X-Git-Url: https://git.camperquake.de/gitweb.cgi?p=quotesite.git;a=blobdiff_plain;f=quotesite%2FTwitterGrabber.pm;h=286b96260a8c3b3b1cd6be4b25ef1b0951293f09;hp=6fab0cd21ccc3dc08b8a62465dd4606a1a8e9f0b;hb=54cdc91228ff95d3b0407baa00adc215272c24a0;hpb=0711e74b084cc99d8f3508d2aac4daf7f933e142 diff --git a/quotesite/TwitterGrabber.pm b/quotesite/TwitterGrabber.pm index 6fab0cd..286b962 100644 --- a/quotesite/TwitterGrabber.pm +++ b/quotesite/TwitterGrabber.pm @@ -11,6 +11,7 @@ use GrabberBase; use LWP::Simple qw(!get); use Data::Dumper; use XML::Simple; +use HTML::Entities qw(decode_entities); use Encode; use strict; @@ -20,7 +21,7 @@ sub new { my $self = $class->SUPER::new(); $self->{'NAME'} = 'twitter.com'; - $self->{'PATTERNS'} = ['(http://twitter.com/(?:\#\!/)?[^/]+/status/(\d+))']; + $self->{'PATTERNS'} = ['(http://twitter.com/(?:\#\!/)?[^/]+/status(?:es)?/(\d+))']; bless($self, $class); $self->_prepare_parameters(); @@ -57,7 +58,7 @@ sub _parse { return undef; } - $metadata->{'CONTENT'} = $t->{'text'}; + $metadata->{'CONTENT'} = encode("utf8", decode_entities(decode("utf8", $t->{'text'}))); $metadata->{'ID'} = $t->{'user'}->{'screen_name'} . '/' . $metadata->{'ID'}; unless(defined($metadata->{'CONTENT'})) {