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

Diff of /code/trunk/ChangeLog

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

revision 59 by nigel, Sat Feb 24 21:39:54 2007 UTC revision 307 by ph10, Sun Jan 20 20:36:45 2008 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 7.6 19-Jan-08
5    ---------------------
6    
7    1.  A character class containing a very large number of characters with
8        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
9        overflow.
10    
11    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
12        HAVE_LONG_LONG is not defined.
13    
14    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
15        bring them up to date and include new features. This patch includes:
16    
17        - Fixed PH's badly added libz and libbz2 support.
18        - Fixed a problem with static linking.
19        - Added pcredemo.
20        - Fixed dftables problem and added an option.
21        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
22            HAVE_LONG_LONG.
23        - Added readline support for pcretest.
24        - Added an listing of the option settings after cmake has run.
25    
26    4.  A user submitted a patch to Makefile that makes it easy to created a dll
27        under mingw. I added stuff to Makefile.am that cause it to include this
28        special target, without affecting anything else.
29    
30    
31    Version 7.5 10-Jan-08
32    ---------------------
33    
34    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
35        values in parens when parsing an RE using the C++ wrapper."
36    
37    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
38        Characters greater than 255 were excluded from the class instead of being
39        included.
40    
41    3.  The same bug as (2) above applied to negated POSIX classes such as
42        [:^space:].
43    
44    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
45        defined or documented. It seems to have been a typo for PCRE_STATIC, so
46        I have changed it.
47    
48    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
49        first named subpattern) and a construct such as (?&a) would reference the
50        first named subpattern whose name started with "a" (in other words, the
51        length check was missing). Both these problems are fixed. "Subpattern name
52        expected" is now given for (?&) (a zero-length name), and this patch also
53        makes it give the same error for \k'' (previously it complained that that
54        was a reference to a non-existent subpattern).
55    
56    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
57        this is right because (?- can be followed by option settings as well as by
58        digits. I have, however, made the messages clearer.
59    
60    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
61        than the number used in the conditional) now cause a compile-time error.
62        This is actually not compatible with Perl, which accepts such patterns, but
63        treats the conditional as always being FALSE (as PCRE used to), but it
64        seems to me that giving a diagnostic is better.
65    
66    8.  Change "alphameric" to the more common word "alphanumeric" in comments
67        and messages.
68    
69    9.  Fix two occurrences of "backslash" in comments that should have been
70        "backspace".
71    
72    10. Remove two redundant lines of code that can never be obeyed (their function
73        was moved elsewhere).
74    
75    11. The program that makes PCRE's Unicode character property table had a bug
76        which caused it to generate incorrect table entries for sequences of
77        characters that have the same character type, but are in different scripts.
78        It amalgamated them into a single range, with the script of the first of
79        them. In other words, some characters were in the wrong script. There were
80        thirteen such cases, affecting characters in the following ranges:
81    
82          U+002b0 - U+002c1
83          U+0060c - U+0060d
84          U+0061e - U+00612
85          U+0064b - U+0065e
86          U+0074d - U+0076d
87          U+01800 - U+01805
88          U+01d00 - U+01d77
89          U+01d9b - U+01dbf
90          U+0200b - U+0200f
91          U+030fc - U+030fe
92          U+03260 - U+0327f
93          U+0fb46 - U+0fbb1
94          U+10450 - U+1049d
95    
96    12. The -o option (show only the matching part of a line) for pcregrep was not
97        compatible with GNU grep in that, if there was more than one match in a
98        line, it showed only the first of them. It now behaves in the same way as
99        GNU grep.
100    
101    13. If the -o and -v options were combined for pcregrep, it printed a blank
102        line for every non-matching line. GNU grep prints nothing, and pcregrep now
103        does the same. The return code can be used to tell if there were any
104        non-matching lines.
105    
106    14. Added --file-offsets and --line-offsets to pcregrep.
107    
108    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
109        infinitely looping recursion. The bug was that positive lookaheads were not
110        being skipped when checking for a possible empty match (negative lookaheads
111        and both kinds of lookbehind were skipped).
112    
113    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
114        inclusion of <windows.h> to before rather than after the definition of
115        INVALID_FILE_ATTRIBUTES (patch from David Byron).
116    
117    17. Specifying a possessive quantifier with a specific limit for a Unicode
118        character property caused pcre_compile() to compile bad code, which led at
119        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
120        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
121        caused the error; without that there was no problem.
122    
123    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
124    
125    19. Added --enable-pcretest-libreadline.
126    
127    20. In pcrecpp.cc, the variable 'count' was incremented twice in
128        RE::GlobalReplace(). As a result, the number of replacements returned was
129        double what it should be. I removed one of the increments, but Craig sent a
130        later patch that removed the other one (the right fix) and added unit tests
131        that check the return values (which was not done before).
132    
133    21. Several CMake things:
134    
135        (1) Arranged that, when cmake is used on Unix, the libraries end up with
136            the names libpcre and libpcreposix, not just pcre and pcreposix.
137    
138        (2) The above change means that pcretest and pcregrep are now correctly
139            linked with the newly-built libraries, not previously installed ones.
140    
141        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
142    
143    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
144        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
145        UTF-8 newline character). The key issue is that the pattern starts .*;
146        this means that the match must be either at the beginning, or after a
147        newline. The bug was in the code for advancing after a failed match and
148        checking that the new position followed a newline. It was not taking
149        account of UTF-8 characters correctly.
150    
151    23. PCRE was behaving differently from Perl in the way it recognized POSIX
152        character classes. PCRE was not treating the sequence [:...:] as a
153        character class unless the ... were all letters. Perl, however, seems to
154        allow any characters between [: and :], though of course it rejects as
155        unknown any "names" that contain non-letters, because all the known class
156        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
157        for example, whereas PCRE did not - it did not recognize a POSIX character
158        class. This seemed a bit dangerous, so the code has been changed to be
159        closer to Perl. The behaviour is not identical to Perl, because PCRE will
160        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
161        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
162        Perl does, and where it didn't before.
163    
164    24. Rewrite so as to remove the single use of %n from pcregrep because in some
165        Windows environments %n is disabled by default.
166    
167    
168    Version 7.4 21-Sep-07
169    ---------------------
170    
171    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
172        means that a class such as [\s] counted as "explicit reference to CR or
173        LF". That isn't really right - the whole point of the change was to try to
174        help when there was an actual mention of one of the two characters. So now
175        the change happens only if \r or \n (or a literal CR or LF) character is
176        encountered.
177    
178    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
179        of both had grown to the point where there were only 3 bits left.
180        Fortunately, there was spare space in the data structure, and so I have
181        moved the internal flags into a new 16-bit field to free up more option
182        bits.
183    
184    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
185        but did not set the internal JCHANGED flag - either of these is enough to
186        control the way the "get" function works - but the PCRE_INFO_JCHANGED
187        facility is supposed to tell if (?J) was ever used, so now (?J) at the
188        start sets both bits.
189    
190    4.  Added options (at build time, compile time, exec time) to change \R from
191        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
192    
193    5.  doc/pcresyntax.html was missing from the distribution.
194    
195    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
196        compatibility, even though it is no longer used.
197    
198    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
199        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
200        windows.h file is present (where different names are used). [This was
201        reversed later after testing - see 16 below.]
202    
203    8.  Changed all #include <config.h> to #include "config.h". There were also
204        some further <pcre.h> cases that I changed to "pcre.h".
205    
206    9.  When pcregrep was used with the --colour option, it missed the line ending
207        sequence off the lines that it output.
208    
209    10. It was pointed out to me that arrays of string pointers cause lots of
210        relocations when a shared library is dynamically loaded. A technique of
211        using a single long string with a table of offsets can drastically reduce
212        these. I have refactored PCRE in four places to do this. The result is
213        dramatic:
214    
215          Originally:                          290
216          After changing UCP table:            187
217          After changing error message table:   43
218          After changing table of "verbs"       36
219          After changing table of Posix names   22
220    
221        Thanks to the folks working on Gregex for glib for this insight.
222    
223    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
224        unicode-properties was also set.
225    
226    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
227    
228    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
229        checked only for CRLF.
230    
231    14. Added casts to pcretest.c to avoid compiler warnings.
232    
233    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
234    
235    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
236        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
237        entirely. This removes changes made in 7 above.
238    
239    17. The CMake files have been updated, and there is now more information about
240        building with CMake in the NON-UNIX-USE document.
241    
242    
243    Version 7.3 28-Aug-07
244    ---------------------
245    
246     1. In the rejigging of the build system that eventually resulted in 7.1, the
247        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
248        brackets there is not right, since it causes compilers to look for an
249        installed pcre.h, not the version that is in the source that is being
250        compiled (which of course may be different). I have changed it back to:
251    
252          #include "pcre.h"
253    
254        I have a vague recollection that the change was concerned with compiling in
255        different directories, but in the new build system, that is taken care of
256        by the VPATH setting the Makefile.
257    
258     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
259        when the subject happened to end in the byte 0x85 (e.g. if the last
260        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
261        characters but of course it shouldn't be taken as a newline when it is part
262        of another character. The bug was that, for an unlimited repeat of . in
263        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
264        characters when looking for a newline.
265    
266     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
267    
268     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
269        in debug output.
270    
271     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
272        long printing in the pcrecpp unittest when running under MinGW.
273    
274     6. ESC_K was left out of the EBCDIC table.
275    
276     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
277        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
278        limit also applies to "virtual nesting" when a pattern is recursive, and in
279        this case 1000 isn't so big. I have been able to remove this limit at the
280        expense of backing off one optimization in certain circumstances. Normally,
281        when pcre_exec() would call its internal match() function recursively and
282        immediately return the result unconditionally, it uses a "tail recursion"
283        feature to save stack. However, when a subpattern that can match an empty
284        string has an unlimited repetition quantifier, it no longer makes this
285        optimization. That gives it a stack frame in which to save the data for
286        checking that an empty string has been matched. Previously this was taken
287        from the 1000-entry workspace that had been reserved. So now there is no
288        explicit limit, but more stack is used.
289    
290     8. Applied Daniel's patches to solve problems with the import/export magic
291        syntax that is required for Windows, and which was going wrong for the
292        pcreposix and pcrecpp parts of the library. These were overlooked when this
293        problem was solved for the main library.
294    
295     9. There were some crude static tests to avoid integer overflow when computing
296        the size of patterns that contain repeated groups with explicit upper
297        limits. As the maximum quantifier is 65535, the maximum group length was
298        set at 30,000 so that the product of these two numbers did not overflow a
299        32-bit integer. However, it turns out that people want to use groups that
300        are longer than 30,000 bytes (though not repeat them that many times).
301        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
302        made it possible to implement the integer overflow checks in a much more
303        dynamic way, which I have now done. The artificial limitation on group
304        length has been removed - we now have only the limit on the total length of
305        the compiled pattern, which depends on the LINK_SIZE setting.
306    
307    10. Fixed a bug in the documentation for get/copy named substring when
308        duplicate names are permitted. If none of the named substrings are set, the
309        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
310        empty string.
311    
312    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
313        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
314        because the ] is interpreted as the first data character and the
315        terminating ] is not found. PCRE has been made compatible with Perl in this
316        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
317        cause memory overwriting.
318    
319    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
320        string has been matched (to stop an infinite loop). It was not recognizing
321        a conditional subpattern that could match an empty string if that
322        subpattern was within another subpattern. For example, it looped when
323        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
324        condition was not nested. This bug has been fixed.
325    
326    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
327        past the start of the subject in the presence of bytes with the top bit
328        set, for example "\x8aBCD".
329    
330    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
331        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
332    
333    14. Optimized (?!) to (*FAIL).
334    
335    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
336        This restricts code points to be within the range 0 to 0x10FFFF, excluding
337        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
338        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
339        does: it's just the validity check that is more restrictive.
340    
341    16. Inserted checks for integer overflows during escape sequence (backslash)
342        processing, and also fixed erroneous offset values for syntax errors during
343        backslash processing.
344    
345    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
346        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
347    
348    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
349        caused an overrun.
350    
351    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
352        something other than just ASCII characters) inside a group that had an
353        unlimited repeat caused a loop at compile time (while checking to see
354        whether the group could match an empty string).
355    
356    20. Debugging a pattern containing \p or \P could cause a crash. For example,
357        [\P{Any}] did so. (Error in the code for printing property names.)
358    
359    21. An orphan \E inside a character class could cause a crash.
360    
361    22. A repeated capturing bracket such as (A)? could cause a wild memory
362        reference during compilation.
363    
364    23. There are several functions in pcre_compile() that scan along a compiled
365        expression for various reasons (e.g. to see if it's fixed length for look
366        behind). There were bugs in these functions when a repeated \p or \P was
367        present in the pattern. These operators have additional parameters compared
368        with \d, etc, and these were not being taken into account when moving along
369        the compiled data. Specifically:
370    
371        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
372            length.
373    
374        (b) An item such as \pL+ within a repeated group could cause crashes or
375            loops.
376    
377        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
378            "reference to non-existent subpattern" error.
379    
380        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
381    
382    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
383        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
384    
385    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
386    
387    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
388        character were causing crashes (broken optimization).
389    
390    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
391        \p or \P) caused a compile-time loop.
392    
393    28. More problems have arisen in unanchored patterns when CRLF is a valid line
394        break. For example, the unstudied pattern [\r\n]A does not match the string
395        "\r\nA" because change 7.0/46 below moves the current point on by two
396        characters after failing to match at the start. However, the pattern \nA
397        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
398        the same is true. There doesn't seem any very clean way out of this, but
399        what I have chosen to do makes the common cases work: PCRE now takes note
400        of whether there can be an explicit match for \r or \n anywhere in the
401        pattern, and if so, 7.0/46 no longer applies. As part of this change,
402        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
403        pattern has explicit CR or LF references.
404    
405    29. Added (*CR) etc for changing newline setting at start of pattern.
406    
407    
408    Version 7.2 19-Jun-07
409    ---------------------
410    
411     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
412        which is apparently normally available under Windows.
413    
414     2. Re-jig the pcregrep tests with different newline settings in an attempt
415        to make them independent of the local environment's newline setting.
416    
417     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
418    
419     4. Some of the "internals" tests were previously cut out when the link size
420        was not 2, because the output contained actual offsets. The recent new
421        "Z" feature of pcretest means that these can be cut out, making the tests
422        usable with all link sizes.
423    
424     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
425        stack recursion. This gives a massive performance boost under BSD, but just
426        a small improvement under Linux. However, it saves one field in the frame
427        in all cases.
428    
429     6. Added more features from the forthcoming Perl 5.10:
430    
431        (a) (?-n) (where n is a string of digits) is a relative subroutine or
432            recursion call. It refers to the nth most recently opened parentheses.
433    
434        (b) (?+n) is also a relative subroutine call; it refers to the nth next
435            to be opened parentheses.
436    
437        (c) Conditions that refer to capturing parentheses can be specified
438            relatively, for example, (?(-2)... or (?(+3)...
439    
440        (d) \K resets the start of the current match so that everything before
441            is not part of it.
442    
443        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
444    
445        (f) \g{name} is another synonym - part of Perl 5.10's unification of
446            reference syntax.
447    
448        (g) (?| introduces a group in which the numbering of parentheses in each
449            alternative starts with the same number.
450    
451        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
452    
453     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
454        PCRE_INFO_JCHANGED.
455    
456     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
457        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
458        for detecting groups that can match an empty string.
459    
460     9. A pattern with a very large number of alternatives (more than several
461        hundred) was running out of internal workspace during the pre-compile
462        phase, where pcre_compile() figures out how much memory will be needed. A
463        bit of new cunning has reduced the workspace needed for groups with
464        alternatives. The 1000-alternative test pattern now uses 12 bytes of
465        workspace instead of running out of the 4096 that are available.
466    
467    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
468    
469    11. Applied patch from Google to remove an optimization that didn't quite work.
470        The report of the bug said:
471    
472          pcrecpp::RE("a*").FullMatch("aaa") matches, while
473          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
474          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
475    
476    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
477        it matched the wrong number of bytes.
478    
479    
480    Version 7.1 24-Apr-07
481    ---------------------
482    
483     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
484        that is more "standard", making use of automake and other Autotools. There
485        is some re-arrangement of the files and adjustment of comments consequent
486        on this.
487    
488     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
489        for recursive directory scanning broke on some systems because the files
490        are not scanned in any specific order and on different systems the order
491        was different. A call to "sort" has been inserted into RunGrepTest for the
492        approprate test as a short-term fix. In the longer term there may be an
493        alternative.
494    
495     3. I had an email from Eric Raymond about problems translating some of PCRE's
496        man pages to HTML (despite the fact that I distribute HTML pages, some
497        people do their own conversions for various reasons). The problems
498        concerned the use of low-level troff macros .br and .in. I have therefore
499        removed all such uses from the man pages (some were redundant, some could
500        be replaced by .nf/.fi pairs). The 132html script that I use to generate
501        HTML has been updated to handle .nf/.fi and to complain if it encounters
502        .br or .in.
503    
504     4. Updated comments in configure.ac that get placed in config.h.in and also
505        arranged for config.h to be included in the distribution, with the name
506        config.h.generic, for the benefit of those who have to compile without
507        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
508    
509     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
510        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
511        makevp.bat for the current PCRE, using the additional files
512        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
513    
514     6. A Windows user reported a minor discrepancy with test 2, which turned out
515        to be caused by a trailing space on an input line that had got lost in his
516        copy. The trailing space was an accident, so I've just removed it.
517    
518     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
519        that is needed.
520    
521     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
522        as "const" (a) because they are and (b) because it helps the PHP
523        maintainers who have recently made a script to detect big data structures
524        in the php code that should be moved to the .rodata section. I remembered
525        to update Builducptable as well, so it won't revert if ucptable.h is ever
526        re-created.
527    
528     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
529        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
530        order to be able to cut out the UTF-8 tables in the latter when UTF-8
531        support is not required. This saves 1.5-2K of code, which is important in
532        some applications.
533    
534        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
535        so as not to refer to the tables, even though these functions will never be
536        called when UTF-8 support is disabled. Otherwise there are problems with a
537        shared library.
538    
539    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
540    
541        (a) It was defining its arguments as char * instead of void *.
542    
543        (b) It was assuming that all moves were upwards in memory; this was true
544            a long time ago when I wrote it, but is no longer the case.
545    
546        The emulated memove() is provided for those environments that have neither
547        memmove() nor bcopy(). I didn't think anyone used it these days, but that
548        is clearly not the case, as these two bugs were recently reported.
549    
550    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
551        and Detrail to create the HTML documentation, the .txt form of the man
552        pages, and it removes trailing spaces from listed files. It also creates
553        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
554        case, it wraps all the #defines with #ifndefs. This script should be run
555        before "make dist".
556    
557    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
558        with Unicode property support.
559    
560        (a) For a maximizing quantifier, if the two different cases of the
561            character were of different lengths in their UTF-8 codings (there are
562            some cases like this - I found 11), and the matching function had to
563            back up over a mixture of the two cases, it incorrectly assumed they
564            were both the same length.
565    
566        (b) When PCRE was configured to use the heap rather than the stack for
567            recursion during matching, it was not correctly preserving the data for
568            the other case of a UTF-8 character when checking ahead for a match
569            while processing a minimizing repeat. If the check also involved
570            matching a wide character, but failed, corruption could cause an
571            erroneous result when trying to check for a repeat of the original
572            character.
573    
574    13. Some tidying changes to the testing mechanism:
575    
576        (a) The RunTest script now detects the internal link size and whether there
577            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
578            values substituted by "configure". (The RunGrepTest script already did
579            this for UTF-8.) The configure.ac script no longer substitutes the
580            relevant variables.
581    
582        (b) The debugging options /B and /D in pcretest show the compiled bytecode
583            with length and offset values. This means that the output is different
584            for different internal link sizes. Test 2 is skipped for link sizes
585            other than 2 because of this, bypassing the problem. Unfortunately,
586            there was also a test in test 3 (the locale tests) that used /B and
587            failed for link sizes other than 2. Rather than cut the whole test out,
588            I have added a new /Z option to pcretest that replaces the length and
589            offset values with spaces. This is now used to make test 3 independent
590            of link size. (Test 2 will be tidied up later.)
591    
592    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
593        segmentation fault instead of returning the appropriate error message.
594    
595    15. In multiline mode when the newline sequence was set to "any", the pattern
596        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
597        This doesn't seem right; it now treats the CRLF combination as the line
598        ending, and so does not match in that case. It's only a pattern such as ^$
599        that would hit this one: something like ^ABC$ would have failed after \r
600        and then tried again after \r\n.
601    
602    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
603        in an attempt to make files that differ only in their line terminators
604        compare equal. This works on Linux.
605    
606    17. Under certain error circumstances pcregrep might try to free random memory
607        as it exited. This is now fixed, thanks to valgrind.
608    
609    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
610        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
611        was because its rules for how to advance for /g after matching an empty
612        string at the end of a line did not allow for this case. They now check for
613        it specially.
614    
615    20. pcretest is supposed to handle patterns and data of any length, by
616        extending its buffers when necessary. It was getting this wrong when the
617        buffer for a data line had to be extended.
618    
619    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
620        CRLF as a newline sequence.
621    
622    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
623        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
624        I have nevertheless tidied it up.
625    
626    23. Added some casts to kill warnings from HP-UX ia64 compiler.
627    
628    24. Added a man page for pcre-config.
629    
630    
631    Version 7.0 19-Dec-06
632    ---------------------
633    
634     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
635        moving to gcc 4.1.1.
636    
637     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
638        sys/time.h, which is documented as needed for this function. It doesn't
639        seem to matter on Linux, but it showed up on some releases of OS X.
640    
641     3. It seems that there are systems where bytes whose values are greater than
642        127 match isprint() in the "C" locale. The "C" locale should be the
643        default when a C program starts up. In most systems, only ASCII printing
644        characters match isprint(). This difference caused the output from pcretest
645        to vary, making some of the tests fail. I have changed pcretest so that:
646    
647        (a) When it is outputting text in the compiled version of a pattern, bytes
648            other than 32-126 are always shown as hex escapes.
649    
650        (b) When it is outputting text that is a matched part of a subject string,
651            it does the same, unless a different locale has been set for the match
652            (using the /L modifier). In this case, it uses isprint() to decide.
