| blob | 74e617fcdfb8ac346c011daf3bd5270fdbceebac |
1 ------------------------------------------------------------------------------
2 -- --
3 -- GNAT LIBRARY COMPONENTS --
4 -- --
5 -- S Y S T E M . R E G P A T --
6 -- --
7 -- S p e c --
8 -- --
9 -- Copyright (C) 1986 by University of Toronto. --
10 -- Copyright (C) 1996-2010, AdaCore --
11 -- --
12 -- GNAT is free software; you can redistribute it and/or modify it under --
13 -- terms of the GNU General Public License as published by the Free Soft- --
14 -- ware Foundation; either version 3, or (at your option) any later ver- --
15 -- sion. GNAT is distributed in the hope that it will be useful, but WITH- --
16 -- OUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY --
17 -- or FITNESS FOR A PARTICULAR PURPOSE. --
18 -- --
19 -- As a special exception under Section 7 of GPL version 3, you are granted --
20 -- additional permissions described in the GCC Runtime Library Exception, --
21 -- version 3.1, as published by the Free Software Foundation. --
22 -- --
23 -- You should have received a copy of the GNU General Public License and --
24 -- a copy of the GCC Runtime Library Exception along with this program; --
25 -- see the files COPYING3 and COPYING.RUNTIME respectively. If not, see --
26 -- <http://www.gnu.org/licenses/>. --
27 -- --
28 -- GNAT was originally developed by the GNAT team at New York University. --
29 -- Extensive contributions were provided by Ada Core Technologies Inc. --
30 -- --
31 ------------------------------------------------------------------------------
33 -- This package implements roughly the same set of regular expressions as
34 -- are available in the Perl or Python programming languages.
36 -- This is an extension of the original V7 style regular expression library
37 -- written in C by Henry Spencer. Apart from the translation to Ada, the
38 -- interface has been considerably changed to use the Ada String type
39 -- instead of C-style nul-terminated strings.
41 -- Note: this package is in the System hierarchy so that it can be directly
42 -- be used by other predefined packages. User access to this package is via
43 -- a renaming of this package in GNAT.Regpat (file g-regpat.ads).
48 -- The grammar is the following:
50 -- regexp ::= expr
51 -- ::= ^ expr -- anchor at the beginning of string
52 -- ::= expr $ -- anchor at the end of string
54 -- expr ::= term
55 -- ::= term | term -- alternation (term or term ...)
57 -- term ::= item
58 -- ::= item item ... -- concatenation (item then item)
60 -- item ::= elmt -- match elmt
61 -- ::= elmt * -- zero or more elmt's
62 -- ::= elmt + -- one or more elmt's
63 -- ::= elmt ? -- matches elmt or nothing
64 -- ::= elmt *? -- zero or more times, minimum number
65 -- ::= elmt +? -- one or more times, minimum number
66 -- ::= elmt ?? -- zero or one time, minimum number
67 -- ::= elmt { num } -- matches elmt exactly num times
68 -- ::= elmt { num , } -- matches elmt at least num times
69 -- ::= elmt { num , num2 } -- matches between num and num2 times
70 -- ::= elmt { num }? -- matches elmt exactly num times
71 -- ::= elmt { num , }? -- matches elmt at least num times
72 -- non-greedy version
73 -- ::= elmt { num , num2 }? -- matches between num and num2 times
74 -- non-greedy version
76 -- elmt ::= nchr -- matches given character
77 -- ::= [range range ...] -- matches any character listed
78 -- ::= [^ range range ...] -- matches any character not listed
79 -- ::= . -- matches any single character
80 -- -- except newlines
81 -- ::= ( expr ) -- parens used for grouping
82 -- ::= \ num -- reference to num-th parenthesis
84 -- range ::= char - char -- matches chars in given range
85 -- ::= nchr
86 -- ::= [: posix :] -- any character in the POSIX range
87 -- ::= [:^ posix :] -- not in the POSIX range
89 -- posix ::= alnum -- alphanumeric characters
90 -- ::= alpha -- alphabetic characters
91 -- ::= ascii -- ascii characters (0 .. 127)
92 -- ::= cntrl -- control chars (0..31, 127..159)
93 -- ::= digit -- digits ('0' .. '9')
94 -- ::= graph -- graphic chars (32..126, 160..255)
95 -- ::= lower -- lower case characters
96 -- ::= print -- printable characters (32..127)
97 -- -- and whitespaces (9 .. 13)
98 -- ::= punct -- printable, except alphanumeric
99 -- ::= space -- space characters
100 -- ::= upper -- upper case characters
101 -- ::= word -- alphanumeric characters
102 -- ::= xdigit -- hexadecimal chars (0..9, a..f)
104 -- char ::= any character, including special characters
105 -- ASCII.NUL is not supported.
