[Imports] Colorado Address Imports

[smocktaylor at gmail.com]

This is a two-phase import project.

I'm going to be updating addresses in Mesa County, Colorado with their
E911 Address Point data (see https://emap.mesacounty.us/DownloadData/ ,
updated January 22, 2020) and eventually importing the state address
database (
https://data.colorado.gov/State/Statewide-Aggregate-Addresses-in-Colorado-2019-Pub/n7je-akky
, updated May 29, 2019).

I've previously gotten permission to use the former, and the latter is
Public Domain (I did confirm that it was in the copyright public domain
with the state). The former dataset is a subset of the latter dataset.

Unfortunately, they are going to require different translations.

I originally downloaded the Mesa County E911 dataset in May, 2019, but
I do not anticipate any significant differences in terms of
translations for tags. There are likely to be new addresses available.

I am currently documenting the process here: 
https://wiki.openstreetmap.org/wiki/Import/Colorado_Addresses .

If you do look at the May 2019 dataset transformation from Mesa County,
please note that (a) everything that needs to be expanded should be
expanded, (b) there is an actual "E Road" and "N Road" (so please don't
ask me to expand them), and (c) streets and addresses really do have
"1/2", "3/4", or some other bizarre fraction ("J 2/10 Road", for
example, *really* is "J 2/10 Road"). I do not intend to convert the
fractions to unicode (casual editors are unlikely to be able to type
the ½). It does make for interesting navigation though ("turn right
onto J 2 10 road" (OsmAnd)).

I'm going to be using the MapWithAI JOSM plugin with 
https://gitlab.com/smocktaylor/serve_osm_files to serve specific
portions of the files up and then add them to OpenStreetMap.

The MapWithAI JOSM plugin can combine addresses and buildings, and I
intend to use that functionality. It will not combine addresses and
buildings if there is more than one address inside the building (*must
be added at the same time for this to work, or one must previously
exist, currently*).

Until I get around to writing a conflation service in python, I do not
currently intend to remove addresses from the dataset that is served to
the MapWithAI plugin. I may, at some point in time, add that
functionality along with a flag to get the "full" dataset (for
debugging purposes for other import projects).

While I am doing the import, I also intend to draw the buildings that
the addresses go with, and merge them together, so I was not intending
to create a separate account for this import (since it is both actual
editing + import of data).

Any feedback would be welcome.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 866 bytes
Desc: This is a digitally signed message part
URL: <http://lists.openstreetmap.org/pipermail/imports/attachments/20200125/3518a850/attachment.sig>