Add zip extracting if file is zipped.

[cimentadaj]

If the the file we're reading is a zip file, we could add an unzipping stage and then just apply the regular reading pipeline for the files inside the zip (if it's a permitted format, then we read, etc..). However, this is trickier because if it's not readable, we always return the attempted URL that failed but this would not have a URL because it's already downloaded.

Something to think about.