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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.61  
changed lines
  Added in v.233

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12