/[pcre]/code/branches/oldtrunk8/ChangeLog
ViewVC logotype

Diff of /code/branches/oldtrunk8/ChangeLog

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

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

Legend:
Removed from v.59  
changed lines
  Added in v.312

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12