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

Diff of /code/trunk/ChangeLog

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

revision 7 by nigel, Sat Feb 24 21:38:09 2007 UTC revision 303 by ph10, Sun Jan 20 19:39:43 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.
16    
17    
18    Version 7.5 10-Jan-08
19    ---------------------
20    
21    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
22        values in parens when parsing an RE using the C++ wrapper."
23    
24    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
25        Characters greater than 255 were excluded from the class instead of being
26        included.
27    
28    3.  The same bug as (2) above applied to negated POSIX classes such as
29        [:^space:].
30    
31    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
32        defined or documented. It seems to have been a typo for PCRE_STATIC, so
33        I have changed it.
34    
35    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
36        first named subpattern) and a construct such as (?&a) would reference the
37        first named subpattern whose name started with "a" (in other words, the
38        length check was missing). Both these problems are fixed. "Subpattern name
39        expected" is now given for (?&) (a zero-length name), and this patch also
40        makes it give the same error for \k'' (previously it complained that that
41        was a reference to a non-existent subpattern).
42    
43    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
44        this is right because (?- can be followed by option settings as well as by
45        digits. I have, however, made the messages clearer.
46    
47    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
48        than the number used in the conditional) now cause a compile-time error.
49        This is actually not compatible with Perl, which accepts such patterns, but
50        treats the conditional as always being FALSE (as PCRE used to), but it
51        seems to me that giving a diagnostic is better.
52    
53    8.  Change "alphameric" to the more common word "alphanumeric" in comments
54        and messages.
55    
56    9.  Fix two occurrences of "backslash" in comments that should have been
57        "backspace".
58    
59    10. Remove two redundant lines of code that can never be obeyed (their function
60        was moved elsewhere).
61    
62    11. The program that makes PCRE's Unicode character property table had a bug
63        which caused it to generate incorrect table entries for sequences of
64        characters that have the same character type, but are in different scripts.
65        It amalgamated them into a single range, with the script of the first of
66        them. In other words, some characters were in the wrong script. There were
67        thirteen such cases, affecting characters in the following ranges:
68    
69          U+002b0 - U+002c1
70          U+0060c - U+0060d
71          U+0061e - U+00612
72          U+0064b - U+0065e
73          U+0074d - U+0076d
74          U+01800 - U+01805
75          U+01d00 - U+01d77
76          U+01d9b - U+01dbf
77          U+0200b - U+0200f
78          U+030fc - U+030fe
79          U+03260 - U+0327f
80          U+0fb46 - U+0fbb1
81          U+10450 - U+1049d
82    
83    12. The -o option (show only the matching part of a line) for pcregrep was not
84        compatible with GNU grep in that, if there was more than one match in a
85        line, it showed only the first of them. It now behaves in the same way as
86        GNU grep.
87    
88    13. If the -o and -v options were combined for pcregrep, it printed a blank
89        line for every non-matching line. GNU grep prints nothing, and pcregrep now
90        does the same. The return code can be used to tell if there were any
91        non-matching lines.
92    
93    14. Added --file-offsets and --line-offsets to pcregrep.
94    
95    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
96        infinitely looping recursion. The bug was that positive lookaheads were not
97        being skipped when checking for a possible empty match (negative lookaheads
98        and both kinds of lookbehind were skipped).
99    
100    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
101        inclusion of <windows.h> to before rather than after the definition of
102        INVALID_FILE_ATTRIBUTES (patch from David Byron).
103    
104    17. Specifying a possessive quantifier with a specific limit for a Unicode
105        character property caused pcre_compile() to compile bad code, which led at
106        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
107        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
108        caused the error; without that there was no problem.
109    
110    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
111    
112    19. Added --enable-pcretest-libreadline.
113    
114    20. In pcrecpp.cc, the variable 'count' was incremented twice in
115        RE::GlobalReplace(). As a result, the number of replacements returned was
116        double what it should be. I removed one of the increments, but Craig sent a
117        later patch that removed the other one (the right fix) and added unit tests
118        that check the return values (which was not done before).
119    
120    21. Several CMake things:
121    
122        (1) Arranged that, when cmake is used on Unix, the libraries end up with
123            the names libpcre and libpcreposix, not just pcre and pcreposix.
124    
125        (2) The above change means that pcretest and pcregrep are now correctly
126            linked with the newly-built libraries, not previously installed ones.
127    
128        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
129    
130    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
131        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
132        UTF-8 newline character). The key issue is that the pattern starts .*;
133        this means that the match must be either at the beginning, or after a
134        newline. The bug was in the code for advancing after a failed match and
135        checking that the new position followed a newline. It was not taking
136        account of UTF-8 characters correctly.
137    
138    23. PCRE was behaving differently from Perl in the way it recognized POSIX
139        character classes. PCRE was not treating the sequence [:...:] as a
140        character class unless the ... were all letters. Perl, however, seems to
141        allow any characters between [: and :], though of course it rejects as
142        unknown any "names" that contain non-letters, because all the known class
143        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
144        for example, whereas PCRE did not - it did not recognize a POSIX character
145        class. This seemed a bit dangerous, so the code has been changed to be
146        closer to Perl. The behaviour is not identical to Perl, because PCRE will
147        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
148        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
149        Perl does, and where it didn't before.
150    
151    24. Rewrite so as to remove the single use of %n from pcregrep because in some
152        Windows environments %n is disabled by default.
153    
154    
155    Version 7.4 21-Sep-07
156    ---------------------
157    
158    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
159        means that a class such as [\s] counted as "explicit reference to CR or
160        LF". That isn't really right - the whole point of the change was to try to
161        help when there was an actual mention of one of the two characters. So now
162        the change happens only if \r or \n (or a literal CR or LF) character is
163        encountered.
164    
165    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
166        of both had grown to the point where there were only 3 bits left.
167        Fortunately, there was spare space in the data structure, and so I have
168        moved the internal flags into a new 16-bit field to free up more option
169        bits.
170    
171    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
172        but did not set the internal JCHANGED flag - either of these is enough to
173        control the way the "get" function works - but the PCRE_INFO_JCHANGED
174        facility is supposed to tell if (?J) was ever used, so now (?J) at the
175        start sets both bits.
176    
177    4.  Added options (at build time, compile time, exec time) to change \R from
178        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
179    
180    5.  doc/pcresyntax.html was missing from the distribution.
181    
182    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
183        compatibility, even though it is no longer used.
184    
185    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
186        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
187        windows.h file is present (where different names are used). [This was
188        reversed later after testing - see 16 below.]
189    
190    8.  Changed all #include <config.h> to #include "config.h". There were also
191        some further <pcre.h> cases that I changed to "pcre.h".
192    
193    9.  When pcregrep was used with the --colour option, it missed the line ending
194        sequence off the lines that it output.
195    
196    10. It was pointed out to me that arrays of string pointers cause lots of
197        relocations when a shared library is dynamically loaded. A technique of
198        using a single long string with a table of offsets can drastically reduce
199        these. I have refactored PCRE in four places to do this. The result is
200        dramatic:
201    
202          Originally:                          290
203          After changing UCP table:            187
204          After changing error message table:   43
205          After changing table of "verbs"       36
206          After changing table of Posix names   22
207    
208        Thanks to the folks working on Gregex for glib for this insight.
209    
210    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
211        unicode-properties was also set.
212    
213    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
214    
215    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
216        checked only for CRLF.
217    
218    14. Added casts to pcretest.c to avoid compiler warnings.
219    
220    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
221    
222    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
223        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
224        entirely. This removes changes made in 7 above.
225    
226    17. The CMake files have been updated, and there is now more information about
227        building with CMake in the NON-UNIX-USE document.
228    
229    
230    Version 7.3 28-Aug-07
231    ---------------------
232    
233     1. In the rejigging of the build system that eventually resulted in 7.1, the
234        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
235        brackets there is not right, since it causes compilers to look for an
236        installed pcre.h, not the version that is in the source that is being
237        compiled (which of course may be different). I have changed it back to:
238    
239          #include "pcre.h"
240    
241        I have a vague recollection that the change was concerned with compiling in
242        different directories, but in the new build system, that is taken care of
243        by the VPATH setting the Makefile.
244    
245     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
246        when the subject happened to end in the byte 0x85 (e.g. if the last
247        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
248        characters but of course it shouldn't be taken as a newline when it is part
249        of another character. The bug was that, for an unlimited repeat of . in
250        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
251        characters when looking for a newline.
252    
253     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
254    
255     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
256        in debug output.
257    
258     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
259        long printing in the pcrecpp unittest when running under MinGW.
260    
261     6. ESC_K was left out of the EBCDIC table.
262    
263     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
264        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
265        limit also applies to "virtual nesting" when a pattern is recursive, and in
266        this case 1000 isn't so big. I have been able to remove this limit at the
267        expense of backing off one optimization in certain circumstances. Normally,
268        when pcre_exec() would call its internal match() function recursively and
269        immediately return the result unconditionally, it uses a "tail recursion"
270        feature to save stack. However, when a subpattern that can match an empty
271        string has an unlimited repetition quantifier, it no longer makes this
272        optimization. That gives it a stack frame in which to save the data for
273        checking that an empty string has been matched. Previously this was taken
274        from the 1000-entry workspace that had been reserved. So now there is no
275        explicit limit, but more stack is used.
276    
277     8. Applied Daniel's patches to solve problems with the import/export magic
278        syntax that is required for Windows, and which was going wrong for the
279        pcreposix and pcrecpp parts of the library. These were overlooked when this
280        problem was solved for the main library.
281    
282     9. There were some crude static tests to avoid integer overflow when computing
283        the size of patterns that contain repeated groups with explicit upper
284        limits. As the maximum quantifier is 65535, the maximum group length was
285        set at 30,000 so that the product of these two numbers did not overflow a
286        32-bit integer. However, it turns out that people want to use groups that
287        are longer than 30,000 bytes (though not repeat them that many times).
288        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
289        made it possible to implement the integer overflow checks in a much more
290        dynamic way, which I have now done. The artificial limitation on group
291        length has been removed - we now have only the limit on the total length of
292        the compiled pattern, which depends on the LINK_SIZE setting.
293    
294    10. Fixed a bug in the documentation for get/copy named substring when
295        duplicate names are permitted. If none of the named substrings are set, the
296        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
297        empty string.
298    
299    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
300        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
301        because the ] is interpreted as the first data character and the
302        terminating ] is not found. PCRE has been made compatible with Perl in this
303        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
304        cause memory overwriting.
305    
306    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
307        string has been matched (to stop an infinite loop). It was not recognizing
308        a conditional subpattern that could match an empty string if that
309        subpattern was within another subpattern. For example, it looped when
310        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
311        condition was not nested. This bug has been fixed.
312    
313    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
314        past the start of the subject in the presence of bytes with the top bit
315        set, for example "\x8aBCD".
316    
317    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
318        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
319    
320    14. Optimized (?!) to (*FAIL).
321    
322    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
323        This restricts code points to be within the range 0 to 0x10FFFF, excluding
324        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
325        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
326        does: it's just the validity check that is more restrictive.
327    
328    16. Inserted checks for integer overflows during escape sequence (backslash)
329        processing, and also fixed erroneous offset values for syntax errors during
330        backslash processing.
331    
332    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
333        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
334    
335    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
336        caused an overrun.
337    
338    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
339        something other than just ASCII characters) inside a group that had an
340        unlimited repeat caused a loop at compile time (while checking to see
341        whether the group could match an empty string).
342    
343    20. Debugging a pattern containing \p or \P could cause a crash. For example,
344        [\P{Any}] did so. (Error in the code for printing property names.)
345    
346    21. An orphan \E inside a character class could cause a crash.
347    
348    22. A repeated capturing bracket such as (A)? could cause a wild memory
349        reference during compilation.
350    
351    23. There are several functions in pcre_compile() that scan along a compiled
352        expression for various reasons (e.g. to see if it's fixed length for look
353        behind). There were bugs in these functions when a repeated \p or \P was
354        present in the pattern. These operators have additional parameters compared
355        with \d, etc, and these were not being taken into account when moving along
356        the compiled data. Specifically:
357    
358        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
359            length.
360    
361        (b) An item such as \pL+ within a repeated group could cause crashes or
362            loops.
363    
364        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
365            "reference to non-existent subpattern" error.
366    
367        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
368    
369    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
370        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
371    
372    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
373    
374    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
375        character were causing crashes (broken optimization).
376    
377    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
378        \p or \P) caused a compile-time loop.
379    
380    28. More problems have arisen in unanchored patterns when CRLF is a valid line
381        break. For example, the unstudied pattern [\r\n]A does not match the string
382        "\r\nA" because change 7.0/46 below moves the current point on by two
383        characters after failing to match at the start. However, the pattern \nA
384        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
385        the same is true. There doesn't seem any very clean way out of this, but
386        what I have chosen to do makes the common cases work: PCRE now takes note
387        of whether there can be an explicit match for \r or \n anywhere in the
388        pattern, and if so, 7.0/46 no longer applies. As part of this change,
389        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
390        pattern has explicit CR or LF references.
391    
392    29. Added (*CR) etc for changing newline setting at start of pattern.
393    
394    
395    Version 7.2 19-Jun-07
396    ---------------------
397    
398     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
399        which is apparently normally available under Windows.
400    
401     2. Re-jig the pcregrep tests with different newline settings in an attempt
402        to make them independent of the local environment's newline setting.
