as the line
1271        ending, and so does not match in that case. It's only a pattern such as ^$
1272        that would hit this one: something like ^ABC$ would have failed after \r
1273        and then tried again after \r\n.
1274    
1275    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1276        in an attempt to make files that differ only in their line terminators
1277        compare equal. This works on Linux.
1278    
1279    17. Under certain error circumstances pcregrep might try to free random memory
1280        as it exited. This is now fixed, thanks to valgrind.
1281    
1282    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1283        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1284        was because its rules for how to advance for /g after matching an empty
1285        string at the end of a line did not allow for this case. They now check for
1286        it specially.
1287    
1288    20. pcretest is supposed to handle patterns and data of any length, by
1289        extending its buffers when necessary. It was getting this wrong when the
1290        buffer for a data line had to be extended.
1291    
1292    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1293        CRLF as a newline sequence.
1294    
1295    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1296        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1297        I have nevertheless tidied it up.
1298    
1299    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1300    
1301    24. Added a man page for pcre-config.
1302    
1303    
1304    Version 7.0 19-Dec-06
1305    ---------------------
1306    
1307     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1308        moving to gcc 4.1.1.
1309    
1310     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1311        sys/time.h, which is documented as needed for this function. It doesn't
1312        seem to matter on Linux, but it showed up on some releases of OS X.
1313    
1314     3. It seems that there are systems where bytes whose values are greater than
1315        127 match isprint() in the "C" locale. The "C" locale should be the
1316        default when a C program starts up. In most systems, only ASCII printing
1317        characters match isprint(). This difference caused the output from pcretest
1318        to vary, making some of the tests fail. I have changed pcretest so that:
1319    
1320        (a) When it is outputting text in the compiled version of a pattern, bytes
1321            other than 32-126 are always shown as hex escapes.
1322    
1323        (b) When it is outputting text that is a matched part of a subject string,
1324            it does the same, unless a different locale has been set for the match
1325            (using the /L modifier). In this case, it uses isprint() to decide.
1326    
1327     4. Fixed a major bug that caused incorrect computation of the amount of memory
1328        required for a compiled pattern when options that changed within the
1329        pattern affected the logic of the preliminary scan that determines the
1330        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1331        that the computed length was too small. The symptoms of this bug were
1332        either the PCRE error "internal error: code overflow" from pcre_compile(),
1333        or a glibc crash with a message such as "pcretest: free(): invalid next
1334        size (fast)". Examples of patterns that provoked this bug (shown in
1335        pcretest format) are:
1336    
1337          /(?-x: )/x
1338          /(?x)(?-x: \s*#\s*)/
1339          /((?i)[\x{c0}])/8
1340          /(?i:[\x{c0}])/8
1341    
1342        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1343        is now done differently.
1344    
1345     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1346        wrapper classes; (b) implement a new function in the C++ scanner that is
1347        more efficient than the old way of doing things because it avoids levels of
1348        recursion in the regex matching; (c) add a paragraph to the documentation
1349        for the FullMatch() function.
1350    
1351     6. The escape sequence \n was being treated as whatever was defined as
1352        "newline". Not only was this contrary to the documentation, which states
1353        that \n is character 10 (hex 0A), but it also went horribly wrong when
1354        "newline" was defined as CRLF. This has been fixed.
1355    
1356     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1357        was being set to -1 for the "end of line" case (supposedly a value that no
1358        character can have). Though this value is never used (the check for end of
1359        line is "zero bytes in current character"), it caused compiler complaints.
1360        I've changed it to 0xffffffff.
1361    
1362     8. In pcre_version.c, the version string was being built by a sequence of
1363        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1364        string (as it is for production releases) called a macro with an empty
1365        argument. The C standard says the result of this is undefined. The gcc
1366        compiler treats it as an empty string (which was what was wanted) but it is
1367        reported that Visual C gives an error. The source has been hacked around to
1368        avoid this problem.
1369    
1370     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1371        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1372        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1373        of them did).
1374    
1375    10. Originally, pcretest opened its input and output without "b"; then I was
1376        told that "b" was needed in some environments, so it was added for release
1377        5.0 to both the input and output. (It makes no difference on Unix-like
1378        systems.) Later I was told that it is wrong for the input on Windows. I've
1379        now abstracted the modes into two macros, to make it easier to fiddle with
1380        them, and removed "b" from the input mode under Windows.
1381    
1382    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1383    
1384    12. Added -help and --help to pcretest as an official way of being reminded
1385        of the options.
1386    
1387    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1388        and pcrecpp.cc because they annoy compilers at high warning levels.
1389    
1390    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1391    
1392    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1393        scripts are not C programs :-) and which was not noticed because it works
1394        on Linux.
1395    
1396    16. pcretest is supposed to handle any length of pattern and data line (as one
1397        line or as a continued sequence of lines) by extending its input buffer if
1398        necessary. This feature was broken for very long pattern lines, leading to
1399        a string of junk being passed to pcre_compile() if the pattern was longer
1400        than about 50K.
1401    
1402    17. I have done a major re-factoring of the way pcre_compile() computes the
1403        amount of memory needed for a compiled pattern. Previously, there was code
1404        that made a preliminary scan of the pattern in order to do this. That was
1405        OK when PCRE was new, but as the facilities have expanded, it has become
1406        harder and harder to keep it in step with the real compile phase, and there
1407        have been a number of bugs (see for example, 4 above). I have now found a
1408        cunning way of running the real compile function in a "fake" mode that
1409        enables it to compute how much memory it would need, while actually only
1410        ever using a few hundred bytes of working memory and without too many
1411        tests of the mode. This should make future maintenance and development
1412        easier. A side effect of this work is that the limit of 200 on the nesting
1413        depth of parentheses has been removed (though this was never a serious
1414        limitation, I suspect). However, there is a downside: pcre_compile() now
1415        runs more slowly than before (30% or more, depending on the pattern). I
1416        hope this isn't a big issue. There is no effect on runtime performance.
1417    
1418    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1419        newline (only possible for the last line of a file) and it was a
1420        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1421    
1422    19. Added additional timing features to pcretest. (1) The -tm option now times
1423        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1424        separate command line item, by a number that specifies the number of
1425        repeats to use when timing. The default is 50000; this gives better
1426        precision, but takes uncomfortably long for very large patterns.
1427    
1428    20. Extended pcre_study() to be more clever in cases where a branch of a
1429        subpattern has no definite first character. For example, (a*|b*)[cd] would
1430        previously give no result from pcre_study(). Now it recognizes that the
1431        first character must be a, b, c, or d.
1432    
1433    21. There was an incorrect error "recursive call could loop indefinitely" if
1434        a subpattern (or the entire pattern) that was being tested for matching an
1435        empty string contained only one non-empty item after a nested subpattern.
1436        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1437        incorrectly, because the \d was being skipped in the check.
1438    
1439    22. The pcretest program now has a new pattern option /B and a command line
1440        option -b, which is equivalent to adding /B to every pattern. This causes
1441        it to show the compiled bytecode, without the additional information that
1442        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1443        is the same as /B/I).
1444    
1445    23. A new optimization is now able automatically to treat some sequences such
1446        as a*b as a*+b. More specifically, if something simple (such as a character
1447        or a simple class like \d) has an unlimited quantifier, and is followed by
1448        something that cannot possibly match the quantified thing, the quantifier
1449        is automatically "possessified".
1450    
1451    24. A recursive reference to a subpattern whose number was greater than 39
1452        went wrong under certain circumstances in UTF-8 mode. This bug could also
1453        have affected the operation of pcre_study().
1454    
1455    25. Realized that a little bit of performance could be had by replacing
1456        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1457    
1458    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1459    
1460    27. Possessive quantifiers such as a++ were previously implemented by turning
1461        them into atomic groups such as ($>a+). Now they have their own opcodes,
1462        which improves performance. This includes the automatically created ones
1463        from 23 above.
1464    
1465    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1466        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1467        the first matched character to be a colon. This applied both to named and
1468        numbered groups.
1469    
1470    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1471    
1472    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1473        building PCRE on Windows easier, so I have included it in the distribution.
1474    
1475    31. There is now a check in pcretest against a ridiculously large number being
1476        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1477        loop, the loop is abandoned.
1478    
1479    32. Forward references to subpatterns in conditions such as (?(2)...) where
1480        subpattern 2 is defined later cause pcre_compile() to search forwards in
1481        the pattern for the relevant set of parentheses. This search went wrong
1482        when there were unescaped parentheses in a character class, parentheses
1483        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1484    
1485    33. "Subroutine" calls and backreferences were previously restricted to
1486        referencing subpatterns earlier in the regex. This restriction has now
1487        been removed.
1488    
1489    34. Added a number of extra features that are going to be in Perl 5.10. On the
1490        whole, these are just syntactic alternatives for features that PCRE had
1491        previously implemented using the Python syntax or my own invention. The
1492        other formats are all retained for compatibility.
1493    
1494        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1495            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1496            also .NET compatible.
1497    
1498        (b) A recursion or subroutine call to a named group can now be defined as
1499            (?&name) as well as (?P>name).
1500    
1501        (c) A backreference to a named group can now be defined as \k<name> or
1502            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1503            5.10, are also .NET compatible.
1504    
1505        (d) A conditional reference to a named group can now use the syntax
1506            (?(<name>) or (?('name') as well as (?(name).
1507    
1508        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1509            groups (named and numbered) that are never evaluated inline, but can be
1510            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1511            is always false. There may be only one alternative in such a group.
1512    
1513        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1514            as the simple (?(R). The condition is true only if the most recent
1515            recursion is that of the given number or name. It does not search out
1516            through the entire recursion stack.
1517    
1518        (g) The escape \gN or \g{N} has been added, where N is a positive or
1519            negative number, specifying an absolute or relative reference.
1520    
1521    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1522        some "unreachable code" warnings.
1523    
1524    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1525        things, this adds five new scripts.
1526    
1527    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1528        There were also incompatibilities regarding the handling of \Q..\E inside
1529        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1530        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1531    
1532    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1533        matches an empty string, and forcibly breaks the loop. There were bugs in
1534        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1535        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1536        separate and independent bugs (that affected different cases) have been
1537        fixed.
1538    
1539    39. Refactored the code to abolish the use of different opcodes for small
1540        capturing bracket numbers. This is a tidy that I avoided doing when I
1541        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1542        The new approach is not only tidier, it makes it possible to reduce the
1543        memory needed to fix the previous bug (38).
1544    
1545    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1546        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1547        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1548        mode.
1549    
1550    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1551        report.
1552    
1553    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1554        copy construction and assignment in the C++ wrapper.
1555    
1556    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1557        couple of bugs that could have given wrong results in the "--newline=crlf"
1558        case.
1559    
1560    44. Added a number of casts and did some reorganization of signed/unsigned int
1561        variables following suggestions from Dair Grant. Also renamed the variable
1562        "this" as "item" because it is a C++ keyword.
1563    
1564    45. Arranged for dftables to add
1565    
1566          #include "pcre_internal.h"
1567    
1568        to pcre_chartables.c because without it, gcc 4.x may remove the array
1569        definition from the final binary if PCRE is built into a static library and
1570        dead code stripping is activated.
1571    
1572    46. For an unanchored pattern, if a match attempt fails at the start of a
1573        newline sequence, and the newline setting is CRLF or ANY, and the next two
1574        characters are CRLF, advance by two characters instead of one.
1575    
1576    
1577    Version 6.7 04-Jul-06
1578    ---------------------
1579    
1580     1. In order to handle tests when input lines are enormously long, pcretest has
1581        been re-factored so that it automatically extends its buffers when
1582        necessary. The code is crude, but this _is_ just a test program. The
1583        default size has been increased from 32K to 50K.
1584    
1585     2. The code in pcre_study() was using the value of the re argument before
1586        testing it for NULL. (Of course, in any sensible call of the function, it
1587        won't be NULL.)
1588    
1589     3. The memmove() emulation function in pcre_internal.h, which is used on
1590        systems that lack both memmove() and bcopy() - that is, hardly ever -
1591        was missing a "static" storage class specifier.
1592    
1593     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1594        containing an extended class (one that cannot be represented by a bitmap
1595        because it contains high-valued characters or Unicode property items, e.g.
1596        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1597        pattern, but PCRE should not loop if you do not (it no longer does).
1598        [Detail: two cases were found: (a) a repeated subpattern containing an
1599        extended class; (b) a recursive reference to a subpattern that followed a
1600        previous extended class. It wasn't skipping over the extended class
1601        correctly when UTF-8 mode was not set.]
1602    
1603     5. A negated single-character class was not being recognized as fixed-length
1604        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1605        compile error "lookbehind assertion is not fixed length".
1606    
1607     6. The RunPerlTest auxiliary script was showing an unexpected difference
1608        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1609        write a Perl script that can interpret lines of an input file either as
1610        byte characters or as UTF-8, which is what "perltest" was being required to
1611        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1612        can't do is switch easily at run time between having the "use utf8;" pragma
1613        or not. In the end, I fudged it by using the RunPerlTest script to insert
1614        "use utf8;" explicitly for the UTF-8 tests.
1615    
1616     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1617        the end of the subject string, contrary to the documentation and to what
1618        Perl does. This was true of both matching functions. Now it matches only at
1619        the start of the subject and immediately after *internal* newlines.
1620    
1621     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1622        a pointer to an int instead of a pointer to an unsigned long int. This
1623        caused problems on 64-bit systems.
1624    
1625     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1626        instance of the 'standard' template library not being so standard".
1627    
1628    10. There was no check on the number of named subpatterns nor the maximum
1629        length of a subpattern name. The product of these values is used to compute
1630        the size of the memory block for a compiled pattern. By supplying a very
1631        long subpattern name and a large number of named subpatterns, the size
1632        computation could be caused to overflow. This is now prevented by limiting
1633        the length of names to 32 characters, and the number of named subpatterns
1634        to 10,000.
1635    
1636    11. Subpatterns that are repeated with specific counts have to be replicated in
1637        the compiled pattern. The size of memory for this was computed from the
1638        length of the subpattern and the repeat count. The latter is limited to
1639        65535, but there was no limit on the former, meaning that integer overflow
1640        could in principle occur. The compiled length of a repeated subpattern is
1641        now limited to 30,000 bytes in order to prevent this.
1642    
1643    12. Added the optional facility to have named substrings with the same name.
1644    
1645    13. Added the ability to use a named substring as a condition, using the
1646        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1647        are numbers (not recommended). Forward references are permitted.
1648    
1649    14. Added forward references in named backreferences (if you see what I mean).
1650    
1651    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1652        pattern could run off the end of the subject. For example, the pattern
1653        "(?s)(.{1,5})"8 did this with the subject "ab".
1654    
1655    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1656        PCRE_CASELESS was set when matching characters that were quantified with ?
1657        or *.
1658    
1659    17. A character class other than a single negated character that had a minimum
1660        but no maximum quantifier - for example [ab]{6,} - was not handled
1661        correctly by pce_dfa_exec(). It would match only one character.
1662    
1663    18. A valid (though odd) pattern that looked like a POSIX character
1664        class but used an invalid character after [ (for example [[,abc,]]) caused
1665        pcre_compile() to give the error "Failed: internal error: code overflow" or
1666        in some cases to crash with a glibc free() error. This could even happen if
1667        the pattern terminated after [[ but there just happened to be a sequence of
1668        letters, a binary zero, and a closing ] in the memory that followed.
1669    
1670    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1671        over the years. Originally (before any Unicode support), just the bottom 8
1672        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1673        output from "man perlunicode" includes this:
1674    
1675          The regular expression compiler produces polymorphic opcodes.  That
1676          is, the pattern adapts to the data and automatically switches to
1677          the Unicode character scheme when presented with Unicode data--or
1678          instead uses a traditional byte scheme when presented with byte
1679          data.
1680    
1681        Sadly, a wide octal escape does not cause a switch, and in a string with
1682        no other multibyte characters, these octal escapes are treated as before.
1683        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1684        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1685        Unicode string.
1686    
1687        I have not perpetrated such confusion in PCRE. Up till now, it took just
1688        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1689        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1690        translate to the appropriate multibyte character.
1691    
1692    29. Applied some refactoring to reduce the number of warnings from Microsoft
1693        and Borland compilers. This has included removing the fudge introduced
1694        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1695        a warning about an unused variable.
1696    
1697    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1698        characters since release 4.0, because Perl (from release 5.004) does not.
1699        [Or at least, is documented not to: some releases seem to be in conflict
1700        with the documentation.] However, when a pattern was studied with
1701        pcre_study() and all its branches started with \s, PCRE still included VT
1702        as a possible starting character. Of course, this did no harm; it just
1703        caused an unnecessary match attempt.
1704    
1705    22. Removed a now-redundant internal flag bit that recorded the fact that case
1706        dependency changed within the pattern. This was once needed for "required
1707        byte" processing, but is no longer used. This recovers a now-scarce options
1708        bit. Also moved the least significant internal flag bit to the most-
1709        significant bit of the word, which was not previously used (hangover from
1710        the days when it was an int rather than a uint) to free up another bit for
1711        the future.
1712    
1713    23. Added support for CRLF line endings as well as CR and LF. As well as the
1714        default being selectable at build time, it can now be changed at runtime
1715        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1716        specify that it is scanning data with non-default line endings.
1717    
1718    24. Changed the definition of CXXLINK to make it agree with the definition of
1719        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1720    
1721    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1722        recursions. This makes a big different to stack usage for some patterns.
1723    
1724    26. If a subpattern containing a named recursion or subroutine reference such
1725        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1726        the space required for the compiled pattern went wrong and gave too small a
1727        value. Depending on the environment, this could lead to "Failed: internal
1728        error: code overflow at offset 49" or "glibc detected double free or
1729        corruption" errors.
1730    
1731    27. Applied patches from Google (a) to support the new newline modes and (b) to
1732        advance over multibyte UTF-8 characters in GlobalReplace.
1733    
1734    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1735        difference for some implementation of PCRE in some Windows version.
1736    
1737    29. Added some extra testing facilities to pcretest:
1738    
1739        \q<number>   in a data line sets the "match limit" value
1740        \Q<number>   in a data line sets the "match recursion limt" value
1741        -S <number>  sets the stack size, where <number> is in megabytes
1742    
1743        The -S option isn't available for Windows.
1744    
1745    
1746    Version 6.6 06-Feb-06
1747    ---------------------
1748    
1749     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1750        in pcreposix.h. I have copied the definition from pcre.h.
1751    
1752     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1753        because pcre.h is no longer a built file.
1754    
1755     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1756        not normally included in the compiled code.
1757    
1758    
1759    Version 6.5 01-Feb-06
1760    ---------------------
1761    
1762     1. When using the partial match feature with pcre_dfa_exec(), it was not
1763        anchoring the second and subsequent partial matches at the new starting
1764        point. This could lead to incorrect results. For example, with the pattern
1765        /1234/, partially matching against "123" and then "a4" gave a match.
1766    
1767     2. Changes to pcregrep:
1768    
1769        (a) All non-match returns from pcre_exec() were being treated as failures
1770            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1771            error message is output. Some extra information is given for the
1772            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1773            probably the only errors that are likely to be caused by users (by
1774            specifying a regex that has nested indefinite repeats, for instance).
1775            If there are more than 20 of these errors, pcregrep is abandoned.
1776    
1777        (b) A binary zero was treated as data while matching, but terminated the
1778            output line if it was written out. This has been fixed: binary zeroes
1779            are now no different to any other data bytes.
1780    
1781        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1782            used to set a locale for matching. The --locale=xxxx long option has
1783            been added (no short equivalent) to specify a locale explicitly on the
1784            pcregrep command, overriding the environment variables.
1785    
1786        (d) When -B was used with -n, some line numbers in the output were one less
1787            than they should have been.
1788    
1789        (e) Added the -o (--only-matching) option.
1790    
1791        (f) If -A or -C was used with -c (count only), some lines of context were
1792            accidentally printed for the final match.
1793    
1794        (g) Added the -H (--with-filename) option.
1795    
1796        (h) The combination of options -rh failed to suppress file names for files
1797            that were found from directory arguments.
1798    
1799        (i) Added the -D (--devices) and -d (--directories) options.
1800    
1801        (j) Added the -F (--fixed-strings) option.
1802    
1803        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1804    
1805        (l) Added the --colo(u)r option.
1806    
1807        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1808            is not present by default.
1809    
1810     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1811        items such as (?R) or (?1), when the recursion could match a number of
1812        alternatives. If it matched one of the alternatives, but subsequently,
1813        outside the recursion, there was a failure, the code tried to back up into
1814        the recursion. However, because of the way PCRE is implemented, this is not
1815        possible, and the result was an incorrect result from the match.
1816    
1817        In order to prevent this happening, the specification of recursion has
1818        been changed so that all such subpatterns are automatically treated as
1819        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1820    
1821     4. I had overlooked the fact that, in some locales, there are characters for
1822        which isalpha() is true but neither isupper() nor islower() are true. In
1823        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1824        and ordfeminine) are like this. This affected the treatment of \w and \W
1825        when they appeared in character classes, but not when they appeared outside
1826        a character class. The bit map for "word" characters is now created
1827        separately from the results of isalnum() instead of just taking it from the
1828        upper, lower, and digit maps. (Plus the underscore character, of course.)
1829    
1830     5. The above bug also affected the handling of POSIX character classes such as
1831        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1832        permanent tables. Instead, the bit maps for such a class were previously
1833        created as the appropriate unions of the upper, lower, and digit bitmaps.
1834        Now they are created by subtraction from the [[:word:]] class, which has
1835        its own bitmap.
1836    
1837     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1838        It is created by subtracting the vertical space characters (\x09, \x0a,
1839        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1840        subtraction was done in the overall bitmap for a character class, meaning
1841        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1842        be recognized. This bug has been fixed.
1843    
1844     7. Patches from the folks at Google:
1845    
1846          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1847          real life, but is still worth protecting against".
1848    
1849          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1850          regular expressions".
1851    
1852          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1853          have it.
1854    
1855          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1856          "configure" and the latter not, in order to fix a problem somebody had
1857          with compiling the Arg class on HP-UX.
1858    
1859          (e) Improve the error-handling of the C++ wrapper a little bit.
1860    
1861          (f) New tests for checking recursion limiting.
1862    
1863     8. The pcre_memmove() function, which is used only if the environment does not
1864        have a standard memmove() function (and is therefore rarely compiled),
1865        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1866        returning a result (though PCRE never actually uses the result).
1867    
1868     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1869        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1870        returned instead of calling malloc() with an overflowing number that would
1871        most likely cause subsequent chaos.
1872    
1873    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1874    
1875    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1876        with this option is matched, the nmatch and pmatch options of regexec() are
1877        ignored.
1878    
1879    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1880        provided in case anyone wants to the the POSIX interface with UTF-8
1881        strings.
1882    
1883    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1884        C++ linking (needed for some HP-UX environments).
1885    
1886    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1887        (unused parameter) and in the pcre_printint() function (omitted "default"
1888        switch label when the default is to do nothing).
1889    
1890    15. Added some code to make it possible, when PCRE is compiled as a C++
1891        library, to replace subject pointers for pcre_exec() with a smart pointer
1892        class, thus making it possible to process discontinuous strings.
1893    
1894    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1895        much the same function. They were added by different people who were trying
1896        to make PCRE easy to compile on non-Unix systems. It has been suggested
1897        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1898        for compiling on Windows systems. I have therefore replaced it with
1899        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1900        defaults to "extern" for C or "extern C" for C++, which works fine on
1901        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1902        SCOPE with something explicit in config.h. In addition:
1903    
1904        (a) pcreposix.h still had just "extern" instead of either of these macros;
1905            I have replaced it with PCRE_DATA_SCOPE.
1906    
1907        (b) Functions such as _pcre_xclass(), which are internal to the library,
1908            but external in the C sense, all had PCRE_EXPORT in their definitions.
1909            This is apparently wrong for the Windows case, so I have removed it.
1910            (It makes no difference on Unix-like systems.)
1911    
1912    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1913        of recursive calls to match(). This is different to MATCH_LIMIT because
1914        that limits the total number of calls to match(), not all of which increase
1915        the depth of recursion. Limiting the recursion depth limits the amount of
1916        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1917        when PCRE is compiled, and changed at run time. A patch from Google adds
1918        this functionality to the C++ interface.
1919    
1920    18. Changes to the handling of Unicode character properties:
1921    
1922        (a) Updated the table to Unicode 4.1.0.
1923    
1924        (b) Recognize characters that are not in the table as "Cn" (undefined).
1925    
1926        (c) I revised the way the table is implemented to a much improved format
1927            which includes recognition of ranges. It now supports the ranges that
1928            are defined in UnicodeData.txt, and it also amalgamates other
1929            characters into ranges. This has reduced the number of entries in the
1930            table from around 16,000 to around 3,000, thus reducing its size
1931            considerably. I realized I did not need to use a tree structure after
1932            all - a binary chop search is just as efficient. Having reduced the
1933            number of entries, I extended their size from 6 bytes to 8 bytes to
1934            allow for more data.
1935    
1936        (d) Added support for Unicode script names via properties such as \p{Han}.
1937    
1938    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1939        matching that character.
1940    
1941    20. When matching a repeated Unicode property with a minimum greater than zero,
1942        (for example \pL{2,}), PCRE could look past the end of the subject if it
1943        reached it while seeking the minimum number of characters. This could
1944        happen only if some of the characters were more than one byte long, because
1945        there is a check for at least the minimum number of bytes.
1946    
1947    21. Refactored the implementation of \p and \P so as to be more general, to
1948        allow for more different types of property in future. This has changed the
1949        compiled form incompatibly. Anybody with saved compiled patterns that use
1950        \p or \P will have to recompile them.
1951    
1952    22. Added "Any" and "L&" to the supported property types.
1953    
1954    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1955        but give a compile time error if the value is greater than 0xff.
1956    
1957    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1958        accidentally not being installed or uninstalled.
1959    
1960    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1961        made were to insert the current release number. This seemed silly, because
1962        it made things harder for people building PCRE on systems that don't run
1963        "configure". I have turned pcre.h into a distributed file, no longer built
1964        by "configure", with the version identification directly included. There is
1965        no longer a pcre.h.in file.
1966    
1967        However, this change necessitated a change to the pcre-config script as
1968        well. It is built from pcre-config.in, and one of the substitutions was the
1969        release number. I have updated configure.ac so that ./configure now finds
1970        the release number by grepping pcre.h.
1971    
1972    26. Added the ability to run the tests under valgrind.
1973    
1974    
1975    Version 6.4 05-Sep-05
1976    ---------------------
1977    
1978     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1979        "--" to be printed when multiple files were scanned, even when none of the
1980        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1981        consider it to be a bug, and have restored the previous behaviour.
1982    
1983     2. A couple of code tidies to get rid of compiler warnings.
1984    
1985     3. The pcretest program used to cheat by referring to symbols in the library
1986        whose names begin with _pcre_. These are internal symbols that are not
1987        really supposed to be visible externally, and in some environments it is
1988        possible to suppress them. The cheating is now confined to including
1989        certain files from the library's source, which is a bit cleaner.
1990    
1991     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1992        file's purpose clearer.
1993    
1994     5. Reorganized pcre_ucp_findchar().
1995    
1996    
1997    Version 6.3 15-Aug-05
1998    ---------------------
1999    
2000     1. The file libpcre.pc.in did not have general read permission in the tarball.
2001    
2002     2. There were some problems when building without C++ support:
2003    
2004        (a) If C++ support was not built, "make install" and "make test" still
2005            tried to test it.
2006    
2007        (b) There were problems when the value of CXX was explicitly set. Some
2008            changes have been made to try to fix these, and ...
2009    
2010        (c) --disable-cpp can now be used to explicitly disable C++ support.
2011    
2012        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2013            backslash in a target when C++ was disabled. This confuses some
2014            versions of "make", apparently. Using an intermediate variable solves
2015            this. (Same for CPP_LOBJ.)
2016    
2017     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2018        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2019        necessary on certain architectures.
2020    
2021     4. Added a setting of -export-symbols-regex to the link command to remove
2022        those symbols that are exported in the C sense, but actually are local
2023        within the library, and not documented. Their names all begin with
2024        "_pcre_". This is not a perfect job, because (a) we have to except some
2025        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2026        available (and never for static libraries). I have made a note to try to
2027        find a way round (a) in the future.
2028    
2029    
2030    Version 6.2 01-Aug-05
2031    ---------------------
2032    
2033     1. There was no test for integer overflow of quantifier values. A construction
2034        such as {1111111111111111} would give undefined results. What is worse, if
2035        a minimum quantifier for a parenthesized subpattern overflowed and became
2036        negative, the calculation of the memory size went wrong. This could have
2037        led to memory overwriting.
2038    
2039     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2040    
2041     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2042        operating environments where this matters.
2043    
2044     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2045        PCRE options from within the C++ wrapper.
2046    
2047     5. Named capturing subpatterns were not being correctly counted when a pattern
2048        was compiled. This caused two problems: (a) If there were more than 100
2049        such subpatterns, the calculation of the memory needed for the whole
2050        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2051        back references of the form \12, where the number was greater than 9, were
2052        not recognized as back references, even though there were sufficient
2053        previous subpatterns.
2054    
2055     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2056        versions of gcc, e.g. 2.95.4.
2057    
2058    
2059    Version 6.1 21-Jun-05
2060    ---------------------
2061    
2062     1. There was one reference to the variable "posix" in pcretest.c that was not
2063        surrounded by "#if !defined NOPOSIX".
2064    
2065     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2066        the cross-check on the old pcre_info() function, for the benefit of the
2067        cut-down version of PCRE that is currently imported into Exim.
2068    
2069     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2070        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2071        patterns that start (?i) but I don't think that matters. The use of (?i) is
2072        just an example; this all applies to the other options as well.
2073    
2074     4. Since libtool seems to echo the compile commands it is issuing, the output
2075        from "make" can be reduced a bit by putting "@" in front of each libtool
2076        compile command.
2077    
2078     5. Patch from the folks at Google for configure.in to be a bit more thorough
2079        in checking for a suitable C++ installation before trying to compile the
2080        C++ stuff. This should fix a reported problem when a compiler was present,
2081        but no suitable headers.
2082    
2083     6. The man pages all had just "PCRE" as their title. I have changed them to
2084        be the relevant file name. I have also arranged that these names are
2085        retained in the file doc/pcre.txt, which is a concatenation in text format
2086        of all the man pages except the little individual ones for each function.
