IdlecloudX commited on
Commit
500dfe6
·
verified ·
1 Parent(s): 6701bf8

Update translator.py

Browse files
Files changed (1) hide show
  1. translator.py +54 -101
translator.py CHANGED
@@ -3,19 +3,20 @@ translator.py
3
  """
4
  import hashlib, hmac, json, os, random, time
5
  from datetime import datetime
6
- from typing import List, Sequence, Optional, Dict, Any
7
 
8
  import requests
9
 
10
- DEFAULT_TENCENT_SECRET_ID = os.environ.get("TENCENT_SECRET_ID")
11
- DEFAULT_TENCENT_SECRET_KEY = os.environ.get("TENCENT_SECRET_KEY")
12
- DEFAULT_TENCENT_URL = os.environ.get("TENCENT_TRANSLATE_URL", "https://tmt.tencentcloudapi.com")
13
 
14
- DEFAULT_BAIDU_URL = os.environ.get("BAIDU_TRANSLATE_URL", "https://fanyi-api.baidu.com/api/trans/vip/translate")
15
- try:
16
- DEFAULT_BAIDU_CREDENTIALS = json.loads(os.environ.get("BAIDU_CREDENTIALS_JSON", "[]"))
17
- except json.JSONDecodeError:
18
- DEFAULT_BAIDU_CREDENTIALS = []
 
 
19
 
20
  def _sign(key: bytes, msg: str) -> bytes:
21
  return hmac.new(key, msg.encode("utf-8"), hashlib.sha256).digest()
@@ -26,123 +27,75 @@ def _tc3_signature(secret_key: str, date: str, service: str, string_to_sign: str
26
  secret_signing = _sign(secret_service, "tc3_request")
27
  return hmac.new(secret_signing, string_to_sign.encode("utf-8"), hashlib.sha256).hexdigest()
28
 
29
- def _translate_with_tencent_dynamic(
30
- texts: Sequence[str],
31
- secret_id: str,
32
- secret_key: str,
33
- url: str,
34
- src="auto",
35
- tgt="zh"
36
- ) -> Optional[List[str]]:
37
  if not (secret_id and secret_key):
38
  return None
 
39
  service, host, action, version, region = "tmt", "tmt.tencentcloudapi.com", "TextTranslate", "2018-03-21", "ap-beijing"
40
- ts = int(time.time())
41
- date = datetime.utcfromtimestamp(ts).strftime("%Y-%m-%d")
42
  algorithm = "TC3-HMAC-SHA256"
43
-
44
- payload = {"SourceText": "\n".join(texts), "Source": src, "Target": tgt, "ProjectId": 0}
45
- payload_str = json.dumps(payload, ensure_ascii=False)
46
-
47
- canonical_request = "\n".join([
48
- "POST", "/", "", f"content-type:application/json; charset=utf-8\nhost:{host}\nx-tc-action:{action.lower()}\n",
49
- "content-type;host;x-tc-action", hashlib.sha256(payload_str.encode()).hexdigest(),
50
- ])
51
  credential_scope = f"{date}/{service}/tc3_request"
52
- string_to_sign = "\n".join([
53
- algorithm, str(ts), credential_scope, hashlib.sha256(canonical_request.encode()).hexdigest(),
54
- ])
55
  signature = _tc3_signature(secret_key, date, service, string_to_sign)
56
- authorization = (
57
- f"{algorithm} Credential={secret_id}/{credential_scope}, "
58
- f"SignedHeaders=content-type;host;x-tc-action, Signature={signature}"
59
- )
60
- headers = {
61
- "Authorization": authorization, "Content-Type": "application/json; charset=utf-8", "Host": host,
62
- "X-TC-Action": action, "X-TC-Timestamp": str(ts), "X-TC-Version": version, "X-TC-Region": region,
63
- }
64
 
65
  try:
66
- resp = requests.post(url, headers=headers, data=payload_str, timeout=8)
67
  resp.raise_for_status()
68
  data = resp.json()
69
- return data["Response"]["TargetText"].split("\n")
 
 
 
70
  except Exception as e:
71
- print(f"[translator] Dynamic Tencent API error {e}")
72
  return None
73
 
74
- def _translate_with_baidu_dynamic(
75
- texts: Sequence[str],
76
- baidu_creds_list: List[Dict[str, str]],
77
- url: str,
78
- src="auto",
79
- tgt="zh"
80
- ) -> Optional[List[str]]:
81
- if not baidu_creds_list:
82
  return None
83
-
84
- cred = random.choice(baidu_creds_list)
85
- app_id, secret_key = cred.get("app_id"), cred.get("secret_key")
86
- if not (app_id and secret_key):
87
  return None
88
 
89
- salt = random.randint(32768, 65536)
90
- query = "\n".join(texts)
91
- sign = hashlib.md5((app_id + query + str(salt) + secret_key).encode()).hexdigest()
92
  params = {"q": query, "from": src, "to": tgt, "appid": app_id, "salt": salt, "sign": sign}
 
93
  try:
94
- resp = requests.get(url, params=params, timeout=8)
95
  resp.raise_for_status()
96
  data = resp.json()
97
- if "trans_result" in data:
98
- return [item["dst"] for item in data["trans_result"]]
99
- else:
100
- print(f"[translator] Baidu API returned error: {data.get('error_msg', 'Unknown error')}")
101
  return None
 
102
  except Exception as e:
103
- print(f"[translator] Dynamic Baidu API error {e}")
104
  return None
105
 
106
- def translate_texts_with_dynamic_keys(
107
- texts: Sequence[str],
108
- tencent_id: str,
109
- tencent_key: str,
110
- baidu_json_str: str,
111
- src_lang: str = "auto",
112
- tgt_lang: str = "zh"
113
- ) -> List[str]:
114
  if not texts:
115
  return []
116
 
117
- if tencent_id and tencent_key:
118
- out = _translate_with_tencent_dynamic(texts, tencent_id, tencent_key, DEFAULT_TENCENT_URL, src_lang, tgt_lang)
119
- if out is not None:
120
- return out
121
 
122
- baidu_creds = []
123
- if baidu_json_str:
124
- try:
125
- baidu_creds = json.loads(baidu_json_str)
126
- if not isinstance(baidu_creds, list): baidu_creds = []
127
- except json.JSONDecodeError:
128
- print("[translator] Invalid Baidu JSON provided by user.")
129
- baidu_creds = []
130
-
131
- if baidu_creds:
132
- out = _translate_with_baidu_dynamic(texts, baidu_creds, DEFAULT_BAIDU_URL, src_lang, tgt_lang)
133
- if out is not None:
134
- return out
135
-
136
- return list(texts)
137
-
138
- def translate_texts(texts: Sequence[str],
139
- src_lang: str = "auto",
140
- tgt_lang: str = "zh") -> List[str]:
141
- return translate_texts_with_dynamic_keys(
142
- texts,
143
- DEFAULT_TENCENT_SECRET_ID,
144
- DEFAULT_TENCENT_SECRET_KEY,
145
- json.dumps(DEFAULT_BAIDU_CREDENTIALS), # 转换回 JSON 字符串以匹配接口
146
- src_lang,
147
- tgt_lang
148
- )
 
3
  """
4
  import hashlib, hmac, json, os, random, time
5
  from datetime import datetime
6
+ from typing import List, Sequence, Optional
7
 
8
  import requests
9
 
10
+ TENCENT_TRANSLATE_URL = os.environ.get("TENCENT_TRANSLATE_URL", "https://tmt.tencentcloudapi.com")
11
+ BAIDU_TRANSLATE_URL = os.environ.get("BAIDU_TRANSLATE_URL", "https://fanyi-api.baidu.com/api/trans/vip/translate")
 
12
 
13
+ _baidu_idx: int = 0
14
+ def _next_baidu_cred(creds_list):
15
+ global _baidu_idx
16
+ if not creds_list: return None
17
+ cred = creds_list[_baidu_idx]
18
+ _baidu_idx = (_baidu_idx + 1) % len(creds_list)
19
+ return cred
20
 
21
  def _sign(key: bytes, msg: str) -> bytes:
22
  return hmac.new(key, msg.encode("utf-8"), hashlib.sha256).digest()
 
27
  secret_signing = _sign(secret_service, "tc3_request")
28
  return hmac.new(secret_signing, string_to_sign.encode("utf-8"), hashlib.sha256).hexdigest()
29
 
30
+ def _translate_with_tencent(texts: Sequence[str], src: str, tgt: str,
31
+ secret_id: Optional[str],
32
+ secret_key: Optional[str]) -> Optional[List[str]]:
 
 
 
 
 
33
  if not (secret_id and secret_key):
34
  return None
35
+
36
  service, host, action, version, region = "tmt", "tmt.tencentcloudapi.com", "TextTranslate", "2018-03-21", "ap-beijing"
37
+ ts, date = int(time.time()), datetime.utcfromtimestamp(int(time.time())).strftime("%Y-%m-%d")
 
38
  algorithm = "TC3-HMAC-SHA256"
39
+ payload_str = json.dumps({"SourceText": "\n".join(texts), "Source": src, "Target": tgt, "ProjectId": 0}, ensure_ascii=False)
40
+
41
+ canonical_request = f"POST\n/\n\ncontent-type:application/json; charset=utf-8\nhost:{host}\nx-tc-action:{action.lower()}\n\ncontent-type;host;x-tc-action\n{hashlib.sha256(payload_str.encode()).hexdigest()}"
 
 
 
 
 
42
  credential_scope = f"{date}/{service}/tc3_request"
43
+ string_to_sign = f"{algorithm}\n{ts}\n{credential_scope}\n{hashlib.sha256(canonical_request.encode()).hexdigest()}"
 
 
44
  signature = _tc3_signature(secret_key, date, service, string_to_sign)
45
+
46
+ authorization = f"{algorithm} Credential={secret_id}/{credential_scope}, SignedHeaders=content-type;host;x-tc-action, Signature={signature}"
47
+ headers = {"Authorization": authorization, "Content-Type": "application/json; charset=utf-8", "Host": host, "X-TC-Action": action, "X-TC-Timestamp": str(ts), "X-TC-Version": version, "X-TC-Region": region}
 
 
 
 
 
48
 
49
  try:
50
+ resp = requests.post(TENCENT_TRANSLATE_URL, headers=headers, data=payload_str, timeout=8)
51
  resp.raise_for_status()
52
  data = resp.json()
53
+ if "Error" in data.get("Response", {}):
54
+ print(f"[translator] Tencent API error: {data['Response']['Error']['Message']}")
55
+ return None
56
+ return data.get("Response", {}).get("TargetText", "").split("\n")
57
  except Exception as e:
58
+ print(f"[translator] Tencent request error: {e}")
59
  return None
60
 
61
+ def _translate_with_baidu(texts: Sequence[str], src: str, tgt: str,
62
+ baidu_credentials_list: Optional[list]) -> Optional[List[str]]:
63
+ if not baidu_credentials_list:
 
 
 
 
 
64
  return None
65
+
66
+ creds = _next_baidu_cred(baidu_credentials_list)
67
+ if not creds or not creds.get("app_id") or not creds.get("secret_key"):
68
+ print("[translator] Baidu credentials format error.")
69
  return None
70
 
71
+ app_id, secret_key = creds["app_id"], creds["secret_key"]
72
+ salt, query = random.randint(32768, 65536), "\n".join(texts)
73
+ sign = hashlib.md5((app_id + query + str(salt) + secret_key).encode()).hexdigest()
74
  params = {"q": query, "from": src, "to": tgt, "appid": app_id, "salt": salt, "sign": sign}
75
+
76
  try:
77
+ resp = requests.get(BAIDU_TRANSLATE_URL, params=params, timeout=8)
78
  resp.raise_for_status()
79
  data = resp.json()
80
+ if "error_code" in data:
81
+ print(f"[translator] Baidu API error: {data.get('error_msg', 'Unknown error')}")
 
 
82
  return None
83
+ return [item["dst"] for item in data.get("trans_result", [])]
84
  except Exception as e:
85
+ print(f"[translator] Baidu request error: {e}")
86
  return None
87
 
88
+ def translate_texts(texts: Sequence[str],
89
+ src_lang: str = "auto",
90
+ tgt_lang: str = "zh",
91
+ tencent_secret_id: Optional[str] = None,
92
+ tencent_secret_key: Optional[str] = None,
93
+ baidu_credentials_list: Optional[list] = None) -> List[str]:
 
 
94
  if not texts:
95
  return []
96
 
97
+ out = _translate_with_tencent(texts, src_lang, tgt_lang, tencent_secret_id, tencent_secret_key)
98
+ if out is None:
99
+ out = _translate_with_baidu(texts, src_lang, tgt_lang, baidu_credentials_list)
 
100
 
101
+ return out or list(texts)