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

Contents of /code/trunk/HACKING

Parent Directory Parent Directory | Revision Log Revision Log


Revision 672 - (show annotations) (download)
Tue Aug 23 16:45:55 2011 UTC (2 years, 8 months ago) by ph10
File size: 19897 byte(s)
Update non-manpage documentation for JIT.

1 Technical Notes about PCRE
2 --------------------------
3
4 These are very rough technical notes that record potentially useful information
5 about PCRE internals. For information about testing PCRE, see the pcretest
6 documentation and the comment at the head of the RunTest file.
7
8
9 Historical note 1
10 -----------------
11
12 Many years ago I implemented some regular expression functions to an algorithm
13 suggested by Martin Richards. These were not Unix-like in form, and were quite
14 restricted in what they could do by comparison with Perl. The interesting part
15 about the algorithm was that the amount of space required to hold the compiled
16 form of an expression was known in advance. The code to apply an expression did
17 not operate by backtracking, as the original Henry Spencer code and current
18 Perl code does, but instead checked all possibilities simultaneously by keeping
19 a list of current states and checking all of them as it advanced through the
20 subject string. In the terminology of Jeffrey Friedl's book, it was a "DFA
21 algorithm", though it was not a traditional Finite State Machine (FSM). When
22 the pattern was all used up, all remaining states were possible matches, and
23 the one matching the longest subset of the subject string was chosen. This did
24 not necessarily maximize the individual wild portions of the pattern, as is
25 expected in Unix and Perl-style regular expressions.
26
27
28 Historical note 2
29 -----------------
30
31 By contrast, the code originally written by Henry Spencer (which was
32 subsequently heavily modified for Perl) compiles the expression twice: once in
33 a dummy mode in order to find out how much store will be needed, and then for
34 real. (The Perl version probably doesn't do this any more; I'm talking about
35 the original library.) The execution function operates by backtracking and
36 maximizing (or, optionally, minimizing in Perl) the amount of the subject that
37 matches individual wild portions of the pattern. This is an "NFA algorithm" in
38 Friedl's terminology.
39
40
41 OK, here's the real stuff
42 -------------------------
43
44 For the set of functions that form the "basic" PCRE library (which are
45 unrelated to those mentioned above), I tried at first to invent an algorithm
46 that used an amount of store bounded by a multiple of the number of characters
47 in the pattern, to save on compiling time. However, because of the greater
48 complexity in Perl regular expressions, I couldn't do this. In any case, a
49 first pass through the pattern is helpful for other reasons.
50
51
52 Computing the memory requirement: how it was
53 --------------------------------------------
54
55 Up to and including release 6.7, PCRE worked by running a very degenerate first
56 pass to calculate a maximum store size, and then a second pass to do the real
57 compile - which might use a bit less than the predicted amount of memory. The
58 idea was that this would turn out faster than the Henry Spencer code because
59 the first pass is degenerate and the second pass can just store stuff straight
60 into the vector, which it knows is big enough.
61
62
63 Computing the memory requirement: how it is
64 -------------------------------------------
65
66 By the time I was working on a potential 6.8 release, the degenerate first pass
67 had become very complicated and hard to maintain. Indeed one of the early
68 things I did for 6.8 was to fix Yet Another Bug in the memory computation. Then
69 I had a flash of inspiration as to how I could run the real compile function in
70 a "fake" mode that enables it to compute how much memory it would need, while
71 actually only ever using a few hundred bytes of working memory, and without too
72 many tests of the mode that might slow it down. So I refactored the compiling
73 functions to work this way. This got rid of about 600 lines of source. It
74 should make future maintenance and development easier. As this was such a major
75 change, I never released 6.8, instead upping the number to 7.0 (other quite
76 major changes were also present in the 7.0 release).
77
78 A side effect of this work was that the previous limit of 200 on the nesting
79 depth of parentheses was removed. However, there is a downside: pcre_compile()
80 runs more slowly than before (30% or more, depending on the pattern) because it
81 is doing a full analysis of the pattern. My hope was that this would not be a
82 big issue, and in the event, nobody has commented on it.
