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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.29  
changed lines
  Added in v.498

  ViewVC Help
Powered by ViewVC 1.1.5