This is a question about the library of parsing combinators which
comes with the hbc compiler.
The library includes the definition of a class, $Token$, with the
following signature:
>class (Text a) => Token a where {
> compareT :: a -> a -> OrderedT;
> stringT :: a -> String;
> positionT :: a -> String;
> eqT :: a -> a -> Bool
> }
where $OrderedT$ is defined by
>data OrderedT = LtT | GtT | EqT | UnT deriving (Eq)
The parsing combinators can only be used to parse lists of Tokens
(fair enough). The problem (for me) is that I don't know what the
class means (I don't have access to the source). I guessed that:
1.) $compareT$ is used to order tokens according to the likelihood of
their appearance.
2.) $stringT$ converts a token to a string.
3.) $eqT$ is just a standard equality test (this guess is very shaky).
I have no idea of the intended meaning of $positionT$.
I used my guesses to define the following stubs:
>instance Token Char where
> compareT x y | x<y = LtT -- probably ok
> | x>y = GtT
> | x==y = EqT
> | otherwise = UnT
>
> stringT x = [x] -- probably ok
>
> eqT = (==) -- a bad guess
>
> positionT = stringT -- a cop-out
(I want to parse a list of characters.)
Obviously, this is rather unsatisfactory, and I *think* that it might
be affecting the parser I've written; it seems to work but it's very
very slow. I know, by the way, that parsers written using combinators
can exhibit strange space-time consumption; it's also fairly likely
that I've made a mistake. Whatever the cause of my problems is, it
would be nice to get the interpretation of $Token$ cleared up; if
anyone should answer my question then I will greatly apreciate it.
thanks,
balan
--------------------------------------------------------------------------
balan muthurajah [EMAIL PROTECTED]
QMW College, London
--------------------------------------------------------------------------
