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

Diff of /code/trunk/ChangeLog

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

revision 324 by ph10, Fri Mar 7 19:48:32 2008 UTC revision 1590 by ph10, Mon Aug 10 14:38:34 2015 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.7 05-Mar-08  Note that the PCRE 8.xx series (PCRE1) is now in a bugfix-only state. All
5    development is happening in the PCRE2 10.xx series.
6    
7    Version 8.38 xx-xxx-xxxx
8    ------------------------
9    
10    1.  If a group that contained a recursive back reference also contained a
11        forward reference subroutine call followed by a non-forward-reference
12        subroutine call, for example /.((?2)(?R)\1)()/, pcre2_compile() failed to
13        compile correct code, leading to undefined behaviour or an internally
14        detected error. This bug was discovered by the LLVM fuzzer.
15    
16    2.  Quantification of certain items (e.g. atomic back references) could cause
17        incorrect code to be compiled when recursive forward references were
18        involved. For example, in this pattern: /(?1)()((((((\1++))\x85)+)|))/.
19        This bug was discovered by the LLVM fuzzer.
20    
21    3.  A repeated conditional group whose condition was a reference by name caused
22        a buffer overflow if there was more than one group with the given name.
23        This bug was discovered by the LLVM fuzzer.
24    
25    4.  A recursive back reference by name within a group that had the same name as
26        another group caused a buffer overflow. For example:
27        /(?J)(?'d'(?'d'\g{d}))/. This bug was discovered by the LLVM fuzzer.
28    
29    5.  A forward reference by name to a group whose number is the same as the
30        current group, for example in this pattern: /(?|(\k'Pm')|(?'Pm'))/, caused
31        a buffer overflow at compile time. This bug was discovered by the LLVM
32        fuzzer.
33    
34    6.  A lookbehind assertion within a set of mutually recursive subpatterns could
35        provoke a buffer overflow. This bug was discovered by the LLVM fuzzer.
36    
37    7.  Another buffer overflow bug involved duplicate named groups with a
38        reference between their definition, with a group that reset capture
39        numbers, for example: /(?J:(?|(?'R')(\k'R')|((?'R'))))/. This has been
40        fixed by always allowing for more memory, even if not needed. (A proper fix
41        is implemented in PCRE2, but it involves more refactoring.)
42    
43    8.  There was no check for integer overflow in subroutine calls such as (?123).
44    
45    9.  The table entry for \l in EBCDIC environments was incorrect, leading to its
46        being treated as a literal 'l' instead of causing an error.
47    
48    10. There was a buffer overflow if pcre_exec() was called with an ovector of
49        size 1. This bug was found by american fuzzy lop.
50    
51    11. If a non-capturing group containing a conditional group that could match
52        an empty string was repeated, it was not identified as matching an empty
53        string itself. For example: /^(?:(?(1)x|)+)+$()/.
54    
55    12. In an EBCDIC environment, pcretest was mishandling the escape sequences
56        \a and \e in test subject lines.
57    
58    13. In an EBCDIC environment, \a in a pattern was converted to the ASCII
59        instead of the EBCDIC value.
60    
61    14. The handling of \c in an EBCDIC environment has been revised so that it is
62        now compatible with the specification in Perl's perlebcdic page.
63    
64    15. The EBCDIC character 0x41 is a non-breaking space, equivalent to 0xa0 in
65        ASCII/Unicode. This has now been added to the list of characters that are
66        recognized as white space in EBCDIC.
67    
68    16. When PCRE was compiled without UCP support, the use of \p and \P gave an
69        error (correctly) when used outside a class, but did not give an error
70        within a class.
71    
72    17. \h within a class was incorrectly compiled in EBCDIC environments.
73    
74    18. A pattern with an unmatched closing parenthesis that contained a backward
75        assertion which itself contained a forward reference caused buffer
76        overflow. And example pattern is: /(?=di(?<=(?1))|(?=(.))))/.
77    
78    19. JIT should return with error when the compiled pattern requires more stack
79        space than the maximum.
80    
81    20. A possessively repeated conditional group that could match an empty string,
82        for example, /(?(R))*+/, was incorrectly compiled.
83    
84    21. Fix infinite recursion in the JIT compiler when certain patterns such as
85        /(?:|a|){100}x/ are analysed.
86    
87    22. Some patterns with character classes involving [: and \\ were incorrectly
88        compiled and could cause reading from uninitialized memory or an incorrect
89        error diagnosis.
90    
91    23. Pathological patterns containing many nested occurrences of [: caused
92        pcre_compile() to run for a very long time.
93    
94    24. A conditional group with only one branch has an implicit empty alternative
95        branch and must therefore be treated as potentially matching an empty
96        string.
97    
98    25. If (?R was followed by - or + incorrect behaviour happened instead of a
99        diagnostic.
100    
101    26. Arrange to give up on finding the minimum matching length for overly
102        complex patterns.
103    
104    27. Similar to (4) above: in a pattern with duplicated named groups and an
105        occurrence of (?| it is possible for an apparently non-recursive back
106        reference to become recursive if a later named group with the relevant
107        number is encountered. This could lead to a buffer overflow. Wen Guanxing
108        from Venustech ADLAB discovered this bug.
109    
110    28. If pcregrep was given the -q option with -c or -l, or when handling a
111        binary file, it incorrectly wrote output to stdout.
112    
113    29. The JIT compiler did not restore the control verb head in case of *THEN
114        control verbs. This issue was found by Karl Skomski with a custom LLVM
115        fuzzer.
116    
117    30. Error messages for syntax errors following \g and \k were giving inaccurate
118        offsets in the pattern.
119    
120    31. Added a check for integer overflow in conditions (?(<digits>) and
121        (?(R<digits>). This omission was discovered by Karl Skomski with the LLVM
122        fuzzer.
123    
124    32. Handling recursive references such as (?2) when the reference is to a group
125        later in the pattern uses code that is very hacked about and error-prone.
126        It has been re-written for PCRE2. Here in PCRE1, a check has been added to
127        give an internal error if it is obvious that compiling has gone wrong.
128    
129    
130    Version 8.37 28-April-2015
131    --------------------------
132    
133    1.  When an (*ACCEPT) is triggered inside capturing parentheses, it arranges
134        for those parentheses to be closed with whatever has been captured so far.
135        However, it was failing to mark any other groups between the hightest
136        capture so far and the currrent group as "unset". Thus, the ovector for
137        those groups contained whatever was previously there. An example is the
138        pattern /(x)|((*ACCEPT))/ when matched against "abcd".
139    
140    2.  If an assertion condition was quantified with a minimum of zero (an odd
141        thing to do, but it happened), SIGSEGV or other misbehaviour could occur.
142    
143    3.  If a pattern in pcretest input had the P (POSIX) modifier followed by an
144        unrecognized modifier, a crash could occur.
145    
146    4.  An attempt to do global matching in pcretest with a zero-length ovector
147        caused a crash.
148    
149    5.  Fixed a memory leak during matching that could occur for a subpattern
150        subroutine call (recursive or otherwise) if the number of captured groups
151        that had to be saved was greater than ten.
152    
153    6.  Catch a bad opcode during auto-possessification after compiling a bad UTF
154        string with NO_UTF_CHECK. This is a tidyup, not a bug fix, as passing bad
155        UTF with NO_UTF_CHECK is documented as having an undefined outcome.
156    
157    7.  A UTF pattern containing a "not" match of a non-ASCII character and a
158        subroutine reference could loop at compile time. Example: /[^\xff]((?1))/.
159    
160    8. When a pattern is compiled, it remembers the highest back reference so that
161       when matching, if the ovector is too small, extra memory can be obtained to
162       use instead. A conditional subpattern whose condition is a check on a
163       capture having happened, such as, for example in the pattern
164       /^(?:(a)|b)(?(1)A|B)/, is another kind of back reference, but it was not
165       setting the highest backreference number. This mattered only if pcre_exec()
166       was called with an ovector that was too small to hold the capture, and there
167       was no other kind of back reference (a situation which is probably quite
168       rare). The effect of the bug was that the condition was always treated as
169       FALSE when the capture could not be consulted, leading to a incorrect
170       behaviour by pcre_exec(). This bug has been fixed.
171    
172    9. A reference to a duplicated named group (either a back reference or a test
173       for being set in a conditional) that occurred in a part of the pattern where
174       PCRE_DUPNAMES was not set caused the amount of memory needed for the pattern
175       to be incorrectly calculated, leading to overwriting.
176    
177    10. A mutually recursive set of back references such as (\2)(\1) caused a
178        segfault at study time (while trying to find the minimum matching length).
179        The infinite loop is now broken (with the minimum length unset, that is,
180        zero).
181    
182    11. If an assertion that was used as a condition was quantified with a minimum
183        of zero, matching went wrong. In particular, if the whole group had
184        unlimited repetition and could match an empty string, a segfault was
185        likely. The pattern (?(?=0)?)+ is an example that caused this. Perl allows
186        assertions to be quantified, but not if they are being used as conditions,
187        so the above pattern is faulted by Perl. PCRE has now been changed so that
188        it also rejects such patterns.
189    
190    12. A possessive capturing group such as (a)*+ with a minimum repeat of zero
191        failed to allow the zero-repeat case if pcre2_exec() was called with an
192        ovector too small to capture the group.
193    
194    13. Fixed two bugs in pcretest that were discovered by fuzzing and reported by
195        Red Hat Product Security:
196    
197        (a) A crash if /K and /F were both set with the option to save the compiled
198        pattern.
199    
200        (b) Another crash if the option to print captured substrings in a callout
201        was combined with setting a null ovector, for example \O\C+ as a subject
202        string.
203    
204    14. A pattern such as "((?2){0,1999}())?", which has a group containing a
205        forward reference repeated a large (but limited) number of times within a
206        repeated outer group that has a zero minimum quantifier, caused incorrect
207        code to be compiled, leading to the error "internal error:
208        previously-checked referenced subpattern not found" when an incorrect
209        memory address was read. This bug was reported as "heap overflow",
210        discovered by Kai Lu of Fortinet's FortiGuard Labs and given the CVE number
211        CVE-2015-2325.
212    
213    23. A pattern such as "((?+1)(\1))/" containing a forward reference subroutine
214        call within a group that also contained a recursive back reference caused
215        incorrect code to be compiled. This bug was reported as "heap overflow",
216        discovered by Kai Lu of Fortinet's FortiGuard Labs, and given the CVE
217        number CVE-2015-2326.
218    
219    24. Computing the size of the JIT read-only data in advance has been a source
220        of various issues, and new ones are still appear unfortunately. To fix
221        existing and future issues, size computation is eliminated from the code,
222        and replaced by on-demand memory allocation.
223    
224    25. A pattern such as /(?i)[A-`]/, where characters in the other case are
225        adjacent to the end of the range, and the range contained characters with
226        more than one other case, caused incorrect behaviour when compiled in UTF
227        mode. In that example, the range a-j was left out of the class.
228    
229    26. Fix JIT compilation of conditional blocks, which assertion
230        is converted to (*FAIL). E.g: /(?(?!))/.
231    
232    27. The pattern /(?(?!)^)/ caused references to random memory. This bug was
233        discovered by the LLVM fuzzer.
234    
235    28. The assertion (?!) is optimized to (*FAIL). This was not handled correctly
236        when this assertion was used as a condition, for example (?(?!)a|b). In
237        pcre2_match() it worked by luck; in pcre2_dfa_match() it gave an incorrect
238        error about an unsupported item.
239    
240    29. For some types of pattern, for example /Z*(|d*){216}/, the auto-
241        possessification code could take exponential time to complete. A recursion
242        depth limit of 1000 has been imposed to limit the resources used by this
243        optimization.
244    
245    30. A pattern such as /(*UTF)[\S\V\H]/, which contains a negated special class
246        such as \S in non-UCP mode, explicit wide characters (> 255) can be ignored
247        because \S ensures they are all in the class. The code for doing this was
248        interacting badly with the code for computing the amount of space needed to
249        compile the pattern, leading to a buffer overflow. This bug was discovered
250        by the LLVM fuzzer.
251    
252    31. A pattern such as /((?2)+)((?1))/ which has mutual recursion nested inside
253        other kinds of group caused stack overflow at compile time. This bug was
254        discovered by the LLVM fuzzer.
255    
256    32. A pattern such as /(?1)(?#?'){8}(a)/ which had a parenthesized comment
257        between a subroutine call and its quantifier was incorrectly compiled,
258        leading to buffer overflow or other errors. This bug was discovered by the
259        LLVM fuzzer.
260    
261    33. The illegal pattern /(?(?<E>.*!.*)?)/ was not being diagnosed as missing an
262        assertion after (?(. The code was failing to check the character after
263        (?(?< for the ! or = that would indicate a lookbehind assertion. This bug
264        was discovered by the LLVM fuzzer.
