104 lines
3.3 KiB
Python
104 lines
3.3 KiB
Python
|
import sys
|
||
|
|
import time
|
||
|
|
import json
|
||
|
|
from os import path
|
||
|
|
|
||
|
|
basepath = "/home/user/mnf/project/MNF/conversion/subtitling"
|
||
|
|
|
||
|
|
# filenames = sys.argv[1]
|
||
|
|
# movie_path = sys.argv[2]
|
||
|
|
# current = basepath + "/"+movie_path
|
||
|
|
|
||
|
|
|
||
|
|
def aws_final(current):
|
||
|
|
|
||
|
|
print("$"*50)
|
||
|
|
|
||
|
|
def formatTime(t):
|
||
|
|
seconds, remainder_of_seconds = t.split('.')
|
||
|
|
result = time.strftime('%H:%M:%S', time.gmtime(int(seconds)))
|
||
|
|
|
||
|
|
return result + "," + remainder_of_seconds.ljust(3, '0')
|
||
|
|
|
||
|
|
chunks = []
|
||
|
|
chunk = {
|
||
|
|
'start_time': '',
|
||
|
|
'end_time': '',
|
||
|
|
'word_index': 1,
|
||
|
|
'sentence': ''
|
||
|
|
}
|
||
|
|
word_break_limit = 11
|
||
|
|
|
||
|
|
# if len(sys.argv) < 2:
|
||
|
|
# sys.exit('Please provide a file name.')
|
||
|
|
|
||
|
|
filename = current + '/output.json'
|
||
|
|
|
||
|
|
if not path.exists(filename):
|
||
|
|
sys.exit('File ' + filename + ' does not exist.')
|
||
|
|
|
||
|
|
with open(filename, encoding='utf8') as f:
|
||
|
|
data = json.load(f)
|
||
|
|
|
||
|
|
items = data['results']['items']
|
||
|
|
|
||
|
|
last_word_end_time = None
|
||
|
|
|
||
|
|
for i, item in enumerate(items):
|
||
|
|
type = item['type']
|
||
|
|
content = item['alternatives'][0]['content']
|
||
|
|
|
||
|
|
if type == "pronunciation":
|
||
|
|
item_start_time = item['start_time']
|
||
|
|
|
||
|
|
if last_word_end_time and item_start_time < last_word_end_time:
|
||
|
|
item_start_time = last_word_end_time
|
||
|
|
|
||
|
|
if chunk['start_time'] == "":
|
||
|
|
chunk['start_time'] = item_start_time
|
||
|
|
|
||
|
|
chunk['end_time'] = item['end_time']
|
||
|
|
# Don't want to start a fresh sentence with a space
|
||
|
|
spacer = '' if chunk['word_index'] == 1 else ' '
|
||
|
|
chunk['sentence'] = chunk['sentence'] + spacer + content
|
||
|
|
chunk['word_index'] = chunk['word_index'] + 1
|
||
|
|
last_word_end_time = item['end_time']
|
||
|
|
|
||
|
|
elif type == "punctuation":
|
||
|
|
# Add punctuation
|
||
|
|
# But don't increment index
|
||
|
|
chunk['sentence'] = chunk['sentence'] + content
|
||
|
|
|
||
|
|
# - we hit the word break limit (provided the NEXT item is not punctuation)
|
||
|
|
item_is_ending_punctuation = content == '.' or content == '?' or content == '!'
|
||
|
|
next_item_is_punctuation = i < len(
|
||
|
|
items) - 1 and items[i+1]['type'] == "punctuation"
|
||
|
|
hit_word_break_limit = chunk['word_index'] >= word_break_limit and not next_item_is_punctuation
|
||
|
|
is_last_item = len(items) - 1 == i
|
||
|
|
|
||
|
|
if item_is_ending_punctuation or hit_word_break_limit or is_last_item:
|
||
|
|
|
||
|
|
chunks.append(chunk)
|
||
|
|
chunk = {
|
||
|
|
'start_time': '',
|
||
|
|
'end_time': '',
|
||
|
|
'word_index': 1,
|
||
|
|
'sentence': ''
|
||
|
|
}
|
||
|
|
|
||
|
|
srt = ''
|
||
|
|
# Build out srt
|
||
|
|
for i, chunk in enumerate(chunks):
|
||
|
|
chunk_index = str(i + 1)
|
||
|
|
srt = srt + chunk_index + "\n"
|
||
|
|
srt = srt + formatTime(chunk['start_time']) + " --> " + \
|
||
|
|
formatTime(chunk['end_time']) + "\n" + chunk['sentence'] + "\n\n"
|
||
|
|
|
||
|
|
# sys.stdout = open(rf"{current}/aws_subtitle.srt", "w", encoding='utf8')
|
||
|
|
# aws_subtitle_file = open(
|
||
|
|
# rf"{current}/aws_subtitle.srt", "w", encoding='utf8').write
|
||
|
|
open(rf"{current}/a_subtitles.srt", "w", encoding='utf8').write(srt)
|
||
|
|
# aws_subtitle_file.write(srt)
|
||
|
|
|
||
|
|
print(srt)
|