A set of tools for writing programming languages in go. Current work is on a PEG parser which will be eventually used to generate itself.
NOTE: The current implementation is missing several necessary features to be able to support actual languages.
go get github.com/Logiraptor/chicken/peg
The language is expressed as a parsing expression grammar. Rules are expressed like so:
rule <- partA partB
partA <- 'a'
partB <- ~'\\d+'
ruleA <- partA*
ruleB <- partA+
ruleC <- partA?
ruleD <- partA / partB
partA above is a string literal.
partB above is defined to recognize a regular expression denoted with a ~
before the quoted regexp.
The library takes a peg description like above, and generates a state machine which will both lex and parse a given input into a parse tree. The Parser can and should be generated only once and reused on multiple input strings.
The following have yet to be implemented.
# Comments