split lexer into separate file, for better readability and ability to use different lexical modes if we need them later
a bit hacky attempt at fixing regexp parsing. It works in the matches {/regexp/} case, but not inside tuples.
fixed multiline strings
removed PCRE grammar for now. It's nice, but you really have to switch to a different lexical mode if you want this - otherwise it will include all the PCRE lexer rules in adl!
made NAMESPACE a fragment so most of the URI token doesn't clash with NAMESPACE
fixed URI rule to not include > , so you can include it between the <> marks in odin
moved id matching to parser rules
uppercased the main grammar names. Otherwise you get java classes with lowercase first characters.
Called it Vocabulary, because AdlLexer clashes with the lexer from the adl-grammar.
To compile
You first have to generate the lexer, then the rest - or it doesn't work:
antlr4 AdlVocabulary.g4
antlr4 Adl
Still broken (not a complete list, probably):
regular expressions in tuples
date_time patterns
probably also the other date/time/duration patterns?
What i did
Called it Vocabulary, because AdlLexer clashes with the lexer from the adl-grammar.
To compile
You first have to generate the lexer, then the rest - or it doesn't work:
Still broken (not a complete list, probably):
fixes #1 fixes #2 fixes #3