83
84
85 Traditional matching function
86 -----------------------------
87
88 The "traditional", and original, matching function is called pcre_exec(), and
89 it implements an NFA algorithm, similar to the original Henry Spencer algorithm
90 and the way that Perl works. This is not surprising, since it is intended to be
91 as compatible with Perl as possible. This is the function most users of PCRE
92 will use most of the time. From release 8.20, if PCRE is compiled with
93 just-in-time (JIT) support, and studying a compiled pattern with JIT is
94 successful, the JIT code is run instead of the normal pcre_exec() code, but the
95 result is the same.
96
97
98 Supplementary matching function
99 -------------------------------
100
101 From PCRE 6.0, there is also a supplementary matching function called
102 pcre_dfa_exec(). This implements a DFA matching algorithm that searches
103 simultaneously for all possible matches that start at one point in the subject
104 string. (Going back to my roots: see Historical Note 1 above.) This function
105 intreprets the same compiled pattern data as pcre_exec(); however, not all the
106 facilities are available, and those that are do not always work in quite the
107 same way. See the user documentation for details.
108
109 The algorithm that is used for pcre_dfa_exec() is not a traditional FSM,
110 because it may have a number of states active at one time. More work would be
111 needed at compile time to produce a traditional FSM where only one state is
112 ever active at once. I believe some other regex matchers work this way.
113
114
115 Changeable options
116 ------------------
117
118 The /i, /m, or /s options (PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL) may
119 change in the middle of patterns. From PCRE 8.13, their processing is handled
120 entirely at compile time by generating different opcodes for the different
121 settings. The runtime functions do not need to keep track of an options state
122 any more.
123
124
125 Format of compiled patterns
126 ---------------------------
127
128 The compiled form of a pattern is a vector of bytes, containing items of
129 variable length. The first byte in an item is an opcode, and the length of the
130 item is either implicit in the opcode or contained in the data bytes that
131 follow it.
132
133 In many cases below LINK_SIZE data values are specified for offsets within the
134 compiled pattern. The default value for LINK_SIZE is 2, but PCRE can be
135 compiled to use 3-byte or 4-byte values for these offsets (impairing the
136 performance). This is necessary only when patterns whose compiled length is
137 greater than 64K are going to be processed. In this description, we assume the
138 "normal" compilation options. Data values that are counts (e.g. for
139 quantifiers) are always just two bytes long.
140
141 Opcodes with no following data
142 ------------------------------
143
144 These items are all just one byte long
145
146 OP_END end of pattern
147 OP_ANY match any one character other than newline
148 OP_ALLANY match any one character, including newline
149 OP_ANYBYTE match any single byte, even in UTF-8 mode
150 OP_SOD match start of data: \A
151 OP_SOM, start of match (subject + offset): \G
152 OP_SET_SOM, set start of match (\K)
153 OP_CIRC ^ (start of data)
154 OP_CIRCM ^ multiline mode (start of data or after newline)
155 OP_NOT_WORD_BOUNDARY \W
156 OP_WORD_BOUNDARY \w
157 OP_NOT_DIGIT \D
158 OP_DIGIT \d
159 OP_NOT_HSPACE \H
160 OP_HSPACE \h
161 OP_NOT_WHITESPACE \S
162 OP_WHITESPACE \s
163 OP_NOT_VSPACE \V
164 OP_VSPACE \v
165 OP_NOT_WORDCHAR \W
166 OP_WORDCHAR \w
167 OP_EODN match end of data or \n at end: \Z
168 OP_EOD match end of data: \z
169 OP_DOLL $ (end of data, or before final newline)
170 OP_DOLLM $ multiline mode (end of data or before newline)
171 OP_EXTUNI match an extended Unicode character
172 OP_ANYNL match any Unicode newline sequence
173
174 OP_ACCEPT ) These are Perl 5.10's "backtracking control
175 OP_COMMIT ) verbs". If OP_ACCEPT is inside capturing
176 OP_FAIL ) parentheses, it may be preceded by one or more
177 OP_PRUNE ) OP_CLOSE, followed by a 2-byte number,
178 OP_SKIP ) indicating which parentheses must be closed.
179
180
181 Backtracking control verbs with data
182 ------------------------------------
183
184 OP_THEN is followed by a LINK_SIZE offset, which is the distance back to the
185 start of the current branch.
