iSpot & EML project blog

Along with all the normal maintenance work of iSpot I’ve had to spend some time this week on fixing the forums on the site. This was to do with the updates of various modules and also the core drupal code to patch potential security vulnerabilities. This left some users unable to post to various of the forums and also changed the breadcrumbs that were displayed for the forums. It took longer than I’d have liked to track down exactly what was happening but once I had it didn’t take too much to fix it.

The MegaLab officially stopped gathering data as of 31st October. This does not mean that the site is no longer available, in fact it’s business as usual for the Evolution MegaLab but it does mean that analysis has now begun on the data that has been collected to date. So to help with this I’ve added two new fields to the full data download that’s available to privileged users, these are: an id field that gives a unique reference to every line of the download, and the date the record was published – this is different from the date the record was made.

It’ll be interesting to hear the outcomes of the data analysis!

I noticed last night that the website had gone down. I’m in the process of working out why. It seems to be that something may have changed with the OU firewall and is no longer allowing the web server to communicate with the database server. Hopefully it will be sorted out soon.

I’ve just noticed the Twitter widget isn’t working, I don’t know who’s twitters it’s showing but it’s not mine!! I can’t get to the code at the moment because the blogs have been moved to a different server which I don’t think I have access to as yet. Once I’ve got permission I’ll have a look at it. Please ignore it for the time being!!

I thought it’d be a good idea to write this down for my own future reference if nothing else.

The sfMessageSource_XLIFF.class.php file in the core files of symfony is responsible for the storage and retrieval of translation text from the XML formated XLIFF files. I’ve found a problem with this which was causing problems with the use of the sfI18nExtractPlugin.

The problem is that the text is not stored as CDATA so consequently the html tags are escaped to produce valid XML while any html entities are left alone. When the XML file is then loaded the next time the object that is created from this decodes ALL the escaped characters including the html entities.

This leads to a problem for the sfI18nExtractPlugin when comparing already stored translation text and the text it has extracted from the pages that render the HTML. The extracted text has entity references for special characters while the stored translation text contains the special characters themselves.

Consequently I’ve modified the sfMessageSource_XLIFF.class.php file to store the text as CDATA, which solves the problem. The loading of the XML that takes place in the same file does not need modifying as the CDATA is taken and and cast as a string which is correct.

the original code was:

$source = $dom->createElement(‘source’, $message);

and I’ve chaged it to:

$cdata = $dom->createCDATASection($message);
$source = $dom->createElement(‘source’);
$source->appendChild($cdata);

I’ve just noticed that the Google maps images aren’t available for either the satellite or hybrid view. I’ve checked other sites and it looks to be a problem with Google.

Hope this doesn’t prove to be a recurrent problem.

Because of the time constraints of building the first version of the MegaLab for the Milton Keynes science week in July I think that the spot recording must be limited to starting from scratch each time the user wishes to enter some spots. This is because there are many issues with how to pick a pre-existing event, whether or not an event is editable, groups linked with an event etc etc… This does mean that each time someone records spots it will be seen as a new event, and that each event will only be associated with one type of habitat.

If I’ve made sufficient progress with the rest of the site I will try and look at this before July, but for the first version I don’t think we can be to ambitious.

The development server is now online, and the current version of the megalab can be viewed at evolutionmegalab-dev.open.ac.uk. This can only viewed from within the OU firewall, and I must stress that this is the actual development version, so it’s constantly changing and it may not work at all from time to time.

The second server – the approval server needs to be moved to a less restricted area of the OU network, which will take a little longer. Once this has happened people will be able to view the site externally and it will not be in active development i.e. I will periodically update this version with incremental improvements/developments, but it should be more stable.

I am now in possession of the servers as of this morning, now all I need to do is install all the relevant software and copy across the website as it is so far from the SVN repository. This will probably take a day, so by the end of today I hope to have the servers online. It will take a further couple of days to put any domains names in place. I also need to consider if external access is needed to the site during development, in which case one of the servers will need to be moved to the DMZ which will take longer.

I will be using this blog to record my progress on the development of this project. I’m going to try and make a weekly post about progress, probably every Friday. In addition I shall be using this to post any ideas I think are relevant and using it as a chance to listen to any comments.