[uf-discuss] Commercial application using hResume import

[Glenn Jones]

> Sarven wrote:
> Neat. Quick feedback:

> I tried it on http://csarven.ca/cv and it seems to pick up only a few
of
> the org vCards. 

The parser is picking up your mark-up. For example I can see that you
have use hcard/org to mark-up educational institution names.
http://ufxtract.com/api/?url=http%3A%2F%2Fcsarven.ca%2Fcv&format=hresume
&output=xml

It's just the mapping from raw microformat data to our CV structure has
not made the best use of your mark-up. I will update our application.  

> It'd be great if it picked up the personal and contact details of
vcard
> with .uid on the page.

The hResume spec looks for a hCard using which is marked up with the
class "contact". I could extend the parser to follow the Representative
hCard rules, but I things its better if you mark-up a hCard for hResume.
The "contact" hCard is required and technically a hResume is invalid
without it.

> Skills didn't pick up.

You need to add rel="tag" to your skill links. This only half the
problem because if you review the http://microformats.org/wiki/rel-tag
page and read the "Tag Spaces" section you will find skills can be very
difficult to define in real world use. This is because the skill is not
the text of the link but the last segment of the URL structure.

Not correct use
<a class="skill" title="Ubuntu" href="http://www.ubuntu.com/">Ubuntu</a>

Correct use - has rel-tag and a tag namespace in the URL structure
<a rel="tag" class="skill" title="Ubuntu"
href="http://en.wikipedia.org/wiki/Ubuntu/">Ubuntu</a>


> Education level didn't pick up.

The education level is not part of the hResume structure. I could infer
an education level by using the NPL function on the education elements
of the hResume. So far I have resisted mixture explicit structured data
from Microformats with the more implicit data parsed by the NPL
functionality. I am worried what user expectation would be. 

> -Sarven

Thanks for the feedback very useful.  

Glenn