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

Diff of /code/trunk/ChangeLog

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

revision 81 by nigel, Sat Feb 24 21:40:59 2007 UTC revision 361 by ph10, Thu Jul 10 16:03:28 2008 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.0 02 Jul-08
5    ---------------------
6    
7    1.  Replaced UCP searching code with optimized version as implemented for Ad
8        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
9        stage table and inline lookup instead of a function, giving speed ups of 2
10        to 5 times on some simple patterns that I tested. Permission was given to
11        distribute the MultiStage2.py script that generates the tables (it's not in
12        the tarball, but is in the Subversion repository).
13    
14    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
15        scripts.
16    
17    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
18        a group with a zero qualifier. The result of the study could be incorrect,
19        or the function might crash, depending on the pattern.
20    
21    4.  Caseless matching was not working for non-ASCII characters in back
22        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
23        It now works when Unicode Property Support is available.
24    
25    5.  In pcretest, an escape such as \x{de} in the data was always generating
26        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
27        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
28        truncation.
29    
30    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
31    
32    7.  Added two (int) casts to pcregrep when printing the difference of two
33        pointers, in case they are 64-bit values.
34    
35    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
36        test 2 if it fails.
37    
38    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
39        and a #define of that name to empty if it is not externally set. This is to
40        allow users of MSVC to set it if necessary.
41    
42    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
43        the convenience functions in the pcre_get.c source file.
44    
45    11. An option change at the start of a pattern that had top-level alternatives
46        could cause overwriting and/or a crash. This command provoked a crash in
47        some environments:
48    
49          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
50    
51        This potential security problem was recorded as CVE-2008-2371.
52    
53    12. For a pattern where the match had to start at the beginning or immediately
54        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
55        pcre_dfa_exec() could read past the end of the passed subject if there was
56        no match. To help with detecting such bugs (e.g. with valgrind), I modified
57        pcretest so that it places the subject at the end of its malloc-ed buffer.
58    
59    
60    Version 7.7 07-May-08
61    ---------------------
62    
63    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
64        a string to a long long, pretend we don't even have a long long." This is
65        done by checking for the strtoq, strtoll, and _strtoi64 functions.
66    
67    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
68        pre-7.6 versions, which defined a global no_arg variable instead of putting
69        it in the RE class. (See also #8 below.)
70    
71    3.  Remove a line of dead code, identified by coverity and reported by Nuno
72        Lopes.
73    
74    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
75    
76        (1) The include/exclude patterns were being applied to the whole pathnames
77            of files, instead of just to the final components.
78    
79        (2) If there was more than one level of directory, the subdirectories were
80            skipped unless they satisfied the include/exclude conditions. This is
81            inconsistent with GNU grep (and could even be seen as contrary to the
82            pcregrep specification - which I improved to make it absolutely clear).
83            The action now is always to scan all levels of directory, and just
84            apply the include/exclude patterns to regular files.
85    
86    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
87        --exclude_dir in the tests to avoid scanning .svn directories.
88    
89    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
90        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
91        doesn't support NULs in patterns.
92    
93    7.  Added some missing "const"s to declarations of static tables in
94        pcre_compile.c and pcre_dfa_exec.c.
95    
96    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
97        caused by fix #2  above. (Subsequently also a second patch to fix the
98        first patch. And a third patch - this was a messy problem.)
99    
100    9.  Applied Craig's patch to remove the use of push_back().
101    
102    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
103        matching function regexec().
104    
105    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
106        which, however, unlike Perl's \g{...}, are subroutine calls, not back
107        references. PCRE supports relative numbers with this syntax (I don't think
108        Oniguruma does).
109    
110    12. Previously, a group with a zero repeat such as (...){0} was completely
111        omitted from the compiled regex. However, this means that if the group
112        was called as a subroutine from elsewhere in the pattern, things went wrong
113        (an internal error was given). Such groups are now left in the compiled
114        pattern, with a new opcode that causes them to be skipped at execution
115        time.
116    
117    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
118        to the way PCRE behaves:
119    
120        (a) A lone ] character is dis-allowed (Perl treats it as data).
121    
122        (b) A back reference to an unmatched subpattern matches an empty string
123            (Perl fails the current match path).
124    
125        (c) A data ] in a character class must be notated as \] because if the
126            first data character in a class is ], it defines an empty class. (In
127            Perl it is not possible to have an empty class.) The empty class []
128            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
129            The negative empty class [^] matches any one character, independently
130            of the DOTALL setting.
131    
132    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
133        non-existent subpattern following a character class starting with ']' and
134        containing () gave an internal compiling error instead of "reference to
135        non-existent subpattern". Fortunately, when the pattern did exist, the
136        compiled code was correct. (When scanning forwards to check for the
137        existencd of the subpattern, it was treating the data ']' as terminating
138        the class, so got the count wrong. When actually compiling, the reference
139        was subsequently set up correctly.)
140    
141    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
142        it was being rejected as not supported by pcre_dfa_exec(), even though
143        other assertions are supported. I have made pcre_dfa_exec() support
144        (*FAIL).
145    
146    16. The implementation of 13c above involved the invention of a new opcode,
147        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
148        cannot be changed at match time, I realized I could make a small
149        improvement to matching performance by compiling OP_ALLANY instead of
150        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
151        on the OP_ANY path.
152    
153    17. Compiling pcretest on Windows with readline support failed without the
154        following two fixes: (1) Make the unistd.h include conditional on
155        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
156    
157    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
158        ncurses library to be included for pcretest when ReadLine support is
159        requested, but also to allow for it to be overridden. This patch came from
160        Daniel Bergström.
161    
162    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
163        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
164        any errors with the current Unicode tables. Thanks to Peter Kankowski for
165        spotting this.
166    
167    
168    Version 7.6 28-Jan-08
169    ---------------------
170    
171    1.  A character class containing a very large number of characters with
172        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
173        overflow.
174    
175    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
176        HAVE_LONG_LONG is not defined.
177    
178    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
179        bring them up to date and include new features. This patch includes:
180    
181        - Fixed PH's badly added libz and libbz2 support.
182        - Fixed a problem with static linking.
183        - Added pcredemo. [But later removed - see 7 below.]
184        - Fixed dftables problem and added an option.
185        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
186            HAVE_LONG_LONG.
187        - Added readline support for pcretest.
188        - Added an listing of the option settings after cmake has run.
189    
190    4.  A user submitted a patch to Makefile that makes it easy to create
191        "pcre.dll" under mingw when using Configure/Make. I added stuff to
192        Makefile.am that cause it to include this special target, without
193        affecting anything else. Note that the same mingw target plus all
194        the other distribution libraries and programs are now supported
195        when configuring with CMake (see 6 below) instead of with
196        Configure/Make.
197    
198    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
199        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
200        exported in the Windows port". It has not yet been confirmed that the patch
201        solves the problem, but it does no harm.
202    
203    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
204        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
205        with CMake, and also correct the comment about stack recursion.
