Bug #109527 for IO-HTML: bug in io::html which ends up in eating all ram when reading a line from the handle

Mon Nov 23 14:58:26 2015 michael.adamcik [...] pro-pos.at - Ticket created

Subject:	bug in io::html which ends up in eating all ram when reading a line from the handle
Date:	Mon, 23 Nov 2015 20:57:58 +0100
To:	bug-IO-HTML [...] rt.cpan.org
From:	Michael Adamcik <michael.adamcik [...] pro-pos.at>

hey there! found a bug in io::html when using "html_file()" which results in eating all ram when reading a line from the handle. i attach the html and a small script which produce this behaviour. i cant explain why it does is with exact this file, and at exact this line.. you will see (hopefully) im using perl v5.20.2 libio-html-perl 1.001-1 on debian stable/testing best regards, michael

Message body is not shown because sender requested not to inline it.

Mon Nov 23 19:37:52 2015 cjm [...] cpan.org - Correspondence added

This does appear to be a bug, but it's not in IO::HTML. I get the exact same results with this program: #!/usr/bin/perl -w use strict; binmode(STDOUT, ":utf8"); open (my $fh, '<:encoding(ISO-2022-JP)', "bug.html"); while (my $l=<$fh>){ print "L: " . $l . "\n"; } The only way IO::HTML is involved is that it correctly detects that the file claims to use ISO-2022-JP and applies that encoding to the filehandle. I suggest reporting this as a bug in Encode (or possibly the Perl core, but I'd try Encode first). Thanks for using IO::HTML, though :-)

Mon Nov 23 19:37:53 2015 The RT System itself - Status changed from 'new' to 'open'

Mon Nov 23 19:37:59 2015 cjm [...] cpan.org - Status changed from 'open' to 'rejected'

Mon Nov 23 20:32:37 2015 michael.adamcik [...] pro-pos.at - Correspondence added

Subject:	Re: [rt.cpan.org #109527] bug in io::html which ends up in eating all ram when reading a line from the handle
Date:	Tue, 24 Nov 2015 02:32:17 +0100
To:	bug-IO-HTML [...] rt.cpan.org
From:	Michael Adamcik <michael.adamcik [...] pro-pos.at>

yes, i figured out the same... allready reported it to perl core. io::html is wonderful if you deal with many html sites ;) On 2015-11-24 01:37, Christopher J. Madsen via RT wrote: Show quoted text

> <URL: https://rt.cpan.org/Ticket/Display.html?id=109527 > > > This does appear to be a bug, but it's not in IO::HTML. I get the exact same results with this program: > > #!/usr/bin/perl -w > use strict; > > binmode(STDOUT, ":utf8"); > > open (my $fh, '<:encoding(ISO-2022-JP)', "bug.html"); > > while (my $l=<$fh>){ > print "L: " . $l . "\n"; > } > > The only way IO::HTML is involved is that it correctly detects that the file claims to use ISO-2022-JP and applies that encoding to the filehandle. > > I suggest reporting this as a bug in Encode (or possibly the Perl core, but I'd try Encode first). > > Thanks for using IO::HTML, though :-) >