xmltv/grab/za tv_grab_za,1.44,1.45

Update of /cvsroot/xmltv/xmltv/grab/za
In directory sfp-cvs-1.v30.ch3.sourceforge.com:/tmp/cvs-serv26227

Modified Files:
	tv_grab_za 
Log Message:
(minor) fix badiso8859

Index: tv_grab_za
===================================================================
RCS file: /cvsroot/xmltv/xmltv/grab/za/tv_grab_za,v
retrieving revision 1.44
retrieving revision 1.45
diff -C2 -d -r1.44 -r1.45
*** tv_grab_za	11 Dec 2013 17:32:18 -0000	1.44
--- tv_grab_za	1 May 2014 19:26:55 -0000	1.45
***************
*** 84,87 ****
--- 84,88 ----
  use IO::File;
  use Digest::MD5 qw(md5 md5_hex);
+ use Encode;
  
  use POSIX qw(strftime);
***************
*** 129,132 ****
--- 130,136 ----
  my $LANG="en";
  
+ # character encoding of output file
+ my $ENCODING = 'ISO-8859-1';
+ 
  our %dstvchannelmap;
  our %mnetchannelmap;
***************
*** 338,342 ****
  	$w_args{OUTPUT} = $fh;
  }
! $w_args{encoding} = 'ISO-8859-1';
  my $writer = new XMLTV::Writer(%w_args);
  $writer->start($HEAD);
--- 342,346 ----
  	$w_args{OUTPUT} = $fh;
  }
! $w_args{encoding} = $ENCODING;
  my $writer = new XMLTV::Writer(%w_args);
  $writer->start($HEAD);
***************
*** 411,423 ****
  }
  
  sub tidy( $ ) {
! 	for (my $tmp = shift) {
! 		if (not defined $tmp) { return };
! 		tr/\t\205/ /d;
! 		s/([^\012\015\040-\176\240-\377]+)//g;
! 		return $_;
! 	}
  }
  
  # The URI to get listings for a given channel.
  sub dstv_channel_uri( $ ) {			# DEPRECATED
--- 415,439 ----
  }
  
+ #sub tidy( $ ) {
+ #	for (my $tmp = shift) {
+ #		if (not defined $tmp) { return };
+ #		tr/\t\205/ /d;
+ #		s/([^\012\015\040-\176\240-\377]+)//g;
+ #		return $_;
+ #	}
+ #}
+ 
+ # Remove bad chars from an element
  sub tidy( $ ) {
!     my $html = shift;
!     return $html if !defined $html;	
!     $html =~ s/(\s)\xA0/$1/g;    # replace 'space- ' with 'space'
!     $html =~ s/\xA0/ /g;         # replace any remaining   with space
!     $html =~ s/\xAD//g;          # delete soft hyphens
!     return $html;
  }
  
+ 	
+ 
  # The URI to get listings for a given channel.
  sub dstv_channel_uri( $ ) {			# DEPRECATED
***************
*** 497,501 ****
  	}
  
! 	if ($desc =~ /^'([^\.]+)'\.\s+(.+)/) {
  		$subtitle = $1;
  		$desc = $2;
--- 513,518 ----
  	}
  
! 	if ($desc =~ /^'([^\.]+)'\.\s+(.+)/) {	   # don't know why this excludes '.'
!                                                #   - means it fails to detect 'S1/E13 - ...A Better Place.'.
  		$subtitle = $1;
  		$desc = $2;
***************
*** 539,543 ****
  	if ($desc =~ /(.*) \((\d{4})\)\s*([^\.]+)\.?\s*$/) {
  		$year = $2;
! 		$director = $3;
  		$desc = $1;
  		t "desc = $desc\n";
--- 556,560 ----
  	if ($desc =~ /(.*) \((\d{4})\)\s*([^\.]+)\.?\s*$/) {
  		$year = $2;
! 		$director = encode($ENCODING, $3);
  		$desc = $1;
  		t "desc = $desc\n";
***************
*** 562,566 ****
  			$actors = [];
  			foreach my $a (@a) {
! 				push @$actors, $a;
  			}
  		}
--- 579,583 ----
  			$actors = [];
  			foreach my $a (@a) {
! 				push @$actors, encode($ENCODING, $a);
  			}
  		}
***************
*** 574,577 ****
--- 591,599 ----
  
  	$desc = "No description available" if ($desc eq "");
+ 	
+ 	# Encode into output charset 
+ 	$desc     = encode($ENCODING, tidy($desc));
+ 	$title    = encode($ENCODING, tidy($title));
+ 	$subtitle = encode($ENCODING, tidy($subtitle));
  
  	$r->{title} = [[$title]];


------------------------------------------------------------------------------
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.  Get 
unparalleled scalability from the best Selenium testing platform available.
Simple to use. Nothing to install. Get started now for free."
http://p.sf.net/sfu/SauceLabs