We've been managing a bunch of cascadenik mml files which contain duplicate data source references; e.g. same thing rendered with different classes. Using XML entities is great, but it's a bit hokey. I've implemented a solution that scratches our itch, but I thought I'd proffer it to the community.
In a nutshell: 1) Define data sources and their SRS separately from layers (parse using python ConfigParser) 2) Layers refer to a data source by name 3) Support variable substitutions through the config parser's builtin mechanism 4) Allow for overrides of datasources and substitution vars (like settings.py and local_settings.py in django) 5) Compile .mml to .xml This makes cascadenik mml more MVCish: .mms == view, .cfg == model, .mml == controller Example: http://github.com/umidev/mapnik-utils/commit/806cfd75eb3774d27bf7a4a2e2e0f0621edec695 Implementation: http://github.com/umidev/mapnik-utils/blob/806cfd75eb3774d27bf7a4a2e2e0f0621edec695/trunk/serverside/cascadenik/cascadenik/source.py http://github.com/umidev/mapnik-utils/blob/806cfd75eb3774d27bf7a4a2e2e0f0621edec695/trunk/serverside/cascadenik/cascadenik/compile.py#L536 This was done quickly; not lots of comments in code, but not a lot of code. Cheers Nino Walker :: Urban Mapping, Inc 690 Fifth Street, Suite 200 :: San Francisco, CA 94107 T. 415.946.8170 x805 :: F. 866.385.8266 :: www.urbanmapping.com<http://www.urbanmapping.com/>
_______________________________________________ Mapnik-users mailing list [email protected] https://lists.berlios.de/mailman/listinfo/mapnik-users
