Annotation of embedaddon/pcre/doc/pcrecpp.3, revision 1.1.1.1
1.1 misho 1: .TH PCRECPP 3
2: .SH NAME
3: PCRE - Perl-compatible regular expressions.
4: .SH "SYNOPSIS OF C++ WRAPPER"
5: .rs
6: .sp
7: .B #include <pcrecpp.h>
8: .
9: .SH DESCRIPTION
10: .rs
11: .sp
12: The C++ wrapper for PCRE was provided by Google Inc. Some additional
13: functionality was added by Giuseppe Maxia. This brief man page was constructed
14: from the notes in the \fIpcrecpp.h\fP file, which should be consulted for
15: further details.
16: .
17: .
18: .SH "MATCHING INTERFACE"
19: .rs
20: .sp
21: The "FullMatch" operation checks that supplied text matches a supplied pattern
22: exactly. If pointer arguments are supplied, it copies matched sub-strings that
23: match sub-patterns into them.
24: .sp
25: Example: successful match
26: pcrecpp::RE re("h.*o");
27: re.FullMatch("hello");
28: .sp
29: Example: unsuccessful match (requires full match):
30: pcrecpp::RE re("e");
31: !re.FullMatch("hello");
32: .sp
33: Example: creating a temporary RE object:
34: pcrecpp::RE("h.*o").FullMatch("hello");
35: .sp
36: You can pass in a "const char*" or a "string" for "text". The examples below
37: tend to use a const char*. You can, as in the different examples above, store
38: the RE object explicitly in a variable or use a temporary RE object. The
39: examples below use one mode or the other arbitrarily. Either could correctly be
40: used for any of these examples.
41: .P
42: You must supply extra pointer arguments to extract matched subpieces.
43: .sp
44: Example: extracts "ruby" into "s" and 1234 into "i"
45: int i;
46: string s;
47: pcrecpp::RE re("(\e\ew+):(\e\ed+)");
48: re.FullMatch("ruby:1234", &s, &i);
49: .sp
50: Example: does not try to extract any extra sub-patterns
51: re.FullMatch("ruby:1234", &s);
52: .sp
53: Example: does not try to extract into NULL
54: re.FullMatch("ruby:1234", NULL, &i);
55: .sp
56: Example: integer overflow causes failure
57: !re.FullMatch("ruby:1234567891234", NULL, &i);
58: .sp
59: Example: fails because there aren't enough sub-patterns:
60: !pcrecpp::RE("\e\ew+:\e\ed+").FullMatch("ruby:1234", &s);
61: .sp
62: Example: fails because string cannot be stored in integer
63: !pcrecpp::RE("(.*)").FullMatch("ruby", &i);
64: .sp
65: The provided pointer arguments can be pointers to any scalar numeric
66: type, or one of:
67: .sp
68: string (matched piece is copied to string)
69: StringPiece (StringPiece is mutated to point to matched piece)
70: T (where "bool T::ParseFrom(const char*, int)" exists)
71: NULL (the corresponding matched sub-pattern is not copied)
72: .sp
73: The function returns true iff all of the following conditions are satisfied:
74: .sp
75: a. "text" matches "pattern" exactly;
76: .sp
77: b. The number of matched sub-patterns is >= number of supplied
78: pointers;
79: .sp
80: c. The "i"th argument has a suitable type for holding the
81: string captured as the "i"th sub-pattern. If you pass in
82: void * NULL for the "i"th argument, or a non-void * NULL
83: of the correct type, or pass fewer arguments than the
84: number of sub-patterns, "i"th captured sub-pattern is
85: ignored.
86: .sp
87: CAVEAT: An optional sub-pattern that does not exist in the matched
88: string is assigned the empty string. Therefore, the following will
89: return false (because the empty string is not a valid number):
90: .sp
91: int number;
92: pcrecpp::RE::FullMatch("abc", "[a-z]+(\e\ed+)?", &number);
93: .sp
94: The matching interface supports at most 16 arguments per call.
95: If you need more, consider using the more general interface
96: \fBpcrecpp::RE::DoMatch\fP. See \fBpcrecpp.h\fP for the signature for
97: \fBDoMatch\fP.
98: .P
99: NOTE: Do not use \fBno_arg\fP, which is used internally to mark the end of a
100: list of optional arguments, as a placeholder for missing arguments, as this can
101: lead to segfaults.
102: .
103: .
104: .SH "QUOTING METACHARACTERS"
105: .rs
106: .sp
107: You can use the "QuoteMeta" operation to insert backslashes before all
108: potentially meaningful characters in a string. The returned string, used as a
109: regular expression, will exactly match the original string.
110: .sp
111: Example:
112: string quoted = RE::QuoteMeta(unquoted);
113: .sp
114: Note that it's legal to escape a character even if it has no special meaning in
115: a regular expression -- so this function does that. (This also makes it
116: identical to the perl function of the same name; see "perldoc -f quotemeta".)
117: For example, "1.5-2.0?" becomes "1\e.5\e-2\e.0\e?".
118: .
