Regular expressions

APaGeD includes a regular expression compiler that generates linear-time matchers.

Here is a short overview of the supported operators:

• | alternation
• (...) non-matching brackets
• (?...) matching brackets (usually not used in lexemes)
• ? zero or one repetition (greedy)
• * zero or more repetitions (greedy)
• + one or more repetitions (greedy)
• ?? zero or one repetition (reluctant)
• *? zero or more repetitions (reluctant)
• +? one or more repetitions (reluctant)
• {x,y} counted occurrence (greedy). At least $x$ , at most $y$ occurrences. Omitting $x$ is equivalent to $x=0$ , omitting $y$ is equivalent to $y=\infty$ .
• {x,y}? counted occurrence (reluctant)
• . any character (precisely: 0x09-0x13, 0x20-0x7e, 0xa0-0xff, 0x0100-0x017f, 0x0180-0x024f, 0x20a3-0x20b5)
• [...] [^...] character classes
• >x negative, single character lookahead (character classes not supported for $x$ , yet)

The lookahead is a speciality of the APaGeD implementation. It is a lot faster than general lookahead but less powerful. For many situations it is powerful enough, though. For example, \*>/ matches any * that is not followed by a /.

The syntax for matching and non-matching brackets has been switched for convenience, since sub-matches are usually not used in lexemes.