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

Legend:
Removed from v.59  
changed lines
  Added in v.100

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12