Theoretically, the parser is matching a stream of terminal symbols from some "alphabet". The primitive matcher for terminal symbols can be qualified by a predicate (trivially TRUE, for "any") that identifies classes (or ranges) of terminal symbols. There is nothing preventing the terminal alphabet from being the tokens from an upstream process. In fact, I've implemented chained parser/transformers, where the output of one phase is a sequence, which is read as a stream of parse-trees by the next parser phase.
On Fri, May 10, 2019 at 5:02 PM Juancarlo Añez <apal...@gmail.com> wrote: > Hello, > > Please point me to the reading list for making PEG work when there's an > existing and required tokenizer? > > Thanks in advance, and regards, > > -- > Juancarlo *Añez* > _______________________________________________ > PEG mailing list > PEG@lists.csail.mit.edu > https://lists.csail.mit.edu/mailman/listinfo/peg >
_______________________________________________ PEG mailing list PEG@lists.csail.mit.edu https://lists.csail.mit.edu/mailman/listinfo/peg