265    
266    34. A pattern such as /X((?2)()*+){2}+/ which has a possessive quantifier with
267        a fixed maximum following a group that contains a subroutine reference was
268        incorrectly compiled and could trigger buffer overflow. This bug was
269        discovered by the LLVM fuzzer.
270    
271    35. A mutual recursion within a lookbehind assertion such as (?<=((?2))((?1)))
272        caused a stack overflow instead of the diagnosis of a non-fixed length
273        lookbehind assertion. This bug was discovered by the LLVM fuzzer.
274    
275    36. The use of \K in a positive lookbehind assertion in a non-anchored pattern
276        (e.g. /(?<=\Ka)/) could make pcregrep loop.
277    
278    37. There was a similar problem to 36 in pcretest for global matches.
279    
280    38. If a greedy quantified \X was preceded by \C in UTF mode (e.g. \C\X*),
281        and a subsequent item in the pattern caused a non-match, backtracking over
282        the repeated \X did not stop, but carried on past the start of the subject,
283        causing reference to random memory and/or a segfault. There were also some
284        other cases where backtracking after \C could crash. This set of bugs was
285        discovered by the LLVM fuzzer.
286    
287    39. The function for finding the minimum length of a matching string could take
288        a very long time if mutual recursion was present many times in a pattern,
289        for example, /((?2){73}(?2))((?1))/. A better mutual recursion detection
290        method has been implemented. This infelicity was discovered by the LLVM
291        fuzzer.
292    
293    40. Static linking against the PCRE library using the pkg-config module was
294        failing on missing pthread symbols.
295    
296    
297    Version 8.36 26-September-2014
298    ------------------------------
299    
300    1.  Got rid of some compiler warnings in the C++ modules that were shown up by
301        -Wmissing-field-initializers and -Wunused-parameter.
302    
303    2.  The tests for quantifiers being too big (greater than 65535) were being
304        applied after reading the number, and stupidly assuming that integer
305        overflow would give a negative number. The tests are now applied as the
306        numbers are read.
307    
308    3.  Tidy code in pcre_exec.c where two branches that used to be different are
309        now the same.
310    
311    4.  The JIT compiler did not generate match limit checks for certain
312        bracketed expressions with quantifiers. This may lead to exponential
313        backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
314        issue should be resolved now.
315    
316    5.  Fixed an issue, which occures when nested alternatives are optimized
317        with table jumps.
318    
319    6.  Inserted two casts and changed some ints to size_t in the light of some
320        reported 64-bit compiler warnings (Bugzilla 1477).
321    
322    7.  Fixed a bug concerned with zero-minimum possessive groups that could match
323        an empty string, which sometimes were behaving incorrectly in the
324        interpreter (though correctly in the JIT matcher). This pcretest input is
325        an example:
326    
327          '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
328          NON QUOTED "QUOT""ED" AFTER "NOT MATCHED
329    
330        the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
331        JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
332        for an empty string was breaking the inner loop and carrying on at a lower
333        level, when possessive repeated groups should always return to a higher
334        level as they have no backtrack points in them. The empty string test now
335        occurs at the outer level.
336    
337    8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
338        ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".
339    
340    9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
341        doesn't).
342    
343    10. Change 8.34/15 introduced a bug that caused the amount of memory needed
344        to hold a pattern to be incorrectly computed (too small) when there were
345        named back references to duplicated names. This could cause "internal
346        error: code overflow" or "double free or corruption" or other memory
347        handling errors.
348    
349    11. When named subpatterns had the same prefixes, back references could be
350        confused. For example, in this pattern:
351    
352          /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/
353    
354        the reference to 'Name' was incorrectly treated as a reference to a
355        duplicate name.
356    
357    12. A pattern such as /^s?c/mi8 where the optional character has more than
358        one "other case" was incorrectly compiled such that it would only try to
359        match starting at "c".
360    
361    13. When a pattern starting with \s was studied, VT was not included in the
362        list of possible starting characters; this should have been part of the
363        8.34/18 patch.
364    
365    14. If a character class started [\Qx]... where x is any character, the class
366        was incorrectly terminated at the ].
367    
368    15. If a pattern that started with a caseless match for a character with more
369        than one "other case" was studied, PCRE did not set up the starting code
370        unit bit map for the list of possible characters. Now it does. This is an
371        optimization improvement, not a bug fix.
372    
373    16. The Unicode data tables have been updated to Unicode 7.0.0.
374    
375    17. Fixed a number of memory leaks in pcregrep.
376    
377    18. Avoid a compiler warning (from some compilers) for a function call with
378        a cast that removes "const" from an lvalue by using an intermediate
379        variable (to which the compiler does not object).
380    
381    19. Incorrect code was compiled if a group that contained an internal recursive
382        back reference was optional (had quantifier with a minimum of zero). This
383        example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
384        caused segmentation faults because of stack overflows at compile time.
385    
386    20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
387        group that is quantified with an indefinite repeat, caused a compile-time
388        loop which used up all the system stack and provoked a segmentation fault.
389        This was not the same bug as 19 above.
390    
391    21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
392        Patch by Mike Frysinger.
393    
394    
395    Version 8.35 04-April-2014
396    --------------------------
397    
398    1.  A new flag is set, when property checks are present in an XCLASS.
399        When this flag is not set, PCRE can perform certain optimizations
400        such as studying these XCLASS-es.
401    
402    2.  The auto-possessification of character sets were improved: a normal
403        and an extended character set can be compared now. Furthermore
404        the JIT compiler optimizes more character set checks.
405    
406    3.  Got rid of some compiler warnings for potentially uninitialized variables
407        that show up only when compiled with -O2.
408    
409    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
410        of a match later then the end of the match. The pcretest program was not
411        handling the case sensibly - it was outputting from the start to the next
412        binary zero. It now reports this situation in a message, and outputs the
413        text from the end to the start.
414    
415    5.  Fast forward search is improved in JIT. Instead of the first three
416        characters, any three characters with fixed position can be searched.
417        Search order: first, last, middle.
418    
419    6.  Improve character range checks in JIT. Characters are read by an inprecise
420        function now, which returns with an unknown value if the character code is
421        above a certain threshold (e.g: 256). The only limitation is that the value
422        must be bigger than the threshold as well. This function is useful when
423        the characters above the threshold are handled in the same way.
424    
425    7.  The macros whose names start with RAWUCHAR are placeholders for a future
426        mode in which only the bottom 21 bits of 32-bit data items are used. To
427        make this more memorable for those maintaining the code, the names have
428        been changed to start with UCHAR21, and an extensive comment has been added
429        to their definition.
430    
431    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
432        to the export list in Makefile.am (they were accidentally omitted from the
433        8.34 tarball).
434    
435    9.  The informational output from pcretest used the phrase "starting byte set"
436        which is inappropriate for the 16-bit and 32-bit libraries. As the output
437        for "first char" and "need char" really means "non-UTF-char", I've changed
438        "byte" to "char", and slightly reworded the output. The documentation about
439        these values has also been (I hope) clarified.
440    
441    10. Another JIT related optimization: use table jumps for selecting the correct
442        backtracking path, when more than four alternatives are present inside a
443        bracket.
444    
445    11. Empty match is not possible, when the minimum length is greater than zero,
446        and there is no \K in the pattern. JIT should avoid empty match checks in
447        such cases.
448    
449    12. In a caseless character class with UCP support, when a character with more
450        than one alternative case was not the first character of a range, not all
451        the alternative cases were added to the class. For example, s and \x{17f}
452        are both alternative cases for S: the class [RST] was handled correctly,
453        but [R-T] was not.
454    
455    13. The configure.ac file always checked for pthread support when JIT was
456        enabled. This is not used in Windows, so I have put this test inside a
457        check for the presence of windows.h (which was already tested for).
458    
459    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
460        The algorithm provides a way to skip certain starting offsets, and usually
461        faster than linear prefix searches.
462    
463    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
464        as for 'fr_FR' and 'french'. For some reason, however, it then used the
465        Windows-specific input and output files, which have 'french' screwed in.
466        So this could never have worked. One of the problems with locales is that
467        they aren't always the same. I have now updated RunTest so that it checks
468        the output of the locale test (test 3) against three different output
469        files, and it allows the test to pass if any one of them matches. With luck
470        this should make the test pass on some versions of Solaris where it was
471        failing. Because of the uncertainty, the script did not used to stop if
472        test 3 failed; it now does. If further versions of a French locale ever
473        come to light, they can now easily be added.
474    
475    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
476        there was a message during ./configure, but it did not stop. This now
477        provokes an error. The invalid example in README has been corrected.
478        If a value less than the minimum is given, the minimum value has always
479        been used, but now a warning is given.
480    
481    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
482        was a bug in the test system, which is now fixed. Also, the list of various
483        configurations that are tested for each release did not have one with both
484        16/32 bits and --enable-bar-anycrlf. It now does.
485    
486    18. pcretest was missing "-C bsr" for displaying the \R default setting.
487    
488    19. Little endian PowerPC systems are supported now by the JIT compiler.
489    
490    20. The fast forward newline mechanism could enter to an infinite loop on
491        certain invalid UTF-8 input. Although we don't support these cases
492        this issue can be fixed by a performance optimization.
493    
494    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
495        not take account if existing stack usage. There is now a new global
496        variable called pcre_stack_guard that can be set to point to an external
497        function to check stack availability. It is called at the start of
498        processing every parenthesized group.
499    
500    22. A typo in the code meant that in ungreedy mode the max/min qualifier
501        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
502        not match "ab".
503    
504    23. When UTF was disabled, the JIT program reported some incorrect compile
505        errors. These messages are silenced now.
506    
507    24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
508        compiler.
509    
510    25. Change all the temporary files used in RunGrepTest to be different to those
511        used by RunTest so that the tests can be run simultaneously, for example by
512        "make -j check".
513    
514    
515    Version 8.34 15-December-2013
516    -----------------------------
517    
518    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
519        executable memory. Patch inspired by Carsten Klein.
520    
521    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
522        this macro is never tested and has no effect, because the work to support
523        coverage involves only compiling and linking options and special targets in
524        the Makefile. The comment in config.h implied that defining the macro would
525        enable coverage support, which is totally false. There was also support for
526        setting this macro in the CMake files (my fault, I just copied it from
527        configure). SUPPORT_GCOV has now been removed.
528    
529    3.  Make a small performance improvement in strlen16() and strlen32() in
530        pcretest.
531    
532    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
533        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
534        these cases). There was also one in pcretest.c.
535    
536    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
537    
538    6.  In UTF mode, the code for checking whether a group could match an empty
539        string (which is used for indefinitely repeated groups to allow for
540        breaking an infinite loop) was broken when the group contained a repeated
541        negated single-character class with a character that occupied more than one
542        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
543        UTF-8 mode). The effect was undefined: the group might or might not be
544        deemed as matching an empty string, or the program might have crashed.
545    
546    7.  The code for checking whether a group could match an empty string was not
547        recognizing that \h, \H, \v, \V, and \R must match a character.
548    
549    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
550        an empty string. If it can, pcretest shows this in its information output.
551    
552    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
553        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
554        matched by pcre_exec() without using JIT:
555    
556        (a) If the rest of the pattern did not match after a maximal run of
557            grapheme clusters, the code for backing up to try with fewer of them
558            did not always back up over a full grapheme when characters that do not
559            have the modifier quality were involved, e.g. Hangul syllables.
560    
561        (b) If the match point in a subject started with modifier character, and
562            there was no match, the code could incorrectly back up beyond the match
563            point, and potentially beyond the first character in the subject,
564            leading to a segfault or an incorrect match result.
565    
566    10. A conditional group with an assertion condition could lead to PCRE
567        recording an incorrect first data item for a match if no other first data
568        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
569        first data item, and therefore matched "ca" after "c" instead of at the
570        start.
571    
572    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
573        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
574    
575    12. The source of pcregrep now includes z/OS-specific code so that it can be
576        compiled for z/OS as part of the special z/OS distribution.
577    
578    13. Added the -T and -TM options to pcretest.
579    
580    14. The code in pcre_compile.c for creating the table of named capturing groups
581        has been refactored. Instead of creating the table dynamically during the
582        actual compiling pass, the information is remembered during the pre-compile
583        pass (on the stack unless there are more than 20 named groups, in which
584        case malloc() is used) and the whole table is created before the actual
585        compile happens. This has simplified the code (it is now nearly 150 lines
586        shorter) and prepared the way for better handling of references to groups
587        with duplicate names.
