Codecov Report

Merging #8958 (f4f542e) into main (2a484a3) will increase coverage by 0.12%.
The diff coverage is 91.80%.

@@            Coverage Diff             @@
##             main    #8958      +/-   ##
==========================================
+ Coverage   68.84%   68.97%   +0.12%     
==========================================
  Files         641      641              
  Lines      102325   102490     +165     
==========================================
+ Hits        70450    70691     +241     
+ Misses      31875    31799      -76     

Should we detect this and then just let them not have to put the space if they don't want to? 🤔

i haven't figure how to do that yet,😅 since defs are parsed twice? paired with completions, i think it would difficult to do that. but it's possible,

i'd probably have to rewrite parse_def() or find a way to deal with spans. so i think, before that happens, a better error message is okay

@1Kinoti I wonder if we could check how many args were passed in, and if it's one less than what we expect, we look to see if we can convert the def foo[] {} to def foo [] {} (and same for def foo() {})

i haven't figure how to do that yet,sweat_smile since defs are parsed twice? paired with completions, i think it would difficult to do that. but it's possible,

Does Nushell not use a tokenizer before parsing? I feel like a tokenizer would make this trivial (but i have zero experience with making a REPL or tokenizer, so...)

I wonder if we could check how many args were passed in, and if it's one less than what we expect, we look to see if we can convert the def foo[] {} to def foo [] {} (and same for def foo() {})

@jntrnr where would we check this? in parse_predecl() or in parse_def()? or in both? i think this is possible.

edit: also there's the --help flag which can appear anywhere, any number of times

Does Nushell not use a tokenizer before parsing? I feel like a tokenizer would make this trivial (but i have zero experience with making a REPL or tokenizer, so...)

@cmidkiff87 there is a tokenizer

@1Kinoti we can probably land this if you fix the conflicts please.

@pingiun, i cannot tell why this test keeps on failing, i see it was introduced in #8956 but before that, this pr passed all CI. can you maybe tell me what i am doing wrong?

@pingiun, i cannot tell why this test keeps on failing, i see it was introduced in #8956 but before that, this pr passed all CI. can you maybe tell me what i am doing wrong?

Just on the face of it I think you reverted some of my changes. I got this error before I changed the parser correctly. That is the rough idea but I'll check your changes to see where it goes wrong

I believe this is where your error comes from. Externs can now have more than 3 spans

aah, because externs can now have bodies!

thanks @pingiun for the catch and @fdncred for the commit

let's move forward with this. thanks.