xmltv/grab Get_nice.pm,1.27,1.28
Geoff <[email protected]>
| Newsgroups | gmane.comp.tv.xmltv.cvs |
|---|---|
| Message-ID | <[email protected]> |
Update of /cvsroot/xmltv/xmltv/grab
In directory sfp-cvs-1.v30.ch3.sourceforge.com:/tmp/cvs-serv16359
Modified Files:
Get_nice.pm
Log Message:
(minor) allow user-specified decode option (see v1.25)
Index: Get_nice.pm
===================================================================
RCS file: /cvsroot/xmltv/xmltv/grab/Get_nice.pm,v
retrieving revision 1.27
retrieving revision 1.28
diff -C2 -d -r1.27 -r1.28
*** Get_nice.pm 21 Jan 2014 17:46:53 -0000 1.27
--- Get_nice.pm 23 Jan 2014 09:25:49 -0000 1.28
***************
*** 32,36 ****
# use version number for feature detection:
# 0.005065 : new methods get_nice_json(), get_nice_xml()
! # 0.005065 : add utf8 decode option to get_nice_tree()
# 0.005065 : expose the LWP response object ($Response)
our $VERSION = 0.005065;
--- 32,36 ----
# use version number for feature detection:
# 0.005065 : new methods get_nice_json(), get_nice_xml()
! # 0.005065 : add decode option to get_nice_tree()
# 0.005065 : expose the LWP response object ($Response)
our $VERSION = 0.005065;
***************
*** 68,82 ****
# i) a function to put the page data through (eg, to clean up bad
# characters) before parsing.
! # ii) convert incoming UTF-8 to UNICODE when set to 1
#
sub get_nice_tree( $;$$ ) {
! my ($uri, $filter, $utf8) = @_;
require HTML::TreeBuilder;
my $content = get_nice $uri;
$content = $filter->($content) if $filter;
! if ($utf8) {
# (note: HTML::Parser->utf8_mode is only available Perl 5.8)
require Encode; import Encode;
! $content = decode('UTF-8', $content); # strict utf-8
}
my $t = HTML::TreeBuilder->new()->parse($content) or die "cannot parse content of $uri\n";
--- 68,82 ----
# i) a function to put the page data through (eg, to clean up bad
# characters) before parsing.
! # ii) convert incoming page to UNICODE using this codepage (use "UTF-8" for strict utf-8)
#
sub get_nice_tree( $;$$ ) {
! my ($uri, $filter, $codepage) = @_;
require HTML::TreeBuilder;
my $content = get_nice $uri;
$content = $filter->($content) if $filter;
! if ($codepage) {
# (note: HTML::Parser->utf8_mode is only available Perl 5.8)
require Encode; import Encode;
! $content = decode($codepage, $content);
}
my $t = HTML::TreeBuilder->new()->parse($content) or die "cannot parse content of $uri\n";
***************
*** 89,102 ****
# i) a function to put the page data through (eg, to clean up bad
# characters) before parsing.
! # ii) convert incoming UTF-8 to UNICODE when set to 1
#
sub get_nice_xml( $;$$ ) {
! my ($uri, $filter, $utf8) = @_;
require XML::Parser;
my $content = get_nice $uri;
$content = $filter->($content) if $filter;
! if ($utf8) {
require Encode; import Encode;
! $content = decode('UTF-8', $content); # strict utf-8
}
my $t = XML::Parser->new(Style => 'Tree')->parse($content) or die "cannot parse content of $uri\n";
--- 89,102 ----
# i) a function to put the page data through (eg, to clean up bad
# characters) before parsing.
! # ii) convert incoming page to UNICODE using this codepage (use "UTF-8" for strict utf-8)
#
sub get_nice_xml( $;$$ ) {
! my ($uri, $filter, $codepage) = @_;
require XML::Parser;
my $content = get_nice $uri;
$content = $filter->($content) if $filter;
! if ($codepage) {
require Encode; import Encode;
! $content = decode($codepage, $content);
}
my $t = XML::Parser->new(Style => 'Tree')->parse($content) or die "cannot parse content of $uri\n";
***************
*** 108,112 ****
# i) a function to put the page data through (eg, to clean up bad
# characters) before parsing.
! # ii) convert incoming UTF-8 to UNICODE when set to 1
#
sub get_nice_json( $;$$ ) {
--- 108,112 ----
# i) a function to put the page data through (eg, to clean up bad
# characters) before parsing.
! # ii) convert incoming UTF-8 to UNICODE
#
sub get_nice_json( $;$$ ) {
***************
*** 115,119 ****
my $content = get_nice $uri;
$content = $filter->($content) if $filter;
! $utf8 = 0 if !defined $utf8; #(not necessary but safer)
my $t = JSON::PP->new()->utf8($utf8)->decode($content) or die "cannot parse content of $uri\n";
return $t;
--- 115,119 ----
my $content = get_nice $uri;
$content = $filter->($content) if $filter;
! $utf8 = defined $utf8 ? 1 : 0;
my $t = JSON::PP->new()->utf8($utf8)->decode($content) or die "cannot parse content of $uri\n";
return $t;
------------------------------------------------------------------------------
CenturyLink Cloud: The Leader in Enterprise Cloud Services.
Learn Why More Businesses Are Choosing CenturyLink Cloud For
Critical Workloads, Development Environments & Everything In Between.
Get a Quote or Start a Free Trial Today.
http://pubads.g.doubleclick.net/gampad/clk?id=119420431&iu=/4140/ostg.clktrk