Re: perl6 tests

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]