[Gs-code-review] MetricsCount broken in Ghostscript?

[Raph Levien]

On Fri, Feb 08, 2002 at 08:40:46PM -0800, Brian Stell wrote:
> Per L. Peter Deutsch's suggestion I am submitting this to
> the gs-code-review at ghostscript.com.

Thanks very much. This is indeed the correct process.

> Would it be possible for someone to apply this fix?

I'll be happy to, once I feel I understand it  fully :)

I agree with your analysis and that your patch is an improvement.
What I don't understand is the first "if" in z11_get_metrics.
Particularly, I don't see why the "wmode > skip >> 2" shouldn't be
"wmode >= (skip >> 2)". It seems to me that the predicate will be
false in the case when skip is 0 and wmode is 0, as well as skip = 4
and wmode = 1, so control will pass down to the code which tries to
read the metrics from pmetrics.

Brian, could you please send your test files? I'd like to get a
better idea of when the z11_get_metrics code gets exercised.

Thanks,

Raph