107 -- nchr ::= any character except \()[].*+?^ or \char to match char
108 -- \n means a newline (ASCII.LF)
109 -- \t means a tab (ASCII.HT)
110 -- \r means a return (ASCII.CR)
111 -- \b matches the empty string at the beginning or end of a
112 -- word. A word is defined as a set of alphanumerical
113 -- characters (see \w below).
114 -- \B matches the empty string only when *not* at the
115 -- beginning or end of a word.
116 -- \d matches any digit character ([0-9])
117 -- \D matches any non digit character ([^0-9])
118 -- \s matches any white space character. This is equivalent
119 -- to [ \t\n\r\f\v] (tab, form-feed, vertical-tab,...
120 -- \S matches any non-white space character.
121 -- \w matches any alphanumeric character or underscore.
122 -- This include accented letters, as defined in the
123 -- package Ada.Characters.Handling.
124 -- \W matches any non-alphanumeric character.
125 -- \A match the empty string only at the beginning of the
126 -- string, whatever flags are used for Compile (the
127 -- behavior of ^ can change, see Regexp_Flags below).
128 -- \G match the empty string only at the end of the
129 -- string, whatever flags are used for Compile (the
130 -- behavior of $ can change, see Regexp_Flags below).
131 -- ... ::= is used to indication repetition (one or more terms)
133 -- Embedded newlines are not matched by the ^ operator.
134 -- It is possible to retrieve the substring matched a parenthesis
135 -- expression. Although the depth of parenthesis is not limited in the
136 -- regexp, only the first 9 substrings can be retrieved.
138 -- The highest value possible for the arguments to the curly operator ({})
139 -- are given by the constant Max_Curly_Repeat below.
141 -- The operators '*', '+', '?' and '{}' always match the longest possible
142 -- substring. They all have a non-greedy version (with an extra ? after the
143 -- operator), which matches the shortest possible substring.
145 -- For instance:
146 -- regexp="<.*>" string="<h1>title</h1>" matches="<h1>title</h1>"
147 -- regexp="<.*?>" string="<h1>title</h1>" matches="<h1>"
148 --
149 -- '{' and '}' are only considered as special characters if they appear
150 -- in a substring that looks exactly like '{n}', '{n,m}' or '{n,}', where
151 -- n and m are digits. No space is allowed. In other contexts, the curly
152 -- braces will simply be treated as normal characters.
154 -- Compiling Regular Expressions
155 -- =============================
157 -- To use this package, you first need to compile the regular expression
158 -- (a string) into a byte-code program, in a Pattern_Matcher structure.
159 -- This first step checks that the regexp is valid, and optimizes the
160 -- matching algorithms of the second step.
162 -- Two versions of the Compile subprogram are given: one in which this
163 -- package will compute itself the best possible size to allocate for the
164 -- byte code; the other where you must allocate enough memory yourself. An
165 -- exception is raised if there is not enough memory.
167 -- declare
168 -- Regexp : String := "a|b";
170 -- Matcher : Pattern_Matcher := Compile (Regexp);
171 -- -- The size for matcher is automatically allocated
173 -- Matcher2 : Pattern_Matcher (1000);
174 -- -- Some space is allocated directly.
176 -- begin
177 -- Compile (Matcher2, Regexp);
178 -- ...
179 -- end;
181 -- Note that the second version is significantly faster, since with the
182 -- first version the regular expression has in fact to be compiled twice
183 -- (first to compute the size, then to generate the byte code).
185 -- Note also that you cannot use the function version of Compile if you
186 -- specify the size of the Pattern_Matcher, since the discriminants will
187 -- most probably be different and you will get a Constraint_Error
189 -- Matching Strings
190 -- ================
192 -- Once the regular expression has been compiled, you can use it as often
193 -- as needed to match strings.
195 -- Several versions of the Match subprogram are provided, with different
196 -- parameters and return results.
198 -- See the description under each of these subprograms
200 -- Here is a short example showing how to get the substring matched by
201 -- the first parenthesis pair.
