Stetl, streaming ETL, pronounced "staedl", is a lightweight ETL-framework for geospatial data conversion.
Notice: the Stetl GH repo is now at the GeoPython GH organization.
Stetl is released under a GNU GPL v3 license (see LICENSE.txt).
The Stetl website and documentation can be found via http://stetl.org. For a quick overview read the 5-minute Stetl-introduction, or a more detailed presentation. Stetl was presented at several events like the FOSS4G 2013 in Nottingham and GeoPython 2016.
Stetl basically glues together existing parsing and transformation tools like GDAL/OGR, Jinja2 and
XSLT with custom Python code. By using native libraries like libxml2
and libxslt
(via Python lxml
) Stetl is speed-optimized.
A configuration file, in Python config .ini
format, specifies a chained sequence of transformation
steps: typically an Input
connected to one or more Filters
, and finally to an Output
.
At runtime, this sequence is instantiated and run as a linked series of Python objects. These objects are
symbolically specified (by their module/class name) and parameterized in the config file.
Via the stetl -c <config file>
command, the transformation is executed.
Stetl has been proven to handle 10's of millions of GML objects without any memory issues.
This is achieved through a technique called "streaming and splitting".
For example: using the OgrPostgisInput
module an GML stream can be generated from the database.
A component called the GmlSplitter
can split this stream into manageable chunks (like 20000 features)
and feed this upstream into the ETL chain.
Stetl has been found particularly useful for complex GML-related ETL-cases, like those found within EU INSPIRE Data Harmonization and the transformation of GML/XML-based National geo-datasets to for example PostGIS.
Most of the data conversions within the Dutch NLExtract Project apply Stetl.
Stetl also proved to be very effective in IoT-related transformations involving the SensorWeb/SOS.
Browse all examples under the examples dir. Best is to start with the basic examples
Stetl can be installed via PyPi pip install stetl
and recently as a Stetl Docker image.
More on installation in the documentation.
Anyone and everyone is welcome to contribute. Please take a moment to review the guidelines for contributing.
Stetl originated in the INSPIRE-FOSS project: 2009-2013 now archived. Since then Stetl evolved into a wider use like transforming Dutch GML-based Open Datasets such as IMGEO/BGT (Large Scale Topography) and IMKAD/BRK (Cadastral Data) and Sensor Data Transformation and Calibration.
The word "stetl" is also an alternative writing for "shtetl": http://en.wikipedia.org/wiki/Stetl : "...Material things were neither disdained nor extremely praised in the shtetl. Learning and education were the ultimate measures of worth in the eyes of the community, while money was secondary to status..."