… not natively - but it does it now with a Dapper API I have created I had a very long email conversation today with Jon Aizen from Dapper (based in Israel) in which we discussed a better support for a Microformats friendly usage of the field and group names you can define within the Dappit creation process … and Jon was so nice to implement my ideas instantly!

What does Dapper do?

With Dapper you scan a webpage and Dapper analyses the structure of the page to cut it into small pieces. But Dapper not only split the page in uncoupled data chunks, but also “unterstand” the relations among the data pieces, which appear several times on a page.

For example: On a blog you will find the header of an article repeated several times - with the Dapper “editor” you click on a headline and Dapper markup all the other headlines on the page and shows you, that it understands what you want. This process and the sensibility of the Dapple “scanner”-algorithmen is full scaleable.

You can now select these fields and define a name for them. The fields can be subsumed into groups, which you can give a name, too. So you markup the structure of a website and create an API around it. Dapper than offers you several interfaces to access this - now structured - data. (XML, HTML, RSS and many more)

One of these interfaces is HTML and there is the point where it gets started with Microformats. For a better explanaition how Dapper works, see this Demo!

How does Dapper support Microformats?

We now reused the values for the field and groupnames as class names for the tags, which enclose the data pieces we have seleted within the Dapper creation process. Jon only had to adjust the HTML transformer and added the values as class names to the div and span tags, which have allready enclosed the data pieces. (There is now a “Microformat Mode” flag, when you choose the HTML output)

If you now know Microformats specifications you can easily create a microformated HTML output of a webpage which doesn’t support Microformats natively … as Timble’s Blog does it now … don’t know why I had his blog instantly in mind … by the way: his alternative Blog output does now support hAtom and hCard

BUT: I have to admit, that this quick’n'easy solution doesn’t cover the whole requirements of a well formed microformat implementation. Until know it has the following diseases:

• you can only create nestings one layer deep
• you don’t have access to the class and rel attributes of a-Tags within the data pieces (so no XFN, no rel-Tag etc.)
• you can’t converte date and time informations in the right dataformat
• abbr-Tags and other “details” can’t be used

… but you can allready do great stuff and I realy like the perspective of a tool like Dapper combined with Microformats.

Why should I use Dapper in this way?

Because you can? No - not this time It’s usefull because microformats are on the right way and many tools to process microformats are build right now. When you wanna do mashups or any other thing with data from microformated websites you will for sure have to handle the problem that a website you wanna use, doesn’t support microformats, yet.

With Dapper and a feature better implementation of microformats you can build a easy “just few clicks needed” API around any website and you can continue using the microformats tools you like!

Sound greats, doesn’t it? When I have our MicroMashup idea in mind, we wanna make it real one day before Barcamp Berlin, I think Dapper will be very usefull. But as you might have noticed we are not at the end of Dappers abilities to handle microformats. But to continue this development, we need you!

If you are interested in helping us with your programming skills, drop me an email!

If you are Tim Berners Lee:
Do you have a Mac and NetNewWire? Because than you could now …
Do you use a contacts application, which supports vCard? You could now …

Also have a look at the API’s I’ve allready created for the Technorati Microformats Search!
Feedback wanted!