403    
404     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
405    
406     4. Some of the "internals" tests were previously cut out when the link size
407        was not 2, because the output contained actual offsets. The recent new
408        "Z" feature of pcretest means that these can be cut out, making the tests
409        usable with all link sizes.
410    
411     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
412        stack recursion. This gives a massive performance boost under BSD, but just
413        a small improvement under Linux. However, it saves one field in the frame
414        in all cases.
415    
416     6. Added more features from the forthcoming Perl 5.10:
417    
418        (a) (?-n) (where n is a string of digits) is a relative subroutine or
419            recursion call. It refers to the nth most recently opened parentheses.
420    
421        (b) (?+n) is also a relative subroutine call; it refers to the nth next
422            to be opened parentheses.
423    
424        (c) Conditions that refer to capturing parentheses can be specified
425            relatively, for example, (?(-2)... or (?(+3)...
426    
427        (d) \K resets the start of the current match so that everything before
428            is not part of it.
429    
430        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
431    
432        (f) \g{name} is another synonym - part of Perl 5.10's unification of
433            reference syntax.
434    
435        (g) (?| introduces a group in which the numbering of parentheses in each
436            alternative starts with the same number.
437    
438        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
439    
440     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
441        PCRE_INFO_JCHANGED.
442    
443     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
444        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
445        for detecting groups that can match an empty string.
446    
447     9. A pattern with a very large number of alternatives (more than several
448        hundred) was running out of internal workspace during the pre-compile
449        phase, where pcre_compile() figures out how much memory will be needed. A
450        bit of new cunning has reduced the workspace needed for groups with
451        alternatives. The 1000-alternative test pattern now uses 12 bytes of
452        workspace instead of running out of the 4096 that are available.
453    
454    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
455    
456    11. Applied patch from Google to remove an optimization that didn't quite work.
457        The report of the bug said:
458    
459          pcrecpp::RE("a*").FullMatch("aaa") matches, while
460          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
461          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
462    
463    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
464        it matched the wrong number of bytes.
465    
466    
467    Version 7.1 24-Apr-07
468    ---------------------
469    
470     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
471        that is more "standard", making use of automake and other Autotools. There
472        is some re-arrangement of the files and adjustment of comments consequent
473        on this.
474    
475     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
476        for recursive directory scanning broke on some systems because the files
477        are not scanned in any specific order and on different systems the order
478        was different. A call to "sort" has been inserted into RunGrepTest for the
479        approprate test as a short-term fix. In the longer term there may be an
480        alternative.
481    
482     3. I had an email from Eric Raymond about problems translating some of PCRE's
483        man pages to HTML (despite the fact that I distribute HTML pages, some
484        people do their own conversions for various reasons). The problems
485        concerned the use of low-level troff macros .br and .in. I have therefore
486        removed all such uses from the man pages (some were redundant, some could
487        be replaced by .nf/.fi pairs). The 132html script that I use to generate
488        HTML has been updated to handle .nf/.fi and to complain if it encounters
489        .br or .in.
490    
491     4. Updated comments in configure.ac that get placed in config.h.in and also
492        arranged for config.h to be included in the distribution, with the name
493        config.h.generic, for the benefit of those who have to compile without
494        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
495    
496     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
497        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
498        makevp.bat for the current PCRE, using the additional files
499        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
500    
501     6. A Windows user reported a minor discrepancy with test 2, which turned out
502        to be caused by a trailing space on an input line that had got lost in his
503        copy. The trailing space was an accident, so I've just removed it.
504    
505     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
506        that is needed.
507    
508     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
509        as "const" (a) because they are and (b) because it helps the PHP
510        maintainers who have recently made a script to detect big data structures
511        in the php code that should be moved to the .rodata section. I remembered
512        to update Builducptable as well, so it won't revert if ucptable.h is ever
513        re-created.
514    
515     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
516        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
517        order to be able to cut out the UTF-8 tables in the latter when UTF-8
518        support is not required. This saves 1.5-2K of code, which is important in
519        some applications.
520    
521        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
522        so as not to refer to the tables, even though these functions will never be
523        called when UTF-8 support is disabled. Otherwise there are problems with a
524        shared library.
525    
526    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
527    
528        (a) It was defining its arguments as char * instead of void *.
529    
530        (b) It was assuming that all moves were upwards in memory; this was true
531            a long time ago when I wrote it, but is no longer the case.
532    
533        The emulated memove() is provided for those environments that have neither
534        memmove() nor bcopy(). I didn't think anyone used it these days, but that
535        is clearly not the case, as these two bugs were recently reported.
536    
537    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
538        and Detrail to create the HTML documentation, the .txt form of the man
539        pages, and it removes trailing spaces from listed files. It also creates
540        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
541        case, it wraps all the #defines with #ifndefs. This script should be run
542        before "make dist".
543    
544    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
545        with Unicode property support.
546    
547        (a) For a maximizing quantifier, if the two different cases of the
548            character were of different lengths in their UTF-8 codings (there are
549            some cases like this - I found 11), and the matching function had to
550            back up over a mixture of the two cases, it incorrectly assumed they
551            were both the same length.
552    
553        (b) When PCRE was configured to use the heap rather than the stack for
554            recursion during matching, it was not correctly preserving the data for
555            the other case of a UTF-8 character when checking ahead for a match
556            while processing a minimizing repeat. If the check also involved
557            matching a wide character, but failed, corruption could cause an
558            erroneous result when trying to check for a repeat of the original
559            character.
560    
561    13. Some tidying changes to the testing mechanism:
562    
563        (a) The RunTest script now detects the internal link size and whether there
564            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
565            values substituted by "configure". (The RunGrepTest script already did
566            this for UTF-8.) The configure.ac script no longer substitutes the
567            relevant variables.
568    
569        (b) The debugging options /B and /D in pcretest show the compiled bytecode
570            with length and offset values. This means that the output is different
571            for different internal link sizes. Test 2 is skipped for link sizes
572            other than 2 because of this, bypassing the problem. Unfortunately,
573            there was also a test in test 3 (the locale tests) that used /B and
574            failed for link sizes other than 2. Rather than cut the whole test out,
575            I have added a new /Z option to pcretest that replaces the length and
576            offset values with spaces. This is now used to make test 3 independent
577            of link size. (Test 2 will be tidied up later.)
578    
579    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
580        segmentation fault instead of returning the appropriate error message.
581    
582    15. In multiline mode when the newline sequence was set to "any", the pattern
583        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
584        This doesn't seem right; it now treats the CRLF combination as the line
585        ending, and so does not match in that case. It's only a pattern such as ^$
586        that would hit this one: something like ^ABC$ would have failed after \r
587        and then tried again after \r\n.
588    
589    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
590        in an attempt to make files that differ only in their line terminators
591        compare equal. This works on Linux.
592    
593    17. Under certain error circumstances pcregrep might try to free random memory
594        as it exited. This is now fixed, thanks to valgrind.
595    
596    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
597        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
598        was because its rules for how to advance for /g after matching an empty
599        string at the end of a line did not allow for this case. They now check for
600        it specially.
601    
602    20. pcretest is supposed to handle patterns and data of any length, by
603        extending its buffers when necessary. It was getting this wrong when the
604        buffer for a data line had to be extended.
605    
606    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
607        CRLF as a newline sequence.
608    
609    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
610        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
611        I have nevertheless tidied it up.
612    
613    23. Added some casts to kill warnings from HP-UX ia64 compiler.
614    
615    24. Added a man page for pcre-config.
616    
617    
618    Version 7.0 19-Dec-06
619    ---------------------
620    
621     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
622        moving to gcc 4.1.1.
623    
624     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
625        sys/time.h, which is documented as needed for this function. It doesn't
626        seem to matter on Linux, but it showed up on some releases of OS X.
627    
628     3. It seems that there are systems where bytes whose values are greater than
629        127 match isprint() in the "C" locale. The "C" locale should be the
630        default when a C program starts up. In most systems, only ASCII printing
631        characters match isprint(). This difference caused the output from pcretest
632        to vary, making some of the tests fail. I have changed pcretest so that:
633    
634        (a) When it is outputting text in the compiled version of a pattern, bytes
635            other than 32-126 are always shown as hex escapes.
636    
637        (b) When it is outputting text that is a matched part of a subject string,
638            it does the same, unless a different locale has been set for the match
639            (using the /L modifier). In this case, it uses isprint() to decide.
640    
641     4. Fixed a major bug that caused incorrect computation of the amount of memory
642        required for a compiled pattern when options that changed within the
643        pattern affected the logic of the preliminary scan that determines the
644        length. The relevant options are -x, and -i in UTF-8 mode. The result was
645        that the computed length was too small. The symptoms of this bug were
646        either the PCRE error "internal error: code overflow" from pcre_compile(),
647        or a glibc crash with a message such as "pcretest: free(): invalid next
648        size (fast)". Examples of patterns that provoked this bug (shown in
649        pcretest format) are:
650    
651          /(?-x: )/x
652          /(?x)(?-x: \s*#\s*)/
653          /((?i)[\x{c0}])/8
654          /(?i:[\x{c0}])/8
655    
656        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
657        is now done differently.
658    
659     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
660        wrapper classes; (b) implement a new function in the C++ scanner that is
661        more efficient than the old way of doing things because it avoids levels of
662        recursion in the regex matching; (c) add a paragraph to the documentation
663        for the FullMatch() function.
664    
665     6. The escape sequence \n was being treated as whatever was defined as
666        "newline". Not only was this contrary to the documentation, which states
667        that \n is character 10 (hex 0A), but it also went horribly wrong when
668        "newline" was defined as CRLF. This has been fixed.
669    
670     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
671        was being set to -1 for the "end of line" case (supposedly a value that no
672        character can have). Though this value is never used (the check for end of
673        line is "zero bytes in current character"), it caused compiler complaints.
674        I've changed it to 0xffffffff.
675    
676     8. In pcre_version.c, the version string was being built by a sequence of
677        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
678        string (as it is for production releases) called a macro with an empty
679        argument. The C standard says the result of this is undefined. The gcc
680        compiler treats it as an empty string (which was what was wanted) but it is
681        reported that Visual C gives an error. The source has been hacked around to
682        avoid this problem.
683    
684     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
685        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
686        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
687        of them did).
688    
689    10. Originally, pcretest opened its input and output without "b"; then I was
690        told that "b" was needed in some environments, so it was added for release
691        5.0 to both the input and output. (It makes no difference on Unix-like
692        systems.) Later I was told that it is wrong for the input on Windows. I've
693        now abstracted the modes into two macros, to make it easier to fiddle with
694        them, and removed "b" from the input mode under Windows.
695    
696    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
697    
698    12. Added -help and --help to pcretest as an official way of being reminded
699        of the options.
700    
701    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
702        and pcrecpp.cc because they annoy compilers at high warning levels.
703    
704    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
705    
706    15. Fixed an occurrence of == in configure.ac that should have been = (shell
707        scripts are not C programs :-) and which was not noticed because it works
708        on Linux.
709    
710    16. pcretest is supposed to handle any length of pattern and data line (as one
711        line or as a continued sequence of lines) by extending its input buffer if
712        necessary. This feature was broken for very long pattern lines, leading to
713        a string of junk being passed to pcre_compile() if the pattern was longer
714        than about 50K.
715    
716    17. I have done a major re-factoring of the way pcre_compile() computes the
717        amount of memory needed for a compiled pattern. Previously, there was code
718        that made a preliminary scan of the pattern in order to do this. That was
719        OK when PCRE was new, but as the facilities have expanded, it has become
720        harder and harder to keep it in step with the real compile phase, and there
721        have been a number of bugs (see for example, 4 above). I have now found a
722        cunning way of running the real compile function in a "fake" mode that
723        enables it to compute how much memory it would need, while actually only
724        ever using a few hundred bytes of working memory and without too many
725        tests of the mode. This should make future maintenance and development
726        easier. A side effect of this work is that the limit of 200 on the nesting
727        depth of parentheses has been removed (though this was never a serious
728        limitation, I suspect). However, there is a downside: pcre_compile() now
729        runs more slowly than before (30% or more, depending on the pattern). I
730        hope this isn't a big issue. There is no effect on runtime performance.
731    
732    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
733        newline (only possible for the last line of a file) and it was a
734        pattern that set a locale (followed by /Lsomething), pcretest crashed.
735    
736    19. Added additional timing features to pcretest. (1) The -tm option now times
737        matching only, not compiling. (2) Both -t and -tm can be followed, as a
738        separate command line item, by a number that specifies the number of
739        repeats to use when timing. The default is 50000; this gives better
740        precision, but takes uncomfortably long for very large patterns.
741    
742    20. Extended pcre_study() to be more clever in cases where a branch of a
743        subpattern has no definite first character. For example, (a*|b*)[cd] would
744        previously give no result from pcre_study(). Now it recognizes that the
745        first character must be a, b, c, or d.
746    
747    21. There was an incorrect error "recursive call could loop indefinitely" if
748        a subpattern (or the entire pattern) that was being tested for matching an
749        empty string contained only one non-empty item after a nested subpattern.
750        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
751        incorrectly, because the \d was being skipped in the check.
752    
753    22. The pcretest program now has a new pattern option /B and a command line
754        option -b, which is equivalent to adding /B to every pattern. This causes
755        it to show the compiled bytecode, without the additional information that
756        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
757        is the same as /B/I).
