Showing posts with label work. Show all posts
Showing posts with label work. Show all posts

10 March 2009

SEEK.COM.AU RSS Feeds – Bad SEEK!

I’ve been trying to adjust the display of a SEEK.COM.AU RSS feed within SharePoint 2007. FYI – SEEK is one of the top job search sites in Australia.

I’ve found that while SEEK’s RSS feed for a job search “complies” with the RSS standard: ie it passes the Markup Validation Service at http://validator.w3.org, it will produce HTML markup that is non-compliant with the HTML standard when your XSLT displays more than one record at a time. 

   1:  <item>


   2:    <title>Network &amp; Desktop Support</title>


   3:    <description><![CDATA[


   4:        <html>


   5:          <head>


   6:          <title>Latest jobs from www.seek.com.au. Australia's #1 job site</title>


   7:          </head>


   8:          <body>


   9:            <div id="rssFeed" style="margin-top:15px; margin-bottom:8px">    


  10:            <div id="rssListedDate"><strong>Date & time listed: </strong>10/03/2009 5:22:30 PM</div>


  11:            <div id="rssJobSummary" style="margin-top:10px">


  12:              <div id="rssDes"><strong>Job ad summary: </strong><br>The Association of Professional Engineers, Scientists and Managers, Australia is seeking a suitably experienced person to work as part of a small team</div>                            


  13:              <div id="rssSalary" style="margin-top:10px">


  14:              


  15:              </div>


  16:              <div id="rssNewFeedLink" style="margin-top:18px">


  17:                <a href="http://www.seek.com.au/rss/index.ascx?catlocation=1002&catindustry=1215&Keywords=Sharepoint&RM=it&RMO=model-main&RUS=SEEKMAIN_USR&RID=1fb856cd-dbe0-4cdc-8530-d1e5300fa83a#RSSReader">Create a new feed with different search criteria</a>


  18:              </div>


  19:            <div id="rssFeedBack"><a href="http://www.seek.com.au/if.asp?loc=rsscontact">Send us feedback about these feeds.</a></div>


  20:            </div>        


  21:            </div>


  22:          </body>


  23:        </html>]]> 


  24:    </description>


  25:    <link>http://www.seek.com.au/users/viewdetails.asp?Action=jobsearch&amp;JobListAction=ViewOneAd&amp;JobSearch=true&amp;AdID=15075376&amp;SearchType=5</link>


  26:    <author>ETM Search &amp; Selection</author>


  27:    <category>Help Desk~Support|I.T. &amp; T</category>


  28:    <pubDate>Tue, 10 Mar 2009 06:22:30 GMT </pubDate>


  29:    <guid isPermaLink="false">15075376</guid>


  30:  </item>




Note within the <description> tags that the DIV elements have ID attributes! As these IDs are the same for each <item> in the RSS file, the HTML that is output will have multiple DIVS with identical ID attributes when displayed with the SharePoint RSS Viewer (multiple results are visible per page). This is not valid markup (see: http://www.w3.org/TR/REC-html40/struct/global.html#adef-id)



I really want to get rid of a couple of these DIVs, so that only the Title, Summary and Salary are listed and am trying to use jQuery to hide the DIVs I don’t want (lines 16 to 19). Initially, I tried to hide DIVs with the relevant IDs. I discovered that Firefox happily parsed the whole document and hid all the DIVs with the correct ID. Internet Explorer, on the other hand, quite correctly found the first instance of the ID and left the others alone.



So thanks to a poor design, we’re left with something unpalatable – trying to locate and hide strings instead. Yuck.

Posted by at No comments:
Labels: , ,

31 July 2008

Confluence and Sharepoint

Well the Confluence - Sharepoint Connector went from beta v.0.5 to release 1.0 just like that. Rather a bit different to open source software I've played with (the connector most definitely isn't open source). Apart from having a flaky laptop with from which to run the demo, it went OK. I think I managed not to make a goose of myself in front of my workmates and boss.

The search function works a treat but only works with MOSS. You set it up as a Shared Service and off you go. One gotcha I found was that you need to keep the internal accounts available in Confluence as the search cannot currently work with anything other than Forms Authentication. (The Atlassian folks may have fixed this by now though).

Once you set up the LDAP authentication in Confluence to enable single sign-on, you can then configure Confluence to automatically pass your credentials to Confluence (automatic single sign-on). Doing this means that you lose access to the internally configured accounts. I quickly realised this as I hadn't configured my own account as a Confluence administrator. Oops... Backtrack, fix and re-enable...

For editing pages, the wiki is far easier (and richer) than dealing with Sharepoint and plug-ins allow for HTML content and for Word and Open Office documents to be used as a content editor (though I wonder how good they are). My boss liked the idea that you could roll up content from various Sharepoint lists, more or less creating "dashboard" style pages. I wonder what we'll do with it?

Cheers!
Posted by at No comments:
Labels: , ,

17 July 2008

What I'm doing at work - Sharepoint and Confluence

I'm getting to know Confluence an enterprise wiki platform. Ultimately we want to be able to connect it to Sharepoint 2007 using the Confluence Sharepoint Connector, currently at v.0.5.


Using this, you can embed Sharepoint list data in Confluence pages and can display Confluence pages and tree structures inside Sharepoint WebParts. The Confluence content is also fully searchable from within Sharepoint too. So far, so good.

I'm doing up a demo for my team at work and expect to deliver it on Friday. I'll blog about it then.

Cheers
Mike
Posted by at No comments:
Labels: , ,
Subscribe to: Posts (Atom)