|This is very much a work in progress.|
Based on ministry of planning and construction of Laos data adquired in 2010 by Jasper Overgaard Waale and donated in December 2015 to the OpenStreetMap community with Public Domain license.
- Proposal to import Lao's district & province administrative boundaries to OpenStreetMap.
- Proposal for a continuous community import of the rivers and roads and 11000 villages (with their original Lao name) of Laos
- To complete the current OSM road and river and villages coverage of Laos.
- To add the current administrative boundaries until district level (that is admin_level=6) with the new official 17th province since 13 December 2013.
The data to be imported comes from a donation by Jasper Overgaard Wale (jasper.o.waale AT gmail.com) which adquired the data from the ministry of planning and construction of Laos around 2010 and was released to OSM with a ODbl permission.
One of the current problems in OpenStreetMap about Lao's data is the lack of volunteers to trace roads and rivers from the Mapbox and BING imaggery sources, also the current administrative boundaries only cover the 17 provinces, not updated to the 2013 official changes and to the level of districts.
OSM Data Files
- Python scripts on the github repository: https://github.com/capiscuas/libreGTFS/blob/master/other/laos_import
- Rivers OSM File (do not import, just use as a background layer and trace the rivers using the imagery from Mapbox or BING, there are around 180 names of rivers in Lao script and english to profit from.)
- 11000 Villages OSM File (same as rivers, only use as background layer and profit names)
- Roads OSM File (same as rivers, only use as background layer while tracing from Mapbox/BING)
- Boundaries OSM File (already imported, fixed with new political boundaries.
- Boundaries: One time import
- Rivers & Roads: Recommendation is to use this files only as a layer support to draw the river based on Mapbox or BING imagery (e.g. sometimes because of the clouds, this files can help us find out what direction does the road or the river heading to)
- Villages: I believe the data contains all of the 11000 current villages of Laos, with their original Lao name in Lao script and their romanization in English. The locations are not exact but are near by each village if you check the imagery. Once you are sure of the position of each village, you can copy paste the node to your workspace and import it.
Data Reduction & Simplification
For the case of the roads and rivers, the data was imported and converted trough JOSM and ways were simplified to save space in the number of nodes per way.
Conversion and translation
Conversion from shapefile to osm
For the case of the boundary files, a python script was created to automatically convert each district polygon, split each segment into a separate way and add those new ways into the corresponding district OSM relations and automatically create also the Province relations with the ways, saving a lot of manual human work. Metadata from the previous provinces will be kept(wikidata id, wikipedia id, ISO codes, foreign names), and for the districts will be added based on wikipedia articles about the districts of Laos, updated by me to visualize the new 17th province since December 2013.
For the case of the rivers, the original source file had some Lao script->Ascii conversion mapping that didn't really comply with any common script(neither IBM-1133 or Unicode) so trough a python script a correlation (ascci-> Lao unicode character) worked to rescue the name of 188 major rivers in Lao-script.
For the case of the 11000 villages, all their Lao names were restored using the script.
Data Merge Workflow
Thanks to Jasper Overgaard Waale for donating the data to OSM.
How to merge? The boundaries will be merged respecting the boundaries of the the neighbouring countries already existing, their wikidata,wikipedia data too. And also each province, district will preserve their metadata and be added more collecting it from the wikidata, wikipedia references.
The following guide has been read: https://github.com/jremillard/osm-import-toolkit
For the rivers and roads, in order for the community to use the files on their benefit, a task could be added in tasks.hotosm.org to use the data as a imagery layer and be able to import ALL the roads and rivers that could be visible by BING and Mapbox using this roads/rivers layer as a approximated reference to find them easily.