Prettify symbol troubles with dot operator

[hmgibson23]

Great fix for the issues with >>=. Thanks for the super fast response. The fix however has unfortunately introduced another bug. Now in imports . is transformed into ∘ so you get the fairly unpleasant import Data∘ByteString∘Char8 instead of Data.ByteString.Char8.

Sorry I feel like the scrooge of prettify-symbols-mode!

[gracjan]

@hmgibson23: If you look through the code you will see that "." requires special handling.

Are you able to look at this yourself and create a pull request?

[hmgibson23]

Yes will have a look

[gracjan]

@hmgibson23: should be pretty easy: just check if a dot has space before and space after.

[hmgibson23]

Yeah good call.

[ivan-m]

@gracjan it can't just be a matter of requiring a space before/after: that means that (.) wouldn't be pretty-fied.

Then again, how do you tell if Foo.Bar is subsequently applying two constructors or just applying one qualified constructor?

[gracjan]

See here:

https://www.haskell.org/onlinereport/haskell2010/haskellch2.html#x7-180002.4

specifically this table:

This Lexes as this

f.g     f . g (three tokens)
F.g     F.g (qualified ‘g’)
f..     f .. (two tokens)
F..     F.. (qualified ‘.’)
F.      F . (two tokens) 

[hmgibson23]

I'm not that familiar with Elisp, so would something like

    (or (re-search-backward "\\<forall\\>[^.\"]*\\="
                            (line-beginning-position) t)
        (re-search-backward "[^ ]\.[^ ]"
                            (line-beginning-position) t))))

work?

Which makes me think the forall regex is actually redundant.

[gracjan]

Close, but no cigar. I guess using char-before and char-after would be better idea. You can read about those here:

http://www.gnu.org/software/emacs/manual/html_node/elisp/Near-Point.html