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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 17 - (show annotations) (download)
Sat Feb 24 21:38:29 2007 UTC (7 years, 6 months ago) by nigel
File size: 10521 byte(s)
Load pcre-1.07 into code/trunk.

1 ChangeLog for PCRE
2 ------------------
3
4
5 Version 1.07 16-Feb-98
6 ----------------------
7
8 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
9 repeat of a potentially empty string).
10
11
12 Version 1.06 23-Jan-98
13 ----------------------
14
15 1. Added Markus Oberhumer's little patches for C++.
16
17 2. Literal strings longer than 255 characters were broken.
18
19
20 Version 1.05 23-Dec-97
21 ----------------------
22
23 1. Negated character classes containing more than one character were failing if
24 PCRE_CASELESS was set at run time.
25
26
27 Version 1.04 19-Dec-97
28 ----------------------
29
30 1. Corrected the man page, where some "const" qualifiers had been omitted.
31
32 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
33 input syntax.
34
35 3. Fixed memory leak which occurred when a regex with back references was
36 matched with an offsets vector that wasn't big enough. The temporary memory
37 that is used in this case wasn't being freed if the match failed.
38
39 4. Tidied pcretest to ensure it frees memory that it gets.
40
41 5. Temporary memory was being obtained in the case where the passed offsets
42 vector was exactly big enough.
43
44 6. Corrected definition of offsetof() from change 5 below.
45
46 7. I had screwed up change 6 below and broken the rules for the use of
47 setjmp(). Now fixed.
48
49
50 Version 1.03 18-Dec-97
51 ----------------------
52
53 1. A erroneous regex with a missing opening parenthesis was correctly
54 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
55 on some systems.
56
57 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
58 it was reported that one broken compiler failed on the former because "code" is
59 also an independent variable.
60
61 3. The erroneous regex a[]b caused an array overrun reference.
62
63 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
64 fail on data ending with that character. (It was going on too far, and checking
65 the next character, typically a binary zero.) This was specific to the
66 optimized code for single-character negative classes.
67
68 5. Added a contributed patch from the TIN world which does the following:
69
70 + Add an undef for memmove, in case the the system defines a macro for it.
71
72 + Add a definition of offsetof(), in case there isn't one. (I don't know
73 the reason behind this - offsetof() is part of the ANSI standard - but
74 it does no harm).
75
76 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
77 most of the places where whitespace preceded '#'. I have given up and
78 allowed the remaining 2 cases to be at the margin.
79
80 + Rename some variables in pcre to eliminate shadowing. This seems very
81 pedantic, but does no harm, of course.
82
83 6. Moved the call to setjmp() into its own function, to get rid of warnings
84 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
85
86 7. Constructs such as \d{8,} were compiling into the equivalent of
87 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
88 outcome, but in this particular case used more store than had been allocated,
89 which caused the bug to be discovered because it threw up an internal error.
90
91 8. The debugging code in both pcre and pcretest for outputting the compiled
92 form of a regex was going wrong in the case of back references followed by
93 curly-bracketed repeats.
94
95
96 Version 1.02 12-Dec-97
97 ----------------------
98
99 1. Typos in pcre.3 and comments in the source fixed.
100
101 2. Applied a contributed patch to get rid of places where it used to remove
102 'const' from variables, and fixed some signed/unsigned and uninitialized
103 variable warnings.
104
105 3. Added the "runtest" target to Makefile.
106
107 4. Set default compiler flag to -O2 rather than just -O.
108
109
110 Version 1.01 19-Nov-97
111 ----------------------
112
113 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
114 like /([ab]*)*/, that is, for classes with more than one character in them.
115
116 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
117 as /((?>a*))*/ (a PCRE_EXTRA facility).
118
119
120 Version 1.00 18-Nov-97
121 ----------------------
122
123 1. Added compile-time macros to support systems such as SunOS4 which don't have
124 memmove() or strerror() but have other things that can be used instead.
125
126 2. Arranged that "make clean" removes the executables.
127
128
129 Version 0.99 27-Oct-97
130 ----------------------
131
132 1. Fixed bug in code for optimizing classes with only one character. It was
133 initializing a 32-byte map regardless, which could cause it to run off the end
134 of the memory it had got.
135
136 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
137
138
139 Version 0.98 22-Oct-97
140 ----------------------
141
142 1. Fixed bug in code for handling temporary memory usage when there are more
143 back references than supplied space in the ovector. This could cause segfaults.