186
187 OP_MARK is followed by the mark name, preceded by a one-byte length, and
188 followed by a binary zero. For (*PRUNE), (*SKIP), and (*THEN) with arguments,
189 the opcodes OP_PRUNE_ARG, OP_SKIP_ARG, and OP_THEN_ARG are used. For the first
190 two, the name follows immediately; for OP_THEN_ARG, it follows the LINK_SIZE
191 offset value.
192
193
194 Matching literal characters
195 ---------------------------
196
197 The OP_CHAR opcode is followed by a single character that is to be matched
198 casefully. For caseless matching, OP_CHARI is used. In UTF-8 mode, the
199 character may be more than one byte long. (Earlier versions of PCRE used
200 multi-character strings, but this was changed to allow some new features to be
201 added.)
202
203
204 Repeating single characters
205 ---------------------------
206
207 The common repeats (*, +, ?) when applied to a single character use the
208 following opcodes, which come in caseful and caseless versions:
209
210 Caseful Caseless
211 OP_STAR OP_STARI
212 OP_MINSTAR OP_MINSTARI
213 OP_POSSTAR OP_POSSTARI
214 OP_PLUS OP_PLUSI
215 OP_MINPLUS OP_MINPLUSI
216 OP_POSPLUS OP_POSPLUSI
217 OP_QUERY OP_QUERYI
218 OP_MINQUERY OP_MINQUERYI
219 OP_POSQUERY OP_POSQUERYI
220
221 In ASCII mode, these are two-byte items; in UTF-8 mode, the length is variable.
222 Those with "MIN" in their name are the minimizing versions. Those with "POS" in
223 their names are possessive versions. Each is followed by the character that is
224 to be repeated. Other repeats make use of these opcodes:
225
226 Caseful Caseless
227 OP_UPTO OP_UPTOI
228 OP_MINUPTO OP_MINUPTOI
229 OP_POSUPTO OP_POSUPTOI
230 OP_EXACT OP_EXACTI
231
232 Each of these is followed by a two-byte count (most significant first) and the
233 repeated character. OP_UPTO matches from 0 to the given number. A repeat with a
234 non-zero minimum and a fixed maximum is coded as an OP_EXACT followed by an
235 OP_UPTO (or OP_MINUPTO or OPT_POSUPTO).
236
237
238 Repeating character types
239 -------------------------
240
241 Repeats of things like \d are done exactly as for single characters, except
242 that instead of a character, the opcode for the type is stored in the data
243 byte. The opcodes are:
244
245 OP_TYPESTAR
246 OP_TYPEMINSTAR
247 OP_TYPEPOSSTAR
248 OP_TYPEPLUS
249 OP_TYPEMINPLUS
250 OP_TYPEPOSPLUS
251 OP_TYPEQUERY
252 OP_TYPEMINQUERY
253 OP_TYPEPOSQUERY
254 OP_TYPEUPTO
255 OP_TYPEMINUPTO
256 OP_TYPEPOSUPTO
257 OP_TYPEEXACT
258
259
260 Match by Unicode property
261 -------------------------
262
263 OP_PROP and OP_NOTPROP are used for positive and negative matches of a
264 character by testing its Unicode property (the \p and \P escape sequences).
265 Each is followed by two bytes that encode the desired property as a type and a
266 value.
267
268 Repeats of these items use the OP_TYPESTAR etc. set of opcodes, followed by
269 three bytes: OP_PROP or OP_NOTPROP and then the desired property type and
270 value.
271
272
273 Character classes
274 -----------------
275
276 If there is only one character, OP_CHAR or OP_CHARI is used for a positive
277 class, and OP_NOT or OP_NOTI for a negative one (that is, for something like
278 [^a]). However, in UTF-8 mode, the use of OP_NOT[I] applies only to characters
279 with values < 128, because OP_NOT[I] is confined to single bytes.
