You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We'll probably want to use LanguageTool. There's a python wrapper that we should try.
But most important is thinking about how we can represent this AST mess as plaintext for LanguageTool to operate on, and then map that plaintext back to the AST nodes so that we can accurately highlight in the source code where errors came from.
Also:
$ time hunspell -l /tmp/test.txt
[...]
real 0m0.065s
user 0m0.062s
sys 0m0.004s
$ time languagetool --language en-US /tmp/test.txt
[...]
real 0m6.373s
user 0m20.568s
sys 0m0.590s
So we can certainly afford to do automatic spell checking, but grammar etc. checking should be reserved for a special command (zoia check, probably) and should download languagetool only then (considering it's like 200MB in size).
Edit: actually, I want zoia build to be as fast as possible - turn code into output (HTML/plaintext) ASAP. zoia check can afford to be slower.
Also think about how to use the system languagetool when possible.
Not a v0.1.0 thing, spellchecking can be added much later.
The text was updated successfully, but these errors were encountered:
We'll probably want to use LanguageTool. There's a python wrapper that we should try.
But most important is thinking about how we can represent this AST mess as plaintext for LanguageTool to operate on, and then map that plaintext back to the AST nodes so that we can accurately highlight in the source code where errors came from.
Also:
So we can certainly afford to do automatic spell checking, but grammar etc. checking should be reserved for a special command (
zoia check
, probably) and should download languagetool only then (considering it's like 200MB in size).Edit: actually, I want
zoia build
to be as fast as possible - turn code into output (HTML/plaintext) ASAP.zoia check
can afford to be slower.Also think about how to use the system languagetool when possible.
Not a v0.1.0 thing, spellchecking can be added much later.
The text was updated successfully, but these errors were encountered: