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

Diff of /code/trunk/ChangeLog

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

revision 59 by nigel, Sat Feb 24 21:39:54 2007 UTC revision 321 by ph10, Wed Mar 5 17:10:45 2008 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 7.7 05-Mar-08
5    ---------------------
6    
7    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
8        a string to a long long, pretend we don't even have a long long." This is
9        done by checking for the strtoq, strtoll, and _strtoi64 functions.
10    
11    
12    Version 7.6 28-Jan-08
13    ---------------------
14    
15    1.  A character class containing a very large number of characters with
16        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
17        overflow.
18    
19    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
20        HAVE_LONG_LONG is not defined.
21    
22    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
23        bring them up to date and include new features. This patch includes:
24    
25        - Fixed PH's badly added libz and libbz2 support.
26        - Fixed a problem with static linking.
27        - Added pcredemo. [But later removed - see 7 below.]
28        - Fixed dftables problem and added an option.
29        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
30            HAVE_LONG_LONG.
31        - Added readline support for pcretest.
32        - Added an listing of the option settings after cmake has run.
33    
34    4.  A user submitted a patch to Makefile that makes it easy to create
35        "pcre.dll" under mingw when using Configure/Make. I added stuff to
36        Makefile.am that cause it to include this special target, without
37        affecting anything else. Note that the same mingw target plus all
38        the other distribution libraries and programs are now supported
39        when configuring with CMake (see 6 below) instead of with
40        Configure/Make.
41    
42    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
43        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
44        exported in the Windows port". It has not yet been confirmed that the patch
45        solves the problem, but it does no harm.
46    
47    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
48        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
49        with CMake, and also correct the comment about stack recursion.
50    
51    7.  Remove the automatic building of pcredemo from the ./configure system and
52        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
53        of a program that users should build themselves after PCRE is installed, so
54        building it automatically is not really right. What is more, it gave
55        trouble in some build environments.
56    
57    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
58    
59    
60    Version 7.5 10-Jan-08
61    ---------------------
62    
63    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
64        values in parens when parsing an RE using the C++ wrapper."
65    
66    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
67        Characters greater than 255 were excluded from the class instead of being
68        included.
69    
70    3.  The same bug as (2) above applied to negated POSIX classes such as
71        [:^space:].
72    
73    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
74        defined or documented. It seems to have been a typo for PCRE_STATIC, so
75        I have changed it.
76    
77    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
78        first named subpattern) and a construct such as (?&a) would reference the
79        first named subpattern whose name started with "a" (in other words, the
80        length check was missing). Both these problems are fixed. "Subpattern name
81        expected" is now given for (?&) (a zero-length name), and this patch also
82        makes it give the same error for \k'' (previously it complained that that
83        was a reference to a non-existent subpattern).
84    
85    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
86        this is right because (?- can be followed by option settings as well as by
87        digits. I have, however, made the messages clearer.
88    
89    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
90        than the number used in the conditional) now cause a compile-time error.
91        This is actually not compatible with Perl, which accepts such patterns, but
92        treats the conditional as always being FALSE (as PCRE used to), but it
93        seems to me that giving a diagnostic is better.
94    
95    8.  Change "alphameric" to the more common word "alphanumeric" in comments
96        and messages.
97    
98    9.  Fix two occurrences of "backslash" in comments that should have been
99        "backspace".
100    
101    10. Remove two redundant lines of code that can never be obeyed (their function
102        was moved elsewhere).
103    
104    11. The program that makes PCRE's Unicode character property table had a bug
105        which caused it to generate incorrect table entries for sequences of
106        characters that have the same character type, but are in different scripts.
107        It amalgamated them into a single range, with the script of the first of
108        them. In other words, some characters were in the wrong script. There were
109        thirteen such cases, affecting characters in the following ranges:
110    
111          U+002b0 - U+002c1
112          U+0060c - U+0060d
113          U+0061e - U+00612
114          U+0064b - U+0065e
115          U+0074d - U+0076d
116          U+01800 - U+01805
117          U+01d00 - U+01d77
118          U+01d9b - U+01dbf
119          U+0200b - U+0200f
120          U+030fc - U+030fe
121          U+03260 - U+0327f
122          U+0fb46 - U+0fbb1
123          U+10450 - U+1049d
124    
125    12. The -o option (show only the matching part of a line) for pcregrep was not
126        compatible with GNU grep in that, if there was more than one match in a
127        line, it showed only the first of them. It now behaves in the same way as
128        GNU grep.
129    
130    13. If the -o and -v options were combined for pcregrep, it printed a blank
131        line for every non-matching line. GNU grep prints nothing, and pcregrep now
132        does the same. The return code can be used to tell if there were any
133        non-matching lines.
134    
135    14. Added --file-offsets and --line-offsets to pcregrep.
136    
137    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
138        infinitely looping recursion. The bug was that positive lookaheads were not
139        being skipped when checking for a possible empty match (negative lookaheads
140        and both kinds of lookbehind were skipped).
141    
142    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
143        inclusion of <windows.h> to before rather than after the definition of
144        INVALID_FILE_ATTRIBUTES (patch from David Byron).
145    
146    17. Specifying a possessive quantifier with a specific limit for a Unicode
147        character property caused pcre_compile() to compile bad code, which led at
148        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
149        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
150        caused the error; without that there was no problem.
151    
152    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
153    
154    19. Added --enable-pcretest-libreadline.
155    
156    20. In pcrecpp.cc, the variable 'count' was incremented twice in
157        RE::GlobalReplace(). As a result, the number of replacements returned was
158        double what it should be. I removed one of the increments, but Craig sent a
159        later patch that removed the other one (the right fix) and added unit tests
160        that check the return values (which was not done before).
161    
162    21. Several CMake things:
163    
164        (1) Arranged that, when cmake is used on Unix, the libraries end up with
165            the names libpcre and libpcreposix, not just pcre and pcreposix.
166    
167        (2) The above change means that pcretest and pcregrep are now correctly
168            linked with the newly-built libraries, not previously installed ones.
169    
170        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
171    
172    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
173        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
174        UTF-8 newline character). The key issue is that the pattern starts .*;
175        this means that the match must be either at the beginning, or after a
176        newline. The bug was in the code for advancing after a failed match and
177        checking that the new position followed a newline. It was not taking
178        account of UTF-8 characters correctly.
179    
180    23. PCRE was behaving differently from Perl in the way it recognized POSIX
181        character classes. PCRE was not treating the sequence [:...:] as a
182        character class unless the ... were all letters. Perl, however, seems to
183        allow any characters between [: and :], though of course it rejects as
184        unknown any "names" that contain non-letters, because all the known class
185        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
186        for example, whereas PCRE did not - it did not recognize a POSIX character
187        class. This seemed a bit dangerous, so the code has been changed to be
188        closer to Perl. The behaviour is not identical to Perl, because PCRE will
189        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
190        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
191        Perl does, and where it didn't before.
192    
193    24. Rewrite so as to remove the single use of %n from pcregrep because in some
194        Windows environments %n is disabled by default.
195    
196    
197    Version 7.4 21-Sep-07
198    ---------------------
199    
200    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
201        means that a class such as [\s] counted as "explicit reference to CR or
202        LF". That isn't really right - the whole point of the change was to try to
203        help when there was an actual mention of one of the two characters. So now
204        the change happens only if \r or \n (or a literal CR or LF) character is
205        encountered.
206    
207    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
208        of both had grown to the point where there were only 3 bits left.
209        Fortunately, there was spare space in the data structure, and so I have
210        moved the internal flags into a new 16-bit field to free up more option
211        bits.
212    
213    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
214        but did not set the internal JCHANGED flag - either of these is enough to
215        control the way the "get" function works - but the PCRE_INFO_JCHANGED
216        facility is supposed to tell if (?J) was ever used, so now (?J) at the
217        start sets both bits.
218    
219    4.  Added options (at build time, compile time, exec time) to change \R from
220        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
221    
222    5.  doc/pcresyntax.html was missing from the distribution.
223    
224    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
225        compatibility, even though it is no longer used.
226    
227    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
228        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
229        windows.h file is present (where different names are used). [This was
230        reversed later after testing - see 16 below.]
231    
232    8.  Changed all #include <config.h> to #include "config.h". There were also
233        some further <pcre.h> cases that I changed to "pcre.h".
234    
235    9.  When pcregrep was used with the --colour option, it missed the line ending
236        sequence off the lines that it output.
237    
238    10. It was pointed out to me that arrays of string pointers cause lots of
239        relocations when a shared library is dynamically loaded. A technique of
240        using a single long string with a table of offsets can drastically reduce
241        these. I have refactored PCRE in four places to do this. The result is
242        dramatic:
243    
244          Originally:                          290
245          After changing UCP table:            187
246          After changing error message table:   43
247          After changing table of "verbs"       36
248          After changing table of Posix names   22
249    
250        Thanks to the folks working on Gregex for glib for this insight.
251    
252    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
253        unicode-properties was also set.
254    
255    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
256    
257    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
258        checked only for CRLF.
259    
260    14. Added casts to pcretest.c to avoid compiler warnings.
261    
262    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
263    
264    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
265        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
266        entirely. This removes changes made in 7 above.
267    
268    17. The CMake files have been updated, and there is now more information about
269        building with CMake in the NON-UNIX-USE document.
270    
271    
272    Version 7.3 28-Aug-07
273    ---------------------
274    
275     1. In the rejigging of the build system that eventually resulted in 7.1, the
276        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
277        brackets there is not right, since it causes compilers to look for an
278        installed pcre.h, not the version that is in the source that is being
279        compiled (which of course may be different). I have changed it back to:
280    
281          #include "pcre.h"
282    
283        I have a vague recollection that the change was concerned with compiling in
284        different directories, but in the new build system, that is taken care of
285        by the VPATH setting the Makefile.
286    
287     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
288        when the subject happened to end in the byte 0x85 (e.g. if the last
289        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
290        characters but of course it shouldn't be taken as a newline when it is part
291        of another character. The bug was that, for an unlimited repeat of . in
292        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
293        characters when looking for a newline.
294    
295     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
296    
297     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
298        in debug output.
299    
300     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
301        long printing in the pcrecpp unittest when running under MinGW.
302    
303     6. ESC_K was left out of the EBCDIC table.
304    
305     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
306        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
307        limit also applies to "virtual nesting" when a pattern is recursive, and in
308        this case 1000 isn't so big. I have been able to remove this limit at the
309        expense of backing off one optimization in certain circumstances. Normally,
310        when pcre_exec() would call its internal match() function recursively and
311        immediately return the result unconditionally, it uses a "tail recursion"
312        feature to save stack. However, when a subpattern that can match an empty
313        string has an unlimited repetition quantifier, it no longer makes this
314        optimization. That gives it a stack frame in which to save the data for
315        checking that an empty string has been matched. Previously this was taken
316        from the 1000-entry workspace that had been reserved. So now there is no
317        explicit limit, but more stack is used.
318    
319     8. Applied Daniel's patches to solve problems with the import/export magic
320        syntax that is required for Windows, and which was going wrong for the
321        pcreposix and pcrecpp parts of the library. These were overlooked when this
322        problem was solved for the main library.
323    
324     9. There were some crude static tests to avoid integer overflow when computing
325        the size of patterns that contain repeated groups with explicit upper
326        limits. As the maximum quantifier is 65535, the maximum group length was
327        set at 30,000 so that the product of these two numbers did not overflow a
328        32-bit integer. However, it turns out that people want to use groups that
329        are longer than 30,000 bytes (though not repeat them that many times).
