6 regex - POSIX 1003.2 regular expressions
9 Regular expressions (``RE''s), as defined in POSIX 1003.2, come in two
10 forms: modern REs (roughly those of egrep; 1003.2 calls these
11 ``extended'' REs) and obsolete REs (roughly those of ed; 1003.2
12 ``basic'' REs). Obsolete REs mostly exist for backward compatibility
13 in some old programs; they will be discussed at the end. 1003.2 leaves
14 some aspects of RE syntax and semantics open; `-' marks decisions on
15 these aspects that may not be fully portable to other 1003.2 implemen-
18 A (modern) RE is one- or more non-empty- branches, separated by `|'.
19 It matches anything that matches one of the branches.
21 A branch is one- or more pieces, concatenated. It matches a match for
22 the first, followed by a match for the second, etc.
24 A piece is an atom possibly followed by a single- `*', `+', `?', or
25 bound. An atom followed by `*' matches a sequence of 0 or more matches
26 of the atom. An atom followed by `+' matches a sequence of 1 or more
27 matches of the atom. An atom followed by `?' matches a sequence of 0
28 or 1 matches of the atom.
30 A bound is `{' followed by an unsigned decimal integer, possibly fol-
31 lowed by `,' possibly followed by another unsigned decimal integer,
32 always followed by `}'. The integers must lie between 0 and RE_DUP_MAX
33 (255-) inclusive, and if there are two of them, the first may not
34 exceed the second. An atom followed by a bound containing one integer
35 i and no comma matches a sequence of exactly i matches of the atom. An
36 atom followed by a bound containing one integer i and a comma matches a
37 sequence of i or more matches of the atom. An atom followed by a bound
38 containing two integers i and j matches a sequence of i through j
39 (inclusive) matches of the atom.
41 An atom is a regular expression enclosed in `()' (matching a match for
42 the regular expression), an empty set of `()' (matching the null
43 string)-, a bracket expression (see below), `.' (matching any single
44 character), `^' (matching the null string at the beginning of a line),
45 `$' (matching the null string at the end of a line), a `\' followed by
46 one of the characters `^.[$()|*+?{\' (matching that character taken as
47 an ordinary character), a `\' followed by any other character- (match-
48 ing that character taken as an ordinary character, as if the `\' had
49 not been present-), or a single character with no other significance
50 (matching that character). A `{' followed by a character other than a
51 digit is an ordinary character, not the beginning of a bound-. It is
52 illegal to end an RE with `\'.
54 A bracket expression is a list of characters enclosed in `[]'. It nor-
55 mally matches any single character from the list (but see below). If
56 the list begins with `^', it matches any single character (but see
57 below) not from the rest of the list. If two characters in the list
58 are separated by `-', this is shorthand for the full range of charac-
59 ters between those two (inclusive) in the collating sequence, e.g.
60 `[0-9]' in ASCII matches any decimal digit. It is illegal- for two
61 ranges to share an endpoint, e.g. `a-c-e'. Ranges are very collating-
62 sequence-dependent, and portable programs should avoid relying on them.
64 To include a literal `]' in the list, make it the first character (fol-
65 lowing a possible `^'). To include a literal `-', make it the first or
66 last character, or the second endpoint of a range. To use a literal
67 `-' as the first endpoint of a range, enclose it in `[.' and `.]' to
68 make it a collating element (see below). With the exception of these
69 and some combinations using `[' (see next paragraphs), all other spe-
70 cial characters, including `\', lose their special significance within
73 Within a bracket expression, a collating element (a character, a multi-
74 character sequence that collates as if it were a single character, or a
75 collating-sequence name for either) enclosed in `[.' and `.]' stands
76 for the sequence of characters of that collating element. The sequence
77 is a single element of the bracket expression's list. A bracket
78 expression containing a multi-character collating element can thus
79 match more than one character, e.g. if the collating sequence includes
80 a `ch' collating element, then the RE `[[.ch.]]*c' matches the first
81 five characters of `chchcc'.
83 Within a bracket expression, a collating element enclosed in `[=' and
84 `=]' is an equivalence class, standing for the sequences of characters
85 of all collating elements equivalent to that one, including itself.
86 (If there are no other equivalent collating elements, the treatment is
87 as if the enclosing delimiters were `[.' and `.]'.) For example, if o
88 and ^ are the members of an equivalence class, then `[[=o=]]',
89 `[[=^=]]', and `[o^]' are all synonymous. An equivalence class may
90 not- be an endpoint of a range.
