When "metadata_regex" does NOT match, the actions in "metadata_postprocessing" are still processed

[kintrupf]

As described in the title, even when the regex from "metadata_regex" does not match anything in the document, the actions in "metadata_postprocessing" are still processed. The results are predictably bad, because any contents from the regex is missing.
When called from the command line paperlessngx_postprocessor.py recognizes the non-matching regex and displays a warning, but processes the actions anyway.

So if the filter in "match" is not specific enough (which in many cases it can't be) the actions in "metadata_postprocessing" will trash the title and the create_date of documents where the regex did not match and for which the documents were not intended.

Please add a rule that documents with no match from "metadata_regex" are not processed.
In this form I find paperlessngx_postprocessor.py much to unpredictable for many automations which would required a regex-match.