2087    
2088     7. The NON-UNIX-USE file had not been updated for the different set of source
2089        files that come with release 6. I also added a few comments about the C++
2090        wrapper.
2091    
2092    
2093    Version 6.0 07-Jun-05
2094    ---------------------
2095    
2096     1. Some minor internal re-organization to help with my DFA experiments.
2097    
2098     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2099        didn't matter for the library itself when fully configured, but did matter
2100        when compiling without UCP support, or within Exim, where the ucp files are
2101        not imported.
2102    
2103     3. Refactoring of the library code to split up the various functions into
2104        different source modules. The addition of the new DFA matching code (see
2105        below) to a single monolithic source would have made it really too
2106        unwieldy, quite apart from causing all the code to be include in a
2107        statically linked application, when only some functions are used. This is
2108        relevant even without the DFA addition now that patterns can be compiled in
2109        one application and matched in another.
2110    
2111        The downside of splitting up is that there have to be some external
2112        functions and data tables that are used internally in different modules of
2113        the library but which are not part of the API. These have all had their
2114        names changed to start with "_pcre_" so that they are unlikely to clash
2115        with other external names.
2116    
2117     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2118        a different (DFA) algorithm. Although it is slower than the original
2119        function, it does have some advantages for certain types of matching
2120        problem.
2121    
2122     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2123        including restarting after a partial match.
2124    
2125     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2126        defined when compiling for Windows was sent to me. I have put it into the
2127        code, though I have no means of testing or verifying it.
2128    
2129     7. Added the pcre_refcount() auxiliary function.
2130    
2131     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2132        match before or at the first newline in the subject string. In pcretest,
2133        the /f option on a pattern can be used to set this.
2134    
2135     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2136        would behave wrongly. This has been present in PCRE since release 4.0.
2137    
2138    10. A number of changes to the pcregrep command:
2139    
2140        (a) Refactored how -x works; insert ^(...)$ instead of setting
2141            PCRE_ANCHORED and checking the length, in preparation for adding
2142            something similar for -w.
2143    
2144        (b) Added the -w (match as a word) option.
2145    
2146        (c) Refactored the way lines are read and buffered so as to have more
2147            than one at a time available.
2148    
2149        (d) Implemented a pcregrep test script.
2150    
2151        (e) Added the -M (multiline match) option. This allows patterns to match
2152            over several lines of the subject. The buffering ensures that at least
2153            8K, or the rest of the document (whichever is the shorter) is available
2154            for matching (and similarly the previous 8K for lookbehind assertions).
2155    
2156        (f) Changed the --help output so that it now says
2157    
2158              -w, --word-regex(p)
2159    
2160            instead of two lines, one with "regex" and the other with "regexp"
2161            because that confused at least one person since the short forms are the
2162            same. (This required a bit of code, as the output is generated
2163            automatically from a table. It wasn't just a text change.)
2164    
2165        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2166            option but starts with a hyphen. Could be a pattern or a path name
2167            starting with a hyphen, for instance.
2168    
2169        (h) "-" can be given as a file name to represent stdin.
2170    
2171        (i) When file names are being printed, "(standard input)" is used for
2172            the standard input, for compatibility with GNU grep. Previously
2173            "<stdin>" was used.
2174    
2175        (j) The option --label=xxx can be used to supply a name to be used for
2176            stdin when file names are being printed. There is no short form.
2177    
2178        (k) Re-factored the options decoding logic because we are going to add
2179            two more options that take data. Such options can now be given in four
2180            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2181    
2182        (l) Added the -A, -B, and -C options for requesting that lines of context
2183            around matches be printed.
2184    
2185        (m) Added the -L option to print the names of files that do not contain
2186            any matching lines, that is, the complement of -l.
2187    
2188        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2189            continue to scan other files.
2190    
2191        (o) The -s option was incorrectly implemented. For compatibility with other
2192            greps, it now suppresses the error message for a non-existent or non-
2193            accessible file (but not the return code). There is a new option called
2194            -q that suppresses the output of matching lines, which was what -s was
2195            previously doing.
2196    
2197        (p) Added --include and --exclude options to specify files for inclusion
2198            and exclusion when recursing.
2199    
2200    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2201        Hopefully, it now does.
2202    
2203    12. Missing cast in pcre_study().
2204    
2205    13. Added an "uninstall" target to the makefile.
2206    
2207    14. Replaced "extern" in the function prototypes in Makefile.in with
2208        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2209        world, but is set differently for Windows.
2210    
2211    15. Added a second compiling function called pcre_compile2(). The only
2212        difference is that it has an extra argument, which is a pointer to an
2213        integer error code. When there is a compile-time failure, this is set
2214        non-zero, in addition to the error test pointer being set to point to an
2215        error message. The new argument may be NULL if no error number is required
2216        (but then you may as well call pcre_compile(), which is now just a
2217        wrapper). This facility is provided because some applications need a
2218        numeric error indication, but it has also enabled me to tidy up the way
2219        compile-time errors are handled in the POSIX wrapper.
2220    
2221    16. Added VPATH=.libs to the makefile; this should help when building with one
2222        prefix path and installing with another. (Or so I'm told by someone who
2223        knows more about this stuff than I do.)
2224    
2225    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2226        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2227        match everything, including newlines. This is not POSIX-compatible, but
2228        somebody wanted the feature. From pcretest it can be activated by using
2229        both the P and the s flags.
2230    
2231    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2232    
2233    19. libpcre.pc was being incorrectly installed as executable.
2234    
2235    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2236        it now also looks for '\r' so that it will work unmodified on Windows.
2237    
2238    21. Added Google's contributed C++ wrapper to the distribution.
2239    
2240    22. Added some untidy missing memory free() calls in pcretest, to keep
2241        Electric Fence happy when testing.
2242    
2243    
2244    
2245    Version 5.0 13-Sep-04
2246    ---------------------
2247    
2248     1. Internal change: literal characters are no longer packed up into items
2249        containing multiple characters in a single byte-string. Each character
2250        is now matched using a separate opcode. However, there may be more than one
2251        byte in the character in UTF-8 mode.
2252    
2253     2. The pcre_callout_block structure has two new fields: pattern_position and
2254        next_item_length. These contain the offset in the pattern to the next match
2255        item, and its length, respectively.
2256    
2257     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2258        insertion of callouts before each pattern item. Added the /C option to
2259        pcretest to make use of this.
2260    
2261     4. On the advice of a Windows user, the lines
2262    
2263          #if defined(_WIN32) || defined(WIN32)
2264          _setmode( _fileno( stdout ), 0x8000 );
2265          #endif  /* defined(_WIN32) || defined(WIN32) */
2266    
2267        have been added to the source of pcretest. This apparently does useful
2268        magic in relation to line terminators.
2269    
2270     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2271        for the benefit of those environments where the "b" makes a difference.
2272    
2273     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2274        to know about it. I have put a hack into configure.in that adds in code
2275        to set GCC=yes if CC=icc. This seems to end up at a point in the
2276        generated configure script that is early enough to affect the setting of
2277        compiler options, which is what is needed, but I have no means of testing
2278        whether it really works. (The user who reported this had patched the
2279        generated configure script, which of course I cannot do.)
2280    
2281        LATER: After change 22 below (new libtool files), the configure script
2282        seems to know about icc (and also ecc). Therefore, I have commented out
2283        this hack in configure.in.
2284    
2285     7. Added support for pkg-config (2 patches were sent in).
2286    
2287     8. Negated POSIX character classes that used a combination of internal tables
2288        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2289        [[:^ascii]]. Typically, they would match almost any characters. The other
2290        POSIX classes were not broken in this way.
2291    
2292     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2293        to find the match, as PCRE was deluded into thinking that the match had to
2294        start at the start point or following a newline. The same bug applied to
2295        patterns with negative forward assertions or any backward assertions
2296        preceding ".*" at the start, unless the pattern required a fixed first
2297        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2298    
2299    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2300        starting at the last subject character, bytes beyond the end of the subject
2301        string were read.
2302    
2303    11. Renamed the variable "class" as "classbits" to make life easier for C++
2304        users. (Previously there was a macro definition, but it apparently wasn't
2305        enough.)
2306    
2307    12. Added the new field "tables" to the extra data so that tables can be passed
2308        in at exec time, or the internal tables can be re-selected. This allows
2309        a compiled regex to be saved and re-used at a later time by a different
2310        program that might have everything at different addresses.
2311    
2312    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2313        -R library as well as a -L library.
2314    
2315    14. The debugging options of pcretest (-d on the command line or D on a
2316        pattern) showed incorrect output for anything following an extended class
2317        that contained multibyte characters and which was followed by a quantifier.
2318    
2319    15. Added optional support for general category Unicode character properties
2320        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2321        support. It adds about 90K to the size of the library. The meanings of the
2322        inbuilt class escapes such as \d and \s have NOT been changed.
2323    
2324    16. Updated pcredemo.c to include calls to free() to release the memory for the
2325        compiled pattern.
2326    
2327    17. The generated file chartables.c was being created in the source directory
2328        instead of in the building directory. This caused the build to fail if the
2329        source directory was different from the building directory, and was
2330        read-only.
2331    
2332    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2333        file. No doubt somebody will tell me if they don't make sense... Also added
2334        Dan Mooney's comments about building on OpenVMS.
2335    
2336    19. Added support for partial matching via the PCRE_PARTIAL option for
2337        pcre_exec() and the \P data escape in pcretest.
2338    
2339    20. Extended pcretest with 3 new pattern features:
2340    
2341        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2342              write the compiled pattern to the file whose name is "rest-of-line".
2343              This is a straight binary dump of the data, with the saved pointer to
2344              the character tables forced to be NULL. The study data, if any, is
2345              written too. After writing, pcretest reads a new pattern.
2346    
2347        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2348              compiled pattern from the given file. There must not be any
2349              occurrences of "<" in the file name (pretty unlikely); if there are,
2350              pcretest will instead treat the initial "<" as a pattern delimiter.
2351              After reading in the pattern, pcretest goes on to read data lines as
2352              usual.
2353    
2354        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2355              and 16-bit fields in a compiled pattern, to simulate a pattern that
2356              was compiled on a host of opposite endianness.