330        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
331        made it possible to implement the integer overflow checks in a much more
332        dynamic way, which I have now done. The artificial limitation on group
333        length has been removed - we now have only the limit on the total length of
334        the compiled pattern, which depends on the LINK_SIZE setting.
335    
336    10. Fixed a bug in the documentation for get/copy named substring when
337        duplicate names are permitted. If none of the named substrings are set, the
338        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
339        empty string.
340    
341    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
342        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
343        because the ] is interpreted as the first data character and the
344        terminating ] is not found. PCRE has been made compatible with Perl in this
345        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
346        cause memory overwriting.
347    
348    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
349        string has been matched (to stop an infinite loop). It was not recognizing
350        a conditional subpattern that could match an empty string if that
351        subpattern was within another subpattern. For example, it looped when
352        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
353        condition was not nested. This bug has been fixed.
354    
355    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
356        past the start of the subject in the presence of bytes with the top bit
357        set, for example "\x8aBCD".
358    
359    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
360        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
361    
362    14. Optimized (?!) to (*FAIL).
363    
364    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
365        This restricts code points to be within the range 0 to 0x10FFFF, excluding
366        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
367        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
368        does: it's just the validity check that is more restrictive.
369    
370    16. Inserted checks for integer overflows during escape sequence (backslash)
371        processing, and also fixed erroneous offset values for syntax errors during
372        backslash processing.
373    
374    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
375        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
376    
377    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
378        caused an overrun.
379    
380    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
381        something other than just ASCII characters) inside a group that had an
382        unlimited repeat caused a loop at compile time (while checking to see
383        whether the group could match an empty string).
384    
385    20. Debugging a pattern containing \p or \P could cause a crash. For example,
386        [\P{Any}] did so. (Error in the code for printing property names.)
387    
388    21. An orphan \E inside a character class could cause a crash.
389    
390    22. A repeated capturing bracket such as (A)? could cause a wild memory
391        reference during compilation.
392    
393    23. There are several functions in pcre_compile() that scan along a compiled
394        expression for various reasons (e.g. to see if it's fixed length for look
395        behind). There were bugs in these functions when a repeated \p or \P was
396        present in the pattern. These operators have additional parameters compared
397        with \d, etc, and these were not being taken into account when moving along
398        the compiled data. Specifically:
399    
400        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
401            length.
402    
403        (b) An item such as \pL+ within a repeated group could cause crashes or
404            loops.
405    
406        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
407            "reference to non-existent subpattern" error.
408    
409        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
410    
411    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
412        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
413    
414    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
415    
416    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
417        character were causing crashes (broken optimization).
418    
419    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
420        \p or \P) caused a compile-time loop.
421    
422    28. More problems have arisen in unanchored patterns when CRLF is a valid line
423        break. For example, the unstudied pattern [\r\n]A does not match the string
424        "\r\nA" because change 7.0/46 below moves the current point on by two
425        characters after failing to match at the start. However, the pattern \nA
426        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
427        the same is true. There doesn't seem any very clean way out of this, but
428        what I have chosen to do makes the common cases work: PCRE now takes note
429        of whether there can be an explicit match for \r or \n anywhere in the
430        pattern, and if so, 7.0/46 no longer applies. As part of this change,
431        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
432        pattern has explicit CR or LF references.
433    
434    29. Added (*CR) etc for changing newline setting at start of pattern.
435    
436    
437    Version 7.2 19-Jun-07
438    ---------------------
439    
440     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
441        which is apparently normally available under Windows.
442    
443     2. Re-jig the pcregrep tests with different newline settings in an attempt
444        to make them independent of the local environment's newline setting.
445    
446     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
447    
448     4. Some of the "internals" tests were previously cut out when the link size
449        was not 2, because the output contained actual offsets. The recent new
450        "Z" feature of pcretest means that these can be cut out, making the tests
451        usable with all link sizes.
452    
453     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
454        stack recursion. This gives a massive performance boost under BSD, but just
455        a small improvement under Linux. However, it saves one field in the frame
456        in all cases.
457    
458     6. Added more features from the forthcoming Perl 5.10:
459    
460        (a) (?-n) (where n is a string of digits) is a relative subroutine or
461            recursion call. It refers to the nth most recently opened parentheses.
462    
463        (b) (?+n) is also a relative subroutine call; it refers to the nth next
464            to be opened parentheses.
465    
466        (c) Conditions that refer to capturing parentheses can be specified
467            relatively, for example, (?(-2)... or (?(+3)...
468    
469        (d) \K resets the start of the current match so that everything before
470            is not part of it.
471    
472        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
473    
474        (f) \g{name} is another synonym - part of Perl 5.10's unification of
475            reference syntax.
476    
477        (g) (?| introduces a group in which the numbering of parentheses in each
478            alternative starts with the same number.
479    
480        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
481    
482     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
483        PCRE_INFO_JCHANGED.
484    
485     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
486        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
487        for detecting groups that can match an empty string.
488    
489     9. A pattern with a very large number of alternatives (more than several
490        hundred) was running out of internal workspace during the pre-compile
491        phase, where pcre_compile() figures out how much memory will be needed. A
492        bit of new cunning has reduced the workspace needed for groups with
493        alternatives. The 1000-alternative test pattern now uses 12 bytes of
494        workspace instead of running out of the 4096 that are available.
495    
496    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
497    
498    11. Applied patch from Google to remove an optimization that didn't quite work.
499        The report of the bug said:
500    
501          pcrecpp::RE("a*").FullMatch("aaa") matches, while
502          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
503          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
504    
505    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
506        it matched the wrong number of bytes.
507    
508    
509    Version 7.1 24-Apr-07
510    ---------------------
511    
512     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
513        that is more "standard", making use of automake and other Autotools. There
514        is some re-arrangement of the files and adjustment of comments consequent
515        on this.
516    
517     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
518        for recursive directory scanning broke on some systems because the files
519        are not scanned in any specific order and on different systems the order
520        was different. A call to "sort" has been inserted into RunGrepTest for the
521        approprate test as a short-term fix. In the longer term there may be an
522        alternative.
523    
524     3. I had an email from Eric Raymond about problems translating some of PCRE's
525        man pages to HTML (despite the fact that I distribute HTML pages, some
526        people do their own conversions for various reasons). The problems
527        concerned the use of low-level troff macros .br and .in. I have therefore
528        removed all such uses from the man pages (some were redundant, some could
529        be replaced by .nf/.fi pairs). The 132html script that I use to generate
530        HTML has been updated to handle .nf/.fi and to complain if it encounters
531        .br or .in.
532    
533     4. Updated comments in configure.ac that get placed in config.h.in and also
534        arranged for config.h to be included in the distribution, with the name
535        config.h.generic, for the benefit of those who have to compile without
536        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
537    
538     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
539        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
540        makevp.bat for the current PCRE, using the additional files
541        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
542    
543     6. A Windows user reported a minor discrepancy with test 2, which turned out
544        to be caused by a trailing space on an input line that had got lost in his
545        copy. The trailing space was an accident, so I've just removed it.
546    
547     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
548        that is needed.
549    
550     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
551        as "const" (a) because they are and (b) because it helps the PHP
552        maintainers who have recently made a script to detect big data structures
553        in the php code that should be moved to the .rodata section. I remembered
554        to update Builducptable as well, so it won't revert if ucptable.h is ever
555        re-created.
556    
557     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
558        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
559        order to be able to cut out the UTF-8 tables in the latter when UTF-8
560        support is not required. This saves 1.5-2K of code, which is important in
561        some applications.
562    
563        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
564        so as not to refer to the tables, even though these functions will never be
565        called when UTF-8 support is disabled. Otherwise there are problems with a
566        shared library.
567    
568    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
569    
570        (a) It was defining its arguments as char * instead of void *.
571    
572        (b) It was assuming that all moves were upwards in memory; this was true
573            a long time ago when I wrote it, but is no longer the case.
574    
575        The emulated memove() is provided for those environments that have neither
576        memmove() nor bcopy(). I didn't think anyone used it these days, but that
577        is clearly not the case, as these two bugs were recently reported.
578    
579    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
580        and Detrail to create the HTML documentation, the .txt form of the man
581        pages, and it removes trailing spaces from listed files. It also creates
582        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
583        case, it wraps all the #defines with #ifndefs. This script should be run
584        before "make dist".
585    
586    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
587        with Unicode property support.
588    
589        (a) For a maximizing quantifier, if the two different cases of the
590            character were of different lengths in their UTF-8 codings (there are
591            some cases like this - I found 11), and the matching function had to
592            back up over a mixture of the two cases, it incorrectly assumed they
593            were both the same length.
594    
595        (b) When PCRE was configured to use the heap rather than the stack for
596            recursion during matching, it was not correctly preserving the data for
597            the other case of a UTF-8 character when checking ahead for a match
598            while processing a minimizing repeat. If the check also involved
599            matching a wide character, but failed, corruption could cause an
600            erroneous result when trying to check for a repeat of the original
601            character.
602    
603    13. Some tidying changes to the testing mechanism:
604    
605        (a) The RunTest script now detects the internal link size and whether there
606            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
607            values substituted by "configure". (The RunGrepTest script already did
608            this for UTF-8.) The configure.ac script no longer substitutes the
609            relevant variables.
610    
611        (b) The debugging options /B and /D in pcretest show the compiled bytecode
612            with length and offset values. This means that the output is different
613            for different internal link sizes. Test 2 is skipped for link sizes
614            other than 2 because of this, bypassing the problem. Unfortunately,
615            there was also a test in test 3 (the locale tests) that used /B and
616            failed for link sizes other than 2. Rather than cut the whole test out,
617            I have added a new /Z option to pcretest that replaces the length and
618            offset values with spaces. This is now used to make test 3 independent
619            of link size. (Test 2 will be tidied up later.)
620    
621    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
622        segmentation fault instead of returning the appropriate error message.
623    
624    15. In multiline mode when the newline sequence was set to "any", the pattern
625        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
626        This doesn't seem right; it now treats the CRLF combination as the line
627        ending, and so does not match in that case. It's only a pattern such as ^$
628        that would hit this one: something like ^ABC$ would have failed after \r
629        and then tried again after \r\n.
630    
631    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
632        in an attempt to make files that differ only in their line terminators
633        compare equal. This works on Linux.
634    
635    17. Under certain error circumstances pcregrep might try to free random memory
636        as it exited. This is now fixed, thanks to valgrind.
637    
638    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
639        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
640        was because its rules for how to advance for /g after matching an empty
641        string at the end of a line did not allow for this case. They now check for
642        it specially.
643    
644    20. pcretest is supposed to handle patterns and data of any length, by
645        extending its buffers when necessary. It was getting this wrong when the
646        buffer for a data line had to be extended.
647    
648    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
649        CRLF as a newline sequence.
650    
651    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
652        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
653        I have nevertheless tidied it up.
654    
655    23. Added some casts to kill warnings from HP-UX ia64 compiler.
656    
657    24. Added a man page for pcre-config.
658    
659    
660    Version 7.0 19-Dec-06
661    ---------------------
662    
663     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
664        moving to gcc 4.1.1.
665    
666     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
667        sys/time.h, which is documented as needed for this function. It doesn't
668        seem to matter on Linux, but it showed up on some releases of OS X.