206    
207    7.  Remove the automatic building of pcredemo from the ./configure system and
208        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
209        of a program that users should build themselves after PCRE is installed, so
210        building it automatically is not really right. What is more, it gave
211        trouble in some build environments.
212    
213    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
214    
215    
216    Version 7.5 10-Jan-08
217    ---------------------
218    
219    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
220        values in parens when parsing an RE using the C++ wrapper."
221    
222    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
223        Characters greater than 255 were excluded from the class instead of being
224        included.
225    
226    3.  The same bug as (2) above applied to negated POSIX classes such as
227        [:^space:].
228    
229    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
230        defined or documented. It seems to have been a typo for PCRE_STATIC, so
231        I have changed it.
232    
233    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
234        first named subpattern) and a construct such as (?&a) would reference the
235        first named subpattern whose name started with "a" (in other words, the
236        length check was missing). Both these problems are fixed. "Subpattern name
237        expected" is now given for (?&) (a zero-length name), and this patch also
238        makes it give the same error for \k'' (previously it complained that that
239        was a reference to a non-existent subpattern).
240    
241    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
242        this is right because (?- can be followed by option settings as well as by
243        digits. I have, however, made the messages clearer.
244    
245    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
246        than the number used in the conditional) now cause a compile-time error.
247        This is actually not compatible with Perl, which accepts such patterns, but
248        treats the conditional as always being FALSE (as PCRE used to), but it
249        seems to me that giving a diagnostic is better.
250    
251    8.  Change "alphameric" to the more common word "alphanumeric" in comments
252        and messages.
253    
254    9.  Fix two occurrences of "backslash" in comments that should have been
255        "backspace".
256    
257    10. Remove two redundant lines of code that can never be obeyed (their function
258        was moved elsewhere).
259    
260    11. The program that makes PCRE's Unicode character property table had a bug
261        which caused it to generate incorrect table entries for sequences of
262        characters that have the same character type, but are in different scripts.
263        It amalgamated them into a single range, with the script of the first of
264        them. In other words, some characters were in the wrong script. There were
265        thirteen such cases, affecting characters in the following ranges:
266    
267          U+002b0 - U+002c1
268          U+0060c - U+0060d
269          U+0061e - U+00612
270          U+0064b - U+0065e
271          U+0074d - U+0076d
272          U+01800 - U+01805
273          U+01d00 - U+01d77
274          U+01d9b - U+01dbf
275          U+0200b - U+0200f
276          U+030fc - U+030fe
277          U+03260 - U+0327f
278          U+0fb46 - U+0fbb1
279          U+10450 - U+1049d
280    
281    12. The -o option (show only the matching part of a line) for pcregrep was not
282        compatible with GNU grep in that, if there was more than one match in a
283        line, it showed only the first of them. It now behaves in the same way as
284        GNU grep.
285    
286    13. If the -o and -v options were combined for pcregrep, it printed a blank
287        line for every non-matching line. GNU grep prints nothing, and pcregrep now
288        does the same. The return code can be used to tell if there were any
289        non-matching lines.
290    
291    14. Added --file-offsets and --line-offsets to pcregrep.
292    
293    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
294        infinitely looping recursion. The bug was that positive lookaheads were not
295        being skipped when checking for a possible empty match (negative lookaheads
296        and both kinds of lookbehind were skipped).
297    
298    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
299        inclusion of <windows.h> to before rather than after the definition of
300        INVALID_FILE_ATTRIBUTES (patch from David Byron).
301    
302    17. Specifying a possessive quantifier with a specific limit for a Unicode
303        character property caused pcre_compile() to compile bad code, which led at
304        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
305        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
306        caused the error; without that there was no problem.
307    
308    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
309    
310    19. Added --enable-pcretest-libreadline.
311    
312    20. In pcrecpp.cc, the variable 'count' was incremented twice in
313        RE::GlobalReplace(). As a result, the number of replacements returned was
314        double what it should be. I removed one of the increments, but Craig sent a
315        later patch that removed the other one (the right fix) and added unit tests
316        that check the return values (which was not done before).
317    
318    21. Several CMake things:
319    
320        (1) Arranged that, when cmake is used on Unix, the libraries end up with
321            the names libpcre and libpcreposix, not just pcre and pcreposix.
322    
323        (2) The above change means that pcretest and pcregrep are now correctly
324            linked with the newly-built libraries, not previously installed ones.
325    
326        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
327    
328    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
329        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
330        UTF-8 newline character). The key issue is that the pattern starts .*;
331        this means that the match must be either at the beginning, or after a
332        newline. The bug was in the code for advancing after a failed match and
333        checking that the new position followed a newline. It was not taking
334        account of UTF-8 characters correctly.
335    
336    23. PCRE was behaving differently from Perl in the way it recognized POSIX
337        character classes. PCRE was not treating the sequence [:...:] as a
338        character class unless the ... were all letters. Perl, however, seems to
339        allow any characters between [: and :], though of course it rejects as
340        unknown any "names" that contain non-letters, because all the known class
341        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
342        for example, whereas PCRE did not - it did not recognize a POSIX character
343        class. This seemed a bit dangerous, so the code has been changed to be
344        closer to Perl. The behaviour is not identical to Perl, because PCRE will
345        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
346        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
347        Perl does, and where it didn't before.
348    
349    24. Rewrite so as to remove the single use of %n from pcregrep because in some
350        Windows environments %n is disabled by default.
351    
352    
353    Version 7.4 21-Sep-07
354    ---------------------
355    
356    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
357        means that a class such as [\s] counted as "explicit reference to CR or
358        LF". That isn't really right - the whole point of the change was to try to
359        help when there was an actual mention of one of the two characters. So now
360        the change happens only if \r or \n (or a literal CR or LF) character is
361        encountered.
362    
363    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
364        of both had grown to the point where there were only 3 bits left.
365        Fortunately, there was spare space in the data structure, and so I have
366        moved the internal flags into a new 16-bit field to free up more option
367        bits.
368    
369    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
370        but did not set the internal JCHANGED flag - either of these is enough to
371        control the way the "get" function works - but the PCRE_INFO_JCHANGED
372        facility is supposed to tell if (?J) was ever used, so now (?J) at the
373        start sets both bits.
374    
375    4.  Added options (at build time, compile time, exec time) to change \R from
376        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
377    
378    5.  doc/pcresyntax.html was missing from the distribution.
379    
380    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
381        compatibility, even though it is no longer used.
382    
383    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
384        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
385        windows.h file is present (where different names are used). [This was
386        reversed later after testing - see 16 below.]
387    
388    8.  Changed all #include <config.h> to #include "config.h". There were also
389        some further <pcre.h> cases that I changed to "pcre.h".
390    
391    9.  When pcregrep was used with the --colour option, it missed the line ending
392        sequence off the lines that it output.
393    
394    10. It was pointed out to me that arrays of string pointers cause lots of
395        relocations when a shared library is dynamically loaded. A technique of
396        using a single long string with a table of offsets can drastically reduce
397        these. I have refactored PCRE in four places to do this. The result is
398        dramatic:
399    
400          Originally:                          290
401          After changing UCP table:            187
402          After changing error message table:   43
403          After changing table of "verbs"       36
404          After changing table of Posix names   22
405    
406        Thanks to the folks working on Gregex for glib for this insight.
407    
408    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
409        unicode-properties was also set.
