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

Diff of /code/trunk/ChangeLog

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

revision 87 by nigel, Sat Feb 24 21:41:21 2007 UTC revision 232 by ph10, Tue Sep 11 11:18:50 2007 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 7.4 10-Sep-07
5    ---------------------
6    
7    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
8        means that a class such as [\s] counted as "explicit reference to CR or
9        LF". That isn't really right - the whole point of the change was to try to
10        help when there was an actual mention of one of the two characters. So now
11        the change happens only if \r or \n (or a literal CR or LF) character is
12        encountered.
13    
14    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
15        of both had grown to the point where there were only 3 bits left.
16        Fortunately, there was spare space in the data structure, and so I have
17        moved the internal flags into a new 16-bit field to free up more option
18        bits.
19    
20    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
21        but did not set the internal JCHANGED flag - either of these is enough to
22        control the way the "get" function works - but the PCRE_INFO_JCHANGED
23        facility is supposed to tell if (?J) was ever used, so now (?J) at the
24        start sets both bits.
25    
26    4.  Added options (at build time, compile time, exec time) to change \R from
27        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
28    
29    5.  doc/pcresyntax.html was missing from the distribution.
30    
31    
32    Version 7.3 28-Aug-07
33    ---------------------
34    
35     1. In the rejigging of the build system that eventually resulted in 7.1, the
36        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
37        brackets there is not right, since it causes compilers to look for an
38        installed pcre.h, not the version that is in the source that is being
39        compiled (which of course may be different). I have changed it back to:
40    
41          #include "pcre.h"
42    
43        I have a vague recollection that the change was concerned with compiling in
44        different directories, but in the new build system, that is taken care of
45        by the VPATH setting the Makefile.
46    
47     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
48        when the subject happened to end in the byte 0x85 (e.g. if the last
49        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
50        characters but of course it shouldn't be taken as a newline when it is part
51        of another character. The bug was that, for an unlimited repeat of . in
52        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
53        characters when looking for a newline.
54    
55     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
56    
57     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
58        in debug output.
59    
60     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
61        long printing in the pcrecpp unittest when running under MinGW.
62    
63     6. ESC_K was left out of the EBCDIC table.
64    
65     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
66        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
67        limit also applies to "virtual nesting" when a pattern is recursive, and in
68        this case 1000 isn't so big. I have been able to remove this limit at the
69        expense of backing off one optimization in certain circumstances. Normally,
70        when pcre_exec() would call its internal match() function recursively and
71        immediately return the result unconditionally, it uses a "tail recursion"
72        feature to save stack. However, when a subpattern that can match an empty
73        string has an unlimited repetition quantifier, it no longer makes this
74        optimization. That gives it a stack frame in which to save the data for
75        checking that an empty string has been matched. Previously this was taken
76        from the 1000-entry workspace that had been reserved. So now there is no
77        explicit limit, but more stack is used.
78    
79     8. Applied Daniel's patches to solve problems with the import/export magic
80        syntax that is required for Windows, and which was going wrong for the
81        pcreposix and pcrecpp parts of the library. These were overlooked when this
82        problem was solved for the main library.
83    
84     9. There were some crude static tests to avoid integer overflow when computing
85        the size of patterns that contain repeated groups with explicit upper
86        limits. As the maximum quantifier is 65535, the maximum group length was
87        set at 30,000 so that the product of these two numbers did not overflow a
88        32-bit integer. However, it turns out that people want to use groups that
89        are longer than 30,000 bytes (though not repeat them that many times).
90        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
91        made it possible to implement the integer overflow checks in a much more
92        dynamic way, which I have now done. The artificial limitation on group
93        length has been removed - we now have only the limit on the total length of
94        the compiled pattern, which depends on the LINK_SIZE setting.
95    
96    10. Fixed a bug in the documentation for get/copy named substring when
97        duplicate names are permitted. If none of the named substrings are set, the
98        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
99        empty string.
100    
101    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
102        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
103        because the ] is interpreted as the first data character and the
104        terminating ] is not found. PCRE has been made compatible with Perl in this
105        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
106        cause memory overwriting.
107    
108    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
109        string has been matched (to stop an infinite loop). It was not recognizing
110        a conditional subpattern that could match an empty string if that
111        subpattern was within another subpattern. For example, it looped when
112        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
113        condition was not nested. This bug has been fixed.
114    
115    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
116        past the start of the subject in the presence of bytes with the top bit
117        set, for example "\x8aBCD".
118    
119    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
120        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
121    
122    14. Optimized (?!) to (*FAIL).
123    
124    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
125        This restricts code points to be within the range 0 to 0x10FFFF, excluding
126        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
127        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
128        does: it's just the validity check that is more restrictive.
129    
130    16. Inserted checks for integer overflows during escape sequence (backslash)
131        processing, and also fixed erroneous offset values for syntax errors during
132        backslash processing.
133    
134    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
135        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
136    
137    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
138        caused an overrun.
139    
140    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
141        something other than just ASCII characters) inside a group that had an
142        unlimited repeat caused a loop at compile time (while checking to see
143        whether the group could match an empty string).
144    
145    20. Debugging a pattern containing \p or \P could cause a crash. For example,
146        [\P{Any}] did so. (Error in the code for printing property names.)
147    
148    21. An orphan \E inside a character class could cause a crash.
149    
150    22. A repeated capturing bracket such as (A)? could cause a wild memory
151        reference during compilation.
152    
153    23. There are several functions in pcre_compile() that scan along a compiled
154        expression for various reasons (e.g. to see if it's fixed length for look
155        behind). There were bugs in these functions when a repeated \p or \P was
156        present in the pattern. These operators have additional parameters compared
157        with \d, etc, and these were not being taken into account when moving along
158        the compiled data. Specifically:
159    
160        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
161            length.
162    
163        (b) An item such as \pL+ within a repeated group could cause crashes or
164            loops.
165    
166        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
167            "reference to non-existent subpattern" error.
168    
169        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
170    
171    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
172        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
173    
174    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
175    
176    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
177        character were causing crashes (broken optimization).
178    
179    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
180        \p or \P) caused a compile-time loop.
181    
182    28. More problems have arisen in unanchored patterns when CRLF is a valid line
183        break. For example, the unstudied pattern [\r\n]A does not match the string
184        "\r\nA" because change 7.0/46 below moves the current point on by two
185        characters after failing to match at the start. However, the pattern \nA
186        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
187        the same is true. There doesn't seem any very clean way out of this, but
188        what I have chosen to do makes the common cases work: PCRE now takes note
189        of whether there can be an explicit match for \r or \n anywhere in the
190        pattern, and if so, 7.0/46 no longer applies. As part of this change,
191        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
192        pattern has explicit CR or LF references.
193    
194    29. Added (*CR) etc for changing newline setting at start of pattern.
195    
196    
197    Version 7.2 19-Jun-07
198    ---------------------
199    
200     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
201        which is apparently normally available under Windows.
202    
203     2. Re-jig the pcregrep tests with different newline settings in an attempt
204        to make them independent of the local environment's newline setting.
205    
206     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
207    
208     4. Some of the "internals" tests were previously cut out when the link size
209        was not 2, because the output contained actual offsets. The recent new
210        "Z" feature of pcretest means that these can be cut out, making the tests
211        usable with all link sizes.
212    
213     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
214        stack recursion. This gives a massive performance boost under BSD, but just
215        a small improvement under Linux. However, it saves one field in the frame
216        in all cases.
217    
218     6. Added more features from the forthcoming Perl 5.10:
219    
220        (a) (?-n) (where n is a string of digits) is a relative subroutine or
221            recursion call. It refers to the nth most recently opened parentheses.
222    
223        (b) (?+n) is also a relative subroutine call; it refers to the nth next
224            to be opened parentheses.
225    
226        (c) Conditions that refer to capturing parentheses can be specified
227            relatively, for example, (?(-2)... or (?(+3)...
228    
229        (d) \K resets the start of the current match so that everything before
230            is not part of it.
231    
232        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
233    
234        (f) \g{name} is another synonym - part of Perl 5.10's unification of
235            reference syntax.
