Allow configuring transformations to be applied to the filename before

parsing. The motivation was that files produced by a Brother scanner
wouldn't match paperless' expectations. At most one transformation
is applied (first matching). It won't affect the filename on disk.

This is generic enough so that it is useful for various purposes.
In my case it allows me to use the different hardware buttons on
the scanner to use different profiles, feeding one instance of
paperless with documents of multiple entities and tagging them
accordingly.

Example:

PAPERLESS_FILENAME_PARSE_TRANSFORMS=[{"pattern":"^([a-z]+)_(\\d{8})_(\\d{6})_([0-9]+)\\.", "repl":"\\2\\3Z - \\4 - \\1."},{"pattern":"^([a-z]+)_([0-9]+)\\.", "repl":" - \\2 - \\1."}]

paperless.conf.example

@@ -135,6 +135,23 @@ PAPERLESS_EMAIL_SECRET=""
 # as normal.
 #PAPERLESS_FILENAME_DATE_ORDER="YMD"
 
+# Sometimes devices won't create filenames which can be parsed properly
+# by the filename parser (see
+# https://paperless.readthedocs.io/en/latest/guesswork.html).
+#
+# This setting allows to specify a list of transformations
+# in regular expression syntax, which are passed in order to re.sub.
+# Transformation stops after the first match, so at most one transformation
+# is applied.
+#
+# Syntax is a JSON array of dictionaries containing "pattern" and "repl"
+# as keys.
+#
+# The example below transforms filenames created by a Brother ADS-2400N
+# document scanner in its standard configuration `Name_Date_Count', so that
+# count is used as title, name as tag and date can be parsed by paperless.
+#PAPERLESS_FILENAME_PARSE_TRANSFORMS=[{"pattern":"^([a-z]+)_(\\d{8})_(\\d{6})_([0-9]+)\\.", "repl":"\\2\\3Z - \\4 - \\1."}]
+
 #
 # The following values use sensible defaults for modern systems, but if you're
 # running Paperless on a low-resource device (like a Raspberry Pi), modifying