Hi Shrini,

On 2 April 2017 at 21:41, Shrinivasan T <[email protected]> wrote:

> How can get them all?
>

Mapzen provides metro extracts[0] and I used xpath here to get list of
nodes without name:ta. Parsing extracted Chennai osm file (400 MB) consumes
lot of memory. Python experts, please feel free to optimize this code[1] in
better ways.  The list of nodes (I will come to ways, relations soon) is
available here[2].

Offlate, I am facing strange unicode issues with my script which used to
work, but will soon get it fixed. I am thinking of trying TensorFlow or
equivalent ML models to provide translation suggestions, aside from google
translate. Wikidata doesnt help when it comes to street level sadly.

[0] https://mapzen.com/data/metro-extracts/
[1] https://github.com/srikanthlogic/osm-ta/blob/master/MetroExtract2.py
[2]
https://docs.google.com/spreadsheets/d/1PJdMVuSHDQ9-11ewj2pT_CNqKMLUiqUurcfyHGgSfoE/edit?usp=sharing


-- 
Regards
Srikanth.L
_______________________________________________
Talk-in mailing list
[email protected]
https://lists.openstreetmap.org/listinfo/talk-in

Reply via email to