236    
237        (g) (?| introduces a group in which the numbering of parentheses in each
238            alternative starts with the same number.
239    
240        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
241    
242     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
243        PCRE_INFO_JCHANGED.
244    
245     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
246        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
247        for detecting groups that can match an empty string.
248    
249     9. A pattern with a very large number of alternatives (more than several
250        hundred) was running out of internal workspace during the pre-compile
251        phase, where pcre_compile() figures out how much memory will be needed. A
252        bit of new cunning has reduced the workspace needed for groups with
253        alternatives. The 1000-alternative test pattern now uses 12 bytes of
254        workspace instead of running out of the 4096 that are available.
255    
256    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
257    
258    11. Applied patch from Google to remove an optimization that didn't quite work.
259        The report of the bug said:
260    
261          pcrecpp::RE("a*").FullMatch("aaa") matches, while
262          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
263          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
264    
265    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
266        it matched the wrong number of bytes.
267    
268    
269    Version 7.1 24-Apr-07
270    ---------------------
271    
272     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
273        that is more "standard", making use of automake and other Autotools. There
274        is some re-arrangement of the files and adjustment of comments consequent
275        on this.
276    
277     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
278        for recursive directory scanning broke on some systems because the files
279        are not scanned in any specific order and on different systems the order
280        was different. A call to "sort" has been inserted into RunGrepTest for the
281        approprate test as a short-term fix. In the longer term there may be an
282        alternative.
283    
284     3. I had an email from Eric Raymond about problems translating some of PCRE's
285        man pages to HTML (despite the fact that I distribute HTML pages, some
286        people do their own conversions for various reasons). The problems
287        concerned the use of low-level troff macros .br and .in. I have therefore
288        removed all such uses from the man pages (some were redundant, some could
289        be replaced by .nf/.fi pairs). The 132html script that I use to generate
290        HTML has been updated to handle .nf/.fi and to complain if it encounters
291        .br or .in.
292    
293     4. Updated comments in configure.ac that get placed in config.h.in and also
294        arranged for config.h to be included in the distribution, with the name
295        config.h.generic, for the benefit of those who have to compile without
296        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
297    
298     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
299        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
300        makevp.bat for the current PCRE, using the additional files
301        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
302    
303     6. A Windows user reported a minor discrepancy with test 2, which turned out
304        to be caused by a trailing space on an input line that had got lost in his
305        copy. The trailing space was an accident, so I've just removed it.
306    
307     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
308        that is needed.
309    
310     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
311        as "const" (a) because they are and (b) because it helps the PHP
312        maintainers who have recently made a script to detect big data structures
313        in the php code that should be moved to the .rodata section. I remembered
314        to update Builducptable as well, so it won't revert if ucptable.h is ever
315        re-created.
316    
317     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
318        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
319        order to be able to cut out the UTF-8 tables in the latter when UTF-8
320        support is not required. This saves 1.5-2K of code, which is important in
321        some applications.
322    
323        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
324        so as not to refer to the tables, even though these functions will never be
325        called when UTF-8 support is disabled. Otherwise there are problems with a
326        shared library.
327    
328    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
329    
330        (a) It was defining its arguments as char * instead of void *.
331    
332        (b) It was assuming that all moves were upwards in memory; this was true
333            a long time ago when I wrote it, but is no longer the case.
334    
335        The emulated memove() is provided for those environments that have neither
336        memmove() nor bcopy(). I didn't think anyone used it these days, but that
337        is clearly not the case, as these two bugs were recently reported.
338    
339    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
340        and Detrail to create the HTML documentation, the .txt form of the man
341        pages, and it removes trailing spaces from listed files. It also creates
342        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
343        case, it wraps all the #defines with #ifndefs. This script should be run
344        before "make dist".
345    
346    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
347        with Unicode property support.
348    
349        (a) For a maximizing quantifier, if the two different cases of the
350            character were of different lengths in their UTF-8 codings (there are
351            some cases like this - I found 11), and the matching function had to
352            back up over a mixture of the two cases, it incorrectly assumed they
353            were both the same length.
