Front page | perl.perl5.porters |
Postings from October 2016
On 10/10/2016 05:09 PM, James E Keenan wrote:
> On 10/10/2016 12:59 PM, pali@cpan.org wrote:
>> On Thursday 29 September 2016 21:22:26 pali@cpan.org wrote:
>>> Hi!
> [snip]
>
>>
>> Hi! I prepared patch for Encode which should fix those problems with
>> readonly, magic, COW, ... scalars. Can you look at it? It is correct now?
>>
>
> Since I'm not very familiar with Encode per se, I decided that, in
> reviewing this patch, I would focus on the tests you have added.
>
> I cloned Encode from its github repository
> (https://github.com/dankogai/p5-encode), built it locally and then ran
> it through Devel-Cover to see how thoroughly its test suite exercised
> the source code. That means I ran:
>
> #####
> HARNESS_PERL_SWITCHES=-MDevel::Cover make test
> cover
> #####
>
> There's clearly room for improvement, as this excerpt from the coverage
> analysis suggests:
>
> #####
> ------------------------------ ------ ------ ------ ------
> File stmt bran cond sub
> ------------------------------ ------ ------ ------ ------
> blib/lib/Encode.pm 70.5 57.0 80.0 74.1
> ...
> blib/lib/encoding.pm 61.8 36.3 30.0 92.8
> #####
>
> But at least we have a baseline.
>
> I then created a branch and used 'git am' to apply Pali's patch, which
> applied cleanly. I then rebuilt Encode and ran 'make test', wherein all
> tests PASSED.
>
> I then called:
>
> #####
> cover -delete
> HARNESS_PERL_SWITCHES=-MDevel::Cover make test
> #####
>
> ... whereat I was surprised to get test failures (attached as
> 'harness.pali.txt').
>
> Can anyone see why the code revisions in the patch and Devel::Cover do
> not play nice with each other?
>
> #####
> $ perl -v | head -2 | tail -1
> This is perl 5, version 24, subversion 0 (v5.24.0) built for x86_64-linux
> $ perl -MDevel::Cover -E 'say $Devel::Cover::VERSION' 2> /dev/null | ack
> '^\d\.\d+$'
> #####
>
Some data gleaned from throwing in some debugging statements into
t/Encode.t near each of the two test failures: under the harness we're
gagging on these characters:
\x{100}
\316
Here are the individual tests failing only under the harness:
"\x{100}" =~ /(.)/;
ok( is_utf8($1)); # ID 20011127.151
...
my $latin1 = find_encoding('latin1');
my $orig = "\316";
$orig =~ /(.)/;
is $latin1->encode($1), $orig, '[cpan #115168] passing magic regex
globals to encode';
Thread Previous
|
Thread Next