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

Diff of /code/trunk/ChangeLog

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

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8        was out-of-date, and there was no check on the pcre_compile() error code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

Legend:
Removed from v.65  
changed lines
  Added in v.422

  ViewVC Help
Powered by ViewVC 1.1.5