410    
411    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
412    
413    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
414        checked only for CRLF.
415    
416    14. Added casts to pcretest.c to avoid compiler warnings.
417    
418    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
419    
420    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
421        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
422        entirely. This removes changes made in 7 above.
423    
424    17. The CMake files have been updated, and there is now more information about
425        building with CMake in the NON-UNIX-USE document.
426    
427    
428    Version 7.3 28-Aug-07
429    ---------------------
430    
431     1. In the rejigging of the build system that eventually resulted in 7.1, the
432        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
433        brackets there is not right, since it causes compilers to look for an
434        installed pcre.h, not the version that is in the source that is being
435        compiled (which of course may be different). I have changed it back to:
436    
437          #include "pcre.h"
438    
439        I have a vague recollection that the change was concerned with compiling in
440        different directories, but in the new build system, that is taken care of
441        by the VPATH setting the Makefile.
442    
443     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
444        when the subject happened to end in the byte 0x85 (e.g. if the last
445        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
446        characters but of course it shouldn't be taken as a newline when it is part
447        of another character. The bug was that, for an unlimited repeat of . in
448        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
449        characters when looking for a newline.
450    
451     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
452    
453     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
454        in debug output.
455    
456     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
457        long printing in the pcrecpp unittest when running under MinGW.
458    
459     6. ESC_K was left out of the EBCDIC table.
460    
461     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
462        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
463        limit also applies to "virtual nesting" when a pattern is recursive, and in
464        this case 1000 isn't so big. I have been able to remove this limit at the
465        expense of backing off one optimization in certain circumstances. Normally,
466        when pcre_exec() would call its internal match() function recursively and
467        immediately return the result unconditionally, it uses a "tail recursion"
468        feature to save stack. However, when a subpattern that can match an empty
469        string has an unlimited repetition quantifier, it no longer makes this
470        optimization. That gives it a stack frame in which to save the data for
471        checking that an empty string has been matched. Previously this was taken
472        from the 1000-entry workspace that had been reserved. So now there is no
473        explicit limit, but more stack is used.
474    
475     8. Applied Daniel's patches to solve problems with the import/export magic
476        syntax that is required for Windows, and which was going wrong for the
477        pcreposix and pcrecpp parts of the library. These were overlooked when this
478        problem was solved for the main library.
479    
480     9. There were some crude static tests to avoid integer overflow when computing
481        the size of patterns that contain repeated groups with explicit upper
482        limits. As the maximum quantifier is 65535, the maximum group length was
483        set at 30,000 so that the product of these two numbers did not overflow a
484        32-bit integer. However, it turns out that people want to use groups that
485        are longer than 30,000 bytes (though not repeat them that many times).
486        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
487        made it possible to implement the integer overflow checks in a much more
488        dynamic way, which I have now done. The artificial limitation on group
489        length has been removed - we now have only the limit on the total length of
490        the compiled pattern, which depends on the LINK_SIZE setting.
491    
492    10. Fixed a bug in the documentation for get/copy named substring when
493        duplicate names are permitted. If none of the named substrings are set, the
494        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
495        empty string.
496    
497    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
498        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
499        because the ] is interpreted as the first data character and the
500        terminating ] is not found. PCRE has been made compatible with Perl in this
501        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
502        cause memory overwriting.
503    
504    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
505        string has been matched (to stop an infinite loop). It was not recognizing
506        a conditional subpattern that could match an empty string if that
507        subpattern was within another subpattern. For example, it looped when
508        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
509        condition was not nested. This bug has been fixed.
510    
511    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
512        past the start of the subject in the presence of bytes with the top bit
513        set, for example "\x8aBCD".
514    
515    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
516        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
517    
518    14. Optimized (?!) to (*FAIL).
519    
520    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
521        This restricts code points to be within the range 0 to 0x10FFFF, excluding
522        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
523        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
524        does: it's just the validity check that is more restrictive.
525    
526    16. Inserted checks for integer overflows during escape sequence (backslash)
527        processing, and also fixed erroneous offset values for syntax errors during
528        backslash processing.
529    
530    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
531        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
532    
533    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
534        caused an overrun.
535    
536    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
537        something other than just ASCII characters) inside a group that had an
538        unlimited repeat caused a loop at compile time (while checking to see
539        whether the group could match an empty string).
540    
541    20. Debugging a pattern containing \p or \P could cause a crash. For example,
542        [\P{Any}] did so. (Error in the code for printing property names.)
543    
544    21. An orphan \E inside a character class could cause a crash.
545    
546    22. A repeated capturing bracket such as (A)? could cause a wild memory
547        reference during compilation.
548    
549    23. There are several functions in pcre_compile() that scan along a compiled
550        expression for various reasons (e.g. to see if it's fixed length for look
551        behind). There were bugs in these functions when a repeated \p or \P was
552        present in the pattern. These operators have additional parameters compared
553        with \d, etc, and these were not being taken into account when moving along
554        the compiled data. Specifically:
555    
556        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
557            length.
558    
559        (b) An item such as \pL+ within a repeated group could cause crashes or
560            loops.
561    
562        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
563            "reference to non-existent subpattern" error.
564    
565        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
566    
567    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
568        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
569    
570    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
571    
572    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
573        character were causing crashes (broken optimization).
574    
575    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
576        \p or \P) caused a compile-time loop.
577    
578    28. More problems have arisen in unanchored patterns when CRLF is a valid line
579        break. For example, the unstudied pattern [\r\n]A does not match the string
580        "\r\nA" because change 7.0/46 below moves the current point on by two
581        characters after failing to match at the start. However, the pattern \nA
582        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
583        the same is true. There doesn't seem any very clean way out of this, but
584        what I have chosen to do makes the common cases work: PCRE now takes note
585        of whether there can be an explicit match for \r or \n anywhere in the
586        pattern, and if so, 7.0/46 no longer applies. As part of this change,
587        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
588        pattern has explicit CR or LF references.
589    
590    29. Added (*CR) etc for changing newline setting at start of pattern.
591    
592    
593    Version 7.2 19-Jun-07
594    ---------------------
595    
596     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
597        which is apparently normally available under Windows.
598    
599     2. Re-jig the pcregrep tests with different newline settings in an attempt
600        to make them independent of the local environment's newline setting.
601    
602     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
603    
604     4. Some of the "internals" tests were previously cut out when the link size
605        was not 2, because the output contained actual offsets. The recent new
606        "Z" feature of pcretest means that these can be cut out, making the tests
607        usable with all link sizes.
608    
609     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
610        stack recursion. This gives a massive performance boost under BSD, but just
611        a small improvement under Linux. However, it saves one field in the frame
612        in all cases.
613    
614     6. Added more features from the forthcoming Perl 5.10:
615    
616        (a) (?-n) (where n is a string of digits) is a relative subroutine or
617            recursion call. It refers to the nth most recently opened parentheses.