588    
589    15. A back reference to a named subpattern when there is more than one of the
590        same name now checks them in the order in which they appear in the pattern.
591        The first one that is set is used for the reference. Previously only the
592        first one was inspected. This change makes PCRE more compatible with Perl.
593    
594    16. Unicode character properties were updated from Unicode 6.3.0.
595    
596    17. The compile-time code for auto-possessification has been refactored, based
597        on a patch by Zoltan Herczeg. It now happens after instead of during
598        compilation. The code is cleaner, and more cases are handled. The option
599        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
600        options in pcretest are provided to set it. It can also be set by
601        (*NO_AUTO_POSSESS) at the start of a pattern.
602    
603    18. The character VT has been added to the default ("C" locale) set of
604        characters that match \s and are generally treated as white space,
605        following this same change in Perl 5.18. There is now no difference between
606        "Perl space" and "POSIX space". Whether VT is treated as white space in
607        other locales depends on the locale.
608    
609    19. The code for checking named groups as conditions, either for being set or
610        for being recursed, has been refactored (this is related to 14 and 15
611        above). Processing unduplicated named groups should now be as fast at
612        numerical groups, and processing duplicated groups should be faster than
613        before.
614    
615    20. Two patches to the CMake build system, by Alexander Barkov:
616    
617          (1) Replace the "source" command by "." in CMakeLists.txt because
618              "source" is a bash-ism.
619    
620          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
621              without these the CMake build does not work on Solaris.
622    
623    21. Perl has changed its handling of \8 and \9. If there is no previously
624        encountered capturing group of those numbers, they are treated as the
625        literal characters 8 and 9 instead of a binary zero followed by the
626        literals. PCRE now does the same.
627    
628    22. Following Perl, added \o{} to specify codepoints in octal, making it
629        possible to specify values greater than 0777 and also making them
630        unambiguous.
631    
632    23. Perl now gives an error for missing closing braces after \x{... instead of
633        treating the string as literal. PCRE now does the same.
634    
635    24. RunTest used to grumble if an inappropriate test was selected explicitly,
636        but just skip it when running all tests. This make it awkward to run ranges
637        of tests when one of them was inappropriate. Now it just skips any
638        inappropriate tests, as it always did when running all tests.
639    
640    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
641        character types such as \d or \w, too many callouts were inserted, and the
642        data that they returned was rubbish.
643    
644    26. In UCP mode, \s was not matching two of the characters that Perl matches,
645        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
646        were matched by \h. The code has now been refactored so that the lists of
647        the horizontal and vertical whitespace characters used for \h and \v (which
648        are defined only in one place) are now also used for \s.
649    
650    27. Add JIT support for the 64 bit TileGX architecture.
651        Patch by Jiong Wang (Tilera Corporation).
652    
653    28. Possessive quantifiers for classes (both explicit and automatically
654        generated) now use special opcodes instead of wrapping in ONCE brackets.
655    
656    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
657        (because it's meaningless), this was not happening when PCRE_CASELESS was
658        set. Not wrong, but inefficient.
659    
660    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
661        properties for \w, \d, etc) is present in a test regex. Otherwise if the
662        test contains no characters greater than 255, Perl doesn't realise it
663        should be using Unicode semantics.
664    
665    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
666        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
667        does in Unicode mode.
668    
669    32. Added the "forbid" facility to pcretest so that putting tests into the
670        wrong test files can sometimes be quickly detected.
671    
672    33. There is now a limit (default 250) on the depth of nesting of parentheses.
673        This limit is imposed to control the amount of system stack used at compile
674        time. It can be changed at build time by --with-parens-nest-limit=xxx or
675        the equivalent in CMake.
676    
677    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
678        errors. Perl warns for these when in warning mode, but PCRE has no facility
679        for giving warnings.
680    
681    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
682        However, this was not working for (?!) because it is optimized to (*FAIL),
683        for which PCRE does not allow quantifiers. The optimization is now disabled
684        when a quantifier follows (?!). I can't see any use for this, but it makes
685        things uniform.
686    
687    36. Perl no longer allows group names to start with digits, so I have made this
688        change also in PCRE. It simplifies the code a bit.
689    
690    37. In extended mode, Perl ignores spaces before a + that indicates a
691        possessive quantifier. PCRE allowed a space before the quantifier, but not
692        before the possessive +. It now does.
693    
694    38. The use of \K (reset reported match start) within a repeated possessive
695        group such as (a\Kb)*+ was not working.
696    
697    40. Document that the same character tables must be used at compile time and
698        run time, and that the facility to pass tables to pcre_exec() and
699        pcre_dfa_exec() is for use only with saved/restored patterns.
700    
701    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
702        features for Builds with MSVC:
703    
704        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
705           builds). The .rc files can be used to set FileDescription and many other
706           attributes.
707    
708        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
709           This allows higher-level build scripts which want .pdb files to avoid
710           hard-coding the exact files needed."
711    
712    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
713        mean "start of word" and "end of word", respectively, as a transition aid.
714    
715    43. A minimizing repeat of a class containing codepoints greater than 255 in
716        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
717        compiled to use the heap for recursion.
718    
719    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
720        is configured.
721    
722    
723    Version 8.33 28-May-2013
724    ------------------------
725    
726    1.  Added 'U' to some constants that are compared to unsigned integers, to
727        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
728        variables that are added to signed variables, to ensure the result is
729        signed and can be negated.
730    
731    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
732        CMake config files.
733    
734    3.  Revise the creation of config.h.generic so that all boolean macros are
735        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
736        overriding via -D on the command line possible.
737    
738    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
739        to unsigned int is reported to make a quite noticeable speed difference in
740        a specific Windows environment. Testing on Linux did also appear to show
741        some benefit (and it is clearly not harmful). Also fixed the definition of
742        Xop which should be unsigned.
743    
744    5.  Related to (4), changing the definition of the intermediate variable cc
745        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
746        improvements.
747    
748    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
749        unnecessary spaces.
750    
751    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
752        and later.
753    
754    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
755    
756    9.  Optimizing fast_forward_start_bits in JIT.
757    
758    10. Adding support for callouts in JIT, and fixing some issues revealed
759        during this work. Namely:
760    
761        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
762    
763        (b) Minimum length was not checked before the matching is started.
764    
765    11. The value of capture_last that is passed to callouts was incorrect in some
766        cases when there was a capture on one path that was subsequently abandoned
767        after a backtrack. Also, the capture_last value is now reset after a
768        recursion, since all captures are also reset in this case.
769    
770    12. The interpreter no longer returns the "too many substrings" error in the
771        case when an overflowing capture is in a branch that is subsequently
772        abandoned after a backtrack.
773    
774    13. In the pathological case when an offset vector of size 2 is used, pcretest
775        now prints out the matched string after a yield of 0 or 1.
776    
777    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
778        Only supported by the JIT compiler at the moment.
779    
780    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
781    
782    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
783        offset of the starting point of the matching process, provided the offsets
784        vector is large enough.
785    
786    17. The \A escape now records a lookbehind value of 1, though its execution
787        does not actually inspect the previous character. This is to ensure that,
788        in partial multi-segment matching, at least one character from the old
789        segment is retained when a new segment is processed. Otherwise, if there
790        are no lookbehinds in the pattern, \A might match incorrectly at the start
791        of a new segment.
792    
793    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
794    
795    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
796        gives some modest performance improvement in 8-bit mode.
797    
798    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
799        be expressed in certain programming languages using Universal Character
800        Names.
801    
802    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
803        which points out that "non characters" are not "characters that may not
804        appear in Unicode strings" but rather "characters that are reserved for
805        internal use and have only local meaning".
806    
807    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
808        there was a conditional group that depended on an assertion, if the
809        assertion was false, the callout that immediately followed the alternation
810        in the condition was skipped when pcre_exec() was used for matching.
811    
812    23. Allow an explicit callout to be inserted before an assertion that is the
813        condition for a conditional group, for compatibility with automatic
814        callouts, which always insert a callout at this point.
815    
816    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
817        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
818    
819    25. (*PRUNE) is now supported by the JIT compiler.
820    
821    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
822    
823    27. Fix the case where there are two or more SKIPs with arguments that may be
824        ignored.
825    
826    28. (*SKIP) is now supported by the JIT compiler.
827    
828    29. (*THEN) is now supported by the JIT compiler.
829    
830    30. Update RunTest with additional test selector options.
831    
832    31. The way PCRE handles backtracking verbs has been changed in two ways.
833    
834        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
835        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
836        backtracking. In some cases this makes it more Perl-compatible, but Perl's
837        rather obscure rules do not always do the same thing.
838    
839        (2) Previously, backtracking verbs were confined within assertions. This is
840        no longer the case for positive assertions, except for (*ACCEPT). Again,
841        this sometimes improves Perl compatibility, and sometimes does not.
842    
843    32. A number of tests that were in test 2 because Perl did things differently
844        have been moved to test 1, because either Perl or PCRE has changed, and
845        these tests are now compatible.
846    
847    32. Backtracking control verbs are now handled in the same way in JIT and
848        interpreter.
849    
850    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
851        contained a forward subroutine reference caused a compile error.
852    
853    34. Auto-detect and optimize limited repetitions in JIT.
854    
855    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
856        blocking (*UTF) etc.
857    
858    36. In the interpreter, maximizing pattern repetitions for characters and
859        character types now use tail recursion, which reduces stack usage.
860    
861    37. The value of the max lookbehind was not correctly preserved if a compiled
862        and saved regex was reloaded on a host of different endianness.
863    
864    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
865        of the compiled pattern block, expand the flags field from 16 to 32 bits
866        because it was almost full.
867    
868    39. Try madvise first before posix_madvise.
869    
870    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
871        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
872        original change locked out all matches of empty strings. This has been
873        changed so that one match of an empty string per line is recognized.
874        Subsequent searches on the same line (for colouring or for --only-matching,
875        for example) do not recognize empty strings.
876    
877    41. Applied a user patch to fix a number of spelling mistakes in comments.
878    
879    42. Data lines longer than 65536 caused pcretest to crash.
880    
881    43. Clarified the data type for length and startoffset arguments for pcre_exec
882        and pcre_dfa_exec in the function-specific man pages, where they were
883        explicitly stated to be in bytes, never having been updated. I also added
884        some clarification to the pcreapi man page.
885    
886    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
887        a segmentation fault.
888    
889    
890    Version 8.32 30-November-2012
891    -----------------------------
892    
893    1.  Improved JIT compiler optimizations for first character search and single
894        character iterators.
895    
896    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
897        Patch by Daniel Richard G.
898    
899    3.  Single character iterator optimizations in the JIT compiler.
900    
901    4.  Improved JIT compiler optimizations for character ranges.
902    
903    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
904        Reported by Giuseppe D'Angelo.
905    
906    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
907        of a line, was being set incorrectly in cases where .* appeared inside
908        atomic brackets at the start of a pattern, or where there was a subsequent
909        *PRUNE or *SKIP.
910    
911    7.  Improved instruction cache flush for POWER/PowerPC.
912        Patch by Daniel Richard G.
913    
914    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
915        grep:
916    
917        (a) There is now no limit to the number of patterns to be matched.
918    
919        (b) An error is given if a pattern is too long.
920    
921        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
922            are now supported.
923    
924        (d) --exclude-from and --include-from (multiple use) have been added.
925    
926        (e) Exclusions and inclusions now apply to all files and directories, not
927            just to those obtained from scanning a directory recursively.
928    
929        (f) Multiple uses of -f and --file-list are now supported.
930    
931        (g) In a Windows environment, the default for -d has been changed from
932            "read" (the GNU grep default) to "skip", because otherwise the presence
933            of a directory in the file list provokes an error.
934    
935        (h) The documentation has been revised and clarified in places.
936    
937    9.  Improve the matching speed of capturing brackets.
938    
939    10. Changed the meaning of \X so that it now matches a Unicode extended
940        grapheme cluster.
941    
942    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
943        out POSIX threads when JIT support is configured.
944    
945    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
946    
947    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
948        structure could go wrong in environments where size_t is not the same size
949        as int.
950    
951    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
952        set.
953    
954    15. The EBCDIC support had decayed; later updates to the code had included
955        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
956        general tidy up of EBCDIC-related issues, and the documentation was also
957        not quite right. There is now a test that can be run on ASCII systems to
958        check some of the EBCDIC-related things (but is it not a full test).
959    
960    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
961        in a small tidy to the code.
