Add hook for input canonicalization

[greschd]

First, we should collect cases where the current method fails, to get a better view of the requirements.

[greschd]

One way of implementing canonicalization would be taking a mapping of (possibly globbed) filenames to functions which canonicalize the file content, and apply this when reading the file content (in the hashing function).

Common canonicalization functions (e.g. strip certain lines) could be provided.