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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.45  
changed lines
  Added in v.231

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12