Hi David, The language specific index code might be nice to mention as well. In short it uses language specific regexes to split the street name into parts.
In Dutch for example you have lots of prefixed surnames which find their way into streets, and you'd want a 'van der Plankstraat' to be indexed under P, not V. That's what the indexer takes care of, because the rules/habits for that differ between localities. Using a Dutch map with English-style indexing would be very confusing for a Dutchman to use, because you'd keep looking up streets under the 'wrong' letter. Cheers, Jeroen On Wed, Nov 6, 2013 at 4:17 PM, David MENTRE <dmen...@linux-france.org>wrote: > Hello, > > I will do a small presentation on MapOSMatic internationalization next > week[1]. I would like to talk about our code but also our > infrastructure and workflow using Transifex. Could someone refresh my > mind on that side? > > As far as I remember : > - we have strings to internationalize in ocitysmap (back-end) and > maposmatic (web front-end) modules; > > - we use language specific Python code for index generation; > > - through Django, we generate PO files, one for each module and > language. We use home made i18n.py script file for that; > > - those PO files are uploaded to transifex (automatically?); > > - translated PO files are downloaded from transifex (automatically?) > and integrated (how?) into our dev git tree (which one exactly)? > > - the dev git tree is pushed on dev.maposmatic.org for tests and > sometimes merged into www.maposmatic.org for word consumption. > > Has anybody any stats on our i18n effort, like number and extent of > supported languages? > > > Best regards, > david > > [1] > http://www.lacantine-rennes.net/2013/11/devcamp-12-coder-pour-plusieurs-langues/ > > -- ↑↑↓↓←→←→BA[Start]
