FW: cvs commit: hugs98/lib IO.hs Prelude.hs

[Simon Peyton-Jones]

Title: Message

Folks

 

I plan to push out a new draft H98 report this week.  Here is one suggested

clarification from Sigbjorn, arising from recent uncertainty about the true

meaning of getLine.

 

I'll incorporate it unless anyone yells.

 

Simon

 

-----Original Message-----
From: Simon Marlow
Sent: 05 June 2002 10:27
To: Sigbjorn Finne; Simon Peyton-Jones
Subject: RE: cvs commit: hugs98/lib IO.hs Prelude.hs

Agree with everything Sigbjorn says below.  Also: change the definition of getLine (section A.3) to be

 

   getLine = hGetLine stdin

 

Cheers,

    Simon

-----Original Message-----
From: Sigbjorn Finne [mailto:[EMAIL PROTECTED]]
Sent: 04 June 2002 16:17
To: Simon Peyton-Jones
Cc: Simon Marlow
Subject: Re: cvs commit: hugs98/lib IO.hs Prelude.hs

How about the following:

 

* change the Report (7.1) to say something along the lines of:

 

   getChar raises an exception (Section 7.3) on end-of-file; a

   predicate isEOFError that identifies this exception is defined

   in the IO library. getLine raises an end-of-file exception

   under the same conditions as hGetLine in the IO library.

 

* in the Library Report, say when hGetLine raises an exception:

 

  Computation hGetLine hdl reads a line from the file or

  channel managed by hdl. The Prelude's getLine is a shorthand

  for hGetLine stdin.

 

   Error reporting: the hWaitForInput, hReady and hGetChar

  computations may fail with: isEOFError if the end of file has

   been reached.

 

   hGetLine will also fail with isEOFError if end-of-file is encountered

   when reading the first character of the line. If hGetLine encounters

   end-of-file at any other point while reading in a line, it is treated as

   a line terminator and the (partial) line is returned.

 

Actually, I'd prefer if 11.7.1 only dealt with hWaitForInput and hReady

(as they're the "checking for input" operations). A second section,

named "Reading input", would then be concerned with hGetChar and

hGetLine.

 

 

--sigbjorn

 

 

----- Original Message -----

From: "Simon Peyton-Jones" <[EMAIL PROTECTED]>

To: "Sigbjorn Finne" <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>

Sent: Tuesday, June 04, 2002 04:03

Subject: RE: cvs commit: hugs98/lib IO.hs Prelude.hs

> Should I change the Report wording to be clearer?
>
> Simon
>
> | -----Original Message-----
> | From: Sigbjorn Finne [mailto:[EMAIL PROTECTED]]
> | Sent: 03 June 2002 15:49
> | To: [EMAIL PROTECTED]
> | Subject: cvs commit: hugs98/lib IO.hs Prelude.hs
> |
> |
> | sof         2002/06/03 07:49:12 PDT
> |
> |   Modified files:
> |     lib                  IO.hs Prelude.hs
> |   Log:
> |   Interpret Report wording for Prelude.getLine and IO.hGetLine
> |   to mean the Right Thing. If EOF is encountered for a partial
> |   line (>= 1 chars), interpret this as '\n'.
> |  
> |   Revision  Changes    Path
> |   1.4       +25 -8     hugs98/lib/IO.hs
> |   1.25      +23 -5     hugs98/lib/Prelude.hs
> | _______________________________________________
> | Cvs-hugs mailing list
> | [EMAIL PROTECTED] http://www.haskell.org/mailman/listinfo/cvs-hugs
> |