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

Diff of /code/trunk/ChangeLog

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

revision 9 by nigel, Sat Feb 24 21:38:13 2007 UTC revision 341 by ph10, Sat Apr 19 16:41:04 2008 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 7.7 05-Mar-08
5    ---------------------
6    
7    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
8        a string to a long long, pretend we don't even have a long long." This is
9        done by checking for the strtoq, strtoll, and _strtoi64 functions.
10    
11    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
12        pre-7.6 versions, which defined a global no_arg variable instead of putting
13        it in the RE class. (See also #8 below.)
14    
15    3.  Remove a line of dead code, identified by coverity and reported by Nuno
16        Lopes.
17    
18    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
19    
20        (1) The include/exclude patterns were being applied to the whole pathnames
21            of files, instead of just to the final components.
22    
23        (2) If there was more than one level of directory, the subdirectories were
24            skipped unless they satisfied the include/exclude conditions. This is
25            inconsistent with GNU grep (and could even be seen as contrary to the
26            pcregrep specification - which I improved to make it absolutely clear).
27            The action now is always to scan all levels of directory, and just
28            apply the include/exclude patterns to regular files.
29    
30    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
31        --exclude_dir in the tests to avoid scanning .svn directories.
32    
33    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
34        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
35        doesn't support NULs in patterns.
36    
37    7.  Added some missing "const"s to declarations of static tables in
38        pcre_compile.c and pcre_dfa_exec.c.
39    
40    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
41        caused by fix #2  above. (Subsequently also a second patch to fix the
42        first patch. And a third patch - this was a messy problem.)
43    
44    9.  Applied Craig's patch to remove the use of push_back().
45    
46    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
47        matching function regexec().
48    
49    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
50        which, however, unlike Perl's \g{...}, are subroutine calls, not back
51        references. PCRE supports relative numbers with this syntax (I don't think
52        Oniguruma does).
53    
54    12. Previously, a group with a zero repeat such as (...){0} was completely
55        omitted from the compiled regex. However, this means that if the group
56        was called as a subroutine from elsewhere in the pattern, things went wrong
57        (an internal error was given). Such groups are now left in the compiled
58        pattern, with a new opcode that causes them to be skipped at execution
59        time.
60    
61    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
62        to the way PCRE behaves:
63    
64        (a) A lone ] character is dis-allowed (Perl treats it as data).
65    
66        (b) A back reference to an unmatched subpattern matches an empty string
67            (Perl fails the current match path).
68    
69        (c) A data ] in a character class must be notated as \] because if the
70            first data character in a class is ], it defines an empty class. (In
71            Perl it is not possible to have an empty class.) The empty class []
72            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
73            The negative empty class [^] matches any one character, independently
74            of the DOTALL setting.
75    
76    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
77        non-existent subpattern following a character class starting with ']' and
78        containing () gave an internal compiling error instead of "reference to
79        non-existent subpattern". Fortunately, when the pattern did exist, the
80        compiled code was correct. (When scanning forwards to check for the
81        existencd of the subpattern, it was treating the data ']' as terminating
82        the class, so got the count wrong. When actually compiling, the reference
83        was subsequently set up correctly.)
84    
85    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
86        it was being rejected as not supported by pcre_dfa_exec(), even though
87        other assertions are supported. I have made pcre_dfa_exec() support
88        (*FAIL).
89    
90    
91    Version 7.6 28-Jan-08
92    ---------------------
93    
94    1.  A character class containing a very large number of characters with
95        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
96        overflow.
97    
98    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
99        HAVE_LONG_LONG is not defined.
100    
101    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
102        bring them up to date and include new features. This patch includes:
103    
104        - Fixed PH's badly added libz and libbz2 support.
105        - Fixed a problem with static linking.
106        - Added pcredemo. [But later removed - see 7 below.]
107        - Fixed dftables problem and added an option.
108        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
109            HAVE_LONG_LONG.
110        - Added readline support for pcretest.
111        - Added an listing of the option settings after cmake has run.
112    
113    4.  A user submitted a patch to Makefile that makes it easy to create
114        "pcre.dll" under mingw when using Configure/Make. I added stuff to
115        Makefile.am that cause it to include this special target, without
116        affecting anything else. Note that the same mingw target plus all
117        the other distribution libraries and programs are now supported
118        when configuring with CMake (see 6 below) instead of with
119        Configure/Make.
120    
121    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
122        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
123        exported in the Windows port". It has not yet been confirmed that the patch
124        solves the problem, but it does no harm.
125    
126    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
127        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
128        with CMake, and also correct the comment about stack recursion.
129    
130    7.  Remove the automatic building of pcredemo from the ./configure system and
131        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
132        of a program that users should build themselves after PCRE is installed, so
133        building it automatically is not really right. What is more, it gave
134        trouble in some build environments.
135    
136    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
137    
138    
139    Version 7.5 10-Jan-08
140    ---------------------
141    
142    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
143        values in parens when parsing an RE using the C++ wrapper."
144    
145    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
146        Characters greater than 255 were excluded from the class instead of being
147        included.
148    
149    3.  The same bug as (2) above applied to negated POSIX classes such as
150        [:^space:].
151    
152    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
153        defined or documented. It seems to have been a typo for PCRE_STATIC, so
154        I have changed it.
155    
156    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
157        first named subpattern) and a construct such as (?&a) would reference the
158        first named subpattern whose name started with "a" (in other words, the
159        length check was missing). Both these problems are fixed. "Subpattern name
160        expected" is now given for (?&) (a zero-length name), and this patch also
161        makes it give the same error for \k'' (previously it complained that that
162        was a reference to a non-existent subpattern).
163    
164    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
165        this is right because (?- can be followed by option settings as well as by
166        digits. I have, however, made the messages clearer.
167    
168    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
169        than the number used in the conditional) now cause a compile-time error.
170        This is actually not compatible with Perl, which accepts such patterns, but
171        treats the conditional as always being FALSE (as PCRE used to), but it
172        seems to me that giving a diagnostic is better.
173    
174    8.  Change "alphameric" to the more common word "alphanumeric" in comments
175        and messages.
176    
177    9.  Fix two occurrences of "backslash" in comments that should have been
178        "backspace".
179    
180    10. Remove two redundant lines of code that can never be obeyed (their function
181        was moved elsewhere).
182    
183    11. The program that makes PCRE's Unicode character property table had a bug
184        which caused it to generate incorrect table entries for sequences of
185        characters that have the same character type, but are in different scripts.
186        It amalgamated them into a single range, with the script of the first of
187        them. In other words, some characters were in the wrong script. There were
188        thirteen such cases, affecting characters in the following ranges:
189    
190          U+002b0 - U+002c1
191          U+0060c - U+0060d
192          U+0061e - U+00612
193          U+0064b - U+0065e
194          U+0074d - U+0076d
195          U+01800 - U+01805
196          U+01d00 - U+01d77
197          U+01d9b - U+01dbf
198          U+0200b - U+0200f
199          U+030fc - U+030fe
200          U+03260 - U+0327f
201          U+0fb46 - U+0fbb1
202          U+10450 - U+1049d
203    
204    12. The -o option (show only the matching part of a line) for pcregrep was not
205        compatible with GNU grep in that, if there was more than one match in a
206        line, it showed only the first of them. It now behaves in the same way as
207        GNU grep.
208    
209    13. If the -o and -v options were combined for pcregrep, it printed a blank
210        line for every non-matching line. GNU grep prints nothing, and pcregrep now
211        does the same. The return code can be used to tell if there were any
212        non-matching lines.
213    
214    14. Added --file-offsets and --line-offsets to pcregrep.
215    
216    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
217        infinitely looping recursion. The bug was that positive lookaheads were not
218        being skipped when checking for a possible empty match (negative lookaheads
219        and both kinds of lookbehind were skipped).
220    
221    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
222        inclusion of <windows.h> to before rather than after the definition of
223        INVALID_FILE_ATTRIBUTES (patch from David Byron).
224    
225    17. Specifying a possessive quantifier with a specific limit for a Unicode
226        character property caused pcre_compile() to compile bad code, which led at
227        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
228        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
229        caused the error; without that there was no problem.
230    
231    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
232    
233    19. Added --enable-pcretest-libreadline.
234    
235    20. In pcrecpp.cc, the variable 'count' was incremented twice in
236        RE::GlobalReplace(). As a result, the number of replacements returned was
237        double what it should be. I removed one of the increments, but Craig sent a
238        later patch that removed the other one (the right fix) and added unit tests
239        that check the return values (which was not done before).
240    
241    21. Several CMake things:
242    
243        (1) Arranged that, when cmake is used on Unix, the libraries end up with
244            the names libpcre and libpcreposix, not just pcre and pcreposix.
245    
246        (2) The above change means that pcretest and pcregrep are now correctly
247            linked with the newly-built libraries, not previously installed ones.
248    
249        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
250    
251    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
252        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
253        UTF-8 newline character). The key issue is that the pattern starts .*;
254        this means that the match must be either at the beginning, or after a
255        newline. The bug was in the code for advancing after a failed match and
256        checking that the new position followed a newline. It was not taking
257        account of UTF-8 characters correctly.
258    
259    23. PCRE was behaving differently from Perl in the way it recognized POSIX
260        character classes. PCRE was not treating the sequence [:...:] as a
261        character class unless the ... were all letters. Perl, however, seems to
262        allow any characters between [: and :], though of course it rejects as
263        unknown any "names" that contain non-letters, because all the known class
264        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
265        for example, whereas PCRE did not - it did not recognize a POSIX character
266        class. This seemed a bit dangerous, so the code has been changed to be
267        closer to Perl. The behaviour is not identical to Perl, because PCRE will
268        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
269        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
270        Perl does, and where it didn't before.
271    
272    24. Rewrite so as to remove the single use of %n from pcregrep because in some
273        Windows environments %n is disabled by default.
274    
275    
276    Version 7.4 21-Sep-07
277    ---------------------
278    
279    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
280        means that a class such as [\s] counted as "explicit reference to CR or
281        LF". That isn't really right - the whole point of the change was to try to
282        help when there was an actual mention of one of the two characters. So now
283        the change happens only if \r or \n (or a literal CR or LF) character is
284        encountered.
285    
286    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
287        of both had grown to the point where there were only 3 bits left.
288        Fortunately, there was spare space in the data structure, and so I have
289        moved the internal flags into a new 16-bit field to free up more option
290        bits.
291    
292    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
293        but did not set the internal JCHANGED flag - either of these is enough to
294        control the way the "get" function works - but the PCRE_INFO_JCHANGED
295        facility is supposed to tell if (?J) was ever used, so now (?J) at the
296        start sets both bits.
297    
298    4.  Added options (at build time, compile time, exec time) to change \R from
299        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
300    
301    5.  doc/pcresyntax.html was missing from the distribution.
302    
303    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
304        compatibility, even though it is no longer used.
305    
306    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
307        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
308        windows.h file is present (where different names are used). [This was
309        reversed later after testing - see 16 below.]
310    
311    8.  Changed all #include <config.h> to #include "config.h". There were also
312        some further <pcre.h> cases that I changed to "pcre.h".
313    
314    9.  When pcregrep was used with the --colour option, it missed the line ending
315        sequence off the lines that it output.
316    
317    10. It was pointed out to me that arrays of string pointers cause lots of
318        relocations when a shared library is dynamically loaded. A technique of
319        using a single long string with a table of offsets can drastically reduce
320        these. I have refactored PCRE in four places to do this. The result is
321        dramatic:
322    
323          Originally:                          290
324          After changing UCP table:            187
325          After changing error message table:   43
326          After changing table of "verbs"       36
327          After changing table of Posix names   22
328    
329        Thanks to the folks working on Gregex for glib for this insight.
330    
331    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
332        unicode-properties was also set.
333    
334    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
335    
336    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
337        checked only for CRLF.
338    
339    14. Added casts to pcretest.c to avoid compiler warnings.
340    
341    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
342    
343    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
344        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
345        entirely. This removes changes made in 7 above.
346    
347    17. The CMake files have been updated, and there is now more information about
348        building with CMake in the NON-UNIX-USE document.
349    
350    
351    Version 7.3 28-Aug-07
352    ---------------------
353    
354     1. In the rejigging of the build system that eventually resulted in 7.1, the
355        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
356        brackets there is not right, since it causes compilers to look for an
357        installed pcre.h, not the version that is in the source that is being
358        compiled (which of course may be different). I have changed it back to:
359    
360          #include "pcre.h"
361    
362        I have a vague recollection that the change was concerned with compiling in
363        different directories, but in the new build system, that is taken care of
364        by the VPATH setting the Makefile.
365    
366     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
367        when the subject happened to end in the byte 0x85 (e.g. if the last
368        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
369        characters but of course it shouldn't be taken as a newline when it is part
370        of another character. The bug was that, for an unlimited repeat of . in
371        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
372        characters when looking for a newline.
373    
374     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
375    
376     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
377        in debug output.
378    
379     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
380        long printing in the pcrecpp unittest when running under MinGW.
381    
382     6. ESC_K was left out of the EBCDIC table.
383    
384     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
385        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
386        limit also applies to "virtual nesting" when a pattern is recursive, and in
387        this case 1000 isn't so big. I have been able to remove this limit at the
388        expense of backing off one optimization in certain circumstances. Normally,
389        when pcre_exec() would call its internal match() function recursively and
390        immediately return the result unconditionally, it uses a "tail recursion"
391        feature to save stack. However, when a subpattern that can match an empty
392        string has an unlimited repetition quantifier, it no longer makes this
393        optimization. That gives it a stack frame in which to save the data for
394        checking that an empty string has been matched. Previously this was taken
395        from the 1000-entry workspace that had been reserved. So now there is no
396        explicit limit, but more stack is used.
397    
398     8. Applied Daniel's patches to solve problems with the import/export magic
399        syntax that is required for Windows, and which was going wrong for the
400        pcreposix and pcrecpp parts of the library. These were overlooked when this
401        problem was solved for the main library.
402    
403     9. There were some crude static tests to avoid integer overflow when computing
404        the size of patterns that contain repeated groups with explicit upper
405        limits. As the maximum quantifier is 65535, the maximum group length was
406        set at 30,000 so that the product of these two numbers did not overflow a
407        32-bit integer. However, it turns out that people want to use groups that
408        are longer than 30,000 bytes (though not repeat them that many times).
409        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
410        made it possible to implement the integer overflow checks in a much more
411        dynamic way, which I have now done. The artificial limitation on group
412        length has been removed - we now have only the limit on the total length of
413        the compiled pattern, which depends on the LINK_SIZE setting.
414    
415    10. Fixed a bug in the documentation for get/copy named substring when
416        duplicate names are permitted. If none of the named substrings are set, the
417        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
418        empty string.
419    
420    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
421        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
422        because the ] is interpreted as the first data character and the
423        terminating ] is not found. PCRE has been made compatible with Perl in this
424        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
425        cause memory overwriting.
426    
427    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
428        string has been matched (to stop an infinite loop). It was not recognizing
429        a conditional subpattern that could match an empty string if that
430        subpattern was within another subpattern. For example, it looped when
431        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
432        condition was not nested. This bug has been fixed.
433    
434    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
435        past the start of the subject in the presence of bytes with the top bit
436        set, for example "\x8aBCD".
437    
438    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
439        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
440    
441    14. Optimized (?!) to (*FAIL).
442    
443    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
444        This restricts code points to be within the range 0 to 0x10FFFF, excluding
445        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
446        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
447        does: it's just the validity check that is more restrictive.
448    
449    16. Inserted checks for integer overflows during escape sequence (backslash)
450        processing, and also fixed erroneous offset values for syntax errors during
451        backslash processing.
452    
453    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
454        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
455    
456    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
457        caused an overrun.
458    
459    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
460        something other than just ASCII characters) inside a group that had an
461        unlimited repeat caused a loop at compile time (while checking to see
462        whether the group could match an empty string).
463    
464    20. Debugging a pattern containing \p or \P could cause a crash. For example,
465        [\P{Any}] did so. (Error in the code for printing property names.)
466    
467    21. An orphan \E inside a character class could cause a crash.
468    
469    22. A repeated capturing bracket such as (A)? could cause a wild memory
470        reference during compilation.
471    
472    23. There are several functions in pcre_compile() that scan along a compiled
473        expression for various reasons (e.g. to see if it's fixed length for look
474        behind). There were bugs in these functions when a repeated \p or \P was
475        present in the pattern. These operators have additional parameters compared
476        with \d, etc, and these were not being taken into account when moving along
477        the compiled data. Specifically:
478    
479        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
480            length.
481    
482        (b) An item such as \pL+ within a repeated group could cause crashes or
483            loops.
484    
485        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
486            "reference to non-existent subpattern" error.
487    
488        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
489    
490    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
491        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
492    
493    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
494    
495    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
496        character were causing crashes (broken optimization).
497    
498    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
499        \p or \P) caused a compile-time loop.
500    
501    28. More problems have arisen in unanchored patterns when CRLF is a valid line
502        break. For example, the unstudied pattern [\r\n]A does not match the string
503        "\r\nA" because change 7.0/46 below moves the current point on by two
504        characters after failing to match at the start. However, the pattern \nA
505        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
506        the same is true. There doesn't seem any very clean way out of this, but
507        what I have chosen to do makes the common cases work: PCRE now takes note
508        of whether there can be an explicit match for \r or \n anywhere in the
509        pattern, and if so, 7.0/46 no longer applies. As part of this change,
510        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
511        pattern has explicit CR or LF references.
512    
513    29. Added (*CR) etc for changing newline setting at start of pattern.
514    
515    
516    Version 7.2 19-Jun-07
517    ---------------------
518    
519     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
520        which is apparently normally available under Windows.
521    
522     2. Re-jig the pcregrep tests with different newline settings in an attempt
523        to make them independent of the local environment's newline setting.
524    
525     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
526    
527     4. Some of the "internals" tests were previously cut out when the link size
528        was not 2, because the output contained actual offsets. The recent new
529        "Z" feature of pcretest means that these can be cut out, making the tests
530        usable with all link sizes.
531    
532     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
533        stack recursion. This gives a massive performance boost under BSD, but just
534        a small improvement under Linux. However, it saves one field in the frame
535        in all cases.
536    
537     6. Added more features from the forthcoming Perl 5.10:
538    
539        (a) (?-n) (where n is a string of digits) is a relative subroutine or
540            recursion call. It refers to the nth most recently opened parentheses.
541    
542        (b) (?+n) is also a relative subroutine call; it refers to the nth next
543            to be opened parentheses.
544    
545        (c) Conditions that refer to capturing parentheses can be specified
546            relatively, for example, (?(-2)... or (?(+3)...
547    
548        (d) \K resets the start of the current match so that everything before
549            is not part of it.
550    
551        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
552    
553        (f) \g{name} is another synonym - part of Perl 5.10's unification of
554            reference syntax.
555    
556        (g) (?| introduces a group in which the numbering of parentheses in each
557            alternative starts with the same number.
558    
559        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
560    
561     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
562        PCRE_INFO_JCHANGED.
563    
564     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
565        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
566        for detecting groups that can match an empty string.
567    
568     9. A pattern with a very large number of alternatives (more than several
569        hundred) was running out of internal workspace during the pre-compile
570        phase, where pcre_compile() figures out how much memory will be needed. A
571        bit of new cunning has reduced the workspace needed for groups with
572        alternatives. The 1000-alternative test pattern now uses 12 bytes of
573        workspace instead of running out of the 4096 that are available.
574    
575    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
576    
577    11. Applied patch from Google to remove an optimization that didn't quite work.
578        The report of the bug said:
579    
580          pcrecpp::RE("a*").FullMatch("aaa") matches, while
581          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
582          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
583    
584    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
585        it matched the wrong number of bytes.
586    
587    
588    Version 7.1 24-Apr-07
589    ---------------------
590    
591     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
592        that is more "standard", making use of automake and other Autotools. There
593        is some re-arrangement of the files and adjustment of comments consequent
594        on this.
595    
596     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
597        for recursive directory scanning broke on some systems because the files
598        are not scanned in any specific order and on different systems the order
599        was different. A call to "sort" has been inserted into RunGrepTest for the
600        approprate test as a short-term fix. In the longer term there may be an
601        alternative.
602    
603     3. I had an email from Eric Raymond about problems translating some of PCRE's
604        man pages to HTML (despite the fact that I distribute HTML pages, some
605        people do their own conversions for various reasons). The problems
606        concerned the use of low-level troff macros .br and .in. I have therefore
607        removed all such uses from the man pages (some were redundant, some could
608        be replaced by .nf/.fi pairs). The 132html script that I use to generate
609        HTML has been updated to handle .nf/.fi and to complain if it encounters
610        .br or .in.
611    
612     4. Updated comments in configure.ac that get placed in config.h.in and also
613        arranged for config.h to be included in the distribution, with the name
614        config.h.generic, for the benefit of those who have to compile without
615        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
616    
617     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
618        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
619        makevp.bat for the current PCRE, using the additional files
620        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
621    
622     6. A Windows user reported a minor discrepancy with test 2, which turned out
623        to be caused by a trailing space on an input line that had got lost in his
624        copy. The trailing space was an accident, so I've just removed it.
625    
626     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
627        that is needed.
628    
629     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
630        as "const" (a) because they are and (b) because it helps the PHP
631        maintainers who have recently made a script to detect big data structures
632        in the php code that should be moved to the .rodata section. I remembered
633        to update Builducptable as well, so it won't revert if ucptable.h is ever
634        re-created.
635    
636     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
637        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
638        order to be able to cut out the UTF-8 tables in the latter when UTF-8
639        support is not required. This saves 1.5-2K of code, which is important in
640        some applications.
641    
642        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
643        so as not to refer to the tables, even though these functions will never be
644        called when UTF-8 support is disabled. Otherwise there are problems with a
645        shared library.
646    
647    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
648    
649        (a) It was defining its arguments as char * instead of void *.
650    
651        (b) It was assuming that all moves were upwards in memory; this was true
652            a long time ago when I wrote it, but is no longer the case.
653    
654        The emulated memove() is provided for those environments that have neither
655        memmove() nor bcopy(). I didn't think anyone used it these days, but that
656        is clearly not the case, as these two bugs were recently reported.
657    
658    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
659        and Detrail to create the HTML documentation, the .txt form of the man
660        pages, and it removes trailing spaces from listed files. It also creates
661        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
662        case, it wraps all the #defines with #ifndefs. This script should be run
663        before "make dist".
664    
665    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
666        with Unicode property support.
667    
668        (a) For a maximizing quantifier, if the two different cases of the
669            character were of different lengths in their UTF-8 codings (there are
670            some cases like this - I found 11), and the matching function had to
671            back up over a mixture of the two cases, it incorrectly assumed they
672            were both the same length.
673    
674        (b) When PCRE was configured to use the heap rather than the stack for
675            recursion during matching, it was not correctly preserving the data for
676            the other case of a UTF-8 character when checking ahead for a match
677            while processing a minimizing repeat. If the check also involved
678            matching a wide character, but failed, corruption could cause an
679            erroneous result when trying to check for a repeat of the original
680            character.
681    
682    13. Some tidying changes to the testing mechanism:
683    
684        (a) The RunTest script now detects the internal link size and whether there
685            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
686            values substituted by "configure". (The RunGrepTest script already did
687            this for UTF-8.) The configure.ac script no longer substitutes the
688            relevant variables.
689    
690        (b) The debugging options /B and /D in pcretest show the compiled bytecode
691            with length and offset values. This means that the output is different
692            for different internal link sizes. Test 2 is skipped for link sizes
693            other than 2 because of this, bypassing the problem. Unfortunately,
694            there was also a test in test 3 (the locale tests) that used /B and
695            failed for link sizes other than 2. Rather than cut the whole test out,
696            I have added a new /Z option to pcretest that replaces the length and
697            offset values with spaces. This is now used to make test 3 independent
698            of link size. (Test 2 will be tidied up later.)
699    
700    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
701        segmentation fault instead of returning the appropriate error message.
702    
703    15. In multiline mode when the newline sequence was set to "any", the pattern
704        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
705        This doesn't seem right; it now treats the CRLF combination as the line
706        ending, and so does not match in that case. It's only a pattern such as ^$
707        that would hit this one: something like ^ABC$ would have failed after \r
708        and then tried again after \r\n.
709    
710    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
711        in an attempt to make files that differ only in their line terminators
712        compare equal. This works on Linux.
713    
714    17. Under certain error circumstances pcregrep might try to free random memory
715        as it exited. This is now fixed, thanks to valgrind.
716    
717    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
718        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
719        was because its rules for how to advance for /g after matching an empty
720        string at the end of a line did not allow for this case. They now check for
721        it specially.
722    
723    20. pcretest is supposed to handle patterns and data of any length, by
724        extending its buffers when necessary. It was getting this wrong when the
725        buffer for a data line had to be extended.
726    
727    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
728        CRLF as a newline sequence.
729    
730    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
731        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
732        I have nevertheless tidied it up.
733    
734    23. Added some casts to kill warnings from HP-UX ia64 compiler.
735    
736    24. Added a man page for pcre-config.
737    
738    
739    Version 7.0 19-Dec-06
740    ---------------------
741    
742     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
743        moving to gcc 4.1.1.
744    
745     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
746        sys/time.h, which is documented as needed for this function. It doesn't
747        seem to matter on Linux, but it showed up on some releases of OS X.
748    
749     3. It seems that there are systems where bytes whose values are greater than
750        127 match isprint() in the "C" locale. The "C" locale should be the
751        default when a C program starts up. In most systems, only ASCII printing
752        characters match isprint(). This difference caused the output from pcretest
753        to vary, making some of the tests fail. I have changed pcretest so that:
754    
755        (a) When it is outputting text in the compiled version of a pattern, bytes
756            other than 32-126 are always shown as hex escapes.
757    
758        (b) When it is outputting text that is a matched part of a subject string,
759            it does the same, unless a different locale has been set for the match
760            (using the /L modifier). In this case, it uses isprint() to decide.
761    
762     4. Fixed a major bug that caused incorrect computation of the amount of memory
763        required for a compiled pattern when options that changed within the
764        pattern affected the logic of the preliminary scan that determines the
765        length. The relevant options are -x, and -i in UTF-8 mode. The result was
766        that the computed length was too small. The symptoms of this bug were
767        either the PCRE error "internal error: code overflow" from pcre_compile(),
768        or a glibc crash with a message such as "pcretest: free(): invalid next
769        size (fast)". Examples of patterns that provoked this bug (shown in
770        pcretest format) are:
771    
772          /(?-x: )/x
773          /(?x)(?-x: \s*#\s*)/
774          /((?i)[\x{c0}])/8
775          /(?i:[\x{c0}])/8
776    
777        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
778        is now done differently.
779    
780     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
781        wrapper classes; (b) implement a new function in the C++ scanner that is
782        more efficient than the old way of doing things because it avoids levels of
783        recursion in the regex matching; (c) add a paragraph to the documentation
784        for the FullMatch() function.
785    
786     6. The escape sequence \n was being treated as whatever was defined as
787        "newline". Not only was this contrary to the documentation, which states
788        that \n is character 10 (hex 0A), but it also went horribly wrong when
789        "newline" was defined as CRLF. This has been fixed.
790    
791     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
792        was being set to -1 for the "end of line" case (supposedly a value that no
793        character can have). Though this value is never used (the check for end of
794        line is "zero bytes in current character"), it caused compiler complaints.
795        I've changed it to 0xffffffff.
796    
797     8. In pcre_version.c, the version string was being built by a sequence of
798        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
799        string (as it is for production releases) called a macro with an empty
800        argument. The C standard says the result of this is undefined. The gcc
801        compiler treats it as an empty string (which was what was wanted) but it is
802        reported that Visual C gives an error. The source has been hacked around to
803        avoid this problem.
804    
805     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
806        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
807        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
808        of them did).
809    
810    10. Originally, pcretest opened its input and output without "b"; then I was
811        told that "b" was needed in some environments, so it was added for release
812        5.0 to both the input and output. (It makes no difference on Unix-like
813        systems.) Later I was told that it is wrong for the input on Windows. I've
814        now abstracted the modes into two macros, to make it easier to fiddle with
815        them, and removed "b" from the input mode under Windows.
816    
817    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
818    
819    12. Added -help and --help to pcretest as an official way of being reminded
820        of the options.
821    
822    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
823        and pcrecpp.cc because they annoy compilers at high warning levels.
824    
825    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
826    
827    15. Fixed an occurrence of == in configure.ac that should have been = (shell
828        scripts are not C programs :-) and which was not noticed because it works
829        on Linux.
830    
831    16. pcretest is supposed to handle any length of pattern and data line (as one
832        line or as a continued sequence of lines) by extending its input buffer if
833        necessary. This feature was broken for very long pattern lines, leading to
834        a string of junk being passed to pcre_compile() if the pattern was longer
835        than about 50K.
836    
837    17. I have done a major re-factoring of the way pcre_compile() computes the
838        amount of memory needed for a compiled pattern. Previously, there was code
839        that made a preliminary scan of the pattern in order to do this. That was
840        OK when PCRE was new, but as the facilities have expanded, it has become
841        harder and harder to keep it in step with the real compile phase, and there
842        have been a number of bugs (see for example, 4 above). I have now found a
843        cunning way of running the real compile function in a "fake" mode that
844        enables it to compute how much memory it would need, while actually only
845        ever using a few hundred bytes of working memory and without too many
846        tests of the mode. This should make future maintenance and development
847        easier. A side effect of this work is that the limit of 200 on the nesting
848        depth of parentheses has been removed (though this was never a serious
849        limitation, I suspect). However, there is a downside: pcre_compile() now
850        runs more slowly than before (30% or more, depending on the pattern). I
851        hope this isn't a big issue. There is no effect on runtime performance.
852    
853    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
854        newline (only possible for the last line of a file) and it was a
855        pattern that set a locale (followed by /Lsomething), pcretest crashed.
856    
857    19. Added additional timing features to pcretest. (1) The -tm option now times
858        matching only, not compiling. (2) Both -t and -tm can be followed, as a
859        separate command line item, by a number that specifies the number of
860        repeats to use when timing. The default is 50000; this gives better
861        precision, but takes uncomfortably long for very large patterns.
862    
863    20. Extended pcre_study() to be more clever in cases where a branch of a
864        subpattern has no definite first character. For example, (a*|b*)[cd] would
865        previously give no result from pcre_study(). Now it recognizes that the
866        first character must be a, b, c, or d.
867    
868    21. There was an incorrect error "recursive call could loop indefinitely" if
869        a subpattern (or the entire pattern) that was being tested for matching an
870        empty string contained only one non-empty item after a nested subpattern.
871        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
872        incorrectly, because the \d was being skipped in the check.
873    
874    22. The pcretest program now has a new pattern option /B and a command line
875        option -b, which is equivalent to adding /B to every pattern. This causes
876        it to show the compiled bytecode, without the additional information that
877        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
878        is the same as /B/I).
879    
880    23. A new optimization is now able automatically to treat some sequences such
881        as a*b as a*+b. More specifically, if something simple (such as a character
882        or a simple class like \d) has an unlimited quantifier, and is followed by
883        something that cannot possibly match the quantified thing, the quantifier
884        is automatically "possessified".
885    
886    24. A recursive reference to a subpattern whose number was greater than 39
887        went wrong under certain circumstances in UTF-8 mode. This bug could also
888        have affected the operation of pcre_study().
889    
890    25. Realized that a little bit of performance could be had by replacing
891        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
892    
893    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
894    
895    27. Possessive quantifiers such as a++ were previously implemented by turning
896        them into atomic groups such as ($>a+). Now they have their own opcodes,
897        which improves performance. This includes the automatically created ones
898        from 23 above.
899    
900    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
901        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
902        the first matched character to be a colon. This applied both to named and
903        numbered groups.
904    
905    29. The ucpinternal.h header file was missing its idempotency #ifdef.
906    
907    30. I was sent a "project" file called libpcre.a.dev which I understand makes
908        building PCRE on Windows easier, so I have included it in the distribution.
909    
910    31. There is now a check in pcretest against a ridiculously large number being
911        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
912        loop, the loop is abandoned.
913    
914    32. Forward references to subpatterns in conditions such as (?(2)...) where
915        subpattern 2 is defined later cause pcre_compile() to search forwards in
916        the pattern for the relevant set of parentheses. This search went wrong
917        when there were unescaped parentheses in a character class, parentheses
918        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
919    
920    33. "Subroutine" calls and backreferences were previously restricted to
921        referencing subpatterns earlier in the regex. This restriction has now
922        been removed.
923    
924    34. Added a number of extra features that are going to be in Perl 5.10. On the
925        whole, these are just syntactic alternatives for features that PCRE had
926        previously implemented using the Python syntax or my own invention. The
927        other formats are all retained for compatibility.
928    
929        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
930            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
931            also .NET compatible.
932    
933        (b) A recursion or subroutine call to a named group can now be defined as
934            (?&name) as well as (?P>name).
935    
936        (c) A backreference to a named group can now be defined as \k<name> or
937            \k'name' as well as (?P=name). The new forms, as well as being in Perl
938            5.10, are also .NET compatible.
939    
940        (d) A conditional reference to a named group can now use the syntax
941            (?(<name>) or (?('name') as well as (?(name).
942    
943        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
944            groups (named and numbered) that are never evaluated inline, but can be
945            called as "subroutines" from elsewhere. In effect, the DEFINE condition
946            is always false. There may be only one alternative in such a group.
947    
948        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
949            as the simple (?(R). The condition is true only if the most recent
950            recursion is that of the given number or name. It does not search out
951            through the entire recursion stack.
952    
953        (g) The escape \gN or \g{N} has been added, where N is a positive or
954            negative number, specifying an absolute or relative reference.
955    
956    35. Tidied to get rid of some further signed/unsigned compiler warnings and
957        some "unreachable code" warnings.
958    
959    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
960        things, this adds five new scripts.
961    
962    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
963        There were also incompatibilities regarding the handling of \Q..\E inside
964        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
965        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
966    
967    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
968        matches an empty string, and forcibly breaks the loop. There were bugs in
969        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
970        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
971        separate and independent bugs (that affected different cases) have been
972        fixed.
973    
974    39. Refactored the code to abolish the use of different opcodes for small
975        capturing bracket numbers. This is a tidy that I avoided doing when I
976        removed the limit on the number of capturing brackets for 3.5 back in 2001.
977        The new approach is not only tidier, it makes it possible to reduce the
978        memory needed to fix the previous bug (38).
979    
980    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
981        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
982        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
983        mode.
984    
985    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
986        report.
987    
988    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
989        copy construction and assignment in the C++ wrapper.
990    
991    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
992        couple of bugs that could have given wrong results in the "--newline=crlf"
993        case.
994    
995    44. Added a number of casts and did some reorganization of signed/unsigned int
996        variables following suggestions from Dair Grant. Also renamed the variable
997        "this" as "item" because it is a C++ keyword.
998    
999    45. Arranged for dftables to add
1000    
1001          #include "pcre_internal.h"
1002    
1003        to pcre_chartables.c because without it, gcc 4.x may remove the array
1004        definition from the final binary if PCRE is built into a static library and
1005        dead code stripping is activated.
1006    
1007    46. For an unanchored pattern, if a match attempt fails at the start of a
1008        newline sequence, and the newline setting is CRLF or ANY, and the next two
1009        characters are CRLF, advance by two characters instead of one.
1010    
1011    
1012    Version 6.7 04-Jul-06
1013    ---------------------
1014    
1015     1. In order to handle tests when input lines are enormously long, pcretest has
1016        been re-factored so that it automatically extends its buffers when
1017        necessary. The code is crude, but this _is_ just a test program. The
1018        default size has been increased from 32K to 50K.
1019    
1020     2. The code in pcre_study() was using the value of the re argument before
1021        testing it for NULL. (Of course, in any sensible call of the function, it
1022        won't be NULL.)
1023    
1024     3. The memmove() emulation function in pcre_internal.h, which is used on
1025        systems that lack both memmove() and bcopy() - that is, hardly ever -
1026        was missing a "static" storage class specifier.
1027    
1028     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1029        containing an extended class (one that cannot be represented by a bitmap
1030        because it contains high-valued characters or Unicode property items, e.g.
1031        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1032        pattern, but PCRE should not loop if you do not (it no longer does).
1033        [Detail: two cases were found: (a) a repeated subpattern containing an
1034        extended class; (b) a recursive reference to a subpattern that followed a
1035        previous extended class. It wasn't skipping over the extended class
1036        correctly when UTF-8 mode was not set.]
1037    
1038     5. A negated single-character class was not being recognized as fixed-length
1039        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1040        compile error "lookbehind assertion is not fixed length".
1041    
1042     6. The RunPerlTest auxiliary script was showing an unexpected difference
1043        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1044        write a Perl script that can interpret lines of an input file either as
1045        byte characters or as UTF-8, which is what "perltest" was being required to
1046        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1047        can't do is switch easily at run time between having the "use utf8;" pragma
1048        or not. In the end, I fudged it by using the RunPerlTest script to insert
1049        "use utf8;" explicitly for the UTF-8 tests.
1050    
1051     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1052        the end of the subject string, contrary to the documentation and to what
1053        Perl does. This was true of both matching functions. Now it matches only at
1054        the start of the subject and immediately after *internal* newlines.
1055    
1056     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1057        a pointer to an int instead of a pointer to an unsigned long int. This
1058        caused problems on 64-bit systems.
1059    
1060     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1061        instance of the 'standard' template library not being so standard".
1062    
1063    10. There was no check on the number of named subpatterns nor the maximum
1064        length of a subpattern name. The product of these values is used to compute
1065        the size of the memory block for a compiled pattern. By supplying a very
1066        long subpattern name and a large number of named subpatterns, the size
1067        computation could be caused to overflow. This is now prevented by limiting
1068        the length of names to 32 characters, and the number of named subpatterns
1069        to 10,000.
1070    
1071    11. Subpatterns that are repeated with specific counts have to be replicated in
1072        the compiled pattern. The size of memory for this was computed from the
1073        length of the subpattern and the repeat count. The latter is limited to
1074        65535, but there was no limit on the former, meaning that integer overflow
1075        could in principle occur. The compiled length of a repeated subpattern is
1076        now limited to 30,000 bytes in order to prevent this.
1077    
1078    12. Added the optional facility to have named substrings with the same name.
1079    
1080    13. Added the ability to use a named substring as a condition, using the
1081        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1082        are numbers (not recommended). Forward references are permitted.
1083    
1084    14. Added forward references in named backreferences (if you see what I mean).
1085    
1086    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1087        pattern could run off the end of the subject. For example, the pattern
1088        "(?s)(.{1,5})"8 did this with the subject "ab".
1089    
1090    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1091        PCRE_CASELESS was set when matching characters that were quantified with ?
1092        or *.
1093    
1094    17. A character class other than a single negated character that had a minimum
1095        but no maximum quantifier - for example [ab]{6,} - was not handled
1096        correctly by pce_dfa_exec(). It would match only one character.
1097    
1098    18. A valid (though odd) pattern that looked like a POSIX character
1099        class but used an invalid character after [ (for example [[,abc,]]) caused
1100        pcre_compile() to give the error "Failed: internal error: code overflow" or
1101        in some cases to crash with a glibc free() error. This could even happen if
1102        the pattern terminated after [[ but there just happened to be a sequence of
1103        letters, a binary zero, and a closing ] in the memory that followed.
1104    
1105    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1106        over the years. Originally (before any Unicode support), just the bottom 8
1107        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1108        output from "man perlunicode" includes this:
1109    
1110          The regular expression compiler produces polymorphic opcodes.  That
1111          is, the pattern adapts to the data and automatically switches to
1112          the Unicode character scheme when presented with Unicode data--or
1113          instead uses a traditional byte scheme when presented with byte
1114          data.
1115    
1116        Sadly, a wide octal escape does not cause a switch, and in a string with
1117        no other multibyte characters, these octal escapes are treated as before.
1118        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1119        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1120        Unicode string.
1121    
1122        I have not perpetrated such confusion in PCRE. Up till now, it took just
1123        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1124        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1125        translate to the appropriate multibyte character.
1126    
1127    29. Applied some refactoring to reduce the number of warnings from Microsoft
1128        and Borland compilers. This has included removing the fudge introduced
1129        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1130        a warning about an unused variable.
1131    
1132    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1133        characters since release 4.0, because Perl (from release 5.004) does not.
1134        [Or at least, is documented not to: some releases seem to be in conflict
1135        with the documentation.] However, when a pattern was studied with
1136        pcre_study() and all its branches started with \s, PCRE still included VT
1137        as a possible starting character. Of course, this did no harm; it just
1138        caused an unnecessary match attempt.
