Hi,
I've been working on getting my html code XHTML compliant and one of the issues I had was with the & (ampersand) sign that is used in GT scripts to generate code for links in dynamic mode/pages.
Pages are not validated if the use the ampersand character "&" but are if they use the &
So I went hunting for the code that generates the strings that are generated and found this :
...
$original .= $IN->escape($_) . "=" . $IN->escape($in_hash->{$_}) . "&";
...
# Fix up the CGI references.
...
$output .= "&$key=" . $IN->escape($in_hash->{$key});
...
in the Links.pm file
Now I replace it with the below which was a bit complicated since the W3C requirements include both an "&" and a ";" (semicolon). Well first of all it works and produces the correct & in the string but I was wondering whether this was a really bad thing to do. Not that I like doing bad things but I thought it may have some adverse affect that I hadn't thought about.
...
$original .= $IN->escape($_) . "=" . $IN->escape($in_hash->{$_}) . "&|";
...
# Fix up the CGI references.
...
$output .= "&$key=" . $IN->escape($in_hash->{$key});
...
Obviously the good thing is that it gets me closer to validating my code
I also wanted to share this with the forum IF someone who knows more than I do about Perl (which is most of the forum members) can say whether it is not a good idea, a part from the "Don't hack core code" alert.
And BTW I have not got a clue why it needs to be "&|" in the first case and "&$key=" in the second but with this code it validates and having the "|" in the second has the actual "|" sign appear in the URL and the d=1 is repeated twice. It's been a while since I coded my own perl but I remembered using this and it worked but it may not be working as efficiently as it could.
Thanks, John
Significant Media
I've been working on getting my html code XHTML compliant and one of the issues I had was with the & (ampersand) sign that is used in GT scripts to generate code for links in dynamic mode/pages.
Pages are not validated if the use the ampersand character "&" but are if they use the &
So I went hunting for the code that generates the strings that are generated and found this :
Code:
# Build a query string. ...
$original .= $IN->escape($_) . "=" . $IN->escape($in_hash->{$_}) . "&";
...
# Fix up the CGI references.
...
$output .= "&$key=" . $IN->escape($in_hash->{$key});
...
in the Links.pm file
Now I replace it with the below which was a bit complicated since the W3C requirements include both an "&" and a ";" (semicolon). Well first of all it works and produces the correct & in the string but I was wondering whether this was a really bad thing to do. Not that I like doing bad things but I thought it may have some adverse affect that I hadn't thought about.
Code:
# Build a query string. ...
$original .= $IN->escape($_) . "=" . $IN->escape($in_hash->{$_}) . "&|";
...
# Fix up the CGI references.
...
$output .= "&$key=" . $IN->escape($in_hash->{$key});
...
Obviously the good thing is that it gets me closer to validating my code
I also wanted to share this with the forum IF someone who knows more than I do about Perl (which is most of the forum members) can say whether it is not a good idea, a part from the "Don't hack core code" alert.
And BTW I have not got a clue why it needs to be "&|" in the first case and "&$key=" in the second but with this code it validates and having the "|" in the second has the actual "|" sign appear in the URL and the d=1 is repeated twice. It's been a while since I coded my own perl but I remembered using this and it worked but it may not be working as efficiently as it could.
Thanks, John
Significant Media