354    
355        (b) When PCRE was configured to use the heap rather than the stack for
356            recursion during matching, it was not correctly preserving the data for
357            the other case of a UTF-8 character when checking ahead for a match
358            while processing a minimizing repeat. If the check also involved
359            matching a wide character, but failed, corruption could cause an
360            erroneous result when trying to check for a repeat of the original
361            character.
362    
363    13. Some tidying changes to the testing mechanism:
364    
365        (a) The RunTest script now detects the internal link size and whether there
366            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
367            values substituted by "configure". (The RunGrepTest script already did
368            this for UTF-8.) The configure.ac script no longer substitutes the
369            relevant variables.
370    
371        (b) The debugging options /B and /D in pcretest show the compiled bytecode
372            with length and offset values. This means that the output is different
373            for different internal link sizes. Test 2 is skipped for link sizes
374            other than 2 because of this, bypassing the problem. Unfortunately,
375            there was also a test in test 3 (the locale tests) that used /B and
376            failed for link sizes other than 2. Rather than cut the whole test out,
377            I have added a new /Z option to pcretest that replaces the length and
378            offset values with spaces. This is now used to make test 3 independent
379            of link size. (Test 2 will be tidied up later.)
380    
381    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
382        segmentation fault instead of returning the appropriate error message.
383    
384    15. In multiline mode when the newline sequence was set to "any", the pattern
385        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
386        This doesn't seem right; it now treats the CRLF combination as the line
387        ending, and so does not match in that case. It's only a pattern such as ^$
388        that would hit this one: something like ^ABC$ would have failed after \r
389        and then tried again after \r\n.
390    
391    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
392        in an attempt to make files that differ only in their line terminators
393        compare equal. This works on Linux.
394    
395    17. Under certain error circumstances pcregrep might try to free random memory
396        as it exited. This is now fixed, thanks to valgrind.
397    
398    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
399        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
400        was because its rules for how to advance for /g after matching an empty
401        string at the end of a line did not allow for this case. They now check for
402        it specially.
403    
404    20. pcretest is supposed to handle patterns and data of any length, by
405        extending its buffers when necessary. It was getting this wrong when the
406        buffer for a data line had to be extended.
407    
408    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
409        CRLF as a newline sequence.
410    
411    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
412        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
413        I have nevertheless tidied it up.
414    
415    23. Added some casts to kill warnings from HP-UX ia64 compiler.
416    
417    24. Added a man page for pcre-config.
418    
419    
420    Version 7.0 19-Dec-06
421    ---------------------
422    
423     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
424        moving to gcc 4.1.1.
425    
426     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
427        sys/time.h, which is documented as needed for this function. It doesn't
428        seem to matter on Linux, but it showed up on some releases of OS X.
429    
430     3. It seems that there are systems where bytes whose values are greater than
431        127 match isprint() in the "C" locale. The "C" locale should be the
432        default when a C program starts up. In most systems, only ASCII printing
433        characters match isprint(). This difference caused the output from pcretest
434        to vary, making some of the tests fail. I have changed pcretest so that:
435    
436        (a) When it is outputting text in the compiled version of a pattern, bytes
437            other than 32-126 are always shown as hex escapes.
438    
439        (b) When it is outputting text that is a matched part of a subject string,
440            it does the same, unless a different locale has been set for the match
441            (using the /L modifier). In this case, it uses isprint() to decide.
442    
443     4. Fixed a major bug that caused incorrect computation of the amount of memory
444        required for a compiled pattern when options that changed within the
445        pattern affected the logic of the preliminary scan that determines the
446        length. The relevant options are -x, and -i in UTF-8 mode. The result was
447        that the computed length was too small. The symptoms of this bug were
448        either the PCRE error "internal error: code overflow" from pcre_compile(),
449        or a glibc crash with a message such as "pcretest: free(): invalid next
450        size (fast)". Examples of patterns that provoked this bug (shown in
451        pcretest format) are:
452    
453          /(?-x: )/x
454          /(?x)(?-x: \s*#\s*)/
455          /((?i)[\x{c0}])/8
456          /(?i:[\x{c0}])/8
457    
458        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
459        is now done differently.
460    
461     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
462        wrapper classes; (b) implement a new function in the C++ scanner that is
463        more efficient than the old way of doing things because it avoids levels of
464        recursion in the regex matching; (c) add a paragraph to the documentation
465        for the FullMatch() function.
466    
467     6. The escape sequence \n was being treated as whatever was defined as
468        "newline". Not only was this contrary to the documentation, which states
469        that \n is character 10 (hex 0A), but it also went horribly wrong when
470        "newline" was defined as CRLF. This has been fixed.
471    
472     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
473        was being set to -1 for the "end of line" case (supposedly a value that no
474        character can have). Though this value is never used (the check for end of
475        line is "zero bytes in current character"), it caused compiler complaints.
476        I've changed it to 0xffffffff.
477    
478     8. In pcre_version.c, the version string was being built by a sequence of
479        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
480        string (as it is for production releases) called a macro with an empty
481        argument. The C standard says the result of this is undefined. The gcc
482        compiler treats it as an empty string (which was what was wanted) but it is
483        reported that Visual C gives an error. The source has been hacked around to
484        avoid this problem.
485    
486     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
487        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
488        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
489        of them did).
490    
491    10. Originally, pcretest opened its input and output without "b"; then I was
492        told that "b" was needed in some environments, so it was added for release
493        5.0 to both the input and output. (It makes no difference on Unix-like
494        systems.) Later I was told that it is wrong for the input on Windows. I've
495        now abstracted the modes into two macros, to make it easier to fiddle with
496        them, and removed "b" from the input mode under Windows.
497    
498    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
499    
500    12. Added -help and --help to pcretest as an official way of being reminded
501        of the options.
502    
503    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
504        and pcrecpp.cc because they annoy compilers at high warning levels.
505    
506    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
507    
508    15. Fixed an occurrence of == in configure.ac that should have been = (shell
509        scripts are not C programs :-) and which was not noticed because it works
510        on Linux.
511    
512    16. pcretest is supposed to handle any length of pattern and data line (as one
513        line or as a continued sequence of lines) by extending its input buffer if
514        necessary. This feature was broken for very long pattern lines, leading to
515        a string of junk being passed to pcre_compile() if the pattern was longer
516        than about 50K.
517    
518    17. I have done a major re-factoring of the way pcre_compile() computes the
519        amount of memory needed for a compiled pattern. Previously, there was code
520        that made a preliminary scan of the pattern in order to do this. That was
521        OK when PCRE was new, but as the facilities have expanded, it has become
522        harder and harder to keep it in step with the real compile phase, and there
523        have been a number of bugs (see for example, 4 above). I have now found a
524        cunning way of running the real compile function in a "fake" mode that
525        enables it to compute how much memory it would need, while actually only
526        ever using a few hundred bytes of working memory and without too many
527        tests of the mode. This should make future maintenance and development
528        easier. A side effect of this work is that the limit of 200 on the nesting
529        depth of parentheses has been removed (though this was never a serious
530        limitation, I suspect). However, there is a downside: pcre_compile() now
531        runs more slowly than before (30% or more, depending on the pattern). I
532        hope this isn't a big issue. There is no effect on runtime performance.
533    
534    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
535        newline (only possible for the last line of a file) and it was a
536        pattern that set a locale (followed by /Lsomething), pcretest crashed.
537    
538    19. Added additional timing features to pcretest. (1) The -tm option now times
539        matching only, not compiling. (2) Both -t and -tm can be followed, as a
540        separate command line item, by a number that specifies the number of
541        repeats to use when timing. The default is 50000; this gives better
542        precision, but takes uncomfortably long for very large patterns.
543    
544    20. Extended pcre_study() to be more clever in cases where a branch of a
545        subpattern has no definite first character. For example, (a*|b*)[cd] would
546        previously give no result from pcre_study(). Now it recognizes that the
547        first character must be a, b, c, or d.
548    
549    21. There was an incorrect error "recursive call could loop indefinitely" if
550        a subpattern (or the entire pattern) that was being tested for matching an
551        empty string contained only one non-empty item after a nested subpattern.
552        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
553        incorrectly, because the \d was being skipped in the check.