758    
759    23. A new optimization is now able automatically to treat some sequences such
760        as a*b as a*+b. More specifically, if something simple (such as a character
761        or a simple class like \d) has an unlimited quantifier, and is followed by
762        something that cannot possibly match the quantified thing, the quantifier
763        is automatically "possessified".
764    
765    24. A recursive reference to a subpattern whose number was greater than 39
766        went wrong under certain circumstances in UTF-8 mode. This bug could also
767        have affected the operation of pcre_study().
768    
769    25. Realized that a little bit of performance could be had by replacing
770        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
771    
772    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
773    
774    27. Possessive quantifiers such as a++ were previously implemented by turning
775        them into atomic groups such as ($>a+). Now they have their own opcodes,
776        which improves performance. This includes the automatically created ones
777        from 23 above.
778    
779    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
780        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
781        the first matched character to be a colon. This applied both to named and
782        numbered groups.
783    
784    29. The ucpinternal.h header file was missing its idempotency #ifdef.
785    
786    30. I was sent a "project" file called libpcre.a.dev which I understand makes
787        building PCRE on Windows easier, so I have included it in the distribution.
788    
789    31. There is now a check in pcretest against a ridiculously large number being
790        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
791        loop, the loop is abandoned.
792    
793    32. Forward references to subpatterns in conditions such as (?(2)...) where
794        subpattern 2 is defined later cause pcre_compile() to search forwards in
795        the pattern for the relevant set of parentheses. This search went wrong
796        when there were unescaped parentheses in a character class, parentheses
797        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
798    
799    33. "Subroutine" calls and backreferences were previously restricted to
800        referencing subpatterns earlier in the regex. This restriction has now
801        been removed.
802    
803    34. Added a number of extra features that are going to be in Perl 5.10. On the
804        whole, these are just syntactic alternatives for features that PCRE had
805        previously implemented using the Python syntax or my own invention. The
806        other formats are all retained for compatibility.
807    
808        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
809            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
810            also .NET compatible.
811    
812        (b) A recursion or subroutine call to a named group can now be defined as
813            (?&name) as well as (?P>name).
814    
815        (c) A backreference to a named group can now be defined as \k<name> or
816            \k'name' as well as (?P=name). The new forms, as well as being in Perl
817            5.10, are also .NET compatible.
818    
819        (d) A conditional reference to a named group can now use the syntax
820            (?(<name>) or (?('name') as well as (?(name).
821    
822        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
823            groups (named and numbered) that are never evaluated inline, but can be
824            called as "subroutines" from elsewhere. In effect, the DEFINE condition
825            is always false. There may be only one alternative in such a group.
826    
827        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
828            as the simple (?(R). The condition is true only if the most recent
829            recursion is that of the given number or name. It does not search out
830            through the entire recursion stack.
831    
832        (g) The escape \gN or \g{N} has been added, where N is a positive or
833            negative number, specifying an absolute or relative reference.
834    
835    35. Tidied to get rid of some further signed/unsigned compiler warnings and
836        some "unreachable code" warnings.
837    
838    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
839        things, this adds five new scripts.
840    
841    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
842        There were also incompatibilities regarding the handling of \Q..\E inside
843        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
844        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
845    
846    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
847        matches an empty string, and forcibly breaks the loop. There were bugs in
848        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
849        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
850        separate and independent bugs (that affected different cases) have been
851        fixed.
852    
853    39. Refactored the code to abolish the use of different opcodes for small
854        capturing bracket numbers. This is a tidy that I avoided doing when I
855        removed the limit on the number of capturing brackets for 3.5 back in 2001.
856        The new approach is not only tidier, it makes it possible to reduce the
857        memory needed to fix the previous bug (38).
858    
859    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
860        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
861        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
862        mode.
863    
864    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
865        report.
866    
867    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
868        copy construction and assignment in the C++ wrapper.
869    
870    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
871        couple of bugs that could have given wrong results in the "--newline=crlf"
872        case.
873    
874    44. Added a number of casts and did some reorganization of signed/unsigned int
875        variables following suggestions from Dair Grant. Also renamed the variable
876        "this" as "item" because it is a C++ keyword.
877    
878    45. Arranged for dftables to add
879    
880          #include "pcre_internal.h"
881    
882        to pcre_chartables.c because without it, gcc 4.x may remove the array
883        definition from the final binary if PCRE is built into a static library and
884        dead code stripping is activated.
885    
886    46. For an unanchored pattern, if a match attempt fails at the start of a
887        newline sequence, and the newline setting is CRLF or ANY, and the next two
888        characters are CRLF, advance by two characters instead of one.
889    
890    
891    Version 6.7 04-Jul-06
892    ---------------------
893    
894     1. In order to handle tests when input lines are enormously long, pcretest has
895        been re-factored so that it automatically extends its buffers when
896        necessary. The code is crude, but this _is_ just a test program. The
897        default size has been increased from 32K to 50K.
898    
899     2. The code in pcre_study() was using the value of the re argument before
900        testing it for NULL. (Of course, in any sensible call of the function, it
901        won't be NULL.)
902    
903     3. The memmove() emulation function in pcre_internal.h, which is used on
904        systems that lack both memmove() and bcopy() - that is, hardly ever -
905        was missing a "static" storage class specifier.
906    
907     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
908        containing an extended class (one that cannot be represented by a bitmap
909        because it contains high-valued characters or Unicode property items, e.g.
910        [\pZ]). Almost always one would set UTF-8 mode when processing such a
911        pattern, but PCRE should not loop if you do not (it no longer does).
912        [Detail: two cases were found: (a) a repeated subpattern containing an
913        extended class; (b) a recursive reference to a subpattern that followed a
914        previous extended class. It wasn't skipping over the extended class
915        correctly when UTF-8 mode was not set.]
916    
917     5. A negated single-character class was not being recognized as fixed-length
918        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
919        compile error "lookbehind assertion is not fixed length".
920    
921     6. The RunPerlTest auxiliary script was showing an unexpected difference
922        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
923        write a Perl script that can interpret lines of an input file either as
924        byte characters or as UTF-8, which is what "perltest" was being required to
925        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
926        can't do is switch easily at run time between having the "use utf8;" pragma
927        or not. In the end, I fudged it by using the RunPerlTest script to insert
928        "use utf8;" explicitly for the UTF-8 tests.
929    
930     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
931        the end of the subject string, contrary to the documentation and to what
932        Perl does. This was true of both matching functions. Now it matches only at
933        the start of the subject and immediately after *internal* newlines.
934    
935     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
936        a pointer to an int instead of a pointer to an unsigned long int. This
937        caused problems on 64-bit systems.
938    
939     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
940        instance of the 'standard' template library not being so standard".
941    
942    10. There was no check on the number of named subpatterns nor the maximum
943        length of a subpattern name. The product of these values is used to compute
944        the size of the memory block for a compiled pattern. By supplying a very
945        long subpattern name and a large number of named subpatterns, the size
946        computation could be caused to overflow. This is now prevented by limiting
947        the length of names to 32 characters, and the number of named subpatterns
948        to 10,000.
949    
950    11. Subpatterns that are repeated with specific counts have to be replicated in
951        the compiled pattern. The size of memory for this was computed from the
952        length of the subpattern and the repeat count. The latter is limited to
953        65535, but there was no limit on the former, meaning that integer overflow
954        could in principle occur. The compiled length of a repeated subpattern is
955        now limited to 30,000 bytes in order to prevent this.
956    
957    12. Added the optional facility to have named substrings with the same name.
958    
959    13. Added the ability to use a named substring as a condition, using the
960        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
961        are numbers (not recommended). Forward references are permitted.
962    
963    14. Added forward references in named backreferences (if you see what I mean).
964    
965    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
966        pattern could run off the end of the subject. For example, the pattern
967        "(?s)(.{1,5})"8 did this with the subject "ab".
968    
969    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
970        PCRE_CASELESS was set when matching characters that were quantified with ?
971        or *.
972    
973    17. A character class other than a single negated character that had a minimum
974        but no maximum quantifier - for example [ab]{6,} - was not handled
975        correctly by pce_dfa_exec(). It would match only one character.
976    
977    18. A valid (though odd) pattern that looked like a POSIX character
978        class but used an invalid character after [ (for example [[,abc,]]) caused
979        pcre_compile() to give the error "Failed: internal error: code overflow" or
980        in some cases to crash with a glibc free() error. This could even happen if
981        the pattern terminated after [[ but there just happened to be a sequence of
982        letters, a binary zero, and a closing ] in the memory that followed.
983    
984    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
985        over the years. Originally (before any Unicode support), just the bottom 8
986        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
987        output from "man perlunicode" includes this:
988    
989          The regular expression compiler produces polymorphic opcodes.  That
990          is, the pattern adapts to the data and automatically switches to
991          the Unicode character scheme when presented with Unicode data--or
992          instead uses a traditional byte scheme when presented with byte
993          data.
994    
995        Sadly, a wide octal escape does not cause a switch, and in a string with
996        no other multibyte characters, these octal escapes are treated as before.
997        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
998        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
999        Unicode string.
1000    
1001        I have not perpetrated such confusion in PCRE. Up till now, it took just
1002        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1003        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1004        translate to the appropriate multibyte character.
1005    
1006    29. Applied some refactoring to reduce the number of warnings from Microsoft
1007        and Borland compilers. This has included removing the fudge introduced
1008        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1009        a warning about an unused variable.
1010    
1011    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1012        characters since release 4.0, because Perl (from release 5.004) does not.
1013        [Or at least, is documented not to: some releases seem to be in conflict
1014        with the documentation.] However, when a pattern was studied with
1015        pcre_study() and all its branches started with \s, PCRE still included VT
1016        as a possible starting character. Of course, this did no harm; it just
1017        caused an unnecessary match attempt.
1018    
1019    22. Removed a now-redundant internal flag bit that recorded the fact that case
1020        dependency changed within the pattern. This was once needed for "required
1021        byte" processing, but is no longer used. This recovers a now-scarce options
1022        bit. Also moved the least significant internal flag bit to the most-
1023        significant bit of the word, which was not previously used (hangover from
1024        the days when it was an int rather than a uint) to free up another bit for
1025        the future.
1026    
1027    23. Added support for CRLF line endings as well as CR and LF. As well as the
1028        default being selectable at build time, it can now be changed at runtime
1029        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1030        specify that it is scanning data with non-default line endings.
1031    
1032    24. Changed the definition of CXXLINK to make it agree with the definition of
1033        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1034    
1035    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1036        recursions. This makes a big different to stack usage for some patterns.
1037    
1038    26. If a subpattern containing a named recursion or subroutine reference such
1039        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1040        the space required for the compiled pattern went wrong and gave too small a
1041        value. Depending on the environment, this could lead to "Failed: internal
1042        error: code overflow at offset 49" or "glibc detected double free or
1043        corruption" errors.
1044    
1045    27. Applied patches from Google (a) to support the new newline modes and (b) to
1046        advance over multibyte UTF-8 characters in GlobalReplace.
1047    
1048    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1049        difference for some implementation of PCRE in some Windows version.
1050    
1051    29. Added some extra testing facilities to pcretest:
1052    
1053        \q<number>   in a data line sets the "match limit" value
1054        \Q<number>   in a data line sets the "match recursion limt" value
1055        -S <number>  sets the stack size, where <number> is in megabytes
1056    
1057        The -S option isn't available for Windows.
1058    
1059    
1060    Version 6.6 06-Feb-06
1061    ---------------------
1062    
1063     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1064        in pcreposix.h. I have copied the definition from pcre.h.
1065    
1066     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1067        because pcre.h is no longer a built file.
1068    
1069     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1070        not normally included in the compiled code.
1071    
1072    
1073    Version 6.5 01-Feb-06
1074    ---------------------
1075    
1076     1. When using the partial match feature with pcre_dfa_exec(), it was not
1077        anchoring the second and subsequent partial matches at the new starting
1078        point. This could lead to incorrect results. For example, with the pattern
1079        /1234/, partially matching against "123" and then "a4" gave a match.
1080    
1081     2. Changes to pcregrep:
1082    
1083        (a) All non-match returns from pcre_exec() were being treated as failures
1084            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1085            error message is output. Some extra information is given for the
1086            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1087            probably the only errors that are likely to be caused by users (by
1088            specifying a regex that has nested indefinite repeats, for instance).
1089            If there are more than 20 of these errors, pcregrep is abandoned.
1090    
1091        (b) A binary zero was treated as data while matching, but terminated the
1092            output line if it was written out. This has been fixed: binary zeroes
1093            are now no different to any other data bytes.
1094    
1095        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1096            used to set a locale for matching. The --locale=xxxx long option has
1097            been added (no short equivalent) to specify a locale explicitly on the
1098            pcregrep command, overriding the environment variables.
1099    
1100        (d) When -B was used with -n, some line numbers in the output were one less
1101            than they should have been.
1102    
1103        (e) Added the -o (--only-matching) option.
1104    
1105        (f) If -A or -C was used with -c (count only), some lines of context were
1106            accidentally printed for the final match.
1107    
1108        (g) Added the -H (--with-filename) option.
1109    
1110        (h) The combination of options -rh failed to suppress file names for files
1111            that were found from directory arguments.
1112    
1113        (i) Added the -D (--devices) and -d (--directories) options.
