Version 0.1 Checklist
- basic parsing algorithm
- basic features
- complete 4 representations (Parsetree, Datatree, Normaltree, Canonicaltree)
- migration to sedlex (no unicode support) #6
- unparsing (Parsetree --> Wirestring) #7
- usable unparsing
- perfect unparsing
- ordering and equality #8
- canonical ordering
- structural equality
- semantical equivalence
- proper testing
- proper unit testings
- (forget . embed) among representations constitute identity
- forgetting functions accepts all inputs
- (x : Datatree) => Parsetree => Wirestriing could be parsed back to x
- (parse . unparse) constitutes identity
- (unparse . parse) constitutes identity, except for parsing errors
Roadmap to Version 1.0-ish
- better string support
-
(str|..|str)strings
-
- phantom element support
- comma, mapsto support
- representational tree traversal library
- datum access via path
- zipper library
- unicode support
- unicode symbols
- unicode string
- proper codepoint range sanity check + normalization handling (NFKC for symbols and NFC for strings)
- binary encoding
- canonical hashing
- hole and partial incorrect input parsing support
- custom lexer support
- preprocessor support
- preprocessor atom
- preprocessor lexing element
- preprocessor special forms
Roadmap to Version 2.0-ish
- pattern support
- pattern language design
- pattern match support (for schema check and info extraction)
- transform via pattern support
Nice to have features
- conversions to and from JSON
- conversions to and from language values
- ocaml record/list/variant
- javascript object/array