Github/guess-filename.py
2
0
Fork 1
mirror of https://github.com/novoid/guess-filename.py.git synced 2026-02-16 13:24:15 +00:00
Code Issues Projects Releases Packages Wiki Activity

added Konica Minolta scan filename timestamp

Browse source
This commit is contained in:
Karl Voit 2020-05-29 16:28:44 +02:00
parent d7fa57777b
commit 71e327bcf6
2 changed files with 19 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

14
guessfilename/__init__.py
View file

@ -1,6 +1,6 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
PROG_VERSION = u"Time-stamp: <2020-03-05 15:12:04 vk>"
PROG_VERSION = u"Time-stamp: <2020-05-29 16:28:30 vk>"
# TODO:
@ -168,6 +168,10 @@ class GuessFilename(object):
VID_REGEX = re.compile('VID_' + DATESTAMP_REGEX + '_' + TIMESTAMP_REGEX + \
'(?P<description>.+)?\.(?P<extension>mp4)', re.UNICODE)
# Konica Minolta scan file-names: YYMMDDHHmmx
KonicaMinolta_TIME_REGEX = re.compile('(?P<truncatedyear>\d{2})(?P<month>[01]\d)(?P<day>[0123]\d)(?P<hour>[012]\d)(?P<minute>[012345]\d)(?P<index>\d).pdf')
# 2019-12-04: "Die Presse (31.10.2019) - Unknown.pdf" -> "2019-10-31 Die Presse.pdf"
NEWSPAPER1_REGEX = re.compile('(?P<description>.+) \((?P<day>\d{2})\.(?P<month>\d{2})\.(?P<year>\d{4})\)(?P<misc>.*)\.(?P<extension>pdf)', re.UNICODE)
@ -683,6 +687,14 @@ class GuessFilename(object):
if extension.upper() == "PDF" and datetimestr and 'Anwesenheitsbest' in oldfilename:
return datetimestr + ' BHAK Anwesenheitsbestaetigung -- scan.' + extension
# 2020-05-29: Konica Minolta scan file-names: YYMMDDHHmmx
# KonicaMinolta_TIME_REGEX = re.compile('(?P<truncatedyear>\d{2})(?P<month>[01]\d)(?P<day>[0123]\d)(?P<hour>[012]\d)(?P<minute>[012345]\d)(?P<index>\d).pdf')
regex_match = re.match(self.KonicaMinolta_TIME_REGEX, oldfilename)
if regex_match:
## re-use index number at the end as first digit of seconds and hope that not more than 5 documents are scanned within a minute:
return '20' + regex_match.group('truncatedyear') + '-' + regex_match.group('month') + '-' + regex_match.group('day') + 'T' + \
regex_match.group('hour') + '.' + regex_match.group('minute') + '.' + regex_match.group('index') + "0 -- scan.pdf"
# FIXXME: more cases!

7
guessfilename_test.py
View file

@ -1,6 +1,6 @@
#!/usr/bin/env python3
# -*- coding: utf-8; mode: python; -*-
# Time-stamp: <2020-03-05 15:12:07 vk>
# Time-stamp: <2020-05-29 16:28:30 vk>
import unittest
import logging
@ -989,6 +989,11 @@ class TestGuessFilename(unittest.TestCase):
self.assertEqual(self.guess_filename.derive_new_filename_from_old_filename('2020-03-03 Anwesenheitsbestaetigung.pdf'),
'2020-03-03 BHAK Anwesenheitsbestaetigung -- scan.pdf')
# 2020-05-29: Konica Minolta scan file-names: YYMMDDHHmmx
self.assertEqual(self.guess_filename.derive_new_filename_from_old_filename('20052915100.pdf'),
'2020-05-29T15.10.00 -- scan.pdf')
# self.assertEqual(self.guess_filename.derive_new_filename_from_old_filename(''),
# '')