1114    
1115        (j) Added the -F (--fixed-strings) option.
1116    
1117        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1118    
1119        (l) Added the --colo(u)r option.
1120    
1121        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1122            is not present by default.
1123    
1124     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1125        items such as (?R) or (?1), when the recursion could match a number of
1126        alternatives. If it matched one of the alternatives, but subsequently,
1127        outside the recursion, there was a failure, the code tried to back up into
1128        the recursion. However, because of the way PCRE is implemented, this is not
1129        possible, and the result was an incorrect result from the match.
1130    
1131        In order to prevent this happening, the specification of recursion has
1132        been changed so that all such subpatterns are automatically treated as
1133        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1134    
1135     4. I had overlooked the fact that, in some locales, there are characters for
1136        which isalpha() is true but neither isupper() nor islower() are true. In
1137        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1138        and ordfeminine) are like this. This affected the treatment of \w and \W
1139        when they appeared in character classes, but not when they appeared outside
1140        a character class. The bit map for "word" characters is now created
1141        separately from the results of isalnum() instead of just taking it from the
1142        upper, lower, and digit maps. (Plus the underscore character, of course.)
1143    
1144     5. The above bug also affected the handling of POSIX character classes such as
1145        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1146        permanent tables. Instead, the bit maps for such a class were previously
1147        created as the appropriate unions of the upper, lower, and digit bitmaps.
1148        Now they are created by subtraction from the [[:word:]] class, which has
1149        its own bitmap.
1150    
1151     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1152        It is created by subtracting the vertical space characters (\x09, \x0a,
1153        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1154        subtraction was done in the overall bitmap for a character class, meaning
1155        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1156        be recognized. This bug has been fixed.
1157    
1158     7. Patches from the folks at Google:
1159    
1160          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1161          real life, but is still worth protecting against".
1162    
1163          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1164          regular expressions".
1165    
1166          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1167          have it.
1168    
1169          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1170          "configure" and the latter not, in order to fix a problem somebody had
1171          with compiling the Arg class on HP-UX.
1172    
1173          (e) Improve the error-handling of the C++ wrapper a little bit.
1174    
1175          (f) New tests for checking recursion limiting.
1176    
1177     8. The pcre_memmove() function, which is used only if the environment does not
1178        have a standard memmove() function (and is therefore rarely compiled),
1179        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1180        returning a result (though PCRE never actually uses the result).
1181    
1182     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1183        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1184        returned instead of calling malloc() with an overflowing number that would
1185        most likely cause subsequent chaos.
1186    
1187    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1188    
1189    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1190        with this option is matched, the nmatch and pmatch options of regexec() are
1191        ignored.
1192    
1193    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1194        provided in case anyone wants to the the POSIX interface with UTF-8
1195        strings.
1196    
1197    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1198        C++ linking (needed for some HP-UX environments).
1199    
1200    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1201        (unused parameter) and in the pcre_printint() function (omitted "default"
1202        switch label when the default is to do nothing).
1203    
1204    15. Added some code to make it possible, when PCRE is compiled as a C++
1205        library, to replace subject pointers for pcre_exec() with a smart pointer
1206        class, thus making it possible to process discontinuous strings.
1207    
1208    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1209        much the same function. They were added by different people who were trying
1210        to make PCRE easy to compile on non-Unix systems. It has been suggested
1211        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1212        for compiling on Windows systems. I have therefore replaced it with
1213        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1214        defaults to "extern" for C or "extern C" for C++, which works fine on
1215        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1216        SCOPE with something explicit in config.h. In addition:
1217    
1218        (a) pcreposix.h still had just "extern" instead of either of these macros;
1219            I have replaced it with PCRE_DATA_SCOPE.
1220    
1221        (b) Functions such as _pcre_xclass(), which are internal to the library,
1222            but external in the C sense, all had PCRE_EXPORT in their definitions.
1223            This is apparently wrong for the Windows case, so I have removed it.
1224            (It makes no difference on Unix-like systems.)
1225    
1226    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1227        of recursive calls to match(). This is different to MATCH_LIMIT because
1228        that limits the total number of calls to match(), not all of which increase
1229        the depth of recursion. Limiting the recursion depth limits the amount of
1230        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1231        when PCRE is compiled, and changed at run time. A patch from Google adds
1232        this functionality to the C++ interface.
1233    
1234    18. Changes to the handling of Unicode character properties:
1235    
1236        (a) Updated the table to Unicode 4.1.0.
1237    
1238        (b) Recognize characters that are not in the table as "Cn" (undefined).
1239    
1240        (c) I revised the way the table is implemented to a much improved format
1241            which includes recognition of ranges. It now supports the ranges that
1242            are defined in UnicodeData.txt, and it also amalgamates other
1243            characters into ranges. This has reduced the number of entries in the
1244            table from around 16,000 to around 3,000, thus reducing its size
1245            considerably. I realized I did not need to use a tree structure after
1246            all - a binary chop search is just as efficient. Having reduced the
1247            number of entries, I extended their size from 6 bytes to 8 bytes to
1248            allow for more data.
1249    
1250        (d) Added support for Unicode script names via properties such as \p{Han}.
1251    
1252    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1253        matching that character.
1254    
1255    20. When matching a repeated Unicode property with a minimum greater than zero,
1256        (for example \pL{2,}), PCRE could look past the end of the subject if it
1257        reached it while seeking the minimum number of characters. This could
1258        happen only if some of the characters were more than one byte long, because
1259        there is a check for at least the minimum number of bytes.
1260    
1261    21. Refactored the implementation of \p and \P so as to be more general, to
1262        allow for more different types of property in future. This has changed the
1263        compiled form incompatibly. Anybody with saved compiled patterns that use
1264        \p or \P will have to recompile them.
1265    
1266    22. Added "Any" and "L&" to the supported property types.
1267    
1268    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1269        but give a compile time error if the value is greater than 0xff.
1270    
1271    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1272        accidentally not being installed or uninstalled.
1273    
1274    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1275        made were to insert the current release number. This seemed silly, because
1276        it made things harder for people building PCRE on systems that don't run
1277        "configure". I have turned pcre.h into a distributed file, no longer built
1278        by "configure", with the version identification directly included. There is
1279        no longer a pcre.h.in file.
1280    
1281        However, this change necessitated a change to the pcre-config script as
1282        well. It is built from pcre-config.in, and one of the substitutions was the
1283        release number. I have updated configure.ac so that ./configure now finds
1284        the release number by grepping pcre.h.
1285    
1286    26. Added the ability to run the tests under valgrind.
1287    
1288    
1289    Version 6.4 05-Sep-05
1290    ---------------------
1291    
1292     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1293        "--" to be printed when multiple files were scanned, even when none of the
1294        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1295        consider it to be a bug, and have restored the previous behaviour.
1296    
1297     2. A couple of code tidies to get rid of compiler warnings.
1298    
1299     3. The pcretest program used to cheat by referring to symbols in the library
1300        whose names begin with _pcre_. These are internal symbols that are not
1301        really supposed to be visible externally, and in some environments it is
1302        possible to suppress them. The cheating is now confined to including
1303        certain files from the library's source, which is a bit cleaner.
1304    
1305     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1306        file's purpose clearer.
1307    
1308     5. Reorganized pcre_ucp_findchar().
1309    
1310    
1311    Version 6.3 15-Aug-05
1312    ---------------------
1313    
1314     1. The file libpcre.pc.in did not have general read permission in the tarball.
1315    
1316     2. There were some problems when building without C++ support:
1317    
1318        (a) If C++ support was not built, "make install" and "make test" still
1319            tried to test it.
1320    
1321        (b) There were problems when the value of CXX was explicitly set. Some
1322            changes have been made to try to fix these, and ...
1323    
1324        (c) --disable-cpp can now be used to explicitly disable C++ support.
1325    
1326        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1327            backslash in a target when C++ was disabled. This confuses some
1328            versions of "make", apparently. Using an intermediate variable solves
1329            this. (Same for CPP_LOBJ.)
1330    
1331     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1332        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1333        necessary on certain architectures.
1334    
1335     4. Added a setting of -export-symbols-regex to the link command to remove
1336        those symbols that are exported in the C sense, but actually are local
1337        within the library, and not documented. Their names all begin with
1338        "_pcre_". This is not a perfect job, because (a) we have to except some
1339        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1340        available (and never for static libraries). I have made a note to try to
1341        find a way round (a) in the future.
1342    
1343    
1344    Version 6.2 01-Aug-05
1345    ---------------------
1346    
1347     1. There was no test for integer overflow of quantifier values. A construction
1348        such as {1111111111111111} would give undefined results. What is worse, if
1349        a minimum quantifier for a parenthesized subpattern overflowed and became
1350        negative, the calculation of the memory size went wrong. This could have
1351        led to memory overwriting.
1352    
1353     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1354    
1355     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1356        operating environments where this matters.
1357    
1358     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1359        PCRE options from within the C++ wrapper.
1360    
1361     5. Named capturing subpatterns were not being correctly counted when a pattern
1362        was compiled. This caused two problems: (a) If there were more than 100
1363        such subpatterns, the calculation of the memory needed for the whole
1364        compiled pattern went wrong, leading to an overflow error. (b) Numerical
1365        back references of the form \12, where the number was greater than 9, were
1366        not recognized as back references, even though there were sufficient
1367        previous subpatterns.
1368    
1369     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1370        versions of gcc, e.g. 2.95.4.
1371    
1372    
1373    Version 6.1 21-Jun-05
1374    ---------------------
1375    
1376     1. There was one reference to the variable "posix" in pcretest.c that was not
1377        surrounded by "#if !defined NOPOSIX".
1378    
1379     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1380        the cross-check on the old pcre_info() function, for the benefit of the
1381        cut-down version of PCRE that is currently imported into Exim.
1382    
1383     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1384        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1385        patterns that start (?i) but I don't think that matters. The use of (?i) is
1386        just an example; this all applies to the other options as well.
1387    
1388     4. Since libtool seems to echo the compile commands it is issuing, the output
1389        from "make" can be reduced a bit by putting "@" in front of each libtool
1390        compile command.
1391    
1392     5. Patch from the folks at Google for configure.in to be a bit more thorough
1393        in checking for a suitable C++ installation before trying to compile the
1394        C++ stuff. This should fix a reported problem when a compiler was present,
1395        but no suitable headers.
1396    
1397     6. The man pages all had just "PCRE" as their title. I have changed them to
1398        be the relevant file name. I have also arranged that these names are
1399        retained in the file doc/pcre.txt, which is a concatenation in text format
1400        of all the man pages except the little individual ones for each function.
1401    
1402     7. The NON-UNIX-USE file had not been updated for the different set of source
1403        files that come with release 6. I also added a few comments about the C++
1404        wrapper.
1405    
1406    
1407    Version 6.0 07-Jun-05
1408    ---------------------
1409    
1410     1. Some minor internal re-organization to help with my DFA experiments.
1411    
1412     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1413        didn't matter for the library itself when fully configured, but did matter
1414        when compiling without UCP support, or within Exim, where the ucp files are
1415        not imported.
1416    
1417     3. Refactoring of the library code to split up the various functions into
1418        different source modules. The addition of the new DFA matching code (see
1419        below) to a single monolithic source would have made it really too
1420        unwieldy, quite apart from causing all the code to be include in a
1421        statically linked application, when only some functions are used. This is
1422        relevant even without the DFA addition now that patterns can be compiled in
1423        one application and matched in another.
1424    
1425        The downside of splitting up is that there have to be some external
1426        functions and data tables that are used internally in different modules of
1427        the library but which are not part of the API. These have all had their
1428        names changed to start with "_pcre_" so that they are unlikely to clash
1429        with other external names.
1430    
1431     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1432        a different (DFA) algorithm. Although it is slower than the original
1433        function, it does have some advantages for certain types of matching
1434        problem.
1435    
1436     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1437        including restarting after a partial match.
1438    
1439     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1440        defined when compiling for Windows was sent to me. I have put it into the
1441        code, though I have no means of testing or verifying it.
1442    
1443     7. Added the pcre_refcount() auxiliary function.
1444    
1445     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1446        match before or at the first newline in the subject string. In pcretest,
1447        the /f option on a pattern can be used to set this.
1448    
1449     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1450        would behave wrongly. This has been present in PCRE since release 4.0.
1451    
1452    10. A number of changes to the pcregrep command:
1453    
1454        (a) Refactored how -x works; insert ^(...)$ instead of setting
1455            PCRE_ANCHORED and checking the length, in preparation for adding
1456            something similar for -w.
1457    
1458        (b) Added the -w (match as a word) option.
1459    
1460        (c) Refactored the way lines are read and buffered so as to have more
1461            than one at a time available.
1462    
1463        (d) Implemented a pcregrep test script.
1464    
1465        (e) Added the -M (multiline match) option. This allows patterns to match
1466            over several lines of the subject. The buffering ensures that at least
1467            8K, or the rest of the document (whichever is the shorter) is available
1468            for matching (and similarly the previous 8K for lookbehind assertions).
1469    
1470        (f) Changed the --help output so that it now says
1471    
1472              -w, --word-regex(p)
1473    
1474            instead of two lines, one with "regex" and the other with "regexp"
1475            because that confused at least one person since the short forms are the
1476            same. (This required a bit of code, as the output is generated
1477            automatically from a table. It wasn't just a text change.)
1478    
1479        (g) -- can be used to terminate pcregrep options if the next thing isn't an
1480            option but starts with a hyphen. Could be a pattern or a path name
1481            starting with a hyphen, for instance.
1482    
1483        (h) "-" can be given as a file name to represent stdin.
1484    
1485        (i) When file names are being printed, "(standard input)" is used for
1486            the standard input, for compatibility with GNU grep. Previously
1487            "<stdin>" was used.
1488    
1489        (j) The option --label=xxx can be used to supply a name to be used for
1490            stdin when file names are being printed. There is no short form.
1491    
1492        (k) Re-factored the options decoding logic because we are going to add
1493            two more options that take data. Such options can now be given in four
1494            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1495    
1496        (l) Added the -A, -B, and -C options for requesting that lines of context
1497            around matches be printed.
1498    
1499        (m) Added the -L option to print the names of files that do not contain
1500            any matching lines, that is, the complement of -l.
1501    
1502        (n) The return code is 2 if any file cannot be opened, but pcregrep does
1503            continue to scan other files.
1504    
1505        (o) The -s option was incorrectly implemented. For compatibility with other
1506            greps, it now suppresses the error message for a non-existent or non-
1507            accessible file (but not the return code). There is a new option called
1508            -q that suppresses the output of matching lines, which was what -s was
1509            previously doing.
1510    
1511        (p) Added --include and --exclude options to specify files for inclusion
1512            and exclusion when recursing.
1513    
1514    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1515        Hopefully, it now does.
1516    
1517    12. Missing cast in pcre_study().
1518    
1519    13. Added an "uninstall" target to the makefile.
1520    
1521    14. Replaced "extern" in the function prototypes in Makefile.in with
1522        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1523        world, but is set differently for Windows.
1524    
1525    15. Added a second compiling function called pcre_compile2(). The only
1526        difference is that it has an extra argument, which is a pointer to an
1527        integer error code. When there is a compile-time failure, this is set
1528        non-zero, in addition to the error test pointer being set to point to an
1529        error message. The new argument may be NULL if no error number is required
1530        (but then you may as well call pcre_compile(), which is now just a
1531        wrapper). This facility is provided because some applications need a
1532        numeric error indication, but it has also enabled me to tidy up the way
1533        compile-time errors are handled in the POSIX wrapper.
1534    
1535    16. Added VPATH=.libs to the makefile; this should help when building with one
1536        prefix path and installing with another. (Or so I'm told by someone who
1537        knows more about this stuff than I do.)
1538    
1539    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1540        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1541        match everything, including newlines. This is not POSIX-compatible, but
1542        somebody wanted the feature. From pcretest it can be activated by using
1543        both the P and the s flags.
1544    
1545    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1546    
1547    19. libpcre.pc was being incorrectly installed as executable.
1548    
1549    20. A couple of places in pcretest check for end-of-line by looking for '\n';
1550        it now also looks for '\r' so that it will work unmodified on Windows.
1551    
1552    21. Added Google's contributed C++ wrapper to the distribution.
1553    
1554    22. Added some untidy missing memory free() calls in pcretest, to keep
1555        Electric Fence happy when testing.
1556    
1557    
1558    
1559    Version 5.0 13-Sep-04
1560    ---------------------
1561    
1562     1. Internal change: literal characters are no longer packed up into items
1563        containing multiple characters in a single byte-string. Each character
1564        is now matched using a separate opcode. However, there may be more than one
1565        byte in the character in UTF-8 mode.
1566    
1567     2. The pcre_callout_block structure has two new fields: pattern_position and
1568        next_item_length. These contain the offset in the pattern to the next match
1569        item, and its length, respectively.
1570    
1571     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1572        insertion of callouts before each pattern item. Added the /C option to
1573        pcretest to make use of this.
1574    
1575     4. On the advice of a Windows user, the lines
1576    
1577          #if defined(_WIN32) || defined(WIN32)
1578          _setmode( _fileno( stdout ), 0x8000 );
1579          #endif  /* defined(_WIN32) || defined(WIN32) */
1580    
1581        have been added to the source of pcretest. This apparently does useful
1582        magic in relation to line terminators.
1583    
1584     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1585        for the benefit of those environments where the "b" makes a difference.
1586    
1587     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1588        to know about it. I have put a hack into configure.in that adds in code
1589        to set GCC=yes if CC=icc. This seems to end up at a point in the
1590        generated configure script that is early enough to affect the setting of
1591        compiler options, which is what is needed, but I have no means of testing
1592        whether it really works. (The user who reported this had patched the
1593        generated configure script, which of course I cannot do.)
1594    
1595        LATER: After change 22 below (new libtool files), the configure script
1596        seems to know about icc (and also ecc). Therefore, I have commented out
1597        this hack in configure.in.
1598    
1599     7. Added support for pkg-config (2 patches were sent in).
1600    
1601     8. Negated POSIX character classes that used a combination of internal tables
1602        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1603        [[:^ascii]]. Typically, they would match almost any characters. The other
1604        POSIX classes were not broken in this way.
1605    
1606     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1607        to find the match, as PCRE was deluded into thinking that the match had to
1608        start at the start point or following a newline. The same bug applied to
1609        patterns with negative forward assertions or any backward assertions
1610        preceding ".*" at the start, unless the pattern required a fixed first
1611        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1612    
1613    10. In UTF-8 mode, when moving forwards in the subject after a failed match
1614        starting at the last subject character, bytes beyond the end of the subject
1615        string were read.
1616    
1617    11. Renamed the variable "class" as "classbits" to make life easier for C++
1618        users. (Previously there was a macro definition, but it apparently wasn't
1619        enough.)
1620    
1621    12. Added the new field "tables" to the extra data so that tables can be passed
1622        in at exec time, or the internal tables can be re-selected. This allows
1623        a compiled regex to be saved and re-used at a later time by a different
1624        program that might have everything at different addresses.
1625    
1626    13. Modified the pcre-config script so that, when run on Solaris, it shows a
1627        -R library as well as a -L library.
1628    
1629    14. The debugging options of pcretest (-d on the command line or D on a
1630        pattern) showed incorrect output for anything following an extended class
1631        that contained multibyte characters and which was followed by a quantifier.
1632    
1633    15. Added optional support for general category Unicode character properties
1634        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1635        support. It adds about 90K to the size of the library. The meanings of the
1636        inbuilt class escapes such as \d and \s have NOT been changed.
1637    
1638    16. Updated pcredemo.c to include calls to free() to release the memory for the
1639        compiled pattern.
1640    
1641    17. The generated file chartables.c was being created in the source directory
1642        instead of in the building directory. This caused the build to fail if the
1643        source directory was different from the building directory, and was
1644        read-only.
1645    
1646    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1647        file. No doubt somebody will tell me if they don't make sense... Also added
1648        Dan Mooney's comments about building on OpenVMS.
1649    
1650    19. Added support for partial matching via the PCRE_PARTIAL option for
1651        pcre_exec() and the \P data escape in pcretest.
1652    
1653    20. Extended pcretest with 3 new pattern features:
1654    
1655        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
1656              write the compiled pattern to the file whose name is "rest-of-line".
1657              This is a straight binary dump of the data, with the saved pointer to
1658              the character tables forced to be NULL. The study data, if any, is
1659              written too. After writing, pcretest reads a new pattern.
1660    
1661        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1662              compiled pattern from the given file. There must not be any
1663              occurrences of "<" in the file name (pretty unlikely); if there are,
1664              pcretest will instead treat the initial "<" as a pattern delimiter.
1665              After reading in the pattern, pcretest goes on to read data lines as
1666              usual.
1667    
1668        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1669              and 16-bit fields in a compiled pattern, to simulate a pattern that
1670              was compiled on a host of opposite endianness.
1671    
1672    21. The pcre-exec() function can now cope with patterns that were compiled on
1673        hosts of opposite endianness, with this restriction:
1674    
1675          As for any compiled expression that is saved and used later, the tables
1676          pointer field cannot be preserved; the extra_data field in the arguments
1677          to pcre_exec() should be used to pass in a tables address if a value
1678          other than the default internal tables were used at compile time.
1679    
1680    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1681        now diagnosed as an error. Previously, most of the time, a negative number
1682        would have been treated as zero, but if in addition "ovector" was passed as
1683        NULL, a crash could occur.
1684    
1685    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1686        new versions from the libtool 1.5 distribution (the last one is a copy of
1687        a file called libtool.m4). This seems to have fixed the need to patch
1688        "configure" to support Darwin 1.3 (which I used to do). However, I still
1689        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1690        workstation).
1691    
1692    24. Changed the PCRE licence to be the more standard "BSD" licence.
1693    
1694    
1695    Version 4.5 01-Dec-03
1696    ---------------------
1697    
1698     1. There has been some re-arrangement of the code for the match() function so
1699        that it can be compiled in a version that does not call itself recursively.
1700        Instead, it keeps those local variables that need separate instances for
1701        each "recursion" in a frame on the heap, and gets/frees frames whenever it
1702        needs to "recurse". Keeping track of where control must go is done by means
1703        of setjmp/longjmp. The whole thing is implemented by a set of macros that
1704        hide most of the details from the main code, and operates only if
1705        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1706        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1707        operating.
1708    
1709        To make it easier for callers to provide specially tailored get/free
1710        functions for this usage, two new functions, pcre_stack_malloc, and
1711        pcre_stack_free, are used. They are always called in strict stacking order,
1712        and the size of block requested is always the same.
1713    
1714        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1715        PCRE has been compiled to use the stack or the heap for recursion. The
1716        -C option of pcretest uses this to show which version is compiled.
1717    
1718        A new data escape \S, is added to pcretest; it causes the amounts of store
1719        obtained and freed by both kinds of malloc/free at match time to be added
1720        to the output.
1721    
1722     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1723        what's available on my current Linux desktop machine.
1724    
1725     3. When matching a UTF-8 string, the test for a valid string at the start has
1726        been extended. If start_offset is not zero, PCRE now checks that it points
1727        to a byte that is the start of a UTF-8 character. If not, it returns
1728        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1729        this is necessary because there may be backward assertions in the pattern.
1730        When matching the same subject several times, it may save resources to use
1731        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1732    
1733     4. The code for checking the validity of UTF-8 strings has been tightened so
1734        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1735        containing "overlong sequences".
1736    
1737     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1738        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1739        should have been "&", but it just so happened that all the cases this let
1740        through by mistake were picked up later in the function.
1741    
1742     6. I had used a variable called "isblank" - this is a C99 function, causing
1743        some compilers to warn. To avoid this, I renamed it (as "blankclass").
1744    
1745     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1746        prompting; (b) run "./pcretest /dev/null" at the start of the test script
1747        so the version is shown; (c) stop "make test" echoing "./RunTest".
1748    
1749     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1750    
1751     9. The prototype for memmove() for systems that don't have it was using
1752        size_t, but the inclusion of the header that defines size_t was later. I've
1753        moved the #includes for the C headers earlier to avoid this.
1754    
1755    10. Added some adjustments to the code to make it easier to compiler on certain
1756        special systems:
1757    
1758          (a) Some "const" qualifiers were missing.
1759          (b) Added the macro EXPORT before all exported functions; by default this
1760              is defined to be empty.
1761          (c) Changed the dftables auxiliary program (that builds chartables.c) so
1762              that it reads its output file name as an argument instead of writing
1763              to the standard output and assuming this can be redirected.
1764    
1765    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1766        class containing characters with values greater than 255, PCRE compilation
1767        went into a loop.
1768    
1769    12. A recursive reference to a subpattern that was within another subpattern
1770        that had a minimum quantifier of zero caused PCRE to crash. For example,
1771        (x(y(?2))z)? provoked this bug with a subject that got as far as the
1772        recursion. If the recursively-called subpattern itself had a zero repeat,
1773        that was OK.
1774    
1775    13. In pcretest, the buffer for reading a data line was set at 30K, but the
1776        buffer into which it was copied (for escape processing) was still set at
1777        1024, so long lines caused crashes.
1778    
1779    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1780        "internal error: code overflow...". This applied to any character class
1781        that was followed by a possessive quantifier.
1782    
1783    15. Modified the Makefile to add libpcre.la as a prerequisite for
1784        libpcreposix.la because I was told this is needed for a parallel build to
1785        work.
1786    
1787    16. If a pattern that contained .* following optional items at the start was
1788        studied, the wrong optimizing data was generated, leading to matching
1789        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1790        matching string must start with a or b or c. The correct conclusion for
1791        this pattern is that a match can start with any character.
1792    
1793    
1794    Version 4.4 13-Aug-03
1795    ---------------------
1796    
1797     1. In UTF-8 mode, a character class containing characters with values between
1798        127 and 255 was not handled correctly if the compiled pattern was studied.
1799        In fixing this, I have also improved the studying algorithm for such
1800        classes (slightly).
1801    
1802     2. Three internal functions had redundant arguments passed to them. Removal
1803        might give a very teeny performance improvement.
1804    
1805     3. Documentation bug: the value of the capture_top field in a callout is *one
1806        more than* the number of the hightest numbered captured substring.
1807    
1808     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1809        in incorrectly linking with a previously installed version. They now link
1810        explicitly with libpcre.la.
1811    
1812     5. configure.in no longer needs to recognize Cygwin specially.
1813    
1814     6. A problem in pcre.in for Windows platforms is fixed.