653    
654     4. Fixed a major bug that caused incorrect computation of the amount of memory
655        required for a compiled pattern when options that changed within the
656        pattern affected the logic of the preliminary scan that determines the
657        length. The relevant options are -x, and -i in UTF-8 mode. The result was
658        that the computed length was too small. The symptoms of this bug were
659        either the PCRE error "internal error: code overflow" from pcre_compile(),
660        or a glibc crash with a message such as "pcretest: free(): invalid next
661        size (fast)". Examples of patterns that provoked this bug (shown in
662        pcretest format) are:
663    
664          /(?-x: )/x
665          /(?x)(?-x: \s*#\s*)/
666          /((?i)[\x{c0}])/8
667          /(?i:[\x{c0}])/8
668    
669        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
670        is now done differently.
671    
672     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
673        wrapper classes; (b) implement a new function in the C++ scanner that is
674        more efficient than the old way of doing things because it avoids levels of
675        recursion in the regex matching; (c) add a paragraph to the documentation
676        for the FullMatch() function.
677    
678     6. The escape sequence \n was being treated as whatever was defined as
679        "newline". Not only was this contrary to the documentation, which states
680        that \n is character 10 (hex 0A), but it also went horribly wrong when
681        "newline" was defined as CRLF. This has been fixed.
682    
683     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
684        was being set to -1 for the "end of line" case (supposedly a value that no
685        character can have). Though this value is never used (the check for end of
686        line is "zero bytes in current character"), it caused compiler complaints.
687        I've changed it to 0xffffffff.
688    
689     8. In pcre_version.c, the version string was being built by a sequence of
690        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
691        string (as it is for production releases) called a macro with an empty
692        argument. The C standard says the result of this is undefined. The gcc
693        compiler treats it as an empty string (which was what was wanted) but it is
694        reported that Visual C gives an error. The source has been hacked around to
695        avoid this problem.
696    
697     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
698        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
699        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
700        of them did).
701    
702    10. Originally, pcretest opened its input and output without "b"; then I was
703        told that "b" was needed in some environments, so it was added for release
704        5.0 to both the input and output. (It makes no difference on Unix-like
705        systems.) Later I was told that it is wrong for the input on Windows. I've
706        now abstracted the modes into two macros, to make it easier to fiddle with
707        them, and removed "b" from the input mode under Windows.
708    
709    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
710    
711    12. Added -help and --help to pcretest as an official way of being reminded
712        of the options.
713    
714    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
715        and pcrecpp.cc because they annoy compilers at high warning levels.
716    
717    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
718    
719    15. Fixed an occurrence of == in configure.ac that should have been = (shell
720        scripts are not C programs :-) and which was not noticed because it works
721        on Linux.
722    
723    16. pcretest is supposed to handle any length of pattern and data line (as one
724        line or as a continued sequence of lines) by extending its input buffer if
725        necessary. This feature was broken for very long pattern lines, leading to
726        a string of junk being passed to pcre_compile() if the pattern was longer
727        than about 50K.
728    
729    17. I have done a major re-factoring of the way pcre_compile() computes the
730        amount of memory needed for a compiled pattern. Previously, there was code
731        that made a preliminary scan of the pattern in order to do this. That was
732        OK when PCRE was new, but as the facilities have expanded, it has become
733        harder and harder to keep it in step with the real compile phase, and there
734        have been a number of bugs (see for example, 4 above). I have now found a
735        cunning way of running the real compile function in a "fake" mode that
736        enables it to compute how much memory it would need, while actually only
737        ever using a few hundred bytes of working memory and without too many
738        tests of the mode. This should make future maintenance and development
739        easier. A side effect of this work is that the limit of 200 on the nesting
740        depth of parentheses has been removed (though this was never a serious
741        limitation, I suspect). However, there is a downside: pcre_compile() now
742        runs more slowly than before (30% or more, depending on the pattern). I
743        hope this isn't a big issue. There is no effect on runtime performance.
744    
745    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
746        newline (only possible for the last line of a file) and it was a
747        pattern that set a locale (followed by /Lsomething), pcretest crashed.
748    
749    19. Added additional timing features to pcretest. (1) The -tm option now times
750        matching only, not compiling. (2) Both -t and -tm can be followed, as a
751        separate command line item, by a number that specifies the number of
752        repeats to use when timing. The default is 50000; this gives better
753        precision, but takes uncomfortably long for very large patterns.
754    
755    20. Extended pcre_study() to be more clever in cases where a branch of a
756        subpattern has no definite first character. For example, (a*|b*)[cd] would
757        previously give no result from pcre_study(). Now it recognizes that the
758        first character must be a, b, c, or d.
759    
760    21. There was an incorrect error "recursive call could loop indefinitely" if
761        a subpattern (or the entire pattern) that was being tested for matching an
762        empty string contained only one non-empty item after a nested subpattern.
763        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
764        incorrectly, because the \d was being skipped in the check.
765    
766    22. The pcretest program now has a new pattern option /B and a command line
767        option -b, which is equivalent to adding /B to every pattern. This causes
768        it to show the compiled bytecode, without the additional information that
769        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
770        is the same as /B/I).
771    
772    23. A new optimization is now able automatically to treat some sequences such
773        as a*b as a*+b. More specifically, if something simple (such as a character
774        or a simple class like \d) has an unlimited quantifier, and is followed by
775        something that cannot possibly match the quantified thing, the quantifier
776        is automatically "possessified".
777    
778    24. A recursive reference to a subpattern whose number was greater than 39
779        went wrong under certain circumstances in UTF-8 mode. This bug could also
780        have affected the operation of pcre_study().
781    
782    25. Realized that a little bit of performance could be had by replacing
783        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
784    
785    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
786    
787    27. Possessive quantifiers such as a++ were previously implemented by turning
788        them into atomic groups such as ($>a+). Now they have their own opcodes,
789        which improves performance. This includes the automatically created ones
790        from 23 above.
791    
792    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
793        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
794        the first matched character to be a colon. This applied both to named and
795        numbered groups.
796    
797    29. The ucpinternal.h header file was missing its idempotency #ifdef.
798    
799    30. I was sent a "project" file called libpcre.a.dev which I understand makes
800        building PCRE on Windows easier, so I have included it in the distribution.
801    
802    31. There is now a check in pcretest against a ridiculously large number being
803        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
804        loop, the loop is abandoned.
805    
806    32. Forward references to subpatterns in conditions such as (?(2)...) where
807        subpattern 2 is defined later cause pcre_compile() to search forwards in
808        the pattern for the relevant set of parentheses. This search went wrong
809        when there were unescaped parentheses in a character class, parentheses
810        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
811    
812    33. "Subroutine" calls and backreferences were previously restricted to
813        referencing subpatterns earlier in the regex. This restriction has now
814        been removed.
815    
816    34. Added a number of extra features that are going to be in Perl 5.10. On the
817        whole, these are just syntactic alternatives for features that PCRE had
818        previously implemented using the Python syntax or my own invention. The
819        other formats are all retained for compatibility.
820    
821        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
822            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
823            also .NET compatible.
824    
825        (b) A recursion or subroutine call to a named group can now be defined as
826            (?&name) as well as (?P>name).
827    
828        (c) A backreference to a named group can now be defined as \k<name> or
829            \k'name' as well as (?P=name). The new forms, as well as being in Perl
830            5.10, are also .NET compatible.
831    
832        (d) A conditional reference to a named group can now use the syntax
833            (?(<name>) or (?('name') as well as (?(name).
834    
835        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
836            groups (named and numbered) that are never evaluated inline, but can be
837            called as "subroutines" from elsewhere. In effect, the DEFINE condition
838            is always false. There may be only one alternative in such a group.
839    
840        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
841            as the simple (?(R). The condition is true only if the most recent
842            recursion is that of the given number or name. It does not search out
843            through the entire recursion stack.
844    
845        (g) The escape \gN or \g{N} has been added, where N is a positive or
846            negative number, specifying an absolute or relative reference.
847    
848    35. Tidied to get rid of some further signed/unsigned compiler warnings and
849        some "unreachable code" warnings.
850    
851    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
852        things, this adds five new scripts.
853    
854    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
855        There were also incompatibilities regarding the handling of \Q..\E inside
856        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
857        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
858    
859    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
860        matches an empty string, and forcibly breaks the loop. There were bugs in
861        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
862        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
863        separate and independent bugs (that affected different cases) have been
864        fixed.
865    
866    39. Refactored the code to abolish the use of different opcodes for small
867        capturing bracket numbers. This is a tidy that I avoided doing when I
868        removed the limit on the number of capturing brackets for 3.5 back in 2001.
869        The new approach is not only tidier, it makes it possible to reduce the
870        memory needed to fix the previous bug (38).
871    
872    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
873        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
874        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
875        mode.
876    
877    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
878        report.
879    
880    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
881        copy construction and assignment in the C++ wrapper.
882    
883    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
884        couple of bugs that could have given wrong results in the "--newline=crlf"
885        case.
886    
887    44. Added a number of casts and did some reorganization of signed/unsigned int
888        variables following suggestions from Dair Grant. Also renamed the variable
889        "this" as "item" because it is a C++ keyword.
890    
891    45. Arranged for dftables to add
892    
893          #include "pcre_internal.h"
894    
895        to pcre_chartables.c because without it, gcc 4.x may remove the array
896        definition from the final binary if PCRE is built into a static library and
897        dead code stripping is activated.
898    
899    46. For an unanchored pattern, if a match attempt fails at the start of a
900        newline sequence, and the newline setting is CRLF or ANY, and the next two
901        characters are CRLF, advance by two characters instead of one.
902    
903    
904    Version 6.7 04-Jul-06
905    ---------------------
906    
907     1. In order to handle tests when input lines are enormously long, pcretest has
908        been re-factored so that it automatically extends its buffers when
909        necessary. The code is crude, but this _is_ just a test program. The
910        default size has been increased from 32K to 50K.
911    
912     2. The code in pcre_study() was using the value of the re argument before
913        testing it for NULL. (Of course, in any sensible call of the function, it
914        won't be NULL.)
915    
916     3. The memmove() emulation function in pcre_internal.h, which is used on
917        systems that lack both memmove() and bcopy() - that is, hardly ever -
918        was missing a "static" storage class specifier.
919    
920     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
921        containing an extended class (one that cannot be represented by a bitmap
922        because it contains high-valued characters or Unicode property items, e.g.
923        [\pZ]). Almost always one would set UTF-8 mode when processing such a
924        pattern, but PCRE should not loop if you do not (it no longer does).
925        [Detail: two cases were found: (a) a repeated subpattern containing an
926        extended class; (b) a recursive reference to a subpattern that followed a
927        previous extended class. It wasn't skipping over the extended class
928        correctly when UTF-8 mode was not set.]
929    
930     5. A negated single-character class was not being recognized as fixed-length
931        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
932        compile error "lookbehind assertion is not fixed length".
933    
934     6. The RunPerlTest auxiliary script was showing an unexpected difference
935        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
936        write a Perl script that can interpret lines of an input file either as
937        byte characters or as UTF-8, which is what "perltest" was being required to
938        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
939        can't do is switch easily at run time between having the "use utf8;" pragma
940        or not. In the end, I fudged it by using the RunPerlTest script to insert
941        "use utf8;" explicitly for the UTF-8 tests.
