For beginners:
Python Regex HOWTO
&
Python Regex Syntax.
For when the details matter:
PCRE Syntax
(what jiten actually uses).
Prefix "Commands"
Queries support prefix "commands" (unrelated to regex syntax):
e.g.
+w foo (word) should give the same results
as
\bfoo\b,
+1 foo (1st word) as
^foo\b, and
+= foo (exact) as
^foo$. (The
+ prefix was chosen
because no valid regex can start with it.)
Quick Reference
-
Most letters and characters will simply match themselves;
e.g.
foo matches "foo".
-
The metacharacters
. ^ $ * + ? { } [ ] \ | ( )
allow matching using sophisticated patterns and rules; they
are escaped by preceding them with a backslash; e.g.
\\ matches an actual backslash.
-
. matches any character (except newline); e.g.
ba. matches "bar", "baz", etc.
-
^ matches at the start of a line; e.g.
^foo matches "foo" at the start of a line.
-
$ matches at the end of a line; e.g.
foo$ matches "foo" at the end of a line.
-
\b matches at a "word boundary" (the start or
end of a "word"); e.g. \bbar\b matches "bar" in
"foo bar baz", but not in "foobarbaz"; \B is
its complement.
-
* matches the preceding thing zero or more
times; e.g. fo* matches "f", "fo", "foo", etc.
-
+ matches the preceding thing one or more
times; e.g. fo+ matches "fo", "foo", etc.
-
? matches the preceding thing optionally; e.g.
fo? matches "f" or "fo".
-
{m,n} (or just {m} instead of
{m,m}) matches the preceding thing
m to n times; e.g.
fo{2,4} matches "foo", "fooo", or "foooo".
-
[...] is a character class; e.g.
[a-z] matches "a" through "z";
[あいうえお] matches "あ", "い", "う", "え", or
"お".
-
[^...] is a complementing character class; e.g.
[^a-z] matches anything but "a" through "z".
-
\d matches any decimal digit (equivalent to
[0-9]); \D is its complement
(equivalent to [^0-9]).
-
\s matches any whitespace character;
\S is its complement.
-
\w matches any alphanumeric (i.e. "word" or
"letter") character; \W is its complement.
-
| is alternation; e.g. foo|bar|baz
matches "foo", "bar", or "baz".
-
(...) is grouping; e.g. ab*
matches "a", "ab", "abb", etc. whereas (ab)*
matches "", "ab", "abab", etc.; ^foo|bar$
matches "foo" at the beginning of a line or "bar" at the
end, whereas ^(foo|bar)$ matches either "foo"
or "bar" as a whole line; a backslash followed by the number
of the group (starting from 1) can be used later in the
pattern to refer back to what it (actually) matched; e.g.
[a-z]{2} matches "aa", "ab", "za", etc. whereas
([a-z])\1 matches "aa", "bb", etc. (but not
"ab" or "za").
-
\p{...} matches a unicode property; e.g.
\p{Han} matches kanji,
\p{Hiragana} matches hiragana, and
\p{Katakana} matches katakana;
\P{...} is its complement.
-
For easy matching of Japanese, jiten supports these
non-standard aliases:
\pK for
\p{Han}, \ph for
\p{Hiragana}, and \pk for
\p{Katakana}.
Examples
-
+w cat (\bcat\b) matches "cat" in
"the cat" (but not in e.g. "indicates").
-
+1 cat (^cat\b) matches "cat" in
"cat" or "cat (esp. the domestic cat, Felis catus)" (but not
in e.g. "category").
-
+= cat (^cat$) matches "cat"
exactly.
-
+= 猫\pK (^猫\pK$) matches "猫"
followed by exactly one other kanji.
-
+= (\pK)\1 (^(\pK)\1$) matches
e.g. "人人".