962    
963    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
964    
965    18. If the --only-matching (-o) option in pcregrep is specified multiple
966        times, each one causes appropriate output. For example, -o1 -o2 outputs the
967        substrings matched by the 1st and 2nd capturing parentheses. A separating
968        string can be specified by --om-separator (default empty).
969    
970    19. Improving the first n character searches.
971    
972    20. Turn case lists for horizontal and vertical white space into macros so that
973        they are defined only once.
974    
975    21. This set of changes together give more compatible Unicode case-folding
976        behaviour for characters that have more than one other case when UCP
977        support is available.
978    
979        (a) The Unicode property table now has offsets into a new table of sets of
980            three or more characters that are case-equivalent. The MultiStage2.py
981            script that generates these tables (the pcre_ucd.c file) now scans
982            CaseFolding.txt instead of UnicodeData.txt for character case
983            information.
984    
985        (b) The code for adding characters or ranges of characters to a character
986            class has been abstracted into a generalized function that also handles
987            case-independence. In UTF-mode with UCP support, this uses the new data
988            to handle characters with more than one other case.
989    
990        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
991            whose other case is greater than 256 are now correctly matched
992            caselessly. Previously, the high codepoint matched the low one, but not
993            vice versa.
994    
995        (d) The processing of \h, \H, \v, and \ in character classes now makes use
996            of the new class addition function, using character lists defined as
997            macros alongside the case definitions of 20 above.
998    
999        (e) Caseless back references now work with characters that have more than
1000            one other case.
1001    
1002        (f) General caseless matching of characters with more than one other case
1003            is supported.
1004    
1005    22. Unicode character properties were updated from Unicode 6.2.0
1006    
1007    23. Improved CMake support under Windows. Patch by Daniel Richard G.
1008    
1009    24. Add support for 32-bit character strings, and UTF-32
1010    
1011    25. Major JIT compiler update (code refactoring and bugfixing).
1012        Experimental Sparc 32 support is added.
1013    
1014    26. Applied a modified version of Daniel Richard G's patch to create
1015        pcre.h.generic and config.h.generic by "make" instead of in the
1016        PrepareRelease script.
1017    
1018    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
1019        pcre_compile.c when checking for a zero character.
1020    
1021    28. Introducing a native interface for JIT. Through this interface, the compiled
1022        machine code can be directly executed. The purpose of this interface is to
1023        provide fast pattern matching, so several sanity checks are not performed.
1024        However, feature tests are still performed. The new interface provides
1025        1.4x speedup compared to the old one.
1026    
1027    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
1028        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
1029        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
1030    
1031    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
1032        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
1033        the "old" RFC 2279). Instead, it ended up passing a negative length to
1034        pcre_exec().
1035    
1036    31. Add support for GCC's visibility feature to hide internal functions.
1037    
1038    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
1039        "unknown -C option" after outputting 0 or 1.
1040    
1041    33. There is now support for generating a code coverage report for the test
1042        suite in environments where gcc is the compiler and lcov is installed. This
1043        is mainly for the benefit of the developers.
1044    
1045    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
1046        unaddressable using valgrind annotations, allowing valgrind to detect
1047        invalid memory accesses. This is mainly for the benefit of the developers.
1048    
1049    25. (*UTF) can now be used to start a pattern in any of the three libraries.
1050    
1051    26. Give configure error if --enable-cpp but no C++ compiler found.
1052    
1053    
1054    Version 8.31 06-July-2012
1055    -------------------------
1056    
1057    1.  Fixing a wrong JIT test case and some compiler warnings.
1058    
1059    2.  Removed a bashism from the RunTest script.
1060    
1061    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
1062        to unsigned type, result still unsigned" that was given by an MS compiler
1063        on encountering the code "-sizeof(xxx)".
1064    
1065    4.  Partial matching support is added to the JIT compiler.
1066    
1067    5.  Fixed several bugs concerned with partial matching of items that consist
1068        of more than one character:
1069    
1070        (a) /^(..)\1/ did not partially match "aba" because checking references was
1071            done on an "all or nothing" basis. This also applied to repeated
1072            references.
1073    
1074        (b) \R did not give a hard partial match if \r was found at the end of the
1075            subject.
1076    
1077        (c) \X did not give a hard partial match after matching one or more
1078            characters at the end of the subject.
1079    
1080        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
1081            a partial match for the string "\r".
1082    
1083        (e) When newline was set to CRLF, the metacharacter "." did not recognize
1084            a partial match for a CR character at the end of the subject string.
1085    
1086    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
1087        running pcretest, the text "(JIT)" added to the output whenever JIT is
1088        actually used to run the match.
1089    
1090    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
1091        or /S+[+] with a digit between 1 and 7.
1092    
1093    8.  OP_NOT now supports any UTF character not just single-byte ones.
1094    
1095    9.  (*MARK) control verb is now supported by the JIT compiler.
1096    
1097    10. The command "./RunTest list" lists the available tests without actually
1098        running any of them. (Because I keep forgetting what they all are.)
1099    
1100    11. Add PCRE_INFO_MAXLOOKBEHIND.
1101    
1102    12. Applied a (slightly modified) user-supplied patch that improves performance
1103        when the heap is used for recursion (compiled with --disable-stack-for-
1104        recursion). Instead of malloc and free for each heap frame each time a
1105        logical recursion happens, frames are retained on a chain and re-used where
1106        possible. This sometimes gives as much as 30% improvement.
1107    
1108    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
1109        call.
1110    
1111    14. As documented, (*COMMIT) is now confined to within a positive assertion.
1112    
1113    15. It is now possible to link pcretest with libedit as an alternative to
1114        libreadline.
1115    
1116    16. (*COMMIT) control verb is now supported by the JIT compiler.
1117    
1118    17. The Unicode data tables have been updated to Unicode 6.1.0.
1119    
1120    18. Added --file-list option to pcregrep.
1121    
1122    19. Added binary file support to pcregrep, including the -a, --binary-files,
1123        -I, and --text options.
1124    
1125    20. The madvise function is renamed for posix_madvise for QNX compatibility
1126        reasons. Fixed by Giuseppe D'Angelo.
1127    
1128    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
1129    
1130    22. Changed the option for creating symbolic links for 16-bit man pages from
1131        -s to -sf so that re-installing does not cause issues.
1132    
1133    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
1134    
1135    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
1136        matches in certain environments (the workspace was not being correctly
1137        retained). Also added to pcre_dfa_exec() a simple plausibility check on
1138        some of the workspace data at the beginning of a restart.
1139    
1140    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
1141        was not doing so when it should - probably a typo introduced by SVN 528
1142        (change 8.10/14).
1143    
1144    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
1145        \w+ when the character tables indicated that \x{c4} was a word character.
1146        There were several related cases, all because the tests for doing a table
1147        lookup were testing for characters less than 127 instead of 255.
1148    
1149    27. If a pattern contains capturing parentheses that are not used in a match,
1150        their slots in the ovector are set to -1. For those that are higher than
1151        any matched groups, this happens at the end of processing. In the case when
1152        there were back references that the ovector was too small to contain
1153        (causing temporary malloc'd memory to be used during matching), and the
1154        highest capturing number was not used, memory off the end of the ovector
1155        was incorrectly being set to -1. (It was using the size of the temporary
1156        memory instead of the true size.)
1157    
1158    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
1159        ovector size, it uses memory at the end of the block that it has got.
1160    
1161    29. Check for an overlong MARK name and give an error at compile time. The
1162        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
1163    
1164    30. JIT compiler update.
1165    
1166    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
1167        Rill for the patch.
1168    
1169    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
1170    
1171    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
1172    
1173    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
1174        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
1175        was enabled.
1176    
1177    35. Improve JIT code generation for greedy plus quantifier.
1178    
1179    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
1180        1 to "aa" instead of to an empty string. The bug affected repeated groups
1181        that could potentially match an empty string.
1182    
1183    37. Optimizing single character iterators in JIT.
1184    
1185    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
1186        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
1187        codepoints that are too big for the mode are faulted, and in a UTF mode,
1188        disallowed codepoints are also faulted.
1189    
1190    39. If PCRE was compiled with UTF support, in three places in the DFA
1191        matcher there was code that should only have been obeyed in UTF mode, but
1192        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
1193        processing when bytes with values greater than 127 were present. In 16-bit
1194        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
1195        both cases the values are those that cannot be the first data item in a UTF
1196        character. The three items that might have provoked this were recursions,
1197        possessively repeated groups, and atomic groups.
1198    
1199    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
1200        and pcregrep, because some OS require shared objects to be explicitly
1201        passed to ld, causing the link step to fail if they are not.
1202    
1203    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
1204        mode, patterns that started with \h* or \R* might be incorrectly matched.
1205    
1206    
1207    Version 8.30 04-February-2012
1208    -----------------------------
1209    
1210    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
1211        name is defined in ctype.h.
1212    
1213    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
1214        only in quite long subpatterns.
1215    
1216    3.  Removed the function pcre_info(), which has been obsolete and deprecated
1217        since it was replaced by pcre_fullinfo() in February 2000.
1218    
1219    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
1220        match a (*MARK), and the match failed at the start of the subject, a
1221        reference to memory before the start of the subject could occur. This bug
1222        was introduced by fix 17 of release 8.21.
1223    
1224    5.  A reference to an unset group with zero minimum repetition was giving
1225        totally wrong answers (in non-JavaScript-compatibility mode). For example,
1226        /(another)?(\1?)test/ matched against "hello world test". This bug was
1227        introduced in release 8.13.
1228    
1229    6.  Add support for 16-bit character strings (a large amount of work involving
1230        many changes and refactorings).
1231    
1232    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
1233        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
1234        from a file.
1235    
1236    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
1237        rounding is not applied in this particular case).
1238    
1239    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
1240        if they appear, or are escaped, in patterns.
1241    
1242    10. Get rid of a number of -Wunused-but-set-variable warnings.
1243    
1244    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
1245        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
1246        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
1247        also returns the mark "x". This bug applied to capturing parentheses,
1248        non-capturing parentheses, and atomic parentheses. It also applied to some
1249        assertions.
1250    
1251    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
1252        information out of configure.ac instead of relying on pcre.h.generic, which
1253        is not stored in the repository.
1254    
1255    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
1256        -lreadline.
1257    
1258    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
1259    
1260    15. Applied Graycode's patch to put the top-level frame on the stack rather
1261        than the heap when not using the stack for recursion. This gives a
1262        performance improvement in many cases when recursion is not deep.
1263    
1264    16. Experimental code added to "pcretest -C" to output the stack frame size.
1265    
1266    
1267    Version 8.21 12-Dec-2011
1268    ------------------------
1269    
1270    1.  Updating the JIT compiler.
1271    
1272    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
1273        are added as well.
1274    
1275    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
1276        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
1277        calling _pcre_jit_exec. Some extra comments are added.
1278    
1279    4.  (*MARK) settings inside atomic groups that do not contain any capturing
1280        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
1281        was introduced by change 18 for 8.20.
1282    
1283    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
1284        ECMA-262 standard.
1285    
1286    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
1287        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
1288        This bug was probably introduced by change 9 of 8.13.
1289    
1290    7.  While fixing 6 above, I noticed that a number of other items were being
1291        incorrectly rejected as "not fixed length". This arose partly because newer
1292        opcodes had not been added to the fixed-length checking code. I have (a)
1293        corrected the bug and added tests for these items, and (b) arranged for an
1294        error to occur if an unknown opcode is encountered while checking for fixed
1295        length instead of just assuming "not fixed length". The items that were
1296        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
1297        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
1298        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
1299    
1300    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
1301        being incorrectly compiled and would have given unpredicatble results.
1302    
1303    9.  A possessively repeated subpattern with minimum repeat count greater than
1304        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
1305        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
1306        the first (A) could occur when it should not.
1307    
1308    10. Add a cast and remove a redundant test from the code.
1309    
1310    11. JIT should use pcre_malloc/pcre_free for allocation.
1311    
1312    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
1313        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
1314        is anything other than /usr, -L is still shown).
1315    
1316    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
1317    
1318    14. Perl does not support \N without a following name in a [] class; PCRE now
1319        also gives an error.
1320    
1321    15. If a forward reference was repeated with an upper limit of around 2000,
1322        it caused the error "internal error: overran compiling workspace". The
1323        maximum number of forward references (including repeats) was limited by the
1324        internal workspace, and dependent on the LINK_SIZE. The code has been
1325        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
1326        the default depends on LINK_SIZE. There is a new upper limit (for safety)
1327        of around 200,000 forward references. While doing this, I also speeded up
1328        the filling in of repeated forward references.
1329    
1330    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
1331        incorrectly expecting the subject to contain another "a" after the start.
1332    
1333    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
1334        in the match, the SKIP should be ignored. This was not happening; instead
1335        the SKIP was being treated as NOMATCH. For patterns such as
1336        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
1337        tested.
1338    
1339    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
1340        now much more compatible with Perl, in particular in cases where the result
1341        is a non-match for a non-anchored pattern. For example, if
1342        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
1343        "m", where previously it did not return a name. A side effect of this
1344        change is that for partial matches, the last encountered mark name is
1345        returned, as for non matches. A number of tests that were previously not
1346        Perl-compatible have been moved into the Perl-compatible test files. The
1347        refactoring has had the pleasing side effect of removing one argument from
1348        the match() function, thus reducing its stack requirements.
1349    
1350    19. If the /S+ option was used in pcretest to study a pattern using JIT,
1351        subsequent uses of /S (without +) incorrectly behaved like /S+.
1352    
1353    21. Retrieve executable code size support for the JIT compiler and fixing
1354        some warnings.
1355    
1356    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
1357        not work when the shorter character appeared right at the end of the
1358        subject string.
1359    
1360    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
1361        systems.
1362    
1363    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
1364        output it when the /M option is used in pcretest.
1365    
1366    25. The CheckMan script was not being included in the distribution. Also, added
1367        an explicit "perl" to run Perl scripts from the PrepareRelease script
1368        because this is reportedly needed in Windows.
1369    
1370    26. If study data was being save in a file and studying had not found a set of
1371        "starts with" bytes for the pattern, the data written to the file (though
1372        never used) was taken from uninitialized memory and so caused valgrind to
1373        complain.
1374    
1375    27. Updated RunTest.bat as provided by Sheri Pierce.
1376    
1377    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
1378    
1379    29. Computation of memory usage for the table of capturing group names was
1380        giving an unnecessarily large value.
1381    
1382    
1383    Version 8.20 21-Oct-2011
1384    ------------------------
1385    
1386    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
1387        a POSIX class. After further experiments with Perl, which convinced me that
1388        Perl has bugs and confusions, a closing square bracket is no longer allowed
1389        in a POSIX name. This bug also affected patterns with classes that started
1390        with full stops.
1391    
1392    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
1393        captured substring, but while checking the failing first alternative,
1394        substring 1 is temporarily captured. If the output vector supplied to
1395        pcre_exec() was not big enough for this capture, the yield of the function
1396        was still zero ("insufficient space for captured substrings"). This cannot
1397        be totally fixed without adding another stack variable, which seems a lot
1398        of expense for a edge case. However, I have improved the situation in cases
1399        such as /(a)(b)x|abc/ matched against "abc", where the return code
1400        indicates that fewer than the maximum number of slots in the ovector have
1401        been set.
1402    
1403    3.  Related to (2) above: when there are more back references in a pattern than
1404        slots in the output vector, pcre_exec() uses temporary memory during
1405        matching, and copies in the captures as far as possible afterwards. It was
1406        using the entire output vector, but this conflicts with the specification
1407        that only 2/3 is used for passing back captured substrings. Now it uses
1408        only the first 2/3, for compatibility. This is, of course, another edge
1409        case.
1410    
1411    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1412        main code base, and can be used by building with --enable-jit. When this is
1413        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1414        runtime --no-jit option is given.
1415    
1416    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1417        ovector, the return from the function was zero, implying that there were
1418        other matches that did not fit. The correct "exactly full" value is now
1419        returned.
1420    
1421    6.  If a subpattern that was called recursively or as a subroutine contained
1422        (*PRUNE) or any other control that caused it to give a non-standard return,
1423        invalid errors such as "Error -26 (nested recursion at the same subject
1424        position)" or even infinite loops could occur.
1425    
1426    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1427        computing the minimum length on reaching *ACCEPT, and so ended up with the
1428        wrong value of 1 rather than 0. Further investigation indicates that
1429        computing a minimum subject length in the presence of *ACCEPT is difficult
1430        (think back references, subroutine calls), and so I have changed the code
1431        so that no minimum is registered for a pattern that contains *ACCEPT.
1432    
1433    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1434        it was not handled as intended. [But see 16 below.]
1435    
1436    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1437        Sheri Pierce.
1438    
1439    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1440        the first byte in a match must be "a".
1441    
1442    11. Change 17 for 8.13 increased the recursion depth for patterns like
1443        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1444        pattern contains any instances of (*THEN). If it does not, the old
1445        optimizations are restored. It would be nice to do this on a per-group
1446        basis, but at the moment that is not feasible.
1447    
1448    12. In some environments, the output of pcretest -C is CRLF terminated. This
1449        broke RunTest's code that checks for the link size. A single white space
1450        character after the value is now allowed for.
1451    
1452    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1453        For "fr", it uses the Windows-specific input and output files.
1454    
1455    14. If (*THEN) appeared in a group that was called recursively or as a
1456        subroutine, it did not work as intended. [But see next item.]
1457    
1458    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1459        pattern fragments (but not containing any | characters). If A and B are
1460        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1461        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1462        D. In other words, Perl considers parentheses that do not contain any |
1463        characters to be part of a surrounding alternative, whereas PCRE was
1464        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1465        differently. PCRE now behaves in the same way as Perl, except in the case
1466        of subroutine/recursion calls such as (?1) which have in any case always
1467        been different (but PCRE had them first :-).
1468    
1469    16. Related to 15 above: Perl does not treat the | in a conditional group as
1470        creating alternatives. Such a group is treated in the same way as an
1471        ordinary group without any | characters when processing (*THEN). PCRE has
1472        been changed to match Perl's behaviour.
1473    
1474    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1475        RunGrepTest script failed.
1476    
1477    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1478        inevitable for groups that contain captures, but it can lead to a lot of
1479        stack use in large patterns. The old behaviour has been restored for atomic
1480        groups that do not contain any capturing parentheses.
1481    
1482    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1483        suppress the check for a minimum subject length at run time. (If it was
1484        given to pcre_exec() or pcre_dfa_exec() it did work.)
1485    
1486    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1487        fail to work when decoding hex characters in data strings in EBCDIC
1488        environments.
1489    
1490    21. It appears that in at least one Mac OS environment, the isxdigit() function
1491        is implemented as a macro that evaluates to its argument more than once,
1492        contravening the C 90 Standard (I haven't checked a later standard). There
1493        was an instance in pcretest which caused it to go wrong when processing
1494        \x{...} escapes in subject strings. The has been rewritten to avoid using
1495        things like p++ in the argument of isxdigit().
1496    
1497    
1498    Version 8.13 16-Aug-2011
1499    ------------------------
1500    
1501    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1502    
1503    2.  Two minor typos in pcre_internal.h have been fixed.
1504    
1505    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1506        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1507        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1508    
1509    4.  There were a number of related bugs in the code for matching backrefences
1510        caselessly in UTF-8 mode when codes for the characters concerned were
1511        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1512        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1513        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1514        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1515        2-byte code at the end of the subject (it thought there wasn't enough data
1516        left).
1517    
1518    5.  Comprehensive information about what went wrong is now returned by
1519        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1520        as the output vector has at least 2 elements. The offset of the start of
1521        the failing character and a reason code are placed in the vector.
1522    
1523    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1524        now returned is for the first byte of the failing character, instead of the
1525        last byte inspected. This is an incompatible change, but I hope it is small
1526        enough not to be a problem. It makes the returned offset consistent with
1527        pcre_exec() and pcre_dfa_exec().
1528    
1529    7.  pcretest now gives a text phrase as well as the error number when
1530        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1531        failure, the offset and reason code are output.
1532    
1533    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1534        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1535        back over a single character (\n). This seems wrong (because it treated the
1536        two characters as a single entity when going forwards), conflicts with the
1537        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1538        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1539        has been changed.
1540    
1541    9.  Some internal refactoring has changed the processing so that the handling
1542        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1543        time (the PCRE_DOTALL option was changed this way some time ago: version
1544        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1545        which was always a bit of a fudge. It also means that there is one less
1546        argument for the match() function, which reduces its stack requirements
1547        slightly. This change also fixes an incompatibility with Perl: the pattern
1548        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1549    
1550    10. More internal refactoring has drastically reduced the number of recursive
1551        calls to match() for possessively repeated groups such as (abc)++ when
1552        using pcre_exec().
1553    
1554    11. While implementing 10, a number of bugs in the handling of groups were
1555        discovered and fixed:
1556    
1557        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1558        (a|)*(?1) gave a compile-time internal error.
1559        ((a|)+)+  did not notice that the outer group could match an empty string.
1560        (^a|^)+   was not marked as anchored.
1561        (.*a|.*)+ was not marked as matching at start or after a newline.
1562    
1563    12. Yet more internal refactoring has removed another argument from the match()
1564        function. Special calls to this function are now indicated by setting a
1565        value in a variable in the "match data" data block.
1566    
1567    13. Be more explicit in pcre_study() instead of relying on "default" for
1568        opcodes that mean there is no starting character; this means that when new
1569        ones are added and accidentally left out of pcre_study(), testing should
1570        pick them up.
1571    
1572    14. The -s option of pcretest has been documented for ages as being an old
1573        synonym of -m (show memory usage). I have changed it to mean "force study
1574        for every regex", that is, assume /S for every regex. This is similar to -i
1575        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1576        using it. It makes it easier to run collections of tests with and without
1577        study enabled, and thereby test pcre_study() more easily. All the standard
1578        tests are now run with and without -s (but some patterns can be marked as
1579        "never study" - see 20 below).
1580    
1581    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1582        restoration of the capturing data to the outer values was not happening
1583        correctly.
1584    
1585    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1586        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1587        pattern had matched an empty string, and so incorrectly returned a no
1588        match.
1589    
1590    17. There was optimizing code for the last branch of non-capturing parentheses,
1591        and also for the obeyed branch of a conditional subexpression, which used
1592        tail recursion to cut down on stack usage. Unfortunately, now that there is
1593        the possibility of (*THEN) occurring in these branches, tail recursion is
1594        no longer possible because the return has to be checked for (*THEN). These
1595        two optimizations have therefore been removed. [But see 8.20/11 above.]
1596    
1597    18. If a pattern containing \R was studied, it was assumed that \R always
1598        matched two bytes, thus causing the minimum subject length to be
1599        incorrectly computed because \R can also match just one byte.
1600    
1601    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1602        was incorrectly computed.
1603    
1604    20. If /S is present twice on a test pattern in pcretest input, it now
1605        *disables* studying, thereby overriding the use of -s on the command line
1606        (see 14 above). This is necessary for one or two tests to keep the output
1607        identical in both cases.
1608    
1609    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1610        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1611    
1612    22. When an atomic group that contained a capturing parenthesis was
1613        successfully matched, but the branch in which it appeared failed, the
1614        capturing was not being forgotten if a higher numbered group was later
1615        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1616        group 1 to "a", when in fact it should be unset. This applied to multi-
1617        branched capturing and non-capturing groups, repeated or not, and also to
1618        positive assertions (capturing in negative assertions does not happen
1619        in PCRE) and also to nested atomic groups.
1620    
1621    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1622        subject after a captured substring, to make it easier to tell which of a
1623        number of identical substrings has been captured.
1624    
1625    24. The way atomic groups are processed by pcre_exec() has been changed so that
1626        if they are repeated, backtracking one repetition now resets captured
1627        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1628        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1629        "aaa". Previously, it would have been "a". As part of this code
1630        refactoring, the way recursive calls are handled has also been changed.
1631    
1632    25. If an assertion condition captured any substrings, they were not passed
1633        back unless some other capturing happened later. For example, if
1634        (?(?=(a))a) was matched against "a", no capturing was returned.
1635    
1636    26. When studying a pattern that contained subroutine calls or assertions,
1637        the code for finding the minimum length of a possible match was handling
1638        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1639        group 1 called group 2 while simultaneously a separate group 2 called group
1640        1). A stack overflow occurred in this case. I have fixed this by limiting
1641        the recursion depth to 10.
1642    
1643    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1644        Fortmann. This supports explicit test numbers on the command line, and has
1645        argument validation and error reporting.
1646    
1647    28. An instance of \X with an unlimited repeat could fail if at any point the
1648        first character it looked at was a mark character.
1649    
1650    29. Some minor code refactoring concerning Unicode properties and scripts
1651        should reduce the stack requirement of match() slightly.
1652    
1653    30. Added the '=' option to pcretest to check the setting of unused capturing
1654        slots at the end of the pattern, which are documented as being -1, but are
1655        not included in the return count.