280
281 Another set of 13 repeating opcodes (called OP_NOTSTAR etc.) are used for a
282 repeated, negated, single-character class. The normal single-character opcodes
283 (OP_STAR, etc.) are used for a repeated positive single-character class.
284
285 When there is more than one character in a class and all the characters are
286 less than 256, OP_CLASS is used for a positive class, and OP_NCLASS for a
287 negative one. In either case, the opcode is followed by a 32-byte bit map
288 containing a 1 bit for every character that is acceptable. The bits are counted
289 from the least significant end of each byte. In caseless mode, bits for both
290 cases are set.
291
292 The reason for having both OP_CLASS and OP_NCLASS is so that, in UTF-8 mode,
293 subject characters with values greater than 256 can be handled correctly. For
294 OP_CLASS they do not match, whereas for OP_NCLASS they do.
295
296 For classes containing characters with values > 255, OP_XCLASS is used. It
297 optionally uses a bit map (if any characters lie within it), followed by a list
298 of pairs (for a range) and single characters. In caseless mode, both cases are
299 explicitly listed. There is a flag character than indicates whether it is a
300 positive or a negative class.
301
302
303 Back references
304 ---------------
305
306 OP_REF (caseful) or OP_REFI (caseless) is followed by two bytes containing the
307 reference number.
308
309
310 Repeating character classes and back references
311 -----------------------------------------------
312
313 Single-character classes are handled specially (see above). This section
314 applies to OP_CLASS and OP_REF[I]. In both cases, the repeat information
315 follows the base item. The matching code looks at the following opcode to see
316 if it is one of
317
318 OP_CRSTAR
319 OP_CRMINSTAR
320 OP_CRPLUS
321 OP_CRMINPLUS
322 OP_CRQUERY
323 OP_CRMINQUERY
324 OP_CRRANGE
325 OP_CRMINRANGE
326
327 All but the last two are just single-byte items. The others are followed by
328 four bytes of data, comprising the minimum and maximum repeat counts. There are
329 no special possessive opcodes for these repeats; a possessive repeat is
330 compiled into an atomic group.
331
332
333 Brackets and alternation
334 ------------------------
335
336 A pair of non-capturing (round) brackets is wrapped round each expression at
337 compile time, so alternation always happens in the context of brackets.
338
339 [Note for North Americans: "bracket" to some English speakers, including
340 myself, can be round, square, curly, or pointy. Hence this usage.]
341
342 Non-capturing brackets use the opcode OP_BRA. Originally PCRE was limited to 99
343 capturing brackets and it used a different opcode for each one. From release
344 3.5, the limit was removed by putting the bracket number into the data for
345 higher-numbered brackets. From release 7.0 all capturing brackets are handled
346 this way, using the single opcode OP_CBRA.
347
348 A bracket opcode is followed by LINK_SIZE bytes which give the offset to the
349 next alternative OP_ALT or, if there aren't any branches, to the matching
350 OP_KET opcode. Each OP_ALT is followed by LINK_SIZE bytes giving the offset to
351 the next one, or to the OP_KET opcode. For capturing brackets, the bracket
352 number immediately follows the offset, always as a 2-byte item.
353
354 OP_KET is used for subpatterns that do not repeat indefinitely, while
355 OP_KETRMIN and OP_KETRMAX are used for indefinite repetitions, minimally or
356 maximally respectively (see below for possessive repetitions). All three are
357 followed by LINK_SIZE bytes giving (as a positive number) the offset back to
358 the matching bracket opcode.
359
360 If a subpattern is quantified such that it is permitted to match zero times, it
361 is preceded by one of OP_BRAZERO, OP_BRAMINZERO, or OP_SKIPZERO. These are
362 single-byte opcodes that tell the matcher that skipping the following
363 subpattern entirely is a valid branch. In the case of the first two, not
364 skipping the pattern is also valid (greedy and non-greedy). The third is used
365 when a pattern has the quantifier {0,0}. It cannot be entirely discarded,
366 because it may be called as a subroutine from elsewhere in the regex.
