Paul,
Fair rant to the wrong list. :) Answer follows, but let's move the
discussion.

A long time ago we decided to use trunk as the live dev branch, and tags to
mark stable points. When in time pressure, we (mostly I) end up committing
snippets that were added for temporary tests, or for new features that
aren't already finished, and shouldn't have made it to SVN. If you want a
stable copy, please feel free to use tag/release-0.4
We are moving to git soon, where it will be easier to manage branches away
from trunk.

This little parameter is for a new feaure that was, in fact, requested by
you, but that we did not have time to fully include for this release. You
can now decide to load the candidate map in memory, but you still need to
touch code for that. It is a working artifact that is not fully configurable
yet. And yes, I got spoiled by Scala that doesn't return "surprise nulls",
rather using the much more elegant Option. Forgot to test for that one null,
and it shouldn't throw with the config file committed, so thank you for
testing it with an outdated config so that I can remove this bug.

About dependency injection, yes, we admit that configuration/dependency
management in DBpedia Spotlight needs work. Please feel free to contribute
that to the project. Or we will have to wait until I am not overloaded so
that I can do it myself.

Cheers
Pablo
On Sep 27, 2011 7:27 PM, "Paul Houle" <[email protected]> wrote:
> I just brought my Spotlight up to date and had a lot of breakage in my
> configuration file because of things like this...
>
> //TODO use separate candidate map
> candidateMapDirectory =
> config.getProperty("org.dbpedia.spotlight.candidateMap.dir").trim();
> if(!new File(candidateMapDirectory).isDirectory()) {
> //throw new ConfigurationException("Cannot find candidate
> map directory "+ candidateMapDirectory);
> }
> LOG.info("Read candidateMap.dir, but not used in this version.");
>
> I get a null pointer exception because I don't define this property,
> which must point to a real directory. The big joke is that this is "not
> used in this version" so it's another nonworking artifact that gets
> people confused when they start working with Spotlight.
>
> Spotlight might benefit from dependency injection or another advanced
> approach to configuration, but it would help a lot to wrap code like
> the above in an existence check so I don't need to maintain a whole
> bunch of bogus configuration parameters that the system doesn't use!
>
>
------------------------------------------------------------------------------
> All the data continuously generated in your IT infrastructure contains a
> definitive record of customers, application performance, security
> threats, fraudulent activity and more. Splunk takes this data and makes
> sense of it. Business sense. IT sense. Common sense.
> http://p.sf.net/sfu/splunk-d2dcopy1
> _______________________________________________
> Dbpedia-discussion mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion
------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure contains a
definitive record of customers, application performance, security
threats, fraudulent activity and more. Splunk takes this data and makes
sense of it. Business sense. IT sense. Common sense.
http://p.sf.net/sfu/splunk-d2dcopy1
_______________________________________________
Dbpedia-discussion mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dbpedia-discussion

Reply via email to