GUIGUTS

Regular Expression Reference Card

Main Menu RegEx Refcard Specific Chars Classes of Chars Quantifiers Anchors Meta-chars Replacements Hotkey Refcard Topical Index

http://cpan.uwinnipeg.ca/htdocs/perl/perlreref.html http://cpan.uwinnipeg.ca/htdocs/perl/perlretut.html http://cpan.uwinnipeg.ca/htdocs/perl/perlunicode.html

Guiguts is implemented in Perl, and so the regular expression (regex) syntax it supports is Perl's regex syntax, with a few twists. In general any tutorial or essay that describes Perl regexes is accurate for Guiguts. For links to some, see this page.

The following tables divide regex syntax into related groups. Scroll, or click in the navigation bar, to go to one of the tables.

Forms that Match Specific Characters

These forms match a single specific character.

When the literal \n appears anywhere in the regex, Guiguts causes Perl to treat the entire document as a single string. The anchors (see below) ^ and $ now refer to the beginning and end of the document. The dot can match \n so an expression like .* can match the entire document. This feature lets you to search for multi-line phrases but requires extra care in using greedy quantifiers.

Forms that Match Classes of Characters

These forms match, not a specific character, but any one of a class of related characters.

Named Character Classes

POSIX (a standards body) and Unicode (another) have defined names for character classes. POSIX names are delimited this way: [:alnum:] and must appear inside class brackets, thus: [[:alpha:][:digit:]] is the class of alphabetics plus digits. To negate a POSIX class, insert a caret after the first colon: [[:^alpha:]] is the class of non-alphabetics.

Unicode names are used so: \p{IsAlnum}, and negated by use of a capital P, so \P{IsAlpha} is the class of non-alphabetics.

The important feature of these named classes is that they are Unicode-aware, so for example \p{IsLower} includes all lowercase letters from all languages, and \p{IsPunct} matches punctuation in Greek and Farsi as well as Latin-1.

Forms that Quantify

These forms always follow an expression and modify how many times the expression should repeat.

Greediness

The basic quantifiers are "greedy"; that is, they always match as many repeats as they can, while still allowing the entire regex to match. Greediness can lead to unexpected results. For example, this search for italic markup:
<i>.+</i>
when applied to a line like this:
<i>What?</i> What do you <i>mean?</i>
will match the entire line: the greedy .+ matches as much as it can while still allowing the entire regex to match, which means it matches everything through the second "?".

To tame the greed of a quantifier, append a question-mark to it. Then it matches as few repetitions as it can, while still allowing the regex to match. The test
<i>.+?</i>
when applied to:
<i>What?</i> What do you <i>mean?</i>
matches only the first marked-up word.

Forms that Match Positions

These forms match particular positions such as "end of line," or particular transitions such as "start of word." They have zero width, but serve to "anchor" the rest of the expression to a fixed spot.

Forms that Control Matching

These forms control how the search is carried out, including grouping and alternates.

Replacement Patterns

These forms are used in the replacement text. They control what text will replace the text that is found. Their use is discussed on this page.

Overview Guiguts and this manual are protected by the Perl Artistic License