618    
619        (b) (?+n) is also a relative subroutine call; it refers to the nth next
620            to be opened parentheses.
621    
622        (c) Conditions that refer to capturing parentheses can be specified
623            relatively, for example, (?(-2)... or (?(+3)...
624    
625        (d) \K resets the start of the current match so that everything before
626            is not part of it.
627    
628        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
629    
630        (f) \g{name} is another synonym - part of Perl 5.10's unification of
631            reference syntax.
632    
633        (g) (?| introduces a group in which the numbering of parentheses in each
634            alternative starts with the same number.
635    
636        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
637    
638     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
639        PCRE_INFO_JCHANGED.
640    
641     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
642        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
643        for detecting groups that can match an empty string.
644    
645     9. A pattern with a very large number of alternatives (more than several
646        hundred) was running out of internal workspace during the pre-compile
647        phase, where pcre_compile() figures out how much memory will be needed. A
648        bit of new cunning has reduced the workspace needed for groups with
649        alternatives. The 1000-alternative test pattern now uses 12 bytes of
650        workspace instead of running out of the 4096 that are available.
651    
652    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
653    
654    11. Applied patch from Google to remove an optimization that didn't quite work.
655        The report of the bug said:
656    
657          pcrecpp::RE("a*").FullMatch("aaa") matches, while
658          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
659          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
660    
661    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
662        it matched the wrong number of bytes.
663    
664    
665    Version 7.1 24-Apr-07
666    ---------------------
667    
668     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
669        that is more "standard", making use of automake and other Autotools. There
670        is some re-arrangement of the files and adjustment of comments consequent
671        on this.
672    
673     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
674        for recursive directory scanning broke on some systems because the files
675        are not scanned in any specific order and on different systems the order
676        was different. A call to "sort" has been inserted into RunGrepTest for the
677        approprate test as a short-term fix. In the longer term there may be an
678        alternative.
679    
680     3. I had an email from Eric Raymond about problems translating some of PCRE's
681        man pages to HTML (despite the fact that I distribute HTML pages, some
682        people do their own conversions for various reasons). The problems
683        concerned the use of low-level troff macros .br and .in. I have therefore
684        removed all such uses from the man pages (some were redundant, some could
685        be replaced by .nf/.fi pairs). The 132html script that I use to generate
686        HTML has been updated to handle .nf/.fi and to complain if it encounters
687        .br or .in.
688    
689     4. Updated comments in configure.ac that get placed in config.h.in and also
690        arranged for config.h to be included in the distribution, with the name
691        config.h.generic, for the benefit of those who have to compile without
692        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
693    
694     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
695        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
696        makevp.bat for the current PCRE, using the additional files
697        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
698    
699     6. A Windows user reported a minor discrepancy with test 2, which turned out
700        to be caused by a trailing space on an input line that had got lost in his
701        copy. The trailing space was an accident, so I've just removed it.
702    
703     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
704        that is needed.
705    
706     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
707        as "const" (a) because they are and (b) because it helps the PHP
708        maintainers who have recently made a script to detect big data structures
709        in the php code that should be moved to the .rodata section. I remembered
710        to update Builducptable as well, so it won't revert if ucptable.h is ever
711        re-created.
712    
713     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
714        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
715        order to be able to cut out the UTF-8 tables in the latter when UTF-8
716        support is not required. This saves 1.5-2K of code, which is important in
717        some applications.
718    
719        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
720        so as not to refer to the tables, even though these functions will never be
721        called when UTF-8 support is disabled. Otherwise there are problems with a
722        shared library.
723    
724    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
725    
726        (a) It was defining its arguments as char * instead of void *.
727    
728        (b) It was assuming that all moves were upwards in memory; this was true
729            a long time ago when I wrote it, but is no longer the case.
730    
731        The emulated memove() is provided for those environments that have neither
732        memmove() nor bcopy(). I didn't think anyone used it these days, but that
733        is clearly not the case, as these two bugs were recently reported.
734    
735    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
736        and Detrail to create the HTML documentation, the .txt form of the man
737        pages, and it removes trailing spaces from listed files. It also creates
738        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
739        case, it wraps all the #defines with #ifndefs. This script should be run
740        before "make dist".
741    
742    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
743        with Unicode property support.
744    
745        (a) For a maximizing quantifier, if the two different cases of the
746            character were of different lengths in their UTF-8 codings (there are
747            some cases like this - I found 11), and the matching function had to
748            back up over a mixture of the two cases, it incorrectly assumed they
749            were both the same length.
750    
751        (b) When PCRE was configured to use the heap rather than the stack for
752            recursion during matching, it was not correctly preserving the data for
753            the other case of a UTF-8 character when checking ahead for a match
754            while processing a minimizing repeat. If the check also involved
755            matching a wide character, but failed, corruption could cause an
756            erroneous result when trying to check for a repeat of the original
757            character.
758    
759    13. Some tidying changes to the testing mechanism:
760    
761        (a) The RunTest script now detects the internal link size and whether there
762            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
763            values substituted by "configure". (The RunGrepTest script already did
764            this for UTF-8.) The configure.ac script no longer substitutes the
765            relevant variables.
766    
767        (b) The debugging options /B and /D in pcretest show the compiled bytecode
768            with length and offset values. This means that the output is different
769            for different internal link sizes. Test 2 is skipped for link sizes
770            other than 2 because of this, bypassing the problem. Unfortunately,
771            there was also a test in test 3 (the locale tests) that used /B and
772            failed for link sizes other than 2. Rather than cut the whole test out,
773            I have added a new /Z option to pcretest that replaces the length and
774            offset values with spaces. This is now used to make test 3 independent
775            of link size. (Test 2 will be tidied up later.)
776    
777    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
778        segmentation fault instead of returning the appropriate error message.
779    
780    15. In multiline mode when the newline sequence was set to "any", the pattern
781        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
782        This doesn't seem right; it now treats the CRLF combination as the line
783        ending, and so does not match in that case. It's only a pattern such as ^$
784        that would hit this one: something like ^ABC$ would have failed after \r
785        and then tried again after \r\n.
786    
787    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
788        in an attempt to make files that differ only in their line terminators
789        compare equal. This works on Linux.
790    
791    17. Under certain error circumstances pcregrep might try to free random memory
792        as it exited. This is now fixed, thanks to valgrind.
793    
794    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
795        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
796        was because its rules for how to advance for /g after matching an empty
797        string at the end of a line did not allow for this case. They now check for
798        it specially.
799    
800    20. pcretest is supposed to handle patterns and data of any length, by
801        extending its buffers when necessary. It was getting this wrong when the
802        buffer for a data line had to be extended.
803    
804    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
805        CRLF as a newline sequence.
806    
807    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
808        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
809        I have nevertheless tidied it up.
810    
811    23. Added some casts to kill warnings from HP-UX ia64 compiler.
812    
813    24. Added a man page for pcre-config.
814    
815    
816    Version 7.0 19-Dec-06
817    ---------------------
818    
819     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
820        moving to gcc 4.1.1.
821    
822     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
823        sys/time.h, which is documented as needed for this function. It doesn't
824        seem to matter on Linux, but it showed up on some releases of OS X.
825    
826     3. It seems that there are systems where bytes whose values are greater than
827        127 match isprint() in the "C" locale. The "C" locale should be the
828        default when a C program starts up. In most systems, only ASCII printing
829        characters match isprint(). This difference caused the output from pcretest
830        to vary, making some of the tests fail. I have changed pcretest so that:
831    
832        (a) When it is outputting text in the compiled version of a pattern, bytes
833            other than 32-126 are always shown as hex escapes.
834    
835        (b) When it is outputting text that is a matched part of a subject string,
836            it does the same, unless a different locale has been set for the match
837            (using the /L modifier). In this case, it uses isprint() to decide.
838    
839     4. Fixed a major bug that caused incorrect computation of the amount of memory
840        required for a compiled pattern when options that changed within the
841        pattern affected the logic of the preliminary scan that determines the
842        length. The relevant options are -x, and -i in UTF-8 mode. The result was
843        that the computed length was too small. The symptoms of this bug were
844        either the PCRE error "internal error: code overflow" from pcre_compile(),
845        or a glibc crash with a message such as "pcretest: free(): invalid next
846        size (fast)". Examples of patterns that provoked this bug (shown in
847        pcretest format) are:
848    
849          /(?-x: )/x
850          /(?x)(?-x: \s*#\s*)/
851          /((?i)[\x{c0}])/8
852          /(?i:[\x{c0}])/8
853    
854        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
855        is now done differently.
856    
857     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
858        wrapper classes; (b) implement a new function in the C++ scanner that is
859        more efficient than the old way of doing things because it avoids levels of
860        recursion in the regex matching; (c) add a paragraph to the documentation
861        for the FullMatch() function.
862    
863     6. The escape sequence \n was being treated as whatever was defined as
864        "newline". Not only was this contrary to the documentation, which states
865        that \n is character 10 (hex 0A), but it also went horribly wrong when
866        "newline" was defined as CRLF. This has been fixed.
867    
868     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
869        was being set to -1 for the "end of line" case (supposedly a value that no
870        character can have). Though this value is never used (the check for end of
871        line is "zero bytes in current character"), it caused compiler complaints.
872        I've changed it to 0xffffffff.
873    
874     8. In pcre_version.c, the version string was being built by a sequence of
875        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
876        string (as it is for production releases) called a macro with an empty
877        argument. The C standard says the result of this is undefined. The gcc
878        compiler treats it as an empty string (which was what was wanted) but it is
879        reported that Visual C gives an error. The source has been hacked around to
880        avoid this problem.
881    
882     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
883        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
884        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
885        of them did).
886    
887    10. Originally, pcretest opened its input and output without "b"; then I was
888        told that "b" was needed in some environments, so it was added for release
889        5.0 to both the input and output. (It makes no difference on Unix-like
890        systems.) Later I was told that it is wrong for the input on Windows. I've
891        now abstracted the modes into two macros, to make it easier to fiddle with
892        them, and removed "b" from the input mode under Windows.
893    
894    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
895    
896    12. Added -help and --help to pcretest as an official way of being reminded
897        of the options.
898    
899    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
900        and pcrecpp.cc because they annoy compilers at high warning levels.
901    
902    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
903    
904    15. Fixed an occurrence of == in configure.ac that should have been = (shell
905        scripts are not C programs :-) and which was not noticed because it works
906        on Linux.
907    
908    16. pcretest is supposed to handle any length of pattern and data line (as one
909        line or as a continued sequence of lines) by extending its input buffer if
910        necessary. This feature was broken for very long pattern lines, leading to
911        a string of junk being passed to pcre_compile() if the pattern was longer
912        than about 50K.
913    
914    17. I have done a major re-factoring of the way pcre_compile() computes the
915        amount of memory needed for a compiled pattern. Previously, there was code
916        that made a preliminary scan of the pattern in order to do this. That was
917        OK when PCRE was new, but as the facilities have expanded, it has become
918        harder and harder to keep it in step with the real compile phase, and there
919        have been a number of bugs (see for example, 4 above). I have now found a
920        cunning way of running the real compile function in a "fake" mode that
921        enables it to compute how much memory it would need, while actually only
922        ever using a few hundred bytes of working memory and without too many
923        tests of the mode. This should make future maintenance and development
924        easier. A side effect of this work is that the limit of 200 on the nesting
925        depth of parentheses has been removed (though this was never a serious
926        limitation, I suspect). However, there is a downside: pcre_compile() now
927        runs more slowly than before (30% or more, depending on the pattern). I
928        hope this isn't a big issue. There is no effect on runtime performance.
929    
930    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
931        newline (only possible for the last line of a file) and it was a
932        pattern that set a locale (followed by /Lsomething), pcretest crashed.
933    
934    19. Added additional timing features to pcretest. (1) The -tm option now times
935        matching only, not compiling. (2) Both -t and -tm can be followed, as a
936        separate command line item, by a number that specifies the number of
937        repeats to use when timing. The default is 50000; this gives better
938        precision, but takes uncomfortably long for very large patterns.
939    
940    20. Extended pcre_study() to be more clever in cases where a branch of a
941        subpattern has no definite first character. For example, (a*|b*)[cd] would
942        previously give no result from pcre_study(). Now it recognizes that the
943        first character must be a, b, c, or d.
944    
945    21. There was an incorrect error "recursive call could loop indefinitely" if
946        a subpattern (or the entire pattern) that was being tested for matching an
947        empty string contained only one non-empty item after a nested subpattern.
948        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
949        incorrectly, because the \d was being skipped in the check.
950    
951    22. The pcretest program now has a new pattern option /B and a command line
952        option -b, which is equivalent to adding /B to every pattern. This causes
953        it to show the compiled bytecode, without the additional information that
954        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
955        is the same as /B/I).
956    
957    23. A new optimization is now able automatically to treat some sequences such
958        as a*b as a*+b. More specifically, if something simple (such as a character
959        or a simple class like \d) has an unlimited quantifier, and is followed by
960        something that cannot possibly match the quantified thing, the quantifier
961        is automatically "possessified".
962    
963    24. A recursive reference to a subpattern whose number was greater than 39
964        went wrong under certain circumstances in UTF-8 mode. This bug could also
965        have affected the operation of pcre_study().
966    
967    25. Realized that a little bit of performance could be had by replacing
968        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
969    
970    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
971    
972    27. Possessive quantifiers such as a++ were previously implemented by turning
973        them into atomic groups such as ($>a+). Now they have their own opcodes,
974        which improves performance. This includes the automatically created ones
975        from 23 above.
976    
977    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
978        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
979        the first matched character to be a colon. This applied both to named and
980        numbered groups.
981    
982    29. The ucpinternal.h header file was missing its idempotency #ifdef.
983    
984    30. I was sent a "project" file called libpcre.a.dev which I understand makes
985        building PCRE on Windows easier, so I have included it in the distribution.
986    
987    31. There is now a check in pcretest against a ridiculously large number being
988        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
989        loop, the loop is abandoned.
990    
991    32. Forward references to subpatterns in conditions such as (?(2)...) where
992        subpattern 2 is defined later cause pcre_compile() to search forwards in
993        the pattern for the relevant set of parentheses. This search went wrong
994        when there were unescaped parentheses in a character class, parentheses
995        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
996    
997    33. "Subroutine" calls and backreferences were previously restricted to
998        referencing subpatterns earlier in the regex. This restriction has now
999        been removed.
1000    
1001    34. Added a number of extra features that are going to be in Perl 5.10. On the
1002        whole, these are just syntactic alternatives for features that PCRE had
1003        previously implemented using the Python syntax or my own invention. The
1004        other formats are all retained for compatibility.
1005    
1006        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1007            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1008            also .NET compatible.
1009    
1010        (b) A recursion or subroutine call to a named group can now be defined as
1011            (?&name) as well as (?P>name).
1012    
1013        (c) A backreference to a named group can now be defined as \k<name> or
1014            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1015            5.10, are also .NET compatible.
1016    
1017        (d) A conditional reference to a named group can now use the syntax
1018            (?(<name>) or (?('name') as well as (?(name).
1019    
1020        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1021            groups (named and numbered) that are never evaluated inline, but can be
1022            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1023            is always false. There may be only one alternative in such a group.
1024    
1025        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1026            as the simple (?(R). The condition is true only if the most recent
1027            recursion is that of the given number or name. It does not search out
1028            through the entire recursion stack.
1029    
1030        (g) The escape \gN or \g{N} has been added, where N is a positive or
1031            negative number, specifying an absolute or relative reference.
1032    
1033    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1034        some "unreachable code" warnings.
1035    
1036    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1037        things, this adds five new scripts.
1038    
1039    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1040        There were also incompatibilities regarding the handling of \Q..\E inside
1041        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1042        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1043    
1044    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1045        matches an empty string, and forcibly breaks the loop. There were bugs in
1046        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1047        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1048        separate and independent bugs (that affected different cases) have been
1049        fixed.
1050    
1051    39. Refactored the code to abolish the use of different opcodes for small
1052        capturing bracket numbers. This is a tidy that I avoided doing when I
1053        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1054        The new approach is not only tidier, it makes it possible to reduce the
1055        memory needed to fix the previous bug (38).
1056    
1057    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1058        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1059        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1060        mode.
1061    
1062    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1063        report.
1064    
1065    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1066        copy construction and assignment in the C++ wrapper.
1067    
1068    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1069        couple of bugs that could have given wrong results in the "--newline=crlf"
1070        case.
1071    
1072    44. Added a number of casts and did some reorganization of signed/unsigned int
1073        variables following suggestions from Dair Grant. Also renamed the variable
1074        "this" as "item" because it is a C++ keyword.
1075    
1076    45. Arranged for dftables to add
1077    
1078          #include "pcre_internal.h"
1079    
1080        to pcre_chartables.c because without it, gcc 4.x may remove the array
1081        definition from the final binary if PCRE is built into a static library and
1082        dead code stripping is activated.
1083    
1084    46. For an unanchored pattern, if a match attempt fails at the start of a
1085        newline sequence, and the newline setting is CRLF or ANY, and the next two
1086        characters are CRLF, advance by two characters instead of one.
1087    
1088    
1089    Version 6.7 04-Jul-06
1090    ---------------------
1091    
1092     1. In order to handle tests when input lines are enormously long, pcretest has
1093        been re-factored so that it automatically extends its buffers when
1094        necessary. The code is crude, but this _is_ just a test program. The
1095        default size has been increased from 32K to 50K.
1096    
1097     2. The code in pcre_study() was using the value of the re argument before
1098        testing it for NULL. (Of course, in any sensible call of the function, it
1099        won't be NULL.)
1100    
1101     3. The memmove() emulation function in pcre_internal.h, which is used on
1102        systems that lack both memmove() and bcopy() - that is, hardly ever -
1103        was missing a "static" storage class specifier.
1104    
1105     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1106        containing an extended class (one that cannot be represented by a bitmap
1107        because it contains high-valued characters or Unicode property items, e.g.
1108        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1109        pattern, but PCRE should not loop if you do not (it no longer does).
1110        [Detail: two cases were found: (a) a repeated subpattern containing an
1111        extended class; (b) a recursive reference to a subpattern that followed a
1112        previous extended class. It wasn't skipping over the extended class
1113        correctly when UTF-8 mode was not set.]
1114    
1115     5. A negated single-character class was not being recognized as fixed-length
1116        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1117        compile error "lookbehind assertion is not fixed length".
1118    
1119     6. The RunPerlTest auxiliary script was showing an unexpected difference
1120        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1121        write a Perl script that can interpret lines of an input file either as
1122        byte characters or as UTF-8, which is what "perltest" was being required to
1123        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1124        can't do is switch easily at run time between having the "use utf8;" pragma
1125        or not. In the end, I fudged it by using the RunPerlTest script to insert
1126        "use utf8;" explicitly for the UTF-8 tests.
1127    
1128     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1129        the end of the subject string, contrary to the documentation and to what
1130        Perl does. This was true of both matching functions. Now it matches only at
1131        the start of the subject and immediately after *internal* newlines.
1132    
1133     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1134        a pointer to an int instead of a pointer to an unsigned long int. This
1135        caused problems on 64-bit systems.
1136    
1137     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1138        instance of the 'standard' template library not being so standard".
1139    
1140    10. There was no check on the number of named subpatterns nor the maximum
1141        length of a subpattern name. The product of these values is used to compute
1142        the size of the memory block for a compiled pattern. By supplying a very
1143        long subpattern name and a large number of named subpatterns, the size
1144        computation could be caused to overflow. This is now prevented by limiting
1145        the length of names to 32 characters, and the number of named subpatterns
1146        to 10,000.
1147    
1148    11. Subpatterns that are repeated with specific counts have to be replicated in
1149        the compiled pattern. The size of memory for this was computed from the
1150        length of the subpattern and the repeat count. The latter is limited to
1151        65535, but there was no limit on the former, meaning that integer overflow
1152        could in principle occur. The compiled length of a repeated subpattern is
1153        now limited to 30,000 bytes in order to prevent this.
1154    
1155    12. Added the optional facility to have named substrings with the same name.
1156    
1157    13. Added the ability to use a named substring as a condition, using the
1158        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1159        are numbers (not recommended). Forward references are permitted.
1160    
1161    14. Added forward references in named backreferences (if you see what I mean).
1162    
1163    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1164        pattern could run off the end of the subject. For example, the pattern
1165        "(?s)(.{1,5})"8 did this with the subject "ab".
1166    
1167    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1168        PCRE_CASELESS was set when matching characters that were quantified with ?
1169        or *.
1170    
1171    17. A character class other than a single negated character that had a minimum
1172        but no maximum quantifier - for example [ab]{6,} - was not handled
1173        correctly by pce_dfa_exec(). It would match only one character.
