Add Scheme mode #1231
Add Scheme mode #1231
Conversation
new file: lib/ace/mode/scheme.js new file: lib/ace/mode/scheme_highlight_rules.js
modified: demo/kitchen-sink/doclist.js modified: demo/kitchen-sink/modelist.js new file: lib/ace/mode/_test/tokens_scheme.json modified: lib/ace/mode/scheme_highlight_rules.js
| }, | ||
| { | ||
| "token" : keywordMapper, | ||
| "regex" : "[a-zA-Z_$][a-zA-Z0-9_$]*\\b" |
There was a problem hiding this comment.
This won't match things like eq? or #t. Lisp mode had the same problem.
Which symbols can be included in scheme keywords?
There was a problem hiding this comment.
I think '?' '#' and ':' at least.
I've tested #f , it's fine, but I didn't test #t
There was a problem hiding this comment.
Anything should I fix for it?
There was a problem hiding this comment.
Looks like you forgot to add docs/scheme.scheme to commit.
I see ["constant.language","#f"] in the test, but #f is highlighted as ["punctuation.definition.constant.character.scheme", "constant.character.scheme"].
I'd suggest to move constant.language into it's own regexp and change keyword regexp to something like [a-zA-Z_$][^\s()[]{}"'+]*
There was a problem hiding this comment.
I think the Lisp-mode was generated by some other format, and there're some strange things, like 0xFFLL which doesn't contained in Lisp/Scheme. And all keywords are missed in the parsing.
Maybe I need to rewrite part of the code rather than modify the keywords name, the Lisp-mode parser is a incorrect one.
There was a problem hiding this comment.
Lisp mode highlighter was converted from textmate syntax, which highlights 0xFFLL as well.
It can be wrong, since probably not many people use textmate/sublime for lisp.
If you can, improve lisp and scheme modes, please do.
new file: demo/kitchen-sink/docs/scheme.scheme
modified: demo/kitchen-sink/docs/scheme.scheme * fix typo modified: lib/ace/mode/scheme_highlight_rules.js * remove '() since it's hard to parse and trivious for coloring. * keyword and constant coloring fine now.
|
Scheme-mode looks fine now. |
|
Much better:) something like this (foo bar baz
(+ '(a b c)))
(position-if-not:ss (+ 1 2) (more (+ 1 2) 3fe #F #f))
(#eeee:ee (+ 1 2) (more (+ 1 2) "
multiline string\n
") |
| }, | ||
| { | ||
| "token": ["punctuation.definition.constant.character.scheme", "constant.character.scheme"], | ||
| "regex": "#:[^ ]+" |
There was a problem hiding this comment.
should this be #:\S+, or tabs are ok here?
There was a problem hiding this comment.
"#:[^ ]+" won't match tabs I think, and I tested, seems ok.
If you know there's an exception, could you show me an example?
There was a problem hiding this comment.
try /#:[^ ]+/.exec("#:s\ts")
maybe you've tested with soft tabs enabled and editor converted tabs to spaces?
|
Actually, we don't parse '(name' for color usually. Just for keywords and 'define' and 'define-syntax'... |
Fixed: 'minus' shouldn't be a delimiter Fixed: '\t' shouldn't be included in keword-type
|
I saw test throw wrong, but I'm not sure what's wrong. |
|
tokens_scheme.json you've added is for old version of highlighter. run |
|
OK, now passes all tests for me ;-) |
|
Merged! Thanks! |
Modified from Lisp mode and tested well.