2357    
2358    21. The pcre-exec() function can now cope with patterns that were compiled on
2359        hosts of opposite endianness, with this restriction:
2360    
2361          As for any compiled expression that is saved and used later, the tables
2362          pointer field cannot be preserved; the extra_data field in the arguments
2363          to pcre_exec() should be used to pass in a tables address if a value
2364          other than the default internal tables were used at compile time.
2365    
2366    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2367        now diagnosed as an error. Previously, most of the time, a negative number
2368        would have been treated as zero, but if in addition "ovector" was passed as
2369        NULL, a crash could occur.
2370    
2371    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2372        new versions from the libtool 1.5 distribution (the last one is a copy of
2373        a file called libtool.m4). This seems to have fixed the need to patch
2374        "configure" to support Darwin 1.3 (which I used to do). However, I still
2375        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2376        workstation).
2377    
2378    24. Changed the PCRE licence to be the more standard "BSD" licence.
2379    
2380    
2381    Version 4.5 01-Dec-03
2382    ---------------------
2383    
2384     1. There has been some re-arrangement of the code for the match() function so
2385        that it can be compiled in a version that does not call itself recursively.
2386        Instead, it keeps those local variables that need separate instances for
2387        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2388        needs to "recurse". Keeping track of where control must go is done by means
2389        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2390        hide most of the details from the main code, and operates only if
2391        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2392        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2393        operating.
2394    
2395        To make it easier for callers to provide specially tailored get/free
2396        functions for this usage, two new functions, pcre_stack_malloc, and
2397        pcre_stack_free, are used. They are always called in strict stacking order,
2398        and the size of block requested is always the same.
2399    
2400        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2401        PCRE has been compiled to use the stack or the heap for recursion. The
2402        -C option of pcretest uses this to show which version is compiled.
2403    
2404        A new data escape \S, is added to pcretest; it causes the amounts of store
2405        obtained and freed by both kinds of malloc/free at match time to be added
2406        to the output.
2407    
2408     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2409        what's available on my current Linux desktop machine.
2410    
2411     3. When matching a UTF-8 string, the test for a valid string at the start has
2412        been extended. If start_offset is not zero, PCRE now checks that it points
2413        to a byte that is the start of a UTF-8 character. If not, it returns
2414        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2415        this is necessary because there may be backward assertions in the pattern.
2416        When matching the same subject several times, it may save resources to use
2417        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2418    
2419     4. The code for checking the validity of UTF-8 strings has been tightened so
2420        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2421        containing "overlong sequences".
2422    
2423     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2424        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2425        should have been "&", but it just so happened that all the cases this let
2426        through by mistake were picked up later in the function.
2427    
2428     6. I had used a variable called "isblank" - this is a C99 function, causing
2429        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2430    
2431     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2432        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2433        so the version is shown; (c) stop "make test" echoing "./RunTest".
2434    
2435     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2436    
2437     9. The prototype for memmove() for systems that don't have it was using
2438        size_t, but the inclusion of the header that defines size_t was later. I've
2439        moved the #includes for the C headers earlier to avoid this.
2440    
2441    10. Added some adjustments to the code to make it easier to compiler on certain
2442        special systems:
2443    
2444          (a) Some "const" qualifiers were missing.
2445          (b) Added the macro EXPORT before all exported functions; by default this
2446              is defined to be empty.
2447          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2448              that it reads its output file name as an argument instead of writing
2449              to the standard output and assuming this can be redirected.
2450    
2451    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2452        class containing characters with values greater than 255, PCRE compilation
2453        went into a loop.
2454    
2455    12. A recursive reference to a subpattern that was within another subpattern
2456        that had a minimum quantifier of zero caused PCRE to crash. For example,
2457        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2458        recursion. If the recursively-called subpattern itself had a zero repeat,
2459        that was OK.
2460    
2461    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2462        buffer into which it was copied (for escape processing) was still set at
2463        1024, so long lines caused crashes.
2464    
2465    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2466        "internal error: code overflow...". This applied to any character class
2467        that was followed by a possessive quantifier.
2468    
2469    15. Modified the Makefile to add libpcre.la as a prerequisite for
2470        libpcreposix.la because I was told this is needed for a parallel build to
2471        work.
2472    
2473    16. If a pattern that contained .* following optional items at the start was
2474        studied, the wrong optimizing data was generated, leading to matching
2475        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2476        matching string must start with a or b or c. The correct conclusion for
2477        this pattern is that a match can start with any character.
2478    
2479    
2480    Version 4.4 13-Aug-03
2481    ---------------------
2482    
2483     1. In UTF-8 mode, a character class containing characters with values between
2484        127 and 255 was not handled correctly if the compiled pattern was studied.
2485        In fixing this, I have also improved the studying algorithm for such
2486        classes (slightly).
2487    
2488     2. Three internal functions had redundant arguments passed to them. Removal
2489        might give a very teeny performance improvement.
2490    
2491     3. Documentation bug: the value of the capture_top field in a callout is *one
2492        more than* the number of the hightest numbered captured substring.
2493    
2494     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2495        in incorrectly linking with a previously installed version. They now link
2496        explicitly with libpcre.la.
2497    
2498     5. configure.in no longer needs to recognize Cygwin specially.
2499    
2500     6. A problem in pcre.in for Windows platforms is fixed.
2501    
2502     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2503        pcretest, it used to include the size of the study block as part of its
2504        output. Unfortunately, the structure contains a field that has a different
2505        size on different hardware architectures. This meant that the tests that
2506        showed this size failed. As the block is currently always of a fixed size,
2507        this information isn't actually particularly useful in pcretest output, so
2508        I have just removed it.
2509    
2510     8. Three pre-processor statements accidentally did not start in column 1.
2511        Sadly, there are *still* compilers around that complain, even though
2512        standard C has not required this for well over a decade. Sigh.
2513    
2514     9. In pcretest, the code for checking callouts passed small integers in the
2515        callout_data field, which is a void * field. However, some picky compilers
2516        complained about the casts involved for this on 64-bit systems. Now
2517        pcretest passes the address of the small integer instead, which should get
2518        rid of the warnings.
2519    
2520    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2521        both compile and run time, and gives an error if an invalid UTF-8 sequence
2522        is found. There is a option for disabling this check in cases where the
2523        string is known to be correct and/or the maximum performance is wanted.
2524    
2525    11. In response to a bug report, I changed one line in Makefile.in from
2526    
2527            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2528        to
2529            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2530    
2531        to look similar to other lines, but I have no way of telling whether this
2532        is the right thing to do, as I do not use Windows. No doubt I'll get told
2533        if it's wrong...
2534    
2535    
2536    Version 4.3 21-May-03
2537    ---------------------
2538    
2539    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2540       Makefile.
2541    
2542    2. Some refactoring to improve the quality of the code:
2543    
2544       (i)   The utf8_table... variables are now declared "const".
2545    
2546       (ii)  The code for \cx, which used the "case flipping" table to upper case
2547             lower case letters, now just substracts 32. This is ASCII-specific,
2548             but the whole concept of \cx is ASCII-specific, so it seems
2549             reasonable.
2550    
2551       (iii) PCRE was using its character types table to recognize decimal and
2552             hexadecimal digits in the pattern. This is silly, because it handles
2553             only 0-9, a-f, and A-F, but the character types table is locale-
2554             specific, which means strange things might happen. A private
2555             table is now used for this - though it costs 256 bytes, a table is
2556             much faster than multiple explicit tests. Of course, the standard
2557             character types table is still used for matching digits in subject
2558             strings against \d.
2559    
2560       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2561             ending in _t are). So I've renamed it as ESC_tee.
2562    
2563    3. The first argument for regexec() in the POSIX wrapper should have been
2564       defined as "const".
2565    
2566    4. Changed pcretest to use malloc() for its buffers so that they can be
2567       Electric Fenced for debugging.
2568    
2569    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2570       to read one or more bytes before the start of the subject string. Often this
2571       had no effect on PCRE's behaviour, but in some circumstances it could
2572       provoke a segmentation fault.
2573    
2574    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2575       to try to read one or more bytes before the start of the subject string.
2576    
2577    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2578       UTF-8 support could misbehave in various ways if the subject string
2579       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2580       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2581       back over UTF-8 characters.)
2582    
2583    
2584    Version 4.2 14-Apr-03
2585    ---------------------
2586    
2587    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2588    
2589    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2590         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2591         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2592         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2593         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2594           and BUILD_EXEEXT
2595         Note: automatic setting of the BUILD variables is not yet working
2596         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2597           compile-time but not at link-time
2598         [LINK]: use for linking executables only
2599         make different versions for Windows and non-Windows
2600         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2601           libraries
2602         [LINK_FOR_BUILD]: new variable
2603         [OBJEXT]: use throughout
2604         [EXEEXT]: use throughout
2605         <winshared>: new target
2606         <wininstall>: new target
2607         <dftables.o>: use native compiler
2608         <dftables>: use native linker
2609         <install>: handle Windows platform correctly
2610         <clean>: ditto
2611         <check>: ditto
2612         copy DLL to top builddir before testing
2613    
2614       As part of these changes, -no-undefined was removed again. This was reported
2615       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2616       in any case.
2617    
2618    3. Some tidies to get rid of compiler warnings:
2619    
2620       . In the match_data structure, match_limit was an unsigned long int, whereas
2621         match_call_count was an int. I've made them both unsigned long ints.
2622    
2623       . In pcretest the fact that a const uschar * doesn't automatically cast to
2624         a void * provoked a warning.
2625    
2626       . Turning on some more compiler warnings threw up some "shadow" variables
2627         and a few more missing casts.
2628    
2629    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2630       option, a class that contained a single character with a value between 128
2631       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2632    
2633    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2634       option, a class that contained several characters, but with at least one
2635       whose value was between 128 and 255 caused PCRE to crash.
2636    
2637    
2638    Version 4.1 12-Mar-03
2639    ---------------------
2640    
2641    1. Compiling with gcc -pedantic found a couple of places where casts were
2642    needed, and a string in dftables.c that was longer than standard compilers are
2643    required to support.
2644    
2645    2. Compiling with Sun's compiler found a few more places where the code could
2646    be tidied up in order to avoid warnings.
