I would like to announce what I think is the first XML application using Links - XMLinks

I have changed the What's New page to a RSS (Rich Site Summary) file, i.e. a RDF/XML file, see: http://www.statskontoret.se/...erna/senaste/rss.xml (will show in MSIE5, but due to my use of Swedish characters, it might not show; look at the source to see the XML)

You can also see it via XMLTree, a XML repository:
http://www.xmltree.com/.../L/3/ResourceID/3946
where you can subscribe to it on several portals or as a HTML-newsletter. Also there, you can (on Tester) get a HTMLified output of the XML source).

How did I do this? Well, I have played with the sub build_new_page in nph_build.cgi, added a new sub to site_html_templates.pl, and XMLified the new.html template. Oh, and created a new template for the RSS/XML links, to which I refer in the modded sub. Laziness also made me turn off spanning pages (will fix that, solution for that is in Eliot and/or Bobsie thread somewhere here...)

I have described these changes in detail at:
http://www.gotzespace.dk/...linksmod/xmlinks.txt

It works. Well, sort of. It creates the RSS-file, but I am having problems with the sorting. In general in this installation of Links, I sort by ID only, and the RSS items should also do this. This is where I need help! - since this is independent of all the XML-stuff, I beg you perl-wizards for assistance -- formulated as a general mod question: How do I get the What's New pages to sort independetly of categories, and only by ID??

(For more information about (that is, links to resources) RSS and XML, see http://www.links.gotzespace.dk
-> XML -> Syndication)

Cheers
John

Hi:

Do you know of any sites using your modifications to create a backend feed. I know that within the next three to four months, we hope to be adding an array of "backends" to various segments of our site "Nonprofit-TechWorld" so that we can feed data to other sources, especially since so many sources are generously feeding data to us.

So I'm very interested in a XML implemtation that will allow us to convert data from the .cgi files to .rdf or .xml format.

Alnisa

Is this MOD still available by any chance ?
>> Nakul Goyal (SEO, Link Building Expert)
Web Site Promotion
Tips 'n' Tricks
SEO News

Here is a copy of the file, hope Mr. Gotze dosen't mind...


Leonard
aka PerlFlunkie

I've been working on this for the past couple of days, and have some changes that others might consider:

1) in site_html_templates get rid of the invalid characters (including the displayed '-' in the record dates):

sub site_html_rsslink {
# --------------------------------------------------------
# This routine is used to display what a link should look
# like in the lastlink.html webpage.
my %rec = @_;
# Set new and pop to either 1 or 0 for templates.
($rec{'isNew'} eq 'Yes') ? ($rec{'isNew'} = 1) : (delete $rec{'isNew'});
($rec{'isPopular'} eq 'Yes') ? ($rec{'isPopular'} = 1) : (delete $rec{'isPopular'});
if ($rec{'Title'}) { ($rec{'Title'}) =~ s/&/&/g; }
if ($rec{'Title'}) { ($rec{'Title'}) =~ s/</&#60;/g; }
if ($rec{'Title'}) { ($rec{'Title'}) =~ s/>/&#62;/g; }
if ($rec{'Title'}) { ($rec{'Title'}) =~ s/’/'/g; }
if ($rec{'Date'}) { ($rec{'Date'}) =~ s/-/ /g; }
return &load_template ('rsslink.html', {
#detailed_url => "$db_detailed_url/$rec{'ID'}$build_extension",
%rec,
%globals
});
}

2) in site_html_templates comment out the standard headers and deliver an xml header instead:


sub site_html_search_rssresults {
# --------------------------------------------------------
# This routine displays the search results.
#
my $term = &urlencode ($in{'query'});
# &html_print_headers;
print "Content-type: application/xml\n\n";
print &load_template ('search_rssresults.html', {
term => $term,
link_results => $link_results,
category_results => $category_results,
next => $next,
cat_hits => $cat_hits,
link_hits => $link_hits,
%in,
%globals
});
}


3) Create a new global to display the date without hyphens (another validation issue):

3a) in site_html_templates change:
%globals = (
date => &get_date,
rssdate => &get_rssdate,
time => &get_time,
db_cgi_url => $db_cgi_url,
build_root_url => $build_root_url,
site_title => $build_site_title,
css => $build_css_url,
banner => ''
);


3b) then in db_utils create a new sub (I put mine under "get_date"):


sub get_rssdate {
# --------------------------------------------------------
# Returns the current date.
#
my ($time1) = $_[0];
($time1) or ($time1 = time());
my ($sec, $min, $hour, $day, $mon, $year, $dweek, $dyear, $daylight) = localtime($time1);
my (@months) = qw!Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec!;
($day < 10) and ($day = "0$day");
$year = $year + 1900;

return "$day $months[$mon] $year";
}


4) Change the templates to RSS 2.0:

4a) search_rssresults.html changes (including new global date):

<?xml version="1.0" encoding="iso-8859-1"?>
<rss version="2.0">
<channel>
<title>Company Name</title>
<link>http://www.mysite.com</link>
<description>Feed Description</description>
<language>en</language>
<lastBuildDate><%rssdate%> <%time%> PST</lastBuildDate>
<copyright>copyright holder</copyright>
<managingEditor>my@email.com</managingEditor>
<webMaster>webmaster@email.com</webMaster>
<%if link_results%>
<%link_results%>
<%endif%>
</channel>
</rss>

4b) rsslink.html changes:

<item>
<title><%Title%></title>
<pubDate><%Date%> <%time%> PST</pubDate>
<description><%Description%></description>
<link><%URL%></link>
</item>

Hope this helps, and I'm sure all the bugs aren't worked out yet. But it's getting to be 'useable' anyway.
tx, Steve

I also installed the 'sort by date' change listed here:
http://gossamer-threads.com/..._flat&post=25659

and here's a link to John's original mod:
http://www.gotzespace.dk/books/xml.txt

Steve