This is only used in logging to identify which file
may have caused problems. While it is probably best
to use unique ids for the sake of debugging, it is not
necessary that the ids be unique. This id
is never used as a hashkey by the batch processors, for example.
This gets the metadata available before the parsing of the file.
This will typically be "external" metadata: file name,
file size, file location, data stream, etc. That is, things
that are known about the file from outside information, not