[Git][ghc/ghc][wip/int-index/emb-type] vdq: wording in docs and comments

[Vladislav Zavialov (@int-index)]

Vladislav Zavialov pushed to branch wip/int-index/emb-type at Glasgow Haskell Compiler / GHC


Commits:
9f1cc0f6 by Vladislav Zavialov at 2023-07-18T22:51:18+02:00
vdq: wording in docs and comments

- - - - -


4 changed files:

- compiler/GHC/Hs/Pat.hs
- compiler/GHC/Rename/Bind.hs
- compiler/GHC/Rename/Expr.hs
- docs/users_guide/exts/required_type_arguments.rst


Changes:

=====================================
compiler/GHC/Hs/Pat.hs
=====================================
@@ -588,8 +588,8 @@ isIrrefutableHsPat is_strict = goL
     -- since we cannot know until the splice is evaluated.
     go (SplicePat {})      = False
 
-    -- Unimportant, as isIrrefutableHsPat is only called in contexts
-    -- where EmbTyPat is rejected for other reasons (see TcRnIllegalTypePattern).
+    -- The behavior of this case is unimportant, as GHC will throw an error shortly
+    -- after reaching this case for other reasons (see TcRnIllegalTypePattern).
     go (EmbTyPat {})       = True
 
     go (XPat ext)          = case ghcPass @p of


=====================================
compiler/GHC/Rename/Bind.hs
=====================================
@@ -580,11 +580,10 @@ isOkNoBindPattern (L _ pat) =
           SumPat _ lp _ _ -> lpatternContainsSplice lp
           ConPat _ _ cpd  -> any lpatternContainsSplice (hsConPatArgs cpd)
           XPat (HsPatExpanded _orig new) -> patternContainsSplice new
-          -- Corner cases
-          EmbTyPat{}      ->
-            -- Unimportant, as isOkNoBindPattern is only called in contexts
-            -- where EmbTyPat is rejected for other reasons (see TcRnIllegalTypePattern).
-            True
+
+          -- The behavior of this case is unimportant, as GHC will throw an error shortly
+          -- after reaching this case for other reasons (see TcRnIllegalTypePattern).
+          EmbTyPat{} -> True
 
 {- Note [Pattern bindings that bind no variables]
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


=====================================
compiler/GHC/Rename/Expr.hs
=====================================
@@ -2305,10 +2305,11 @@ isStrictPattern (L loc pat) =
     NPat{}          -> True
     NPlusKPat{}     -> True
     SplicePat{}     -> True
-    EmbTyPat{}      ->
-        -- Unimportant, as isStrictPattern is only called in contexts
-        -- where EmbTyPat is rejected for other reasons (see TcRnIllegalTypePattern).
-        False
+
+    -- The behavior of this case is unimportant, as GHC will throw an error shortly
+    -- after reaching this case for other reasons (see TcRnIllegalTypePattern).
+    EmbTyPat{}  -> False
+
     XPat ext        -> case ghcPass @p of
 #if __GLASGOW_HASKELL__ < 811
       GhcPs -> dataConCantHappen ext


=====================================
docs/users_guide/exts/required_type_arguments.rst
=====================================
@@ -58,6 +58,7 @@ At use sites, we also instantiate this type variable explicitly::
 
 Relation to :extension:`TypeApplications`
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
 :extension:`RequiredTypeArguments` are similar to :extension:`TypeApplications`
 in that we pass a type to a function as an explicit argument. The difference is
 that type applications are optional: it is up to the caller whether to write
@@ -105,7 +106,8 @@ indistinguishble from ordinary function arguments::
 
   n = id_vdq Integer 42
 
-In this example we pass ``Integer`` as opposed to ``(type Integer)``. **This is
-not currently implemented**. But for reasons of forward-compatibility,
-:extension:`RequiredTypeArguments` does not allow the ``type`` syntax; nor does
-it imply that :extension:`ExplicitNamespaces` does.
\ No newline at end of file
+In this example we pass ``Integer`` as opposed to ``(type Integer)``.
+**This is not currently implemented**, but it demonstrates how
+:extension:`RequiredTypeArguments` is not tied to the ``type`` syntax.
+For this reason, using the ``type`` syntax requires enabling the
+:extension:`ExplicitNamespaces` extension separately.
\ No newline at end of file



View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9f1cc0f6bc2135b471937725805e35ef9158d2a8

-- 
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/9f1cc0f6bc2135b471937725805e35ef9158d2a8
You're receiving this email because of your account on gitlab.haskell.org.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.haskell.org/pipermail/ghc-commits/attachments/20230718/5a9d73a4/attachment-0001.html>