554    
555    22. The pcretest program now has a new pattern option /B and a command line
556        option -b, which is equivalent to adding /B to every pattern. This causes
557        it to show the compiled bytecode, without the additional information that
558        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
559        is the same as /B/I).
560    
561    23. A new optimization is now able automatically to treat some sequences such
562        as a*b as a*+b. More specifically, if something simple (such as a character
563        or a simple class like \d) has an unlimited quantifier, and is followed by
564        something that cannot possibly match the quantified thing, the quantifier
565        is automatically "possessified".
566    
567    24. A recursive reference to a subpattern whose number was greater than 39
568        went wrong under certain circumstances in UTF-8 mode. This bug could also
569        have affected the operation of pcre_study().
570    
571    25. Realized that a little bit of performance could be had by replacing
572        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
573    
574    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
575    
576    27. Possessive quantifiers such as a++ were previously implemented by turning
577        them into atomic groups such as ($>a+). Now they have their own opcodes,
578        which improves performance. This includes the automatically created ones
579        from 23 above.
580    
581    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
582        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
583        the first matched character to be a colon. This applied both to named and
584        numbered groups.
585    
586    29. The ucpinternal.h header file was missing its idempotency #ifdef.
587    
588    30. I was sent a "project" file called libpcre.a.dev which I understand makes
589        building PCRE on Windows easier, so I have included it in the distribution.
590    
591    31. There is now a check in pcretest against a ridiculously large number being
592        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
593        loop, the loop is abandoned.
594    
595    32. Forward references to subpatterns in conditions such as (?(2)...) where
596        subpattern 2 is defined later cause pcre_compile() to search forwards in
597        the pattern for the relevant set of parentheses. This search went wrong
598        when there were unescaped parentheses in a character class, parentheses
599        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
600    
601    33. "Subroutine" calls and backreferences were previously restricted to
602        referencing subpatterns earlier in the regex. This restriction has now
603        been removed.
604    
605    34. Added a number of extra features that are going to be in Perl 5.10. On the
606        whole, these are just syntactic alternatives for features that PCRE had
607        previously implemented using the Python syntax or my own invention. The
608        other formats are all retained for compatibility.
609    
610        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
611            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
612            also .NET compatible.
613    
614        (b) A recursion or subroutine call to a named group can now be defined as
615            (?&name) as well as (?P>name).
616    
617        (c) A backreference to a named group can now be defined as \k<name> or
618            \k'name' as well as (?P=name). The new forms, as well as being in Perl
619            5.10, are also .NET compatible.
620    
621        (d) A conditional reference to a named group can now use the syntax
622            (?(<name>) or (?('name') as well as (?(name).
623    
624        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
625            groups (named and numbered) that are never evaluated inline, but can be
626            called as "subroutines" from elsewhere. In effect, the DEFINE condition
627            is always false. There may be only one alternative in such a group.
628    
629        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
630            as the simple (?(R). The condition is true only if the most recent
631            recursion is that of the given number or name. It does not search out
632            through the entire recursion stack.
633    
634        (g) The escape \gN or \g{N} has been added, where N is a positive or
635            negative number, specifying an absolute or relative reference.
636    
637    35. Tidied to get rid of some further signed/unsigned compiler warnings and
638        some "unreachable code" warnings.
639    
640    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
641        things, this adds five new scripts.
642    
643    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
644        There were also incompatibilities regarding the handling of \Q..\E inside
645        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
646        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
647    
648    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
649        matches an empty string, and forcibly breaks the loop. There were bugs in
650        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
651        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
652        separate and independent bugs (that affected different cases) have been
653        fixed.
654    
655    39. Refactored the code to abolish the use of different opcodes for small
656        capturing bracket numbers. This is a tidy that I avoided doing when I
657        removed the limit on the number of capturing brackets for 3.5 back in 2001.
658        The new approach is not only tidier, it makes it possible to reduce the
659        memory needed to fix the previous bug (38).
660    
661    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
662        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
663        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
664        mode.
665    
666    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
667        report.
668    
669    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
670        copy construction and assignment in the C++ wrapper.
671    
672    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
673        couple of bugs that could have given wrong results in the "--newline=crlf"
674        case.
