download, convert, clean, import US tiger shapefiles into RethinkDB tables.

geojson

objects=zcta5 county concity
tiger_url=ftp://ftp.census.gov/geo/tiger/TIGER2015

zcta5: zcta5.import
county: county.import
concity: concity.import

%.import: %.geo.json
	rethinkdb import -f $< --table geo.$*
	@echo '$* import complete'

%.geo.json: %.shp
	ogr2ogr -t_srs crs:84 -f "GeoJSON" /vsistdout/ $< | \
		./clean_collection.js > [email protected]
	mv [email protected] $@

%.shp: %.zip
	unzip -d $@ $<

%.zip: %.manifest
	curl $(shell head -n 1 $<) -o $@

%.manifest:
	$(eval url := $(tiger_url)/$(shell echo $* | tr -s '[:lower:]' '[:upper:]')/)
	curl -l $(url) | \
		sort -n | \
		sed 's,^,$(url)/,' > $@

all: $(objects)

clean:
	rm *.manifest

.PRECIOUS: %.zip %.geo.json

.PHONY: all clean %.import

Figured out this is a makefile. But attempts to execute result in curl: no URL specified!. Looks like line 25, if this is a parameter I'm supposed to be supplying, I'm not sure what to.

@brwnll, you were right that I didn't get a notification about gist comments :P Thanks for the email!

Are you importing into rethinkdb? I would suggest mongodb for a few reasons, but if you're already set on rethink, I've fixed up this makefile with the updated ftp url and a few little other fixes (take a look at the revisions for more info).

There are more examples of the usage of this makefile (this particular gist is pretty old) and imports into rethink / mongo in this repo and I use this makefile for actual production imports for this project app. Take a look at the Makefile for either of those if you're interested and if you'd like some specific help with importing this data into either mongo or rethinkdb I can certainly help out! This gist also references a script clean_collection.js which is not present here. This makefile is a better example of simply generating the geojson feature collection (delimited by line breaks) for general usage in any import process. I had some issues with the rethinkdb import via commandline consuming too much memory so a js streaming script was the solution I eventually used.