# Install dependencies
$ pip install -r requirements
# Run the script, which outputs an analysis to stdout, and also generates `better.kml`, which is meant to be added as overlay to a Google Map
$ ./find_parking_meters.pyDefinitions are drawn using this polyline tool website, which allows one to draw polyline definitions on a Google Maps-like map- https://www.keene.edu/campus/maps/tool/
It produces two things: a list of lat/lon coordinates, and a JSON containing those coordinates.
That format is preserved when storing those definitions in this project, because: the JSON is imported by find_parking_meters.py (via load_boundary_file() from utils.py, which retrieves only the JSON.) The lat/lon list super-useful whenever the boundaries need adjusting. The list can be copy-pasted directly to the import box of the Keene State Polyline tool.
Much of San Francisco's data regime is provided as single tables that can be exported as TSV. Those files are downloaded to ./data, and imported using load_tsv().
SF Data files of physical objects and boundaries typically contain the columns LATITUDE & LONGITUDE. Some contain WKT definitions of points, polylines, or polygons. Those are imported using wkt_to_kml(), which generates a simplekml LineString whether or not the original called for a polygon or linestring. Seems to work fine, but the only use-case so far is parking meter Point locations. I suppose this may call for a cascading if/elif to produce specific kml objects, in the future.