669    
670     3. It seems that there are systems where bytes whose values are greater than
671        127 match isprint() in the "C" locale. The "C" locale should be the
672        default when a C program starts up. In most systems, only ASCII printing
673        characters match isprint(). This difference caused the output from pcretest
674        to vary, making some of the tests fail. I have changed pcretest so that:
675    
676        (a) When it is outputting text in the compiled version of a pattern, bytes
677            other than 32-126 are always shown as hex escapes.
678    
679        (b) When it is outputting text that is a matched part of a subject string,
680            it does the same, unless a different locale has been set for the match
681            (using the /L modifier). In this case, it uses isprint() to decide.
682    
683     4. Fixed a major bug that caused incorrect computation of the amount of memory
684        required for a compiled pattern when options that changed within the
685        pattern affected the logic of the preliminary scan that determines the
686        length. The relevant options are -x, and -i in UTF-8 mode. The result was
687        that the computed length was too small. The symptoms of this bug were
688        either the PCRE error "internal error: code overflow" from pcre_compile(),
689        or a glibc crash with a message such as "pcretest: free(): invalid next
690        size (fast)". Examples of patterns that provoked this bug (shown in
691        pcretest format) are:
692    
693          /(?-x: )/x
694          /(?x)(?-x: \s*#\s*)/
695          /((?i)[\x{c0}])/8
696          /(?i:[\x{c0}])/8
697    
698        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
699        is now done differently.
700    
701     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
702        wrapper classes; (b) implement a new function in the C++ scanner that is
703        more efficient than the old way of doing things because it avoids levels of
704        recursion in the regex matching; (c) add a paragraph to the documentation
705        for the FullMatch() function.
706    
707     6. The escape sequence \n was being treated as whatever was defined as
708        "newline". Not only was this contrary to the documentation, which states
709        that \n is character 10 (hex 0A), but it also went horribly wrong when
710        "newline" was defined as CRLF. This has been fixed.
711    
712     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
713        was being set to -1 for the "end of line" case (supposedly a value that no
714        character can have). Though this value is never used (the check for end of
715        line is "zero bytes in current character"), it caused compiler complaints.
716        I've changed it to 0xffffffff.
717    
718     8. In pcre_version.c, the version string was being built by a sequence of
719        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
720        string (as it is for production releases) called a macro with an empty
721        argument. The C standard says the result of this is undefined. The gcc
722        compiler treats it as an empty string (which was what was wanted) but it is
723        reported that Visual C gives an error. The source has been hacked around to
724        avoid this problem.
725    
726     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
727        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
728        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
729        of them did).
730    
731    10. Originally, pcretest opened its input and output without "b"; then I was
732        told that "b" was needed in some environments, so it was added for release
733        5.0 to both the input and output. (It makes no difference on Unix-like
734        systems.) Later I was told that it is wrong for the input on Windows. I've
735        now abstracted the modes into two macros, to make it easier to fiddle with
736        them, and removed "b" from the input mode under Windows.
737    
738    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
739    
740    12. Added -help and --help to pcretest as an official way of being reminded
741        of the options.
742    
743    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
744        and pcrecpp.cc because they annoy compilers at high warning levels.
745    
746    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
747    
748    15. Fixed an occurrence of == in configure.ac that should have been = (shell
749        scripts are not C programs :-) and which was not noticed because it works
750        on Linux.
751    
752    16. pcretest is supposed to handle any length of pattern and data line (as one
753        line or as a continued sequence of lines) by extending its input buffer if
754        necessary. This feature was broken for very long pattern lines, leading to
755        a string of junk being passed to pcre_compile() if the pattern was longer
756        than about 50K.
757    
758    17. I have done a major re-factoring of the way pcre_compile() computes the
759        amount of memory needed for a compiled pattern. Previously, there was code
760        that made a preliminary scan of the pattern in order to do this. That was
761        OK when PCRE was new, but as the facilities have expanded, it has become
762        harder and harder to keep it in step with the real compile phase, and there
763        have been a number of bugs (see for example, 4 above). I have now found a
764        cunning way of running the real compile function in a "fake" mode that
765        enables it to compute how much memory it would need, while actually only
766        ever using a few hundred bytes of working memory and without too many
767        tests of the mode. This should make future maintenance and development
768        easier. A side effect of this work is that the limit of 200 on the nesting
769        depth of parentheses has been removed (though this was never a serious
770        limitation, I suspect). However, there is a downside: pcre_compile() now
771        runs more slowly than before (30% or more, depending on the pattern). I
772        hope this isn't a big issue. There is no effect on runtime performance.
773    
774    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
775        newline (only possible for the last line of a file) and it was a
776        pattern that set a locale (followed by /Lsomething), pcretest crashed.
777    
778    19. Added additional timing features to pcretest. (1) The -tm option now times
779        matching only, not compiling. (2) Both -t and -tm can be followed, as a
780        separate command line item, by a number that specifies the number of
781        repeats to use when timing. The default is 50000; this gives better
782        precision, but takes uncomfortably long for very large patterns.
783    
784    20. Extended pcre_study() to be more clever in cases where a branch of a
785        subpattern has no definite first character. For example, (a*|b*)[cd] would
786        previously give no result from pcre_study(). Now it recognizes that the
787        first character must be a, b, c, or d.
788    
789    21. There was an incorrect error "recursive call could loop indefinitely" if
790        a subpattern (or the entire pattern) that was being tested for matching an
791        empty string contained only one non-empty item after a nested subpattern.
792        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
793        incorrectly, because the \d was being skipped in the check.
794    
795    22. The pcretest program now has a new pattern option /B and a command line
796        option -b, which is equivalent to adding /B to every pattern. This causes
797        it to show the compiled bytecode, without the additional information that
798        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
799        is the same as /B/I).
800    
801    23. A new optimization is now able automatically to treat some sequences such
802        as a*b as a*+b. More specifically, if something simple (such as a character
803        or a simple class like \d) has an unlimited quantifier, and is followed by
804        something that cannot possibly match the quantified thing, the quantifier
805        is automatically "possessified".
806    
807    24. A recursive reference to a subpattern whose number was greater than 39
808        went wrong under certain circumstances in UTF-8 mode. This bug could also
809        have affected the operation of pcre_study().
810    
811    25. Realized that a little bit of performance could be had by replacing
812        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
813    
814    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
815    
816    27. Possessive quantifiers such as a++ were previously implemented by turning
817        them into atomic groups such as ($>a+). Now they have their own opcodes,
818        which improves performance. This includes the automatically created ones
819        from 23 above.
820    
821    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
822        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
823        the first matched character to be a colon. This applied both to named and
824        numbered groups.
825    
826    29. The ucpinternal.h header file was missing its idempotency #ifdef.
827    
828    30. I was sent a "project" file called libpcre.a.dev which I understand makes
829        building PCRE on Windows easier, so I have included it in the distribution.
830    
831    31. There is now a check in pcretest against a ridiculously large number being
832        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
833        loop, the loop is abandoned.
834    
835    32. Forward references to subpatterns in conditions such as (?(2)...) where
836        subpattern 2 is defined later cause pcre_compile() to search forwards in
837        the pattern for the relevant set of parentheses. This search went wrong
838        when there were unescaped parentheses in a character class, parentheses
839        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
840    
841    33. "Subroutine" calls and backreferences were previously restricted to
842        referencing subpatterns earlier in the regex. This restriction has now
843        been removed.
844    
845    34. Added a number of extra features that are going to be in Perl 5.10. On the
846        whole, these are just syntactic alternatives for features that PCRE had
847        previously implemented using the Python syntax or my own invention. The
848        other formats are all retained for compatibility.
849    
850        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
851            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
852            also .NET compatible.
853    
854        (b) A recursion or subroutine call to a named group can now be defined as
855            (?&name) as well as (?P>name).
856    
857        (c) A backreference to a named group can now be defined as \k<name> or
858            \k'name' as well as (?P=name). The new forms, as well as being in Perl
859            5.10, are also .NET compatible.
860    
861        (d) A conditional reference to a named group can now use the syntax
862            (?(<name>) or (?('name') as well as (?(name).
863    
864        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
865            groups (named and numbered) that are never evaluated inline, but can be
866            called as "subroutines" from elsewhere. In effect, the DEFINE condition
867            is always false. There may be only one alternative in such a group.
868    
869        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
870            as the simple (?(R). The condition is true only if the most recent
871            recursion is that of the given number or name. It does not search out
872            through the entire recursion stack.
873    
874        (g) The escape \gN or \g{N} has been added, where N is a positive or
875            negative number, specifying an absolute or relative reference.
876    
877    35. Tidied to get rid of some further signed/unsigned compiler warnings and
878        some "unreachable code" warnings.
879    
880    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
881        things, this adds five new scripts.
882    
883    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
884        There were also incompatibilities regarding the handling of \Q..\E inside
885        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
886        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
887    
888    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
889        matches an empty string, and forcibly breaks the loop. There were bugs in
890        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
891        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
892        separate and independent bugs (that affected different cases) have been
893        fixed.
894    
895    39. Refactored the code to abolish the use of different opcodes for small
896        capturing bracket numbers. This is a tidy that I avoided doing when I
897        removed the limit on the number of capturing brackets for 3.5 back in 2001.
898        The new approach is not only tidier, it makes it possible to reduce the
899        memory needed to fix the previous bug (38).
900    
901    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
902        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
903        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
904        mode.
905    
906    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
907        report.
908    
909    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
910        copy construction and assignment in the C++ wrapper.
911    
912    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
913        couple of bugs that could have given wrong results in the "--newline=crlf"
914        case.
915    
916    44. Added a number of casts and did some reorganization of signed/unsigned int
917        variables following suggestions from Dair Grant. Also renamed the variable
918        "this" as "item" because it is a C++ keyword.
919    
920    45. Arranged for dftables to add
921    
922          #include "pcre_internal.h"
923    
924        to pcre_chartables.c because without it, gcc 4.x may remove the array
925        definition from the final binary if PCRE is built into a static library and
926        dead code stripping is activated.
927    
928    46. For an unanchored pattern, if a match attempt fails at the start of a
929        newline sequence, and the newline setting is CRLF or ANY, and the next two
930        characters are CRLF, advance by two characters instead of one.
931    
932    
933    Version 6.7 04-Jul-06
934    ---------------------
935    
936     1. In order to handle tests when input lines are enormously long, pcretest has
937        been re-factored so that it automatically extends its buffers when
938        necessary. The code is crude, but this _is_ just a test program. The
939        default size has been increased from 32K to 50K.
940    
941     2. The code in pcre_study() was using the value of the re argument before
942        testing it for NULL. (Of course, in any sensible call of the function, it
943        won't be NULL.)
944    
945     3. The memmove() emulation function in pcre_internal.h, which is used on
946        systems that lack both memmove() and bcopy() - that is, hardly ever -
947        was missing a "static" storage class specifier.
948    
949     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
950        containing an extended class (one that cannot be represented by a bitmap
951        because it contains high-valued characters or Unicode property items, e.g.
952        [\pZ]). Almost always one would set UTF-8 mode when processing such a
953        pattern, but PCRE should not loop if you do not (it no longer does).
954        [Detail: two cases were found: (a) a repeated subpattern containing an
955        extended class; (b) a recursive reference to a subpattern that followed a
956        previous extended class. It wasn't skipping over the extended class
957        correctly when UTF-8 mode was not set.]
958    
959     5. A negated single-character class was not being recognized as fixed-length
960        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
961        compile error "lookbehind assertion is not fixed length".
962    
963     6. The RunPerlTest auxiliary script was showing an unexpected difference
964        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
965        write a Perl script that can interpret lines of an input file either as
966        byte characters or as UTF-8, which is what "perltest" was being required to
967        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
968        can't do is switch easily at run time between having the "use utf8;" pragma
969        or not. In the end, I fudged it by using the RunPerlTest script to insert
970        "use utf8;" explicitly for the UTF-8 tests.
971    
972     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
973        the end of the subject string, contrary to the documentation and to what
974        Perl does. This was true of both matching functions. Now it matches only at
975        the start of the subject and immediately after *internal* newlines.
976    
977     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
978        a pointer to an int instead of a pointer to an unsigned long int. This
979        caused problems on 64-bit systems.
980    
981     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
982        instance of the 'standard' template library not being so standard".
983    
984    10. There was no check on the number of named subpatterns nor the maximum
985        length of a subpattern name. The product of these values is used to compute
986        the size of the memory block for a compiled pattern. By supplying a very
987        long subpattern name and a large number of named subpatterns, the size
988        computation could be caused to overflow. This is now prevented by limiting
989        the length of names to 32 characters, and the number of named subpatterns
990        to 10,000.
991    
992    11. Subpatterns that are repeated with specific counts have to be replicated in
993        the compiled pattern. The size of memory for this was computed from the
994        length of the subpattern and the repeat count. The latter is limited to
995        65535, but there was no limit on the former, meaning that integer overflow
996        could in principle occur. The compiled length of a repeated subpattern is
997        now limited to 30,000 bytes in order to prevent this.
998    
999    12. Added the optional facility to have named substrings with the same name.
1000    
1001    13. Added the ability to use a named substring as a condition, using the
1002        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1003        are numbers (not recommended). Forward references are permitted.
1004    
1005    14. Added forward references in named backreferences (if you see what I mean).
1006    
1007    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1008        pattern could run off the end of the subject. For example, the pattern
1009        "(?s)(.{1,5})"8 did this with the subject "ab".
1010    
1011    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1012        PCRE_CASELESS was set when matching characters that were quantified with ?
1013        or *.
1014    
1015    17. A character class other than a single negated character that had a minimum
1016        but no maximum quantifier - for example [ab]{6,} - was not handled
1017        correctly by pce_dfa_exec(). It would match only one character.
1018    
1019    18. A valid (though odd) pattern that looked like a POSIX character
1020        class but used an invalid character after [ (for example [[,abc,]]) caused
1021        pcre_compile() to give the error "Failed: internal error: code overflow" or
1022        in some cases to crash with a glibc free() error. This could even happen if
1023        the pattern terminated after [[ but there just happened to be a sequence of
1024        letters, a binary zero, and a closing ] in the memory that followed.
1025    
1026    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1027        over the years. Originally (before any Unicode support), just the bottom 8
1028        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1029        output from "man perlunicode" includes this:
1030    
1031          The regular expression compiler produces polymorphic opcodes.  That
1032          is, the pattern adapts to the data and automatically switches to
1033          the Unicode character scheme when presented with Unicode data--or
1034          instead uses a traditional byte scheme when presented with byte
1035          data.
1036    
1037        Sadly, a wide octal escape does not cause a switch, and in a string with
1038        no other multibyte characters, these octal escapes are treated as before.
1039        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1040        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1041        Unicode string.
1042    
1043        I have not perpetrated such confusion in PCRE. Up till now, it took just
1044        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1045        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1046        translate to the appropriate multibyte character.
1047    
1048    29. Applied some refactoring to reduce the number of warnings from Microsoft
1049        and Borland compilers. This has included removing the fudge introduced
1050        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1051        a warning about an unused variable.
1052    
1053    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1054        characters since release 4.0, because Perl (from release 5.004) does not.
1055        [Or at least, is documented not to: some releases seem to be in conflict
1056        with the documentation.] However, when a pattern was studied with
1057        pcre_study() and all its branches started with \s, PCRE still included VT
1058        as a possible starting character. Of course, this did no harm; it just
1059        caused an unnecessary match attempt.
1060    
1061    22. Removed a now-redundant internal flag bit that recorded the fact that case
1062        dependency changed within the pattern. This was once needed for "required
1063        byte" processing, but is no longer used. This recovers a now-scarce options
1064        bit. Also moved the least significant internal flag bit to the most-
1065        significant bit of the word, which was not previously used (hangover from
1066        the days when it was an int rather than a uint) to free up another bit for
1067        the future.
1068    
1069    23. Added support for CRLF line endings as well as CR and LF. As well as the
1070        default being selectable at build time, it can now be changed at runtime
1071        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1072        specify that it is scanning data with non-default line endings.
1073    
1074    24. Changed the definition of CXXLINK to make it agree with the definition of
1075        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1076    
1077    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1078        recursions. This makes a big different to stack usage for some patterns.
1079    
1080    26. If a subpattern containing a named recursion or subroutine reference such
1081        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1082        the space required for the compiled pattern went wrong and gave too small a
1083        value. Depending on the environment, this could lead to "Failed: internal
1084        error: code overflow at offset 49" or "glibc detected double free or
1085        corruption" errors.
1086    
1087    27. Applied patches from Google (a) to support the new newline modes and (b) to
1088        advance over multibyte UTF-8 characters in GlobalReplace.
1089    
1090    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1091        difference for some implementation of PCRE in some Windows version.
1092    
1093    29. Added some extra testing facilities to pcretest:
1094    
1095        \q<number>   in a data line sets the "match limit" value
1096        \Q<number>   in a data line sets the "match recursion limt" value
1097        -S <number>  sets the stack size, where <number> is in megabytes
1098    
1099        The -S option isn't available for Windows.
1100    
1101    
1102    Version 6.6 06-Feb-06
1103    ---------------------
1104    
1105     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1106        in pcreposix.h. I have copied the definition from pcre.h.
1107    
1108     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1109        because pcre.h is no longer a built file.
1110    
1111     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1112        not normally included in the compiled code.
1113    
1114    
1115    Version 6.5 01-Feb-06
1116    ---------------------
1117    
1118     1. When using the partial match feature with pcre_dfa_exec(), it was not
1119        anchoring the second and subsequent partial matches at the new starting
1120        point. This could lead to incorrect results. For example, with the pattern
1121        /1234/, partially matching against "123" and then "a4" gave a match.
1122    
1123     2. Changes to pcregrep:
1124    
1125        (a) All non-match returns from pcre_exec() were being treated as failures
1126            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1127            error message is output. Some extra information is given for the
1128            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1129            probably the only errors that are likely to be caused by users (by
1130            specifying a regex that has nested indefinite repeats, for instance).
1131            If there are more than 20 of these errors, pcregrep is abandoned.
1132    
1133        (b) A binary zero was treated as data while matching, but terminated the
1134            output line if it was written out. This has been fixed: binary zeroes
1135            are now no different to any other data bytes.
1136    
1137        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1138            used to set a locale for matching. The --locale=xxxx long option has
1139            been added (no short equivalent) to specify a locale explicitly on the
1140            pcregrep command, overriding the environment variables.
1141    
1142        (d) When -B was used with -n, some line numbers in the output were one less
1143            than they should have been.
1144    
1145        (e) Added the -o (--only-matching) option.
1146    
1147        (f) If -A or -C was used with -c (count only), some lines of context were
1148            accidentally printed for the final match.
1149    
1150        (g) Added the -H (--with-filename) option.
1151    
1152        (h) The combination of options -rh failed to suppress file names for files
1153            that were found from directory arguments.
1154    
1155        (i) Added the -D (--devices) and -d (--directories) options.
1156    
1157        (j) Added the -F (--fixed-strings) option.
1158    
1159        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1160    
1161        (l) Added the --colo(u)r option.
1162    
1163        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1164            is not present by default.
1165    
1166     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1167        items such as (?R) or (?1), when the recursion could match a number of
1168        alternatives. If it matched one of the alternatives, but subsequently,
1169        outside the recursion, there was a failure, the code tried to back up into
1170        the recursion. However, because of the way PCRE is implemented, this is not
1171        possible, and the result was an incorrect result from the match.
1172    
1173        In order to prevent this happening, the specification of recursion has
1174        been changed so that all such subpatterns are automatically treated as
1175        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1176    
1177     4. I had overlooked the fact that, in some locales, there are characters for
1178        which isalpha() is true but neither isupper() nor islower() are true. In
1179        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1180        and ordfeminine) are like this. This affected the treatment of \w and \W
1181        when they appeared in character classes, but not when they appeared outside
1182        a character class. The bit map for "word" characters is now created
1183        separately from the results of isalnum() instead of just taking it from the
1184        upper, lower, and digit maps. (Plus the underscore character, of course.)
1185    
1186     5. The above bug also affected the handling of POSIX character classes such as
1187        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1188        permanent tables. Instead, the bit maps for such a class were previously
1189        created as the appropriate unions of the upper, lower, and digit bitmaps.
1190        Now they are created by subtraction from the [[:word:]] class, which has
1191        its own bitmap.
1192    
1193     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1194        It is created by subtracting the vertical space characters (\x09, \x0a,
1195        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1196        subtraction was done in the overall bitmap for a character class, meaning
1197        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1198        be recognized. This bug has been fixed.
1199    
1200     7. Patches from the folks at Google:
1201    
1202          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1203          real life, but is still worth protecting against".
1204    
1205          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1206          regular expressions".
1207    
1208          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1209          have it.
1210    
1211          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1212          "configure" and the latter not, in order to fix a problem somebody had
1213          with compiling the Arg class on HP-UX.
1214    
1215          (e) Improve the error-handling of the C++ wrapper a little bit.
1216    
1217          (f) New tests for checking recursion limiting.
1218    
1219     8. The pcre_memmove() function, which is used only if the environment does not
1220        have a standard memmove() function (and is therefore rarely compiled),
1221        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1222        returning a result (though PCRE never actually uses the result).
1223    
1224     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1225        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1226        returned instead of calling malloc() with an overflowing number that would
1227        most likely cause subsequent chaos.
1228    
1229    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1230    
1231    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1232        with this option is matched, the nmatch and pmatch options of regexec() are
1233        ignored.
1234    
1235    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1236        provided in case anyone wants to the the POSIX interface with UTF-8
1237        strings.
1238    
1239    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1240        C++ linking (needed for some HP-UX environments).
1241    
1242    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1243        (unused parameter) and in the pcre_printint() function (omitted "default"
1244        switch label when the default is to do nothing).
1245    
1246    15. Added some code to make it possible, when PCRE is compiled as a C++
1247        library, to replace subject pointers for pcre_exec() with a smart pointer
1248        class, thus making it possible to process discontinuous strings.
1249    
1250    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1251        much the same function. They were added by different people who were trying
1252        to make PCRE easy to compile on non-Unix systems. It has been suggested
1253        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1254        for compiling on Windows systems. I have therefore replaced it with
1255        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1256        defaults to "extern" for C or "extern C" for C++, which works fine on
1257        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1258        SCOPE with something explicit in config.h. In addition:
1259    
1260        (a) pcreposix.h still had just "extern" instead of either of these macros;
1261            I have replaced it with PCRE_DATA_SCOPE.
1262    
1263        (b) Functions such as _pcre_xclass(), which are internal to the library,
1264            but external in the C sense, all had PCRE_EXPORT in their definitions.
