1
00:00:22,105 --> 00:00:23,148
♫ (Hayden Magnum - Flicker) Love was my religion, ♫
2
00:00:23,148 --> 00:00:24,357
♫ (Hayden Magnum - Flicker) but i lost my faith ♫
3
00:00:24,357 --> 00:00:26,735
♫ My line is going straight, but wait ♫
- 翻译逻辑：
- 跳过空行、数字行、箭头行，直接原文写入新文件
- 剩余的是需要翻译的有效行，调接口翻译，翻译完把原文和译文写入新文件
- 有效行中可能出现一句对话拆分多行的情况，采用多行合并后再翻译的处理方式
- API 限制：
- 标准版QPS=1，不限字符，高级版QPS=10，但是每月只有两百万字符免费额度。我使用的标准版，调用时加延时
import hashlib
import re
import time
import requests
def translate(from_, to, text):
base_url = 'https://fanyi-api.baidu.com/api/trans/vip/translate'
payload = {
'q': text,
'from': from_,
'to': to,
'appid': 'xxx',
'salt': 'xxx'
}
raw = payload['appid'] + payload['q'] + payload['salt'] + 'xxx'
sign = hashlib.md5(raw.encode()).hexdigest()
try:
r = requests.post(
base_url,
params=dict(**payload, sign=sign)
)
trans_result = r.json().get('trans_result', [{}])
if r.status_code != 200:
print(f'api return error: {r.status_code}')
time.sleep(1)
return trans_result[0].get('dst', '')
except Exception as e:
print(f'translate error: {e}')
return ''
with open(f'es.srt', 'r', encoding='utf-8') as f:
with open(f'en.srt', 'w', encoding='utf-8') as fc:
lastlines = []
for i, line in enumerate(f.readlines()):
line = re.sub(r'\(.*\)', '', line)
line = line.strip().lstrip('-').strip().lstrip('\ufeff')
if not line or line.isdigit() or '-->' in line:
if lastlines:
line_ = ' '.join(lastlines)
print(f'{i} start')
fc.write(f'{translate(from_="spa", to="en", text=line_)}\n')
print(f'{i} done')
lastlines.clear()
fc.write(f'{line}\n')
else:
lastlines.append(line)
