next | previous | forward | backward | up | top | index | toc | home

Parsing > Parser > orP

orP -- parsing alternatives

Synopsis

Usage:
r = orP(p,q,...)
Inputs:
- (p,q,...), a sequence of parsers (of type Parser)
Outputs:
- r, a parser, a parser that accepts any sequence of tokens acceptable to one of the parsers p,q,...

Description

An abbreviation for orP(p,q) is p|q.

In case of ambiguity, the value returned by the left-most accepting parser is provided.

If one of the arguments is a string then constParser is used to convert it into a parser.

In an efficient grammar, the first token presented to r will be acceptable to at most one of the input parsers, and then the parser returned by r will be the parser returned by the single accepting input parser.

i1 : (constParser "abc" | constParser "def" : charAnalyzer) "abc"

o1 = abc

i2 : (constParser "abc" | constParser "def" : charAnalyzer) "def"

o2 = def

i3 : (constParser "abc" | "def" : charAnalyzer) "def"

o3 = def

See also

constParser -- produce a parser that accepts a fixed string, one character at a time
charAnalyzer -- a lexical analyzer that provides characters from a string one at a time
Parser : Analyzer -- combine a parser with a lexical analyzer to make a complete system

Code

-- ../../../../Macaulay2/packages/Parsing.m2:77-83
Parser | Parser := (p,q) -> new Parser from (
     c -> (
          p' := p c;
          q' := q c;
          if p' === null then q'
          else if q' === null then p'
          else if c === null then p' else p'|q'))