367
368 A subpattern with an indefinite maximum repetition is replicated in the
369 compiled data its minimum number of times (or once with OP_BRAZERO if the
370 minimum is zero), with the final copy terminating with OP_KETRMIN or OP_KETRMAX
371 as appropriate.
372
373 A subpattern with a bounded maximum repetition is replicated in a nested
374 fashion up to the maximum number of times, with OP_BRAZERO or OP_BRAMINZERO
375 before each replication after the minimum, so that, for example, (abc){2,5} is
376 compiled as (abc)(abc)((abc)((abc)(abc)?)?)?, except that each bracketed group
377 has the same number.
378
379 When a repeated subpattern has an unbounded upper limit, it is checked to see
380 whether it could match an empty string. If this is the case, the opcode in the
381 final replication is changed to OP_SBRA or OP_SCBRA. This tells the matcher
382 that it needs to check for matching an empty string when it hits OP_KETRMIN or
383 OP_KETRMAX, and if so, to break the loop.
384
385 Possessive brackets
386 -------------------
387
388 When a repeated group (capturing or non-capturing) is marked as possessive by
389 the "+" notation, e.g. (abc)++, different opcodes are used. Their names all
390 have POS on the end, e.g. OP_BRAPOS instead of OP_BRA and OP_SCPBRPOS instead
391 of OP_SCBRA. The end of such a group is marked by OP_KETRPOS. If the minimum
392 repetition is zero, the group is preceded by OP_BRAPOSZERO.
393
394
395 Assertions
396 ----------
397
398 Forward assertions are just like other subpatterns, but starting with one of
399 the opcodes OP_ASSERT or OP_ASSERT_NOT. Backward assertions use the opcodes
400 OP_ASSERTBACK and OP_ASSERTBACK_NOT, and the first opcode inside the assertion
401 is OP_REVERSE, followed by a two byte count of the number of characters to move
402 back the pointer in the subject string. When operating in UTF-8 mode, the count
403 is a character count rather than a byte count. A separate count is present in
404 each alternative of a lookbehind assertion, allowing them to have different
405 fixed lengths.
406
407
408 Once-only (atomic) subpatterns
409 ------------------------------
410
411 These are also just like other subpatterns, but they start with the opcode
412 OP_ONCE. The check for matching an empty string in an unbounded repeat is
413 handled entirely at runtime, so there is just this one opcode.
414
415
416 Conditional subpatterns
417 -----------------------
418
419 These are like other subpatterns, but they start with the opcode OP_COND, or
420 OP_SCOND for one that might match an empty string in an unbounded repeat. If
421 the condition is a back reference, this is stored at the start of the
422 subpattern using the opcode OP_CREF followed by two bytes containing the
423 reference number. OP_NCREF is used instead if the reference was generated by
424 name (so that the runtime code knows to check for duplicate names).
425
426 If the condition is "in recursion" (coded as "(?(R)"), or "in recursion of
427 group x" (coded as "(?(Rx)"), the group number is stored at the start of the
428 subpattern using the opcode OP_RREF or OP_NRREF (cf OP_NCREF), and a value of
429 zero for "the whole pattern". For a DEFINE condition, just the single byte
430 OP_DEF is used (it has no associated data). Otherwise, a conditional subpattern
431 always starts with one of the assertions.
432
433
434 Recursion
435 ---------
436
437 Recursion either matches the current regex, or some subexpression. The opcode
438 OP_RECURSE is followed by an value which is the offset to the starting bracket
439 from the start of the whole pattern. From release 6.5, OP_RECURSE is
440 automatically wrapped inside OP_ONCE brackets (because otherwise some patterns
441 broke it). OP_RECURSE is also used for "subroutine" calls, even though they
442 are not strictly a recursion.
443
444
445 Callout
446 -------
447
448 OP_CALLOUT is followed by one byte of data that holds a callout number in the
449 range 0 to 254 for manual callouts, or 255 for an automatic callout. In both
450 cases there follows a two-byte value giving the offset in the pattern to the
451 start of the following item, and another two-byte item giving the length of the
452 next item.
453
454
455 Philip Hazel
456 August 2011

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12