Support downloading seed file from URL

[tw4l]

Fixes #841

Crawler work toward long URL lists in Browsertrix. This PR moves seed handling from the arg parser's validation step to the crawler's bootstrap step in order to be able to async fetch the seed file from a URL.

[tw4l]

It looks like GitHub always returns a content-type of text/plain for raw files such as https://raw.githubusercontent.com/webrecorder/browsertrix-crawler/refs/heads/main/tests/custom-behaviors/custom-2.js, so we may want to compute the mime type from what we receive rather than relying on what the web server says it is.

This might be harder than it appears at first glance. mime and mime-types check only by extension, while file-type checks by magic number and so doesn't work well for text-based files. We may be better off reverting back to checking file extension, or at least doing so as a fallback when the content-type header doesn't match expected values.

Alternatively we could use a file characterization tool like Siegfried, which may net much more accurate results once the content is written to a file.

[tw4l]

@ikreymer I reverted the MIME check changes and think this should be good to go now.

I'm not sure that we need to enforce a .txt extension for seed files, since the crawler will later fail anyway if it's not able to parse any valid seeds from what's passed. It seems like it might be nice for the crawler to be flexible enough to accept a file that doesn't have an extension, even if we do enforce that in the Browsertrix UI.