Deploy on Heroku

[banesullivan]

Its live at https://localtileserver-demo.herokuapp.com

I'm using a free dyno so if this gets any traffic at all, it will probably crash

The changes in this PR make it so that any URL file can be passed as a query param at that URL for viewing. For example:

https://localtileserver-demo.herokuapp.com?filename=https://opendata.digitalglobe.com/events/california-fire-2020/pre-event/2018-02-16/pine-gulch-fire20/1030010076004E00.tif

And another file hosted on S3: https://localtileserver-demo.herokuapp.com/?filename=s3://sentinel-cogs/sentinel-s2-l2a-cogs/2020/S2A_31QHU_20200714_0_L2A/B01.tif

[codecov-commenter]

Codecov Report

Merging #30 (7d18c47) into main (fd3436d) will increase coverage by 0.64%.
The diff coverage is 82.03%.

@@            Coverage Diff             @@
##             main      #30      +/-   ##
==========================================
+ Coverage   80.06%   80.70%   +0.64%     
==========================================
  Files          14       16       +2     
  Lines         647      705      +58     
==========================================
+ Hits          518      569      +51     
- Misses        129      136       +7     

[giswqs]

Wow, this is exciting! BTW, Is the sample dataset used in the demo a low-resolution image?

[banesullivan]

Yep, that sample image is super low resolution

[banesullivan]

Here is a demo with the changes on this branch:

from localtileserver import get_leaflet_tile_layer
from localtileserver import RemoteTileClient
from ipyleaflet import Map

# Connect to a remote instance of localtileserver
tile_client = RemoteTileClient(
    'https://opendata.digitalglobe.com/events/california-fire-2020/pre-event/2018-02-16/pine-gulch-fire20/1030010076004E00.tif',
    #'https://localtileserver-demo.herokuapp.com/',
)

# Create folium tile layer from that server
t = get_leaflet_tile_layer(tile_client)

m = Map(center=tile_client.center())
m.add_layer(t)
m

RemoteTileClient will use my demo production server by default but it throws a warning about the fact that this server will not be able to handle that many requests

[banesullivan]

Another demo:

from localtileserver import get_leaflet_tile_layer
from localtileserver import RemoteTileClient
from ipyleaflet import Map

# Connect to a remote instance of localtileserver
tile_client = RemoteTileClient(
    's3://sentinel-cogs/sentinel-s2-l2a-cogs/2020/S2A_31QHU_20200714_0_L2A/B01.tif',
)

# Create folium tile layer from that server
t = get_leaflet_tile_layer(tile_client)

m = Map(center=tile_client.center())
m.add_layer(t)
m

[banesullivan]

I'm not going to document this in the README to avoid people overloading my demo server.

I will however try to add some instructions on how to deploy this to heroku on your own and perhaps make a cookiecutter to make this super easy

[banesullivan]

I just added a template for deploying your own here: https://github.com/banesullivan/remotetileserver