A framework in Python to process OpenStreetMap data
Python Other
Switch branches/tags
Nothing to show
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.
minimongo @ 2521e6e Update minimongo: fixes #1 Feb 17, 2014
osm Update pbar.py Oct 30, 2014
utils Merge branch 'master' of github.com:larroy/osmcompiler Nov 5, 2013
.gitmodules Initial commit Aug 26, 2012
LICENSE.txt Initial commit Aug 26, 2012
README.md Fix documentation Sep 1, 2012
bootstrap.sh bootstrap.sh Jul 31, 2013
mongocredentials.py Using argparse Oct 23, 2012
osm_mongo_compiler.py Using argparse Oct 23, 2012


Open Street Map to Mongodb compiler

This is a framework to process OSM data dumps in PBF format.

There's an included backend to store data in mongodb, to implement a new backend it's very simple, just implement osm.sink which does an action on the processed entities and osm.factory which creates the instances of OSM entitities. Then call the parser with your implementations of sink and factory.


Initialize the submodules:

git submodule init
git submodule update


python-pymongo python-protobuf


./osm_mongo_compiler.py spain.osm.pbf 
Loading: spain.osm.pbf
Counting blobs...  3829.
Bounding Box: (-9.779014,35.91539) (5.098525,44.14855)
[      0% 7/3829 blocks. 56 K nodes 2012-08-26 18:58      ]


Count the number of data blocks with:

./osm_mongo_compiler.py file --count yes

By running with -f and -n options the file can be devided in ranges and thus processing be done in a map kind of job across several processors or nodes.

./osm_mongo_compiler.py file -f 0 -n 1000
./osm_mongo_compiler.py file -f 1000 -n 1000


Feedback welcome to pedro.larroy.lists@gmail.com please put [osmcompiler] on the subject or your mails will be probably ignored. Based initially on the parsepbf tool by Chris Hill.