144
145
146 Version 0.97 21-Oct-97
147 ----------------------
148
149 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
150
151 2. Optimized negated single characters not to use a bit map.
152
153 3. Brought error texts together as macro definitions; clarified some of them;
154 fixed one that was wrong - it said "range out of order" when it meant "invalid
155 escape sequence".
156
157 4. Changed some char * arguments to const char *.
158
159 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
160
161 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
162 pcretest.
163
164
165 Version 0.96 16-Oct-97
166 ----------------------
167
168 1. Added a simple "pgrep" utility to the distribution.
169
170 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
171 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
172 where "ddd" means "one or more decimal digits".
173
174 3. Fixed serious bug. If a pattern had a back reference, but the call to
175 pcre_exec() didn't supply a large enough ovector to record the related
176 identifying subpattern, the match always failed. PCRE now remembers the number
177 of the largest back reference, and gets some temporary memory in which to save
178 the offsets during matching if necessary, in order to ensure that
179 backreferences always work.
180
181 4. Increased the compatibility with Perl in a number of ways:
182
183 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
184 to request this handling. The option can be set at compile or exec time.
185
186 (b) $ matches before a terminating newline by default; an option
187 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
188 mode). The option can be set at compile or exec time.
189
190 (c) The handling of \ followed by a digit other than 0 is now supposed to be
191 the same as Perl's. If the decimal number it represents is less than 10
192 or there aren't that many previous left capturing parentheses, an octal
193 escape is read. Inside a character class, it's always an octal escape,
194 even if it is a single digit.
195
196 (d) An escaped but undefined alphabetic character is taken as a literal,
197 unless PCRE_EXTRA is set. Currently this just reserves the remaining
198 escapes.
199
200 (e) {0} is now permitted. (The previous item is removed from the compiled
201 pattern).
202
203 5. Changed all the names of code files so that the basic parts are no longer
204 than 10 characters, and abolished the teeny "globals.c" file.
205
206 6. Changed the handling of character classes; they are now done with a 32-byte
207 bit map always.
208
209 7. Added the -d and /D options to pcretest to make it possible to look at the
210 internals of compilation without having to recompile pcre.
211
212
213 Version 0.95 23-Sep-97
214 ----------------------
215
216 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
217 \x20 at the start of a run of normal characters. These were being treated as
218 real characters, instead of the source characters being re-checked.
219
220
221 Version 0.94 18-Sep-97
222 ----------------------
223
224 1. The functions are now thread-safe, with the caveat that the global variables
225 containing pointers to malloc() and free() or alternative functions are the
226 same for all threads.
227
228 2. Get pcre_study() to generate a bitmap of initial characters for non-
229 anchored patterns when this is possible, and use it if passed to pcre_exec().
230
231
232 Version 0.93 15-Sep-97
233 ----------------------
234
235 1. /(b)|(:+)/ was computing an incorrect first character.
236
237 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
238 but not actually doing anything yet.
239
240 3. Treat "-" characters in classes that cannot be part of ranges as literals,
241 as Perl does (e.g. [-az] or [az-]).
242
243 4. Set the anchored flag if a branch starts with .* or .*? because that tests
244 all possible positions.
245
246 5. Split up into different modules to avoid including unneeded functions in a
247 compiled binary. However, compile and exec are still in one module. The "study"
248 function is split off.
249
250 6. The character tables are now in a separate module whose source is generated
251 by an auxiliary program - but can then be edited by hand if required. There are
252 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
253 toupper() in the code.
254
255 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
256 make them global. Abolish the function for setting them, as the caller can now
257 set them directly.
258
259
260 Version 0.92 11-Sep-97
261 ----------------------
262
263 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
264 (e.g. /a{1,3}/) was broken (I mis-optimized it).
265
266 2. Caseless matching was not working in character classes if the characters in
267 the pattern were in upper case.
268
269 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
270
271 4. Make PCRE_ANCHORED public and accept as a compile option.
272
273 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
274 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
275 pass them.
276
277 6. Give an error if bad option bits passed at compile or run time.
278
279 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
280 pcretest to cause it to pass that flag.
281
282 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
283 options, and the first character, if set.
284
285 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
286
287
288 Version 0.91 10-Sep-97
289 ----------------------
290
291 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
292 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
293
294 2. PCRE was looping on encountering an indefinitely repeated back reference to
295 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
296 Perl does - treats the match as successful.
297
298 ****

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12