1174    
1175    18. A valid (though odd) pattern that looked like a POSIX character
1176        class but used an invalid character after [ (for example [[,abc,]]) caused
1177        pcre_compile() to give the error "Failed: internal error: code overflow" or
1178        in some cases to crash with a glibc free() error. This could even happen if
1179        the pattern terminated after [[ but there just happened to be a sequence of
1180        letters, a binary zero, and a closing ] in the memory that followed.
1181    
1182    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1183        over the years. Originally (before any Unicode support), just the bottom 8
1184        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1185        output from "man perlunicode" includes this:
1186    
1187          The regular expression compiler produces polymorphic opcodes.  That
1188          is, the pattern adapts to the data and automatically switches to
1189          the Unicode character scheme when presented with Unicode data--or
1190          instead uses a traditional byte scheme when presented with byte
1191          data.
1192    
1193        Sadly, a wide octal escape does not cause a switch, and in a string with
1194        no other multibyte characters, these octal escapes are treated as before.
1195        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1196        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1197        Unicode string.
1198    
1199        I have not perpetrated such confusion in PCRE. Up till now, it took just
1200        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1201        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1202        translate to the appropriate multibyte character.
1203    
1204    29. Applied some refactoring to reduce the number of warnings from Microsoft
1205        and Borland compilers. This has included removing the fudge introduced
1206        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1207        a warning about an unused variable.
1208    
1209    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1210        characters since release 4.0, because Perl (from release 5.004) does not.
1211        [Or at least, is documented not to: some releases seem to be in conflict
1212        with the documentation.] However, when a pattern was studied with
1213        pcre_study() and all its branches started with \s, PCRE still included VT
1214        as a possible starting character. Of course, this did no harm; it just
1215        caused an unnecessary match attempt.
1216    
1217    22. Removed a now-redundant internal flag bit that recorded the fact that case
1218        dependency changed within the pattern. This was once needed for "required
1219        byte" processing, but is no longer used. This recovers a now-scarce options
1220        bit. Also moved the least significant internal flag bit to the most-
1221        significant bit of the word, which was not previously used (hangover from
1222        the days when it was an int rather than a uint) to free up another bit for
1223        the future.
1224    
1225    23. Added support for CRLF line endings as well as CR and LF. As well as the
1226        default being selectable at build time, it can now be changed at runtime
1227        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1228        specify that it is scanning data with non-default line endings.
1229    
1230    24. Changed the definition of CXXLINK to make it agree with the definition of
1231        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1232    
1233    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1234        recursions. This makes a big different to stack usage for some patterns.
1235    
1236    26. If a subpattern containing a named recursion or subroutine reference such
1237        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1238        the space required for the compiled pattern went wrong and gave too small a
1239        value. Depending on the environment, this could lead to "Failed: internal
1240        error: code overflow at offset 49" or "glibc detected double free or
1241        corruption" errors.
1242    
1243    27. Applied patches from Google (a) to support the new newline modes and (b) to
1244        advance over multibyte UTF-8 characters in GlobalReplace.
1245    
1246    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1247        difference for some implementation of PCRE in some Windows version.
1248    
1249    29. Added some extra testing facilities to pcretest:
1250    
1251        \q<number>   in a data line sets the "match limit" value
1252        \Q<number>   in a data line sets the "match recursion limt" value
1253        -S <number>  sets the stack size, where <number> is in megabytes
1254    
1255        The -S option isn't available for Windows.
1256    
1257    
1258    Version 6.6 06-Feb-06
1259    ---------------------
1260    
1261     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1262        in pcreposix.h. I have copied the definition from pcre.h.
1263    
1264     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1265        because pcre.h is no longer a built file.
1266    
1267     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1268        not normally included in the compiled code.
1269    
1270    
1271    Version 6.5 01-Feb-06
1272    ---------------------
1273    
1274     1. When using the partial match feature with pcre_dfa_exec(), it was not
1275        anchoring the second and subsequent partial matches at the new starting
1276        point. This could lead to incorrect results. For example, with the pattern
1277        /1234/, partially matching against "123" and then "a4" gave a match.
1278    
1279     2. Changes to pcregrep:
1280    
1281        (a) All non-match returns from pcre_exec() were being treated as failures
1282            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1283            error message is output. Some extra information is given for the
1284            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1285            probably the only errors that are likely to be caused by users (by
1286            specifying a regex that has nested indefinite repeats, for instance).
1287            If there are more than 20 of these errors, pcregrep is abandoned.
1288    
1289        (b) A binary zero was treated as data while matching, but terminated the
1290            output line if it was written out. This has been fixed: binary zeroes
1291            are now no different to any other data bytes.
1292    
1293        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1294            used to set a locale for matching. The --locale=xxxx long option has
1295            been added (no short equivalent) to specify a locale explicitly on the
1296            pcregrep command, overriding the environment variables.
1297    
1298        (d) When -B was used with -n, some line numbers in the output were one less
1299            than they should have been.
1300    
1301        (e) Added the -o (--only-matching) option.
1302    
1303        (f) If -A or -C was used with -c (count only), some lines of context were
1304            accidentally printed for the final match.
1305    
1306        (g) Added the -H (--with-filename) option.
1307    
1308        (h) The combination of options -rh failed to suppress file names for files
1309            that were found from directory arguments.
1310    
1311        (i) Added the -D (--devices) and -d (--directories) options.
1312    
1313        (j) Added the -F (--fixed-strings) option.
1314    
1315        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1316    
1317        (l) Added the --colo(u)r option.
1318    
1319        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1320            is not present by default.
1321    
1322     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1323        items such as (?R) or (?1), when the recursion could match a number of
1324        alternatives. If it matched one of the alternatives, but subsequently,
1325        outside the recursion, there was a failure, the code tried to back up into
1326        the recursion. However, because of the way PCRE is implemented, this is not
1327        possible, and the result was an incorrect result from the match.
1328    
1329        In order to prevent this happening, the specification of recursion has
1330        been changed so that all such subpatterns are automatically treated as
1331        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1332    
1333     4. I had overlooked the fact that, in some locales, there are characters for
1334        which isalpha() is true but neither isupper() nor islower() are true. In
1335        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1336        and ordfeminine) are like this. This affected the treatment of \w and \W
1337        when they appeared in character classes, but not when they appeared outside
1338        a character class. The bit map for "word" characters is now created
1339        separately from the results of isalnum() instead of just taking it from the
1340        upper, lower, and digit maps. (Plus the underscore character, of course.)
1341    
1342     5. The above bug also affected the handling of POSIX character classes such as
1343        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1344        permanent tables. Instead, the bit maps for such a class were previously
1345        created as the appropriate unions of the upper, lower, and digit bitmaps.
1346        Now they are created by subtraction from the [[:word:]] class, which has
1347        its own bitmap.
1348    
1349     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1350        It is created by subtracting the vertical space characters (\x09, \x0a,
1351        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1352        subtraction was done in the overall bitmap for a character class, meaning
1353        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1354        be recognized. This bug has been fixed.
