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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 15 - (show annotations) (download)
Sat Feb 24 21:38:25 2007 UTC (7 years, 7 months ago) by nigel
File size: 10355 byte(s)
Load pcre-1.06 into code/trunk.

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12