119: .SH "PARTIAL MATCHES"
120: .rs
121: .sp
122: You can use the "PartialMatch" operation when you want the pattern
123: to match any substring of the text.
124: .sp
125: Example: simple search for a string:
126: pcrecpp::RE("ell").PartialMatch("hello");
127: .sp
128: Example: find first number in a string:
129: int number;
130: pcrecpp::RE re("(\e\ed+)");
131: re.PartialMatch("x*100 + 20", &number);
132: assert(number == 100);
133: .
134: .
135: .SH "UTF-8 AND THE MATCHING INTERFACE"
136: .rs
137: .sp
138: By default, pattern and text are plain text, one byte per character. The UTF8
139: flag, passed to the constructor, causes both pattern and string to be treated
140: as UTF-8 text, still a byte stream but potentially multiple bytes per
141: character. In practice, the text is likelier to be UTF-8 than the pattern, but
142: the match returned may depend on the UTF8 flag, so always use it when matching
143: UTF8 text. For example, "." will match one byte normally but with UTF8 set may
144: match up to three bytes of a multi-byte character.
145: .sp
146: Example:
147: pcrecpp::RE_Options options;
148: options.set_utf8();
149: pcrecpp::RE re(utf8_pattern, options);
150: re.FullMatch(utf8_string);
151: .sp
152: Example: using the convenience function UTF8():
153: pcrecpp::RE re(utf8_pattern, pcrecpp::UTF8());
154: re.FullMatch(utf8_string);
155: .sp
156: NOTE: The UTF8 flag is ignored if pcre was not configured with the
157: --enable-utf8 flag.
158: .
159: .
160: .SH "PASSING MODIFIERS TO THE REGULAR EXPRESSION ENGINE"
161: .rs
162: .sp
163: PCRE defines some modifiers to change the behavior of the regular expression
164: engine. The C++ wrapper defines an auxiliary class, RE_Options, as a vehicle to
165: pass such modifiers to a RE class. Currently, the following modifiers are
166: supported:
167: .sp
168: modifier description Perl corresponding
169: .sp
170: PCRE_CASELESS case insensitive match /i
171: PCRE_MULTILINE multiple lines match /m
172: PCRE_DOTALL dot matches newlines /s
173: PCRE_DOLLAR_ENDONLY $ matches only at end N/A
174: PCRE_EXTRA strict escape parsing N/A
175: PCRE_EXTENDED ignore whitespaces /x
176: PCRE_UTF8 handles UTF8 chars built-in
177: PCRE_UNGREEDY reverses * and *? N/A
178: PCRE_NO_AUTO_CAPTURE disables capturing parens N/A (*)
179: .sp
180: (*) Both Perl and PCRE allow non capturing parentheses by means of the
181: "?:" modifier within the pattern itself. e.g. (?:ab|cd) does not
182: capture, while (ab|cd) does.
183: .P
184: For a full account on how each modifier works, please check the
185: PCRE API reference page.
186: .P
187: For each modifier, there are two member functions whose name is made
188: out of the modifier in lowercase, without the "PCRE_" prefix. For
189: instance, PCRE_CASELESS is handled by
190: .sp
191: bool caseless()
192: .sp
193: which returns true if the modifier is set, and
194: .sp
195: RE_Options & set_caseless(bool)
196: .sp
197: which sets or unsets the modifier. Moreover, PCRE_EXTRA_MATCH_LIMIT can be
198: accessed through the \fBset_match_limit()\fP and \fBmatch_limit()\fP member
199: functions. Setting \fImatch_limit\fP to a non-zero value will limit the
200: execution of pcre to keep it from doing bad things like blowing the stack or
201: taking an eternity to return a result. A value of 5000 is good enough to stop
202: stack blowup in a 2MB thread stack. Setting \fImatch_limit\fP to zero disables
203: match limiting. Alternatively, you can call \fBmatch_limit_recursion()\fP
204: which uses PCRE_EXTRA_MATCH_LIMIT_RECURSION to limit how much PCRE
205: recurses. \fBmatch_limit()\fP limits the number of matches PCRE does;
206: \fBmatch_limit_recursion()\fP limits the depth of internal recursion, and
207: therefore the amount of stack that is used.
208: .P
209: Normally, to pass one or more modifiers to a RE class, you declare
210: a \fIRE_Options\fP object, set the appropriate options, and pass this
211: object to a RE constructor. Example:
212: .sp
213: RE_Options opt;
214: opt.set_caseless(true);
215: if (RE("HELLO", opt).PartialMatch("hello world")) ...
216: .sp
217: RE_options has two constructors. The default constructor takes no arguments and
218: creates a set of flags that are off by default. The optional parameter
219: \fIoption_flags\fP is to facilitate transfer of legacy code from C programs.
