def test_date(self):
self.assertEqual(search_year(' in the year 2000... '), (2000, (13, 17)))
self.assertEqual(search_year(' they arrived in 1492. '), (None, None))
today = date.today()
today_year_2 = int(str(today.year)[2:])
future = today + timedelta(days=1000)
future_year_2 = int(str(future.year)[2:])
past = today - timedelta(days=10000)
past_year_2 = int(str(past.year)[2:])
self.assertEqual(search_date(' Something before 2002-04-22 '), (date(2002, 4, 22), (18, 28)))
self.assertEqual(search_date(' 2002-04-22 Something after '), (date(2002, 4, 22), (1, 11)))
self.assertEqual(search_date(' This happened on 2002-04-22. '), (date(2002, 4, 22), (18, 28)))
self.assertEqual(search_date(' This happened on 22-04-2002. '), (date(2002, 4, 22), (18, 28)))
self.assertEqual(search_date(' This happened on 13-04-%s. ' % (today_year_2,)), (date(today.year, 4, 13), (18, 26)))
self.assertEqual(search_date(' This happened on 22-04-%s. ' % (future_year_2,)), (date(future.year, 4, 22), (18, 26)))
self.assertEqual(search_date(' This happened on 20-04-%s. ' % (past_year_2)), (date(past.year, 4, 20), (18, 26)))
self.assertEqual(search_date(' This happened on 13-06-14. ', year_first=True), (date(2013, 6, 14), (18, 26)))
self.assertEqual(search_date(' This happened on 13-05-14. ', year_first=False), (date(2014, 5, 13), (18, 26)))
self.assertEqual(search_date(' This happened on 04-13-%s. ' % (today_year_2,)), (date(today.year, 4, 13), (18, 26)))
self.assertEqual(search_date(' This happened on 04-22-%s. ' % (future_year_2,)), (date(future.year, 4, 22), (18, 26)))
self.assertEqual(search_date(' This happened on 04-20-%s. ' % (past_year_2)), (date(past.year, 4, 20), (18, 26)))
self.assertEqual(search_date(' This happened on 35-12-%s. ' % (today_year_2,)), (None, None))
self.assertEqual(search_date(' This happened on 37-18-%s. ' % (future_year_2,)), (None, None))
self.assertEqual(search_date(' This happened on 44-42-%s. ' % (past_year_2)), (None, None))
self.assertEqual(search_date(' This happened on %s. ' % (today, )), (today, (18, 28)))
self.assertEqual(search_date(' This happened on %s. ' % (future, )), (future, (18, 28)))
self.assertEqual(search_date(' This happened on %s. ' % (past, )), (past, (18, 28)))
self.assertEqual(search_date(' released date: 04-03-1901? '), (None, None))
self.assertEqual(search_date(' There\'s no date in here. '), (None, None))
self.assertEqual(search_date(' Something 01-02-03 '), (date(2003, 2, 1), (11, 19)))
self.assertEqual(search_date(' Something 01-02-03 ', year_first=False, day_first=True), (date(2003, 2, 1), (11, 19)))
self.assertEqual(search_date(' Something 01-02-03 ', year_first=True), (date(2001, 2, 3), (11, 19)))
self.assertEqual(search_date(' Something 01-02-03 ', day_first=False), (date(2003, 1, 2), (11, 19)))
def guess_year_skip_first(string):
year, span = search_year(string)
if year:
year2, span2 = guess_year(string[span[1]:])
if year2:
return year2, (span2[0] + span[1], span2[1] + span[1])
return None, None
def guess_year_skip_first(string):
year, span = search_year(string)
if year:
year2, span2 = guess_year(string[span[1]:])
if year2:
return year2, (span2[0]+span[1], span2[1]+span[1])
return None, None
def guess_year(string):
year, span = search_year(string)
if year:
return {'year': year}, span
else:
return None, None
def guess_year(string, node=None, options=None):
year, span = search_year(string)
if year:
return {"year": year}, span
else:
return None, None
def guess_year(string):
year, span = search_year(string)
if year:
return { 'year': year }, span
else:
return None, None
def guess_year(self, string, node=None, options=None):
year, span = search_year(string)
if year:
return {'year': year}, span
else:
return None, None
def guess_year(self, string):
year, span = search_year(string)
if year:
return {"year": year}, span
else:
return None, None
def guess_groups(string, result, filetype):
# add sentinels so we can match a separator char at either end of
# our groups, even when they are at the beginning or end of the string
# we will adjust the span accordingly later
#
# filetype can either be movie, moviesubtitle, episode, episodesubtitle
current = " " + string + " "
regions = [] # list of (start, end) of matched regions
def guessed(match_dict, confidence):
guess = format_guess(Guess(match_dict, confidence=confidence))
result.append(guess)
log.debug("Found with confidence %.2f: %s" % (confidence, guess))
return guess
def update_found(string, guess, span, span_adjust=(0, 0)):
span = (span[0] + span_adjust[0], span[1] + span_adjust[1])
regions.append((span, guess))
return blank_region(string, span)
# try to find dates first, as they are very specific
date, span = search_date(current)
if date:
guess = guessed({"date": date}, confidence=1.0)
current = update_found(current, guess, span)
# for non episodes only, look for year information
if filetype not in ("episode", "episodesubtitle"):
year, span = search_year(current)
if year:
guess = guessed({"year": year}, confidence=1.0)
current = update_found(current, guess, span)
# specific regexps (ie: cd number, season X episode, ...)
for rexp, confidence, span_adjust in video_rexps:
match = re.search(rexp, current, re.IGNORECASE)
if match:
metadata = match.groupdict()
# is this the better place to put it? (maybe, as it is at least the soonest that we can catch it)
if "cdNumberTotal" in metadata and metadata["cdNumberTotal"] is None:
del metadata["cdNumberTotal"]
guess = guessed(metadata, confidence=confidence)
current = update_found(current, guess, match.span(), span_adjust)
if filetype in ("episode", "episodesubtitle"):
for rexp, confidence, span_adjust in episode_rexps:
match = re.search(rexp, current, re.IGNORECASE)
if match:
metadata = match.groupdict()
guess = guessed(metadata, confidence=confidence)
current = update_found(current, guess, match.span(), span_adjust)
# Now websites, but as exact string instead of regexps
clow = current.lower()
for site in websites:
pos = clow.find(site.lower())
if pos != -1:
guess = guessed({"website": site}, confidence=confidence)
current = update_found(current, guess, (pos, pos + len(site)))
clow = current.lower()
# release groups have certain constraints, cannot be included in the previous general regexps
group_names = [
r"\.(Xvid)-(?P<releaseGroup>.*?)[ \.]",
r"\.(DivX)-(?P<releaseGroup>.*?)[\. ]",
r"\.(DVDivX)-(?P<releaseGroup>.*?)[\. ]",
]
for rexp in group_names:
match = re.search(rexp, current, re.IGNORECASE)
if match:
metadata = match.groupdict()
metadata.update({"videoCodec": match.group(1)})
guess = guessed(metadata, confidence=0.8)
current = update_found(current, guess, match.span(), span_adjust=(1, -1))
# common well-defined words and regexps
confidence = 1.0 # for all of them
for prop, value, pos, end in find_properties(current):
guess = guessed({prop: value}, confidence=confidence)
current = update_found(current, guess, (pos, end))
# weak guesses for episode number, only run it if we don't have an estimate already
if filetype in ("episode", "episodesubtitle"):
if not any("episodeNumber" in match for match in result):
for rexp, _, span_adjust in weak_episode_rexps:
match = re.search(rexp, current, re.IGNORECASE)
if match:
metadata = match.groupdict()
epnum = int(metadata["episodeNumber"])
if epnum > 100:
guess = guessed({"season": epnum // 100, "episodeNumber": epnum % 100}, confidence=0.6)
else:
guess = guessed(metadata, confidence=0.3)
current = update_found(current, guess, match.span(), span_adjust)
# try to find languages now
language, span, confidence = search_language(current)
while language:
# is it a subtitle language?
if "sub" in clean_string(current[: span[0]]).lower().split(" "):
guess = guessed({"subtitleLanguage": language}, confidence=confidence)
else:
guess = guessed({"language": language}, confidence=confidence)
current = update_found(current, guess, span)
language, span, confidence = search_language(current)
# remove our sentinels now and ajust spans accordingly
assert current[0] == " " and current[-1] == " "
current = current[1:-1]
regions = [((start - 1, end - 1), guess) for (start, end), guess in regions]
# split into '-' separated subgroups (with required separator chars
# around the dash)
didx = current.find("-")
while didx > 0:
regions.append(((didx, didx), None))
didx = current.find("-", didx + 1)
# cut our final groups, and rematch the guesses to the group that created
# id, None if it is a leftover group
region_spans = [span for span, guess in regions]
string_groups = split_on_groups(string, region_spans)
remaining_groups = split_on_groups(current, region_spans)
guesses = []
pos = 0
for group in string_groups:
found = False
for span, guess in regions:
if span[0] == pos:
guesses.append(guess)
found = True
if not found:
guesses.append(None)
pos += len(group)
return zip(string_groups, remaining_groups, guesses)
