|
|
Boost.RegexPOSIX Basic Regular Expression Syntax |
|
The POSIX-Basic regular expression syntax is used by the Unix utility sed, and variations are used by grep and emacs. You can construct POSIX basic regular expressions in Boost.Regex by passing the flag basic to the regex constructor, for example:
// e1 is a case sensitive POSIX-Basic expression: boost::regex e1(my_expression, boost::regex::basic); // e2 a case insensitive POSIX-Basic expression: boost::regex e2(my_expression, boost::regex::basic|boost::regex::icase);
In POSIX-Basic regular expressions, all characters are match themselves except for the following special characters:
.[\*^$
The single character '.' when used outside of a character set will match any single character except:
The NULL character when the flag match_no_dot_null is passed to the matching algorithms.
The newline character when the flag match_not_dot_newline is passed to the matching algorithms.
A '^' character shall match the start of a line when used as the first character of an expression, or the first character of a sub-expression.
A '$' character shall match the end of a line when used as the last character of an expression, or the last character of a sub-expression.
A section beginning \( and ending \) acts as a marked sub-expression. Whatever matched the sub-expression is split out in a separate field by the matching algorithms. Marked sub-expressions can also repeated, or referred-to by a back-reference.
Any atom (a single character, a marked sub-expression, or a character class) can be repeated with the * operator.
For example a* will match any number of letter a's repeated zero or more times (an atom repeated zero times matches an empty string), so the expression a*b will match any of the following:
b ab aaaaaaaab
An atom can also be repeated with a bounded repeat:
a\{n\} Matches 'a' repeated exactly n times.
a\{n,\} Matches 'a' repeated n or more times.
a\{n, m\} Matches 'a' repeated between n and m times inclusive.
For example:
^a\{2,3\}$
Will match either of:
aa aaa
But neither of:
a aaaa
It is an error to use a repeat operator, if the preceding construct can not be repeated, for example:
a\(*\)
Will raise an error, as there is nothing for the * operator to be applied to.
An escape character followed by a digit n, where n is in the range 1-9, matches the same string that was matched by sub-expression n. For example the expression:
^\(a*\).*\1$
Will match the string:
aaabbaaa
But not the string:
aaabba
A character set is a bracket-expression starting with [ and ending with ], it defines a set of characters, and matches any single character that is a member of that set.
A bracket expression may contain any combination of the following:
Single characters:
For example [abc], will match any of the characters 'a', 'b', or 'c'.
Character ranges:
For example [a-c] will match any single character in the range 'a' to 'c'. By default, for POSIX-Basic regular expressions, a character x is within the range y to z, if it collates within that range; this results in locale specific behavior. This behavior can be turned off by unsetting the collate option flag - in which case whether a character appears within a range is determined by comparing the code points of the characters only
Negation:
If the bracket-expression begins with the ^ character, then it matches the complement of the characters it contains, for example [^a-c] matches any character that is not in the range a-c.
Character classes:
An expression of the form [[:name:]] matches the named character class "name", for example [[:lower:]] matches any lower case character. See character class names.
Collating Elements:
An expression of the form [[.col.] matches the collating element col. A collating element is any single character, or any sequence of characters that collates as a single unit. Collating elements may also be used as the end point of a range, for example: [[.ae.]-c] matches the character sequence "ae", plus any single character in the rangle "ae"-c, assuming that "ae" is treated as a single collating element in the current locale.
Collating elements may be used in place of escapes (which are not normally allowed inside character sets), for example [[.^.]abc] would match either one of the characters 'abc^'.
As an extension, a collating element may also be specified via its symbolic name, for example:
[[.NUL.]]
matches a NUL character.
Equivalence classes:
An expression of theform[[=col=]], matches any character or collating element whose primary sort key is the same as that for collating element col, as with collating elements the name col may be a symbolic name. A primary sort key is one that ignores case, accentation, or locale-specific tailorings; so for example [[=a=]] matches any of the characters: a, à, á, â, ã, ä, å, A, À, Á, Â, Ã, Ä and Å. Unfortunately implementation of this is reliant on the platform's collation and localisation support; this feature can not be relied upon to work portably across all platforms, or even all locales on one platform.
All of the above can be combined in one character set declaration, for example: [[:digit:]a-c[.NUL.]].
With the exception of the escape sequences \{, \}, \(, and \), which are documented above, an escape followed by any character matches that character. This can be used to make the special characters .[\*^$, "ordinary". Note that the escape character loses its special meaning inside a character set, so [\^] will match either a literal '\' or a '^'.
When there is more that one way to match a regular expression, the "best" possible match is obtained using the leftmost-longest rule.
When an expression is compiled with the flag grep set, then the expression is treated as a newline separated list of POSIX-Basic expressions, a match is found if any of the expressions in the list match, for example:
boost::regex e("abc\ndef", boost::regex::grep);
will match either of the POSIX-Basic expressions "abc" or "def".
As its name suggests, this behavior is consistent with the Unix utility grep.
In addition to the POSIX-Basic features the following characters are also special:
+ repeats the preceding atom one or more times.
? repeats the preceding atom zero or one times.
*? A non-greedy version of *.
+? A non-greedy version of +.
?? A non-greedy version of ?.
And the following escape sequences are also recognised:
\| specifies an alternative.
\(?: ... \) is a non-marking grouping construct - allows you to lexically group something without spitting out an extra sub-expression.
\w matches any word character.
\W matches any non-word character.
\sx matches any character in the syntax group x, the following emacs groupings are supported: 's', ' ', '_', 'w', '.', ')', '(', '"', '\'', '>' and '<'. Refer to the emacs docs for details.
\Sx matches any character not in the syntax grouping x.
\c and \C are not supported.
\` matches zero characters only at the start of a buffer (or string being matched).
\' matches zero characters only at the end of a buffer (or string being matched).
\b matches zero characters at a word boundary.
\B matches zero characters, not at a word boundary.
\< matches zero characters only at the start of a word.
\> matches zero characters only at the end of a word.
Finally, you should note that emacs style regular expressions are matched according to the Perl "depth first search" rules. Emacs expressions are matched this way because they contain Perl-like extensions, that do not interact well with the POSIX-style leftmost-longest rule.
There are a variety of flags that may be combined with the basic and grep options when constructing the regular expression, in particular note that the newline_alt, no_char_classes, no-intervals, bk_plus_qm and bk_plus_vbar options all alter the syntax, while the collate and icase options modify how the case and locale sensitivity are to be applied.
Revised 21 Aug 2004
© Copyright John Maddock 2004
Use, modification and distribution are subject to the Boost Software License, Version 1.0. (See accompanying file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt).