pandora_cdosea/subtitles.py

105 lines
2.7 KiB
Python
Raw Normal View History

2017-05-16 12:59:32 +00:00
#!/usr/bin/python3
import os
import sys
import json
import subprocess
from collections import defaultdict
import string
from glob import glob
from copy import deepcopy
import ox
import ox.web.auth
base_url = 'http://127.0.0.1:2620'
FRAME_DURATION = 1/60
MAX_DURATION = 40
HIDDEN_TAGS = [
"women with white males",
"gene z hanrahan"
]
# items to not use at all
BLACKLIST = [
'XN'
]
api = None
def get_api():
global api
if not api:
api = ox.API(base_url + '/api/')
api.signin(**ox.web.auth.get('cdosea'))
def get_subtitles():
get_api()
items = api.find({
'query': {
'conditions': [{'key': 'tags', 'value': 'Vocal', 'operator': '=='}]
},
'keys': ['id', 'title'],
'range': [0, 1000]})['data']['items']
for item in items:
'''
info = api.findMedia({
'query': {
'conditions': [
{'key': 'id', 'operator': '==', 'value': item['id']}
]
},
'keys': ['id', 'extension'],
'range': [0, 1]
})['data']['items'][0]
'''
item['subtitles'] = api.get({'id': item['id'], 'keys': ['layers']})['data']['layers']['subtitles']
return items
def get_subtitles(items, id):
for item in items:
if item['title'].startswith(id):
return deepcopy(item['subtitles'])
def render_subtitles(item_json, output_json, output_srt):
with open(item_json) as fd:
item = json.load(fd)
subtitles = []
position = 0
for clip in item['vocals']:
if not clip.get('blank'):
# vocals/A/A4_chaton.wav
id = clip['path'].split('/')[-1][:2]
clip_subtitles = get_subtitles(items, id)
for sub in clip_subtitles:
srt = {}
srt['in'] = sub['in'] + position
srt['out'] = sub['out'] + position
srt['value'] = sub['value'].replace('<br>', '')
subtitles.append(srt)
position += clip['duration']
with open(output_srt, 'wb') as fd:
fd.write(ox.srt.encode(subtitles))
with open(output_json, 'w') as fd:
json.dump(subtitles, fd, indent=4, ensure_ascii=False)
if __name__ == '__main__':
if os.path.exists('subtitles.json'):
items = json.load(open('subtitles.json'))
else:
items = get_subtitles()
with open('subtitles.json', 'w') as fd:
json.dump(items, fd, indent=4, ensure_ascii=False)
for item_json in glob('output/*/*.json'):
prefix = 'public/' + item_json.split('/')[-1][0].lower() + item_json.split('/')[-2] + '.1080p.'
output_json = prefix + 'json'
output_srt = prefix + 'srt'
render_subtitles(item_json, output_json, output_srt)