[tap] Octothorpe/hash "#" in descriptions?

[Salve J Nilsen <[e-mail]>]

This is Re: the current version of the draft, 
<http://github.com/gaurav/test-anything-protocol/blob/master/draft-vaidya-test-anything-protocol-00.xml>

Ovid said:
>
> Why are '#' characters not allowed in descriptions?
>
>  my $builder = Test::Builder->new;
>  $builder->todo_start('hash mark filtering not implemented')
>  unlike $result, qr/#/, 'Result does not contain #';
>  $builder->todo_end;
>
> Sometimes that's something may want in a description and it reads 
> naturally.  I'd like to see the hash mark (octothorpe, whatever), 
> escaped:
>
>  ok 3 - Result does not contain \# # TODO hash mark filtering unimplemented

FWIW, I think escaping is completely reasonable. It's already 
implemented like this in Test::Builder and Test::Harness, and could at 
least be considered "established practice" in the Perl community. (Not 
sure since when though.)

We have four options, I think.

1) Keep disallowing # in the description (breaking with established
    practice, but perhaps gain a small simplification in the draft
    text.)

2) Allow escaped # in the description, using the \# syntax already
    established in the Perl community's most used testing modules.

3) Allow escaped # in the description, using some other syntax,
    e.g. the printf(3)-like "##" way with double-characters ("%%" to
    represent a literal "%".) This breaks with both established
    practice and doesn't seem to give an obvious benefit otherwise.


I'm for 2).


- Salve

-- 
#!/usr/bin/perl
sub AUTOLOAD{$AUTOLOAD=~/.*::(\d+)/;seek(DATA,$1,0);print#  Salve Joshua Nilsen
getc DATA}$"="'};&{'";@_=unpack("C*",unpack("u*",':4@,$'.#     <[e-mail]>
'2!--"5-(50P%$PL,!0X354UC-PP%/0\`'."\n"));eval "&{'@_'}";   __END__ is near! :)