1265            This is apparently wrong for the Windows case, so I have removed it.
1266            (It makes no difference on Unix-like systems.)
1267    
1268    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1269        of recursive calls to match(). This is different to MATCH_LIMIT because
1270        that limits the total number of calls to match(), not all of which increase
1271        the depth of recursion. Limiting the recursion depth limits the amount of
1272        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1273        when PCRE is compiled, and changed at run time. A patch from Google adds
1274        this functionality to the C++ interface.
1275    
1276    18. Changes to the handling of Unicode character properties:
1277    
1278        (a) Updated the table to Unicode 4.1.0.
1279    
1280        (b) Recognize characters that are not in the table as "Cn" (undefined).
1281    
1282        (c) I revised the way the table is implemented to a much improved format
1283            which includes recognition of ranges. It now supports the ranges that
1284            are defined in UnicodeData.txt, and it also amalgamates other
1285            characters into ranges. This has reduced the number of entries in the
1286            table from around 16,000 to around 3,000, thus reducing its size
1287            considerably. I realized I did not need to use a tree structure after
1288            all - a binary chop search is just as efficient. Having reduced the
1289            number of entries, I extended their size from 6 bytes to 8 bytes to
1290            allow for more data.
1291    
1292        (d) Added support for Unicode script names via properties such as \p{Han}.
1293    
1294    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1295        matching that character.
1296    
1297    20. When matching a repeated Unicode property with a minimum greater than zero,
1298        (for example \pL{2,}), PCRE could look past the end of the subject if it
1299        reached it while seeking the minimum number of characters. This could
1300        happen only if some of the characters were more than one byte long, because
1301        there is a check for at least the minimum number of bytes.
1302    
1303    21. Refactored the implementation of \p and \P so as to be more general, to
1304        allow for more different types of property in future. This has changed the
1305        compiled form incompatibly. Anybody with saved compiled patterns that use
1306        \p or \P will have to recompile them.
1307    
1308    22. Added "Any" and "L&" to the supported property types.
1309    
1310    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1311        but give a compile time error if the value is greater than 0xff.
1312    
1313    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1314        accidentally not being installed or uninstalled.
1315    
1316    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1317        made were to insert the current release number. This seemed silly, because
1318        it made things harder for people building PCRE on systems that don't run
1319        "configure". I have turned pcre.h into a distributed file, no longer built
1320        by "configure", with the version identification directly included. There is
1321        no longer a pcre.h.in file.
1322    
1323        However, this change necessitated a change to the pcre-config script as
1324        well. It is built from pcre-config.in, and one of the substitutions was the
1325        release number. I have updated configure.ac so that ./configure now finds
1326        the release number by grepping pcre.h.
1327    
1328    26. Added the ability to run the tests under valgrind.
1329    
1330    
1331    Version 6.4 05-Sep-05
1332    ---------------------
1333    
1334     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1335        "--" to be printed when multiple files were scanned, even when none of the
1336        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1337        consider it to be a bug, and have restored the previous behaviour.
1338    
1339     2. A couple of code tidies to get rid of compiler warnings.
1340    
1341     3. The pcretest program used to cheat by referring to symbols in the library
1342        whose names begin with _pcre_. These are internal symbols that are not
1343        really supposed to be visible externally, and in some environments it is
1344        possible to suppress them. The cheating is now confined to including
1345        certain files from the library's source, which is a bit cleaner.
1346    
1347     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1348        file's purpose clearer.
1349    
1350     5. Reorganized pcre_ucp_findchar().
1351    
1352    
1353    Version 6.3 15-Aug-05
1354    ---------------------
1355    
1356     1. The file libpcre.pc.in did not have general read permission in the tarball.
1357    
1358     2. There were some problems when building without C++ support:
1359    
1360        (a) If C++ support was not built, "make install" and "make test" still
1361            tried to test it.
1362    
1363        (b) There were problems when the value of CXX was explicitly set. Some
1364            changes have been made to try to fix these, and ...
1365    
1366        (c) --disable-cpp can now be used to explicitly disable C++ support.
1367    
1368        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1369            backslash in a target when C++ was disabled. This confuses some
1370            versions of "make", apparently. Using an intermediate variable solves
1371            this. (Same for CPP_LOBJ.)
1372    
1373     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1374        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1375        necessary on certain architectures.
1376    
1377     4. Added a setting of -export-symbols-regex to the link command to remove
1378        those symbols that are exported in the C sense, but actually are local
1379        within the library, and not documented. Their names all begin with
1380        "_pcre_". This is not a perfect job, because (a) we have to except some
1381        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1382        available (and never for static libraries). I have made a note to try to
1383        find a way round (a) in the future.
1384    
1385    
1386    Version 6.2 01-Aug-05
1387    ---------------------
1388    
1389     1. There was no test for integer overflow of quantifier values. A construction
1390        such as {1111111111111111} would give undefined results. What is worse, if
1391        a minimum quantifier for a parenthesized subpattern overflowed and became
1392        negative, the calculation of the memory size went wrong. This could have
1393        led to memory overwriting.
1394    
1395     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1396    
1397     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1398        operating environments where this matters.
1399    
1400     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1401        PCRE options from within the C++ wrapper.
1402    
1403     5. Named capturing subpatterns were not being correctly counted when a pattern
1404        was compiled. This caused two problems: (a) If there were more than 100
1405        such subpatterns, the calculation of the memory needed for the whole
1406        compiled pattern went wrong, leading to an overflow error. (b) Numerical
1407        back references of the form \12, where the number was greater than 9, were
1408        not recognized as back references, even though there were sufficient
1409        previous subpatterns.
1410    
1411     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1412        versions of gcc, e.g. 2.95.4.
1413    
1414    
1415    Version 6.1 21-Jun-05
1416    ---------------------
1417    
1418     1. There was one reference to the variable "posix" in pcretest.c that was not
1419        surrounded by "#if !defined NOPOSIX".
1420    
1421     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1422        the cross-check on the old pcre_info() function, for the benefit of the
1423        cut-down version of PCRE that is currently imported into Exim.
1424    
1425     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1426        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1427        patterns that start (?i) but I don't think that matters. The use of (?i) is
1428        just an example; this all applies to the other options as well.
1429    
1430     4. Since libtool seems to echo the compile commands it is issuing, the output
1431        from "make" can be reduced a bit by putting "@" in front of each libtool
1432        compile command.
1433    
1434     5. Patch from the folks at Google for configure.in to be a bit more thorough
1435        in checking for a suitable C++ installation before trying to compile the
1436        C++ stuff. This should fix a reported problem when a compiler was present,
1437        but no suitable headers.
1438    
1439     6. The man pages all had just "PCRE" as their title. I have changed them to
1440        be the relevant file name. I have also arranged that these names are
1441        retained in the file doc/pcre.txt, which is a concatenation in text format
1442        of all the man pages except the little individual ones for each function.
1443    
1444     7. The NON-UNIX-USE file had not been updated for the different set of source
1445        files that come with release 6. I also added a few comments about the C++
1446        wrapper.
1447    
1448    
1449    Version 6.0 07-Jun-05
1450    ---------------------
1451    
1452     1. Some minor internal re-organization to help with my DFA experiments.
1453    
1454     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1455        didn't matter for the library itself when fully configured, but did matter
1456        when compiling without UCP support, or within Exim, where the ucp files are
1457        not imported.
1458    
1459     3. Refactoring of the library code to split up the various functions into
1460        different source modules. The addition of the new DFA matching code (see
1461        below) to a single monolithic source would have made it really too
1462        unwieldy, quite apart from causing all the code to be include in a
1463        statically linked application, when only some functions are used. This is
1464        relevant even without the DFA addition now that patterns can be compiled in
1465        one application and matched in another.
1466    
1467        The downside of splitting up is that there have to be some external
1468        functions and data tables that are used internally in different modules of
1469        the library but which are not part of the API. These have all had their
1470        names changed to start with "_pcre_" so that they are unlikely to clash
1471        with other external names.
1472    
1473     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1474        a different (DFA) algorithm. Although it is slower than the original
1475        function, it does have some advantages for certain types of matching
1476        problem.
1477    
1478     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1479        including restarting after a partial match.
1480    
1481     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1482        defined when compiling for Windows was sent to me. I have put it into the
1483        code, though I have no means of testing or verifying it.
1484    
1485     7. Added the pcre_refcount() auxiliary function.
1486    
1487     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1488        match before or at the first newline in the subject string. In pcretest,
1489        the /f option on a pattern can be used to set this.
1490    
1491     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1492        would behave wrongly. This has been present in PCRE since release 4.0.
1493    
1494    10. A number of changes to the pcregrep command:
1495    
1496        (a) Refactored how -x works; insert ^(...)$ instead of setting
1497            PCRE_ANCHORED and checking the length, in preparation for adding
1498            something similar for -w.
1499    
1500        (b) Added the -w (match as a word) option.
1501    
1502        (c) Refactored the way lines are read and buffered so as to have more
1503            than one at a time available.
1504    
1505        (d) Implemented a pcregrep test script.
1506    
1507        (e) Added the -M (multiline match) option. This allows patterns to match
1508            over several lines of the subject. The buffering ensures that at least
1509            8K, or the rest of the document (whichever is the shorter) is available
1510            for matching (and similarly the previous 8K for lookbehind assertions).
1511    
1512        (f) Changed the --help output so that it now says
1513    
1514              -w, --word-regex(p)
1515    
1516            instead of two lines, one with "regex" and the other with "regexp"
1517            because that confused at least one person since the short forms are the
1518            same. (This required a bit of code, as the output is generated
1519            automatically from a table. It wasn't just a text change.)
1520    
1521        (g) -- can be used to terminate pcregrep options if the next thing isn't an
1522            option but starts with a hyphen. Could be a pattern or a path name
1523            starting with a hyphen, for instance.
1524    
1525        (h) "-" can be given as a file name to represent stdin.
1526    
1527        (i) When file names are being printed, "(standard input)" is used for
1528            the standard input, for compatibility with GNU grep. Previously
1529            "<stdin>" was used.
1530    
1531        (j) The option --label=xxx can be used to supply a name to be used for
1532            stdin when file names are being printed. There is no short form.
1533    
1534        (k) Re-factored the options decoding logic because we are going to add
1535            two more options that take data. Such options can now be given in four
1536            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1537    
1538        (l) Added the -A, -B, and -C options for requesting that lines of context
1539            around matches be printed.
1540    
1541        (m) Added the -L option to print the names of files that do not contain
1542            any matching lines, that is, the complement of -l.
1543    
1544        (n) The return code is 2 if any file cannot be opened, but pcregrep does
1545            continue to scan other files.
1546    
1547        (o) The -s option was incorrectly implemented. For compatibility with other
1548            greps, it now suppresses the error message for a non-existent or non-
1549            accessible file (but not the return code). There is a new option called
1550            -q that suppresses the output of matching lines, which was what -s was
1551            previously doing.
1552    
1553        (p) Added --include and --exclude options to specify files for inclusion
1554            and exclusion when recursing.
1555    
1556    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1557        Hopefully, it now does.
1558    
1559    12. Missing cast in pcre_study().
1560    
1561    13. Added an "uninstall" target to the makefile.
1562    
1563    14. Replaced "extern" in the function prototypes in Makefile.in with
1564        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1565        world, but is set differently for Windows.