942    
943     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
944        the end of the subject string, contrary to the documentation and to what
945        Perl does. This was true of both matching functions. Now it matches only at
946        the start of the subject and immediately after *internal* newlines.
947    
948     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
949        a pointer to an int instead of a pointer to an unsigned long int. This
950        caused problems on 64-bit systems.
951    
952     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
953        instance of the 'standard' template library not being so standard".
954    
955    10. There was no check on the number of named subpatterns nor the maximum
956        length of a subpattern name. The product of these values is used to compute
957        the size of the memory block for a compiled pattern. By supplying a very
958        long subpattern name and a large number of named subpatterns, the size
959        computation could be caused to overflow. This is now prevented by limiting
960        the length of names to 32 characters, and the number of named subpatterns
961        to 10,000.
962    
963    11. Subpatterns that are repeated with specific counts have to be replicated in
964        the compiled pattern. The size of memory for this was computed from the
965        length of the subpattern and the repeat count. The latter is limited to
966        65535, but there was no limit on the former, meaning that integer overflow
967        could in principle occur. The compiled length of a repeated subpattern is
968        now limited to 30,000 bytes in order to prevent this.
969    
970    12. Added the optional facility to have named substrings with the same name.
971    
972    13. Added the ability to use a named substring as a condition, using the
973        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
974        are numbers (not recommended). Forward references are permitted.
975    
976    14. Added forward references in named backreferences (if you see what I mean).
977    
978    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
979        pattern could run off the end of the subject. For example, the pattern
980        "(?s)(.{1,5})"8 did this with the subject "ab".
981    
982    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
983        PCRE_CASELESS was set when matching characters that were quantified with ?
984        or *.
985    
986    17. A character class other than a single negated character that had a minimum
987        but no maximum quantifier - for example [ab]{6,} - was not handled
988        correctly by pce_dfa_exec(). It would match only one character.
989    
990    18. A valid (though odd) pattern that looked like a POSIX character
991        class but used an invalid character after [ (for example [[,abc,]]) caused
992        pcre_compile() to give the error "Failed: internal error: code overflow" or
993        in some cases to crash with a glibc free() error. This could even happen if
994        the pattern terminated after [[ but there just happened to be a sequence of
995        letters, a binary zero, and a closing ] in the memory that followed.
996    
997    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
998        over the years. Originally (before any Unicode support), just the bottom 8
999        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1000        output from "man perlunicode" includes this:
1001    
1002          The regular expression compiler produces polymorphic opcodes.  That
1003          is, the pattern adapts to the data and automatically switches to
1004          the Unicode character scheme when presented with Unicode data--or
1005          instead uses a traditional byte scheme when presented with byte
1006          data.
1007    
1008        Sadly, a wide octal escape does not cause a switch, and in a string with
1009        no other multibyte characters, these octal escapes are treated as before.
1010        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1011        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1012        Unicode string.
1013    
1014        I have not perpetrated such confusion in PCRE. Up till now, it took just
1015        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1016        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1017        translate to the appropriate multibyte character.
1018    
1019    29. Applied some refactoring to reduce the number of warnings from Microsoft
1020        and Borland compilers. This has included removing the fudge introduced
1021        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1022        a warning about an unused variable.
1023    
1024    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1025        characters since release 4.0, because Perl (from release 5.004) does not.
1026        [Or at least, is documented not to: some releases seem to be in conflict
1027        with the documentation.] However, when a pattern was studied with
1028        pcre_study() and all its branches started with \s, PCRE still included VT
1029        as a possible starting character. Of course, this did no harm; it just
1030        caused an unnecessary match attempt.
1031    
1032    22. Removed a now-redundant internal flag bit that recorded the fact that case
1033        dependency changed within the pattern. This was once needed for "required
1034        byte" processing, but is no longer used. This recovers a now-scarce options
1035        bit. Also moved the least significant internal flag bit to the most-
1036        significant bit of the word, which was not previously used (hangover from
1037        the days when it was an int rather than a uint) to free up another bit for
1038        the future.
1039    
1040    23. Added support for CRLF line endings as well as CR and LF. As well as the
1041        default being selectable at build time, it can now be changed at runtime
1042        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1043        specify that it is scanning data with non-default line endings.
1044    
1045    24. Changed the definition of CXXLINK to make it agree with the definition of
1046        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1047    
1048    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1049        recursions. This makes a big different to stack usage for some patterns.
1050    
1051    26. If a subpattern containing a named recursion or subroutine reference such
1052        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1053        the space required for the compiled pattern went wrong and gave too small a
1054        value. Depending on the environment, this could lead to "Failed: internal
1055        error: code overflow at offset 49" or "glibc detected double free or
1056        corruption" errors.
1057    
1058    27. Applied patches from Google (a) to support the new newline modes and (b) to
1059        advance over multibyte UTF-8 characters in GlobalReplace.
1060    
1061    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1062        difference for some implementation of PCRE in some Windows version.
1063    
1064    29. Added some extra testing facilities to pcretest:
1065    
1066        \q<number>   in a data line sets the "match limit" value
1067        \Q<number>   in a data line sets the "match recursion limt" value
1068        -S <number>  sets the stack size, where <number> is in megabytes
1069    
1070        The -S option isn't available for Windows.
1071    
1072    
1073    Version 6.6 06-Feb-06
1074    ---------------------
1075    
1076     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1077        in pcreposix.h. I have copied the definition from pcre.h.
1078    
1079     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1080        because pcre.h is no longer a built file.
1081    
1082     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1083        not normally included in the compiled code.
1084    
1085    
1086    Version 6.5 01-Feb-06
1087    ---------------------
1088    
1089     1. When using the partial match feature with pcre_dfa_exec(), it was not
1090        anchoring the second and subsequent partial matches at the new starting
1091        point. This could lead to incorrect results. For example, with the pattern
1092        /1234/, partially matching against "123" and then "a4" gave a match.
1093    
1094     2. Changes to pcregrep:
1095    
1096        (a) All non-match returns from pcre_exec() were being treated as failures
1097            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1098            error message is output. Some extra information is given for the
1099            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1100            probably the only errors that are likely to be caused by users (by
1101            specifying a regex that has nested indefinite repeats, for instance).
1102            If there are more than 20 of these errors, pcregrep is abandoned.
1103    
1104        (b) A binary zero was treated as data while matching, but terminated the
1105            output line if it was written out. This has been fixed: binary zeroes
1106            are now no different to any other data bytes.
1107    
1108        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1109            used to set a locale for matching. The --locale=xxxx long option has
1110            been added (no short equivalent) to specify a locale explicitly on the
1111            pcregrep command, overriding the environment variables.
1112    
1113        (d) When -B was used with -n, some line numbers in the output were one less
1114            than they should have been.
1115    
1116        (e) Added the -o (--only-matching) option.
1117    
1118        (f) If -A or -C was used with -c (count only), some lines of context were
1119            accidentally printed for the final match.
1120    
1121        (g) Added the -H (--with-filename) option.
1122    
1123        (h) The combination of options -rh failed to suppress file names for files
1124            that were found from directory arguments.
1125    
1126        (i) Added the -D (--devices) and -d (--directories) options.
1127    
1128        (j) Added the -F (--fixed-strings) option.
1129    
1130        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1131    
1132        (l) Added the --colo(u)r option.
1133    
1134        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1135            is not present by default.
1136    
1137     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1138        items such as (?R) or (?1), when the recursion could match a number of
1139        alternatives. If it matched one of the alternatives, but subsequently,
1140        outside the recursion, there was a failure, the code tried to back up into
1141        the recursion. However, because of the way PCRE is implemented, this is not
1142        possible, and the result was an incorrect result from the match.
1143    
1144        In order to prevent this happening, the specification of recursion has
1145        been changed so that all such subpatterns are automatically treated as
1146        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1147    
1148     4. I had overlooked the fact that, in some locales, there are characters for
1149        which isalpha() is true but neither isupper() nor islower() are true. In
1150        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1151        and ordfeminine) are like this. This affected the treatment of \w and \W
1152        when they appeared in character classes, but not when they appeared outside
1153        a character class. The bit map for "word" characters is now created
1154        separately from the results of isalnum() instead of just taking it from the
1155        upper, lower, and digit maps. (Plus the underscore character, of course.)
1156    
1157     5. The above bug also affected the handling of POSIX character classes such as
1158        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1159        permanent tables. Instead, the bit maps for such a class were previously
1160        created as the appropriate unions of the upper, lower, and digit bitmaps.
1161        Now they are created by subtraction from the [[:word:]] class, which has
1162        its own bitmap.
1163    
1164     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1165        It is created by subtracting the vertical space characters (\x09, \x0a,
1166        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1167        subtraction was done in the overall bitmap for a character class, meaning
1168        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1169        be recognized. This bug has been fixed.
1170    
1171     7. Patches from the folks at Google:
1172    
1173          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1174          real life, but is still worth protecting against".
1175    
1176          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1177          regular expressions".
1178    
1179          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1180          have it.
1181    
1182          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1183          "configure" and the latter not, in order to fix a problem somebody had
1184          with compiling the Arg class on HP-UX.
1185    
1186          (e) Improve the error-handling of the C++ wrapper a little bit.
1187    
1188          (f) New tests for checking recursion limiting.
1189    
1190     8. The pcre_memmove() function, which is used only if the environment does not
1191        have a standard memmove() function (and is therefore rarely compiled),
1192        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1193        returning a result (though PCRE never actually uses the result).
1194    
1195     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1196        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1197        returned instead of calling malloc() with an overflowing number that would
1198        most likely cause subsequent chaos.
1199    
1200    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1201    
1202    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1203        with this option is matched, the nmatch and pmatch options of regexec() are
1204        ignored.
1205    
1206    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1207        provided in case anyone wants to the the POSIX interface with UTF-8
1208        strings.
1209    
1210    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1211        C++ linking (needed for some HP-UX environments).
1212    
1213    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1214        (unused parameter) and in the pcre_printint() function (omitted "default"
1215        switch label when the default is to do nothing).
1216    
1217    15. Added some code to make it possible, when PCRE is compiled as a C++
1218        library, to replace subject pointers for pcre_exec() with a smart pointer
1219        class, thus making it possible to process discontinuous strings.
1220    
1221    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1222        much the same function. They were added by different people who were trying
1223        to make PCRE easy to compile on non-Unix systems. It has been suggested
1224        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1225        for compiling on Windows systems. I have therefore replaced it with
1226        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1227        defaults to "extern" for C or "extern C" for C++, which works fine on
1228        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1229        SCOPE with something explicit in config.h. In addition:
1230    
1231        (a) pcreposix.h still had just "extern" instead of either of these macros;
1232            I have replaced it with PCRE_DATA_SCOPE.
1233    
1234        (b) Functions such as _pcre_xclass(), which are internal to the library,
1235            but external in the C sense, all had PCRE_EXPORT in their definitions.
1236            This is apparently wrong for the Windows case, so I have removed it.
1237            (It makes no difference on Unix-like systems.)
1238    
1239    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1240        of recursive calls to match(). This is different to MATCH_LIMIT because
1241        that limits the total number of calls to match(), not all of which increase
1242        the depth of recursion. Limiting the recursion depth limits the amount of
1243        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1244        when PCRE is compiled, and changed at run time. A patch from Google adds
1245        this functionality to the C++ interface.
1246    
1247    18. Changes to the handling of Unicode character properties:
1248    
1249        (a) Updated the table to Unicode 4.1.0.
1250    
1251        (b) Recognize characters that are not in the table as "Cn" (undefined).
1252    
1253        (c) I revised the way the table is implemented to a much improved format
1254            which includes recognition of ranges. It now supports the ranges that
1255            are defined in UnicodeData.txt, and it also amalgamates other
1256            characters into ranges. This has reduced the number of entries in the
1257            table from around 16,000 to around 3,000, thus reducing its size
1258            considerably. I realized I did not need to use a tree structure after
1259            all - a binary chop search is just as efficient. Having reduced the
1260            number of entries, I extended their size from 6 bytes to 8 bytes to
1261            allow for more data.
1262    
1263        (d) Added support for Unicode script names via properties such as \p{Han}.
1264    
1265    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1266        matching that character.
1267    
1268    20. When matching a repeated Unicode property with a minimum greater than zero,
1269        (for example \pL{2,}), PCRE could look past the end of the subject if it
1270        reached it while seeking the minimum number of characters. This could
1271        happen only if some of the characters were more than one byte long, because
1272        there is a check for at least the minimum number of bytes.
1273    
1274    21. Refactored the implementation of \p and \P so as to be more general, to
1275        allow for more different types of property in future. This has changed the
1276        compiled form incompatibly. Anybody with saved compiled patterns that use
1277        \p or \P will have to recompile them.
1278    
1279    22. Added "Any" and "L&" to the supported property types.
1280    
1281    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1282        but give a compile time error if the value is greater than 0xff.
1283    
1284    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1285        accidentally not being installed or uninstalled.
1286    
1287    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1288        made were to insert the current release number. This seemed silly, because
1289        it made things harder for people building PCRE on systems that don't run
1290        "configure". I have turned pcre.h into a distributed file, no longer built
1291        by "configure", with the version identification directly included. There is
1292        no longer a pcre.h.in file.
1293    
1294        However, this change necessitated a change to the pcre-config script as
1295        well. It is built from pcre-config.in, and one of the substitutions was the
1296        release number. I have updated configure.ac so that ./configure now finds
1297        the release number by grepping pcre.h.
1298    
1299    26. Added the ability to run the tests under valgrind.
1300    
1301    
1302    Version 6.4 05-Sep-05
1303    ---------------------
1304    
1305     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1306        "--" to be printed when multiple files were scanned, even when none of the
1307        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1308        consider it to be a bug, and have restored the previous behaviour.
1309    
1310     2. A couple of code tidies to get rid of compiler warnings.
1311    
1312     3. The pcretest program used to cheat by referring to symbols in the library
1313        whose names begin with _pcre_. These are internal symbols that are not
1314        really supposed to be visible externally, and in some environments it is
1315        possible to suppress them. The cheating is now confined to including
1316        certain files from the library's source, which is a bit cleaner.
1317    
1318     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1319        file's purpose clearer.
1320    
1321     5. Reorganized pcre_ucp_findchar().
1322    
1323    
1324    Version 6.3 15-Aug-05
1325    ---------------------
1326    
1327     1. The file libpcre.pc.in did not have general read permission in the tarball.
1328    
1329     2. There were some problems when building without C++ support:
1330    
1331        (a) If C++ support was not built, "make install" and "make test" still
1332            tried to test it.
1333    
1334        (b) There were problems when the value of CXX was explicitly set. Some
1335            changes have been made to try to fix these, and ...
1336    
1337        (c) --disable-cpp can now be used to explicitly disable C++ support.
1338    
1339        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1340            backslash in a target when C++ was disabled. This confuses some
1341            versions of "make", apparently. Using an intermediate variable solves
1342            this. (Same for CPP_LOBJ.)
1343    
1344     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1345        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1346        necessary on certain architectures.
1347    
1348     4. Added a setting of -export-symbols-regex to the link command to remove
1349        those symbols that are exported in the C sense, but actually are local
1350        within the library, and not documented. Their names all begin with
1351        "_pcre_". This is not a perfect job, because (a) we have to except some
1352        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1353        available (and never for static libraries). I have made a note to try to
1354        find a way round (a) in the future.
1355    
1356    
1357    Version 6.2 01-Aug-05
1358    ---------------------
1359    
1360     1. There was no test for integer overflow of quantifier values. A construction
1361        such as {1111111111111111} would give undefined results. What is worse, if
1362        a minimum quantifier for a parenthesized subpattern overflowed and became
1363        negative, the calculation of the memory size went wrong. This could have
1364        led to memory overwriting.
1365    
1366     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1367    
1368     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1369        operating environments where this matters.
1370    
1371     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1372        PCRE options from within the C++ wrapper.
1373    
1374     5. Named capturing subpatterns were not being correctly counted when a pattern
1375        was compiled. This caused two problems: (a) If there were more than 100
1376        such subpatterns, the calculation of the memory needed for the whole
1377        compiled pattern went wrong, leading to an overflow error. (b) Numerical
1378        back references of the form \12, where the number was greater than 9, were
1379        not recognized as back references, even though there were sufficient
1380        previous subpatterns.
1381    
1382     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1383        versions of gcc, e.g. 2.95.4.
1384    
1385    
1386    Version 6.1 21-Jun-05
1387    ---------------------
1388    
1389     1. There was one reference to the variable "posix" in pcretest.c that was not
1390        surrounded by "#if !defined NOPOSIX".
1391    
1392     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1393        the cross-check on the old pcre_info() function, for the benefit of the
1394        cut-down version of PCRE that is currently imported into Exim.
1395    
1396     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1397        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1398        patterns that start (?i) but I don't think that matters. The use of (?i) is
1399        just an example; this all applies to the other options as well.
1400    
1401     4. Since libtool seems to echo the compile commands it is issuing, the output
1402        from "make" can be reduced a bit by putting "@" in front of each libtool
1403        compile command.
1404    
1405     5. Patch from the folks at Google for configure.in to be a bit more thorough
1406        in checking for a suitable C++ installation before trying to compile the
1407        C++ stuff. This should fix a reported problem when a compiler was present,
1408        but no suitable headers.
1409    
1410     6. The man pages all had just "PCRE" as their title. I have changed them to
1411        be the relevant file name. I have also arranged that these names are
1412        retained in the file doc/pcre.txt, which is a concatenation in text format
1413        of all the man pages except the little individual ones for each function.
1414    
1415     7. The NON-UNIX-USE file had not been updated for the different set of source
1416        files that come with release 6. I also added a few comments about the C++
1417        wrapper.
1418    
1419    
1420    Version 6.0 07-Jun-05
1421    ---------------------
1422    
1423     1. Some minor internal re-organization to help with my DFA experiments.
1424    
1425     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1426        didn't matter for the library itself when fully configured, but did matter
1427        when compiling without UCP support, or within Exim, where the ucp files are
1428        not imported.
1429    
1430     3. Refactoring of the library code to split up the various functions into
1431        different source modules. The addition of the new DFA matching code (see
1432        below) to a single monolithic source would have made it really too
1433        unwieldy, quite apart from causing all the code to be include in a
1434        statically linked application, when only some functions are used. This is
1435        relevant even without the DFA addition now that patterns can be compiled in
1436        one application and matched in another.
1437    
1438        The downside of splitting up is that there have to be some external
1439        functions and data tables that are used internally in different modules of
1440        the library but which are not part of the API. These have all had their
1441        names changed to start with "_pcre_" so that they are unlikely to clash
1442        with other external names.
1443    
1444     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1445        a different (DFA) algorithm. Although it is slower than the original
1446        function, it does have some advantages for certain types of matching
1447        problem.
1448    
1449     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1450        including restarting after a partial match.
1451    
1452     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1453        defined when compiling for Windows was sent to me. I have put it into the
1454        code, though I have no means of testing or verifying it.
1455    
1456     7. Added the pcre_refcount() auxiliary function.
1457    
1458     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1459        match before or at the first newline in the subject string. In pcretest,
1460        the /f option on a pattern can be used to set this.
1461    
1462     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1463        would behave wrongly. This has been present in PCRE since release 4.0.
1464    
1465    10. A number of changes to the pcregrep command:
1466    
1467        (a) Refactored how -x works; insert ^(...)$ instead of setting
1468            PCRE_ANCHORED and checking the length, in preparation for adding
1469            something similar for -w.
1470    
1471        (b) Added the -w (match as a word) option.
1472    
1473        (c) Refactored the way lines are read and buffered so as to have more
1474            than one at a time available.
1475    
1476        (d) Implemented a pcregrep test script.
1477    
1478        (e) Added the -M (multiline match) option. This allows patterns to match
1479            over several lines of the subject. The buffering ensures that at least
1480            8K, or the rest of the document (whichever is the shorter) is available
1481            for matching (and similarly the previous 8K for lookbehind assertions).
1482    
1483        (f) Changed the --help output so that it now says
1484    
1485              -w, --word-regex(p)
1486    
1487            instead of two lines, one with "regex" and the other with "regexp"
1488            because that confused at least one person since the short forms are the
1489            same. (This required a bit of code, as the output is generated
1490            automatically from a table. It wasn't just a text change.)
1491    
1492        (g) -- can be used to terminate pcregrep options if the next thing isn't an
1493            option but starts with a hyphen. Could be a pattern or a path name
1494            starting with a hyphen, for instance.
1495    
1496        (h) "-" can be given as a file name to represent stdin.
1497    
1498        (i) When file names are being printed, "(standard input)" is used for
1499            the standard input, for compatibility with GNU grep. Previously
1500            "<stdin>" was used.
1501    
1502        (j) The option --label=xxx can be used to supply a name to be used for
1503            stdin when file names are being printed. There is no short form.
1504    
1505        (k) Re-factored the options decoding logic because we are going to add
1506            two more options that take data. Such options can now be given in four
1507            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1508    
1509        (l) Added the -A, -B, and -C options for requesting that lines of context
1510            around matches be printed.
1511    
1512        (m) Added the -L option to print the names of files that do not contain
1513            any matching lines, that is, the complement of -l.
1514    
1515        (n) The return code is 2 if any file cannot be opened, but pcregrep does
1516            continue to scan other files.
1517    
1518        (o) The -s option was incorrectly implemented. For compatibility with other
1519            greps, it now suppresses the error message for a non-existent or non-
1520            accessible file (but not the return code). There is a new option called
1521            -q that suppresses the output of matching lines, which was what -s was
1522            previously doing.
1523    
1524        (p) Added --include and --exclude options to specify files for inclusion
1525            and exclusion when recursing.
1526    
1527    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1528        Hopefully, it now does.
1529    
1530    12. Missing cast in pcre_study().
1531    
1532    13. Added an "uninstall" target to the makefile.
1533    
1534    14. Replaced "extern" in the function prototypes in Makefile.in with
1535        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1536        world, but is set differently for Windows.
1537    
1538    15. Added a second compiling function called pcre_compile2(). The only
1539        difference is that it has an extra argument, which is a pointer to an
1540        integer error code. When there is a compile-time failure, this is set
1541        non-zero, in addition to the error test pointer being set to point to an
1542        error message. The new argument may be NULL if no error number is required
1543        (but then you may as well call pcre_compile(), which is now just a
1544        wrapper). This facility is provided because some applications need a
1545        numeric error indication, but it has also enabled me to tidy up the way
1546        compile-time errors are handled in the POSIX wrapper.