1355    
1356     7. Patches from the folks at Google:
1357    
1358          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1359          real life, but is still worth protecting against".
1360    
1361          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1362          regular expressions".
1363    
1364          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1365          have it.
1366    
1367          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1368          "configure" and the latter not, in order to fix a problem somebody had
1369          with compiling the Arg class on HP-UX.
1370    
1371          (e) Improve the error-handling of the C++ wrapper a little bit.
1372    
1373          (f) New tests for checking recursion limiting.
1374    
1375     8. The pcre_memmove() function, which is used only if the environment does not
1376        have a standard memmove() function (and is therefore rarely compiled),
1377        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1378        returning a result (though PCRE never actually uses the result).
1379    
1380     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1381        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1382        returned instead of calling malloc() with an overflowing number that would
1383        most likely cause subsequent chaos.
1384    
1385    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1386    
1387    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1388        with this option is matched, the nmatch and pmatch options of regexec() are
1389        ignored.
1390    
1391    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1392        provided in case anyone wants to the the POSIX interface with UTF-8
1393        strings.
1394    
1395    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1396        C++ linking (needed for some HP-UX environments).
1397    
1398    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1399        (unused parameter) and in the pcre_printint() function (omitted "default"
1400        switch label when the default is to do nothing).
1401    
1402    15. Added some code to make it possible, when PCRE is compiled as a C++
1403        library, to replace subject pointers for pcre_exec() with a smart pointer
1404        class, thus making it possible to process discontinuous strings.
1405    
1406    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1407        much the same function. They were added by different people who were trying
1408        to make PCRE easy to compile on non-Unix systems. It has been suggested
1409        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1410        for compiling on Windows systems. I have therefore replaced it with
1411        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1412        defaults to "extern" for C or "extern C" for C++, which works fine on
1413        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1414        SCOPE with something explicit in config.h. In addition:
1415    
1416        (a) pcreposix.h still had just "extern" instead of either of these macros;
1417            I have replaced it with PCRE_DATA_SCOPE.
1418    
1419        (b) Functions such as _pcre_xclass(), which are internal to the library,
1420            but external in the C sense, all had PCRE_EXPORT in their definitions.
1421            This is apparently wrong for the Windows case, so I have removed it.
1422            (It makes no difference on Unix-like systems.)
1423    
1424    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1425        of recursive calls to match(). This is different to MATCH_LIMIT because
1426        that limits the total number of calls to match(), not all of which increase
1427        the depth of recursion. Limiting the recursion depth limits the amount of
1428        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1429        when PCRE is compiled, and changed at run time. A patch from Google adds
1430        this functionality to the C++ interface.
1431    
1432    18. Changes to the handling of Unicode character properties:
1433    
1434        (a) Updated the table to Unicode 4.1.0.
1435    
1436        (b) Recognize characters that are not in the table as "Cn" (undefined).
1437    
1438        (c) I revised the way the table is implemented to a much improved format
1439            which includes recognition of ranges. It now supports the ranges that
1440            are defined in UnicodeData.txt, and it also amalgamates other
1441            characters into ranges. This has reduced the number of entries in the
1442            table from around 16,000 to around 3,000, thus reducing its size
1443            considerably. I realized I did not need to use a tree structure after
1444            all - a binary chop search is just as efficient. Having reduced the
1445            number of entries, I extended their size from 6 bytes to 8 bytes to
1446            allow for more data.
1447    
1448        (d) Added support for Unicode script names via properties such as \p{Han}.
1449    
1450    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1451        matching that character.
1452    
1453    20. When matching a repeated Unicode property with a minimum greater than zero,
1454        (for example \pL{2,}), PCRE could look past the end of the subject if it
1455        reached it while seeking the minimum number of characters. This could
1456        happen only if some of the characters were more than one byte long, because
1457        there is a check for at least the minimum number of bytes.
1458    
1459    21. Refactored the implementation of \p and \P so as to be more general, to
1460        allow for more different types of property in future. This has changed the
1461        compiled form incompatibly. Anybody with saved compiled patterns that use
1462        \p or \P will have to recompile them.
1463    
1464    22. Added "Any" and "L&" to the supported property types.
1465    
1466    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1467        but give a compile time error if the value is greater than 0xff.
1468    
1469    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1470        accidentally not being installed or uninstalled.
1471    
1472    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1473        made were to insert the current release number. This seemed silly, because
1474        it made things harder for people building PCRE on systems that don't run
1475        "configure". I have turned pcre.h into a distributed file, no longer built
1476        by "configure", with the version identification directly included. There is
1477        no longer a pcre.h.in file.
1478    
1479        However, this change necessitated a change to the pcre-config script as
1480        well. It is built from pcre-config.in, and one of the substitutions was the
1481        release number. I have updated configure.ac so that ./configure now finds
1482        the release number by grepping pcre.h.
1483    
1484    26. Added the ability to run the tests under valgrind.
1485    
1486    
1487    Version 6.4 05-Sep-05
1488    ---------------------
1489    
1490     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1491        "--" to be printed when multiple files were scanned, even when none of the
1492        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1493        consider it to be a bug, and have restored the previous behaviour.
1494    
1495     2. A couple of code tidies to get rid of compiler warnings.
1496    
1497     3. The pcretest program used to cheat by referring to symbols in the library
1498        whose names begin with _pcre_. These are internal symbols that are not
1499        really supposed to be visible externally, and in some environments it is
1500        possible to suppress them. The cheating is now confined to including
1501        certain files from the library's source, which is a bit cleaner.
1502    
1503     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1504        file's purpose clearer.
1505    
1506     5. Reorganized pcre_ucp_findchar().
1507    
1508    
1509    Version 6.3 15-Aug-05
1510    ---------------------
1511    
1512     1. The file libpcre.pc.in did not have general read permission in the tarball.
1513    
1514     2. There were some problems when building without C++ support:
1515    
1516        (a) If C++ support was not built, "make install" and "make test" still
1517            tried to test it.
1518    
1519        (b) There were problems when the value of CXX was explicitly set. Some
1520            changes have been made to try to fix these, and ...
1521    
1522        (c) --disable-cpp can now be used to explicitly disable C++ support.
1523    
1524        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1525            backslash in a target when C++ was disabled. This confuses some
1526            versions of "make", apparently. Using an intermediate variable solves
1527            this. (Same for CPP_LOBJ.)
1528    
1529     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1530        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1531        necessary on certain architectures.
1532    
1533     4. Added a setting of -export-symbols-regex to the link command to remove
1534        those symbols that are exported in the C sense, but actually are local
1535        within the library, and not documented. Their names all begin with
1536        "_pcre_". This is not a perfect job, because (a) we have to except some
1537        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1538        available (and never for static libraries). I have made a note to try to
1539        find a way round (a) in the future.
1540    
1541    
1542  Version 6.2 01-Aug-05  Version 6.2 01-Aug-05
1543  ---------------------  ---------------------
1544    

Legend:
Removed from v.81  
changed lines
  Added in v.361

  ViewVC Help
Powered by ViewVC 1.1.5