This tool helps process the French open data on radio sites > 5W and import it into an SQLite database that is easy to process and query. The original CSV actually has a some duplicate data, and the tables and field names as well as the schema are not always very straightfoward, therefore some processing is done in order to obtain the following target schema that is (hopefully) easier to work with:

It is to be noted that

• Table sites was created because there may be several supports at the same location
• One support obviously has several stations, but conversely the current definition of "station" may encompass several transmitters that (due to historical reasons) may be on different supports
• bandgroups.id is not a PK (it's not unique). I could have defined another table "bands_id" where bands would have been unique, but this would add complexity and overall the amount of band combinations is only 2x the amount of distinct bands

Then, the tool also enable the creation of tables gen_sites and gen_sectors (possibly in another DB) to summarize within a single table some useful information from the standpoint of sites (i.e. geographical address) and sectors (i.e. with a boresight), e.g. for the purpose of importing in a GeoDataFrame and mapping with leaflet.js.

• fradiodb.py -d : downloads the data (last version that was committed in this repository. More recent data may be available in data.gouv.fr). By default the two .zip files are stored in anfr/ subdirectory
• fradiodb.py -i foo.db [-p mydir/] : imports and process the CSV data from directory mydir (by default anfr) into the SQLite DB foo.db
• fradiodb.py -g foo.db : generate tables gen_sites and gen_sectors in foo.db

This is a work-in-progress, and both the schema and the processing may still have bugs and/or change