1547    
1548    16. Added VPATH=.libs to the makefile; this should help when building with one
1549        prefix path and installing with another. (Or so I'm told by someone who
1550        knows more about this stuff than I do.)
1551    
1552    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1553        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1554        match everything, including newlines. This is not POSIX-compatible, but
1555        somebody wanted the feature. From pcretest it can be activated by using
1556        both the P and the s flags.
1557    
1558    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1559    
1560    19. libpcre.pc was being incorrectly installed as executable.
1561    
1562    20. A couple of places in pcretest check for end-of-line by looking for '\n';
1563        it now also looks for '\r' so that it will work unmodified on Windows.
1564    
1565    21. Added Google's contributed C++ wrapper to the distribution.
1566    
1567    22. Added some untidy missing memory free() calls in pcretest, to keep
1568        Electric Fence happy when testing.
1569    
1570    
1571    
1572    Version 5.0 13-Sep-04
1573    ---------------------
1574    
1575     1. Internal change: literal characters are no longer packed up into items
1576        containing multiple characters in a single byte-string. Each character
1577        is now matched using a separate opcode. However, there may be more than one
1578        byte in the character in UTF-8 mode.
1579    
1580     2. The pcre_callout_block structure has two new fields: pattern_position and
1581        next_item_length. These contain the offset in the pattern to the next match
1582        item, and its length, respectively.
1583    
1584     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1585        insertion of callouts before each pattern item. Added the /C option to
1586        pcretest to make use of this.
1587    
1588     4. On the advice of a Windows user, the lines
1589    
1590          #if defined(_WIN32) || defined(WIN32)
1591          _setmode( _fileno( stdout ), 0x8000 );
1592          #endif  /* defined(_WIN32) || defined(WIN32) */
1593    
1594        have been added to the source of pcretest. This apparently does useful
1595        magic in relation to line terminators.
1596    
1597     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1598        for the benefit of those environments where the "b" makes a difference.
1599    
1600     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1601        to know about it. I have put a hack into configure.in that adds in code
1602        to set GCC=yes if CC=icc. This seems to end up at a point in the
1603        generated configure script that is early enough to affect the setting of
1604        compiler options, which is what is needed, but I have no means of testing
1605        whether it really works. (The user who reported this had patched the
1606        generated configure script, which of course I cannot do.)
1607    
1608        LATER: After change 22 below (new libtool files), the configure script
1609        seems to know about icc (and also ecc). Therefore, I have commented out
1610        this hack in configure.in.
1611    
1612     7. Added support for pkg-config (2 patches were sent in).
1613    
1614     8. Negated POSIX character classes that used a combination of internal tables
1615        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1616        [[:^ascii]]. Typically, they would match almost any characters. The other
1617        POSIX classes were not broken in this way.
1618    
1619     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1620        to find the match, as PCRE was deluded into thinking that the match had to
1621        start at the start point or following a newline. The same bug applied to
1622        patterns with negative forward assertions or any backward assertions
1623        preceding ".*" at the start, unless the pattern required a fixed first
1624        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1625    
1626    10. In UTF-8 mode, when moving forwards in the subject after a failed match
1627        starting at the last subject character, bytes beyond the end of the subject
1628        string were read.
1629    
1630    11. Renamed the variable "class" as "classbits" to make life easier for C++
1631        users. (Previously there was a macro definition, but it apparently wasn't
1632        enough.)
1633    
1634    12. Added the new field "tables" to the extra data so that tables can be passed
1635        in at exec time, or the internal tables can be re-selected. This allows
1636        a compiled regex to be saved and re-used at a later time by a different
1637        program that might have everything at different addresses.
1638    
1639    13. Modified the pcre-config script so that, when run on Solaris, it shows a
1640        -R library as well as a -L library.
1641    
1642    14. The debugging options of pcretest (-d on the command line or D on a
1643        pattern) showed incorrect output for anything following an extended class
1644        that contained multibyte characters and which was followed by a quantifier.
1645    
1646    15. Added optional support for general category Unicode character properties
1647        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1648        support. It adds about 90K to the size of the library. The meanings of the
1649        inbuilt class escapes such as \d and \s have NOT been changed.
1650    
1651    16. Updated pcredemo.c to include calls to free() to release the memory for the
1652        compiled pattern.
1653    
1654    17. The generated file chartables.c was being created in the source directory
1655        instead of in the building directory. This caused the build to fail if the
1656        source directory was different from the building directory, and was
1657        read-only.
1658    
1659    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1660        file. No doubt somebody will tell me if they don't make sense... Also added
1661        Dan Mooney's comments about building on OpenVMS.
1662    
1663    19. Added support for partial matching via the PCRE_PARTIAL option for
1664        pcre_exec() and the \P data escape in pcretest.
1665    
1666    20. Extended pcretest with 3 new pattern features:
1667    
1668        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
1669              write the compiled pattern to the file whose name is "rest-of-line".
1670              This is a straight binary dump of the data, with the saved pointer to
1671              the character tables forced to be NULL. The study data, if any, is
1672              written too. After writing, pcretest reads a new pattern.
1673    
1674        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1675              compiled pattern from the given file. There must not be any
1676              occurrences of "<" in the file name (pretty unlikely); if there are,
1677              pcretest will instead treat the initial "<" as a pattern delimiter.
1678              After reading in the pattern, pcretest goes on to read data lines as
1679              usual.
1680    
1681        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1682              and 16-bit fields in a compiled pattern, to simulate a pattern that
1683              was compiled on a host of opposite endianness.
1684    
1685    21. The pcre-exec() function can now cope with patterns that were compiled on
1686        hosts of opposite endianness, with this restriction:
1687    
1688          As for any compiled expression that is saved and used later, the tables
1689          pointer field cannot be preserved; the extra_data field in the arguments
1690          to pcre_exec() should be used to pass in a tables address if a value
1691          other than the default internal tables were used at compile time.
1692    
1693    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1694        now diagnosed as an error. Previously, most of the time, a negative number
1695        would have been treated as zero, but if in addition "ovector" was passed as
1696        NULL, a crash could occur.
1697    
1698    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1699        new versions from the libtool 1.5 distribution (the last one is a copy of
1700        a file called libtool.m4). This seems to have fixed the need to patch
1701        "configure" to support Darwin 1.3 (which I used to do). However, I still
1702        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1703        workstation).
1704    
1705    24. Changed the PCRE licence to be the more standard "BSD" licence.
1706    
1707    
1708    Version 4.5 01-Dec-03
1709    ---------------------
1710    
1711     1. There has been some re-arrangement of the code for the match() function so
1712        that it can be compiled in a version that does not call itself recursively.
1713        Instead, it keeps those local variables that need separate instances for
1714        each "recursion" in a frame on the heap, and gets/frees frames whenever it
1715        needs to "recurse". Keeping track of where control must go is done by means
1716        of setjmp/longjmp. The whole thing is implemented by a set of macros that
1717        hide most of the details from the main code, and operates only if
1718        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1719        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1720        operating.
1721    
1722        To make it easier for callers to provide specially tailored get/free
1723        functions for this usage, two new functions, pcre_stack_malloc, and
1724        pcre_stack_free, are used. They are always called in strict stacking order,
1725        and the size of block requested is always the same.
1726    
1727        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1728        PCRE has been compiled to use the stack or the heap for recursion. The
1729        -C option of pcretest uses this to show which version is compiled.
1730    
1731        A new data escape \S, is added to pcretest; it causes the amounts of store
1732        obtained and freed by both kinds of malloc/free at match time to be added
1733        to the output.
1734    
1735     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1736        what's available on my current Linux desktop machine.
1737    
1738     3. When matching a UTF-8 string, the test for a valid string at the start has
1739        been extended. If start_offset is not zero, PCRE now checks that it points
1740        to a byte that is the start of a UTF-8 character. If not, it returns
1741        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1742        this is necessary because there may be backward assertions in the pattern.
1743        When matching the same subject several times, it may save resources to use
1744        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1745    
1746     4. The code for checking the validity of UTF-8 strings has been tightened so
1747        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1748        containing "overlong sequences".
1749    
1750     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1751        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1752        should have been "&", but it just so happened that all the cases this let
1753        through by mistake were picked up later in the function.
1754    
1755     6. I had used a variable called "isblank" - this is a C99 function, causing
1756        some compilers to warn. To avoid this, I renamed it (as "blankclass").
1757    
1758     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1759        prompting; (b) run "./pcretest /dev/null" at the start of the test script
1760        so the version is shown; (c) stop "make test" echoing "./RunTest".
1761    
1762     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1763    
1764     9. The prototype for memmove() for systems that don't have it was using
1765        size_t, but the inclusion of the header that defines size_t was later. I've
1766        moved the #includes for the C headers earlier to avoid this.
1767    
1768    10. Added some adjustments to the code to make it easier to compiler on certain
1769        special systems:
1770    
1771          (a) Some "const" qualifiers were missing.
1772          (b) Added the macro EXPORT before all exported functions; by default this
1773              is defined to be empty.
1774          (c) Changed the dftables auxiliary program (that builds chartables.c) so
1775              that it reads its output file name as an argument instead of writing
1776              to the standard output and assuming this can be redirected.
1777    
1778    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1779        class containing characters with values greater than 255, PCRE compilation
1780        went into a loop.
1781    
1782    12. A recursive reference to a subpattern that was within another subpattern
1783        that had a minimum quantifier of zero caused PCRE to crash. For example,
1784        (x(y(?2))z)? provoked this bug with a subject that got as far as the
1785        recursion. If the recursively-called subpattern itself had a zero repeat,
1786        that was OK.
1787    
1788    13. In pcretest, the buffer for reading a data line was set at 30K, but the
1789        buffer into which it was copied (for escape processing) was still set at
1790        1024, so long lines caused crashes.
1791    
1792    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1793        "internal error: code overflow...". This applied to any character class
1794        that was followed by a possessive quantifier.
1795    
1796    15. Modified the Makefile to add libpcre.la as a prerequisite for
1797        libpcreposix.la because I was told this is needed for a parallel build to
1798        work.
1799    
1800    16. If a pattern that contained .* following optional items at the start was
1801        studied, the wrong optimizing data was generated, leading to matching
1802        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1803        matching string must start with a or b or c. The correct conclusion for
1804        this pattern is that a match can start with any character.
1805    
1806    
1807    Version 4.4 13-Aug-03
1808    ---------------------
1809    
1810     1. In UTF-8 mode, a character class containing characters with values between
1811        127 and 255 was not handled correctly if the compiled pattern was studied.
1812        In fixing this, I have also improved the studying algorithm for such
1813        classes (slightly).
1814    
1815     2. Three internal functions had redundant arguments passed to them. Removal
1816        might give a very teeny performance improvement.
1817    
1818     3. Documentation bug: the value of the capture_top field in a callout is *one
1819        more than* the number of the hightest numbered captured substring.
1820    
1821     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1822        in incorrectly linking with a previously installed version. They now link
1823        explicitly with libpcre.la.
1824    
1825     5. configure.in no longer needs to recognize Cygwin specially.
1826    
1827     6. A problem in pcre.in for Windows platforms is fixed.