1566    
1567    15. Added a second compiling function called pcre_compile2(). The only
1568        difference is that it has an extra argument, which is a pointer to an
1569        integer error code. When there is a compile-time failure, this is set
1570        non-zero, in addition to the error test pointer being set to point to an
1571        error message. The new argument may be NULL if no error number is required
1572        (but then you may as well call pcre_compile(), which is now just a
1573        wrapper). This facility is provided because some applications need a
1574        numeric error indication, but it has also enabled me to tidy up the way
1575        compile-time errors are handled in the POSIX wrapper.
1576    
1577    16. Added VPATH=.libs to the makefile; this should help when building with one
1578        prefix path and installing with another. (Or so I'm told by someone who
1579        knows more about this stuff than I do.)
1580    
1581    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1582        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1583        match everything, including newlines. This is not POSIX-compatible, but
1584        somebody wanted the feature. From pcretest it can be activated by using
1585        both the P and the s flags.
1586    
1587    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1588    
1589    19. libpcre.pc was being incorrectly installed as executable.
1590    
1591    20. A couple of places in pcretest check for end-of-line by looking for '\n';
1592        it now also looks for '\r' so that it will work unmodified on Windows.
1593    
1594    21. Added Google's contributed C++ wrapper to the distribution.
1595    
1596    22. Added some untidy missing memory free() calls in pcretest, to keep
1597        Electric Fence happy when testing.
1598    
1599    
1600    
1601    Version 5.0 13-Sep-04
1602    ---------------------
1603    
1604     1. Internal change: literal characters are no longer packed up into items
1605        containing multiple characters in a single byte-string. Each character
1606        is now matched using a separate opcode. However, there may be more than one
1607        byte in the character in UTF-8 mode.
1608    
1609     2. The pcre_callout_block structure has two new fields: pattern_position and
1610        next_item_length. These contain the offset in the pattern to the next match
1611        item, and its length, respectively.
1612    
1613     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1614        insertion of callouts before each pattern item. Added the /C option to
1615        pcretest to make use of this.
1616    
1617     4. On the advice of a Windows user, the lines
1618    
1619          #if defined(_WIN32) || defined(WIN32)
1620          _setmode( _fileno( stdout ), 0x8000 );
1621          #endif  /* defined(_WIN32) || defined(WIN32) */
1622    
1623        have been added to the source of pcretest. This apparently does useful
1624        magic in relation to line terminators.
1625    
1626     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1627        for the benefit of those environments where the "b" makes a difference.
1628    
1629     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1630        to know about it. I have put a hack into configure.in that adds in code
1631        to set GCC=yes if CC=icc. This seems to end up at a point in the
1632        generated configure script that is early enough to affect the setting of
1633        compiler options, which is what is needed, but I have no means of testing
1634        whether it really works. (The user who reported this had patched the
1635        generated configure script, which of course I cannot do.)
1636    
1637        LATER: After change 22 below (new libtool files), the configure script
1638        seems to know about icc (and also ecc). Therefore, I have commented out
1639        this hack in configure.in.
1640    
1641     7. Added support for pkg-config (2 patches were sent in).
1642    
1643     8. Negated POSIX character classes that used a combination of internal tables
1644        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1645        [[:^ascii]]. Typically, they would match almost any characters. The other
1646        POSIX classes were not broken in this way.
1647    
1648     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1649        to find the match, as PCRE was deluded into thinking that the match had to
1650        start at the start point or following a newline. The same bug applied to
1651        patterns with negative forward assertions or any backward assertions
1652        preceding ".*" at the start, unless the pattern required a fixed first
1653        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1654    
1655    10. In UTF-8 mode, when moving forwards in the subject after a failed match
1656        starting at the last subject character, bytes beyond the end of the subject
1657        string were read.
1658    
1659    11. Renamed the variable "class" as "classbits" to make life easier for C++
1660        users. (Previously there was a macro definition, but it apparently wasn't
1661        enough.)
1662    
1663    12. Added the new field "tables" to the extra data so that tables can be passed
1664        in at exec time, or the internal tables can be re-selected. This allows
1665        a compiled regex to be saved and re-used at a later time by a different
1666        program that might have everything at different addresses.
1667    
1668    13. Modified the pcre-config script so that, when run on Solaris, it shows a
1669        -R library as well as a -L library.
1670    
1671    14. The debugging options of pcretest (-d on the command line or D on a
1672        pattern) showed incorrect output for anything following an extended class
1673        that contained multibyte characters and which was followed by a quantifier.
1674    
1675    15. Added optional support for general category Unicode character properties
1676        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1677        support. It adds about 90K to the size of the library. The meanings of the
1678        inbuilt class escapes such as \d and \s have NOT been changed.
1679    
1680    16. Updated pcredemo.c to include calls to free() to release the memory for the
1681        compiled pattern.
1682    
1683    17. The generated file chartables.c was being created in the source directory
1684        instead of in the building directory. This caused the build to fail if the
1685        source directory was different from the building directory, and was
1686        read-only.
1687    
1688    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1689        file. No doubt somebody will tell me if they don't make sense... Also added
1690        Dan Mooney's comments about building on OpenVMS.
1691    
1692    19. Added support for partial matching via the PCRE_PARTIAL option for
1693        pcre_exec() and the \P data escape in pcretest.
1694    
1695    20. Extended pcretest with 3 new pattern features:
1696    
1697        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
1698              write the compiled pattern to the file whose name is "rest-of-line".
1699              This is a straight binary dump of the data, with the saved pointer to
1700              the character tables forced to be NULL. The study data, if any, is
1701              written too. After writing, pcretest reads a new pattern.
1702    
1703        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1704              compiled pattern from the given file. There must not be any
1705              occurrences of "<" in the file name (pretty unlikely); if there are,
1706              pcretest will instead treat the initial "<" as a pattern delimiter.
1707              After reading in the pattern, pcretest goes on to read data lines as
1708              usual.
1709    
1710        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1711              and 16-bit fields in a compiled pattern, to simulate a pattern that
1712              was compiled on a host of opposite endianness.
1713    
1714    21. The pcre-exec() function can now cope with patterns that were compiled on
1715        hosts of opposite endianness, with this restriction:
1716    
1717          As for any compiled expression that is saved and used later, the tables
1718          pointer field cannot be preserved; the extra_data field in the arguments
1719          to pcre_exec() should be used to pass in a tables address if a value
1720          other than the default internal tables were used at compile time.
1721    
1722    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1723        now diagnosed as an error. Previously, most of the time, a negative number
1724        would have been treated as zero, but if in addition "ovector" was passed as
1725        NULL, a crash could occur.
1726    
1727    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1728        new versions from the libtool 1.5 distribution (the last one is a copy of
1729        a file called libtool.m4). This seems to have fixed the need to patch
1730        "configure" to support Darwin 1.3 (which I used to do). However, I still
1731        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1732        workstation).
1733    
1734    24. Changed the PCRE licence to be the more standard "BSD" licence.
1735    
1736    
1737    Version 4.5 01-Dec-03
1738    ---------------------
1739    
1740     1. There has been some re-arrangement of the code for the match() function so
1741        that it can be compiled in a version that does not call itself recursively.
1742        Instead, it keeps those local variables that need separate instances for
1743        each "recursion" in a frame on the heap, and gets/frees frames whenever it
1744        needs to "recurse". Keeping track of where control must go is done by means
1745        of setjmp/longjmp. The whole thing is implemented by a set of macros that
1746        hide most of the details from the main code, and operates only if
1747        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1748        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1749        operating.
1750    
1751        To make it easier for callers to provide specially tailored get/free
1752        functions for this usage, two new functions, pcre_stack_malloc, and
1753        pcre_stack_free, are used. They are always called in strict stacking order,
1754        and the size of block requested is always the same.
1755    
1756        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1757        PCRE has been compiled to use the stack or the heap for recursion. The
1758        -C option of pcretest uses this to show which version is compiled.
1759    
1760        A new data escape \S, is added to pcretest; it causes the amounts of store
1761        obtained and freed by both kinds of malloc/free at match time to be added
1762        to the output.
1763    
1764     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1765        what's available on my current Linux desktop machine.
1766    
1767     3. When matching a UTF-8 string, the test for a valid string at the start has
1768        been extended. If start_offset is not zero, PCRE now checks that it points
1769        to a byte that is the start of a UTF-8 character. If not, it returns
1770        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1771        this is necessary because there may be backward assertions in the pattern.
1772        When matching the same subject several times, it may save resources to use
1773        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1774    
1775     4. The code for checking the validity of UTF-8 strings has been tightened so
1776        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1777        containing "overlong sequences".
1778    
1779     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1780        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1781        should have been "&", but it just so happened that all the cases this let
1782        through by mistake were picked up later in the function.
1783    
1784     6. I had used a variable called "isblank" - this is a C99 function, causing
1785        some compilers to warn. To avoid this, I renamed it (as "blankclass").
1786    
1787     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1788        prompting; (b) run "./pcretest /dev/null" at the start of the test script
1789        so the version is shown; (c) stop "make test" echoing "./RunTest".
1790    
1791     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1792    
1793     9. The prototype for memmove() for systems that don't have it was using
1794        size_t, but the inclusion of the header that defines size_t was later. I've
1795        moved the #includes for the C headers earlier to avoid this.
1796    
1797    10. Added some adjustments to the code to make it easier to compiler on certain
1798        special systems:
1799    
1800          (a) Some "const" qualifiers were missing.
1801          (b) Added the macro EXPORT before all exported functions; by default this
1802              is defined to be empty.
1803          (c) Changed the dftables auxiliary program (that builds chartables.c) so
1804              that it reads its output file name as an argument instead of writing
1805              to the standard output and assuming this can be redirected.
1806    
1807    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1808        class containing characters with values greater than 255, PCRE compilation
1809        went into a loop.
1810    
1811    12. A recursive reference to a subpattern that was within another subpattern
1812        that had a minimum quantifier of zero caused PCRE to crash. For example,
1813        (x(y(?2))z)? provoked this bug with a subject that got as far as the
1814        recursion. If the recursively-called subpattern itself had a zero repeat,
1815        that was OK.
1816    
1817    13. In pcretest, the buffer for reading a data line was set at 30K, but the
1818        buffer into which it was copied (for escape processing) was still set at
1819        1024, so long lines caused crashes.
1820    
1821    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1822        "internal error: code overflow...". This applied to any character class
1823        that was followed by a possessive quantifier.
1824    
1825    15. Modified the Makefile to add libpcre.la as a prerequisite for
1826        libpcreposix.la because I was told this is needed for a parallel build to
1827        work.
1828    
1829    16. If a pattern that contained .* following optional items at the start was
1830        studied, the wrong optimizing data was generated, leading to matching
1831        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1832        matching string must start with a or b or c. The correct conclusion for
1833        this pattern is that a match can start with any character.
1834    
1835    
1836    Version 4.4 13-Aug-03
1837    ---------------------
1838    
1839     1. In UTF-8 mode, a character class containing characters with values between
1840        127 and 255 was not handled correctly if the compiled pattern was studied.
1841        In fixing this, I have also improved the studying algorithm for such
1842        classes (slightly).
1843    
1844     2. Three internal functions had redundant arguments passed to them. Removal
1845        might give a very teeny performance improvement.
1846    
1847     3. Documentation bug: the value of the capture_top field in a callout is *one
1848        more than* the number of the hightest numbered captured substring.
1849    
1850     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1851        in incorrectly linking with a previously installed version. They now link
1852        explicitly with libpcre.la.