220: This lets you do
221: .sp
222: RE(pattern,
223: RE_Options(PCRE_CASELESS|PCRE_MULTILINE)).PartialMatch(str);
224: .sp
225: However, new code is better off doing
226: .sp
227: RE(pattern,
228: RE_Options().set_caseless(true).set_multiline(true))
229: .PartialMatch(str);
230: .sp
231: If you are going to pass one of the most used modifiers, there are some
232: convenience functions that return a RE_Options class with the
233: appropriate modifier already set: \fBCASELESS()\fP, \fBUTF8()\fP,
234: \fBMULTILINE()\fP, \fBDOTALL\fP(), and \fBEXTENDED()\fP.
235: .P
236: If you need to set several options at once, and you don't want to go through
237: the pains of declaring a RE_Options object and setting several options, there
238: is a parallel method that give you such ability on the fly. You can concatenate
239: several \fBset_xxxxx()\fP member functions, since each of them returns a
240: reference to its class object. For example, to pass PCRE_CASELESS,
241: PCRE_EXTENDED, and PCRE_MULTILINE to a RE with one statement, you may write:
242: .sp
243: RE(" ^ xyz \e\es+ .* blah$",
244: RE_Options()
245: .set_caseless(true)
246: .set_extended(true)
247: .set_multiline(true)).PartialMatch(sometext);
248: .sp
249: .
250: .
251: .SH "SCANNING TEXT INCREMENTALLY"
252: .rs
253: .sp
254: The "Consume" operation may be useful if you want to repeatedly
255: match regular expressions at the front of a string and skip over
256: them as they match. This requires use of the "StringPiece" type,
257: which represents a sub-range of a real string. Like RE, StringPiece
258: is defined in the pcrecpp namespace.
259: .sp
260: Example: read lines of the form "var = value" from a string.
261: string contents = ...; // Fill string somehow
262: pcrecpp::StringPiece input(contents); // Wrap in a StringPiece
263: .sp
264: string var;
265: int value;
266: pcrecpp::RE re("(\e\ew+) = (\e\ed+)\en");
267: while (re.Consume(&input, &var, &value)) {
268: ...;
269: }
270: .sp
271: Each successful call to "Consume" will set "var/value", and also
272: advance "input" so it points past the matched text.
273: .P
274: The "FindAndConsume" operation is similar to "Consume" but does not
275: anchor your match at the beginning of the string. For example, you
276: could extract all words from a string by repeatedly calling
277: .sp
278: pcrecpp::RE("(\e\ew+)").FindAndConsume(&input, &word)
279: .
280: .
281: .SH "PARSING HEX/OCTAL/C-RADIX NUMBERS"
282: .rs
283: .sp
284: By default, if you pass a pointer to a numeric value, the
285: corresponding text is interpreted as a base-10 number. You can
286: instead wrap the pointer with a call to one of the operators Hex(),
287: Octal(), or CRadix() to interpret the text in another base. The
288: CRadix operator interprets C-style "0" (base-8) and "0x" (base-16)
289: prefixes, but defaults to base-10.
290: .sp
291: Example:
292: int a, b, c, d;
293: pcrecpp::RE re("(.*) (.*) (.*) (.*)");
294: re.FullMatch("100 40 0100 0x40",
295: pcrecpp::Octal(&a), pcrecpp::Hex(&b),
296: pcrecpp::CRadix(&c), pcrecpp::CRadix(&d));
297: .sp
298: will leave 64 in a, b, c, and d.
299: .
300: .
301: .SH "REPLACING PARTS OF STRINGS"
302: .rs
303: .sp
304: You can replace the first match of "pattern" in "str" with "rewrite".
305: Within "rewrite", backslash-escaped digits (\e1 to \e9) can be
306: used to insert text matching corresponding parenthesized group
307: from the pattern. \e0 in "rewrite" refers to the entire matching
308: text. For example:
309: .sp
310: string s = "yabba dabba doo";
311: pcrecpp::RE("b+").Replace("d", &s);
312: .sp
313: will leave "s" containing "yada dabba doo". The result is true if the pattern
314: matches and a replacement occurs, false otherwise.
315: .P
316: \fBGlobalReplace\fP is like \fBReplace\fP except that it replaces all
317: occurrences of the pattern in the string with the rewrite. Replacements are
318: not subject to re-matching. For example:
319: .sp
320: string s = "yabba dabba doo";
321: pcrecpp::RE("b+").GlobalReplace("d", &s);
322: .sp
323: will leave "s" containing "yada dada doo". It returns the number of
324: replacements made.
325: .P
326: \fBExtract\fP is like \fBReplace\fP, except that if the pattern matches,
327: "rewrite" is copied into "out" (an additional argument) with substitutions.
328: The non-matching portions of "text" are ignored. Returns true iff a match
329: occurred and the extraction happened successfully; if no match occurs, the
330: string is left unaffected.
331: .
332: .
333: .SH AUTHOR
334: .rs
335: .sp
336: .nf
337: The C++ wrapper was contributed by Google Inc.
338: Copyright (c) 2007 Google Inc.
339: .fi
340: .
341: .
342: .SH REVISION
343: .rs
344: .sp
345: .nf
346: Last updated: 17 March 2009
347: Minor typo fixed: 25 July 2011
348: .fi
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>