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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.19  
changed lines
  Added in v.302

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12