675    
676    44. Added a number of casts and did some reorganization of signed/unsigned int
677        variables following suggestions from Dair Grant. Also renamed the variable
678        "this" as "item" because it is a C++ keyword.
679    
680    45. Arranged for dftables to add
681    
682          #include "pcre_internal.h"
683    
684        to pcre_chartables.c because without it, gcc 4.x may remove the array
685        definition from the final binary if PCRE is built into a static library and
686        dead code stripping is activated.
687    
688    46. For an unanchored pattern, if a match attempt fails at the start of a
689        newline sequence, and the newline setting is CRLF or ANY, and the next two
690        characters are CRLF, advance by two characters instead of one.
691    
692    
693    Version 6.7 04-Jul-06
694    ---------------------
695    
696     1. In order to handle tests when input lines are enormously long, pcretest has
697        been re-factored so that it automatically extends its buffers when
698        necessary. The code is crude, but this _is_ just a test program. The
699        default size has been increased from 32K to 50K.
700    
701     2. The code in pcre_study() was using the value of the re argument before
702        testing it for NULL. (Of course, in any sensible call of the function, it
703        won't be NULL.)
704    
705     3. The memmove() emulation function in pcre_internal.h, which is used on
706        systems that lack both memmove() and bcopy() - that is, hardly ever -
707        was missing a "static" storage class specifier.
708    
709     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
710        containing an extended class (one that cannot be represented by a bitmap
711        because it contains high-valued characters or Unicode property items, e.g.
712        [\pZ]). Almost always one would set UTF-8 mode when processing such a
713        pattern, but PCRE should not loop if you do not (it no longer does).
714        [Detail: two cases were found: (a) a repeated subpattern containing an
715        extended class; (b) a recursive reference to a subpattern that followed a
716        previous extended class. It wasn't skipping over the extended class
717        correctly when UTF-8 mode was not set.]
718    
719     5. A negated single-character class was not being recognized as fixed-length
720        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
721        compile error "lookbehind assertion is not fixed length".
722    
723     6. The RunPerlTest auxiliary script was showing an unexpected difference
724        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
725        write a Perl script that can interpret lines of an input file either as
726        byte characters or as UTF-8, which is what "perltest" was being required to
727        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
728        can't do is switch easily at run time between having the "use utf8;" pragma
729        or not. In the end, I fudged it by using the RunPerlTest script to insert
730        "use utf8;" explicitly for the UTF-8 tests.
731    
732     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
733        the end of the subject string, contrary to the documentation and to what
734        Perl does. This was true of both matching functions. Now it matches only at
735        the start of the subject and immediately after *internal* newlines.
736    
737     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
738        a pointer to an int instead of a pointer to an unsigned long int. This
739        caused problems on 64-bit systems.
740    
741     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
742        instance of the 'standard' template library not being so standard".
743    
744    10. There was no check on the number of named subpatterns nor the maximum
745        length of a subpattern name. The product of these values is used to compute
746        the size of the memory block for a compiled pattern. By supplying a very
747        long subpattern name and a large number of named subpatterns, the size
748        computation could be caused to overflow. This is now prevented by limiting
749        the length of names to 32 characters, and the number of named subpatterns
750        to 10,000.
751    
752    11. Subpatterns that are repeated with specific counts have to be replicated in
753        the compiled pattern. The size of memory for this was computed from the
754        length of the subpattern and the repeat count. The latter is limited to
755        65535, but there was no limit on the former, meaning that integer overflow
756        could in principle occur. The compiled length of a repeated subpattern is
757        now limited to 30,000 bytes in order to prevent this.
758    
759    12. Added the optional facility to have named substrings with the same name.
760    
761    13. Added the ability to use a named substring as a condition, using the
762        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
763        are numbers (not recommended). Forward references are permitted.
764    
765    14. Added forward references in named backreferences (if you see what I mean).
766    
767    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
768        pattern could run off the end of the subject. For example, the pattern
769        "(?s)(.{1,5})"8 did this with the subject "ab".
770    
771    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
772        PCRE_CASELESS was set when matching characters that were quantified with ?
773        or *.
