“Media Extractor” weekly report #12

julio 21, 2014

14/07/2014 -> 18/07/2014

Last week I (almost) completely implemented FlickrWrappr2, with a semi “pluggable” architecture very similar to the one we discussed earlier on with my mentor.

However, I still have some problems correctly implementing the singleton (again!). Correctly accessing the member of the companion object/class in Scala is not as trivial as it would seem…

While I fix that, I started to research on adding other media providers (the result of this, to appear on next week’s resume).


“Media Extractor” weekly report #11

julio 15, 2014

07/07/2014 –> 12/07/2014

This week, I worked a lot. I managed to:

-refactor all FlickrService-related scripts into proper tests.

-discarded (for now) the “complex-singleton” approach for the FlickrOAuth authentication. It is simply too complex to be implemented easily. In any case, I reused lots of its non-OAuth-related code into the “not-so-complex” implementation…

Next step is refactoring the RDF generation-related code


“Media Extractor” weekly report #10

julio 15, 2014

02/07/2014 –> 04/07/2014

Due to a personal urgency, I could not work much this week. However, I managed to:

-define and patially implement a “complex-singleton” implementation for the persistent OAuth authentication
-keep on working on the “simple” implementation of the persistent OAuth AuthenticationCorrectly implementing Singletons in Scala is taking me much more work than expected…8


“Media Extractor” weekly report #09

junio 27, 2014

22/06/2014 –> 27/06/2014

-After a short discussion with Alexandru, realized that the GeoSearch RDF generation code I ended up with last week was right all along… So, I extended it to DBpediaSearch RDF (which is quite similar).

-Even though some bugs persist, I have now managed to (mostly) replicated all FlickrWrappr functionality via experimenting and testing in Scala.

-However, the components are dispersed. I still have to make them work correctly together into a “Scalaized” FlickrWrappr script. Enter FlickrWrappr2.

Branching out…
-Started refactoring my tests together into FlickWrappr2
-Defined and implemented a preliminary “pluggable” architecture for Media-Extractor, based on the “Strategy” design pattern. Alexandru says it is a good idea to try.

-I will work in parallel in both branches, as solving one problem only for Flickr will provide me with useful insight when working on the “pluggable” version (e. g. I realized that I need to use a Singleton object to make the Flickr authentications credentials persistent across invocations)


“Media Extractor” weekly report #08

junio 27, 2014

16/06/2014 –> 20/06/2014

Not a nice week:

-Took a short break after being working nine days without stopping.

-Got back to work, and lost lots of time trying to correctly generate RDFs from Flickr GeoSearch (not only coding, but also studying)… Until Alexandru confirmed me later on that I had been doing it right all along. Hmmm…


“Media Extractor” weekly report #07

junio 16, 2014

09/06/2014 –> 15/06/2014

Started to experiment with Jena (for RDF). Managed to generate RDF files partially alike as those generated by flickrwrappr (image and location elements -metadata still to go).

Started to study SPARQL.

Almost done with replicating flickrwrappr’s functionallity…


“Media Extractor” weekly report #06

junio 14, 2014

02/05/2014 –> 08/06/2014

Researched on XML I/O (to work with Flickr search results) and on OAutth (to automate Flickr authentication). Managed to generate Flickr links to the photos and to the pages.