1815    
1816     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1817        pcretest, it used to include the size of the study block as part of its
1818        output. Unfortunately, the structure contains a field that has a different
1819        size on different hardware architectures. This meant that the tests that
1820        showed this size failed. As the block is currently always of a fixed size,
1821        this information isn't actually particularly useful in pcretest output, so
1822        I have just removed it.
1823    
1824     8. Three pre-processor statements accidentally did not start in column 1.
1825        Sadly, there are *still* compilers around that complain, even though
1826        standard C has not required this for well over a decade. Sigh.
1827    
1828     9. In pcretest, the code for checking callouts passed small integers in the
1829        callout_data field, which is a void * field. However, some picky compilers
1830        complained about the casts involved for this on 64-bit systems. Now
1831        pcretest passes the address of the small integer instead, which should get
1832        rid of the warnings.
1833    
1834    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1835        both compile and run time, and gives an error if an invalid UTF-8 sequence
1836        is found. There is a option for disabling this check in cases where the
1837        string is known to be correct and/or the maximum performance is wanted.
1838    
1839    11. In response to a bug report, I changed one line in Makefile.in from
1840    
1841            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1842        to
1843            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1844    
1845        to look similar to other lines, but I have no way of telling whether this
1846        is the right thing to do, as I do not use Windows. No doubt I'll get told
1847        if it's wrong...
1848    
1849    
1850    Version 4.3 21-May-03
1851    ---------------------
1852    
1853    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1854       Makefile.
1855    
1856    2. Some refactoring to improve the quality of the code:
1857    
1858       (i)   The utf8_table... variables are now declared "const".
1859    
1860       (ii)  The code for \cx, which used the "case flipping" table to upper case
1861             lower case letters, now just substracts 32. This is ASCII-specific,
1862             but the whole concept of \cx is ASCII-specific, so it seems
1863             reasonable.
1864    
1865       (iii) PCRE was using its character types table to recognize decimal and
1866             hexadecimal digits in the pattern. This is silly, because it handles
1867             only 0-9, a-f, and A-F, but the character types table is locale-
1868             specific, which means strange things might happen. A private
1869             table is now used for this - though it costs 256 bytes, a table is
1870             much faster than multiple explicit tests. Of course, the standard
1871             character types table is still used for matching digits in subject
1872             strings against \d.
1873    
1874       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1875             ending in _t are). So I've renamed it as ESC_tee.
1876    
1877    3. The first argument for regexec() in the POSIX wrapper should have been
1878       defined as "const".
1879    
1880    4. Changed pcretest to use malloc() for its buffers so that they can be
1881       Electric Fenced for debugging.
1882    
1883    5. There were several places in the code where, in UTF-8 mode, PCRE would try
1884       to read one or more bytes before the start of the subject string. Often this
1885       had no effect on PCRE's behaviour, but in some circumstances it could
1886       provoke a segmentation fault.
1887    
1888    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1889       to try to read one or more bytes before the start of the subject string.
1890    
1891    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1892       UTF-8 support could misbehave in various ways if the subject string
1893       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1894       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1895       back over UTF-8 characters.)
1896    
1897    
1898    Version 4.2 14-Apr-03
1899    ---------------------
1900    
1901    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1902    
1903    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1904         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1905         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1906         [WIN_PREFIX]: new variable, "cyg" for Cygwin
1907         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1908           and BUILD_EXEEXT
1909         Note: automatic setting of the BUILD variables is not yet working
1910         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1911           compile-time but not at link-time
1912         [LINK]: use for linking executables only
1913         make different versions for Windows and non-Windows
1914         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1915           libraries
1916         [LINK_FOR_BUILD]: new variable
1917         [OBJEXT]: use throughout
1918         [EXEEXT]: use throughout
1919         <winshared>: new target
1920         <wininstall>: new target
1921         <dftables.o>: use native compiler
1922         <dftables>: use native linker
1923         <install>: handle Windows platform correctly
1924         <clean>: ditto
1925         <check>: ditto
1926         copy DLL to top builddir before testing
1927    
1928       As part of these changes, -no-undefined was removed again. This was reported
1929       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1930       in any case.
1931    
1932    3. Some tidies to get rid of compiler warnings:
1933    
1934       . In the match_data structure, match_limit was an unsigned long int, whereas
1935         match_call_count was an int. I've made them both unsigned long ints.
1936    
1937       . In pcretest the fact that a const uschar * doesn't automatically cast to
1938         a void * provoked a warning.
1939    
1940       . Turning on some more compiler warnings threw up some "shadow" variables
1941         and a few more missing casts.
1942    
1943    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1944       option, a class that contained a single character with a value between 128
1945       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1946    
1947    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1948       option, a class that contained several characters, but with at least one
1949       whose value was between 128 and 255 caused PCRE to crash.
1950    
1951    
1952    Version 4.1 12-Mar-03
1953    ---------------------
1954    
1955    1. Compiling with gcc -pedantic found a couple of places where casts were
1956    needed, and a string in dftables.c that was longer than standard compilers are
1957    required to support.
1958    
1959    2. Compiling with Sun's compiler found a few more places where the code could
1960    be tidied up in order to avoid warnings.
1961    
1962    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1963    first of these names is deprecated in the latest Autoconf in favour of the name
1964    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1965    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1966    analogy I have changed it to CFLAGS_FOR_BUILD.
1967    
1968    4. Added -no-undefined to the linking command in the Makefile, because this is
1969    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1970    linking step for the pcreposix library.
1971    
1972    5. PCRE was failing to diagnose the case of two named groups with the same
1973    name.
1974    
1975    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1976    literal character that is needed in the subject for a match, and scans along to
1977    ensure that it is present before embarking on the full matching process. This
1978    saves time in cases of nested unlimited repeats that are never going to match.
1979    Problem: the scan can take a lot of time if the subject is very long (e.g.
1980    megabytes), thus penalizing straightforward matches. It is now done only if the
1981    amount of subject to be scanned is less than 1000 bytes.
1982    
1983    7. A lesser problem with the same optimization is that it was recording the
1984    first character of an anchored pattern as "needed", thus provoking a search
1985    right along the subject, even when the first match of the pattern was going to
1986    fail. The "needed" character is now not set for anchored patterns, unless it
1987    follows something in the pattern that is of non-fixed length. Thus, it still
1988    fulfils its original purpose of finding quick non-matches in cases of nested
1989    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1990    
1991    
1992    Version 4.0 17-Feb-03
1993    ---------------------
1994    
1995    1. If a comment in an extended regex that started immediately after a meta-item
1996    extended to the end of string, PCRE compiled incorrect data. This could lead to
1997    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1998    
1999    2. Moved to autoconf 2.53 and libtool 1.4.2.
2000    
2001    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2002    the special perltest8 script is no longer needed - all the tests can be run
2003    from a single perltest script.
2004    
2005    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2006    by \s. It has now been removed in PCRE. This means it isn't recognized as
2007    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2008    class [:space:] *does* include VT, thereby creating a mess.
2009    
2010    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2011    space and tab.
2012    
2013    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2014    its new features into the main test script, reducing the number of scripts.
2015    
2016    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2017    were backward compatible, and made the (?i) apply to the whole pattern, as if
2018    /i were given. Now it behaves more logically, and applies the option setting
2019    only to what follows. PCRE has been changed to follow suit. However, if it
2020    finds options settings right at the start of the pattern, it extracts them into
2021    the global options, as before. Thus, they show up in the info data.
2022    
2023    8. Added support for the \Q...\E escape sequence. Characters in between are
2024    treated as literals. This is slightly different from Perl in that $ and @ are
2025    also handled as literals inside the quotes. In Perl, they will cause variable
2026    interpolation. Note the following examples:
2027    
2028        Pattern            PCRE matches      Perl matches
2029    
2030        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2031        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2032        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2033    
2034    For compatibility with Perl, \Q...\E sequences are recognized inside character
2035    classes as well as outside them.
2036    
2037    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2038    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2039    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2040    signed/unsigned warnings.
2041    
2042    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2043    option for pcretest, so I've replaced it by a simple function that does just
2044    that job.
2045    
2046    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2047    "pcregrep -".
2048    
2049    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2050    Java package. This provides some syntactic sugar for simple cases of what my
2051    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2052    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2053    item, you can use this simplified notation. Note that only makes sense with
2054    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2055    greediness, whatever the setting of the PCRE_UNGREEDY option.
2056    
2057    13. A change of greediness default within a pattern was not taking effect at
2058    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2059    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2060    was abstracted outside.
2061    
2062    14. PCRE now supports the \G assertion. It is true when the current matching
2063    position is at the start point of the match. This differs from \A when the
2064    starting offset is non-zero. Used with the /g option of pcretest (or similar
2065    code), it works in the same way as it does for Perl's /g option. If all
2066    alternatives of a regex begin with \G, the expression is anchored to the start
2067    match position, and the "anchored" flag is set in the compiled expression.
2068    
2069    15. Some bugs concerning the handling of certain option changes within patterns
2070    have been fixed. These applied to options other than (?ims). For example,
2071    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2072    been the other way round. Some of this was related to change 7 above.
2073    
2074    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2075    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2076    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2077    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2078    
2079    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2080    mode. Unlike ".", it always matches newline, whatever the setting of
2081    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2082    assertions. Perl allows it, but it doesn't (in general) work because it can't
2083    calculate the length of the lookbehind. At least, that's the case for Perl
2084    5.8.0 - I've been told they are going to document that it doesn't work in
2085    future.
2086    
2087    18. Added an error diagnosis for escapes that PCRE does not support: these are
2088    \L, \l, \N, \P, \p, \U, \u, and \X.
2089    
2090    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2091    reading past the end of the pattern in cases such as /[abcd/.
2092    
2093    20. PCRE was getting more memory than necessary for patterns with classes that
2094    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2095    
2096    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2097    compiling PCRE for use with Virtual Pascal.
2098    
2099    22. Small fix to the Makefile to make it work properly if the build is done
2100    outside the source tree.
2101    
2102    23. Added a new extension: a condition to go with recursion. If a conditional
2103    subpattern starts with (?(R) the "true" branch is used if recursion has
2104    happened, whereas the "false" branch is used only at the top level.
2105    
2106    24. When there was a very long string of literal characters (over 255 bytes
2107    without UTF support, over 250 bytes with UTF support), the computation of how
2108    much memory was required could be incorrect, leading to segfaults or other
2109    strange effects.
2110    
2111    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2112    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2113    there was a subsequent back reference to those brackets. This meant that, for
2114    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2115    possible to check for precisely this case. All we can do is abandon the
2116    optimization if .* occurs inside capturing brackets when there are any back
2117    references whatsoever. (See below for a better fix that came later.)
2118    
2119    26. The handling of the optimization for finding the first character of a
2120    non-anchored pattern, and for finding a character that is required later in the
2121    match were failing in some cases. This didn't break the matching; it just
2122    failed to optimize when it could. The way this is done has been re-implemented.
2123    
2124    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2125    
2126    28. Added a new feature that provides some of the functionality that Perl
2127    provides with (?{...}). The facility is termed a "callout". The way it is done
2128    in PCRE is for the caller to provide an optional function, by setting
2129    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2130    global variable. By default it is unset, which disables all calling out. To get
2131    the function called, the regex must include (?C) at appropriate points. This
2132    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2133    This provides a means of identifying different callout points. When PCRE
2134    reaches such a point in the regex, if pcre_callout has been set, the external
2135    function is called. It is provided with data in a structure called
2136    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2137    matching continues; if it returns a non-zero value, the match at the current
2138    point fails. However, backtracking will occur if possible. [This was changed
2139    later and other features added - see item 49 below.]
2140    
2141    29. pcretest is upgraded to test the callout functionality. It provides a
2142    callout function that displays information. By default, it shows the start of
2143    the match and the current position in the text. There are some new data escapes
2144    to vary what happens:
2145    
2146        \C+         in addition, show current contents of captured substrings
2147        \C-         do not supply a callout function
2148        \C!n        return 1 when callout number n is reached
2149        \C!n!m      return 1 when callout number n is reached for the mth time
2150    
2151    30. If pcregrep was called with the -l option and just a single file name, it
2152    output "<stdin>" if a match was found, instead of the file name.
2153    
2154    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2155    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2156    pcre_exec(). This saves a malloc/free per call. The default value of
2157    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2158    when configuring.
2159    
2160    32. The default maximum size of a compiled pattern is 64K. There have been a
2161    few cases of people hitting this limit. The code now uses macros to handle the
2162    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2163    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2164    configuring. Tests 2 and 5 work only with 2-byte links because they output
2165    debugging information about compiled patterns.
2166    
2167    33. Internal code re-arrangements:
2168    
2169    (a) Moved the debugging function for printing out a compiled regex into
2170        its own source file (printint.c) and used #include to pull it into
2171        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2172        separate copies.
2173    
2174    (b) Defined the list of op-code names for debugging as a macro in
2175        internal.h so that it is next to the definition of the opcodes.
2176    
2177    (c) Defined a table of op-code lengths for simpler skipping along compiled
2178        code. This is again a macro in internal.h so that it is next to the
2179        definition of the opcodes.
2180    
2181    34. Added support for recursive calls to individual subpatterns, along the
2182    lines of Robin Houston's patch (but implemented somewhat differently).
2183    
2184    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2185    allow it to read and process whole directories in Win32. This code was
2186    contributed by Lionel Fourquaux; it has not been tested by me.