1828    
1829     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1830        pcretest, it used to include the size of the study block as part of its
1831        output. Unfortunately, the structure contains a field that has a different
1832        size on different hardware architectures. This meant that the tests that
1833        showed this size failed. As the block is currently always of a fixed size,
1834        this information isn't actually particularly useful in pcretest output, so
1835        I have just removed it.
1836    
1837     8. Three pre-processor statements accidentally did not start in column 1.
1838        Sadly, there are *still* compilers around that complain, even though
1839        standard C has not required this for well over a decade. Sigh.
1840    
1841     9. In pcretest, the code for checking callouts passed small integers in the
1842        callout_data field, which is a void * field. However, some picky compilers
1843        complained about the casts involved for this on 64-bit systems. Now
1844        pcretest passes the address of the small integer instead, which should get
1845        rid of the warnings.
1846    
1847    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1848        both compile and run time, and gives an error if an invalid UTF-8 sequence
1849        is found. There is a option for disabling this check in cases where the
1850        string is known to be correct and/or the maximum performance is wanted.
1851    
1852    11. In response to a bug report, I changed one line in Makefile.in from
1853    
1854            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1855        to
1856            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1857    
1858        to look similar to other lines, but I have no way of telling whether this
1859        is the right thing to do, as I do not use Windows. No doubt I'll get told
1860        if it's wrong...
1861    
1862    
1863    Version 4.3 21-May-03
1864    ---------------------
1865    
1866    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1867       Makefile.
1868    
1869    2. Some refactoring to improve the quality of the code:
1870    
1871       (i)   The utf8_table... variables are now declared "const".
1872    
1873       (ii)  The code for \cx, which used the "case flipping" table to upper case
1874             lower case letters, now just substracts 32. This is ASCII-specific,
1875             but the whole concept of \cx is ASCII-specific, so it seems
1876             reasonable.
1877    
1878       (iii) PCRE was using its character types table to recognize decimal and
1879             hexadecimal digits in the pattern. This is silly, because it handles
1880             only 0-9, a-f, and A-F, but the character types table is locale-
1881             specific, which means strange things might happen. A private
1882             table is now used for this - though it costs 256 bytes, a table is
1883             much faster than multiple explicit tests. Of course, the standard
1884             character types table is still used for matching digits in subject
1885             strings against \d.
1886    
1887       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1888             ending in _t are). So I've renamed it as ESC_tee.
1889    
1890    3. The first argument for regexec() in the POSIX wrapper should have been
1891       defined as "const".
1892    
1893    4. Changed pcretest to use malloc() for its buffers so that they can be
1894       Electric Fenced for debugging.
1895    
1896    5. There were several places in the code where, in UTF-8 mode, PCRE would try
1897       to read one or more bytes before the start of the subject string. Often this
1898       had no effect on PCRE's behaviour, but in some circumstances it could
1899       provoke a segmentation fault.
1900    
1901    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1902       to try to read one or more bytes before the start of the subject string.
1903    
1904    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1905       UTF-8 support could misbehave in various ways if the subject string
1906       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1907       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1908       back over UTF-8 characters.)
1909    
1910    
1911    Version 4.2 14-Apr-03
1912    ---------------------
1913    
1914    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1915    
1916    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1917         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1918         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1919         [WIN_PREFIX]: new variable, "cyg" for Cygwin
1920         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1921           and BUILD_EXEEXT
1922         Note: automatic setting of the BUILD variables is not yet working
1923         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1924           compile-time but not at link-time
1925         [LINK]: use for linking executables only
1926         make different versions for Windows and non-Windows
1927         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1928           libraries
1929         [LINK_FOR_BUILD]: new variable
1930         [OBJEXT]: use throughout
1931         [EXEEXT]: use throughout
1932         <winshared>: new target
1933         <wininstall>: new target
1934         <dftables.o>: use native compiler
1935         <dftables>: use native linker
1936         <install>: handle Windows platform correctly
1937         <clean>: ditto
1938         <check>: ditto
1939         copy DLL to top builddir before testing
1940    
1941       As part of these changes, -no-undefined was removed again. This was reported
1942       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1943       in any case.
1944    
1945    3. Some tidies to get rid of compiler warnings:
1946    
1947       . In the match_data structure, match_limit was an unsigned long int, whereas
1948         match_call_count was an int. I've made them both unsigned long ints.
1949    
1950       . In pcretest the fact that a const uschar * doesn't automatically cast to
1951         a void * provoked a warning.
1952    
1953       . Turning on some more compiler warnings threw up some "shadow" variables
1954         and a few more missing casts.
1955    
1956    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1957       option, a class that contained a single character with a value between 128
1958       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1959    
1960    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1961       option, a class that contained several characters, but with at least one
1962       whose value was between 128 and 255 caused PCRE to crash.
1963    
1964    
1965    Version 4.1 12-Mar-03
1966    ---------------------
1967    
1968    1. Compiling with gcc -pedantic found a couple of places where casts were
1969    needed, and a string in dftables.c that was longer than standard compilers are
1970    required to support.
1971    
1972    2. Compiling with Sun's compiler found a few more places where the code could
1973    be tidied up in order to avoid warnings.
1974    
1975    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1976    first of these names is deprecated in the latest Autoconf in favour of the name
1977    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1978    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1979    analogy I have changed it to CFLAGS_FOR_BUILD.
1980    
1981    4. Added -no-undefined to the linking command in the Makefile, because this is
1982    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1983    linking step for the pcreposix library.
1984    
1985    5. PCRE was failing to diagnose the case of two named groups with the same
1986    name.
1987    
1988    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1989    literal character that is needed in the subject for a match, and scans along to
1990    ensure that it is present before embarking on the full matching process. This
1991    saves time in cases of nested unlimited repeats that are never going to match.
1992    Problem: the scan can take a lot of time if the subject is very long (e.g.
1993    megabytes), thus penalizing straightforward matches. It is now done only if the
1994    amount of subject to be scanned is less than 1000 bytes.
1995    
1996    7. A lesser problem with the same optimization is that it was recording the
1997    first character of an anchored pattern as "needed", thus provoking a search
1998    right along the subject, even when the first match of the pattern was going to
1999    fail. The "needed" character is now not set for anchored patterns, unless it
2000    follows something in the pattern that is of non-fixed length. Thus, it still
2001    fulfils its original purpose of finding quick non-matches in cases of nested
2002    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2003    
2004    
2005    Version 4.0 17-Feb-03
2006    ---------------------
2007    
2008    1. If a comment in an extended regex that started immediately after a meta-item
2009    extended to the end of string, PCRE compiled incorrect data. This could lead to
2010    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2011    
2012    2. Moved to autoconf 2.53 and libtool 1.4.2.
2013    
2014    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2015    the special perltest8 script is no longer needed - all the tests can be run
2016    from a single perltest script.
2017    
2018    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2019    by \s. It has now been removed in PCRE. This means it isn't recognized as
2020    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2021    class [:space:] *does* include VT, thereby creating a mess.
2022    
2023    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2024    space and tab.
2025    
2026    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2027    its new features into the main test script, reducing the number of scripts.
2028    
2029    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2030    were backward compatible, and made the (?i) apply to the whole pattern, as if
2031    /i were given. Now it behaves more logically, and applies the option setting
2032    only to what follows. PCRE has been changed to follow suit. However, if it
2033    finds options settings right at the start of the pattern, it extracts them into
2034    the global options, as before. Thus, they show up in the info data.
2035    
2036    8. Added support for the \Q...\E escape sequence. Characters in between are
2037    treated as literals. This is slightly different from Perl in that $ and @ are
2038    also handled as literals inside the quotes. In Perl, they will cause variable
2039    interpolation. Note the following examples:
2040    
2041        Pattern            PCRE matches      Perl matches
2042    
2043        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2044        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2045        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2046    
2047    For compatibility with Perl, \Q...\E sequences are recognized inside character
2048    classes as well as outside them.
2049    
2050    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2051    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2052    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2053    signed/unsigned warnings.
2054    
2055    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2056    option for pcretest, so I've replaced it by a simple function that does just
2057    that job.
2058    
2059    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2060    "pcregrep -".
2061    
2062    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2063    Java package. This provides some syntactic sugar for simple cases of what my
2064    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2065    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2066    item, you can use this simplified notation. Note that only makes sense with
2067    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2068    greediness, whatever the setting of the PCRE_UNGREEDY option.
2069    
2070    13. A change of greediness default within a pattern was not taking effect at
2071    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2072    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2073    was abstracted outside.
2074    
2075    14. PCRE now supports the \G assertion. It is true when the current matching
2076    position is at the start point of the match. This differs from \A when the
2077    starting offset is non-zero. Used with the /g option of pcretest (or similar
2078    code), it works in the same way as it does for Perl's /g option. If all
2079    alternatives of a regex begin with \G, the expression is anchored to the start
2080    match position, and the "anchored" flag is set in the compiled expression.
2081    
2082    15. Some bugs concerning the handling of certain option changes within patterns
2083    have been fixed. These applied to options other than (?ims). For example,
2084    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2085    been the other way round. Some of this was related to change 7 above.
2086    
2087    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2088    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2089    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2090    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2091    
2092    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2093    mode. Unlike ".", it always matches newline, whatever the setting of
2094    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2095    assertions. Perl allows it, but it doesn't (in general) work because it can't
2096    calculate the length of the lookbehind. At least, that's the case for Perl
2097    5.8.0 - I've been told they are going to document that it doesn't work in
2098    future.
2099    
2100    18. Added an error diagnosis for escapes that PCRE does not support: these are
2101    \L, \l, \N, \P, \p, \U, \u, and \X.
2102    
2103    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2104    reading past the end of the pattern in cases such as /[abcd/.
2105    
2106    20. PCRE was getting more memory than necessary for patterns with classes that
2107    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2108    
2109    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2110    compiling PCRE for use with Virtual Pascal.
2111    
2112    22. Small fix to the Makefile to make it work properly if the build is done
2113    outside the source tree.
2114    
2115    23. Added a new extension: a condition to go with recursion. If a conditional
2116    subpattern starts with (?(R) the "true" branch is used if recursion has
2117    happened, whereas the "false" branch is used only at the top level.
2118    
2119    24. When there was a very long string of literal characters (over 255 bytes
2120    without UTF support, over 250 bytes with UTF support), the computation of how
2121    much memory was required could be incorrect, leading to segfaults or other
2122    strange effects.
2123    
2124    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2125    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2126    there was a subsequent back reference to those brackets. This meant that, for
2127    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2128    possible to check for precisely this case. All we can do is abandon the
2129    optimization if .* occurs inside capturing brackets when there are any back
2130    references whatsoever. (See below for a better fix that came later.)
2131    
2132    26. The handling of the optimization for finding the first character of a
2133    non-anchored pattern, and for finding a character that is required later in the
2134    match were failing in some cases. This didn't break the matching; it just
2135    failed to optimize when it could. The way this is done has been re-implemented.
2136    
2137    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2138    
2139    28. Added a new feature that provides some of the functionality that Perl
2140    provides with (?{...}). The facility is termed a "callout". The way it is done
2141    in PCRE is for the caller to provide an optional function, by setting
2142    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2143    global variable. By default it is unset, which disables all calling out. To get
2144    the function called, the regex must include (?C) at appropriate points. This
2145    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2146    This provides a means of identifying different callout points. When PCRE
2147    reaches such a point in the regex, if pcre_callout has been set, the external
2148    function is called. It is provided with data in a structure called
2149    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2150    matching continues; if it returns a non-zero value, the match at the current
2151    point fails. However, backtracking will occur if possible. [This was changed
2152    later and other features added - see item 49 below.]
2153    
2154    29. pcretest is upgraded to test the callout functionality. It provides a
2155    callout function that displays information. By default, it shows the start of
2156    the match and the current position in the text. There are some new data escapes
2157    to vary what happens:
2158    
2159        \C+         in addition, show current contents of captured substrings
2160        \C-         do not supply a callout function
2161        \C!n        return 1 when callout number n is reached
2162        \C!n!m      return 1 when callout number n is reached for the mth time
2163    
2164    30. If pcregrep was called with the -l option and just a single file name, it
2165    output "<stdin>" if a match was found, instead of the file name.
2166    
2167    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2168    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2169    pcre_exec(). This saves a malloc/free per call. The default value of
2170    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2171    when configuring.
2172    
2173    32. The default maximum size of a compiled pattern is 64K. There have been a
2174    few cases of people hitting this limit. The code now uses macros to handle the
2175    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2176    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2177    configuring. Tests 2 and 5 work only with 2-byte links because they output
2178    debugging information about compiled patterns.
2179    
2180    33. Internal code re-arrangements:
2181    
2182    (a) Moved the debugging function for printing out a compiled regex into
2183        its own source file (printint.c) and used #include to pull it into
2184        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2185        separate copies.
2186    
2187    (b) Defined the list of op-code names for debugging as a macro in
2188        internal.h so that it is next to the definition of the opcodes.
2189    
2190    (c) Defined a table of op-code lengths for simpler skipping along compiled
2191        code. This is again a macro in internal.h so that it is next to the
2192        definition of the opcodes.
2193    
2194    34. Added support for recursive calls to individual subpatterns, along the
2195    lines of Robin Houston's patch (but implemented somewhat differently).
2196    
2197    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2198    allow it to read and process whole directories in Win32. This code was
2199    contributed by Lionel Fourquaux; it has not been tested by me.
2200    
2201    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2202    used to name a group. Names consist of alphanumerics and underscores, and must
2203    be unique. Back references use the syntax (?P=name) and recursive calls use
2204    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2205    numbers. The function pcre_fullinfo() can be used after compilation to extract
2206    a name/number map. There are three relevant calls:
2207    
2208      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2209      PCRE_INFO_NAMECOUNT            yields the number of entries
2210      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2211    
2212    The map is a vector of fixed-size entries. The size of each entry depends on
2213    the length of the longest name used. The first two bytes of each entry are the
2214    group number, most significant byte first. There follows the corresponding
2215    name, zero terminated. The names are in alphabetical order.
2216    
2217    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2218    case instead of 255. Making it the same both with and without UTF-8 support
2219    means that the same test output works with both.
2220    
2221    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2222    calling malloc() with a zero argument.
2223    
2224    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2225    optimization. I've improved things by keeping a bitmap of backreferences with
2226    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2227    fact referenced, the optimization can be applied. It is unlikely that a
2228    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2229    the match to follow \n) will appear inside brackets with a number greater than
2230    31, but if it does, any back reference > 31 suppresses the optimization.
2231    
2232    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2233    of disabling numbered capturing parentheses. Any opening parenthesis that is
2234    not followed by ? behaves as if it were followed by ?: but named parentheses
2235    can still be used for capturing (and they will acquire numbers in the usual
2236    way).
2237    
2238    41. Redesigned the return codes from the match() function into yes/no/error so
2239    that errors can be passed back from deep inside the nested calls. A malloc
2240    failure while inside a recursive subpattern call now causes the
2241    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2242    
2243    42. It is now possible to set a limit on the number of times the match()
2244    function is called in a call to pcre_exec(). This facility makes it possible to
2245    limit the amount of recursion and backtracking, though not in a directly
2246    obvious way, because the match() function is used in a number of different
2247    circumstances. The count starts from zero for each position in the subject
2248    string (for non-anchored patterns). The default limit is, for compatibility, a
2249    large number, namely 10 000 000. You can change this in two ways:
2250    
2251    (a) When configuring PCRE before making, you can use --with-match-limit=n
2252        to set a default value for the compiled library.
2253    
2254    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2255        a different value is set. See 45 below.
2256    
2257    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2258    
2259    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2260    of things that can be changed at compile time. The first argument specifies
2261    what is wanted and the second points to where the information is to be placed.
2262    The current list of available information is:
2263    
2264      PCRE_CONFIG_UTF8
2265    
2266    The output is an integer that is set to one if UTF-8 support is available;
2267    otherwise it is set to zero.
2268    
2269      PCRE_CONFIG_NEWLINE
2270    
2271    The output is an integer that it set to the value of the code that is used for
2272    newline. It is either LF (10) or CR (13).
2273    
2274      PCRE_CONFIG_LINK_SIZE
2275    
2276    The output is an integer that contains the number of bytes used for internal
2277    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2278    
2279      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2280    
2281    The output is an integer that contains the threshold above which the POSIX
2282    interface uses malloc() for output vectors. See item 31 above.
2283    
2284      PCRE_CONFIG_MATCH_LIMIT
2285    
2286    The output is an unsigned integer that contains the default limit of the number
2287    of match() calls in a pcre_exec() execution. See 42 above.
2288    
2289    44. pcretest has been upgraded by the addition of the -C option. This causes it
2290    to extract all the available output from the new pcre_config() function, and to
2291    output it. The program then exits immediately.
2292    
2293    45. A need has arisen to pass over additional data with calls to pcre_exec() in
2294    order to support additional features. One way would have been to define
2295    pcre_exec2() (for example) with extra arguments, but this would not have been
2296    extensible, and would also have required all calls to the original function to
2297    be mapped to the new one. Instead, I have chosen to extend the mechanism that
2298    is used for passing in "extra" data from pcre_study().
2299    
2300    The pcre_extra structure is now exposed and defined in pcre.h. It currently
2301    contains the following fields:
2302    
2303      flags         a bitmap indicating which of the following fields are set
2304      study_data    opaque data from pcre_study()
2305      match_limit   a way of specifying a limit on match() calls for a specific
2306                      call to pcre_exec()
2307      callout_data  data for callouts (see 49 below)
2308    
2309    The flag bits are also defined in pcre.h, and are
2310    
2311      PCRE_EXTRA_STUDY_DATA
2312      PCRE_EXTRA_MATCH_LIMIT
2313      PCRE_EXTRA_CALLOUT_DATA
2314    
2315    The pcre_study() function now returns one of these new pcre_extra blocks, with
2316    the actual study data pointed to by the study_data field, and the
2317    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2318    before. That is, this change is entirely upwards-compatible and requires no
2319    change to existing code.
2320    
2321    If you want to pass in additional data to pcre_exec(), you can either place it
2322    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2323    block.
2324    
2325    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2326    data string contains the escape sequence \M, pcretest calls pcre_exec() several
2327    times with different match limits, until it finds the minimum value needed for
2328    pcre_exec() to complete. The value is then output. This can be instructive; for
2329    most simple matches the number is quite small, but for pathological cases it
2330    gets very large very quickly.
2331    
2332    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2333    returns the size of the data block pointed to by the study_data field in a
2334    pcre_extra block, that is, the value that was passed as the argument to
2335    pcre_malloc() when PCRE was getting memory in which to place the information
2336    created by pcre_study(). The fourth argument should point to a size_t variable.
2337    pcretest has been extended so that this information is shown after a successful
2338    pcre_study() call when information about the compiled regex is being displayed.
2339    
2340    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2341    because what follows is always an absolute path. (Later: it turns out that this
2342    is more than cosmetic for MinGW, because it doesn't like empty path
2343    components.)
2344    
2345    49. Some changes have been made to the callout feature (see 28 above):
2346    
2347    (i)  A callout function now has three choices for what it returns:
2348    
2349           0  =>  success, carry on matching
2350         > 0  =>  failure at this point, but backtrack if possible
2351         < 0  =>  serious error, return this value from pcre_exec()
2352    
2353         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2354         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2355         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2356         use by callout functions. It will never be used by PCRE itself.
2357    
2358    (ii) The pcre_extra structure (see 45 above) has a void * field called
2359         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2360         pcre_callout_block structure has a field of the same name. The contents of
2361         the field passed in the pcre_extra structure are passed to the callout
2362         function in the corresponding field in the callout block. This makes it
2363         easier to use the same callout-containing regex from multiple threads. For
2364         testing, the pcretest program has a new data escape
2365    
2366           \C*n        pass the number n (may be negative) as callout_data
2367    
2368         If the callout function in pcretest receives a non-zero value as
2369         callout_data, it returns that value.
2370    
2371    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2372    there were some redundant $(CFLAGS) in commands that are now specified as
2373    $(LINK), which already includes $(CFLAGS).
2374    
2375    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2376    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2377    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2378    one-byte characters throughout. Note that case-insensitive matching applies
2379    only to characters whose values are less than 256. PCRE doesn't support the
2380    notion of cases for higher-valued characters.
2381    
2382    (i)   A character class whose characters are all within 0-255 is handled as
2383          a bit map, and the map is inverted for negative classes. Previously, a
2384          character > 255 always failed to match such a class; however it should
2385          match if the class was a negative one (e.g. [^ab]). This has been fixed.
2386    
2387    (ii)  A negated character class with a single character < 255 is coded as
2388          "not this character" (OP_NOT). This wasn't working properly when the test
2389          character was multibyte, either singly or repeated.
2390    
2391    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2392          mode, for example: \x{100}{2,3}.
2393    
2394    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2395          singly or repeated) now correctly test multibyte characters. However,
2396          PCRE doesn't recognize any characters with values greater than 255 as
2397          digits, spaces, or word characters. Such characters always match \D, \S,
2398          and \W, and never match \d, \s, or \w.
2399    
2400    (v)   Classes may now contain characters and character ranges with values
2401          greater than 255. For example: [ab\x{100}-\x{400}].
2402    
2403    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
2404          PCRE in UTF-8 mode.
2405    
2406    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2407    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2408    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2409    value.)
2410    
2411    53. The single man page has become too large. I have therefore split it up into
2412    a number of separate man pages. These also give rise to individual HTML pages;
2413    these are now put in a separate directory, and there is an index.html page that
2414    lists them all. Some hyperlinking between the pages has been installed.
2415    
2416    54. Added convenience functions for handling named capturing parentheses.
2417    
2418    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2419    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2420    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2421    are faulted.
2422    
2423    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2424    calling configure. These values are used when compiling the dftables.c program
2425    which is run to generate the source of the default character tables. They
2426    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2427    you will need to set these values.
2428    
2429    57. Updated the building process for Windows DLL, as provided by Fred Cox.
2430    
2431    
2432    Version 3.9 02-Jan-02
2433    ---------------------
2434    
2435    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2436    
2437    2. If --disable-static was given, the building process failed when trying to
2438    build pcretest and pcregrep. (For some reason it was using libtool to compile
2439    them, which is not right, as they aren't part of the library.)
2440    
2441    
2442  Version 3.8 18-Dec-01  Version 3.8 18-Dec-01
2443  ---------------------  ---------------------
2444    

Legend:
Removed from v.59  
changed lines
  Added in v.307

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12