2647    
2648    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2649    first of these names is deprecated in the latest Autoconf in favour of the name
2650    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2651    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2652    analogy I have changed it to CFLAGS_FOR_BUILD.
2653    
2654    4. Added -no-undefined to the linking command in the Makefile, because this is
2655    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2656    linking step for the pcreposix library.
2657    
2658    5. PCRE was failing to diagnose the case of two named groups with the same
2659    name.
2660    
2661    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2662    literal character that is needed in the subject for a match, and scans along to
2663    ensure that it is present before embarking on the full matching process. This
2664    saves time in cases of nested unlimited repeats that are never going to match.
2665    Problem: the scan can take a lot of time if the subject is very long (e.g.
2666    megabytes), thus penalizing straightforward matches. It is now done only if the
2667    amount of subject to be scanned is less than 1000 bytes.
2668    
2669    7. A lesser problem with the same optimization is that it was recording the
2670    first character of an anchored pattern as "needed", thus provoking a search
2671    right along the subject, even when the first match of the pattern was going to
2672    fail. The "needed" character is now not set for anchored patterns, unless it
2673    follows something in the pattern that is of non-fixed length. Thus, it still
2674    fulfils its original purpose of finding quick non-matches in cases of nested
2675    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2676    
2677    
2678    Version 4.0 17-Feb-03
2679    ---------------------
2680    
2681    1. If a comment in an extended regex that started immediately after a meta-item
2682    extended to the end of string, PCRE compiled incorrect data. This could lead to
2683    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2684    
2685    2. Moved to autoconf 2.53 and libtool 1.4.2.
2686    
2687    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2688    the special perltest8 script is no longer needed - all the tests can be run
2689    from a single perltest script.
2690    
2691    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2692    by \s. It has now been removed in PCRE. This means it isn't recognized as
2693    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2694    class [:space:] *does* include VT, thereby creating a mess.
2695    
2696    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2697    space and tab.
2698    
2699    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2700    its new features into the main test script, reducing the number of scripts.
2701    
2702    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2703    were backward compatible, and made the (?i) apply to the whole pattern, as if
2704    /i were given. Now it behaves more logically, and applies the option setting
2705    only to what follows. PCRE has been changed to follow suit. However, if it
2706    finds options settings right at the start of the pattern, it extracts them into
2707    the global options, as before. Thus, they show up in the info data.
2708    
2709    8. Added support for the \Q...\E escape sequence. Characters in between are
2710    treated as literals. This is slightly different from Perl in that $ and @ are
2711    also handled as literals inside the quotes. In Perl, they will cause variable
2712    interpolation. Note the following examples:
2713    
2714        Pattern            PCRE matches      Perl matches
2715    
2716        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2717        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2718        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2719    
2720    For compatibility with Perl, \Q...\E sequences are recognized inside character
2721    classes as well as outside them.
2722    
2723    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2724    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2725    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2726    signed/unsigned warnings.
2727    
2728    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2729    option for pcretest, so I've replaced it by a simple function that does just
2730    that job.
2731    
2732    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2733    "pcregrep -".
2734    
2735    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2736    Java package. This provides some syntactic sugar for simple cases of what my
2737    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2738    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2739    item, you can use this simplified notation. Note that only makes sense with
2740    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2741    greediness, whatever the setting of the PCRE_UNGREEDY option.
2742    
2743    13. A change of greediness default within a pattern was not taking effect at
2744    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2745    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2746    was abstracted outside.
2747    
2748    14. PCRE now supports the \G assertion. It is true when the current matching
2749    position is at the start point of the match. This differs from \A when the
2750    starting offset is non-zero. Used with the /g option of pcretest (or similar
2751    code), it works in the same way as it does for Perl's /g option. If all
2752    alternatives of a regex begin with \G, the expression is anchored to the start
2753    match position, and the "anchored" flag is set in the compiled expression.
2754    
2755    15. Some bugs concerning the handling of certain option changes within patterns
2756    have been fixed. These applied to options other than (?ims). For example,
2757    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2758    been the other way round. Some of this was related to change 7 above.
2759    
2760    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2761    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2762    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2763    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2764    
2765    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2766    mode. Unlike ".", it always matches newline, whatever the setting of
2767    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2768    assertions. Perl allows it, but it doesn't (in general) work because it can't
2769    calculate the length of the lookbehind. At least, that's the case for Perl
2770    5.8.0 - I've been told they are going to document that it doesn't work in
2771    future.
2772    
2773    18. Added an error diagnosis for escapes that PCRE does not support: these are
2774    \L, \l, \N, \P, \p, \U, \u, and \X.
2775    
2776    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2777    reading past the end of the pattern in cases such as /[abcd/.
2778    
2779    20. PCRE was getting more memory than necessary for patterns with classes that
2780    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2781    
2782    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2783    compiling PCRE for use with Virtual Pascal.
2784    
2785    22. Small fix to the Makefile to make it work properly if the build is done
2786    outside the source tree.
2787    
2788    23. Added a new extension: a condition to go with recursion. If a conditional
2789    subpattern starts with (?(R) the "true" branch is used if recursion has
2790    happened, whereas the "false" branch is used only at the top level.
2791    
2792    24. When there was a very long string of literal characters (over 255 bytes
2793    without UTF support, over 250 bytes with UTF support), the computation of how
2794    much memory was required could be incorrect, leading to segfaults or other
2795    strange effects.
2796    
2797    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2798    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2799    there was a subsequent back reference to those brackets. This meant that, for
2800    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2801    possible to check for precisely this case. All we can do is abandon the
2802    optimization if .* occurs inside capturing brackets when there are any back
2803    references whatsoever. (See below for a better fix that came later.)
2804    
2805    26. The handling of the optimization for finding the first character of a
2806    non-anchored pattern, and for finding a character that is required later in the
2807    match were failing in some cases. This didn't break the matching; it just
2808    failed to optimize when it could. The way this is done has been re-implemented.
2809    
2810    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2811    
2812    28. Added a new feature that provides some of the functionality that Perl
2813    provides with (?{...}). The facility is termed a "callout". The way it is done
2814    in PCRE is for the caller to provide an optional function, by setting
2815    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2816    global variable. By default it is unset, which disables all calling out. To get
2817    the function called, the regex must include (?C) at appropriate points. This
2818    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2819    This provides a means of identifying different callout points. When PCRE
2820    reaches such a point in the regex, if pcre_callout has been set, the external
2821    function is called. It is provided with data in a structure called
2822    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2823    matching continues; if it returns a non-zero value, the match at the current
2824    point fails. However, backtracking will occur if possible. [This was changed
2825    later and other features added - see item 49 below.]
2826    
2827    29. pcretest is upgraded to test the callout functionality. It provides a
2828    callout function that displays information. By default, it shows the start of
2829    the match and the current position in the text. There are some new data escapes
2830    to vary what happens:
2831    
2832        \C+         in addition, show current contents of captured substrings
2833        \C-         do not supply a callout function
2834        \C!n        return 1 when callout number n is reached
2835        \C!n!m      return 1 when callout number n is reached for the mth time
2836    
2837    30. If pcregrep was called with the -l option and just a single file name, it
2838    output "<stdin>" if a match was found, instead of the file name.
2839    
2840    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2841    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2842    pcre_exec(). This saves a malloc/free per call. The default value of
2843    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2844    when configuring.
2845    
2846    32. The default maximum size of a compiled pattern is 64K. There have been a
2847    few cases of people hitting this limit. The code now uses macros to handle the
2848    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2849    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2850    configuring. Tests 2 and 5 work only with 2-byte links because they output
2851    debugging information about compiled patterns.
2852    
2853    33. Internal code re-arrangements:
2854    
2855    (a) Moved the debugging function for printing out a compiled regex into
2856        its own source file (printint.c) and used #include to pull it into
2857        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2858        separate copies.
2859    
2860    (b) Defined the list of op-code names for debugging as a macro in
2861        internal.h so that it is next to the definition of the opcodes.
2862    
2863    (c) Defined a table of op-code lengths for simpler skipping along compiled
2864        code. This is again a macro in internal.h so that it is next to the
2865        definition of the opcodes.
2866    
2867    34. Added support for recursive calls to individual subpatterns, along the
2868    lines of Robin Houston's patch (but implemented somewhat differently).
2869    
2870    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2871    allow it to read and process whole directories in Win32. This code was
2872    contributed by Lionel Fourquaux; it has not been tested by me.
2873    
2874    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2875    used to name a group. Names consist of alphanumerics and underscores, and must
2876    be unique. Back references use the syntax (?P=name) and recursive calls use
2877    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2878    numbers. The function pcre_fullinfo() can be used after compilation to extract
2879    a name/number map. There are three relevant calls:
2880    
2881      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2882      PCRE_INFO_NAMECOUNT            yields the number of entries
2883      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2884    
2885    The map is a vector of fixed-size entries. The size of each entry depends on
2886    the length of the longest name used. The first two bytes of each entry are the
2887    group number, most significant byte first. There follows the corresponding
2888    name, zero terminated. The names are in alphabetical order.
2889    
2890    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2891    case instead of 255. Making it the same both with and without UTF-8 support
2892    means that the same test output works with both.
2893    
2894    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2895    calling malloc() with a zero argument.
2896    
2897    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2898    optimization. I've improved things by keeping a bitmap of backreferences with
2899    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2900    fact referenced, the optimization can be applied. It is unlikely that a
2901    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2902    the match to follow \n) will appear inside brackets with a number greater than
2903    31, but if it does, any back reference > 31 suppresses the optimization.
2904    
2905    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2906    of disabling numbered capturing parentheses. Any opening parenthesis that is
2907    not followed by ? behaves as if it were followed by ?: but named parentheses
2908    can still be used for capturing (and they will acquire numbers in the usual
2909    way).
2910    
2911    41. Redesigned the return codes from the match() function into yes/no/error so
2912    that errors can be passed back from deep inside the nested calls. A malloc
2913    failure while inside a recursive subpattern call now causes the
2914    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2915    
2916    42. It is now possible to set a limit on the number of times the match()
2917    function is called in a call to pcre_exec(). This facility makes it possible to
2918    limit the amount of recursion and backtracking, though not in a directly
2919    obvious way, because the match() function is used in a number of different
2920    circumstances. The count starts from zero for each position in the subject
2921    string (for non-anchored patterns). The default limit is, for compatibility, a
2922    large number, namely 10 000 000. You can change this in two ways:
2923    
2924    (a) When configuring PCRE before making, you can use --with-match-limit=n
2925        to set a default value for the compiled library.
