emacs mode minor issues / wishes

[jberdine]

• when refmt reports a syntax error, the cursor is not moved to the location of the error, and next-error also does not jump there
• documentation comments don't have a distinct face from normal comments
• binding occurrence of identifiers in default face (see screenshot)
• type constructors don't have a distinct face from types
• try not in keyword face
• assert not in keyword face
• rec in let rec not in keyword face
• lazy not in keyword face
• when not in keyword face
• and isn't in keyword face
• types defined using and are not in type face
• value identifiers defined using and are in type face
• punned record fields are in default face
• calling reason-mode-indent-line at the beginning of a comment line produces: indent-line-to: Wrong type argument: wholenump, -2
• reason-mode-indent-line does not always agree with refmt (seems to indent all comments 2 columns less than refmt)
• I missed the default binding of C-c C-c to compile
• I miss tuareg-find-alternate-file normally bound to C-c C-a
• wrong face for code following identifiers ending in static, see screenshot in comment below

[yunxing]

Thanks! I haven't been dogfooding emacs mode a lot so these feedbacks are definitely helpful

[fxfactorial]

I'm trying out the emacs mode and getting error of:

error in process filter: tq-process-buffer: Spurious communication from process merlin, see buffer *spurious*

and spurious buffer:

Raised at file "pervasives.ml", line 30, characters 22-33
Called from file "arg.ml", line 207, characters 11-20
Called from file "arg.ml", line 214, characters 2-73

[yunxing]

@fxfactorial Thanks for reporting. Do you have any instruction on how to reproduce this?

[jberdine]

I suspect that this is a mismatch between versions of the reason emacs mode and the version of merlin. The backtrace is from merlin's command line argument parsing.

[jberdine]

Another issue: defining occurrences of identifiers are shown in the same face as other occurrences. See e.g.:

[fxfactorial]

@yunxing Ah, sorry for lateness, I couldn't find this thread and the github update via @jberdine finally gave me a link.

I just copy pasted in the elisp provided in the documentation, nothing special, kept rest of emacs the same.

[jberdine]

wrong face of code following identifiers ending in static, see hpred and e:

[jberdine]

value identifiers defined using and are in type face:

[yunxing]

Type face for raw text:

[anmonteiro]

transferred to reasonml-editor/reason-mode#59