breaks on long filenames about shrine-url

[jrochkind]

Yeah, so the long filename is just a symptom, really the problem is that when used with shrine-url, the query params are making their way into the id (and thus subsequent storage path), when they shoudln't be.

I worked around in my custom code with a custom generate_location override that includes this (among other things)

        basename = super
          if io.respond_to?(:storage) && io.storage.is_a?(Shrine::Storage::Url)
            basename = URI.parse(basename).path
          end

I am not quite sure what to fix how in shrine itself though, more cleanly.

[janko]

I would try to add a failing test... but I notice now shrine-url doesn't even have any tests, so not sure where to add it.

The tests are here: https://github.com/shrinerb/shrine-url/blob/master/test/url_test.rb 😉

Or wait, maybe it looks like the part of shrine that's extracting the filename suffix (to preserve it as it gets promoted), needs to be smart enough to ignore query parameter at the end, in order to be compatible with shrine-url? That might be it?

Yeah, good catch, that seems to be it:

[1] pry(main)> File.extname("filename.tif?x-amz-algorithm=aws4-hmac-sha256&x-amz-credential=akiajorcl6txivhsdrhq%2f20181119%2fus-east-1%2fs3%2faws4_request&x-amz-date=20181119t182037z&x-amz-expires=14400&x-amz-signedheaders=host&x-amz-signature=fakefake45db944222adfadfadfadf7d20eacce3e6d062979f993db0ffa1e6e")
=> ".tif?x-amz-algorithm=aws4-hmac-sha256&x-amz-credential=akiajorcl6txivhsdrhq%2f20181119%2fus-east-1%2fs3%2faws4_request&x-amz-date=20181119t182037z&x-amz-expires=14400&x-amz-signedheaders=host&x-amz-signature=fakefake45db944222adfadfadfadf7d20eacce3e6d062979f993db0ffa1e6e"

Feel free to send the PR if you can, thanks 😃

[jrochkind]

I wonder if a question mark could hypothetically be a real file suffix. This gets weirdly complicated to do reliably, but I'm going to PR something cause I think it needs fixing.