Add SSL Certificate for examples.get_path()

[belletva]

Currently, xDEM does not support adding an SSL context for urllib.request.urlopen calls, which is required in some cases to authenticate with the server. When trying to fetch data using xdem.examples.get_path, the function fails because it does not support adding an SSL context. The error occurs because the server requires a valid SSL certificate for authentication.

The xdem.examples.get_path function should support adding an SSL context to urllib.request.urlopen calls. This would allow users to provide a certificate file path (cafile) to validate the connection.

def get_path(self, example_name, cafile=None):
 ....
 if cafile:
        context = ssl.create_default_context(cafile=cafile)
        response = urllib.request.urlopen(url, context=context)
    else:
        response = urllib.request.urlopen(url)

[rhugonnet]

Great idea! Linked to GlacioHack/geoutils#607, I think? 🙂
It's supported upstream by Rasterio/Geopandas, so should be a relatively easy addition.

I think you mean adding support to file opening in Raster(), Vector() and PointCloud(), instead of get_path()?

The function get_path() is only used internally to get the path to local examples for tests/documentation. And it's not recommended to depend on anything hosted online (that is not GitHub) for CI/doc, so we shouldn't use it there. We even have a plugin pytest-socket that checks we don't access anything online during tests, otherwise it makes the test fail! 😉

[belletva]

My problem was concerning the following line: filename_dem = xdem.examples.get_path("longyearbyen_ref_dem"). I need a certificate in order to have access to the example data in https://github.com/GlacioHack/xdem-data/tree/main/data/Longyearbyen. Maybe most people will want to download the data from the documentation so they can test the examples in the doc ?

[rhugonnet]

Hmm, that's strange... I've never seen this locally or in CI. I just tested with the latest main and it works fine (i.e. xdem.examples.get_path triggers the download of the examples data automatically, if files don't exist yet).

Is it because you're behind a proxy, or have a complex network configuration? Or other? 🤔
Need more info.