def simplify_ass(cfg):
# subs = SSAFile.load('S03E09.ass', encoding="utf-16-le")
subs = SSAFile.load(cfg.OUT)
last_txt = None
last_index = 0
i = 0
length = len(subs)
while i < length:
line = subs[i]
if not is_contain_chinese(line.text):
subs.__delitem__(i)
length -= 1
elif line.text == last_txt:
subs[last_index].end = line.end
subs.__delitem__(i)
length -= 1
else:
last_index = i
last_txt = line.text
i += 1
subs.save(cfg.OUT)
subtitle_ass: SSAFile = None
# work_dir = 'D:\BaiduYunDownload\Hyouka\Hyouka-01'
# work_dir = 'D:\BaiduYunDownload\Oreimo\Oreimo-02'
# work_dir = 'D:\BaiduYunDownload\GirlsLastTour\GirlsLastTour-01'
# work_dir = 'D:\BaiduYunDownload\SAOII\mkv\SAOII-08'
work_dir = 'D:\BaiduYunDownload\Children\Children-02'
file_list = os.listdir(work_dir)
i = 1
for filename in file_list:
file_path = os.path.join(work_dir, filename)
if os.path.isfile(file_path):
i += 1
sub = SSAFile.load(file_path)
if subtitle_ass:
if filename == 'c.ass' or filename == 'm.ass':
sub.shift(s=1)
subtitle_ass += sub
else:
if filename == 'c.ass' or filename == 'm.ass':
sub.shift(s=1)
subtitle_ass = sub
# for parent, dirnames, filenames in os.walk(work_dir, followlinks=True):
# for filename in filenames:
# file_path = os.path.join(parent, filename)
# sub = SSAFile.load(file_path)
# if subtitle_ass:
# subtitle_ass += sub
# else:
def main(argv=None):
parser = argparse.ArgumentParser(
prog=ENTRYPOINT,
description=
"Checks whether all characters in ASS file exist in declared fonts")
parser.add_argument("files", nargs="+")
parser.add_argument("--log-level",
"-l",
default="info",
choices=("debug", "info", "warn", "error"))
args = parser.parse_args(argv)
logging.basicConfig(level=args.log_level.upper())
log = logging.getLogger(ENTRYPOINT)
contexts = {}
have_nonexistent_char = False
def get_context(key):
if key not in contexts:
log.debug("New font: %s", key)
contexts[key] = FontContext(**key._asdict())
return contexts[key]
for name in args.files:
ssa = SSAFile.load(name)
line_number = 0
for event in ssa.events:
line_number += 1
if "type=Comment" in str(event):
continue
style = ssa.styles[event.style]
key = key_from_style(style)
context = get_context(key)
text = event.text
while text:
m = TAG_RE.search(text)
display_text = text if not m else text[:m.start()]
log.debug("Text block: %s", display_text)
result = context.check(display_text)
if result:
have_nonexistent_char = True
log.warning("%s Dialogue #%s: [%s] does not exist in %s",
name, line_number, "".join(result), key)
text = "" if not m else text[m.end():]
if not text:
break
ovr_tags = m.group(0)[1:-1].split("\\")
log.debug("Override tags: %s", ovr_tags)
for tag in ovr_tags:
tag = tag.rstrip()
if tag[:1].lower() == "r":
if len(tag) == 1:
key = key_from_style(style)
else:
style_str = tag[1:]
key = key_from_style(ssa.styles[style_str])
continue
tag_match = TAG_PART_RE.match(tag)
if not tag_match:
continue
type = tag_match.group(1).lower()
if type == "fn":
key = key._replace(name=tag_match.group(2))
elif type == "b":
key = key._replace(bold=bool(int(tag_match.group(2))))
elif type == "i":
key = key._replace(
italics=bool(int(tag_match.group(2))))
context = get_context(key)
sys.exit(1 if have_nonexistent_char else 0)
def add_word(self,
word,
collection,
start,
end,
name,
add_type,
word_type,
group,
word_id='',
wordset_id=''):
clean_word = word.strip()
puresave_filename = name.split('.')[0] + "~" + clean_word
# row = {
# "videaname":puresave_filename,
# "wordbase_type":"video"
# }
# w = WordbaseHelper()
# w.init_word(row,clean_word)
# w.insert(row,collection)
data = {
'filename': puresave_filename,
'wordbase_collection': collection,
'word': word,
'add_type': add_type,
'word_type': word_type,
'group': group,
'word_id': word_id,
'wordset_id': wordset_id,
}
work_dir = "D:\BaiduYunDownload"
file_path = ""
parent_path = ""
double_loop_flag = False
for parent, dirnames, filenames in os.walk(work_dir, followlinks=True):
for filename in filenames:
if filename == name:
parent_path = parent
file_path = os.path.join(parent, filename)
double_loop_flag = True
break
if double_loop_flag:
break
start_time = float(start)
end_time = float(end)
pure_filename = name.split('.')[0]
subfile_path = os.path.join(parent_path, pure_filename + ".srt")