2187    
2188    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2189    used to name a group. Names consist of alphanumerics and underscores, and must
2190    be unique. Back references use the syntax (?P=name) and recursive calls use
2191    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2192    numbers. The function pcre_fullinfo() can be used after compilation to extract
2193    a name/number map. There are three relevant calls:
2194    
2195      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2196      PCRE_INFO_NAMECOUNT            yields the number of entries
2197      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2198    
2199    The map is a vector of fixed-size entries. The size of each entry depends on
2200    the length of the longest name used. The first two bytes of each entry are the
2201    group number, most significant byte first. There follows the corresponding
2202    name, zero terminated. The names are in alphabetical order.
2203    
2204    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2205    case instead of 255. Making it the same both with and without UTF-8 support
2206    means that the same test output works with both.
2207    
2208    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2209    calling malloc() with a zero argument.
2210    
2211    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2212    optimization. I've improved things by keeping a bitmap of backreferences with
2213    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2214    fact referenced, the optimization can be applied. It is unlikely that a
2215    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2216    the match to follow \n) will appear inside brackets with a number greater than
2217    31, but if it does, any back reference > 31 suppresses the optimization.
2218    
2219    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2220    of disabling numbered capturing parentheses. Any opening parenthesis that is
2221    not followed by ? behaves as if it were followed by ?: but named parentheses
2222    can still be used for capturing (and they will acquire numbers in the usual
2223    way).
2224    
2225    41. Redesigned the return codes from the match() function into yes/no/error so
2226    that errors can be passed back from deep inside the nested calls. A malloc
2227    failure while inside a recursive subpattern call now causes the
2228    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2229    
2230    42. It is now possible to set a limit on the number of times the match()
2231    function is called in a call to pcre_exec(). This facility makes it possible to
2232    limit the amount of recursion and backtracking, though not in a directly
2233    obvious way, because the match() function is used in a number of different
2234    circumstances. The count starts from zero for each position in the subject
2235    string (for non-anchored patterns). The default limit is, for compatibility, a
2236    large number, namely 10 000 000. You can change this in two ways:
2237    
2238    (a) When configuring PCRE before making, you can use --with-match-limit=n
2239        to set a default value for the compiled library.
2240    
2241    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2242        a different value is set. See 45 below.
2243    
2244    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2245    
2246    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2247    of things that can be changed at compile time. The first argument specifies
2248    what is wanted and the second points to where the information is to be placed.
2249    The current list of available information is:
2250    
2251      PCRE_CONFIG_UTF8
2252    
2253    The output is an integer that is set to one if UTF-8 support is available;
2254    otherwise it is set to zero.
2255    
2256      PCRE_CONFIG_NEWLINE
2257    
2258    The output is an integer that it set to the value of the code that is used for
2259    newline. It is either LF (10) or CR (13).
2260    
2261      PCRE_CONFIG_LINK_SIZE
2262    
2263    The output is an integer that contains the number of bytes used for internal
2264    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2265    
2266      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2267    
2268    The output is an integer that contains the threshold above which the POSIX
2269    interface uses malloc() for output vectors. See item 31 above.
2270    
2271      PCRE_CONFIG_MATCH_LIMIT
2272    
2273    The output is an unsigned integer that contains the default limit of the number
2274    of match() calls in a pcre_exec() execution. See 42 above.
2275    
2276    44. pcretest has been upgraded by the addition of the -C option. This causes it
2277    to extract all the available output from the new pcre_config() function, and to
2278    output it. The program then exits immediately.
2279    
2280    45. A need has arisen to pass over additional data with calls to pcre_exec() in
2281    order to support additional features. One way would have been to define
2282    pcre_exec2() (for example) with extra arguments, but this would not have been
2283    extensible, and would also have required all calls to the original function to
2284    be mapped to the new one. Instead, I have chosen to extend the mechanism that
2285    is used for passing in "extra" data from pcre_study().
2286    
2287    The pcre_extra structure is now exposed and defined in pcre.h. It currently
2288    contains the following fields:
2289    
2290      flags         a bitmap indicating which of the following fields are set
2291      study_data    opaque data from pcre_study()
2292      match_limit   a way of specifying a limit on match() calls for a specific
2293                      call to pcre_exec()
2294      callout_data  data for callouts (see 49 below)
2295    
2296    The flag bits are also defined in pcre.h, and are
2297    
2298      PCRE_EXTRA_STUDY_DATA
2299      PCRE_EXTRA_MATCH_LIMIT
2300      PCRE_EXTRA_CALLOUT_DATA
2301    
2302    The pcre_study() function now returns one of these new pcre_extra blocks, with
2303    the actual study data pointed to by the study_data field, and the
2304    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2305    before. That is, this change is entirely upwards-compatible and requires no
2306    change to existing code.
2307    
2308    If you want to pass in additional data to pcre_exec(), you can either place it
2309    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2310    block.
2311    
2312    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2313    data string contains the escape sequence \M, pcretest calls pcre_exec() several
2314    times with different match limits, until it finds the minimum value needed for
2315    pcre_exec() to complete. The value is then output. This can be instructive; for
2316    most simple matches the number is quite small, but for pathological cases it
2317    gets very large very quickly.
2318    
2319    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2320    returns the size of the data block pointed to by the study_data field in a
2321    pcre_extra block, that is, the value that was passed as the argument to
2322    pcre_malloc() when PCRE was getting memory in which to place the information
2323    created by pcre_study(). The fourth argument should point to a size_t variable.
2324    pcretest has been extended so that this information is shown after a successful
2325    pcre_study() call when information about the compiled regex is being displayed.
2326    
2327    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2328    because what follows is always an absolute path. (Later: it turns out that this
2329    is more than cosmetic for MinGW, because it doesn't like empty path
2330    components.)
2331    
2332    49. Some changes have been made to the callout feature (see 28 above):
2333    
2334    (i)  A callout function now has three choices for what it returns:
2335    
2336           0  =>  success, carry on matching
2337         > 0  =>  failure at this point, but backtrack if possible
2338         < 0  =>  serious error, return this value from pcre_exec()
2339    
2340         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2341         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2342         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2343         use by callout functions. It will never be used by PCRE itself.
2344    
2345    (ii) The pcre_extra structure (see 45 above) has a void * field called
2346         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2347         pcre_callout_block structure has a field of the same name. The contents of
2348         the field passed in the pcre_extra structure are passed to the callout
2349         function in the corresponding field in the callout block. This makes it
2350         easier to use the same callout-containing regex from multiple threads. For
2351         testing, the pcretest program has a new data escape
2352    
2353           \C*n        pass the number n (may be negative) as callout_data
2354    
2355         If the callout function in pcretest receives a non-zero value as
2356         callout_data, it returns that value.
2357    
2358    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2359    there were some redundant $(CFLAGS) in commands that are now specified as
2360    $(LINK), which already includes $(CFLAGS).
2361    
2362    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2363    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2364    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2365    one-byte characters throughout. Note that case-insensitive matching applies
2366    only to characters whose values are less than 256. PCRE doesn't support the
2367    notion of cases for higher-valued characters.
2368    
2369    (i)   A character class whose characters are all within 0-255 is handled as
2370          a bit map, and the map is inverted for negative classes. Previously, a
2371          character > 255 always failed to match such a class; however it should
2372          match if the class was a negative one (e.g. [^ab]). This has been fixed.
2373    
2374    (ii)  A negated character class with a single character < 255 is coded as
2375          "not this character" (OP_NOT). This wasn't working properly when the test
2376          character was multibyte, either singly or repeated.
2377    
2378    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2379          mode, for example: \x{100}{2,3}.
2380    
2381    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2382          singly or repeated) now correctly test multibyte characters. However,
2383          PCRE doesn't recognize any characters with values greater than 255 as
2384          digits, spaces, or word characters. Such characters always match \D, \S,
2385          and \W, and never match \d, \s, or \w.
2386    
2387    (v)   Classes may now contain characters and character ranges with values
2388          greater than 255. For example: [ab\x{100}-\x{400}].
2389    
2390    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
2391          PCRE in UTF-8 mode.
2392    
2393    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2394    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2395    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2396    value.)
2397    
2398    53. The single man page has become too large. I have therefore split it up into
2399    a number of separate man pages. These also give rise to individual HTML pages;
2400    these are now put in a separate directory, and there is an index.html page that
2401    lists them all. Some hyperlinking between the pages has been installed.
2402    
2403    54. Added convenience functions for handling named capturing parentheses.
2404    
2405    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2406    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2407    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2408    are faulted.
2409    
2410    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2411    calling configure. These values are used when compiling the dftables.c program
2412    which is run to generate the source of the default character tables. They
2413    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2414    you will need to set these values.
2415    
2416    57. Updated the building process for Windows DLL, as provided by Fred Cox.
2417    
2418    
2419    Version 3.9 02-Jan-02
2420    ---------------------
2421    
2422    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2423    
2424    2. If --disable-static was given, the building process failed when trying to
2425    build pcretest and pcregrep. (For some reason it was using libtool to compile
2426    them, which is not right, as they aren't part of the library.)
2427    
2428    
2429    Version 3.8 18-Dec-01
2430    ---------------------
2431    
2432    1. The experimental UTF-8 code was completely screwed up. It was packing the
2433    bytes in the wrong order. How dumb can you get?
2434    
2435    
2436    Version 3.7 29-Oct-01
2437    ---------------------
2438    
2439    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2440    This caused pcretest, when used on the test data, to segfault. Unfortunately,
2441    this didn't happen under Solaris 8, where I normally test things.
2442    
2443    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2444    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2445    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2446    this fix an hour or so after the initial 3.7 release.)
2447    
2448    
2449    Version 3.6 23-Oct-01
2450    ---------------------
2451    
2452    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2453    offsets passed as NULL with zero offset count.
2454    
2455    2. The config.guess and config.sub files had not been updated when I moved to
2456    the latest autoconf.
2457    
2458    
2459    Version 3.5 15-Aug-01
2460    ---------------------
2461    
2462    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2463    had been forgotten.
2464    
2465    2. By using declared but undefined structures, we can avoid using "void"
2466    definitions in pcre.h while keeping the internal definitions of the structures
2467    private.
2468    
2469    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2470    user point of view, this means that both static and shared libraries are built
2471    by default, but this can be individually controlled. More of the work of
2472    handling this static/shared cases is now inside libtool instead of PCRE's make
2473    file.
2474    
2475    4. The pcretest utility is now installed along with pcregrep because it is
2476    useful for users (to test regexs) and by doing this, it automatically gets
2477    relinked by libtool. The documentation has been turned into a man page, so
2478    there are now .1, .txt, and .html versions in /doc.
2479    
2480    5. Upgrades to pcregrep:
2481       (i)   Added long-form option names like gnu grep.
2482       (ii)  Added --help to list all options with an explanatory phrase.
2483       (iii) Added -r, --recursive to recurse into sub-directories.
2484       (iv)  Added -f, --file to read patterns from a file.
2485    
2486    6. pcre_exec() was referring to its "code" argument before testing that
2487    argument for NULL (and giving an error if it was NULL).
2488    
2489    7. Upgraded Makefile.in to allow for compiling in a different directory from
2490    the source directory.
2491    
2492    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2493    options bits, the pointer it was passed was to an int instead of to an unsigned
2494    long int. This mattered only on 64-bit systems.
2495    
2496    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2497    generated) instead of pcre.in, which it its source. Also made the same change
2498    in several of the .c files.
2499    
2500    10. A new release of gcc defines printf() as a macro, which broke pcretest
2501    because it had an ifdef in the middle of a string argument for printf(). Fixed
2502    by using separate calls to printf().
2503    
2504    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2505    script, to force use of CR or LF instead of \n in the source. On non-Unix
2506    systems, the value can be set in config.h.
2507    
2508    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2509    absolute limit. Changed the text of the error message to make this clear, and
2510    likewise updated the man page.
2511    
2512    13. The limit of 99 on the number of capturing subpatterns has been removed.
2513    The new limit is 65535, which I hope will not be a "real" limit.
2514    
2515    
2516    Version 3.4 22-Aug-00
2517    ---------------------
2518    
2519    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2520    
2521    2. Diagnose condition (?(0) as an error instead of crashing on matching.
2522    
2523    
2524    Version 3.3 01-Aug-00
2525    ---------------------
2526    
2527    1. If an octal character was given, but the value was greater than \377, it
2528    was not getting masked to the least significant bits, as documented. This could
2529    lead to crashes in some systems.
2530    
2531    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2532    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2533    
2534    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2535    These just pass their arguments on to (pcre_free)(), but they are provided
2536    because some uses of PCRE bind it to non-C systems that can call its functions,
2537    but cannot call free() or pcre_free() directly.
2538    
2539    4. Add "make test" as a synonym for "make check". Corrected some comments in
2540    the Makefile.
2541    
2542    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2543    Makefile.
2544    
2545    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2546    command called pgrep for grepping around the active processes.
2547    
2548    7. Added the beginnings of support for UTF-8 character strings.
2549    
2550    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2551    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2552    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2553    out for the ar command.)
2554    
2555    
2556    Version 3.2 12-May-00
2557    ---------------------
2558    
2559    This is purely a bug fixing release.
2560    
2561    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2562    of ZA. This was just one example of several cases that could provoke this bug,
2563    which was introduced by change 9 of version 2.00. The code for breaking
2564    infinite loops after an iteration that matches an empty string was't working
2565    correctly.