2926    
2927    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2928        a different value is set. See 45 below.
2929    
2930    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2931    
2932    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2933    of things that can be changed at compile time. The first argument specifies
2934    what is wanted and the second points to where the information is to be placed.
2935    The current list of available information is:
2936    
2937      PCRE_CONFIG_UTF8
2938    
2939    The output is an integer that is set to one if UTF-8 support is available;
2940    otherwise it is set to zero.
2941    
2942      PCRE_CONFIG_NEWLINE
2943    
2944    The output is an integer that it set to the value of the code that is used for
2945    newline. It is either LF (10) or CR (13).
2946    
2947      PCRE_CONFIG_LINK_SIZE
2948    
2949    The output is an integer that contains the number of bytes used for internal
2950    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2951    
2952      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2953    
2954    The output is an integer that contains the threshold above which the POSIX
2955    interface uses malloc() for output vectors. See item 31 above.
2956    
2957      PCRE_CONFIG_MATCH_LIMIT
2958    
2959    The output is an unsigned integer that contains the default limit of the number
2960    of match() calls in a pcre_exec() execution. See 42 above.
2961    
2962    44. pcretest has been upgraded by the addition of the -C option. This causes it
2963    to extract all the available output from the new pcre_config() function, and to
2964    output it. The program then exits immediately.
2965    
2966    45. A need has arisen to pass over additional data with calls to pcre_exec() in
2967    order to support additional features. One way would have been to define
2968    pcre_exec2() (for example) with extra arguments, but this would not have been
2969    extensible, and would also have required all calls to the original function to
2970    be mapped to the new one. Instead, I have chosen to extend the mechanism that
2971    is used for passing in "extra" data from pcre_study().
2972    
2973    The pcre_extra structure is now exposed and defined in pcre.h. It currently
2974    contains the following fields:
2975    
2976      flags         a bitmap indicating which of the following fields are set
2977      study_data    opaque data from pcre_study()
2978      match_limit   a way of specifying a limit on match() calls for a specific
2979                      call to pcre_exec()
2980      callout_data  data for callouts (see 49 below)
2981    
2982    The flag bits are also defined in pcre.h, and are
2983    
2984      PCRE_EXTRA_STUDY_DATA
2985      PCRE_EXTRA_MATCH_LIMIT
2986      PCRE_EXTRA_CALLOUT_DATA
2987    
2988    The pcre_study() function now returns one of these new pcre_extra blocks, with
2989    the actual study data pointed to by the study_data field, and the
2990    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2991    before. That is, this change is entirely upwards-compatible and requires no
2992    change to existing code.
2993    
2994    If you want to pass in additional data to pcre_exec(), you can either place it
2995    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2996    block.
2997    
2998    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2999    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3000    times with different match limits, until it finds the minimum value needed for
3001    pcre_exec() to complete. The value is then output. This can be instructive; for
3002    most simple matches the number is quite small, but for pathological cases it
3003    gets very large very quickly.
3004    
3005    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3006    returns the size of the data block pointed to by the study_data field in a
3007    pcre_extra block, that is, the value that was passed as the argument to
3008    pcre_malloc() when PCRE was getting memory in which to place the information
3009    created by pcre_study(). The fourth argument should point to a size_t variable.
3010    pcretest has been extended so that this information is shown after a successful
3011    pcre_study() call when information about the compiled regex is being displayed.
3012    
3013    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3014    because what follows is always an absolute path. (Later: it turns out that this
3015    is more than cosmetic for MinGW, because it doesn't like empty path
3016    components.)
3017    
3018    49. Some changes have been made to the callout feature (see 28 above):
3019    
3020    (i)  A callout function now has three choices for what it returns:
3021    
3022           0  =>  success, carry on matching
3023         > 0  =>  failure at this point, but backtrack if possible
3024         < 0  =>  serious error, return this value from pcre_exec()
3025    
3026         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3027         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3028         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3029         use by callout functions. It will never be used by PCRE itself.
3030    
3031    (ii) The pcre_extra structure (see 45 above) has a void * field called
3032         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3033         pcre_callout_block structure has a field of the same name. The contents of
3034         the field passed in the pcre_extra structure are passed to the callout
3035         function in the corresponding field in the callout block. This makes it
3036         easier to use the same callout-containing regex from multiple threads. For
3037         testing, the pcretest program has a new data escape
3038    
3039           \C*n        pass the number n (may be negative) as callout_data
3040    
3041         If the callout function in pcretest receives a non-zero value as
3042         callout_data, it returns that value.
3043    
3044    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3045    there were some redundant $(CFLAGS) in commands that are now specified as
3046    $(LINK), which already includes $(CFLAGS).
3047    
3048    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3049    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3050    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3051    one-byte characters throughout. Note that case-insensitive matching applies
3052    only to characters whose values are less than 256. PCRE doesn't support the
3053    notion of cases for higher-valued characters.
3054    
3055    (i)   A character class whose characters are all within 0-255 is handled as
3056          a bit map, and the map is inverted for negative classes. Previously, a
3057          character > 255 always failed to match such a class; however it should
3058          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3059    
3060    (ii)  A negated character class with a single character < 255 is coded as
3061          "not this character" (OP_NOT). This wasn't working properly when the test
3062          character was multibyte, either singly or repeated.
3063    
3064    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3065          mode, for example: \x{100}{2,3}.
3066    
3067    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3068          singly or repeated) now correctly test multibyte characters. However,
3069          PCRE doesn't recognize any characters with values greater than 255 as
3070          digits, spaces, or word characters. Such characters always match \D, \S,
3071          and \W, and never match \d, \s, or \w.
3072    
3073    (v)   Classes may now contain characters and character ranges with values
3074          greater than 255. For example: [ab\x{100}-\x{400}].
3075    
3076    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3077          PCRE in UTF-8 mode.
3078    
3079    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3080    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3081    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3082    value.)
3083    
3084    53. The single man page has become too large. I have therefore split it up into
3085    a number of separate man pages. These also give rise to individual HTML pages;
3086    these are now put in a separate directory, and there is an index.html page that
3087    lists them all. Some hyperlinking between the pages has been installed.
3088    
3089    54. Added convenience functions for handling named capturing parentheses.
3090    
3091    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3092    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3093    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3094    are faulted.
3095    
3096    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3097    calling configure. These values are used when compiling the dftables.c program
3098    which is run to generate the source of the default character tables. They
3099    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3100    you will need to set these values.
3101    
3102    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3103    
3104    
3105    Version 3.9 02-Jan-02
3106    ---------------------
3107    
3108    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3109    
3110    2. If --disable-static was given, the building process failed when trying to
3111    build pcretest and pcregrep. (For some reason it was using libtool to compile
3112    them, which is not right, as they aren't part of the library.)
3113    
3114    
3115    Version 3.8 18-Dec-01
3116    ---------------------
3117    
3118    1. The experimental UTF-8 code was completely screwed up. It was packing the
3119    bytes in the wrong order. How dumb can you get?
3120    
3121    
3122    Version 3.7 29-Oct-01
3123    ---------------------
3124    
3125    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3126    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3127    this didn't happen under Solaris 8, where I normally test things.
3128    
3129    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3130    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3131    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3132    this fix an hour or so after the initial 3.7 release.)
3133    
3134    
3135    Version 3.6 23-Oct-01
3136    ---------------------
3137    
3138    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3139    offsets passed as NULL with zero offset count.
3140    
3141    2. The config.guess and config.sub files had not been updated when I moved to
3142    the latest autoconf.
3143    
3144    
3145    Version 3.5 15-Aug-01
3146    ---------------------
3147    
3148    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3149    had been forgotten.
3150    
3151    2. By using declared but undefined structures, we can avoid using "void"
3152    definitions in pcre.h while keeping the internal definitions of the structures
3153    private.
3154    
3155    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3156    user point of view, this means that both static and shared libraries are built
3157    by default, but this can be individually controlled. More of the work of
3158    handling this static/shared cases is now inside libtool instead of PCRE's make
3159    file.
3160    
3161    4. The pcretest utility is now installed along with pcregrep because it is
3162    useful for users (to test regexs) and by doing this, it automatically gets
3163    relinked by libtool. The documentation has been turned into a man page, so
3164    there are now .1, .txt, and .html versions in /doc.
3165    
3166    5. Upgrades to pcregrep:
3167       (i)   Added long-form option names like gnu grep.
3168       (ii)  Added --help to list all options with an explanatory phrase.
3169       (iii) Added -r, --recursive to recurse into sub-directories.
3170       (iv)  Added -f, --file to read patterns from a file.
3171    
3172    6. pcre_exec() was referring to its "code" argument before testing that
3173    argument for NULL (and giving an error if it was NULL).
3174    
3175    7. Upgraded Makefile.in to allow for compiling in a different directory from
3176    the source directory.
3177    
3178    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3179    options bits, the pointer it was passed was to an int instead of to an unsigned
3180    long int. This mattered only on 64-bit systems.
3181    
3182    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3183    generated) instead of pcre.in, which it its source. Also made the same change
3184    in several of the .c files.
3185    
3186    10. A new release of gcc defines printf() as a macro, which broke pcretest
3187    because it had an ifdef in the middle of a string argument for printf(). Fixed
3188    by using separate calls to printf().
3189    
3190    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3191    script, to force use of CR or LF instead of \n in the source. On non-Unix
3192    systems, the value can be set in config.h.
3193    
3194    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3195    absolute limit. Changed the text of the error message to make this clear, and
3196    likewise updated the man page.
3197    
3198    13. The limit of 99 on the number of capturing subpatterns has been removed.
3199    The new limit is 65535, which I hope will not be a "real" limit.
3200    
3201    
3202    Version 3.4 22-Aug-00
3203    ---------------------
3204    
3205    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3206    
3207    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3208    
3209    
3210    Version 3.3 01-Aug-00
3211    ---------------------
3212    
3213    1. If an octal character was given, but the value was greater than \377, it
3214    was not getting masked to the least significant bits, as documented. This could
3215    lead to crashes in some systems.