92 Within a bracket expression, the name of a character class enclosed in
93 `[:' and `:]' stands for the list of all characters belonging to that
94 class. Standard character class names are:
101 These stand for the character classes defined in ctype(3). A locale
102 may provide others. A character class may not be used as an endpoint
105 There are two special cases- of bracket expressions: the bracket
106 expressions `[[:<:]]' and `[[:>:]]' match the null string at the begin-
107 ning and end of a word respectively. A word is defined as a sequence
108 of word characters which is neither preceded nor followed by word char-
109 acters. A word character is an alnum character (as defined by
110 ctype(3)) or an underscore. This is an extension, compatible with but
111 not specified by POSIX 1003.2, and should be used with caution in soft-
112 ware intended to be portable to other systems.
114 In the event that an RE could match more than one substring of a given
115 string, the RE matches the one starting earliest in the string. If the
116 RE could match more than one substring starting at that point, it
117 matches the longest. Subexpressions also match the longest possible
118 substrings, subject to the constraint that the whole match be as long
119 as possible, with subexpressions starting earlier in the RE taking pri-
120 ority over ones starting later. Note that higher-level subexpressions
121 thus take priority over their lower-level component subexpressions.
123 Match lengths are measured in characters, not collating elements. A
124 null string is considered longer than no match at all. For example,
125 `bb*' matches the three middle characters of `abbbc',
126 `(wee|week)(knights|nights)' matches all ten characters of `week-
127 nights', when `(.*).*' is matched against `abc' the parenthesized sub-
128 expression matches all three characters, and when `(a*)*' is matched
129 against `bc' both the whole RE and the parenthesized subexpression
130 match the null string.
132 If case-independent matching is specified, the effect is much as if all
133 case distinctions had vanished from the alphabet. When an alphabetic
134 that exists in multiple cases appears as an ordinary character outside
135 a bracket expression, it is effectively transformed into a bracket
136 expression containing both cases, e.g. `x' becomes `[xX]'. When it
137 appears inside a bracket expression, all case counterparts of it are
138 added to the bracket expression, so that (e.g.) `[x]' becomes `[xX]'
139 and `[^x]' becomes `[^xX]'.
141 No particular limit is imposed on the length of REs-. Programs
142 intended to be portable should not employ REs longer than 256 bytes, as
143 an implementation can refuse to accept such REs and remain POSIX-com-
146 Obsolete (``basic'') regular expressions differ in several respects.
147 `|', `+', and `?' are ordinary characters and there is no equivalent
148 for their functionality. The delimiters for bounds are `\{' and `\}',
149 with `{' and `}' by themselves ordinary characters. The parentheses
150 for nested subexpressions are `\(' and `\)', with `(' and `)' by them-
151 selves ordinary characters. `^' is an ordinary character except at the
152 beginning of the RE or- the beginning of a parenthesized subexpression,
153 `$' is an ordinary character except at the end of the RE or- the end of
154 a parenthesized subexpression, and `*' is an ordinary character if it
155 appears at the beginning of the RE or the beginning of a parenthesized
156 subexpression (after a possible leading `^'). Finally, there is one
157 new type of atom, a back reference: `\' followed by a non-zero decimal
158 digit d matches the same sequence of characters matched by the dth
159 parenthesized subexpression (numbering subexpressions by the positions
160 of their opening parentheses, left to right), so that (e.g.)
161 `\([bc]\)\1' matches `bb' or `cc' but not `bc'.
166 POSIX 1003.2, section 2.8 (Regular Expression Notation).
169 Written by Henry Spencer, based on the 1003.2 spec.
172 Having two kinds of REs is a botch.
174 The current 1003.2 spec says that `)' is an ordinary character in the
175 absence of an unmatched `('; this was an unintentional result of a
176 wording error, and change is likely. Avoid relying on it.
178 Back references are a dreadful botch, posing major problems for effi-
179 cient implementations. They are also somewhat vaguely defined (does
180 `a\(\(b\)*\2\)*d' match `abbbd'?). Avoid using them.
182 1003.2's specification of case-independent matching is vague. The
183 ``one case implies all cases'' definition given above is current con-
184 sensus among implementors as to the right interpretation.
186 The syntax for word boundaries is incredibly ugly.