Re: perl6 tests

[Joseph F. Ryan]

Tanton Gibbs wrote:

The tests look Great!  A couple of remarks
The exponential test in numeric.t
I think the last two numbers should be
-1230
and
-12300

Woops.

The Infinity test in numeric.t
Shouldn't you print $a...if not, why have it?

Double woops :)

The Binary test in radii.t
I think Larry ruled that 0B0110 was an error...it had to be 0b0110.
Also, if the number is represented in binary, won't it still print in
decimal?  You have it
printing in binary.
The same comments apply to the octal test (also, is it 0c or 0o?), and the
hex test.

I wasn't sure what the behavior was to be when writing the test,
so I figured it would be easier to do it the easy way for now and
change it later.

Also, with all tests, I thought that the number should be printed in
decimal, not the radix it
is written in.  This will also fill in the ???? in the dotted notation as it
will also print in decimal.

In string.t

Does q)hi( really DWIM??  I haven't kept up with it, so I didn't know...that
is really odd if so.

Yep, Apoc 5.  Personally, I'm not a big fan, but I'm not a language
designer.

String concat is now ~ instead of _

Hmmm, that needs to be fixed in P6C asap then, or thats going
to break a lot of tests.  I'll put it in a TODO block for now.

In interpolation.t

There is an extra ' at the end of the single quoted test's output
The print should not be in the output of
print '$x' ... it should just be $x

Woops.

In the one two three example, I thought arrays, when interpolated, did not
add spaces between their elements, such that the output should be
onetwothree
instead of
one two three

Well, in perl5, Interpolated lists, join with $", which by default is a space. So this:
"@a"

is actually:

join($",@a)

Whereas "@(@a)" should be treated like:

"".@a.""

Which simply flattens out the list elements. I might be wrong, but at least thats how
I had P6C behave :)

You seem to agree with this in the later array interpolation section where
"@(1, 2)" becomes
12
instead of
1 2

Does a list still return its last element in scalar context?  I thought I
remembered something about that changing?

I think you may be right, but thats how P6C behaves.
Time to send another bug to RT :)

Also conversion.t is just a messed up version of interpolation.t in my
version.

That's because my brilliant self saved over the file. Good thing I had a backup.

Thanks!
Tanton

Joseph F. Ryan
[EMAIL PROTECTED]