def rip(self, post_process: Callable[[str], str]):
subs = SubRipFile(
path=str(self.pgs.media_path.translate(extension='srt')))
items = self.pgs.items
confidence = self.confidence
max_width = self.max_tess_width
previous_size = len(items)
while previous_size > 0:
items = self.process(subs, items, post_process, confidence,
max_width)
if not items:
break
current_size = len(items)
if current_size < 20:
max_width = min(
sum([item.width + self.gap[1] for item in items]),
self.max_tess_width)
confidence = 0
remaining_items = self.process(subs, items, post_process,
confidence, max_width)
if remaining_items:
logger.warning(
f'Subtitles were not ripped: {remaining_items!r}')
break
elif current_size > previous_size * 0.8:
max_width = min(
sum([item.width + self.gap[1]
for item in items]), self.max_tess_width) // 2
confidence = max(0, confidence - 5)
previous_size = current_size
subs.clean_indexes()
return subs
def merge_subtitle(sub_a, sub_b, delta, encoding='utf-8'):
"""
合并两种不同言语的srt字幕
因为两个字幕文件的时间轴不一样，所以合并后的字幕会在某一字幕文件转换时生成新的一条字幕，
导致双语字幕并不是同时变化，不过这也是没有办法的事，无法避免
参考https://github.com/byroot/pysrt/issues/17
https://github.com/byroot/pysrt/issues/15
:param sub_a: 使用sub_a = SubRipFile.open(sub_a_path, encoding=encoding)
:param sub_b:
:param delta:
:return:
"""
out = SubRipFile()
intervals = [item.start.ordinal for item in sub_a]
intervals.extend([item.end.ordinal for item in sub_a])
intervals.extend([item.start.ordinal for item in sub_b])
intervals.extend([item.end.ordinal for item in sub_b])
intervals.sort()
j = k = 0
for i in xrange(1, len(intervals)):
start = SubRipTime.from_ordinal(intervals[i - 1])
end = SubRipTime.from_ordinal(intervals[i])
if (end - start) > delta:
text_a, j = find_subtitle(sub_a, start, end, j)
text_b, k = find_subtitle(sub_b, start, end, k)
text = join_lines(text_a, text_b)
if len(text) > 0:
item = SubRipItem(0, start, end, text)
out.append(item)
out.clean_indexes()
return out
def merge_subtitle(sub_a, sub_b, delta, encoding='utf-8'):
"""
合并两种不同言语的srt字幕
因为两个字幕文件的时间轴不一样，所以合并后的字幕会在某一字幕文件转换时生成新的一条字幕，
导致双语字幕并不是同时变化，不过这也是没有办法的事，无法避免
参考https://github.com/byroot/pysrt/issues/17
https://github.com/byroot/pysrt/issues/15
:param sub_a: 使用sub_a = SubRipFile.open(sub_a_path, encoding=encoding)
:param sub_b:
:param delta:
:return:
"""
out = SubRipFile()
intervals = [item.start.ordinal for item in sub_a]
intervals.extend([item.end.ordinal for item in sub_a])
intervals.extend([item.start.ordinal for item in sub_b])
intervals.extend([item.end.ordinal for item in sub_b])
intervals.sort()
j = k = 0
for i in xrange(1, len(intervals)):
start = SubRipTime.from_ordinal(intervals[i - 1])
end = SubRipTime.from_ordinal(intervals[i])
if (end - start) > delta:
text_a, j = find_subtitle(sub_a, start, end, j)
text_b, k = find_subtitle(sub_b, start, end, k)
text = join_lines(text_a, text_b)
if len(text) > 0:
item = SubRipItem(0, start, end, text)
out.append(item)
out.clean_indexes()
return out
def merge_subtitle(sub_a, sub_b, delta):
out = SubRipFile()
intervals = [item.start.ordinal for item in sub_a]
intervals.extend([item.end.ordinal for item in sub_a])
intervals.extend([item.start.ordinal for item in sub_b])
intervals.extend([item.end.ordinal for item in sub_b])
intervals.sort()
j = k = 0
for i in xrange(1, len(intervals)):
start = SubRipTime.from_ordinal(intervals[i-1])
end = SubRipTime.from_ordinal(intervals[i])
if (end-start) > delta:
text_a, j = find_subtitle(sub_a, start, end, j)
text_b, k = find_subtitle(sub_b, start, end, k)
text = join_lines(text_a, text_b)
if len(text) > 0:
item = SubRipItem(0, start, end, text)
out.append(item)
out.clean_indexes()
return out
def merge_subtitle(sub_a, sub_b, delta):
out = SubRipFile()
intervals = [item.start.ordinal for item in sub_a]
intervals.extend([item.end.ordinal for item in sub_a])
intervals.extend([item.start.ordinal for item in sub_b])
intervals.extend([item.end.ordinal for item in sub_b])
intervals.sort()
j = k = 0
for i in range(1, len(intervals)):
start = SubRipTime.from_ordinal(intervals[i - 1])
end = SubRipTime.from_ordinal(intervals[i])
if (end - start) > delta:
text_a, j = find_subtitle(sub_a, start, end, j)
text_b, k = find_subtitle(sub_b, start, end, k)
text = join_lines(text_a, text_b)
if len(text) > 0:
item = SubRipItem(0, start, end, text)
out.append(item)
out.clean_indexes()
return out
chat_channel,
