[m-rev.] for review: add string.line type

[Jonathan Morgan]

On 4/19/07, Peter Wang <wangp at students.csse.unimelb.edu.au> wrote:
> On 2007-04-19, Jonathan Morgan <jonmmorgan at gmail.com> wrote:
> > >+    % This type is used for defining stream typeclass instances where the
> > >raw
> > >+    % string type would be ambiguous.
> > >+    %
> > >+:- type line
> > >+    --->    line(string).
> > >+
> >
> > I don't think that this comment is sufficient.  I would like a
> > definition of what the term `line' actually means in this context.
>
> How about:
>
>     % A line is:
>     %
>     % - a possibly empty sequence of non-newline characters terminated by a
>     %   newline character; or
>     % - a non-empty sequence of non-newline characters terminated by the end
>     %   of the file.

Sounds reasonable.

> > For example, a line may contain the end-of-line character, or it may
> > not.  I have seen the term used both ways (and, as you already know
> > that the end-of-line character must have occurred at the end of the
> > string containing the line, I would argue that preserving it is not
> > necessary when you are using the line type).
>
> The problem with that is you can't tell if a "line" at the end of a file
> was terminated with a newline or not.

That is a little unfortunate (though I wonder whether it really matters).

Jon
--------------------------------------------------------------------------
mercury-reviews mailing list
Post messages to:       mercury-reviews at csse.unimelb.edu.au
Administrative Queries: owner-mercury-reviews at csse.unimelb.edu.au
Subscriptions:          mercury-reviews-request at csse.unimelb.edu.au
--------------------------------------------------------------------------