Hi every one here, as you can guess, my Name is Benjamin Birkenhake. I live in two of germanys most beautifull cities: Bielfeld and Hamburg. In Bielefeld I spend the lovely weekends and in Hamburg I work for a german Newspaper. We are currently evaluating Drupal as Community-Framework.
As we are not planing to replace our major CMS with Drupal, Drupal is only an addition to our Mainsite. So my main interrestes are API and "cultural" Interfaces to other sites.
Another point of interrest is how to make use of huge masses of old content, which we still keep online, but which are not used by our users. We are aiming to implement Semantic Web technologies, to improve the relations between, our Articles and make them thus mor usefull. One part of this strategy contains user and community-centered approaches.
Comments
On-demand republishing
One concept I've been playing with in Drupal is the ability to import content when requested.
For example, one can write a module that has predictive behavior. Take this example.
1) In the old CMS, you have content.
2) Each content item has a unique ID and URL.
3) Write a drupal module that handles requests for /cms/IDSTRING
4) When that path is invoked, reach out and grab the old content via HTTP or XML transfer.
5) Create a new Drupal node with the old content.
5a) Teach the module to correlate cms/IDSTRING to the new /node/NID.
5b) Publish your new Drupal node.
6) Enjoy interactivity with your archive content.
This approach works best when importing from a database-driven system or API. Then use XML parsing to create the nodes.
--
http://ken.therickards.com/
http://savannahnow.com/user/2
http://blufftontoday.com/user/3
--
http://ken.therickards.com/
Jawollja
That's exactly what we have planed.
As our Main CMS ist xml file based, each article is already available online as an XML-Resource,
so all we need to do is to define a content type (node type), that saves the url and "imports" the article live
every time it is requested.
Instances of the content type (node type?) may be created when the first user want's to add a comment to that article.
A good way to do this, would be to create something like a virtual-node, which contains only the URL, as descirbed, and some kind of rule, how to transform the XML-Format the URL is stored in into HTML. Than you could even create Subclasses of that type simply using different xml2html transformations (which is, what XSLT was made for and by the way: is there some point in Drupal where XSLT ist used?)
Wee keep thinking about this.
Thanks alot!
--
anmut und demut
--
anmut und demut
I don't know
I'm not technical enough to answer that. In my implementations, I simply use Curl calls to retrieve XML documents, then parse them and create nodes as needed, storing the extra node data in a separate node type and table.
Perhaps you mean http://api.drupal.org/api/5/function/hook_xmlrpc
--
http://ken.therickards.com/
http://savannahnow.com/user/2
http://blufftontoday.com/user/3
--
http://ken.therickards.com/
Hum..
I'm not that familiar with Drupal yet. Waht I describes was more or less the way, I would have made it, if I had used "regular" OO-PHP.
The advantage of just storing the URL in a node and getting the xml-content each time the node is accessed, is that updates of the URL appear in the node immideatelly. That's quite important for us, as Articles may change quite often.
How do you parse your XML-Documents? SimpleXML?
--
anmut und demut
--
anmut und demut
Yes
Yes, SimpleXML, because it works on my tech level :-).
When I worry about updates from the external, I use Drupal's cache mechanism to save the XML data (or the parsed output) to the local system. Then I update daily or weekly, depending on the data.
http://api.drupal.org/apis/5/cache
It is fairly trivial to store the external, uneditable data separately from Drupal-accessible data.
--
http://ken.therickards.com/
http://savannahnow.com/user/2
http://blufftontoday.com/user/3
--
http://ken.therickards.com/