chat_server[0],
chat_server[1],
twitchclient_version=twitchclient_version)
outsrt = SubRipFile()
text = ''
while 1:
raw_msg_list = bot.get_message()
if len(raw_msg_list) > 0:
if len(text) > 0:
end = SubRipTime.from_time(datetime.now())
item = SubRipItem(0, start, end, text)
outsrt.append(item)
start = SubRipTime.from_time(datetime.now())
text = ''
timestamp = get_timestamp(timestamp_format)
for item in raw_msg_list:
if record_raw:
log_add(raw_log_path, timestamp + ' ' + item + '\n')
username, message = irc_bot.parse_user(item)
if username != '':
safe_print(chat_channel + " " + username + ": " + message)
log_add(log_path,
timestamp + ' ' + username + ': ' + message + '\n')
text += username + ": " + message + '\n'
outsrt.clean_indexes()
outsrt.save(srt_log_path, encoding='utf-8')
srt_log_path = current_directory + '/comment_log/' + chat_channel + '.srt' bot = irc_bot.irc_bot(username, oauth, chat_channel, chat_server[0], chat_server[1], twitchclient_version = twitchclient_version) outsrt = SubRipFile() text = '' while 1: raw_msg_list = bot.get_message() if len(raw_msg_list) > 0: if len(text) > 0: end = SubRipTime.from_time(datetime.now()) item = SubRipItem(0, start, end, text) outsrt.append(item) start = SubRipTime.from_time(datetime.now()) text = '' timestamp = get_timestamp(timestamp_format) for item in raw_msg_list: if record_raw: log_add(raw_log_path, timestamp + ' ' + item + '\n') username, message = irc_bot.parse_user(item) if username != '': safe_print(chat_channel + " " + username + ": " + message) log_add(log_path, timestamp + ' ' + username + ': ' + message + '\n') text += username + ": " + message + '\n' outsrt.clean_indexes() outsrt.save(srt_log_path, encoding='utf-8')
def syncSrts(subs_L1, subs_L2):
"""Sync subs_L1 by subs_L2 timings and return a SubRipFile.
"""
out = SubRipFile()
subs_L2_out = SubRipFile()
j = 0
last_j = -1
dupes = 0
L2_ind = -1
for L2_sub in subs_L2:
L2_ind = L2_ind + 1
start = L2_sub.start
end = L2_sub.end
j = matchSubtitle(subs_L1, start, end, max(last_j, 0))
L1_sub = subs_L1[j] if (j > -1) else None
if L1_sub is None:
text = L2_sub.text
print("---- Missing: {}: {}".format(
L2_sub.index, L2_sub.text.replace("\n", "[[NL]]")))
else:
text = L1_sub.text
if j - 1 > last_j and last_j > -1:
# we skipped a sub in L1_subs
if isSubMatch(subs_L1[j - 1], subs_L2[L2_ind - 1].start,
subs_L2[L2_ind - 1].end):
out[len(out) -
1].text = out[len(out) -
1].text + "\n" + subs_L1[j - 1].text
elif isSubMatch(subs_L1[j - 1], start, end):
text = subs_L1[j - 1].text + "\n" + text
else:
# A sub line in L1 does not match any in L2
# We add it to synced L1, and add an empty one to subs L2
item = SubRipItem(0, subs_L1[j - 1].start,
subs_L1[j - 1].end, subs_L1[j - 1].text)
out.append(item)
item2 = SubRipItem(0, subs_L1[j - 1].start,
subs_L1[j - 1].end, " ")
subs_L2_out.append(item2)
if j == last_j:
dupes = dupes + 1
#print("---- OOPS. {}: {} - {}".format(L2_sub.index, L2_sub.text.replace("\n",""), L1_sub.text.replace("\n","")))
last_j = j
item = SubRipItem(0, start, end, text)
out.append(item)
item2 = SubRipItem(0, start, end, L2_sub.text)
subs_L2_out.append(item2)
out.clean_indexes()
subs_L2_out.clean_indexes()
fixed = 0
for i in range(1, len(out)):
sub1 = out[i - 1].text
sub2 = out[i].text
if ((sub1 == sub2)
and (subs_L2_out[i - 1].text != subs_L2_out[i].text)):
if (trySplitLine(out, i, sub1)):
fixed = fixed + 1
i = i + 1
else:
print("---- Oy. {}: {} not fixed".format(
i, sub1.replace("\n", "[[NL]]")))
return out, dupes, fixed, subs_L2_out
srt = SubRipFile(eol='\n', encoding='utf-8')
i = 1
for line in sublog:
line = line.split(",", 1)
if (line[0] and line[0][0] == '-'):
if (START_TIME == None and line[0][:8] == '- start '):
START_TIME = datetime.strptime(line[0], '- start ' + TIMEFORMAT +
'\n')
continue
no = datetime.strptime(line[0], TIMEFORMAT) - START_TIME
if (abs(no) > timedelta(1)):
print("\nCan't go over a day in a subtitle! Delete non-used lines in" + \
" log.\nLet there only be one '- start' line at the top of" + \
" the log-file.")
sys.exit(1)
time = SubRipTime.from_ordinal(no.seconds*1000 + no.microseconds*0.001)
item = SubRipItem(i, start=time, end=time + 30*1000,
text=unicode(line[1], 'utf-8'))
srt.append(item)
i += 1
srt.clean_indexes()
#srt.save(path=sys.stdout)
for line in srt:
sys.stdout.write(unicode(line).encode('utf-8'))