video_clip = VideoFileClip(file_path)
clip = video_clip.subclip(start_time, end_time)
target = "D:\BaiduYunDownload\\videos\\" + puresave_filename + ".mp4"
clip.write_videofile(target,
codec='libx264',
verbose=False,
audio=True)
video_clip.close()
subtitle = SSAFile.load(subfile_path)
text = '''
1
00:00:00,000 --> 00:00:00,000
'''
temp = SSAFile().from_string(text)
for sub in subtitle:
if sub.start >= start_time * 1000 and sub.end <= end_time * 1000:
text = sub.text.replace(
clean_word, '<c.video-heightlight>' + clean_word + '</c>')
sub.text = text
sub.shift(s=-start_time)
temp.append(sub)
sub_target = "D:\BaiduYunDownload\\videos\\" + puresave_filename
temp.save(sub_target + '.srt')
vtt = WebVTT().from_srt(sub_target + '.srt')
vtt.save(sub_target + '.vtt')
files = {
"video": open(target, "rb"),
"subtitle": open(sub_target + '.vtt', "rb")
}
# print(files)
# r = requests.post('http://127.0.0.1:5000/video', data=data,files=files)
r = requests.post('http://' + server_ip + '/video',
data=data,
files=files)
# print(r.request)
return "true"
def getalltext(filename):
relustarray = []
subs = SSAFile.load(filename)
for line in subs:
relustarray.append(line.text)
return relustarray
def ocr_with_timeline(video, video_path, box, ocr_reader, lang, main_window, progress_bar):
fps = video.get(5)
_, video_name = os.path.split(video_path)
video_name = video_name.split('.')[0]
frame_dir = 'frame/'+video_name+'/'
subs = SSAFile.load(frame_dir+'/split_vision.ass')
prefix = video_path[:-len(video_path.split('.')[-1])]
total = len(subs)
start = time.time()
count = 0
re_chinese = re.compile(u"[\u4e00-\u9fa5]+")
re_ascii = re.compile(r'\w+', re.ASCII)
i = 0
length = len(subs)
while i < length:
line = subs[i]
srt_start = line.start / 1000
srt_end = line.end / 1000
# print(fmt_time(srt_start), fmt_time(srt_end))
srt_mid = (srt_start + srt_end) / 2
frame_id = int(srt_mid * fps)
video.set(cv.CAP_PROP_POS_FRAMES, frame_id) # 设置要获取的帧号
_, frame = video.read()
clipped_frame = frame[box[0][0]:box[0][1], box[1][0]:box[1][1]]
result = ocr_reader.ocr(clipped_frame)
print(result)
if len(result) == 0:
subs.__delitem__(i)
length -= 1
else:
# subs[i].text = result[0]
if len(lang) == 2:
split = list(map(lambda x: len(x), [re.findall(re_chinese,result[i]) for i in range(len(result))]))
eng_str = []
ch_str = []
iseng = 1
for str_ind in reversed(range(len(result))):
iseng += split[str_ind]
if iseng == 1:
eng_str.append(result[str_ind])
else:
ch_str.append(result[str_ind])
eng_str = ' '.join(reversed(eng_str))
ch_str = ''.join(reversed(ch_str))
subs[i].text = re.findall(re_chinese, ch_str)[0]
subs[i+1].text = ' '.join(re.findall(re_ascii, eng_str))
else:
subtitle = ''.join(result)
subs[i].text = re.findall(re_chinese, subtitle)[0]
i += len(lang)
count += len(lang)
progress_bar.setValue(int(count / total * 100) + 1)
if (count % 1) == 0 or count == total:
elapsed = time.time() - start
eta = (total - count) / count * elapsed
print("[{}/{}], Elapsed: {}, ETA: {}".format(count, total, fmt_time(elapsed), fmt_time(eta)))
subs.save(prefix+'ass', format_='ass')
progress_bar.setValue(100)
QMessageBox.information(main_window, "提示", "字幕生成成功！", QMessageBox.Yes | QMessageBox.No, QMessageBox.Yes)
progress_bar.setValue(0)
def split_subtitle():
path = './CHS_test.srt'
subs = SSAFile.load(path)
print(subs[0].plaintext)
cv.imwrite(
"{}/f{}_l{}.jpg".format(frame_dir, fc_start,
fc - fc_start), last_srt_frame)
if output_segged_frame:
cv.imwrite(
"{}/f{}_l{}_segged.jpg".format(
frame_dir, fc_start, fc - fc_start),
last_srt_seg_frame)
srt_count += 1
fc_start = 0
progress_bar.setValue(int(fc / frames_num * 100) + 1)
# progress_bar.format(Q)
elapsed = time.time() - time_start
eta = (frames_num - fc) * elapsed / fc if fc > 0 else 0
print('[%d/%d] Elapsed: %s, ETA: %s' %
(fc, frames_num, fmt_time(elapsed), fmt_time(eta)))
video_path.split('/')
subs.save(frame_dir + '/split_vision.ass')
progress_bar.setValue(100)
QMessageBox.information(main_window, "提示", "时间轴生成成功！",
QMessageBox.Yes | QMessageBox.No, QMessageBox.Yes)
progress_bar.setValue(0)
if __name__ == "__main__":
sub = SSAFile.load('demo/empty.ass')
sub.append(
SSAEvent(start=0, end=make_time(s=2.5), text="New first subtitle"))
print("test")
