添加腾讯翻译官方接口

This commit is contained in:
John Smith
2021-04-11 22:11:13 +08:00
parent f6e38c8804
commit 8aa80fadd3
3 changed files with 185 additions and 13 deletions

View File

@@ -78,6 +78,16 @@ class AppConfig:
if type_ == 'TencentTranslateFree':
translator_config['source_language'] = section['source_language']
translator_config['target_language'] = section['target_language']
elif type_ == 'BilibiliTranslateFree':
pass
elif type_ == 'TencentTranslate':
translator_config['source_language'] = section['source_language']
translator_config['target_language'] = section['target_language']
translator_config['secret_id'] = section['secret_id']
translator_config['secret_key'] = section['secret_key']
translator_config['region'] = section['region']
else:
raise ValueError(f'Invalid translator type: {type_}')
translator_configs.append(translator_config)
self.translator_configs = translator_configs

View File

@@ -25,19 +25,23 @@ loader_url = https://xfgryujk.sinacloud.net/blivechat/loader.html
# -------------------------------------------------------------------------------------------------
# 以下是给字幕组看的实在懒得翻译了_(:з」∠)_。如果你不了解以下参数的意思使用默认值就好
# **The following is for translation team. Leave it default if you don't know its meaning**
# -------------------------------------------------------------------------------------------------
# 翻译器配置,索引到下面的配置节。可以配置多个翻译器,翻译时会使用当前第一个可用的翻译器
# 配置多个翻译器可以增加QPS、容灾。不同配置可以使用同一个类型,但要使用不同的账号,否则还是会遇到调用频率限制
# 翻译器配置,索引到下面的配置节。可以以逗号分隔配置多个翻译器,翻译时会使用当前第一个可用的翻译器
# 配置多个翻译器可以增加额度、增加QPS、容灾
# 不同配置可以使用同一个类型,但要使用不同的账号,否则还是会遇到额度、调用频率限制
translator_configs = tencent_translate_free,bilibili_translate_free
[tencent_translate_free]
# 类型:腾讯翻译白嫖版。使用了网页版的接口,**将来可能失效**
type = TencentTranslateFree
# 请求间隔时间(秒),等于 1 / QPS。目前没有遇到此接口有调用频率限制10QPS应该够用了
query_interval = 0.1
# 最大队列长度,注意最长等待时间等于 最大队列长度 * 请求间隔时间
max_queue_size = 100
# 自动auto中文zh日语jp英语en韩语kr
# 完整语言列表见文档https://cloud.tencent.com/document/product/551/15619
# 源语言
@@ -49,7 +53,42 @@ target_language = jp
[bilibili_translate_free]
# 类型B站翻译白嫖版。使用了B站直播网页的接口**将来可能失效**。目前B站翻译后端是百度翻译
type = BilibiliTranslateFree
# 请求间隔时间(秒),等于 1 / QPS。目前此接口频率限制是3秒一次
query_interval = 3.1
# 最大队列长度,注意最长等待时间等于 最大队列长度 * 请求间隔时间
max_queue_size = 3
[tencent_translate]
# 文档https://cloud.tencent.com/product/tmt
# 定价https://cloud.tencent.com/document/product/551/35017
# * 文本翻译的每月免费额度为5百万字符
# * 文本翻译当月需付费字符数小于100百万字符1亿字符刊例价为58元/每百万字符
# * 文本翻译当月需付费字符数大于等于100百万字符1亿字符刊例价为50元/每百万字符
# 限制https://cloud.tencent.com/document/product/551/32572
# * 文本翻译最高QPS为5
# 类型:腾讯翻译
type = TencentTranslate
# 请求间隔时间(秒),等于 1 / QPS。理论上最高QPS为5实际测试是3
query_interval = 0.333
# 最大队列长度,注意最长等待时间等于 最大队列长度 * 请求间隔时间
max_queue_size = 30
# 自动auto中文zh日语jp英语en韩语kr
# 完整语言列表见文档https://cloud.tencent.com/document/product/551/15619
# 源语言
source_language = zh
# 目标语言
target_language = jp
# 腾讯云API密钥
secret_id =
secret_key =
# 腾讯云地域参数,用来标识希望操作哪个地域的数据
# 北京ap-beijing上海ap-shanghai香港ap-hongkong首尔ap-seoul
# 完整地域列表见文档https://cloud.tencent.com/document/api/551/15615#.E5.9C.B0.E5.9F.9F.E5.88.97.E8.A1.A8
region = ap-shanghai

View File

@@ -1,7 +1,11 @@
# -*- coding: utf-8 -*-
import asyncio
import datetime
import functools
import hashlib
import hmac
import json
import logging
import re
from typing import *
@@ -32,19 +36,11 @@ def init():
async def _do_init():
cfg = config.get_config()
if not cfg.enable_translate:
return
providers = []
for trans_cfg in cfg.translator_configs:
provider = None
type_ = trans_cfg['type']
if type_ == 'TencentTranslateFree':
provider = TencentTranslateFree(
trans_cfg['query_interval'], trans_cfg['max_queue_size'], trans_cfg['source_language'],
trans_cfg['target_language']
)
elif type_ == 'BilibiliTranslateFree':
provider = BilibiliTranslateFree(trans_cfg['query_interval'], trans_cfg['max_queue_size'])
provider = create_translate_provider(trans_cfg)
if provider is not None:
providers.append(provider)
await asyncio.gather(*(provider.init() for provider in providers))
@@ -52,6 +48,24 @@ async def _do_init():
_translate_providers = providers
def create_translate_provider(cfg):
type_ = cfg['type']
if type_ == 'TencentTranslateFree':
return TencentTranslateFree(
cfg['query_interval'], cfg['max_queue_size'], cfg['source_language'],
cfg['target_language']
)
elif type_ == 'BilibiliTranslateFree':
return BilibiliTranslateFree(cfg['query_interval'], cfg['max_queue_size'])
elif type_ == 'TencentTranslate':
return TencentTranslate(
cfg['query_interval'], cfg['max_queue_size'], cfg['source_language'],
cfg['target_language'], cfg['secret_id'], cfg['secret_key'],
cfg['region']
)
return None
def need_translate(text):
text = text.strip()
# 没有中文,平时打不出的字不管
@@ -321,3 +335,112 @@ class BilibiliTranslateFree(FlowControlTranslateProvider):
logger.warning('BilibiliTranslateFree failed: %d %s', data['code'], data['msg'])
return None
return data['data']['message_trans']
class TencentTranslate(FlowControlTranslateProvider):
def __init__(self, query_interval, max_queue_size, source_language, target_language,
secret_id, secret_key, region):
super().__init__(query_interval, max_queue_size)
self._source_language = source_language
self._target_language = target_language
self._secret_id = secret_id
self._secret_key = secret_key
self._region = region
self._cool_down_timer_handle = None
@property
def is_available(self):
return self._cool_down_timer_handle is None and super().is_available
async def _do_translate(self, text):
try:
async with self._request_tencent_cloud(
'TextTranslate',
'2018-03-21',
{
'SourceText': text,
'Source': self._source_language,
'Target': self._target_language,
'ProjectId': 0
}
) as r:
if r.status != 200:
logger.warning('TencentTranslate request failed: status=%d %s', r.status, r.reason)
return None
data = (await r.json())['Response']
except (aiohttp.ClientConnectionError, asyncio.TimeoutError):
return None
error = data.get('Error', None)
if error is not None:
logger.warning('TencentTranslate failed: %s %s, RequestId=%s', error['Code'],
error['Message'], data['RequestId'])
self._on_fail(error['Code'])
return None
return data['TargetText']
def _request_tencent_cloud(self, action, version, body):
body_bytes = json.dumps(body).encode('utf-8')
canonical_headers = 'content-type:application/json; charset=utf-8\nhost:tmt.tencentcloudapi.com\n'
signed_headers = 'content-type;host'
hashed_request_payload = hashlib.sha256(body_bytes).hexdigest()
canonical_request = f'POST\n/\n\n{canonical_headers}\n{signed_headers}\n{hashed_request_payload}'
request_timestamp = int(datetime.datetime.now().timestamp())
date = datetime.datetime.utcfromtimestamp(request_timestamp).strftime('%Y-%m-%d')
credential_scope = f'{date}/tmt/tc3_request'
hashed_canonical_request = hashlib.sha256(canonical_request.encode('utf-8')).hexdigest()
string_to_sign = f'TC3-HMAC-SHA256\n{request_timestamp}\n{credential_scope}\n{hashed_canonical_request}'
def sign(key, msg):
return hmac.new(key, msg.encode('utf-8'), hashlib.sha256).digest()
secret_date = sign(('TC3' + self._secret_key).encode('utf-8'), date)
secret_service = sign(secret_date, 'tmt')
secret_signing = sign(secret_service, 'tc3_request')
signature = hmac.new(secret_signing, string_to_sign.encode('utf-8'), hashlib.sha256).hexdigest()
authorization = (
f'TC3-HMAC-SHA256 Credential={self._secret_id}/{credential_scope}, '
f'SignedHeaders={signed_headers}, Signature={signature}'
)
headers = {
'Authorization': authorization,
'Content-Type': 'application/json; charset=utf-8',
'X-TC-Action': action,
'X-TC-Version': version,
'X-TC-Timestamp': str(request_timestamp),
'X-TC-Region': self._region
}
return _http_session.post('https://tmt.tencentcloudapi.com/', headers=headers, data=body_bytes)
def _on_fail(self, code):
if self._cool_down_timer_handle is not None:
return
sleep_time = 0
if code == 'FailedOperation.NoFreeAmount':
# 下个月恢复免费额度
cur_time = datetime.datetime.now()
year = cur_time.year
month = cur_time.month + 1
if month > 12:
year += 1
month = 1
next_month_time = datetime.datetime(year, month, 1, minute=5)
sleep_time = (next_month_time - cur_time).total_seconds()
# Python 3.8之前不能超过一天
sleep_time = min(sleep_time, 24 * 60 * 60 - 1)
elif code in ('FailedOperation.ServiceIsolate', 'LimitExceeded'):
# 需要手动处理等5分钟
sleep_time = 5 * 60
if sleep_time != 0:
self._cool_down_timer_handle = asyncio.get_event_loop().call_later(
sleep_time, self._on_cool_down_timeout
)
def _on_cool_down_timeout(self):
self._cool_down_timer_handle = None