Author: | Daniel Belasco Rogers <[email protected]>, Peter Vasil <[email protected]> |
---|
A script for importing GPX files into a SpatiaLite database.
Uses Tomo Krajina's gpx module. gpx2spatialite takes a single or folder full of gpx files and imports them into a spatialite spatial database. On importing, a file is hashed and entered into a table to make sure that a file with identical contents is not added twice. The spatial information from trackpoints is stored in two tables in the spatialite database, trackpoints and tracklines which are related. You can read more detail about the reason to represent data from a GPX file like this here: <http://planbperformance.net/dan/blog/?p=984>
The compulsory 'user' argument is to import gpx files from different users into the same database. This is because gpx2spatialite was written as an importer for DrawingLife <https://github.com/ptrv/drawinglife> which is a visualisation software written by Peter Vasil for Daniel Belasco Rogers' and Sophia New's art project of recording everywhere they go with a GPS since 2003 and 2007 respectively (<http://belasconew.com/works/lifedrawing/>)
If you only have one user to enter, just pick any name and import all files under this name.
Because DrawingLife has text at the top of the screen which shows the current location of the animation screen, this information is provided by the citydef_uid column in the trackpoints table which points to the citydefs table. This column is also populated on import by default. If you do not require this, you can set the option -s --skip-locations which will speed up importing considerably.
- Python:
- gpxpy
- pyspatialite, pysqlite2 (with support for loading extensions)
- Other libraries:
- libspatialite
- libsqlite
gpx2spatialite is available via pip. You can simply run:
pip install gpx2spatialite
Or to install it with user scope run the following command:
pip install gpx2spatialite --user
Make sure that $HOME/.local/bin
is available in your PATH
environment variable.
Otherwise the gpx2spatialite executable will not be found when you run it from the shell:
PATH=$PATH:$HOME/.local/bin
- Download python 2.7 32bit for windows
- Install pip. (See instructions here)
- Install gpxpy with
pip install gpxpy
- Download mod_spatialte from here
and extract all dlls to
C:\Python27\Scripts
- Download sqlite3.dll from sqlite website
(something like 'sqlite-dll-win32-x86-3080704.zip')
and overwrite
sqlite3.dll
inC:\Python27\Scripts\DLLs
with the downloaded version (This ensures thatsqlite3.dll
is capable of loading extensions) - Install gpx2spatialite with
pip install gpx2spatialite
It is possible to read in single files:
gpx2spatialite import -d <path/to/database> -u <user_id> <path/to/gpx>
Or multiple folders:
gpx2spatialite import -d <path/to/database> -u <user_id> <path/to/folder1> <path/to/folder2>
Files and folders can be specified both at the same time:
gpx2spatialite import -d <path/to/database> -u <user_id> <path/to/folder1> <path/to/gpx>
Run script with subcommand to create a new database and initialize it:
gpx2spatialite create_db <path/to/new/database>
Each trackpoint has a location assigned to it which is used for the drawinglife animation.
These are defined at import time unless the option -s or --skip-locations is passed.
Locations are defined in the citydefs table in the database, created and populated automatically by running 'gpx2spatialite create_db'.
Import citydefs into existing database:
gpx2spatialite citydefs -i <path/to/input.sql> <path/to/database>
Export citydefs table:
gpx2spatialite citydefs -e <path/to/output.sql> <path/to/database>
After adding new locations to the citydefs table, you can look for currently unknown trackpoints and assign them to any relevant, newly defined locations with:
gpx2spatialite update_locs <path/to/database>
If you have redefined currently assigned locations or completely changed the citydefs table, you will want to redefine every trackpoint in the database, for which you use the -a or -all-locations option to the above script:
gpx2spatialite update_locs -a <path/to/database>
The repository contains the standalone py.test (version 2.5.2) script runtests.py.
Run the following command to run the tests:
python setup.py test
Local Variables: mode: rst End: