Extract: Skip streams if all files are present.

Migrated Jira Issue CDDSO-89
Jira creation date 22/04/2022 13:47
-------
When re-running extract, following failure, if the same set of streams are used it is possible for extract to spend a long time (several hours in the case of the piControl) attempting to "fill in gaps". This is inefficient in both our time and load on MASS resources.

I propose the following solution; At the start of processing a stream/substream, extract checks for existing files for that stream/substream and if all expected files are present skips on to the next.
-------

Start by locating the extract entry point that processes streams and substreams, then trace how expected output files are identified after a failed run. Confirm how the current gap-filling work is triggered and identify the relevant tests, if any. Done means a rerun skips a stream or substream only when all expected files are already present.

Written by the indexing model from the issue text.