Re: df_readline() question

[Ethan M. <merritt@u.washington.edu>]

On Monday 02 November 2009 08:06:04 Philipp K. Janert wrote:
> 
> Comments at the end.
> 
> On Sunday 01 November 2009 10:38:04 pm Ethan Merritt (sfeam) wrote:
> > On Sunday 01 November 2009, Philipp K. Janert wrote:
> > > > Simple test: feed it a file junk.dat:
> > > >  1
> > > >  2
> > > >  junk
> > > >  3
> > > >  4
> > > >
> > > > gnuplot> plot 'junk.dat' with lp
> > > >               ^
> > > >          Bad data on line 3
> > > > gnuplot>
> > >
> > > Does plot do anything special here?
> > >
> > > I have a situation (stolen from the fit function),
> > > which basically says:
> > >
> > > while( (i = readline(...)) != EOF ) {
> > > 	...
> > >
> > > And now I feed it your data file and it does NOT
> > > return either DF_MISSING or DF_UNDEFINED!
> >
> > No, it returns 0.  But that's not EOF, so it should
> > be catchable.
> 
> No, it doesn't. It does not return anything for the
> line that contains "junk", when run with "using 1".
> 
> It does return -2 (DF_UNDEFINED) for the line
> with "junk", but only when run with "using ($1)".

Hence the entry on my TODO list that I quoted earlier.
You could evaluate the following simple-minded patch:

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
--- gnuplot/src/datafile.c	2009-10-10 11:43:30.000000000 -0700
+++ gnuplot-cvs/src/datafile.c	2009-11-02 08:53:58.000000000 -0800
@@ -1870,7 +1870,7 @@ df_readascii(double v[], int max)
 			/* line bad only if user explicitly asked
 			 * for this column */
 			if (df_no_use_specs)
-			    line_okay = 0;
+			    return DF_UNDEFINED;
 			break;      /* return or ignore depending on line_okay */
 		    }
 		}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%5




-- 
Ethan A Merritt