initializing types.ok (was Re: new parser)

On Tue, 11 Feb 2003, Stephen Schaefer wrote:

> So I deleted the echo-$(IPx) lines.  Next:
>
> I've got a chicken and egg problem: if types.ok isn't there, ischeck
> insists it be retrieved from CVS.  If it is there, it dies:
> "oracle_8.1.7.3_server is an invalid variable name.".  I suppose I'll
> hand edit the types.ok file, then generate the checksum by nefarious use
> of perl -d, but it shouldn't be like this...

Okay, I just checked a new version of ischeck in; it's not tested, but it
was a very small change, so it should work just fine.  It just initializes
types.ok to a checksummed copy of types if types.ok doesn't exist.

Do you actually have 'oracle_8.1.7.3_server' as a variable name?  If so,
then yeah, that's a problem, because you can't have '.' in variables, but
I expect (since that looks like one of my types) that this just a type and
is causing a problem somewhere.

Try out the new ischeck, and let me know what happens.

In regards to manually setting the checksum, though, yeah, that's
necessary sometimes.  I've had to do it to reorder things, for instance.
I should probably add a flag to ischeck to allow the checksum to be
ignored...

Of course, I've also thought about just getting rid of the checksums and
the timestamps, but...

So what's your status now?

-- 
    The Washington Bullets are changing their name.  The owners no
    longer want their team's name to be associated with crime.  So from
    now on the team will be known as The Bullets.
                -- Paul Harvey, quoting Argus Hamilton


-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com