203 -- declare
204 -- Matches : Match_Array (0 .. 1);
205 -- Regexp : String := "a(b|c)d";
206 -- Str : String := "gacdg";
208 -- begin
209 -- Match (Compile (Regexp), Str, Matches);
210 -- return Str (Matches (1).First .. Matches (1).Last);
211 -- -- returns 'c'
212 -- end;
214 -- Finding all occurrences
215 -- =======================
217 -- Finding all the occurrences of a regular expression in a string cannot
218 -- be done by simply passing a slice of the string. This wouldn't work for
219 -- anchored regular expressions (the ones starting with "^" or ending with
220 -- "$").
221 -- Instead, you need to use the last parameter to Match (Data_First), as in
222 -- the following loop:
224 -- declare
225 -- Str : String :=
226 -- "-- first line" & ASCII.LF & "-- second line";
227 -- Matches : Match_Array (0 .. 0);
228 -- Regexp : Pattern_Matcher := Compile ("^--", Multiple_Lines);
229 -- Current : Natural := Str'First;
230 -- begin
231 -- loop
232 -- Match (Regexp, Str, Matches, Current);
233 -- exit when Matches (0) = No_Match;
234 --
235 -- -- Process the match at position Matches (0).First
236 --
237 -- Current := Matches (0).Last + 1;
238 -- end loop;
239 -- end;
241 -- String Substitution
242 -- ===================
244 -- No subprogram is currently provided for string substitution.
245 -- However, this is easy to simulate with the parenthesis groups, as
246 -- shown below.
248 -- This example swaps the first two words of the string:
250 -- declare
251 -- Regexp : String := "([a-z]+) +([a-z]+)";
252 -- Str : String := " first second third ";
253 -- Matches : Match_Array (0 .. 2);
255 -- begin
256 -- Match (Compile (Regexp), Str, Matches);
257 -- return Str (Str'First .. Matches (1).First - 1)
258 -- & Str (Matches (2).First .. Matches (2).Last)
259 -- & " "
260 -- & Str (Matches (1).First .. Matches (1).Last)
261 -- & Str (Matches (2).Last + 1 .. Str'Last);
262 -- -- returns " second first third "
263 -- end;
265 ---------------
266 -- Constants --
267 ---------------
270 -- This exception is raised when trying to compile an invalid regular
271 -- expression. All subprograms taking an expression as parameter may raise
272 -- Expression_Error.
275 -- Maximum number of parenthesis in a regular expression. This is limited
276 -- by the size of a Character, as found in the byte-compiled version of
277 -- regular expressions.
280 -- Maximum number of repetition for the curly operator. The digits in the
281 -- {n}, {n,} and {n,m } operators cannot be higher than this constant,
282 -- since they have to fit on two characters in the byte-compiled version of
283 -- regular expressions.
286 -- Maximum size that can be allocated for a program
290 -- Number of bytes allocated for the byte-compiled version of a regular
291 -- expression. The size required depends on the complexity of the regular
292 -- expression in a complex manner that is undocumented (other than in the
293 -- body of the Compile procedure). Normally the size is automatically set
294 -- and the programmer need not be concerned about it. There are two
295 -- exceptions to this. First in the calls to Match, it is possible to
296 -- specify a non-zero size that is known to be large enough. This can
297 -- slightly increase the efficiency by avoiding a copy. Second, in the case
298 -- of calling compile, it is possible using the procedural form of Compile
299 -- to use a single Pattern_Matcher variable for several different
300 -- expressions by setting its size sufficiently large.
303 -- Used in calls to Match to indicate that the Size should be set to
304 -- a value appropriate to the expression being used automatically.
308 -- Flags that can be given at compile time to specify default
309 -- properties for the regular expression.
313 -- The automaton is optimized so that the matching is done in a case
314 -- insensitive manner (upper case characters and lower case characters
315 -- are all treated the same way).
318 -- Treat the Data we are matching as a single line. This means that
319 -- ^ and $ will ignore \n (unless Multiple_Lines is also specified),
320 -- and that '.' will match \n.
323 -- Treat the Data as multiple lines. This means that ^ and $ will also
324 -- match on internal newlines (ASCII.LF), in addition to the beginning
325 -- and end of the string.
326 --
327 -- This can be combined with Single_Line.
