Feature Request: Support for truncated hashes in DYNAMIC modes

It would be useful to have a standard function to call when creating a custom DYNAMIC mode that would allow truncating generated hashes, either before being further processed, or compared against the final target hash-list supplied by the user.

Reasons:

1. This would be useful to generate 'vanity hashes', such as finding a hash that starts with '123456'.
2. This would also be useful to quickly target malformed and weird/custom hash formats as they are encountered. An example of this was the original LinkedIn password dump where many of the Sha1 hashes had the first couple of bytes be 0. Note, in my head it would be up to the user to format the target hash list, (for example cutting off the leading 0's in the LinkedIn list), before running JtR against them. The DYNAMIC enhancement would just in this instance only check a reduced Sha1 hash against the target values.

Start by reading the existing DYNAMIC mode entry points and how generated hashes are processed and compared with the user-supplied target hash list. Done means custom DYNAMIC modes can request truncation before further processing or final comparison, supporting vanity hashes and malformed or custom hash formats.