On Wednesday 30 January 2008 16:35:59 [EMAIL PROTECTED] wrote:
> Author: jkeenan
> Date: Wed Jan 30 16:35:59 2008
> New Revision: 25374
>
> Modified:
> trunk/t/manifest/02-regenerate_file.t
> trunk/t/manifest/03-regenerate_skip.t
> trunk/t/steps/init_headers-01.t
> trunk/t/steps/init_manifest-02.t
>
> Log:
> Delete copies of MANIFEST in tempdirs after testing.
> Modified: trunk/t/manifest/02-regenerate_file.t
> ========================================================================
> --- trunk/t/manifest/02-regenerate_file.t (original)
> +++ trunk/t/manifest/02-regenerate_file.t Wed Jan 30 16:35:59 2008
> @@ -40,6 +40,7 @@
> ok( !$need_for_file, "No need to regenerate $f" );
> chdir $cwd
> or croak "Unable to change back from temporary directory after
> testing";
> + unlink qq{$tdir/$f} or croak "Unable to delete file from
> tempdir"; }
I can't find anything in PBP that says "Despite the qq{} quoting form being
almost as horrifically ugly as slicing a hash reference, use it and pray your
sanity rolls are all 17+." Why mix string delimiters on the same line?
(Ironically, PBP p. 51 suggests that the second string on that line needs
single quotes.)
That concatenation of paths probably ought to use File::Spec as well, if
portability is important.
Of course, File::Temp (already in use) can create a temporary directory and
and clean it up when the program ends... when used exactly as this test file
already does, in which case I don't understand the purpose of this patch at
all.
-- c