1656    
1657    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1658        compiled something random. Now it gives a compile-time error (as does
1659        Perl).
1660    
1661    32. A *MARK encountered during the processing of a positive assertion is now
1662        recorded and passed back (compatible with Perl).
1663    
1664    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1665        had alternative anchored branches, the search for a second match in a line
1666        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1667        matched the line "0102" twice. The same bug affected patterns that started
1668        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1669        twice.
1670    
1671    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1672        does, and because of capturing effects, quantifying parenthesized
1673        assertions may at times be useful. Quantifiers are now allowed for
1674        parenthesized assertions.
1675    
1676    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1677    
1678    36. \g was being checked for fancy things in a character class, when it should
1679        just be a literal "g".
1680    
1681    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1682        appearance of a nested POSIX class supersedes an apparent external class.
1683        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1684        unescaped square brackets may also appear as part of class names. For
1685        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1686        more like Perl. (But see 8.20/1 above.)
1687    
1688    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1689        was because it thought it was \N{name}, which is not supported).
1690    
1691    39. Add minix to OS list not supporting the -S option in pcretest.
1692    
1693    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1694        cannot analyze patterns in sufficient detail to catch mutual recursions
1695        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1696        subgroup is called recursively as a subpattern for a second time at the
1697        same position in the subject string. In previous releases this might have
1698        been caught by the recursion limit, or it might have run out of stack.
1699    
1700    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1701        happen only once. PCRE was, however incorrectly giving a compile time error
1702        "recursive call could loop indefinitely" because it cannot analyze the
1703        pattern in sufficient detail. The compile time test no longer happens when
1704        PCRE is compiling a conditional subpattern, but actual runaway loops are
1705        now caught at runtime (see 40 above).
1706    
1707    42. It seems that Perl allows any characters other than a closing parenthesis
1708        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1709        has been changed to be the same.
1710    
1711    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1712        as not to get warnings when autogen.sh is called. Also changed
1713        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1714    
1715    44. To help people who use pcregrep to scan files containing exceedingly long
1716        lines, the following changes have been made:
1717    
1718        (a) The default value of the buffer size parameter has been increased from
1719            8K to 20K. (The actual buffer used is three times this size.)
1720    
1721        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1722            PCRE is built.
1723    
1724        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1725            to be set at run time.
1726    
1727        (d) Numerical values in pcregrep options can be followed by K or M, for
1728            example --buffer-size=50K.
1729    
1730        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1731            given and the return code is set to 2.
1732    
1733    45. Add a pointer to the latest mark to the callout data block.
1734    
1735    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1736        partial match of an empty string instead of no match. This was specific to
1737        the use of ".".
1738    
1739    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1740        complete match instead of a partial match. This bug was dependent on both
1741        the PCRE_UTF8 and PCRE_DOTALL options being set.
1742    
1743    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1744        starting byte set, because \b was not being ignored.
1745    
1746    
1747    Version 8.12 15-Jan-2011
1748    ------------------------
1749    
1750    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1751        checks for such things as part of the documentation building process.
1752    
1753    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1754        --match-limit and --recursion-limit options (added for 8.11). In
1755        particular, this made one of the standard tests fail. (The integer value
1756        went into the wrong half of a long int.)
1757    
1758    3.  If the --colour option was given to pcregrep with -v (invert match), it
1759        did strange things, either producing crazy output, or crashing. It should,
1760        of course, ignore a request for colour when reporting lines that do not
1761        match.
1762    
1763    4.  Another pcregrep bug caused similar problems if --colour was specified with
1764        -M (multiline) and the pattern match finished with a line ending.
1765    
1766    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1767        matched in multiline mode, the following line was shown as part of the
1768        match. This seems wrong, so I have changed it.
1769    
1770    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1771        the check for further matches in the same line (so they could be coloured)
1772        to overrun the end of the current line. If another match was found, it was
1773        incorrectly shown (and then shown again when found in the next line).
1774    
1775    7.  If pcregrep was compiled under Windows, there was a reference to the
1776        function pcregrep_exit() before it was defined. I am assuming this was
1777        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1778        reported by a user. I've moved the definition above the reference.
1779    
1780    
1781    Version 8.11 10-Dec-2010
1782    ------------------------
1783    
1784    1.  (*THEN) was not working properly if there were untried alternatives prior
1785        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1786        backtracked to try for "b" instead of moving to the next alternative branch
1787        at the same level (in this case, to look for "c"). The Perl documentation
1788        is clear that when (*THEN) is backtracked onto, it goes to the "next
1789        alternative in the innermost enclosing group".
1790    
1791    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1792        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1793        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1794        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1795        (*THEN).
1796    
1797    3.  If \s appeared in a character class, it removed the VT character from
1798        the class, even if it had been included by some previous item, for example
1799        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1800        of \s, but is part of the POSIX "space" class.)
1801    
1802    4.  A partial match never returns an empty string (because you can always
1803        match an empty string at the end of the subject); however the checking for
1804        an empty string was starting at the "start of match" point. This has been
1805        changed to the "earliest inspected character" point, because the returned
1806        data for a partial match starts at this character. This means that, for
1807        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1808        (previously it gave "no match").
1809    
1810    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1811        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1812        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1813        has an implication that the given string is incomplete (because a partial
1814        match is preferred over a full match). For this reason, these items now
1815        give a partial match in this situation. [Aside: previously, the one case
1816        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1817        match rather than a full match, which was wrong by the old rules, but is
1818        now correct.]
1819    
1820    6.  There was a bug in the handling of #-introduced comments, recognized when
1821        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1822        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1823        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1824        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1825        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1826        places in pcre_compile().
1827    
1828    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1829        comments when looking ahead for named forward references to subpatterns,
1830        the only newline sequence it recognized was NL. It now handles newlines
1831        according to the set newline convention.
1832    
1833    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1834        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1835        cater for a lack of strerror(). These oversights have been fixed.
1836    
1837    9.  Added --match-limit and --recursion-limit to pcregrep.
1838    
1839    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1840    
1841    11. When the -o option was used, pcregrep was setting a return code of 1, even
1842        when matches were found, and --line-buffered was not being honoured.
1843    
1844    12. Added an optional parentheses number to the -o and --only-matching options
1845        of pcregrep.
1846    
1847    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1848        can match an empty string. The code to do it in pcretest and pcredemo
1849        needed fixing:
1850    
1851        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1852            only one byte after an empty string match just before CRLF (this case
1853            just got forgotten; "any" and "anycrlf" were OK).
1854    
1855        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1856            mode when an empty string match preceded an ASCII character followed by
1857            a non-ASCII character. (The code for advancing by one character rather
1858            than one byte was nonsense.)
1859    
1860        (c) The pcredemo.c sample program did not have any code at all to handle
1861            the cases when CRLF is a valid newline sequence.
1862    
1863    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1864        as a starting offset was within the subject string. There is now a new
1865        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1866        negative or greater than the length of the string. In order to test this,
1867        pcretest is extended to allow the setting of negative starting offsets.
1868    
1869    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1870        starting offset points to the beginning of a UTF-8 character was
1871        unnecessarily clumsy. I tidied it up.
1872    
1873    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1874        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1875    
1876    17. Nobody had reported that the --include_dir option, which was added in
1877        release 7.7 should have been called --include-dir (hyphen, not underscore)
1878        for compatibility with GNU grep. I have changed it to --include-dir, but
1879        left --include_dir as an undocumented synonym, and the same for
1880        --exclude-dir, though that is not available in GNU grep, at least as of
1881        release 2.5.4.
1882    
1883    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1884        characters from a string of bytes) have been redefined so as not to use
1885        loops, in order to improve performance in some environments. At the same
1886        time, I abstracted some of the common code into auxiliary macros to save
1887        repetition (this should not affect the compiled code).
1888    
1889    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1890        compile-time error is now given if \c is not followed by an ASCII
1891        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1892        different, and any byte value is allowed.)
1893    
1894    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1895        START_OPTIMIZE option, which is now allowed at compile time - but just
1896        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1897        to pcregrep and other applications that have no direct access to PCRE
1898        options. The new /Y option in pcretest sets this option when calling
1899        pcre_compile().
1900    
1901    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1902        back references. Groups containing recursive back references were forced to
1903        be atomic by that change, but in the case of named groups, the amount of
1904        memory required was incorrectly computed, leading to "Failed: internal
1905        error: code overflow". This has been fixed.
1906    
1907    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1908        pcretest.c, to avoid build problems in some Borland environments.
1909    
1910    
1911    Version 8.10 25-Jun-2010
1912    ------------------------
1913    
1914    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1915        THEN.
1916    
1917    2.  (*ACCEPT) was not working when inside an atomic group.
1918    
1919    3.  Inside a character class, \B is treated as a literal by default, but
1920        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1921        causes the error). The code is unchanged, but I tidied the documentation.
1922    
1923    4.  Inside a character class, PCRE always treated \R and \X as literals,
1924        whereas Perl faults them if its -w option is set. I have changed PCRE so
1925        that it faults them when PCRE_EXTRA is set.
1926    
1927    5.  Added support for \N, which always matches any character other than
1928        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1929    
1930    6.  When compiling pcregrep with newer versions of gcc which may have
1931        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1932        declared with attribute warn_unused_result" were given. Just casting the
1933        result to (void) does not stop the warnings; a more elaborate fudge is
1934        needed. I've used a macro to implement this.
1935    
1936    7.  Minor change to pcretest.c to avoid a compiler warning.
1937    
1938    8.  Added four artifical Unicode properties to help with an option to make
1939        \s etc use properties (see next item). The new properties are: Xan
1940        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1941    
1942    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1943        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1944        this option. Modified pcretest to add /W to test this facility. Added
1945        REG_UCP to make it available via the POSIX interface.
1946    
1947    10. Added --line-buffered to pcregrep.
1948    
1949    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1950        studied, and the match started with a letter with a code point greater than
1951        127 whose first byte was different to the first byte of the other case of
1952        the letter, the other case of this starting letter was not recognized
1953        (#976).
1954    
1955    12. If a pattern that was studied started with a repeated Unicode property
1956        test, for example, \p{Nd}+, there was the theoretical possibility of
1957        setting up an incorrect bitmap of starting bytes, but fortunately it could
1958        not have actually happened in practice until change 8 above was made (it
1959        added property types that matched character-matching opcodes).
1960    
1961    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1962        possible starting bytes for non-anchored patterns.
1963    
1964    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1965        \R, and also a number of cases that involve Unicode properties, both
1966        explicit and implicit when PCRE_UCP is set.
1967    
1968    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1969        input, it could crash or give wrong results if characters with values
1970        greater than 0xc0 were present in the subject string. (Detail: it assumed
1971        UTF-8 input when processing these items.)
1972    
1973    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1974        size_t is 64-bit (#991).
1975    
1976    17. Added a check for running out of memory when PCRE is compiled with
1977        --disable-stack-for-recursion (#990).
1978    
1979    18. If the last data line in a file for pcretest does not have a newline on
1980        the end, a newline was missing in the output.
1981    
1982    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1983        less than 128) in its various bitmaps. However, there is a facility for
1984        generating tables according to the current locale when PCRE is compiled. It
1985        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1986        characters, are recognized by isspace() and therefore were getting set in
1987        these tables, and indeed these tables seem to approximate to ISO 8859. This
1988        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1989        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1990        which of course cannot start UTF-8 characters. I have changed the code so
1991        that only real ASCII characters (less than 128) and the correct starting
1992        bytes for UTF-8 encodings are set for characters greater than 127 when in
1993        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1994        altogether.)
1995    
1996    20. Added the /T option to pcretest so as to be able to run tests with non-
1997        standard character tables, thus making it possible to include the tests
1998        used for 19 above in the standard set of tests.
1999    
2000    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
2001        reference to a subpattern the other side of a comment that contains an
2002        opening parenthesis caused either an internal compiling error, or a
2003        reference to the wrong subpattern.
2004    
2005    
2006    Version 8.02 19-Mar-2010
2007    ------------------------
2008    
2009    1.  The Unicode data tables have been updated to Unicode 5.2.0.
2010    
2011    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
2012        configured.
2013    
2014    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
2015        original author of that file, following a query about its status.
2016    
2017    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
2018        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
2019    
2020    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
2021        quantifier applied to a forward-referencing subroutine call, could compile
2022        incorrect code or give the error "internal error: previously-checked
2023        referenced subpattern not found".
2024    
2025    6.  Both MS Visual Studio and Symbian OS have problems with initializing
2026        variables to point to external functions. For these systems, therefore,
2027        pcre_malloc etc. are now initialized to local functions that call the
2028        relevant global functions.
2029    
2030    7.  There were two entries missing in the vectors called coptable and poptable
2031        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
2032        I've fixed the data, and added a kludgy way of testing at compile time that
2033        the lengths are correct (equal to the number of opcodes).
2034    
2035    8.  Following on from 7, I added a similar kludge to check the length of the
2036        eint vector in pcreposix.c.
2037    
2038    9.  Error texts for pcre_compile() are held as one long string to avoid too
2039        much relocation at load time. To find a text, the string is searched,
2040        counting zeros. There was no check for running off the end of the string,
2041        which could happen if a new error number was added without updating the
2042        string.
2043    
2044    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
2045    
2046    11. \K was not working if it appeared in an atomic group or in a group that
2047        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
2048        \K is "not well defined" if used in an assertion. PCRE now accepts it if
2049        the assertion is positive, but not if it is negative.
2050    
2051    12. Change 11 fortuitously reduced the size of the stack frame used in the
2052        "match()" function of pcre_exec.c by one pointer. Forthcoming
2053        implementation of support for (*MARK) will need an extra pointer on the
2054        stack; I have reserved it now, so that the stack frame size does not
2055        decrease.
2056    
2057    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
2058        item in branch that calls a recursion is a subroutine call - as in the
2059        second branch in the above example - was incorrectly given the compile-
2060        time error "recursive call could loop indefinitely" because pcre_compile()
2061        was not correctly checking the subroutine for matching a non-empty string.
2062    
2063    14. The checks for overrunning compiling workspace could trigger after an
2064        overrun had occurred. This is a "should never occur" error, but it can be
2065        triggered by pathological patterns such as hundreds of nested parentheses.
2066        The checks now trigger 100 bytes before the end of the workspace.
2067    
2068    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
2069    
2070    
2071    Version 8.01 19-Jan-2010
2072    ------------------------
2073    
2074    1.  If a pattern contained a conditional subpattern with only one branch (in
2075        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
2076        computed the wrong minimum data length (which is of course zero for such
2077        subpatterns). This could cause incorrect "no match" results.
2078    
2079    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
2080        the pattern is reset in the first branch, pcre_compile() failed with
2081        "internal error: code overflow at offset...". This happened only when
2082        the reset was to the original external option setting. (An optimization
2083        abstracts leading options settings into an external setting, which was the
2084        cause of this.)
2085    
2086    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
2087        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
2088        assertion pattern did not match (meaning that the assertion was true), it
2089        was incorrectly treated as false if the SKIP had been reached during the
2090        matching. This also applied to assertions used as conditions.
2091    
2092    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
2093        assertion subpattern, including such a pattern used as a condition,
2094        unpredictable results occurred, instead of the error return
2095        PCRE_ERROR_DFA_UITEM.
2096    
2097    5.  The C++ GlobalReplace function was not working like Perl for the special
2098        situation when an empty string is matched. It now does the fancy magic
2099        stuff that is necessary.
2100    
2101    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
2102        removed. (These were left over from very, very early versions of PCRE.)
2103    
2104    7.  Some cosmetic changes to the code to make life easier when compiling it
2105        as part of something else:
2106    
2107        (a) Change DEBUG to PCRE_DEBUG.
2108    
2109        (b) In pcre_compile(), rename the member of the "branch_chain" structure
2110            called "current" as "current_branch", to prevent a collision with the
2111            Linux macro when compiled as a kernel module.
2112    
2113        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
2114            prevent a collision with the Linux macro when compiled as a kernel
2115            module.
2116    
2117    8.  In pcre_compile() there are some checks for integer overflows that used to
2118        cast potentially large values to (double). This has been changed to that
2119        when building, a check for int64_t is made, and if it is found, it is used
2120        instead, thus avoiding the use of floating point arithmetic. (There is no
2121        other use of FP in PCRE.) If int64_t is not found, the fallback is to
2122        double.
2123    
2124    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
2125        2005 (difference between two addresses compared to an unsigned value).
2126    
2127    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
2128        custom one, because of the following reported problem in Windows:
2129    
2130          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
2131              under Win32.
2132          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
2133              therefore missing the function definition.
2134          - The compiler thus generates a "C" signature for the test function.
2135          - The linker fails to find the "C" function.
2136          - PCRE fails to configure if asked to do so against libbz2.
2137    
2138    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
2139        messages were output:
2140    
2141          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
2142          rerunning libtoolize, to keep the correct libtool macros in-tree.
2143          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
2144    
2145        I have done both of these things.
2146    
2147    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
2148        most of the time, it *can* run out if it is given a pattern that contains a
2149        runaway infinite recursion. I updated the discussion in the pcrestack man
2150        page.
2151    
2152    13. Now that we have gone to the x.xx style of version numbers, the minor
2153        version may start with zero. Using 08 or 09 is a bad idea because users
2154        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
2155        interpreted as invalid octal numbers. I've updated the previous comment in
2156        configure.ac, and also added a check that gives an error if 08 or 09 are
2157        used.
2158    
2159    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
2160        causing partial matching to fail when the end of the subject matched \W
2161        in a UTF-8 pattern where \W was quantified with a minimum of 3.
2162    
2163    15. There were some discrepancies between the declarations in pcre_internal.h
2164        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
2165        their definitions. The declarations used "const uschar *" and the
2166        definitions used USPTR. Even though USPTR is normally defined as "const
2167        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
2168        reported that: "This difference in casting confuses some C++ compilers, for
2169        example, SunCC recognizes above declarations as different functions and
2170        generates broken code for hbpcre." I have changed the declarations to use
2171        USPTR.
2172    
2173    16. GNU libtool is named differently on some systems. The autogen.sh script now
2174        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
2175        (FreeBSD).
2176    
2177    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
2178        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
2179        comment: "Figure out how to create a longlong from a string: strtoll and
2180        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
2181        instance, but it only takes 2 args instead of 3!"
2182    
2183    18. A subtle bug concerned with back references has been fixed by a change of
2184        specification, with a corresponding code fix. A pattern such as
2185        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
2186        refers, was giving matches when it shouldn't. For example, xa=xaaa would
2187        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
2188        same bug. Such groups have to be quantified to be useful, or contained
2189        inside another quantified group. (If there's no repetition, the reference
2190        can never match.) The problem arises because, having left the group and
2191        moved on to the rest of the pattern, a later failure that backtracks into
2192        the group uses the captured value from the final iteration of the group
2193        rather than the correct earlier one. I have fixed this in PCRE by forcing
2194        any group that contains a reference to itself to be an atomic group; that
2195        is, there cannot be any backtracking into it once it has completed. This is
2196        similar to recursive and subroutine calls.
2197    
2198    
2199    Version 8.00 19-Oct-09
2200    ----------------------
2201    
2202    1.  The table for translating pcre_compile() error codes into POSIX error codes
2203        was out-of-date, and there was no check on the pcre_compile() error code
2204        being within the table. This could lead to an OK return being given in
2205        error.
2206    
2207    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
2208        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
2209        in a Windows environment.
2210    
2211    3.  The pcregrep --count option prints the count for each file even when it is
2212        zero, as does GNU grep. However, pcregrep was also printing all files when
2213        --files-with-matches was added. Now, when both options are given, it prints
2214        counts only for those files that have at least one match. (GNU grep just
2215        prints the file name in this circumstance, but including the count seems
2216        more useful - otherwise, why use --count?) Also ensured that the
2217        combination -clh just lists non-zero counts, with no names.
2218    
2219    4.  The long form of the pcregrep -F option was incorrectly implemented as
2220        --fixed_strings instead of --fixed-strings. This is an incompatible change,
2221        but it seems right to fix it, and I didn't think it was worth preserving
2222        the old behaviour.
2223    
2224    5.  The command line items --regex=pattern and --regexp=pattern were not
2225        recognized by pcregrep, which required --regex pattern or --regexp pattern
2226        (with a space rather than an '='). The man page documented the '=' forms,
2227        which are compatible with GNU grep; these now work.
2228    
2229    6.  No libpcreposix.pc file was created for pkg-config; there was just
2230        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
2231    
2232    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
2233        when UCP support is not needed, by modifying the Python script that
2234        generates it from Unicode data files. This should not matter if the module
2235        is correctly used as a library, but I received one complaint about 50K of
2236        unwanted data. My guess is that the person linked everything into his
2237        program rather than using a library. Anyway, it does no harm.
2238    
2239    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
2240        was a minimum greater than 1 for a wide character in a possessive
2241        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
2242        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
2243        character. Chaos in the form of incorrect output or a compiling loop could
2244        result.
2245    
2246    9.  The restrictions on what a pattern can contain when partial matching is
2247        requested for pcre_exec() have been removed. All patterns can now be
2248        partially matched by this function. In addition, if there are at least two
2249        slots in the offset vector, the offset of the earliest inspected character
2250        for the match and the offset of the end of the subject are set in them when
2251        PCRE_ERROR_PARTIAL is returned.
2252    
2253    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
2254        synonymous with PCRE_PARTIAL, for backwards compatibility, and
2255        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
2256        and may be more useful for multi-segment matching.
2257    
2258    11. Partial matching with pcre_exec() is now more intuitive. A partial match
2259        used to be given if ever the end of the subject was reached; now it is
2260        given only if matching could not proceed because another character was
2261        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
2262        string "Z", which now yields "no match" instead of "partial match". In the
2263        case of pcre_dfa_exec(), "no match" is given if every matching path for the
2264        final character ended with (*FAIL).
2265    
2266    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
2267        if the pattern had a "must contain" character that was already found in the
2268        earlier partial match, unless partial matching was again requested. For
2269        example, with the pattern /dog.(body)?/, the "must contain" character is
2270        "g". If the first part-match was for the string "dog", restarting with
2271        "sbody" failed. This bug has been fixed.
2272    
2273    13. The string returned by pcre_dfa_exec() after a partial match has been
2274        changed so that it starts at the first inspected character rather than the
2275        first character of the match. This makes a difference only if the pattern
2276        starts with a lookbehind assertion or \b or \B (\K is not supported by
2277        pcre_dfa_exec()). It's an incompatible change, but it makes the two
2278        matching functions compatible, and I think it's the right thing to do.
2279    
2280    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
2281        so that the demonstration program is easily available in environments where
2282        PCRE has not been installed from source.
2283    
2284    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
2285        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
2286        library.
2287    
2288    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
2289        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
2290        is not the first non-POSIX option to be added. Clearly some people find
2291        these options useful.
2292    
2293    17. If a caller to the POSIX matching function regexec() passes a non-zero
2294        value for nmatch with a NULL value for pmatch, the value of
2295        nmatch is forced to zero.
2296    
2297    18. RunGrepTest did not have a test for the availability of the -u option of
2298        the diff command, as RunTest does. It now checks in the same way as
2299        RunTest, and also checks for the -b option.
2300    
2301    19. If an odd number of negated classes containing just a single character
2302        interposed, within parentheses, between a forward reference to a named
2303        subpattern and the definition of the subpattern, compilation crashed with
2304        an internal error, complaining that it could not find the referenced
2305        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
2306        [The bug was that it was starting one character too far in when skipping
2307        over the character class, thus treating the ] as data rather than
2308        terminating the class. This meant it could skip too much.]
2309    
2310    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
2311        /g option in pcretest when the pattern contains \K, which makes it possible
2312        to have an empty string match not at the start, even when the pattern is
2313        anchored. Updated pcretest and pcredemo to use this option.
2314    
2315    21. If the maximum number of capturing subpatterns in a recursion was greater
2316        than the maximum at the outer level, the higher number was returned, but
2317        with unset values at the outer level. The correct (outer level) value is
2318        now given.
2319    
2320    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
2321        PCRE did not set those parentheses (unlike Perl). I have now found a way to
2322        make it do so. The string so far is captured, making this feature
2323        compatible with Perl.
2324    
2325    23. The tests have been re-organized, adding tests 11 and 12, to make it
2326        possible to check the Perl 5.10 features against Perl 5.10.
2327    
2328    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
2329        pattern matches a fixed length string. PCRE did not allow this; now it
2330        does. Neither allows recursion.
2331    
2332    25. I finally figured out how to implement a request to provide the minimum
2333        length of subject string that was needed in order to match a given pattern.
2334        (It was back references and recursion that I had previously got hung up
2335        on.) This code has now been added to pcre_study(); it finds a lower bound
2336        to the length of subject needed. It is not necessarily the greatest lower
2337        bound, but using it to avoid searching strings that are too short does give
2338        some useful speed-ups. The value is available to calling programs via
2339        pcre_fullinfo().