2566    
2567    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2568    when matched against abbab (for example). After matching an empty string, it
2569    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2570    caused it to match further down the string than it should.
2571    
2572    3. The code contained an inclusion of sys/types.h. It isn't clear why this
2573    was there because it doesn't seem to be needed, and it causes trouble on some
2574    systems, as it is not a Standard C header. It has been removed.
2575    
2576    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2577    were reported on the Macintosh. The changes were from
2578    
2579      while ((c = *(++ptr)) != 0 && c != '\n');
2580    to
2581      while ((c = *(++ptr)) != 0 && c != '\n') ;
2582    
2583    Totally extraordinary, but if that's what it takes...
2584    
2585    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2586    available. Added HAVE_BCOPY and an autoconf test for it; if neither
2587    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2588    assumes the way PCRE uses memmove() (always moving upwards).
2589    
2590    6. PCRE is being used in one environment where strchr() is not available. There
2591    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2592    faster code anyway.
2593    
2594    
2595    Version 3.1 09-Feb-00
2596    ---------------------
2597    
2598    The only change in this release is the fixing of some bugs in Makefile.in for
2599    the "install" target:
2600    
2601    (1) It was failing to install pcreposix.h.
2602    
2603    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2604    
2605    
2606    Version 3.0 01-Feb-00
2607    ---------------------
2608    
2609    1. Add support for the /+ modifier to perltest (to output $` like it does in
2610    pcretest).
2611    
2612    2. Add support for the /g modifier to perltest.
2613    
2614    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2615    matches null strings.
2616    
2617    4. Fix perltest so that it doesn't do unwanted things when fed an empty
2618    pattern. Perl treats empty patterns specially - it reuses the most recent
2619    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2620    effect.
2621    
2622    5. The POSIX interface was broken in that it was just handing over the POSIX
2623    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2624    required a bigger vector, with some working space on the end. This means that
2625    the POSIX wrapper now has to get and free some memory, and copy the results.
2626    
2627    6. Added some simple autoconf support, placing the test data and the
2628    documentation in separate directories, re-organizing some of the
2629    information files, and making it build pcre-config (a GNU standard). Also added
2630    libtool support for building PCRE as a shared library, which is now the
2631    default.
2632    
2633    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2634    09 are not valid octal constants. Single digits will be used for minor values
2635    less than 10.
2636    
2637    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2638    existing programs that set these in the POSIX interface can use PCRE without
2639    modification.
2640    
2641    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2642    return all that pcre_info() returns, plus additional data. The pcre_info()
2643    function is retained for compatibility, but is considered to be obsolete.
2644    
2645    10. Added experimental recursion feature (?R) to handle one common case that
2646    Perl 5.6 will be able to do with (?p{...}).
2647    
2648    11. Added support for POSIX character classes like [:alpha:], which Perl is
2649    adopting.
2650    
2651    
2652    Version 2.08 31-Aug-99
2653    ----------------------
2654    
2655    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2656    trying to match at the startoffset position, but instead was moving forward to
2657    the next newline as if a previous match had failed.
2658    
2659    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2660    and could get into a loop if a null string was matched other than at the start
2661    of the subject.
2662    
2663    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2664    be distinguished at compile time, and for completeness also added PCRE_DATE.
2665    
2666    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2667    in GnuWin32 environments.
2668    
2669    
2670    Version 2.07 29-Jul-99
2671    ----------------------
2672    
2673    1. The documentation is now supplied in plain text form and HTML as well as in
2674    the form of man page sources.
2675    
2676    2. C++ compilers don't like assigning (void *) values to other pointer types.
2677    In particular this affects malloc(). Although there is no problem in Standard
2678    C, I've put in casts to keep C++ compilers happy.
2679    
2680    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2681    should be (const char *).
2682    
2683    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2684    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2685    However, I haven't made this a standard facility. The documentation doesn't
2686    mention it, and the Makefile doesn't support it.
2687    
2688    5. The Makefile now contains an "install" target, with editable destinations at
2689    the top of the file. The pcretest program is not installed.
2690    
2691    6. pgrep -V now gives the PCRE version number and date.
2692    
2693    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2694    causing the entire string to be ignored, instead of just the last character.
2695    
2696    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2697    non-matching string, it can take a very, very long time, even for strings of
2698    quite modest length, because of the nested recursion. PCRE now does better in
2699    some of these cases. It does this by remembering the last required literal
2700    character in the pattern, and pre-searching the subject to ensure it is present
2701    before running the real match. In other words, it applies a heuristic to detect
2702    some types of certain failure quickly, and in the above example, if presented
2703    with a string that has no trailing " it gives "no match" very quickly.
2704    
2705    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2706    other alternatives are tried instead.
2707    
2708    
2709    Version 2.06 09-Jun-99
2710    ----------------------
2711    
2712    1. Change pcretest's output for amount of store used to show just the code
2713    space, because the remainder (the data block) varies in size between 32-bit and
2714    64-bit systems.
2715    
2716    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2717    start matching at. This allows lookbehinds to work when searching for multiple
2718    occurrences in a string.
2719    
2720    3. Added additional options to pcretest for testing multiple occurrences:
2721    
2722       /+   outputs the rest of the string that follows a match
2723       /g   loops for multiple occurrences, using the new startoffset argument
2724       /G   loops for multiple occurrences by passing an incremented pointer
2725    
2726    4. PCRE wasn't doing the "first character" optimization for patterns starting
2727    with \b or \B, though it was doing it for other lookbehind assertions. That is,
2728    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2729    the letter 'x'. On long subject strings, this gives a significant speed-up.
2730    
2731    
2732    Version 2.05 21-Apr-99
2733    ----------------------
2734    
2735    1. Changed the type of magic_number from int to long int so that it works
2736    properly on 16-bit systems.
2737    
2738    2. Fixed a bug which caused patterns starting with .* not to work correctly
2739    when the subject string contained newline characters. PCRE was assuming
2740    anchoring for such patterns in all cases, which is not correct because .* will
2741    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2742    DOTALL is set at top level; otherwise it knows that patterns starting with .*
2743    must be retried after every newline in the subject.
2744    
2745    
2746    Version 2.04 18-Feb-99
2747    ----------------------
2748    
2749    1. For parenthesized subpatterns with repeats whose minimum was zero, the
2750    computation of the store needed to hold the pattern was incorrect (too large).
2751    If such patterns were nested a few deep, this could multiply and become a real
2752    problem.
2753    
2754    2. Added /M option to pcretest to show the memory requirement of a specific
2755    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2756    
2757    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2758    compiled in such a way that the backtracking after subsequent failure was
2759    pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2760    ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2761    
2762    
2763    Version 2.03 02-Feb-99
2764    ----------------------
2765    
2766    1. Fixed typo and small mistake in man page.
2767    
2768    2. Added 4th condition (GPL supersedes if conflict) and created separate
2769    LICENCE file containing the conditions.
2770    
2771    3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2772    Perl, that is the internal \ allows the delimiter to be included in the
2773    pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2774    the final delimiter, add \ to the end of the pattern (to test the error).
2775    
2776    4. Added the convenience functions for extracting substrings after a successful
2777    match. Updated pcretest to make it able to test these functions.
2778    
2779    
2780    Version 2.02 14-Jan-99
2781    ----------------------
2782    
2783    1. Initialized the working variables associated with each extraction so that
2784    their saving and restoring doesn't refer to uninitialized store.
2785    
2786    2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2787    compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2788    fix the problem.
2789    
2790    3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2791    calls, and wasn't printing the correct value for compiling calls. Increased the
2792    default value of LOOPREPEAT, and the number of significant figures in the
2793    times.
2794    
2795    4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2796    
2797    5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2798    a building problem on Windows NT with a FAT file system.
2799    
2800    
2801    Version 2.01 21-Oct-98
2802    ----------------------
2803    
2804    1. Changed the API for pcre_compile() to allow for the provision of a pointer
2805    to character tables built by pcre_maketables() in the current locale. If NULL
2806    is passed, the default tables are used.
2807    
2808    
2809    Version 2.00 24-Sep-98
2810    ----------------------
2811    
2812    1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2813    it any more.
2814    
2815    2. Allow quantification of (?>) groups, and make it work correctly.
2816    
2817    3. The first character computation wasn't working for (?>) groups.
2818    
2819    4. Correct the implementation of \Z (it is permitted to match on the \n at the
2820    end of the subject) and add 5.005's \z, which really does match only at the
2821    very end of the subject.
2822    
2823    5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2824    
2825    6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2826    DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2827    localized options. All options to pcre_study() were also removed.
2828    
2829    7. Add other new features from 5.005:
2830    
2831       $(?<=           positive lookbehind
2832       $(?<!           negative lookbehind
2833       (?imsx-imsx)    added the unsetting capability
2834                       such a setting is global if at outer level; local otherwise
2835       (?imsx-imsx:)   non-capturing groups with option setting
2836       (?(cond)re|re)  conditional pattern matching
2837    
2838       A backreference to itself in a repeated group matches the previous
2839       captured string.
2840    
2841    8. General tidying up of studying (both automatic and via "study")
2842    consequential on the addition of new assertions.
2843    
2844    9. As in 5.005, unlimited repeated groups that could match an empty substring
2845    are no longer faulted at compile time. Instead, the loop is forcibly broken at
2846    runtime if any iteration does actually match an empty substring.
2847    
2848    10. Include the RunTest script in the distribution.
2849    
2850    11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2851    discrepancies, some of which were old and were also with respect to 5.004. They
2852    have now been fixed.
2853    
2854    
2855    Version 1.09 28-Apr-98
2856    ----------------------
2857    
2858    1. A negated single character class followed by a quantifier with a minimum
2859    value of one (e.g.  [^x]{1,6}  ) was not compiled correctly. This could lead to
2860    program crashes, or just wrong answers. This did not apply to negated classes
2861    containing more than one character, or to minima other than one.
2862    
2863    
2864    Version 1.08 27-Mar-98
2865    ----------------------
2866    
2867    1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2868    
2869    2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2870    latter must appear before anything that relies on it in the pattern.
2871    
2872    
2873    Version 1.07 16-Feb-98
2874    ----------------------
2875    
2876    1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2877    repeat of a potentially empty string).
2878    
2879    
2880    Version 1.06 23-Jan-98
2881    ----------------------
2882    
2883    1. Added Markus Oberhumer's little patches for C++.
2884    
2885    2. Literal strings longer than 255 characters were broken.
2886    
2887    
2888    Version 1.05 23-Dec-97
2889    ----------------------
2890    
2891    1. Negated character classes containing more than one character were failing if
2892    PCRE_CASELESS was set at run time.
2893    
2894    
2895    Version 1.04 19-Dec-97
2896    ----------------------
2897    
2898    1. Corrected the man page, where some "const" qualifiers had been omitted.
2899    
2900    2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2901    input syntax.
2902    
2903    3. Fixed memory leak which occurred when a regex with back references was
2904    matched with an offsets vector that wasn't big enough. The temporary memory
2905    that is used in this case wasn't being freed if the match failed.
2906    
2907    4. Tidied pcretest to ensure it frees memory that it gets.
2908    
2909    5. Temporary memory was being obtained in the case where the passed offsets
2910    vector was exactly big enough.
2911    
2912    6. Corrected definition of offsetof() from change 5 below.
2913    
2914    7. I had screwed up change 6 below and broken the rules for the use of
2915    setjmp(). Now fixed.
2916    
2917    
2918    Version 1.03 18-Dec-97
2919    ----------------------
2920    
2921    1. A erroneous regex with a missing opening parenthesis was correctly
2922    diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2923    on some systems.
2924    
2925    2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2926    it was reported that one broken compiler failed on the former because "code" is
2927    also an independent variable.
2928    
2929    3. The erroneous regex a[]b caused an array overrun reference.
2930    
2931    4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2932    fail on data ending with that character. (It was going on too far, and checking
2933    the next character, typically a binary zero.) This was specific to the
2934    optimized code for single-character negative classes.
2935    
2936    5. Added a contributed patch from the TIN world which does the following:
2937    
2938      + Add an undef for memmove, in case the the system defines a macro for it.
2939    
2940      + Add a definition of offsetof(), in case there isn't one. (I don't know
2941        the reason behind this - offsetof() is part of the ANSI standard - but
2942        it does no harm).
2943    
2944      + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2945        most of the places where whitespace preceded '#'. I have given up and
2946        allowed the remaining 2 cases to be at the margin.
2947    
2948      + Rename some variables in pcre to eliminate shadowing. This seems very
2949        pedantic, but does no harm, of course.
2950    
2951    6. Moved the call to setjmp() into its own function, to get rid of warnings
2952    from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2953    
2954    7. Constructs such as \d{8,} were compiling into the equivalent of
2955    \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2956    outcome, but in this particular case used more store than had been allocated,
2957    which caused the bug to be discovered because it threw up an internal error.
2958    
2959    8. The debugging code in both pcre and pcretest for outputting the compiled
2960    form of a regex was going wrong in the case of back references followed by
2961    curly-bracketed repeats.
2962    
2963    
2964  Version 1.02 12-Dec-97  Version 1.02 12-Dec-97
2965  ----------------------  ----------------------

Legend:
Removed from v.7  
changed lines
  Added in v.303

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12