xmltv/grab/nl tv_grab_nl,1.77,1.78

Update of /cvsroot/xmltv/xmltv/grab/nl
In directory sfp-cvs-1.v30.ch3.sourceforge.com:/tmp/cvs-serv18522

Modified Files:
	tv_grab_nl 
Log Message:
(minor) fix badiso8859 warning (faster method)
(minor) extend web page cache to 4 hours

Index: tv_grab_nl
===================================================================
RCS file: /cvsroot/xmltv/xmltv/grab/nl/tv_grab_nl,v
retrieving revision 1.77
retrieving revision 1.78
diff -C2 -d -r1.77 -r1.78
*** tv_grab_nl	30 Apr 2014 18:28:23 -0000	1.77
--- tv_grab_nl	1 May 2014 06:19:26 -0000	1.78
***************
*** 183,187 ****
  HTTP::Cache::Transparent::init( { 
      BasePath => get_default_cachedir(),
!     NoUpdate => 60*60,          # cache time in seconds
      MaxAge => 24,               # flush time in hours
      Verbose => $opt_debug,
--- 183,187 ----
  HTTP::Cache::Transparent::init( { 
      BasePath => get_default_cachedir(),
!     NoUpdate => 4*3600,         # cache time in seconds
      MaxAge => 24,               # flush time in hours
      Verbose => $opt_debug,
***************
*** 413,417 ****
  
  #my $warned_bad_chars;
! #sub tidy( $ ) {
  #    for (my $tmp = shift) {
  #    	tr/\221\222/''/;
--- 413,417 ----
  
  #my $warned_bad_chars;
! #sub tidypage( $ ) {
  #    for (my $tmp = shift) {
  #    	tr/\221\222/''/;
***************
*** 423,432 ****
  #}
  
! # Remove bad chars
  sub tidy( $ ) {
!     $_[0] =~ s/(\s) /$1/og;    # replace 'space- ' with 'space'
!     $_[0] =~ s/ / /og;         # replace any remaining   with space
!     $_[0] =~ s/­//og;           # delete soft hyphens
!     return $_[0];
  }
  
--- 423,442 ----
  #}
  
! # Remove bad chars from entire page
! sub tidypage( $ ) {
!     my $html = shift;
!     $html =~ s/(\s) /$1/og;    # replace 'space- ' with 'space'
!     $html =~ s/ / /og;         # replace any remaining   with space
!     $html =~ s/­//og;           # delete soft hyphens
!     return $html;
! }
! 
! # Remove bad chars from an element
  sub tidy( $ ) {
!     my $html = shift;
!     $html =~ s/(\s)\xA0/$1/og;    # replace 'space- ' with 'space'
!     $html =~ s/\xA0/ /og;         # replace any remaining   with space
!     $html =~ s/\xAD//og;           # delete soft hyphens
!     return $html;
  }
  
***************
*** 684,692 ****
      # Get HTML::TreeBuilder object.
      my $t;
!     # why "tidy"? what's wrong with them?
!     #       eval { $t = get_nice_tree $url, \&tidy };
      #
      # page is "charset=iso-8859-1"
!     $t = get_nice_tree($url, \&tidy);
      if ($@) {
      	warn "error getting/parsing $url: $@";
--- 694,701 ----
      # Get HTML::TreeBuilder object.
      my $t;
!     #       eval { $t = get_nice_tree $url, \&tidypage };
      #
      # page is "charset=iso-8859-1"
!     $t = get_nice_tree($url);
      if ($@) {
      	warn "error getting/parsing $url: $@";
***************
*** 706,710 ****
                  $_s->detach();
              }
!             $$p->{'desc'} = trim( $_p->as_text() );
              $_p->detach();
          }
--- 715,719 ----
                  $_s->detach();
              }
!             $$p->{'desc'} = trim( tidy( $_p->as_text() ) );
              $_p->detach();
          }
***************
*** 714,718 ****
              $html =~ s/(<br>|<br \/>)/ /sg;    # insert spaces between paragraphs
              $html =~ s/<.+?>//sg;              # dump the html
!             $$p->{'desc_add'} = trim( $html );
          }
      }
--- 723,727 ----
              $html =~ s/(<br>|<br \/>)/ /sg;    # insert spaces between paragraphs
              $html =~ s/<.+?>//sg;              # dump the html
!             $$p->{'desc_add'} = trim( tidy( $html ) );
          }
      }
***************
*** 736,740 ****
              #
              foreach my $_li (@_li) {
!                 my $li = $_li->as_text();
                  
                  if ( $li =~ /Genre:(.*)/ ) {
--- 745,749 ----
              #
              foreach my $_li (@_li) {
!                 my $li = tidy( $_li->as_text() );
                  
                  if ( $li =~ /Genre:(.*)/ ) {
***************
*** 761,765 ****
              #
              foreach my $_li (@_li) {
!                 my $li = $_li->as_text();
  
                  if ( $li =~ /Jaar.*:(.*)/ ) {
--- 770,774 ----
              #
              foreach my $_li (@_li) {
!                 my $li = tidy( $_li->as_text() );
  
                  if ( $li =~ /Jaar.*:(.*)/ ) {


------------------------------------------------------------------------------
"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