1139    
1140    22. Removed a now-redundant internal flag bit that recorded the fact that case
1141        dependency changed within the pattern. This was once needed for "required
1142        byte" processing, but is no longer used. This recovers a now-scarce options
1143        bit. Also moved the least significant internal flag bit to the most-
1144        significant bit of the word, which was not previously used (hangover from
1145        the days when it was an int rather than a uint) to free up another bit for
1146        the future.
1147    
1148    23. Added support for CRLF line endings as well as CR and LF. As well as the
1149        default being selectable at build time, it can now be changed at runtime
1150        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1151        specify that it is scanning data with non-default line endings.
1152    
1153    24. Changed the definition of CXXLINK to make it agree with the definition of
1154        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1155    
1156    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1157        recursions. This makes a big different to stack usage for some patterns.
1158    
1159    26. If a subpattern containing a named recursion or subroutine reference such
1160        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1161        the space required for the compiled pattern went wrong and gave too small a
1162        value. Depending on the environment, this could lead to "Failed: internal
1163        error: code overflow at offset 49" or "glibc detected double free or
1164        corruption" errors.
1165    
1166    27. Applied patches from Google (a) to support the new newline modes and (b) to
1167        advance over multibyte UTF-8 characters in GlobalReplace.
1168    
1169    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1170        difference for some implementation of PCRE in some Windows version.
1171    
1172    29. Added some extra testing facilities to pcretest:
1173    
1174        \q<number>   in a data line sets the "match limit" value
1175        \Q<number>   in a data line sets the "match recursion limt" value
1176        -S <number>  sets the stack size, where <number> is in megabytes
1177    
1178        The -S option isn't available for Windows.
1179    
1180    
1181    Version 6.6 06-Feb-06
1182    ---------------------
1183    
1184     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1185        in pcreposix.h. I have copied the definition from pcre.h.
1186    
1187     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1188        because pcre.h is no longer a built file.
1189    
1190     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1191        not normally included in the compiled code.
1192    
1193    
1194    Version 6.5 01-Feb-06
1195    ---------------------
1196    
1197     1. When using the partial match feature with pcre_dfa_exec(), it was not
1198        anchoring the second and subsequent partial matches at the new starting
1199        point. This could lead to incorrect results. For example, with the pattern
1200        /1234/, partially matching against "123" and then "a4" gave a match.
1201    
1202     2. Changes to pcregrep:
1203    
1204        (a) All non-match returns from pcre_exec() were being treated as failures
1205            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1206            error message is output. Some extra information is given for the
1207            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1208            probably the only errors that are likely to be caused by users (by
1209            specifying a regex that has nested indefinite repeats, for instance).
1210            If there are more than 20 of these errors, pcregrep is abandoned.
1211    
1212        (b) A binary zero was treated as data while matching, but terminated the
1213            output line if it was written out. This has been fixed: binary zeroes
1214            are now no different to any other data bytes.
1215    
1216        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1217            used to set a locale for matching. The --locale=xxxx long option has
1218            been added (no short equivalent) to specify a locale explicitly on the
1219            pcregrep command, overriding the environment variables.
1220    
1221        (d) When -B was used with -n, some line numbers in the output were one less
1222            than they should have been.
1223    
1224        (e) Added the -o (--only-matching) option.
1225    
1226        (f) If -A or -C was used with -c (count only), some lines of context were
1227            accidentally printed for the final match.
1228    
1229        (g) Added the -H (--with-filename) option.
1230    
1231        (h) The combination of options -rh failed to suppress file names for files
1232            that were found from directory arguments.
1233    
1234        (i) Added the -D (--devices) and -d (--directories) options.
1235    
1236        (j) Added the -F (--fixed-strings) option.
1237    
1238        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1239    
1240        (l) Added the --colo(u)r option.
1241    
1242        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1243            is not present by default.
1244    
1245     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1246        items such as (?R) or (?1), when the recursion could match a number of
1247        alternatives. If it matched one of the alternatives, but subsequently,
1248        outside the recursion, there was a failure, the code tried to back up into
1249        the recursion. However, because of the way PCRE is implemented, this is not
1250        possible, and the result was an incorrect result from the match.
1251    
1252        In order to prevent this happening, the specification of recursion has
1253        been changed so that all such subpatterns are automatically treated as
1254        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1255    
1256     4. I had overlooked the fact that, in some locales, there are characters for
1257        which isalpha() is true but neither isupper() nor islower() are true. In
1258        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1259        and ordfeminine) are like this. This affected the treatment of \w and \W
1260        when they appeared in character classes, but not when they appeared outside
1261        a character class. The bit map for "word" characters is now created
1262        separately from the results of isalnum() instead of just taking it from the
1263        upper, lower, and digit maps. (Plus the underscore character, of course.)
1264    
1265     5. The above bug also affected the handling of POSIX character classes such as
1266        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1267        permanent tables. Instead, the bit maps for such a class were previously
1268        created as the appropriate unions of the upper, lower, and digit bitmaps.
1269        Now they are created by subtraction from the [[:word:]] class, which has
1270        its own bitmap.
1271    
1272     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1273        It is created by subtracting the vertical space characters (\x09, \x0a,
1274        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1275        subtraction was done in the overall bitmap for a character class, meaning
1276        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1277        be recognized. This bug has been fixed.
1278    
1279     7. Patches from the folks at Google:
1280    
1281          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1282          real life, but is still worth protecting against".
1283    
1284          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1285          regular expressions".
1286    
1287          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1288          have it.
1289    
1290          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1291          "configure" and the latter not, in order to fix a problem somebody had
1292          with compiling the Arg class on HP-UX.
1293    
1294          (e) Improve the error-handling of the C++ wrapper a little bit.
1295    
1296          (f) New tests for checking recursion limiting.
1297    
1298     8. The pcre_memmove() function, which is used only if the environment does not
1299        have a standard memmove() function (and is therefore rarely compiled),
1300        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1301        returning a result (though PCRE never actually uses the result).
1302    
1303     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1304        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1305        returned instead of calling malloc() with an overflowing number that would
1306        most likely cause subsequent chaos.
1307    
1308    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1309    
1310    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1311        with this option is matched, the nmatch and pmatch options of regexec() are
1312        ignored.
1313    
1314    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1315        provided in case anyone wants to the the POSIX interface with UTF-8
1316        strings.
1317    
1318    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1319        C++ linking (needed for some HP-UX environments).
1320    
1321    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1322        (unused parameter) and in the pcre_printint() function (omitted "default"
1323        switch label when the default is to do nothing).
1324    
1325    15. Added some code to make it possible, when PCRE is compiled as a C++
1326        library, to replace subject pointers for pcre_exec() with a smart pointer
1327        class, thus making it possible to process discontinuous strings.
1328    
1329    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1330        much the same function. They were added by different people who were trying
1331        to make PCRE easy to compile on non-Unix systems. It has been suggested
1332        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1333        for compiling on Windows systems. I have therefore replaced it with
1334        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1335        defaults to "extern" for C or "extern C" for C++, which works fine on
1336        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1337        SCOPE with something explicit in config.h. In addition:
1338    
1339        (a) pcreposix.h still had just "extern" instead of either of these macros;
1340            I have replaced it with PCRE_DATA_SCOPE.
1341    
1342        (b) Functions such as _pcre_xclass(), which are internal to the library,
1343            but external in the C sense, all had PCRE_EXPORT in their definitions.
1344            This is apparently wrong for the Windows case, so I have removed it.
1345            (It makes no difference on Unix-like systems.)
1346    
1347    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1348        of recursive calls to match(). This is different to MATCH_LIMIT because
1349        that limits the total number of calls to match(), not all of which increase
1350        the depth of recursion. Limiting the recursion depth limits the amount of
1351        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1352        when PCRE is compiled, and changed at run time. A patch from Google adds
1353        this functionality to the C++ interface.
1354    
1355    18. Changes to the handling of Unicode character properties:
1356    
1357        (a) Updated the table to Unicode 4.1.0.
1358    
1359        (b) Recognize characters that are not in the table as "Cn" (undefined).
1360    
1361        (c) I revised the way the table is implemented to a much improved format
1362            which includes recognition of ranges. It now supports the ranges that
1363            are defined in UnicodeData.txt, and it also amalgamates other
1364            characters into ranges. This has reduced the number of entries in the
1365            table from around 16,000 to around 3,000, thus reducing its size
1366            considerably. I realized I did not need to use a tree structure after
1367            all - a binary chop search is just as efficient. Having reduced the
1368            number of entries, I extended their size from 6 bytes to 8 bytes to
1369            allow for more data.
1370    
1371        (d) Added support for Unicode script names via properties such as \p{Han}.
1372    
1373    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1374        matching that character.
1375    
1376    20. When matching a repeated Unicode property with a minimum greater than zero,
1377        (for example \pL{2,}), PCRE could look past the end of the subject if it
1378        reached it while seeking the minimum number of characters. This could
1379        happen only if some of the characters were more than one byte long, because
1380        there is a check for at least the minimum number of bytes.
1381    
1382    21. Refactored the implementation of \p and \P so as to be more general, to
1383        allow for more different types of property in future. This has changed the
1384        compiled form incompatibly. Anybody with saved compiled patterns that use
1385        \p or \P will have to recompile them.
1386    
1387    22. Added "Any" and "L&" to the supported property types.
1388    
1389    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1390        but give a compile time error if the value is greater than 0xff.
1391    
1392    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1393        accidentally not being installed or uninstalled.
1394    
1395    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1396        made were to insert the current release number. This seemed silly, because
1397        it made things harder for people building PCRE on systems that don't run
1398        "configure". I have turned pcre.h into a distributed file, no longer built
1399        by "configure", with the version identification directly included. There is
1400        no longer a pcre.h.in file.
1401    
1402        However, this change necessitated a change to the pcre-config script as
1403        well. It is built from pcre-config.in, and one of the substitutions was the
1404        release number. I have updated configure.ac so that ./configure now finds
1405        the release number by grepping pcre.h.
1406    
1407    26. Added the ability to run the tests under valgrind.
1408    
1409    
1410    Version 6.4 05-Sep-05
1411    ---------------------
1412    
1413     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1414        "--" to be printed when multiple files were scanned, even when none of the
1415        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1416        consider it to be a bug, and have restored the previous behaviour.
1417    
1418     2. A couple of code tidies to get rid of compiler warnings.
1419    
1420     3. The pcretest program used to cheat by referring to symbols in the library
1421        whose names begin with _pcre_. These are internal symbols that are not
1422        really supposed to be visible externally, and in some environments it is
1423        possible to suppress them. The cheating is now confined to including
1424        certain files from the library's source, which is a bit cleaner.
1425    
1426     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1427        file's purpose clearer.
1428    
1429     5. Reorganized pcre_ucp_findchar().
1430    
1431    
1432    Version 6.3 15-Aug-05
1433    ---------------------
1434    
1435     1. The file libpcre.pc.in did not have general read permission in the tarball.
1436    
1437     2. There were some problems when building without C++ support:
1438    
1439        (a) If C++ support was not built, "make install" and "make test" still
1440            tried to test it.
1441    
1442        (b) There were problems when the value of CXX was explicitly set. Some
1443            changes have been made to try to fix these, and ...
1444    
1445        (c) --disable-cpp can now be used to explicitly disable C++ support.
1446    
1447        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1448            backslash in a target when C++ was disabled. This confuses some
1449            versions of "make", apparently. Using an intermediate variable solves
1450            this. (Same for CPP_LOBJ.)
1451    
1452     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1453        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1454        necessary on certain architectures.
1455    
1456     4. Added a setting of -export-symbols-regex to the link command to remove
1457        those symbols that are exported in the C sense, but actually are local
1458        within the library, and not documented. Their names all begin with
1459        "_pcre_". This is not a perfect job, because (a) we have to except some
1460        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1461        available (and never for static libraries). I have made a note to try to
1462        find a way round (a) in the future.
1463    
1464    
1465    Version 6.2 01-Aug-05
1466    ---------------------
1467    
1468     1. There was no test for integer overflow of quantifier values. A construction
1469        such as {1111111111111111} would give undefined results. What is worse, if
1470        a minimum quantifier for a parenthesized subpattern overflowed and became
1471        negative, the calculation of the memory size went wrong. This could have
1472        led to memory overwriting.
1473    
1474     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1475    
1476     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1477        operating environments where this matters.
1478    
1479     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1480        PCRE options from within the C++ wrapper.
1481    
1482     5. Named capturing subpatterns were not being correctly counted when a pattern
1483        was compiled. This caused two problems: (a) If there were more than 100
1484        such subpatterns, the calculation of the memory needed for the whole
1485        compiled pattern went wrong, leading to an overflow error. (b) Numerical
1486        back references of the form \12, where the number was greater than 9, were
1487        not recognized as back references, even though there were sufficient
1488        previous subpatterns.
1489    
1490     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1491        versions of gcc, e.g. 2.95.4.
1492    
1493    
1494    Version 6.1 21-Jun-05
1495    ---------------------
1496    
1497     1. There was one reference to the variable "posix" in pcretest.c that was not
1498        surrounded by "#if !defined NOPOSIX".
1499    
1500     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1501        the cross-check on the old pcre_info() function, for the benefit of the
1502        cut-down version of PCRE that is currently imported into Exim.
1503    
1504     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1505        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1506        patterns that start (?i) but I don't think that matters. The use of (?i) is
1507        just an example; this all applies to the other options as well.
1508    
1509     4. Since libtool seems to echo the compile commands it is issuing, the output
1510        from "make" can be reduced a bit by putting "@" in front of each libtool
1511        compile command.
1512    
1513     5. Patch from the folks at Google for configure.in to be a bit more thorough
1514        in checking for a suitable C++ installation before trying to compile the
1515        C++ stuff. This should fix a reported problem when a compiler was present,
1516        but no suitable headers.
1517    
1518     6. The man pages all had just "PCRE" as their title. I have changed them to
1519        be the relevant file name. I have also arranged that these names are
1520        retained in the file doc/pcre.txt, which is a concatenation in text format
1521        of all the man pages except the little individual ones for each function.
1522    
1523     7. The NON-UNIX-USE file had not been updated for the different set of source
1524        files that come with release 6. I also added a few comments about the C++
1525        wrapper.
1526    
1527    
1528    Version 6.0 07-Jun-05
1529    ---------------------
1530    
1531     1. Some minor internal re-organization to help with my DFA experiments.
1532    
1533     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1534        didn't matter for the library itself when fully configured, but did matter
1535        when compiling without UCP support, or within Exim, where the ucp files are
1536        not imported.
1537    
1538     3. Refactoring of the library code to split up the various functions into
1539        different source modules. The addition of the new DFA matching code (see
1540        below) to a single monolithic source would have made it really too
1541        unwieldy, quite apart from causing all the code to be include in a
1542        statically linked application, when only some functions are used. This is
1543        relevant even without the DFA addition now that patterns can be compiled in
1544        one application and matched in another.
1545    
1546        The downside of splitting up is that there have to be some external
1547        functions and data tables that are used internally in different modules of
1548        the library but which are not part of the API. These have all had their
1549        names changed to start with "_pcre_" so that they are unlikely to clash
1550        with other external names.
1551    
1552     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1553        a different (DFA) algorithm. Although it is slower than the original
1554        function, it does have some advantages for certain types of matching
1555        problem.
1556    
1557     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1558        including restarting after a partial match.
1559    
1560     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1561        defined when compiling for Windows was sent to me. I have put it into the
1562        code, though I have no means of testing or verifying it.
1563    
1564     7. Added the pcre_refcount() auxiliary function.
1565    
1566     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1567        match before or at the first newline in the subject string. In pcretest,
1568        the /f option on a pattern can be used to set this.
1569    
1570     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1571        would behave wrongly. This has been present in PCRE since release 4.0.
1572    
1573    10. A number of changes to the pcregrep command:
1574    
1575        (a) Refactored how -x works; insert ^(...)$ instead of setting
1576            PCRE_ANCHORED and checking the length, in preparation for adding
1577            something similar for -w.
1578    
1579        (b) Added the -w (match as a word) option.
1580    
1581        (c) Refactored the way lines are read and buffered so as to have more
1582            than one at a time available.
1583    
1584        (d) Implemented a pcregrep test script.
1585    
1586        (e) Added the -M (multiline match) option. This allows patterns to match
1587            over several lines of the subject. The buffering ensures that at least
1588            8K, or the rest of the document (whichever is the shorter) is available
1589            for matching (and similarly the previous 8K for lookbehind assertions).
1590    
1591        (f) Changed the --help output so that it now says
1592    
1593              -w, --word-regex(p)
1594    
1595            instead of two lines, one with "regex" and the other with "regexp"
1596            because that confused at least one person since the short forms are the
1597            same. (This required a bit of code, as the output is generated
1598            automatically from a table. It wasn't just a text change.)
1599    
1600        (g) -- can be used to terminate pcregrep options if the next thing isn't an
1601            option but starts with a hyphen. Could be a pattern or a path name
1602            starting with a hyphen, for instance.
1603    
1604        (h) "-" can be given as a file name to represent stdin.
1605    
1606        (i) When file names are being printed, "(standard input)" is used for
1607            the standard input, for compatibility with GNU grep. Previously
1608            "<stdin>" was used.
1609    
1610        (j) The option --label=xxx can be used to supply a name to be used for
1611            stdin when file names are being printed. There is no short form.
1612    
1613        (k) Re-factored the options decoding logic because we are going to add
1614            two more options that take data. Such options can now be given in four
1615            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1616    
1617        (l) Added the -A, -B, and -C options for requesting that lines of context
1618            around matches be printed.
1619    
1620        (m) Added the -L option to print the names of files that do not contain
1621            any matching lines, that is, the complement of -l.
1622    
1623        (n) The return code is 2 if any file cannot be opened, but pcregrep does
1624            continue to scan other files.
1625    
1626        (o) The -s option was incorrectly implemented. For compatibility with other
1627            greps, it now suppresses the error message for a non-existent or non-
1628            accessible file (but not the return code). There is a new option called
1629            -q that suppresses the output of matching lines, which was what -s was
1630            previously doing.
1631    
1632        (p) Added --include and --exclude options to specify files for inclusion
1633            and exclusion when recursing.
1634    
1635    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1636        Hopefully, it now does.
1637    
1638    12. Missing cast in pcre_study().
1639    
1640    13. Added an "uninstall" target to the makefile.
1641    
1642    14. Replaced "extern" in the function prototypes in Makefile.in with
1643        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1644        world, but is set differently for Windows.
1645    
1646    15. Added a second compiling function called pcre_compile2(). The only
1647        difference is that it has an extra argument, which is a pointer to an
1648        integer error code. When there is a compile-time failure, this is set
1649        non-zero, in addition to the error test pointer being set to point to an
1650        error message. The new argument may be NULL if no error number is required
1651        (but then you may as well call pcre_compile(), which is now just a
1652        wrapper). This facility is provided because some applications need a
1653        numeric error indication, but it has also enabled me to tidy up the way
1654        compile-time errors are handled in the POSIX wrapper.
1655    
1656    16. Added VPATH=.libs to the makefile; this should help when building with one
1657        prefix path and installing with another. (Or so I'm told by someone who
1658        knows more about this stuff than I do.)
1659    
1660    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1661        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1662        match everything, including newlines. This is not POSIX-compatible, but
1663        somebody wanted the feature. From pcretest it can be activated by using
1664        both the P and the s flags.
1665    
1666    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1667    
1668    19. libpcre.pc was being incorrectly installed as executable.
1669    
1670    20. A couple of places in pcretest check for end-of-line by looking for '\n';
1671        it now also looks for '\r' so that it will work unmodified on Windows.
1672    
1673    21. Added Google's contributed C++ wrapper to the distribution.
1674    
1675    22. Added some untidy missing memory free() calls in pcretest, to keep
1676        Electric Fence happy when testing.
1677    
1678    
1679    
1680    Version 5.0 13-Sep-04
1681    ---------------------
1682    
1683     1. Internal change: literal characters are no longer packed up into items
1684        containing multiple characters in a single byte-string. Each character
1685        is now matched using a separate opcode. However, there may be more than one
1686        byte in the character in UTF-8 mode.
1687    
1688     2. The pcre_callout_block structure has two new fields: pattern_position and
1689        next_item_length. These contain the offset in the pattern to the next match
1690        item, and its length, respectively.
1691    
1692     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1693        insertion of callouts before each pattern item. Added the /C option to
1694        pcretest to make use of this.
1695    
1696     4. On the advice of a Windows user, the lines
1697    
1698          #if defined(_WIN32) || defined(WIN32)
1699          _setmode( _fileno( stdout ), 0x8000 );
1700          #endif  /* defined(_WIN32) || defined(WIN32) */
1701    
1702        have been added to the source of pcretest. This apparently does useful
1703        magic in relation to line terminators.
1704    
1705     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1706        for the benefit of those environments where the "b" makes a difference.
1707    
1708     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1709        to know about it. I have put a hack into configure.in that adds in code
1710        to set GCC=yes if CC=icc. This seems to end up at a point in the
1711        generated configure script that is early enough to affect the setting of
1712        compiler options, which is what is needed, but I have no means of testing
1713        whether it really works. (The user who reported this had patched the
1714        generated configure script, which of course I cannot do.)
1715    
1716        LATER: After change 22 below (new libtool files), the configure script
1717        seems to know about icc (and also ecc). Therefore, I have commented out
1718        this hack in configure.in.
1719    
1720     7. Added support for pkg-config (2 patches were sent in).
1721    
1722     8. Negated POSIX character classes that used a combination of internal tables
1723        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1724        [[:^ascii]]. Typically, they would match almost any characters. The other
1725        POSIX classes were not broken in this way.
1726    
1727     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1728        to find the match, as PCRE was deluded into thinking that the match had to
1729        start at the start point or following a newline. The same bug applied to
1730        patterns with negative forward assertions or any backward assertions
1731        preceding ".*" at the start, unless the pattern required a fixed first
1732        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1733    
1734    10. In UTF-8 mode, when moving forwards in the subject after a failed match
1735        starting at the last subject character, bytes beyond the end of the subject
1736        string were read.
1737    
1738    11. Renamed the variable "class" as "classbits" to make life easier for C++
1739        users. (Previously there was a macro definition, but it apparently wasn't
1740        enough.)
1741    
1742    12. Added the new field "tables" to the extra data so that tables can be passed
1743        in at exec time, or the internal tables can be re-selected. This allows
1744        a compiled regex to be saved and re-used at a later time by a different
1745        program that might have everything at different addresses.
1746    
1747    13. Modified the pcre-config script so that, when run on Solaris, it shows a
1748        -R library as well as a -L library.
1749    
1750    14. The debugging options of pcretest (-d on the command line or D on a
1751        pattern) showed incorrect output for anything following an extended class
1752        that contained multibyte characters and which was followed by a quantifier.
1753    
1754    15. Added optional support for general category Unicode character properties
1755        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1756        support. It adds about 90K to the size of the library. The meanings of the
1757        inbuilt class escapes such as \d and \s have NOT been changed.
1758    
1759    16. Updated pcredemo.c to include calls to free() to release the memory for the
1760        compiled pattern.
1761    
1762    17. The generated file chartables.c was being created in the source directory
1763        instead of in the building directory. This caused the build to fail if the
1764        source directory was different from the building directory, and was
1765        read-only.
1766    
1767    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1768        file. No doubt somebody will tell me if they don't make sense... Also added
1769        Dan Mooney's comments about building on OpenVMS.
1770    
1771    19. Added support for partial matching via the PCRE_PARTIAL option for
1772        pcre_exec() and the \P data escape in pcretest.
1773    
1774    20. Extended pcretest with 3 new pattern features:
1775    
1776        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
1777              write the compiled pattern to the file whose name is "rest-of-line".
1778              This is a straight binary dump of the data, with the saved pointer to
1779              the character tables forced to be NULL. The study data, if any, is
1780              written too. After writing, pcretest reads a new pattern.
1781    
1782        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1783              compiled pattern from the given file. There must not be any
1784              occurrences of "<" in the file name (pretty unlikely); if there are,
1785              pcretest will instead treat the initial "<" as a pattern delimiter.
1786              After reading in the pattern, pcretest goes on to read data lines as
1787              usual.
1788    
1789        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1790              and 16-bit fields in a compiled pattern, to simulate a pattern that
1791              was compiled on a host of opposite endianness.
1792    
1793    21. The pcre-exec() function can now cope with patterns that were compiled on
1794        hosts of opposite endianness, with this restriction:
1795    
1796          As for any compiled expression that is saved and used later, the tables
1797          pointer field cannot be preserved; the extra_data field in the arguments
1798          to pcre_exec() should be used to pass in a tables address if a value
1799          other than the default internal tables were used at compile time.
1800    
1801    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1802        now diagnosed as an error. Previously, most of the time, a negative number
1803        would have been treated as zero, but if in addition "ovector" was passed as
1804        NULL, a crash could occur.
1805    
1806    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1807        new versions from the libtool 1.5 distribution (the last one is a copy of
1808        a file called libtool.m4). This seems to have fixed the need to patch
1809        "configure" to support Darwin 1.3 (which I used to do). However, I still
1810        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1811        workstation).
1812    
1813    24. Changed the PCRE licence to be the more standard "BSD" licence.
1814    
1815    
1816    Version 4.5 01-Dec-03
1817    ---------------------
1818    
1819     1. There has been some re-arrangement of the code for the match() function so
1820        that it can be compiled in a version that does not call itself recursively.
1821        Instead, it keeps those local variables that need separate instances for
1822        each "recursion" in a frame on the heap, and gets/frees frames whenever it
1823        needs to "recurse". Keeping track of where control must go is done by means
1824        of setjmp/longjmp. The whole thing is implemented by a set of macros that
1825        hide most of the details from the main code, and operates only if
1826        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1827        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1828        operating.
1829    
1830        To make it easier for callers to provide specially tailored get/free
1831        functions for this usage, two new functions, pcre_stack_malloc, and
1832        pcre_stack_free, are used. They are always called in strict stacking order,
1833        and the size of block requested is always the same.
1834    
1835        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1836        PCRE has been compiled to use the stack or the heap for recursion. The
1837        -C option of pcretest uses this to show which version is compiled.
1838    
1839        A new data escape \S, is added to pcretest; it causes the amounts of store
1840        obtained and freed by both kinds of malloc/free at match time to be added
1841        to the output.
1842    
1843     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1844        what's available on my current Linux desktop machine.
1845    
1846     3. When matching a UTF-8 string, the test for a valid string at the start has
1847        been extended. If start_offset is not zero, PCRE now checks that it points
1848        to a byte that is the start of a UTF-8 character. If not, it returns
1849        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1850        this is necessary because there may be backward assertions in the pattern.
1851        When matching the same subject several times, it may save resources to use
1852        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1853    
1854     4. The code for checking the validity of UTF-8 strings has been tightened so
1855        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1856        containing "overlong sequences".
1857    
1858     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1859        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1860        should have been "&", but it just so happened that all the cases this let
1861        through by mistake were picked up later in the function.
1862    
1863     6. I had used a variable called "isblank" - this is a C99 function, causing
1864        some compilers to warn. To avoid this, I renamed it (as "blankclass").
1865    
1866     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1867        prompting; (b) run "./pcretest /dev/null" at the start of the test script
1868        so the version is shown; (c) stop "make test" echoing "./RunTest".
1869    
1870     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1871    
1872     9. The prototype for memmove() for systems that don't have it was using
1873        size_t, but the inclusion of the header that defines size_t was later. I've
1874        moved the #includes for the C headers earlier to avoid this.
1875    
1876    10. Added some adjustments to the code to make it easier to compiler on certain
1877        special systems:
1878    
1879          (a) Some "const" qualifiers were missing.
1880          (b) Added the macro EXPORT before all exported functions; by default this
1881              is defined to be empty.
1882          (c) Changed the dftables auxiliary program (that builds chartables.c) so
1883              that it reads its output file name as an argument instead of writing
1884              to the standard output and assuming this can be redirected.
1885    
1886    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1887        class containing characters with values greater than 255, PCRE compilation
1888        went into a loop.
1889    
1890    12. A recursive reference to a subpattern that was within another subpattern
1891        that had a minimum quantifier of zero caused PCRE to crash. For example,
1892        (x(y(?2))z)? provoked this bug with a subject that got as far as the
1893        recursion. If the recursively-called subpattern itself had a zero repeat,
1894        that was OK.
1895    
1896    13. In pcretest, the buffer for reading a data line was set at 30K, but the
1897        buffer into which it was copied (for escape processing) was still set at
1898        1024, so long lines caused crashes.
1899    
1900    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1901        "internal error: code overflow...". This applied to any character class
1902        that was followed by a possessive quantifier.
1903    
1904    15. Modified the Makefile to add libpcre.la as a prerequisite for
1905        libpcreposix.la because I was told this is needed for a parallel build to
1906        work.
1907    
1908    16. If a pattern that contained .* following optional items at the start was
1909        studied, the wrong optimizing data was generated, leading to matching
1910        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1911        matching string must start with a or b or c. The correct conclusion for
1912        this pattern is that a match can start with any character.
1913    
1914    
1915    Version 4.4 13-Aug-03
1916    ---------------------
1917    
1918     1. In UTF-8 mode, a character class containing characters with values between
1919        127 and 255 was not handled correctly if the compiled pattern was studied.
1920        In fixing this, I have also improved the studying algorithm for such
1921        classes (slightly).
1922    
1923     2. Three internal functions had redundant arguments passed to them. Removal
1924        might give a very teeny performance improvement.
1925    
1926     3. Documentation bug: the value of the capture_top field in a callout is *one
1927        more than* the number of the hightest numbered captured substring.
1928    
1929     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1930        in incorrectly linking with a previously installed version. They now link
1931        explicitly with libpcre.la.
1932    
1933     5. configure.in no longer needs to recognize Cygwin specially.
1934    
1935     6. A problem in pcre.in for Windows platforms is fixed.
1936    
1937     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1938        pcretest, it used to include the size of the study block as part of its
1939        output. Unfortunately, the structure contains a field that has a different
1940        size on different hardware architectures. This meant that the tests that
1941        showed this size failed. As the block is currently always of a fixed size,
1942        this information isn't actually particularly useful in pcretest output, so
1943        I have just removed it.
1944    
1945     8. Three pre-processor statements accidentally did not start in column 1.
1946        Sadly, there are *still* compilers around that complain, even though
1947        standard C has not required this for well over a decade. Sigh.
1948    
1949     9. In pcretest, the code for checking callouts passed small integers in the
1950        callout_data field, which is a void * field. However, some picky compilers
1951        complained about the casts involved for this on 64-bit systems. Now
1952        pcretest passes the address of the small integer instead, which should get
1953        rid of the warnings.
1954    
1955    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1956        both compile and run time, and gives an error if an invalid UTF-8 sequence
1957        is found. There is a option for disabling this check in cases where the
1958        string is known to be correct and/or the maximum performance is wanted.
1959    
1960    11. In response to a bug report, I changed one line in Makefile.in from
1961    
1962            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1963        to
1964            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1965    
1966        to look similar to other lines, but I have no way of telling whether this
1967        is the right thing to do, as I do not use Windows. No doubt I'll get told
1968        if it's wrong...
1969    
1970    
1971    Version 4.3 21-May-03
1972    ---------------------
1973    
1974    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1975       Makefile.
1976    
1977    2. Some refactoring to improve the quality of the code:
1978    
1979       (i)   The utf8_table... variables are now declared "const".
1980    
1981       (ii)  The code for \cx, which used the "case flipping" table to upper case
1982             lower case letters, now just substracts 32. This is ASCII-specific,
1983             but the whole concept of \cx is ASCII-specific, so it seems
1984             reasonable.
1985    
1986       (iii) PCRE was using its character types table to recognize decimal and
1987             hexadecimal digits in the pattern. This is silly, because it handles
1988             only 0-9, a-f, and A-F, but the character types table is locale-
1989             specific, which means strange things might happen. A private
1990             table is now used for this - though it costs 256 bytes, a table is
1991             much faster than multiple explicit tests. Of course, the standard
1992             character types table is still used for matching digits in subject
1993             strings against \d.
1994    
1995       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1996             ending in _t are). So I've renamed it as ESC_tee.
1997    
1998    3. The first argument for regexec() in the POSIX wrapper should have been
1999       defined as "const".
2000    
2001    4. Changed pcretest to use malloc() for its buffers so that they can be
2002       Electric Fenced for debugging.
2003    
2004    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2005       to read one or more bytes before the start of the subject string. Often this
2006       had no effect on PCRE's behaviour, but in some circumstances it could
2007       provoke a segmentation fault.
2008    
2009    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2010       to try to read one or more bytes before the start of the subject string.
2011    
2012    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2013       UTF-8 support could misbehave in various ways if the subject string
2014       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2015       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2016       back over UTF-8 characters.)
2017    
2018    
2019    Version 4.2 14-Apr-03
2020    ---------------------
2021    
2022    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2023    
2024    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2025         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2026         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2027         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2028         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2029           and BUILD_EXEEXT
2030         Note: automatic setting of the BUILD variables is not yet working
2031         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2032           compile-time but not at link-time
2033         [LINK]: use for linking executables only
2034         make different versions for Windows and non-Windows
2035         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2036           libraries
2037         [LINK_FOR_BUILD]: new variable
2038         [OBJEXT]: use throughout
2039         [EXEEXT]: use throughout
2040         <winshared>: new target
2041         <wininstall>: new target
2042         <dftables.o>: use native compiler
2043         <dftables>: use native linker
2044         <install>: handle Windows platform correctly
2045         <clean>: ditto
2046         <check>: ditto
2047         copy DLL to top builddir before testing
2048    
2049       As part of these changes, -no-undefined was removed again. This was reported
2050       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2051       in any case.
2052    
2053    3. Some tidies to get rid of compiler warnings:
2054    
2055       . In the match_data structure, match_limit was an unsigned long int, whereas
2056         match_call_count was an int. I've made them both unsigned long ints.
2057    
2058       . In pcretest the fact that a const uschar * doesn't automatically cast to
2059         a void * provoked a warning.
2060    
2061       . Turning on some more compiler warnings threw up some "shadow" variables
2062         and a few more missing casts.
2063    
2064    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2065       option, a class that contained a single character with a value between 128
2066       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2067    
2068    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2069       option, a class that contained several characters, but with at least one
2070       whose value was between 128 and 255 caused PCRE to crash.
2071    
2072    
2073    Version 4.1 12-Mar-03
2074    ---------------------
2075    
2076    1. Compiling with gcc -pedantic found a couple of places where casts were
2077    needed, and a string in dftables.c that was longer than standard compilers are
2078    required to support.
2079    
2080    2. Compiling with Sun's compiler found a few more places where the code could
2081    be tidied up in order to avoid warnings.
2082    
2083    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2084    first of these names is deprecated in the latest Autoconf in favour of the name
2085    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2086    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2087    analogy I have changed it to CFLAGS_FOR_BUILD.
2088    
2089    4. Added -no-undefined to the linking command in the Makefile, because this is
2090    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2091    linking step for the pcreposix library.
2092    
2093    5. PCRE was failing to diagnose the case of two named groups with the same
2094    name.
2095    
2096    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2097    literal character that is needed in the subject for a match, and scans along to
2098    ensure that it is present before embarking on the full matching process. This
2099    saves time in cases of nested unlimited repeats that are never going to match.
2100    Problem: the scan can take a lot of time if the subject is very long (e.g.
2101    megabytes), thus penalizing straightforward matches. It is now done only if the
2102    amount of subject to be scanned is less than 1000 bytes.
2103    
2104    7. A lesser problem with the same optimization is that it was recording the
2105    first character of an anchored pattern as "needed", thus provoking a search
2106    right along the subject, even when the first match of the pattern was going to
2107    fail. The "needed" character is now not set for anchored patterns, unless it
2108    follows something in the pattern that is of non-fixed length. Thus, it still
2109    fulfils its original purpose of finding quick non-matches in cases of nested
2110    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2111    
2112    
2113    Version 4.0 17-Feb-03
2114    ---------------------
2115    
2116    1. If a comment in an extended regex that started immediately after a meta-item
2117    extended to the end of string, PCRE compiled incorrect data. This could lead to
2118    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2119    
2120    2. Moved to autoconf 2.53 and libtool 1.4.2.
2121    
2122    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2123    the special perltest8 script is no longer needed - all the tests can be run
2124    from a single perltest script.
2125    
2126    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2127    by \s. It has now been removed in PCRE. This means it isn't recognized as
2128    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2129    class [:space:] *does* include VT, thereby creating a mess.
2130    
2131    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2132    space and tab.
2133    
2134    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2135    its new features into the main test script, reducing the number of scripts.
2136    
2137    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2138    were backward compatible, and made the (?i) apply to the whole pattern, as if
2139    /i were given. Now it behaves more logically, and applies the option setting
2140    only to what follows. PCRE has been changed to follow suit. However, if it
2141    finds options settings right at the start of the pattern, it extracts them into
2142    the global options, as before. Thus, they show up in the info data.
2143    
2144    8. Added support for the \Q...\E escape sequence. Characters in between are
2145    treated as literals. This is slightly different from Perl in that $ and @ are
2146    also handled as literals inside the quotes. In Perl, they will cause variable
2147    interpolation. Note the following examples:
2148    
2149        Pattern            PCRE matches      Perl matches
2150    
2151        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2152        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2153        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2154    
2155    For compatibility with Perl, \Q...\E sequences are recognized inside character
2156    classes as well as outside them.
2157    
2158    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2159    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2160    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2161    signed/unsigned warnings.
2162    
2163    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2164    option for pcretest, so I've replaced it by a simple function that does just
2165    that job.
2166    
2167    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2168    "pcregrep -".
2169    
2170    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2171    Java package. This provides some syntactic sugar for simple cases of what my
2172    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2173    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2174    item, you can use this simplified notation. Note that only makes sense with
2175    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2176    greediness, whatever the setting of the PCRE_UNGREEDY option.
2177    
2178    13. A change of greediness default within a pattern was not taking effect at
2179    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2180    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2181    was abstracted outside.
2182    
2183    14. PCRE now supports the \G assertion. It is true when the current matching
2184    position is at the start point of the match. This differs from \A when the
2185    starting offset is non-zero. Used with the /g option of pcretest (or similar
2186    code), it works in the same way as it does for Perl's /g option. If all
2187    alternatives of a regex begin with \G, the expression is anchored to the start
2188    match position, and the "anchored" flag is set in the compiled expression.
2189    
2190    15. Some bugs concerning the handling of certain option changes within patterns
2191    have been fixed. These applied to options other than (?ims). For example,
2192    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2193    been the other way round. Some of this was related to change 7 above.
2194    
2195    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2196    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2197    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2198    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2199    
2200    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2201    mode. Unlike ".", it always matches newline, whatever the setting of
2202    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2203    assertions. Perl allows it, but it doesn't (in general) work because it can't
2204    calculate the length of the lookbehind. At least, that's the case for Perl
2205    5.8.0 - I've been told they are going to document that it doesn't work in
2206    future.
2207    
2208    18. Added an error diagnosis for escapes that PCRE does not support: these are
2209    \L, \l, \N, \P, \p, \U, \u, and \X.
2210    
2211    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2212    reading past the end of the pattern in cases such as /[abcd/.
2213    
2214    20. PCRE was getting more memory than necessary for patterns with classes that
2215    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2216    
2217    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2218    compiling PCRE for use with Virtual Pascal.
2219    
2220    22. Small fix to the Makefile to make it work properly if the build is done
2221    outside the source tree.
2222    
2223    23. Added a new extension: a condition to go with recursion. If a conditional
2224    subpattern starts with (?(R) the "true" branch is used if recursion has
2225    happened, whereas the "false" branch is used only at the top level.
2226    
2227    24. When there was a very long string of literal characters (over 255 bytes
2228    without UTF support, over 250 bytes with UTF support), the computation of how
2229    much memory was required could be incorrect, leading to segfaults or other
2230    strange effects.
2231    
2232    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2233    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2234    there was a subsequent back reference to those brackets. This meant that, for
2235    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2236    possible to check for precisely this case. All we can do is abandon the
2237    optimization if .* occurs inside capturing brackets when there are any back
2238    references whatsoever. (See below for a better fix that came later.)
2239    
2240    26. The handling of the optimization for finding the first character of a
2241    non-anchored pattern, and for finding a character that is required later in the
2242    match were failing in some cases. This didn't break the matching; it just
2243    failed to optimize when it could. The way this is done has been re-implemented.
2244    
2245    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2246    
2247    28. Added a new feature that provides some of the functionality that Perl
2248    provides with (?{...}). The facility is termed a "callout". The way it is done
2249    in PCRE is for the caller to provide an optional function, by setting
2250    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2251    global variable. By default it is unset, which disables all calling out. To get
2252    the function called, the regex must include (?C) at appropriate points. This
2253    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2254    This provides a means of identifying different callout points. When PCRE
2255    reaches such a point in the regex, if pcre_callout has been set, the external
2256    function is called. It is provided with data in a structure called
2257    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2258    matching continues; if it returns a non-zero value, the match at the current
2259    point fails. However, backtracking will occur if possible. [This was changed
2260    later and other features added - see item 49 below.]
