#5255: String literals cause runtime crashes when OverloadedStrings is in effect
---------------------------------+------------------------------------------
Reporter: YitzGale | Owner:
Type: feature request | Status: new
Priority: normal | Milestone:
Component: Compiler | Version: 7.0.3
Keywords: | Testcase:
Blockedby: | Difficulty:
Os: Unknown/Multiple | Blocking:
Architecture: Unknown/Multiple | Failure: None/Unknown
---------------------------------+------------------------------------------
Comment(by snoyberg):
Replying to [comment:8 simonpj]:
> I'm having trouble following the details of this discussion; some
examples would help.
>
> As I understand it, what you want is to be able to write the Haskell
expression
> {{{
> "<foo>mumble</blah>" :: XML
> }}}
> and have a ''compile-time'' error saying that the string is ill-formed.
Is that right.
>
> If so, the solution is to hand, in the form of quasi-quotation
([http://www.haskell.org/ghc/docs/7.0-latest/html/users_guide/template-
haskell.html#th-quasiquotation]), Geoff Mainland's enhancement to Template
Haskell. You say
> {{{
> [xml| <foo>mumble</blah> |]
> }}}
We're not using IsString for the XML datatype, but rather for the Name
datatype. An element or attribute name is composed of a local name, a
namespace and a prefix, resulting in the datatype:
{{{
data Name = Name { local :: Text, namespace :: Maybe Text, prefix :: Maybe
Text }
}}}
The idea of the current IsString instance for Name is to be able to encode
both the namespace and the local name simultaneously, via "clark
notation." In other words, "foo" => Name "foo" Nothing Nothing, and
"{bar}foo" => Name "foo" (Just "bar") Nothing. The problem arises when the
programmer enters the string literal "{barfoo", which currently errors
out.
The reason we'd like to avoid using a quasi-quoter is that it's simply
much bigger syntax overhead than a string literal. My code base makes use
of the IsString instance all over the place, and having to turn "{bar}foo"
into [name|{bar}foo|] everywhere would just be inconvenient.
--
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/5255#comment:11>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
_______________________________________________
Glasgow-haskell-bugs mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
