OpenSearch Validator

[Tatham Oddie]

Hi folks,

 

Just letting you know that I’ve released a validator specifically for OpenSearch feeds:

 

                http://opensearchvalidator.com/

 

The source code for the validator (including a bunch of custom XSDs) is online here:

 

http://opensearchvalidator.codeplex.com/

 

I’ve also documented the test process in plain English here:

 

http://opensearchvalidator.codeplex.com/Wiki/View.aspx?title=general%20test%20outline

 

I’d love to gather any feedback that the group might have. J

 

 

Thanks,

 

Tatham Oddie

au mob: +61 414 275 989, us cell: +1 213 422 7068, skype: tathamoddie, landline: +61 2 8011 3982, fax: +61 2 9475 5172

my business: tixi.com.au – Ticketing without the dramas

 

[Ed Summers]

On Wed, Aug 5, 2009 at 6:44 PM, Tatham Oddie<tat...@oddie.com.au> wrote:
> Just letting you know that I’ve released a validator specifically for
> OpenSearch feeds:

Hey, this is really nice, and fills an important niche in the
opensearch ecosystem I think. In your test plan [1] you have:

4. Expects to find only one matching link tag

I have a website [2] that references two opensearch descriptions, each
of which sit on top of different types of content (newspaper metadata
and actual newspaper page content). Am I doing something technically
wrong, or is your validator being too strict [3]. FWIW IE and Firefox
seem to be able to handle multiple descriptions just fine.

//Ed

[1] http://opensearchvalidator.codeplex.com/Wiki/View.aspx?title=general%20test%20outline
[2] http://chroniclingamerica.loc.gov
[3] http://opensearchvalidator.com/validate?p=http%3A%2F%2Fchroniclingamerica.loc.gov%2F&st=roosevelt

[DeWitt Clinton]

Ed's right -- multiple rel="search" links on a page are definitely allowed.  Some clients distinguish between them, others accept only one, but in any case they should be considered valid.

-DeWitt

[DeWitt Clinton]

The Validator is very neat, btw.  Could someone add a Tools page to the opensearch.org wiki and link to it, please?

[Tatham Oddie]

Hi DeWitt,

 

Thanks for the vote of support.

 

Do you by chance know who owns/manages this page?

 

http://www.opensearch.org/Community/OpenSearch_software

 

 

Thanks,

 

Tatham Oddie

au mob: +61 414 275 989, us cell: +1 213 422 7068, skype: tathamoddie, landline: +61 2 8011 3982, fax: +61 2 9475 5172

my business: tixi.com.au – Ticketing without the dramas

 

[Tatham Oddie]

Hi Ed and DeWitt,

 

Thanks for your feedback, and sorry for the delayed response.

 

I’ve updated the validator to support multiple link tags now. The validator was born out of frustration from the clients (*cough* IE *cough*) not following their own standards properly and thus it can be overly restrictive in some cases. I’ve erred on the side of caution so that a PASS result from the validator can indicate that your feed is guaranteed to work. In this case, it was too aggressive and I’ve corrected it.

 

It’s now surfacing a real problem with your feed though. J The ShortName node has a max length of 16 characters which you are exceeding in both of your descriptions.

 

http://www.opensearch.org/Specifications/OpenSearch/1.1#The_.22ShortName.22_element

 

Thanks again, and please keep telling me about other issues you find. J

 

 

Thanks,

 

Tatham Oddie

au mob: +61 414 275 989, us cell: +1 213 422 7068, skype: tathamoddie, landline: +61 2 8011 3982, fax: +61 2 9475 5172

my business: tixi.com.au – Ticketing without the dramas

 

 

From: opens...@googlegroups.com [mailto:opens...@googlegroups.com] On Behalf Of DeWitt Clinton

Sent: Saturday, 8 August 2009 1:54 AM
To: opens...@googlegroups.com

[Ed Summers]

On Thu, Aug 13, 2009 at 2:21 AM, Tatham Oddie<tat...@oddie.com.au> wrote:
> It’s now surfacing a real problem with your feed though. J The ShortName
> node has a max length of 16 characters which you are exceeding in both of
> your descriptions.

Hey nice, thanks for the heads up re: ShortName. And of course,
thanks for putting the time and effort into
http://www.opensearchvalidator.com.

//Ed

PS. Another random idea for the service: capture the description file
URIs that are validated, and expose them via an rss/atom feed. Might
be a nice way to keep tabs on the growth of the opensearch community.

[Andrew Turner]

I would particularly like to see the uptake of extensions,
specifically OpenSearch-Geo.

Thanks for the great tool!

Andrew

--
Andrew Turner
mobile: 248.982.3609
and...@fortiusone.com
http://highearthorbit.com

http://geocommons.com Helping build the Geospatial Web
Introduction to Neogeography - http://oreilly.com/catalog/neogeography