3216    
3217    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3218    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3219    
3220    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3221    These just pass their arguments on to (pcre_free)(), but they are provided
3222    because some uses of PCRE bind it to non-C systems that can call its functions,
3223    but cannot call free() or pcre_free() directly.
3224    
3225    4. Add "make test" as a synonym for "make check". Corrected some comments in
3226    the Makefile.
3227    
3228    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3229    Makefile.
3230    
3231    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3232    command called pgrep for grepping around the active processes.
3233    
3234    7. Added the beginnings of support for UTF-8 character strings.
3235    
3236    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3237    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3238    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3239    out for the ar command.)
3240    
3241    
3242    Version 3.2 12-May-00
3243    ---------------------
3244    
3245    This is purely a bug fixing release.
3246    
3247    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3248    of ZA. This was just one example of several cases that could provoke this bug,
3249    which was introduced by change 9 of version 2.00. The code for breaking
3250    infinite loops after an iteration that matches an empty string was't working
3251    correctly.
3252    
3253    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3254    when matched against abbab (for example). After matching an empty string, it
3255    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3256    caused it to match further down the string than it should.
3257    
3258    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3259    was there because it doesn't seem to be needed, and it causes trouble on some
3260    systems, as it is not a Standard C header. It has been removed.
3261    
3262    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3263    were reported on the Macintosh. The changes were from
3264    
3265      while ((c = *(++ptr)) != 0 && c != '\n');
3266    to
3267      while ((c = *(++ptr)) != 0 && c != '\n') ;
3268    
3269    Totally extraordinary, but if that's what it takes...
3270    
3271    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3272    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3273    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3274    assumes the way PCRE uses memmove() (always moving upwards).
3275    
3276    6. PCRE is being used in one environment where strchr() is not available. There
3277    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3278    faster code anyway.
3279    
3280    
3281    Version 3.1 09-Feb-00
3282    ---------------------
3283    
3284    The only change in this release is the fixing of some bugs in Makefile.in for
3285    the "install" target:
3286    
3287    (1) It was failing to install pcreposix.h.
3288    
3289    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3290    
3291    
3292    Version 3.0 01-Feb-00
3293    ---------------------
3294    
3295    1. Add support for the /+ modifier to perltest (to output $` like it does in
3296    pcretest).
3297    
3298    2. Add support for the /g modifier to perltest.
3299    
3300    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3301    matches null strings.
3302    
3303    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3304    pattern. Perl treats empty patterns specially - it reuses the most recent
3305    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3306    effect.
3307    
3308    5. The POSIX interface was broken in that it was just handing over the POSIX
3309    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3310    required a bigger vector, with some working space on the end. This means that
3311    the POSIX wrapper now has to get and free some memory, and copy the results.
3312    
3313    6. Added some simple autoconf support, placing the test data and the
3314    documentation in separate directories, re-organizing some of the
3315    information files, and making it build pcre-config (a GNU standard). Also added
3316    libtool support for building PCRE as a shared library, which is now the
3317    default.
3318    
3319    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3320    09 are not valid octal constants. Single digits will be used for minor values
3321    less than 10.
3322    
3323    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3324    existing programs that set these in the POSIX interface can use PCRE without
3325    modification.
3326    
3327    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3328    return all that pcre_info() returns, plus additional data. The pcre_info()
3329    function is retained for compatibility, but is considered to be obsolete.
3330    
3331    10. Added experimental recursion feature (?R) to handle one common case that
3332    Perl 5.6 will be able to do with (?p{...}).
3333    
3334    11. Added support for POSIX character classes like [:alpha:], which Perl is
3335    adopting.
3336    
3337    
3338    Version 2.08 31-Aug-99
3339    ----------------------
3340    
3341    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3342    trying to match at the startoffset position, but instead was moving forward to
3343    the next newline as if a previous match had failed.
3344    
3345    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3346    and could get into a loop if a null string was matched other than at the start
3347    of the subject.
3348    
3349    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3350    be distinguished at compile time, and for completeness also added PCRE_DATE.
3351    
3352    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3353    in GnuWin32 environments.
3354    
3355    
3356    Version 2.07 29-Jul-99
3357    ----------------------
3358    
3359    1. The documentation is now supplied in plain text form and HTML as well as in
3360    the form of man page sources.
3361    
3362    2. C++ compilers don't like assigning (void *) values to other pointer types.
3363    In particular this affects malloc(). Although there is no problem in Standard
3364    C, I've put in casts to keep C++ compilers happy.
3365    
3366    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3367    should be (const char *).
3368    
3369    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3370    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3371    However, I haven't made this a standard facility. The documentation doesn't
3372    mention it, and the Makefile doesn't support it.
3373    
3374    5. The Makefile now contains an "install" target, with editable destinations at
3375    the top of the file. The pcretest program is not installed.
3376    
3377    6. pgrep -V now gives the PCRE version number and date.
3378    
3379    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3380    causing the entire string to be ignored, instead of just the last character.
3381    
3382    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3383    non-matching string, it can take a very, very long time, even for strings of
3384    quite modest length, because of the nested recursion. PCRE now does better in
3385    some of these cases. It does this by remembering the last required literal
3386    character in the pattern, and pre-searching the subject to ensure it is present
3387    before running the real match. In other words, it applies a heuristic to detect
3388    some types of certain failure quickly, and in the above example, if presented
3389    with a string that has no trailing " it gives "no match" very quickly.
3390    
3391    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3392    other alternatives are tried instead.
3393    
3394    
3395    Version 2.06 09-Jun-99
3396    ----------------------
3397    
3398    1. Change pcretest's output for amount of store used to show just the code
3399    space, because the remainder (the data block) varies in size between 32-bit and
3400    64-bit systems.
3401    
3402    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3403    start matching at. This allows lookbehinds to work when searching for multiple
3404    occurrences in a string.
3405    
3406    3. Added additional options to pcretest for testing multiple occurrences:
3407    
3408       /+   outputs the rest of the string that follows a match
3409       /g   loops for multiple occurrences, using the new startoffset argument
3410       /G   loops for multiple occurrences by passing an incremented pointer
3411    
3412    4. PCRE wasn't doing the "first character" optimization for patterns starting
3413    with \b or \B, though it was doing it for other lookbehind assertions. That is,
3414    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3415    the letter 'x'. On long subject strings, this gives a significant speed-up.
3416    
3417    
3418    Version 2.05 21-Apr-99
3419    ----------------------
3420    
3421    1. Changed the type of magic_number from int to long int so that it works
3422    properly on 16-bit systems.
3423    
3424    2. Fixed a bug which caused patterns starting with .* not to work correctly
3425    when the subject string contained newline characters. PCRE was assuming
3426    anchoring for such patterns in all cases, which is not correct because .* will
3427    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3428    DOTALL is set at top level; otherwise it knows that patterns starting with .*
3429    must be retried after every newline in the subject.
3430    
3431    
3432    Version 2.04 18-Feb-99
3433    ----------------------
3434    
3435    1. For parenthesized subpatterns with repeats whose minimum was zero, the
3436    computation of the store needed to hold the pattern was incorrect (too large).
3437    If such patterns were nested a few deep, this could multiply and become a real
3438    problem.
3439    
3440    2. Added /M option to pcretest to show the memory requirement of a specific
3441    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3442    
3443    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3444    compiled in such a way that the backtracking after subsequent failure was
3445    pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3446    ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3447    
3448    
3449    Version 2.03 02-Feb-99
3450    ----------------------
3451    
3452    1. Fixed typo and small mistake in man page.
3453    
3454    2. Added 4th condition (GPL supersedes if conflict) and created separate
3455    LICENCE file containing the conditions.
3456    
3457    3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3458    Perl, that is the internal \ allows the delimiter to be included in the
3459    pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3460    the final delimiter, add \ to the end of the pattern (to test the error).
3461    
3462    4. Added the convenience functions for extracting substrings after a successful
3463    match. Updated pcretest to make it able to test these functions.
3464    
3465    
3466    Version 2.02 14-Jan-99
3467    ----------------------
3468    
3469    1. Initialized the working variables associated with each extraction so that
3470    their saving and restoring doesn't refer to uninitialized store.
3471    
3472    2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3473    compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3474    fix the problem.
3475    
3476    3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3477    calls, and wasn't printing the correct value for compiling calls. Increased the
3478    default value of LOOPREPEAT, and the number of significant figures in the
3479    times.
3480    
3481    4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3482    
3483    5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3484    a building problem on Windows NT with a FAT file system.
3485    
3486    
3487    Version 2.01 21-Oct-98
3488    ----------------------
3489    
3490    1. Changed the API for pcre_compile() to allow for the provision of a pointer
3491    to character tables built by pcre_maketables() in the current locale. If NULL
3492    is passed, the default tables are used.
3493    
3494    
3495    Version 2.00 24-Sep-98
3496    ----------------------
3497    
3498    1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3499    it any more.
3500    
3501    2. Allow quantification of (?>) groups, and make it work correctly.
3502    
3503    3. The first character computation wasn't working for (?>) groups.
3504    
3505    4. Correct the implementation of \Z (it is permitted to match on the \n at the
3506    end of the subject) and add 5.005's \z, which really does match only at the
3507    very end of the subject.
3508    
3509    5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3510    
3511    6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3512    DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3513    localized options. All options to pcre_study() were also removed.
3514    
3515    7. Add other new features from 5.005:
3516    
3517       $(?<=           positive lookbehind
3518       $(?<!           negative lookbehind
3519       (?imsx-imsx)    added the unsetting capability
3520                       such a setting is global if at outer level; local otherwise
3521       (?imsx-imsx:)   non-capturing groups with option setting
3522       (?(cond)re|re)  conditional pattern matching
3523    
3524       A backreference to itself in a repeated group matches the previous
3525       captured string.
3526    
3527    8. General tidying up of studying (both automatic and via "study")
3528    consequential on the addition of new assertions.
3529    
3530    9. As in 5.005, unlimited repeated groups that could match an empty substring
3531    are no longer faulted at compile time. Instead, the loop is forcibly broken at
3532    runtime if any iteration does actually match an empty substring.
3533    
3534    10. Include the RunTest script in the distribution.
3535    
3536    11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3537    discrepancies, some of which were old and were also with respect to 5.004. They
3538    have now been fixed.
3539