Re: Tentative Term::ReadLine patch/test

[Richard Clamp]

On Sat, Dec 08, 2001 at 09:48:24PM -0700, chromatic wrote:
> On Sat, 08 Dec 2001 12:04:27 -0700, Richard Clamp wrote:
>
> When I was writing tests for libnet, Graham recommended not to chdir() into 't'
> for testing outside of the core.  I'd probably move that inside the if block.

This is a piece of cargo I scavenged from lib/Term/Cap.t, so maybe
that could take some poking at too.  [cc to jns as I think this may
well be his domain]

In a way it's a little academic since Term::ReadLine doesn't live a
life outside of the core, but I've tweaked it up anyhow, and left the
test for PERL_CORE in to save someone the trouble of adding it later.

> +can_ok($t,   qw( ReadLine readline addhistory IN OUT MinLine
> + findConsole Attribs Features new ));
> 
> I'd loop over these:
> 
> foreach my $method (qw( ReadLine readline addhistory IN OUT MinLine
>   findConsole Attribs Features new) ) {
> can_ok( $t, $method );
> }
> 
> It seems easier to find the one failure if each method has a separate test.
> (My assumption is that can_ok() does all in the list, looking at the test plan
> earlier.)

Good point, I forget that most people aren't going to be running tests
against the core verbosely.

That way it also looks like I wrote a bunch more tests :)

> Neither is a big deal, just stylistic things.  Thanks for sending it along!

Glad to, thanks for the input.

-- 
Richard Clamp <[EMAIL PROTECTED]>

Re: Tentative Term::ReadLine patch/test

[chromatic]

On Sat, 08 Dec 2001 12:04:27 -0700, Richard Clamp wrote:

> Does it look like I'm heading down the right road with this?  If so let me
> know and I'll submit it to p5p for applying to blead.
 

+BEGIN {
+chdir 't' if -d 't';
+
+if ( $ENV{PERL_CORE} ) {
+@INC = '../lib';
+}
+}

When I was writing tests for libnet, Graham recommended not to chdir() into 't'
for testing outside of the core.  I'd probably move that inside the if block.


+can_ok($t,   qw( ReadLine readline addhistory IN OUT MinLine
+ findConsole Attribs Features new ));

I'd loop over these:

foreach my $method (qw( ReadLine readline addhistory IN OUT MinLine
findConsole Attribs Features new) ) {
can_ok( $t, $method );
}

It seems easier to find the one failure if each method has a separate test.
(My assumption is that can_ok() does all in the list, looking at the test plan
earlier.)

Neither is a big deal, just stylistic things.  Thanks for sending it along!

-- c