Haskell signature highlighting

[NickHu]

Currently, the syntax highlighting for Haskell only highlights the first word before ::.

When I'm writing, I often like to provide a long list of type calculations; for instance, consider below

I believe that it would be better to highlight f x and f . g also; in general, I think it should highlight up to the preceding parenthesis (i.e. id should not be highlighted).

Another strange thing is that if the function before the double colon is fmap, no highlighting is applied...

[jgm]

I've verified that this matches kate's highlighting behavior, so this isn't a bug in the implementation. Perhaps the haskell.xml file could be improved, though, to give better behavior.

Note that we apply the following local patch to the stock haskell.xml (and this explains why fmap isn't highlighted as a function, as it is in kate):

diff --git a/xml/haskell.xml b/xml/haskell.xml
index 27ca018..8997722 100644
--- a/xml/haskell.xml
+++ b/xml/haskell.xml
@@ -476,9 +476,9 @@
 
     <itemData name="Keyword"          defStyleNum="dsKeyword"  spellChecking="false" />
     <itemData name="Type Prelude"     defStyleNum="dsDataType" spellChecking="false" />
-    <itemData name="Function Prelude" defStyleNum="dsFunction" spellChecking="false" />
-    <itemData name="Data Prelude"     defStyleNum="dsKeyword"  spellChecking="false" />
-    <itemData name="Class Prelude"    defStyleNum="dsKeyword"  spellChecking="false" />
+    <itemData name="Function Prelude" defStyleNum="dsNormal"   spellChecking="false" />
+    <itemData name="Data Prelude"     defStyleNum="dsDataType" spellChecking="false" />
+    <itemData name="Class Prelude"    defStyleNum="dsDataType" spellChecking="false" />
 
     <itemData name="Signature"        defStyleNum="dsOthers"  spellChecking="false" />
     <itemData name="Function"         defStyleNum="dsNormal"   spellChecking="false" />

[NickHu]

Yes, I had assumed this was a detail of the implementation rather than a bug. I started this issue to start a dialogue so that a consensus could be reached on whether or not this change is actually something that should be actioned. What is the rationale behind that patch?

I would submit a pull request myself, but I don't really know how the highlighting XML style works.

[jgm]

The patch dates to https://github.com/jgm/highlighting-kate/pull/32 I'm happy to reconsider it; perhaps you should comment there to see if @mekeor can elaborate on the motivation. I think I agreed at the time that the effect was more pleasing, but it does seem a bit odd to distinguish Prelude functions from others.

There are references on the xml format at the end of the skylighting README.md, if you want to play with revisions. Note that you can manually load an xml definition with the skylighting executable; so you can iterate and test without recompiling the library.

+++ Nick Hu [Aug 03 17 08:27 ]:

Yes, I had assumed this was a detail of the implementation rather than a bug. I started this issue to start a dialogue so that a consensus could be reached on whether or not this change is actually something that should be actioned. What is the rationale behind that patch?

I would submit a pull request myself, but I don't really know how the highlighting XML style works.

— You are receiving this because you commented. Reply to this email directly, [1]view it on GitHub, or [2]mute the thread.

References

1. https://github.com/jgm/skylighting/issues/12#issuecomment-319903513
2. https://github.com/notifications/unsubscribe-auth/AAAL5Gyv1TaRNbIu_wyCC7z7DgEvlN8wks5sUYSCgaJpZM4OrnYg

[ghost]

Sorry, I don't remember my motivation for that patch. It has been a long time. So I don't know why I changed the XML like that. I can't help you here.

(Anyway, it's funny that we meet here again, @NickHu, after meeting at the xmonad-contrib repository. 😃 )

[ghost]

You can see the effect of that patch between https://imgur.com/3FK5bOo and https://imgur.com/w3fAWZP