774    
775    17. A character class other than a single negated character that had a minimum
776        but no maximum quantifier - for example [ab]{6,} - was not handled
777        correctly by pce_dfa_exec(). It would match only one character.
778    
779    18. A valid (though odd) pattern that looked like a POSIX character
780        class but used an invalid character after [ (for example [[,abc,]]) caused
781        pcre_compile() to give the error "Failed: internal error: code overflow" or
782        in some cases to crash with a glibc free() error. This could even happen if
783        the pattern terminated after [[ but there just happened to be a sequence of
784        letters, a binary zero, and a closing ] in the memory that followed.
785    
786    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
787        over the years. Originally (before any Unicode support), just the bottom 8
788        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
789        output from "man perlunicode" includes this:
790    
791          The regular expression compiler produces polymorphic opcodes.  That
792          is, the pattern adapts to the data and automatically switches to
793          the Unicode character scheme when presented with Unicode data--or
794          instead uses a traditional byte scheme when presented with byte
795          data.
796    
797        Sadly, a wide octal escape does not cause a switch, and in a string with
798        no other multibyte characters, these octal escapes are treated as before.
799        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
800        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
801        Unicode string.
802    
803        I have not perpetrated such confusion in PCRE. Up till now, it took just
804        the bottom 8 bits, as in old Perl. I have now made octal escapes with
805        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
806        translate to the appropriate multibyte character.
807    
808    29. Applied some refactoring to reduce the number of warnings from Microsoft
809        and Borland compilers. This has included removing the fudge introduced
810        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
811        a warning about an unused variable.
812    
813    21. PCRE has not included VT (character 0x0b) in the set of whitespace
814        characters since release 4.0, because Perl (from release 5.004) does not.
815        [Or at least, is documented not to: some releases seem to be in conflict
816        with the documentation.] However, when a pattern was studied with
817        pcre_study() and all its branches started with \s, PCRE still included VT
818        as a possible starting character. Of course, this did no harm; it just
819        caused an unnecessary match attempt.
820    
821    22. Removed a now-redundant internal flag bit that recorded the fact that case
822        dependency changed within the pattern. This was once needed for "required
823        byte" processing, but is no longer used. This recovers a now-scarce options
824        bit. Also moved the least significant internal flag bit to the most-
825        significant bit of the word, which was not previously used (hangover from
826        the days when it was an int rather than a uint) to free up another bit for
827        the future.
828    
829    23. Added support for CRLF line endings as well as CR and LF. As well as the
830        default being selectable at build time, it can now be changed at runtime
831        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
832        specify that it is scanning data with non-default line endings.
833    
834    24. Changed the definition of CXXLINK to make it agree with the definition of
835        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
836    
837    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
838        recursions. This makes a big different to stack usage for some patterns.
839    
840    26. If a subpattern containing a named recursion or subroutine reference such
841        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
842        the space required for the compiled pattern went wrong and gave too small a
843        value. Depending on the environment, this could lead to "Failed: internal
844        error: code overflow at offset 49" or "glibc detected double free or
845        corruption" errors.
846    
847    27. Applied patches from Google (a) to support the new newline modes and (b) to
848        advance over multibyte UTF-8 characters in GlobalReplace.
849    
850    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
851        difference for some implementation of PCRE in some Windows version.
852    
853    29. Added some extra testing facilities to pcretest:
854    
855        \q<number>   in a data line sets the "match limit" value
856        \Q<number>   in a data line sets the "match recursion limt" value
857        -S <number>  sets the stack size, where <number> is in megabytes
858    
859        The -S option isn't available for Windows.
860    
861    
862    Version 6.6 06-Feb-06
863    ---------------------
864    
865     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
866        in pcreposix.h. I have copied the definition from pcre.h.
867    
868     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
869        because pcre.h is no longer a built file.
870    
871     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
872        not normally included in the compiled code.
873    
874    
875  Version 6.5 01-Feb-06  Version 6.5 01-Feb-06
876  ---------------------  ---------------------
877    

Legend:
Removed from v.87  
changed lines
  Added in v.232

  ViewVC Help
Powered by ViewVC 1.1.5