python-geosupport

A Python library for geocoding with NYC Planning's Geosupport Desktop Edition. Developed using version 17C of Geosupport Desktop Edition.

Getting Started

Prerequisites

Install Geosupport Desktop Edition:

Windows users: Ensure you select the correct Geosupport installation that corresponds to the Python interpreter you are using. Ex., Python 32-bit will only work with Geosupport 32-bit.

Additionally, you should check that the following environmental variables are set:

GEOFILES = the fls directory of your Geosupport installation. (ex. C:\Program Files (x86)\Geosupport Desktop Edition\fls\)
The Geosupport bin directory is in the PATH. (ex. C:\Program Files (x86)\Geosupport Desktop Edition\bin)

Linux users: Extract the .zip to a folder of your choice and set the GEOFILES and LD_LIBRARY_PATH environmental variables of the fls and lib directories like so:

$ export GEOFILES=/var/geosupport/version-17c/fls
$ export LD_LIBRARY_PATH=/var/geosupport/version-17c/lib/

How to use

Install via pip:
```
pip install python-geosupport
```

Import the library and create an instance:

from geosupport import Geocode

g = Geocode()

Geocode with:

Single input street address

result = g.address(address="125 Worth st, NY, NY")

Parsed street address (Must provide zip code or borough)

result = g.address(house_number="125", street="Worth st", zip=10013)

result = g.address(house_number="125", street="Worth st", boro='MANHATTAN')

Borough, Block, and Lot

result = g.bbl(boro='1', block='00168', lot='0032')

BIN

result = g.bin(bin='1001831')

Placename

result = g.place(place='VAN CRTLANDT MANSION', boro='bronx')

Intersection

result = g.intersection(street_1='ST NICHOLAS AVENUE', street_2='MENAHAN STREET', boro='QUEENS')

Street Segment

result = g.street_segment(on_street='STORY AVENUE',
                          cross_street_1='EVERGREEN AVENUE',
                          cross_street_2='WHEELER AVENUE',
                          boro='BRONX')

Blockface

result = g.blockface(on_street='STORY AVENUE',
                     cross_street_1='EVERGREEN AVENUE',
                     cross_street_2='WHEELER AVENUE',
                     boro='BRONX',
                     compass_direction='N')

Street Stretch

result = g.street_stretch(on_street='CLIFTON PLACE',
                          cross_street_1='SAINT JAMES PLACE',
                          cross_street_2='GRAND AVENUE',
                          boro='BK',
                          compass_direction='N',
                          compass_direction_2='S')

*The boro argument can be borough code, borough name or a common borough abbreviation.

Running tests

>> python setup.py test

Known Issues

Single address inputs containing Grand Concourse are not parsed correctly if they do not contain NY or New York in the string.
Single address inputs with a space instead of a hyphen are not parsed correctly. Ex. 180 05 HILLSIDE AVE, QUEENS, 11432

TODO

Improve single address input parsing
Better data with compass direction for testing blockface and street_stretch methods.

License

This project is licensed under the MIT License - see the license.txt file for details

Acknowledgments

Charles Wang for his initial proof of concept in python
Chris Whong and Noah Veltman for their work using Node.js
Datamade's usaddress

aepyornis / python-geosupport