1853    
1854     5. configure.in no longer needs to recognize Cygwin specially.
1855    
1856     6. A problem in pcre.in for Windows platforms is fixed.
1857    
1858     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1859        pcretest, it used to include the size of the study block as part of its
1860        output. Unfortunately, the structure contains a field that has a different
1861        size on different hardware architectures. This meant that the tests that
1862        showed this size failed. As the block is currently always of a fixed size,
1863        this information isn't actually particularly useful in pcretest output, so
1864        I have just removed it.
1865    
1866     8. Three pre-processor statements accidentally did not start in column 1.
1867        Sadly, there are *still* compilers around that complain, even though
1868        standard C has not required this for well over a decade. Sigh.
1869    
1870     9. In pcretest, the code for checking callouts passed small integers in the
1871        callout_data field, which is a void * field. However, some picky compilers
1872        complained about the casts involved for this on 64-bit systems. Now
1873        pcretest passes the address of the small integer instead, which should get
1874        rid of the warnings.
1875    
1876    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1877        both compile and run time, and gives an error if an invalid UTF-8 sequence
1878        is found. There is a option for disabling this check in cases where the
1879        string is known to be correct and/or the maximum performance is wanted.
1880    
1881    11. In response to a bug report, I changed one line in Makefile.in from
1882    
1883            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1884        to
1885            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1886    
1887        to look similar to other lines, but I have no way of telling whether this
1888        is the right thing to do, as I do not use Windows. No doubt I'll get told
1889        if it's wrong...
1890    
1891    
1892    Version 4.3 21-May-03
1893    ---------------------
1894    
1895    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1896       Makefile.
1897    
1898    2. Some refactoring to improve the quality of the code:
1899    
1900       (i)   The utf8_table... variables are now declared "const".
1901    
1902       (ii)  The code for \cx, which used the "case flipping" table to upper case
1903             lower case letters, now just substracts 32. This is ASCII-specific,
1904             but the whole concept of \cx is ASCII-specific, so it seems
1905             reasonable.
1906    
1907       (iii) PCRE was using its character types table to recognize decimal and
1908             hexadecimal digits in the pattern. This is silly, because it handles
1909             only 0-9, a-f, and A-F, but the character types table is locale-
1910             specific, which means strange things might happen. A private
1911             table is now used for this - though it costs 256 bytes, a table is
1912             much faster than multiple explicit tests. Of course, the standard
1913             character types table is still used for matching digits in subject
1914             strings against \d.
1915    
1916       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1917             ending in _t are). So I've renamed it as ESC_tee.
1918    
1919    3. The first argument for regexec() in the POSIX wrapper should have been
1920       defined as "const".
1921    
1922    4. Changed pcretest to use malloc() for its buffers so that they can be
1923       Electric Fenced for debugging.
1924    
1925    5. There were several places in the code where, in UTF-8 mode, PCRE would try
1926       to read one or more bytes before the start of the subject string. Often this
1927       had no effect on PCRE's behaviour, but in some circumstances it could
1928       provoke a segmentation fault.
1929    
1930    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1931       to try to read one or more bytes before the start of the subject string.
1932    
1933    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1934       UTF-8 support could misbehave in various ways if the subject string
1935       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1936       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1937       back over UTF-8 characters.)
1938    
1939    
1940    Version 4.2 14-Apr-03
1941    ---------------------
1942    
1943    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1944    
1945    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1946         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1947         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1948         [WIN_PREFIX]: new variable, "cyg" for Cygwin
1949         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1950           and BUILD_EXEEXT
1951         Note: automatic setting of the BUILD variables is not yet working
1952         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1953           compile-time but not at link-time
1954         [LINK]: use for linking executables only
1955         make different versions for Windows and non-Windows
1956         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1957           libraries
1958         [LINK_FOR_BUILD]: new variable
1959         [OBJEXT]: use throughout
1960         [EXEEXT]: use throughout
1961         <winshared>: new target
1962         <wininstall>: new target
1963         <dftables.o>: use native compiler
1964         <dftables>: use native linker
1965         <install>: handle Windows platform correctly
1966         <clean>: ditto
1967         <check>: ditto
1968         copy DLL to top builddir before testing
1969    
1970       As part of these changes, -no-undefined was removed again. This was reported
1971       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1972       in any case.
1973    
1974    3. Some tidies to get rid of compiler warnings:
1975    
1976       . In the match_data structure, match_limit was an unsigned long int, whereas
1977         match_call_count was an int. I've made them both unsigned long ints.
1978    
1979       . In pcretest the fact that a const uschar * doesn't automatically cast to
1980         a void * provoked a warning.
1981    
1982       . Turning on some more compiler warnings threw up some "shadow" variables
1983         and a few more missing casts.
1984    
1985    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1986       option, a class that contained a single character with a value between 128
1987       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1988    
1989    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1990       option, a class that contained several characters, but with at least one
1991       whose value was between 128 and 255 caused PCRE to crash.
1992    
1993    
1994    Version 4.1 12-Mar-03
1995    ---------------------
1996    
1997    1. Compiling with gcc -pedantic found a couple of places where casts were
1998    needed, and a string in dftables.c that was longer than standard compilers are
1999    required to support.
2000    
2001    2. Compiling with Sun's compiler found a few more places where the code could
2002    be tidied up in order to avoid warnings.
2003    
2004    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2005    first of these names is deprecated in the latest Autoconf in favour of the name
2006    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2007    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2008    analogy I have changed it to CFLAGS_FOR_BUILD.
2009    
2010    4. Added -no-undefined to the linking command in the Makefile, because this is
2011    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2012    linking step for the pcreposix library.
2013    
2014    5. PCRE was failing to diagnose the case of two named groups with the same
2015    name.
2016    
2017    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2018    literal character that is needed in the subject for a match, and scans along to
2019    ensure that it is present before embarking on the full matching process. This
2020    saves time in cases of nested unlimited repeats that are never going to match.
2021    Problem: the scan can take a lot of time if the subject is very long (e.g.
2022    megabytes), thus penalizing straightforward matches. It is now done only if the
2023    amount of subject to be scanned is less than 1000 bytes.
2024    
2025    7. A lesser problem with the same optimization is that it was recording the
2026    first character of an anchored pattern as "needed", thus provoking a search
2027    right along the subject, even when the first match of the pattern was going to
2028    fail. The "needed" character is now not set for anchored patterns, unless it
2029    follows something in the pattern that is of non-fixed length. Thus, it still
2030    fulfils its original purpose of finding quick non-matches in cases of nested
2031    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2032    
2033    
2034    Version 4.0 17-Feb-03
2035    ---------------------
2036    
2037    1. If a comment in an extended regex that started immediately after a meta-item
2038    extended to the end of string, PCRE compiled incorrect data. This could lead to
2039    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2040    
2041    2. Moved to autoconf 2.53 and libtool 1.4.2.
2042    
2043    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2044    the special perltest8 script is no longer needed - all the tests can be run
2045    from a single perltest script.
2046    
2047    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2048    by \s. It has now been removed in PCRE. This means it isn't recognized as
2049    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2050    class [:space:] *does* include VT, thereby creating a mess.
2051    
2052    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2053    space and tab.
2054    
2055    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2056    its new features into the main test script, reducing the number of scripts.
2057    
2058    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2059    were backward compatible, and made the (?i) apply to the whole pattern, as if
2060    /i were given. Now it behaves more logically, and applies the option setting
2061    only to what follows. PCRE has been changed to follow suit. However, if it
2062    finds options settings right at the start of the pattern, it extracts them into
2063    the global options, as before. Thus, they show up in the info data.
2064    
2065    8. Added support for the \Q...\E escape sequence. Characters in between are
2066    treated as literals. This is slightly different from Perl in that $ and @ are
2067    also handled as literals inside the quotes. In Perl, they will cause variable
2068    interpolation. Note the following examples:
2069    
2070        Pattern            PCRE matches      Perl matches
2071    
2072        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2073        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2074        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2075    
2076    For compatibility with Perl, \Q...\E sequences are recognized inside character
2077    classes as well as outside them.
2078    
2079    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2080    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2081    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2082    signed/unsigned warnings.
2083    
2084    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2085    option for pcretest, so I've replaced it by a simple function that does just
2086    that job.
2087    
2088    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2089    "pcregrep -".
2090    
2091    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2092    Java package. This provides some syntactic sugar for simple cases of what my
2093    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2094    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2095    item, you can use this simplified notation. Note that only makes sense with
2096    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2097    greediness, whatever the setting of the PCRE_UNGREEDY option.
2098    
2099    13. A change of greediness default within a pattern was not taking effect at
2100    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2101    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2102    was abstracted outside.
2103    
2104    14. PCRE now supports the \G assertion. It is true when the current matching
2105    position is at the start point of the match. This differs from \A when the
2106    starting offset is non-zero. Used with the /g option of pcretest (or similar
2107    code), it works in the same way as it does for Perl's /g option. If all
2108    alternatives of a regex begin with \G, the expression is anchored to the start
2109    match position, and the "anchored" flag is set in the compiled expression.
2110    
2111    15. Some bugs concerning the handling of certain option changes within patterns
2112    have been fixed. These applied to options other than (?ims). For example,
2113    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2114    been the other way round. Some of this was related to change 7 above.
2115    
2116    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2117    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2118    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2119    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2120    
2121    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2122    mode. Unlike ".", it always matches newline, whatever the setting of
2123    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2124    assertions. Perl allows it, but it doesn't (in general) work because it can't
2125    calculate the length of the lookbehind. At least, that's the case for Perl
2126    5.8.0 - I've been told they are going to document that it doesn't work in
2127    future.
2128    
2129    18. Added an error diagnosis for escapes that PCRE does not support: these are
2130    \L, \l, \N, \P, \p, \U, \u, and \X.
2131    
2132    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2133    reading past the end of the pattern in cases such as /[abcd/.
2134    
2135    20. PCRE was getting more memory than necessary for patterns with classes that
2136    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2137    
2138    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2139    compiling PCRE for use with Virtual Pascal.
2140    
2141    22. Small fix to the Makefile to make it work properly if the build is done
2142    outside the source tree.
2143    
2144    23. Added a new extension: a condition to go with recursion. If a conditional
2145    subpattern starts with (?(R) the "true" branch is used if recursion has
2146    happened, whereas the "false" branch is used only at the top level.
2147    
2148    24. When there was a very long string of literal characters (over 255 bytes
2149    without UTF support, over 250 bytes with UTF support), the computation of how
2150    much memory was required could be incorrect, leading to segfaults or other
2151    strange effects.
2152    
2153    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2154    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2155    there was a subsequent back reference to those brackets. This meant that, for
2156    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2157    possible to check for precisely this case. All we can do is abandon the
2158    optimization if .* occurs inside capturing brackets when there are any back
2159    references whatsoever. (See below for a better fix that came later.)
2160    
2161    26. The handling of the optimization for finding the first character of a
2162    non-anchored pattern, and for finding a character that is required later in the
2163    match were failing in some cases. This didn't break the matching; it just
2164    failed to optimize when it could. The way this is done has been re-implemented.