329 -----------------
330 -- Match_Array --
331 -----------------
341 -- Used for regular expressions that can contain parenthesized
342 -- subexpressions. Certain Match subprograms below produce Matches of type
343 -- Match_Array. Each component of Matches is set to the subrange of the
344 -- matches substring, or to No_Match if no match. Matches (N) is for the
345 -- N'th parenthesized subexpressions; Matches (0) is for the whole
346 -- expression.
347 --
348 -- For instance, if your regular expression is: "a((b*)c+)(d+)", then
349 -- 12 3
350 -- Matches (0) is for "a((b*)c+)(d+)" (the entire expression)
351 -- Matches (1) is for "(b*)c+"
352 -- Matches (2) is for "b*"
353 -- Matches (3) is for "d+"
354 --
355 -- The number of parenthesis groups that can be retrieved is limited only
356 -- by Max_Paren_Count.
357 --
358 -- Normally, the bounds of the Matches actual parameter will be
359 -- 0 .. Paren_Count (Regexp), to get all the matches. However, it is fine
360 -- if Matches is shorter than that on either end; missing components will
361 -- be ignored. Thus, in the above example, you could use 2 .. 2 if all you
362 -- care about it the second parenthesis pair "b*". Likewise, if
363 -- Matches'Last > Paren_Count (Regexp), the extra components will be set to
364 -- No_Match.
367 -- The No_Match constant is (0, 0) to differentiate between matching a null
368 -- string at position 1, which uses (1, 0) and no match at all.
370 ---------------------------------
371 -- Pattern_Matcher Compilation --
372 ---------------------------------
374 -- The subprograms here are used to precompile regular expressions for use
375 -- in subsequent Match calls. Precompilation improves efficiency if the
376 -- same regular expression is to be used in more than one Match call.
379 -- Type used to represent a regular expression compiled into byte code
382 -- A regular expression that never matches anything
384 function Compile
387 -- Compile a regular expression into internal code
388 --
389 -- Raises Expression_Error if Expression is not a legal regular expression
390 --
391 -- The appropriate size is calculated automatically to correspond to the
392 -- provided expression. This is the normal default method of compilation.
393 -- Note that it is generally not possible to assign the result of two
394 -- different calls to this Compile function to the same Pattern_Matcher
395 -- variable, since the sizes will differ.
396 --
397 -- Flags is the default value to use to set properties for Expression
398 -- (e.g. case sensitivity,...).
400 procedure Compile
405 -- Compile a regular expression into internal code
407 -- This procedure is significantly faster than the Compile function since
408 -- it avoids the extra step of precomputing the required size.
409 --
410 -- However, it requires the user to provide a Pattern_Matcher variable
411 -- whose size is preset to a large enough value. One advantage of this
412 -- approach, in addition to the improved efficiency, is that the same
413 -- Pattern_Matcher variable can be used to hold the compiled code for
414 -- several different regular expressions by setting a size that is large
415 -- enough to accommodate all possibilities.
416 --
417 -- In this version of the procedure call, the actual required code size is
418 -- returned. Also if Matcher.Size is zero on entry, then the resulting code
419 -- is not stored. A call with Matcher.Size set to Auto_Size can thus be
420 -- used to determine the space required for compiling the given regular
421 -- expression.
422 --
423 -- This function raises Storage_Error if Matcher is too small to hold
424 -- the resulting code (i.e. Matcher.Size has too small a value).
425 --
426 -- Expression_Error is raised if the string Expression does not contain
427 -- a valid regular expression.
428 --
429 -- Flags is the default value to use to set properties for Expression (case
430 -- sensitivity,...).
432 procedure Compile
436 -- Same procedure as above, expect it does not return the final
437 -- program size, and Matcher.Size cannot be Auto_Size.
441 -- Return the number of parenthesis pairs in Regexp.
442 --
443 -- This is the maximum index that will be filled if a Match_Array is
444 -- used as an argument to Match.
445 --
446 -- Thus, if you want to be sure to get all the parenthesis, you should
447 -- do something like:
448 --
449 -- declare
450 -- Regexp : Pattern_Matcher := Compile ("a(b*)(c+)");
451 -- Matched : Match_Array (0 .. Paren_Count (Regexp));
452 -- begin
453 -- Match (Regexp, "a string", Matched);
454 -- end;
456 -------------
457 -- Quoting --
458 -------------
461 -- Return a version of Str so that every special character is quoted.
462 -- The resulting string can be used in a regular expression to match
463 -- exactly Str, whatever character was present in Str.
465 --------------
466 -- Matching --
467 --------------
469 -- The Match subprograms are given a regular expression in string
470 -- form, and perform the corresponding match. The following parameters
471 -- are present in all forms of the Match call.