2340    
2341    26. While implementing 25, I discovered to my embarrassment that pcretest had
2342        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
2343        study optimizations had never been tested with that matching function.
2344        Oops. What is worse, even when it was passed study data, there was a bug in
2345        pcre_dfa_exec() that meant it never actually used it. Double oops. There
2346        were also very few tests of studied patterns with pcre_dfa_exec().
2347    
2348    27. If (?| is used to create subpatterns with duplicate numbers, they are now
2349        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
2350        on the other side of the coin, they are no longer allowed to have different
2351        names, because these cannot be distinguished in PCRE, and this has caused
2352        confusion. (This is a difference from Perl.)
2353    
2354    28. When duplicate subpattern names are present (necessarily with different
2355        numbers, as required by 27 above), and a test is made by name in a
2356        conditional pattern, either for a subpattern having been matched, or for
2357        recursion in such a pattern, all the associated numbered subpatterns are
2358        tested, and the overall condition is true if the condition is true for any
2359        one of them. This is the way Perl works, and is also more like the way
2360        testing by number works.
2361    
2362    
2363    Version 7.9 11-Apr-09
2364    ---------------------
2365    
2366    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
2367        (pcretest), all targets were linked against these libraries. This included
2368        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
2369        libraries. This caused unwanted dependencies to be created. This problem
2370        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
2371        pcretest is linked with readline.
2372    
2373    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
2374        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
2375        moved outside it again, because FALSE and TRUE are already defined in AIX,
2376        but BOOL is not.
2377    
2378    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
2379        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
2380    
2381    4.  The pcregrep documentation said spaces were inserted as well as colons (or
2382        hyphens) following file names and line numbers when outputting matching
2383        lines. This is not true; no spaces are inserted. I have also clarified the
2384        wording for the --colour (or --color) option.
2385    
2386    5.  In pcregrep, when --colour was used with -o, the list of matching strings
2387        was not coloured; this is different to GNU grep, so I have changed it to be
2388        the same.
2389    
2390    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
2391        each matching line was coloured. Now it goes on to look for further matches
2392        of any of the test patterns, which is the same behaviour as GNU grep.
2393    
2394    7.  A pattern that could match an empty string could cause pcregrep to loop; it
2395        doesn't make sense to accept an empty string match in pcregrep, so I have
2396        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
2397        seems to be how GNU grep behaves. [But see later change 40 for release
2398        8.33.]
2399    
2400    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
2401        start or after a newline", because the conditional assertion was not being
2402        correctly handled. The rule now is that both the assertion and what follows
2403        in the first alternative must satisfy the test.
2404    
2405    9.  If auto-callout was enabled in a pattern with a conditional group whose
2406        condition was an assertion, PCRE could crash during matching, both with
2407        pcre_exec() and pcre_dfa_exec().
2408    
2409    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
2410        used for matching.
2411    
2412    11. Unicode property support in character classes was not working for
2413        characters (bytes) greater than 127 when not in UTF-8 mode.
2414    
2415    12. Added the -M command line option to pcretest.
2416    
2417    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
2418    
2419    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
2420    
2421    16. Added comments and documentation about mis-use of no_arg in the C++
2422        wrapper.
2423    
2424    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
2425        from Martin Jerabek that uses macro names for all relevant character and
2426        string constants.
2427    
2428    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
2429        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
2430        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
2431        these, but not everybody uses configure.
2432    
2433    19. A conditional group that had only one branch was not being correctly
2434        recognized as an item that could match an empty string. This meant that an
2435        enclosing group might also not be so recognized, causing infinite looping
2436        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
2437        with the subject "ab", where knowledge that the repeated group can match
2438        nothing is needed in order to break the loop.
2439    
2440    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
2441        exec(), but without supplying a callout function, matching went wrong.
2442    
2443    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
2444        leak if the size of the offset vector was greater than 30. When the vector
2445        is smaller, the saved offsets during recursion go onto a local stack
2446        vector, but for larger vectors malloc() is used. It was failing to free
2447        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
2448        error, in fact).
2449    
2450    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
2451        heapframe that is used only when UTF-8 support is enabled. This caused no
2452        problem, but was untidy.
2453    
2454    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
2455        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
2456        included within another project.
2457    
2458    24. Steven Van Ingelgem's patches to add more options to the CMake support,
2459        slightly modified by me:
2460    
2461          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
2462              not building pcregrep.
2463    
2464          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
2465              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
2466    
2467    25. Forward references, both numeric and by name, in patterns that made use of
2468        duplicate group numbers, could behave incorrectly or give incorrect errors,
2469        because when scanning forward to find the reference group, PCRE was not
2470        taking into account the duplicate group numbers. A pattern such as
2471        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
2472    
2473    26. Changed a few more instances of "const unsigned char *" to USPTR, making
2474        the feature of a custom pointer more persuasive (as requested by a user).
2475    
2476    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
2477        pcretest.c, inside #ifndefs, because it seems they are sometimes already
2478        pre-defined.
2479    
2480    28. Added support for (*UTF8) at the start of a pattern.
2481    
2482    29. Arrange for flags added by the "release type" setting in CMake to be shown
2483        in the configuration summary.
2484    
2485    
2486    Version 7.8 05-Sep-08
2487    ---------------------
2488    
2489    1.  Replaced UCP searching code with optimized version as implemented for Ad
2490        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2491        stage table and inline lookup instead of a function, giving speed ups of 2
2492        to 5 times on some simple patterns that I tested. Permission was given to
2493        distribute the MultiStage2.py script that generates the tables (it's not in
2494        the tarball, but is in the Subversion repository).
2495    
2496    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2497        scripts.
2498    
2499    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2500        a group with a zero qualifier. The result of the study could be incorrect,
2501        or the function might crash, depending on the pattern.
2502    
2503    4.  Caseless matching was not working for non-ASCII characters in back
2504        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2505        It now works when Unicode Property Support is available.
2506    
2507    5.  In pcretest, an escape such as \x{de} in the data was always generating
2508        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2509        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2510        truncation.
2511    
2512    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2513    
2514    7.  Added two (int) casts to pcregrep when printing the difference of two
2515        pointers, in case they are 64-bit values.
2516    
2517    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
2518        test 2 if it fails.
2519    
2520    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2521        and a #define of that name to empty if it is not externally set. This is to
2522        allow users of MSVC to set it if necessary.
2523    
2524    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2525        the convenience functions in the pcre_get.c source file.
2526    
2527    11. An option change at the start of a pattern that had top-level alternatives
2528        could cause overwriting and/or a crash. This command provoked a crash in
2529        some environments:
2530    
2531          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2532    
2533        This potential security problem was recorded as CVE-2008-2371.
2534    
2535    12. For a pattern where the match had to start at the beginning or immediately
2536        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2537        pcre_dfa_exec() could read past the end of the passed subject if there was
2538        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2539        pcretest so that it places the subject at the end of its malloc-ed buffer.
2540    
2541    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2542        exec() might read past the end of the data buffer in UTF-8 mode.
2543    
2544    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2545        the data contained the byte 0x85 as part of a UTF-8 character within its
2546        first line. This applied both to normal and DFA matching.
2547    
2548    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2549        /^[^d]*?$/8 failed to match "abc".
2550    
2551    16. Added a missing copyright notice to pcrecpp_internal.h.
2552    
2553    17. Make it more clear in the documentation that values returned from
2554        pcre_exec() in ovector are byte offsets, not character counts.
2555    
2556    18. Tidied a few places to stop certain compilers from issuing warnings.
2557    
2558    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2559        supplied by Stefan Weber. I made a further small update for 7.8 because
2560        there is a change of source arrangements: the pcre_searchfuncs.c module is
2561        replaced by pcre_ucd.c.
2562    
2563    
2564    Version 7.7 07-May-08
2565  ---------------------  ---------------------
2566    
2567  1.  Applied Craig's patch to sort out a long long problem: "If we can't convert  1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2568      a string to a long long, pretend we don't even have a long long." This is      a string to a long long, pretend we don't even have a long long." This is
2569      done by checking for the strtoq, strtoll, and _strtoi64 functions.      done by checking for the strtoq, strtoll, and _strtoi64 functions.
2570    
2571  2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with  2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2572      pre-7.6 versions, which defined a global no_arg variable instead of putting      pre-7.6 versions, which defined a global no_arg variable instead of putting
2573      it in the RE class.      it in the RE class. (See also #8 below.)
2574    
2575  3.  Remove a line of dead code, identified by coverity and reported by Nuno  3.  Remove a line of dead code, identified by coverity and reported by Nuno
2576      Lopes.      Lopes.
2577    
2578  4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:  4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2579    
2580      (1) The include/exclude patterns were being applied to the whole pathnames      (1) The include/exclude patterns were being applied to the whole pathnames
2581          of files, instead of just to the final components.          of files, instead of just to the final components.
2582    
2583      (2) If there was more than one level of directory, the subdirectories were      (2) If there was more than one level of directory, the subdirectories were
2584          skipped unless they satisfied the include/exclude conditions. This is          skipped unless they satisfied the include/exclude conditions. This is
2585          inconsistent with GNU grep (and could even be seen as contrary to the          inconsistent with GNU grep (and could even be seen as contrary to the
# Line 27  Version 7.7 05-Mar-08 Line 2587  Version 7.7 05-Mar-08
2587          The action now is always to scan all levels of directory, and just          The action now is always to scan all levels of directory, and just
2588          apply the include/exclude patterns to regular files.          apply the include/exclude patterns to regular files.
2589    
2590    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2591        --exclude_dir in the tests to avoid scanning .svn directories.
2592    
2593    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2594        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2595        doesn't support NULs in patterns.
2596    
2597    7.  Added some missing "const"s to declarations of static tables in
2598        pcre_compile.c and pcre_dfa_exec.c.
2599    
2600    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2601        caused by fix #2  above. (Subsequently also a second patch to fix the
2602        first patch. And a third patch - this was a messy problem.)
2603    
2604    9.  Applied Craig's patch to remove the use of push_back().
2605    
2606    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2607        matching function regexec().
2608    
2609    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2610        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2611        references. PCRE supports relative numbers with this syntax (I don't think
2612        Oniguruma does).
2613    
2614    12. Previously, a group with a zero repeat such as (...){0} was completely
2615        omitted from the compiled regex. However, this means that if the group
2616        was called as a subroutine from elsewhere in the pattern, things went wrong
2617        (an internal error was given). Such groups are now left in the compiled
2618        pattern, with a new opcode that causes them to be skipped at execution
2619        time.
2620    
2621    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2622        to the way PCRE behaves:
2623    
2624        (a) A lone ] character is dis-allowed (Perl treats it as data).
2625    
2626        (b) A back reference to an unmatched subpattern matches an empty string
2627            (Perl fails the current match path).
2628    
2629        (c) A data ] in a character class must be notated as \] because if the
2630            first data character in a class is ], it defines an empty class. (In
2631            Perl it is not possible to have an empty class.) The empty class []
2632            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2633            The negative empty class [^] matches any one character, independently
2634            of the DOTALL setting.
2635    
2636    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2637        non-existent subpattern following a character class starting with ']' and
2638        containing () gave an internal compiling error instead of "reference to
2639        non-existent subpattern". Fortunately, when the pattern did exist, the
2640        compiled code was correct. (When scanning forwards to check for the
2641        existence of the subpattern, it was treating the data ']' as terminating
2642        the class, so got the count wrong. When actually compiling, the reference
2643        was subsequently set up correctly.)
2644    
2645    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2646        it was being rejected as not supported by pcre_dfa_exec(), even though
2647        other assertions are supported. I have made pcre_dfa_exec() support
2648        (*FAIL).
2649    
2650    16. The implementation of 13c above involved the invention of a new opcode,
2651        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2652        cannot be changed at match time, I realized I could make a small
2653        improvement to matching performance by compiling OP_ALLANY instead of
2654        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2655        on the OP_ANY path.
2656    
2657    17. Compiling pcretest on Windows with readline support failed without the
2658        following two fixes: (1) Make the unistd.h include conditional on
2659        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2660    
2661    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2662        ncurses library to be included for pcretest when ReadLine support is
2663        requested, but also to allow for it to be overridden. This patch came from
2664        Daniel Bergström.
2665    
2666    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2667        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2668        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2669        spotting this.
2670    
2671    
2672  Version 7.6 28-Jan-08  Version 7.6 28-Jan-08
2673  ---------------------  ---------------------

Legend:
Removed from v.324  
changed lines
  Added in v.1590

  ViewVC Help
Powered by ViewVC 1.1.5