2165    
2166    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2167    
2168    28. Added a new feature that provides some of the functionality that Perl
2169    provides with (?{...}). The facility is termed a "callout". The way it is done
2170    in PCRE is for the caller to provide an optional function, by setting
2171    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2172    global variable. By default it is unset, which disables all calling out. To get
2173    the function called, the regex must include (?C) at appropriate points. This
2174    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2175    This provides a means of identifying different callout points. When PCRE
2176    reaches such a point in the regex, if pcre_callout has been set, the external
2177    function is called. It is provided with data in a structure called
2178    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2179    matching continues; if it returns a non-zero value, the match at the current
2180    point fails. However, backtracking will occur if possible. [This was changed
2181    later and other features added - see item 49 below.]
2182    
2183    29. pcretest is upgraded to test the callout functionality. It provides a
2184    callout function that displays information. By default, it shows the start of
2185    the match and the current position in the text. There are some new data escapes
2186    to vary what happens:
2187    
2188        \C+         in addition, show current contents of captured substrings
2189        \C-         do not supply a callout function
2190        \C!n        return 1 when callout number n is reached
2191        \C!n!m      return 1 when callout number n is reached for the mth time
2192    
2193    30. If pcregrep was called with the -l option and just a single file name, it
2194    output "<stdin>" if a match was found, instead of the file name.
2195    
2196    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2197    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2198    pcre_exec(). This saves a malloc/free per call. The default value of
2199    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2200    when configuring.
2201    
2202    32. The default maximum size of a compiled pattern is 64K. There have been a
2203    few cases of people hitting this limit. The code now uses macros to handle the
2204    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2205    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2206    configuring. Tests 2 and 5 work only with 2-byte links because they output
2207    debugging information about compiled patterns.
2208    
2209    33. Internal code re-arrangements:
2210    
2211    (a) Moved the debugging function for printing out a compiled regex into
2212        its own source file (printint.c) and used #include to pull it into
2213        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2214        separate copies.
2215    
2216    (b) Defined the list of op-code names for debugging as a macro in
2217        internal.h so that it is next to the definition of the opcodes.
2218    
2219    (c) Defined a table of op-code lengths for simpler skipping along compiled
2220        code. This is again a macro in internal.h so that it is next to the
2221        definition of the opcodes.
2222    
2223    34. Added support for recursive calls to individual subpatterns, along the
2224    lines of Robin Houston's patch (but implemented somewhat differently).
2225    
2226    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2227    allow it to read and process whole directories in Win32. This code was
2228    contributed by Lionel Fourquaux; it has not been tested by me.
2229    
2230    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2231    used to name a group. Names consist of alphanumerics and underscores, and must
2232    be unique. Back references use the syntax (?P=name) and recursive calls use
2233    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2234    numbers. The function pcre_fullinfo() can be used after compilation to extract
2235    a name/number map. There are three relevant calls:
2236    
2237      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2238      PCRE_INFO_NAMECOUNT            yields the number of entries
2239      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2240    
2241    The map is a vector of fixed-size entries. The size of each entry depends on
2242    the length of the longest name used. The first two bytes of each entry are the
2243    group number, most significant byte first. There follows the corresponding
2244    name, zero terminated. The names are in alphabetical order.
2245    
2246    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2247    case instead of 255. Making it the same both with and without UTF-8 support
2248    means that the same test output works with both.
2249    
2250    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2251    calling malloc() with a zero argument.
2252    
2253    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2254    optimization. I've improved things by keeping a bitmap of backreferences with
2255    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2256    fact referenced, the optimization can be applied. It is unlikely that a
2257    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2258    the match to follow \n) will appear inside brackets with a number greater than
2259    31, but if it does, any back reference > 31 suppresses the optimization.
2260    
2261    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2262    of disabling numbered capturing parentheses. Any opening parenthesis that is
2263    not followed by ? behaves as if it were followed by ?: but named parentheses
2264    can still be used for capturing (and they will acquire numbers in the usual
2265    way).
2266    
2267    41. Redesigned the return codes from the match() function into yes/no/error so
2268    that errors can be passed back from deep inside the nested calls. A malloc
2269    failure while inside a recursive subpattern call now causes the
2270    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2271    
2272    42. It is now possible to set a limit on the number of times the match()
2273    function is called in a call to pcre_exec(). This facility makes it possible to
2274    limit the amount of recursion and backtracking, though not in a directly
2275    obvious way, because the match() function is used in a number of different
2276    circumstances. The count starts from zero for each position in the subject
2277    string (for non-anchored patterns). The default limit is, for compatibility, a
2278    large number, namely 10 000 000. You can change this in two ways:
2279    
2280    (a) When configuring PCRE before making, you can use --with-match-limit=n
2281        to set a default value for the compiled library.
2282    
2283    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2284        a different value is set. See 45 below.
2285    
2286    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2287    
2288    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2289    of things that can be changed at compile time. The first argument specifies
2290    what is wanted and the second points to where the information is to be placed.
2291    The current list of available information is:
2292    
2293      PCRE_CONFIG_UTF8
2294    
2295    The output is an integer that is set to one if UTF-8 support is available;
2296    otherwise it is set to zero.
2297    
2298      PCRE_CONFIG_NEWLINE
2299    
2300    The output is an integer that it set to the value of the code that is used for
2301    newline. It is either LF (10) or CR (13).
2302    
2303      PCRE_CONFIG_LINK_SIZE
2304    
2305    The output is an integer that contains the number of bytes used for internal
2306    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2307    
2308      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2309    
2310    The output is an integer that contains the threshold above which the POSIX
2311    interface uses malloc() for output vectors. See item 31 above.
2312    
2313      PCRE_CONFIG_MATCH_LIMIT
2314    
2315    The output is an unsigned integer that contains the default limit of the number
2316    of match() calls in a pcre_exec() execution. See 42 above.
2317    
2318    44. pcretest has been upgraded by the addition of the -C option. This causes it
2319    to extract all the available output from the new pcre_config() function, and to
2320    output it. The program then exits immediately.
2321    
2322    45. A need has arisen to pass over additional data with calls to pcre_exec() in
2323    order to support additional features. One way would have been to define
2324    pcre_exec2() (for example) with extra arguments, but this would not have been
2325    extensible, and would also have required all calls to the original function to
2326    be mapped to the new one. Instead, I have chosen to extend the mechanism that
2327    is used for passing in "extra" data from pcre_study().
2328    
2329    The pcre_extra structure is now exposed and defined in pcre.h. It currently
2330    contains the following fields:
2331    
2332      flags         a bitmap indicating which of the following fields are set
2333      study_data    opaque data from pcre_study()
2334      match_limit   a way of specifying a limit on match() calls for a specific
2335                      call to pcre_exec()
2336      callout_data  data for callouts (see 49 below)
2337    
2338    The flag bits are also defined in pcre.h, and are
2339    
2340      PCRE_EXTRA_STUDY_DATA
2341      PCRE_EXTRA_MATCH_LIMIT
2342      PCRE_EXTRA_CALLOUT_DATA
2343    
2344    The pcre_study() function now returns one of these new pcre_extra blocks, with
2345    the actual study data pointed to by the study_data field, and the
2346    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2347    before. That is, this change is entirely upwards-compatible and requires no
2348    change to existing code.
2349    
2350    If you want to pass in additional data to pcre_exec(), you can either place it
2351    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2352    block.
2353    
2354    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2355    data string contains the escape sequence \M, pcretest calls pcre_exec() several
2356    times with different match limits, until it finds the minimum value needed for
2357    pcre_exec() to complete. The value is then output. This can be instructive; for
2358    most simple matches the number is quite small, but for pathological cases it
2359    gets very large very quickly.
2360    
2361    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2362    returns the size of the data block pointed to by the study_data field in a
2363    pcre_extra block, that is, the value that was passed as the argument to
2364    pcre_malloc() when PCRE was getting memory in which to place the information
2365    created by pcre_study(). The fourth argument should point to a size_t variable.
2366    pcretest has been extended so that this information is shown after a successful
2367    pcre_study() call when information about the compiled regex is being displayed.
2368    
2369    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2370    because what follows is always an absolute path. (Later: it turns out that this
2371    is more than cosmetic for MinGW, because it doesn't like empty path
2372    components.)
2373    
2374    49. Some changes have been made to the callout feature (see 28 above):
2375    
2376    (i)  A callout function now has three choices for what it returns:
2377    
2378           0  =>  success, carry on matching
2379         > 0  =>  failure at this point, but backtrack if possible
2380         < 0  =>  serious error, return this value from pcre_exec()
2381    
2382         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2383         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2384         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2385         use by callout functions. It will never be used by PCRE itself.
2386    
2387    (ii) The pcre_extra structure (see 45 above) has a void * field called
2388         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2389         pcre_callout_block structure has a field of the same name. The contents of
2390         the field passed in the pcre_extra structure are passed to the callout
2391         function in the corresponding field in the callout block. This makes it
2392         easier to use the same callout-containing regex from multiple threads. For
2393         testing, the pcretest program has a new data escape
2394    
2395           \C*n        pass the number n (may be negative) as callout_data
2396    
2397         If the callout function in pcretest receives a non-zero value as
2398         callout_data, it returns that value.
2399    
2400    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2401    there were some redundant $(CFLAGS) in commands that are now specified as
2402    $(LINK), which already includes $(CFLAGS).
2403    
2404    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2405    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2406    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2407    one-byte characters throughout. Note that case-insensitive matching applies
2408    only to characters whose values are less than 256. PCRE doesn't support the
2409    notion of cases for higher-valued characters.
2410    
2411    (i)   A character class whose characters are all within 0-255 is handled as
2412          a bit map, and the map is inverted for negative classes. Previously, a
2413          character > 255 always failed to match such a class; however it should
2414          match if the class was a negative one (e.g. [^ab]). This has been fixed.
2415    
2416    (ii)  A negated character class with a single character < 255 is coded as
2417          "not this character" (OP_NOT). This wasn't working properly when the test
2418          character was multibyte, either singly or repeated.
2419    
2420    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2421          mode, for example: \x{100}{2,3}.
2422    
2423    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2424          singly or repeated) now correctly test multibyte characters. However,
2425          PCRE doesn't recognize any characters with values greater than 255 as
2426          digits, spaces, or word characters. Such characters always match \D, \S,
2427          and \W, and never match \d, \s, or \w.
2428    
2429    (v)   Classes may now contain characters and character ranges with values
2430          greater than 255. For example: [ab\x{100}-\x{400}].
2431    
2432    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
2433          PCRE in UTF-8 mode.
2434    
2435    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2436    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2437    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2438    value.)
2439    
2440    53. The single man page has become too large. I have therefore split it up into
2441    a number of separate man pages. These also give rise to individual HTML pages;
2442    these are now put in a separate directory, and there is an index.html page that
2443    lists them all. Some hyperlinking between the pages has been installed.
2444    
2445    54. Added convenience functions for handling named capturing parentheses.
2446    
2447    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2448    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2449    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2450    are faulted.
2451    
2452    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2453    calling configure. These values are used when compiling the dftables.c program
2454    which is run to generate the source of the default character tables. They
2455    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2456    you will need to set these values.
2457    
2458    57. Updated the building process for Windows DLL, as provided by Fred Cox.
2459    
2460    
2461    Version 3.9 02-Jan-02
2462    ---------------------
2463    
2464    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2465    
2466    2. If --disable-static was given, the building process failed when trying to
2467    build pcretest and pcregrep. (For some reason it was using libtool to compile
2468    them, which is not right, as they aren't part of the library.)
2469    
2470    
2471  Version 3.8 18-Dec-01  Version 3.8 18-Dec-01
2472  ---------------------  ---------------------
2473    

Legend:
Removed from v.59  
changed lines
  Added in v.321

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12