473 -- Expression contains the regular expression to be matched as a string
475 -- Data contains the string to be matched
477 -- Data_First is the lower bound for the match, i.e. Data (Data_First)
478 -- will be the first character to be examined. If Data_First is set to
479 -- the special value of -1 (the default), then the first character to
480 -- be examined is Data (Data_First). However, the regular expression
481 -- character ^ (start of string) still refers to the first character
482 -- of the full string (Data (Data'First)), which is why there is a
483 -- separate mechanism for specifying Data_First.
485 -- Data_Last is the upper bound for the match, i.e. Data (Data_Last)
486 -- will be the last character to be examined. If Data_Last is set to
487 -- the special value of Positive'Last (the default), then the last
488 -- character to be examined is Data (Data_Last). However, the regular
489 -- expression character $ (end of string) still refers to the last
490 -- character of the full string (Data (Data'Last)), which is why there
491 -- is a separate mechanism for specifying Data_Last.
493 -- Note: the use of Data_First and Data_Last is not equivalent to
494 -- simply passing a slice as Expression because of the handling of
495 -- regular expression characters ^ and $.
497 -- Size is the size allocated for the compiled byte code. Normally
498 -- this is defaulted to Auto_Size which means that the appropriate
499 -- size is allocated automatically. It is possible to specify an
500 -- explicit size, which must be sufficiently large. This slightly
501 -- increases the efficiency by avoiding the extra step of computing
502 -- the appropriate size.
504 -- The following exceptions can be raised in calls to Match
505 --
506 -- Storage_Error is raised if a non-zero value is given for Size
507 -- and it is too small to hold the compiled byte code.
508 --
509 -- Expression_Error is raised if the given expression is not a legal
510 -- regular expression.
512 procedure Match
519 -- This version returns the result of the match stored in Match_Array;
520 -- see comments under Match_Array above for details.
522 function Match
528 -- This version returns the position where Data matches, or if there is
529 -- no match, then the value Data'First - 1.
531 function Match
537 -- This version returns True if the match succeeds, False otherwise
539 ------------------------------------------------
540 -- Matching a Pre-Compiled Regular Expression --
541 ------------------------------------------------
543 -- The following functions are significantly faster if you need to reuse
544 -- the same regular expression multiple times, since you only have to
545 -- compile it once. For these functions you must first compile the
546 -- expression with a call to Compile as previously described.
548 -- The parameters Data, Data_First and Data_Last are as described
549 -- in the previous section.
551 function Match
556 -- Match Data using the given pattern matcher. Returns the position
557 -- where Data matches, or (Data'First - 1) if there is no match.
559 function Match
564 -- Return True if Data matches using the given pattern matcher
567 -- All except the last one below
569 procedure Match
575 -- Match Data using the given pattern matcher and store result in Matches;
576 -- see comments under Match_Array above for details.
578 -----------
579 -- Debug --
580 -----------
583 -- Dump the compiled version of the regular expression matched by Self
585 --------------------------
586 -- Private Declarations --
587 --------------------------
589 private
592 -- The Pointer type is used to point into Program_Data
594 -- Note that the pointer type is not necessarily 2 bytes
595 -- although it is stored in the program using 2 bytes
601 -- The "internal use only" fields in regexp are present to pass info from
602 -- compile to execute that permits the execute phase to run lots faster on
603 -- simple cases. They are:
605 -- First character that must begin a match or ASCII.NUL
606 -- Anchored true iff match must start at beginning of line
607 -- Must_Have pointer to string that match must include or null
608 -- Must_Have_Length length of Must_Have string
610 -- First and Anchored permit very fast decisions on suitable starting
611 -- points for a match, cutting down the work a lot. Must_Have permits fast
612 -- rejection of lines that cannot possibly match.
614 -- The Must_Have tests are costly enough that Optimize supplies a Must_Have
615 -- only if the r.e. contains something potentially expensive (at present,
616 -- the only such thing detected is * or at the start of the r.e., which can
617 -- involve a lot of backup). The length is supplied because the test in
618 -- Execute needs it and Optimize is computing it anyway.
620 -- The initialization is meant to fail-safe in case the user of this
621 -- package tries to use an uninitialized matcher. This takes advantage
622 -- of the knowledge that ASCII.NUL translates to the end-of-program (EOP)
623 -- instruction code of the state machine.