2261    
2262    29. pcretest is upgraded to test the callout functionality. It provides a
2263    callout function that displays information. By default, it shows the start of
2264    the match and the current position in the text. There are some new data escapes
2265    to vary what happens:
2266    
2267        \C+         in addition, show current contents of captured substrings
2268        \C-         do not supply a callout function
2269        \C!n        return 1 when callout number n is reached
2270        \C!n!m      return 1 when callout number n is reached for the mth time
2271    
2272    30. If pcregrep was called with the -l option and just a single file name, it
2273    output "<stdin>" if a match was found, instead of the file name.
2274    
2275    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2276    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2277    pcre_exec(). This saves a malloc/free per call. The default value of
2278    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2279    when configuring.
2280    
2281    32. The default maximum size of a compiled pattern is 64K. There have been a
2282    few cases of people hitting this limit. The code now uses macros to handle the
2283    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2284    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2285    configuring. Tests 2 and 5 work only with 2-byte links because they output
2286    debugging information about compiled patterns.
2287    
2288    33. Internal code re-arrangements:
2289    
2290    (a) Moved the debugging function for printing out a compiled regex into
2291        its own source file (printint.c) and used #include to pull it into
2292        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2293        separate copies.
2294    
2295    (b) Defined the list of op-code names for debugging as a macro in
2296        internal.h so that it is next to the definition of the opcodes.
2297    
2298    (c) Defined a table of op-code lengths for simpler skipping along compiled
2299        code. This is again a macro in internal.h so that it is next to the
2300        definition of the opcodes.
2301    
2302    34. Added support for recursive calls to individual subpatterns, along the
2303    lines of Robin Houston's patch (but implemented somewhat differently).
2304    
2305    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2306    allow it to read and process whole directories in Win32. This code was
2307    contributed by Lionel Fourquaux; it has not been tested by me.
2308    
2309    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2310    used to name a group. Names consist of alphanumerics and underscores, and must
2311    be unique. Back references use the syntax (?P=name) and recursive calls use
2312    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2313    numbers. The function pcre_fullinfo() can be used after compilation to extract
2314    a name/number map. There are three relevant calls:
2315    
2316      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2317      PCRE_INFO_NAMECOUNT            yields the number of entries
2318      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2319    
2320    The map is a vector of fixed-size entries. The size of each entry depends on
2321    the length of the longest name used. The first two bytes of each entry are the
2322    group number, most significant byte first. There follows the corresponding
2323    name, zero terminated. The names are in alphabetical order.
2324    
2325    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2326    case instead of 255. Making it the same both with and without UTF-8 support
2327    means that the same test output works with both.
2328    
2329    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2330    calling malloc() with a zero argument.
2331    
2332    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2333    optimization. I've improved things by keeping a bitmap of backreferences with
2334    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2335    fact referenced, the optimization can be applied. It is unlikely that a
2336    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2337    the match to follow \n) will appear inside brackets with a number greater than
2338    31, but if it does, any back reference > 31 suppresses the optimization.
2339    
2340    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2341    of disabling numbered capturing parentheses. Any opening parenthesis that is
2342    not followed by ? behaves as if it were followed by ?: but named parentheses
2343    can still be used for capturing (and they will acquire numbers in the usual
2344    way).
2345    
2346    41. Redesigned the return codes from the match() function into yes/no/error so
2347    that errors can be passed back from deep inside the nested calls. A malloc
2348    failure while inside a recursive subpattern call now causes the
2349    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2350    
2351    42. It is now possible to set a limit on the number of times the match()
2352    function is called in a call to pcre_exec(). This facility makes it possible to
2353    limit the amount of recursion and backtracking, though not in a directly
2354    obvious way, because the match() function is used in a number of different
2355    circumstances. The count starts from zero for each position in the subject
2356    string (for non-anchored patterns). The default limit is, for compatibility, a
2357    large number, namely 10 000 000. You can change this in two ways:
2358    
2359    (a) When configuring PCRE before making, you can use --with-match-limit=n
2360        to set a default value for the compiled library.
2361    
2362    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2363        a different value is set. See 45 below.
2364    
2365    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2366    
2367    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2368    of things that can be changed at compile time. The first argument specifies
2369    what is wanted and the second points to where the information is to be placed.
2370    The current list of available information is:
2371    
2372      PCRE_CONFIG_UTF8
2373    
2374    The output is an integer that is set to one if UTF-8 support is available;
2375    otherwise it is set to zero.
2376    
2377      PCRE_CONFIG_NEWLINE
2378    
2379    The output is an integer that it set to the value of the code that is used for
2380    newline. It is either LF (10) or CR (13).
2381    
2382      PCRE_CONFIG_LINK_SIZE
2383    
2384    The output is an integer that contains the number of bytes used for internal
2385    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2386    
2387      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2388    
2389    The output is an integer that contains the threshold above which the POSIX
2390    interface uses malloc() for output vectors. See item 31 above.
2391    
2392      PCRE_CONFIG_MATCH_LIMIT
2393    
2394    The output is an unsigned integer that contains the default limit of the number
2395    of match() calls in a pcre_exec() execution. See 42 above.
2396    
2397    44. pcretest has been upgraded by the addition of the -C option. This causes it
2398    to extract all the available output from the new pcre_config() function, and to
2399    output it. The program then exits immediately.
2400    
2401    45. A need has arisen to pass over additional data with calls to pcre_exec() in
2402    order to support additional features. One way would have been to define
2403    pcre_exec2() (for example) with extra arguments, but this would not have been
2404    extensible, and would also have required all calls to the original function to
2405    be mapped to the new one. Instead, I have chosen to extend the mechanism that
2406    is used for passing in "extra" data from pcre_study().
2407    
2408    The pcre_extra structure is now exposed and defined in pcre.h. It currently
2409    contains the following fields:
2410    
2411      flags         a bitmap indicating which of the following fields are set
2412      study_data    opaque data from pcre_study()
2413      match_limit   a way of specifying a limit on match() calls for a specific
2414                      call to pcre_exec()
2415      callout_data  data for callouts (see 49 below)
2416    
2417    The flag bits are also defined in pcre.h, and are
2418    
2419      PCRE_EXTRA_STUDY_DATA
2420      PCRE_EXTRA_MATCH_LIMIT
2421      PCRE_EXTRA_CALLOUT_DATA
2422    
2423    The pcre_study() function now returns one of these new pcre_extra blocks, with
2424    the actual study data pointed to by the study_data field, and the
2425    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2426    before. That is, this change is entirely upwards-compatible and requires no
2427    change to existing code.
2428    
2429    If you want to pass in additional data to pcre_exec(), you can either place it
2430    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2431    block.
2432    
2433    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2434    data string contains the escape sequence \M, pcretest calls pcre_exec() several
2435    times with different match limits, until it finds the minimum value needed for
2436    pcre_exec() to complete. The value is then output. This can be instructive; for
2437    most simple matches the number is quite small, but for pathological cases it
2438    gets very large very quickly.
2439    
2440    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2441    returns the size of the data block pointed to by the study_data field in a
2442    pcre_extra block, that is, the value that was passed as the argument to
2443    pcre_malloc() when PCRE was getting memory in which to place the information
2444    created by pcre_study(). The fourth argument should point to a size_t variable.
2445    pcretest has been extended so that this information is shown after a successful
2446    pcre_study() call when information about the compiled regex is being displayed.
2447    
2448    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2449    because what follows is always an absolute path. (Later: it turns out that this
2450    is more than cosmetic for MinGW, because it doesn't like empty path
2451    components.)
2452    
2453    49. Some changes have been made to the callout feature (see 28 above):
2454    
2455    (i)  A callout function now has three choices for what it returns:
2456    
2457           0  =>  success, carry on matching
2458         > 0  =>  failure at this point, but backtrack if possible
2459         < 0  =>  serious error, return this value from pcre_exec()
2460    
2461         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2462         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2463         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2464         use by callout functions. It will never be used by PCRE itself.
2465    
2466    (ii) The pcre_extra structure (see 45 above) has a void * field called
2467         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2468         pcre_callout_block structure has a field of the same name. The contents of
2469         the field passed in the pcre_extra structure are passed to the callout
2470         function in the corresponding field in the callout block. This makes it
2471         easier to use the same callout-containing regex from multiple threads. For
2472         testing, the pcretest program has a new data escape
2473    
2474           \C*n        pass the number n (may be negative) as callout_data
2475    
2476         If the callout function in pcretest receives a non-zero value as
2477         callout_data, it returns that value.
2478    
2479    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2480    there were some redundant $(CFLAGS) in commands that are now specified as
2481    $(LINK), which already includes $(CFLAGS).
2482    
2483    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2484    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2485    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2486    one-byte characters throughout. Note that case-insensitive matching applies
2487    only to characters whose values are less than 256. PCRE doesn't support the
2488    notion of cases for higher-valued characters.
2489    
2490    (i)   A character class whose characters are all within 0-255 is handled as
2491          a bit map, and the map is inverted for negative classes. Previously, a
2492          character > 255 always failed to match such a class; however it should
2493          match if the class was a negative one (e.g. [^ab]). This has been fixed.
2494    
2495    (ii)  A negated character class with a single character < 255 is coded as
2496          "not this character" (OP_NOT). This wasn't working properly when the test
2497          character was multibyte, either singly or repeated.
2498    
2499    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2500          mode, for example: \x{100}{2,3}.
2501    
2502    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2503          singly or repeated) now correctly test multibyte characters. However,
2504          PCRE doesn't recognize any characters with values greater than 255 as
2505          digits, spaces, or word characters. Such characters always match \D, \S,
2506          and \W, and never match \d, \s, or \w.
2507    
2508    (v)   Classes may now contain characters and character ranges with values
2509          greater than 255. For example: [ab\x{100}-\x{400}].
2510    
2511    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
2512          PCRE in UTF-8 mode.
2513    
2514    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2515    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2516    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2517    value.)
2518    
2519    53. The single man page has become too large. I have therefore split it up into
2520    a number of separate man pages. These also give rise to individual HTML pages;
2521    these are now put in a separate directory, and there is an index.html page that
2522    lists them all. Some hyperlinking between the pages has been installed.
2523    
2524    54. Added convenience functions for handling named capturing parentheses.
2525    
2526    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2527    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2528    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2529    are faulted.
2530    
2531    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2532    calling configure. These values are used when compiling the dftables.c program
2533    which is run to generate the source of the default character tables. They
2534    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2535    you will need to set these values.
2536    
2537    57. Updated the building process for Windows DLL, as provided by Fred Cox.
2538    
2539    
2540    Version 3.9 02-Jan-02
2541    ---------------------
2542    
2543    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2544    
2545    2. If --disable-static was given, the building process failed when trying to
2546    build pcretest and pcregrep. (For some reason it was using libtool to compile
2547    them, which is not right, as they aren't part of the library.)
2548    
2549    
2550    Version 3.8 18-Dec-01
2551    ---------------------
2552    
2553    1. The experimental UTF-8 code was completely screwed up. It was packing the
2554    bytes in the wrong order. How dumb can you get?
2555    
2556    
2557    Version 3.7 29-Oct-01
2558    ---------------------
2559    
2560    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2561    This caused pcretest, when used on the test data, to segfault. Unfortunately,
2562    this didn't happen under Solaris 8, where I normally test things.
2563    
2564    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2565    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2566    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2567    this fix an hour or so after the initial 3.7 release.)
2568    
2569    
2570    Version 3.6 23-Oct-01
2571    ---------------------
2572    
2573    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2574    offsets passed as NULL with zero offset count.
2575    
2576    2. The config.guess and config.sub files had not been updated when I moved to
2577    the latest autoconf.
2578    
2579    
2580    Version 3.5 15-Aug-01
2581    ---------------------
2582    
2583    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2584    had been forgotten.
2585    
2586    2. By using declared but undefined structures, we can avoid using "void"
2587    definitions in pcre.h while keeping the internal definitions of the structures
2588    private.
2589    
2590    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2591    user point of view, this means that both static and shared libraries are built
2592    by default, but this can be individually controlled. More of the work of
2593    handling this static/shared cases is now inside libtool instead of PCRE's make
2594    file.
2595    
2596    4. The pcretest utility is now installed along with pcregrep because it is
2597    useful for users (to test regexs) and by doing this, it automatically gets
2598    relinked by libtool. The documentation has been turned into a man page, so
2599    there are now .1, .txt, and .html versions in /doc.
2600    
2601    5. Upgrades to pcregrep:
2602       (i)   Added long-form option names like gnu grep.
2603       (ii)  Added --help to list all options with an explanatory phrase.
2604       (iii) Added -r, --recursive to recurse into sub-directories.
2605       (iv)  Added -f, --file to read patterns from a file.
2606    
2607    6. pcre_exec() was referring to its "code" argument before testing that
2608    argument for NULL (and giving an error if it was NULL).
2609    
2610    7. Upgraded Makefile.in to allow for compiling in a different directory from
2611    the source directory.
2612    
2613    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2614    options bits, the pointer it was passed was to an int instead of to an unsigned
2615    long int. This mattered only on 64-bit systems.
2616    
2617    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2618    generated) instead of pcre.in, which it its source. Also made the same change
2619    in several of the .c files.
2620    
2621    10. A new release of gcc defines printf() as a macro, which broke pcretest
2622    because it had an ifdef in the middle of a string argument for printf(). Fixed
2623    by using separate calls to printf().
2624    
2625    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2626    script, to force use of CR or LF instead of \n in the source. On non-Unix
2627    systems, the value can be set in config.h.
2628    
2629    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2630    absolute limit. Changed the text of the error message to make this clear, and
2631    likewise updated the man page.
2632    
2633    13. The limit of 99 on the number of capturing subpatterns has been removed.
2634    The new limit is 65535, which I hope will not be a "real" limit.
2635    
2636    
2637    Version 3.4 22-Aug-00
2638    ---------------------
2639    
2640    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2641    
2642    2. Diagnose condition (?(0) as an error instead of crashing on matching.
2643    
2644    
2645    Version 3.3 01-Aug-00
2646    ---------------------
2647    
2648    1. If an octal character was given, but the value was greater than \377, it
2649    was not getting masked to the least significant bits, as documented. This could
2650    lead to crashes in some systems.
2651    
2652    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2653    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2654    
2655    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2656    These just pass their arguments on to (pcre_free)(), but they are provided
2657    because some uses of PCRE bind it to non-C systems that can call its functions,
2658    but cannot call free() or pcre_free() directly.
2659    
2660    4. Add "make test" as a synonym for "make check". Corrected some comments in
2661    the Makefile.
2662    
2663    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2664    Makefile.
2665    
2666    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2667    command called pgrep for grepping around the active processes.
2668    
2669    7. Added the beginnings of support for UTF-8 character strings.
2670    
2671    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2672    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2673    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2674    out for the ar command.)
2675    
2676    
2677    Version 3.2 12-May-00
2678    ---------------------
2679    
2680    This is purely a bug fixing release.
2681    
2682    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2683    of ZA. This was just one example of several cases that could provoke this bug,
2684    which was introduced by change 9 of version 2.00. The code for breaking
2685    infinite loops after an iteration that matches an empty string was't working
2686    correctly.
2687    
2688    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2689    when matched against abbab (for example). After matching an empty string, it
2690    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2691    caused it to match further down the string than it should.
2692    
2693    3. The code contained an inclusion of sys/types.h. It isn't clear why this
2694    was there because it doesn't seem to be needed, and it causes trouble on some
2695    systems, as it is not a Standard C header. It has been removed.
2696    
2697    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2698    were reported on the Macintosh. The changes were from
2699    
2700      while ((c = *(++ptr)) != 0 && c != '\n');
2701    to
2702      while ((c = *(++ptr)) != 0 && c != '\n') ;
2703    
2704    Totally extraordinary, but if that's what it takes...
2705    
2706    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2707    available. Added HAVE_BCOPY and an autoconf test for it; if neither
2708    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2709    assumes the way PCRE uses memmove() (always moving upwards).
2710    
2711    6. PCRE is being used in one environment where strchr() is not available. There
2712    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2713    faster code anyway.
2714    
2715    
2716    Version 3.1 09-Feb-00
2717    ---------------------
2718    
2719    The only change in this release is the fixing of some bugs in Makefile.in for
2720    the "install" target:
2721    
2722    (1) It was failing to install pcreposix.h.
2723    
2724    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2725    
2726    
2727    Version 3.0 01-Feb-00
2728    ---------------------
2729    
2730    1. Add support for the /+ modifier to perltest (to output $` like it does in
2731    pcretest).
2732    
2733    2. Add support for the /g modifier to perltest.
2734    
2735    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2736    matches null strings.
2737    
2738    4. Fix perltest so that it doesn't do unwanted things when fed an empty
2739    pattern. Perl treats empty patterns specially - it reuses the most recent
2740    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2741    effect.
2742    
2743    5. The POSIX interface was broken in that it was just handing over the POSIX
2744    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2745    required a bigger vector, with some working space on the end. This means that
2746    the POSIX wrapper now has to get and free some memory, and copy the results.
2747    
2748    6. Added some simple autoconf support, placing the test data and the
2749    documentation in separate directories, re-organizing some of the
2750    information files, and making it build pcre-config (a GNU standard). Also added
2751    libtool support for building PCRE as a shared library, which is now the
2752    default.
2753    
2754    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2755    09 are not valid octal constants. Single digits will be used for minor values
2756    less than 10.
2757    
2758    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2759    existing programs that set these in the POSIX interface can use PCRE without
2760    modification.
2761    
2762    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2763    return all that pcre_info() returns, plus additional data. The pcre_info()
2764    function is retained for compatibility, but is considered to be obsolete.
2765    
2766    10. Added experimental recursion feature (?R) to handle one common case that
2767    Perl 5.6 will be able to do with (?p{...}).
2768    
2769    11. Added support for POSIX character classes like [:alpha:], which Perl is
2770    adopting.
2771    
2772    
2773    Version 2.08 31-Aug-99
2774    ----------------------
2775    
2776    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2777    trying to match at the startoffset position, but instead was moving forward to
2778    the next newline as if a previous match had failed.
2779    
2780    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2781    and could get into a loop if a null string was matched other than at the start
2782    of the subject.
2783    
2784    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2785    be distinguished at compile time, and for completeness also added PCRE_DATE.
2786    
2787    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2788    in GnuWin32 environments.
2789    
2790    
2791    Version 2.07 29-Jul-99
2792    ----------------------
2793    
2794    1. The documentation is now supplied in plain text form and HTML as well as in
2795    the form of man page sources.
2796    
2797    2. C++ compilers don't like assigning (void *) values to other pointer types.
2798    In particular this affects malloc(). Although there is no problem in Standard
2799    C, I've put in casts to keep C++ compilers happy.
2800    
2801    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2802    should be (const char *).
2803    
2804    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2805    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2806    However, I haven't made this a standard facility. The documentation doesn't
2807    mention it, and the Makefile doesn't support it.
2808    
2809    5. The Makefile now contains an "install" target, with editable destinations at
2810    the top of the file. The pcretest program is not installed.
2811    
2812    6. pgrep -V now gives the PCRE version number and date.
2813    
2814    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2815    causing the entire string to be ignored, instead of just the last character.
2816    
2817    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2818    non-matching string, it can take a very, very long time, even for strings of
2819    quite modest length, because of the nested recursion. PCRE now does better in
2820    some of these cases. It does this by remembering the last required literal
2821    character in the pattern, and pre-searching the subject to ensure it is present
2822    before running the real match. In other words, it applies a heuristic to detect
2823    some types of certain failure quickly, and in the above example, if presented
2824    with a string that has no trailing " it gives "no match" very quickly.
2825    
2826    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2827    other alternatives are tried instead.
2828    
2829    
2830    Version 2.06 09-Jun-99
2831    ----------------------
2832    
2833    1. Change pcretest's output for amount of store used to show just the code
2834    space, because the remainder (the data block) varies in size between 32-bit and
2835    64-bit systems.
2836    
2837    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2838    start matching at. This allows lookbehinds to work when searching for multiple
2839    occurrences in a string.
2840    
2841    3. Added additional options to pcretest for testing multiple occurrences:
2842    
2843       /+   outputs the rest of the string that follows a match
2844       /g   loops for multiple occurrences, using the new startoffset argument
2845       /G   loops for multiple occurrences by passing an incremented pointer
2846    
2847    4. PCRE wasn't doing the "first character" optimization for patterns starting
2848    with \b or \B, though it was doing it for other lookbehind assertions. That is,
2849    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2850    the letter 'x'. On long subject strings, this gives a significant speed-up.
2851    
2852    
2853    Version 2.05 21-Apr-99
2854    ----------------------
2855    
2856    1. Changed the type of magic_number from int to long int so that it works
2857    properly on 16-bit systems.
2858    
2859    2. Fixed a bug which caused patterns starting with .* not to work correctly
2860    when the subject string contained newline characters. PCRE was assuming
2861    anchoring for such patterns in all cases, which is not correct because .* will
2862    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2863    DOTALL is set at top level; otherwise it knows that patterns starting with .*
2864    must be retried after every newline in the subject.
2865    
2866    
2867    Version 2.04 18-Feb-99
2868    ----------------------
2869    
2870    1. For parenthesized subpatterns with repeats whose minimum was zero, the
2871    computation of the store needed to hold the pattern was incorrect (too large).
2872    If such patterns were nested a few deep, this could multiply and become a real
2873    problem.
2874    
2875    2. Added /M option to pcretest to show the memory requirement of a specific
2876    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2877    
2878    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2879    compiled in such a way that the backtracking after subsequent failure was
2880    pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2881    ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2882    
2883    
2884    Version 2.03 02-Feb-99
2885    ----------------------
2886    
2887    1. Fixed typo and small mistake in man page.
2888    
2889    2. Added 4th condition (GPL supersedes if conflict) and created separate
2890    LICENCE file containing the conditions.
2891    
2892    3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2893    Perl, that is the internal \ allows the delimiter to be included in the
2894    pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2895    the final delimiter, add \ to the end of the pattern (to test the error).
2896    
2897    4. Added the convenience functions for extracting substrings after a successful
2898    match. Updated pcretest to make it able to test these functions.
2899    
2900    
2901    Version 2.02 14-Jan-99
2902    ----------------------
2903    
2904    1. Initialized the working variables associated with each extraction so that
2905    their saving and restoring doesn't refer to uninitialized store.
2906    
2907    2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2908    compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2909    fix the problem.
2910    
2911    3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2912    calls, and wasn't printing the correct value for compiling calls. Increased the
2913    default value of LOOPREPEAT, and the number of significant figures in the
2914    times.
2915    
2916    4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2917    
2918    5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2919    a building problem on Windows NT with a FAT file system.
2920    
2921    
2922    Version 2.01 21-Oct-98
2923    ----------------------
2924    
2925    1. Changed the API for pcre_compile() to allow for the provision of a pointer
2926    to character tables built by pcre_maketables() in the current locale. If NULL
2927    is passed, the default tables are used.
2928    
2929    
2930    Version 2.00 24-Sep-98
2931    ----------------------
2932    
2933    1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2934    it any more.
2935    
2936    2. Allow quantification of (?>) groups, and make it work correctly.
2937    
2938    3. The first character computation wasn't working for (?>) groups.
2939    
2940    4. Correct the implementation of \Z (it is permitted to match on the \n at the
2941    end of the subject) and add 5.005's \z, which really does match only at the
2942    very end of the subject.
2943    
2944    5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2945    
2946    6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2947    DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2948    localized options. All options to pcre_study() were also removed.
2949    
2950    7. Add other new features from 5.005:
2951    
2952       $(?<=           positive lookbehind
2953       $(?<!           negative lookbehind
2954       (?imsx-imsx)    added the unsetting capability
2955                       such a setting is global if at outer level; local otherwise
2956       (?imsx-imsx:)   non-capturing groups with option setting
2957       (?(cond)re|re)  conditional pattern matching
2958    
2959       A backreference to itself in a repeated group matches the previous
2960       captured string.
2961    
2962    8. General tidying up of studying (both automatic and via "study")
2963    consequential on the addition of new assertions.
2964    
2965    9. As in 5.005, unlimited repeated groups that could match an empty substring
2966    are no longer faulted at compile time. Instead, the loop is forcibly broken at
2967    runtime if any iteration does actually match an empty substring.
2968    
2969    10. Include the RunTest script in the distribution.
2970    
2971    11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2972    discrepancies, some of which were old and were also with respect to 5.004. They
2973    have now been fixed.
2974    
2975    
2976    Version 1.09 28-Apr-98
2977    ----------------------
2978    
2979    1. A negated single character class followed by a quantifier with a minimum
2980    value of one (e.g.  [^x]{1,6}  ) was not compiled correctly. This could lead to
2981    program crashes, or just wrong answers. This did not apply to negated classes
2982    containing more than one character, or to minima other than one.
2983    
2984    
2985    Version 1.08 27-Mar-98
2986    ----------------------
2987    
2988    1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2989    
2990    2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2991    latter must appear before anything that relies on it in the pattern.
2992    
2993    
2994    Version 1.07 16-Feb-98
2995    ----------------------
2996    
2997    1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2998    repeat of a potentially empty string).
2999    
3000    
3001    Version 1.06 23-Jan-98
3002    ----------------------
3003    
3004    1. Added Markus Oberhumer's little patches for C++.
3005    
3006    2. Literal strings longer than 255 characters were broken.
3007    
3008    
3009    Version 1.05 23-Dec-97
3010    ----------------------
3011    
3012    1. Negated character classes containing more than one character were failing if
3013    PCRE_CASELESS was set at run time.
3014    
3015    
3016    Version 1.04 19-Dec-97
3017    ----------------------
3018    
3019    1. Corrected the man page, where some "const" qualifiers had been omitted.
3020    
3021    2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3022    input syntax.
3023    
3024    3. Fixed memory leak which occurred when a regex with back references was
3025    matched with an offsets vector that wasn't big enough. The temporary memory
3026    that is used in this case wasn't being freed if the match failed.
3027    
3028    4. Tidied pcretest to ensure it frees memory that it gets.
3029    
3030    5. Temporary memory was being obtained in the case where the passed offsets
3031    vector was exactly big enough.
3032    
3033    6. Corrected definition of offsetof() from change 5 below.
3034    
3035    7. I had screwed up change 6 below and broken the rules for the use of
3036    setjmp(). Now fixed.
3037    
3038    
3039  Version 1.03 18-Dec-97  Version 1.03 18-Dec-97
3040  ----------------------  ----------------------
# Line 39  optimized code for single-character nega Line 3073  optimized code for single-character nega
3073  from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.  from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3074    
3075  7. Constructs such as \d{8,} were compiling into the equivalent of  7. Constructs such as \d{8,} were compiling into the equivalent of
3076  \d{8}\d{65527} instead of \d{8}\d* which didn't make much difference to the  \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3077  outcome, but in this particular case used more store than had been allocated,  outcome, but in this particular case used more store than had been allocated,
3078  which caused the bug to be discovered because it threw up an internal error.  which caused the bug to be discovered because it threw up an internal error.
3079    

Legend:
Removed from v.9  
changed lines
  Added in v.341

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12