XciD HF staff commited on
Commit
fb231c9
1 Parent(s): 999a35e

Fix query params for spaces

Browse files
Files changed (1) hide show
  1. detector/server.py +11 -6
detector/server.py CHANGED
@@ -7,13 +7,17 @@ from transformers import RobertaForSequenceClassification, RobertaTokenizer
7
  import json
8
  import fire
9
  import torch
10
- from urllib.parse import urlparse, unquote, parse_qs
11
-
12
 
13
  model: RobertaForSequenceClassification = None
14
  tokenizer: RobertaTokenizer = None
15
  device: str = None
16
 
 
 
 
 
17
  def log(*args):
18
  print(f"[{os.environ.get('RANK', '')}]", *args, file=sys.stderr)
19
 
@@ -45,10 +49,11 @@ class RequestHandler(SimpleHTTPRequestHandler):
45
  self.wfile.write(json.dumps({"error": str(e)}).encode('utf-8'))
46
 
47
  def do_GET(self):
48
- parsed = urlparse(self.path)
49
- query_params = parse_qs(parsed.query)
 
50
 
51
- if 'text' not in query_params:
52
  self.begin_content('text/html')
53
 
54
  html = os.path.join(os.path.dirname(__file__), 'index.html')
@@ -57,7 +62,7 @@ class RequestHandler(SimpleHTTPRequestHandler):
57
 
58
  self.begin_content('application/json;charset=UTF-8')
59
 
60
- all_tokens, used_tokens, fake, real = self.infer(unquote(query_params['text'][0]))
61
 
62
  self.wfile.write(json.dumps(dict(
63
  all_tokens=all_tokens,
 
7
  import json
8
  import fire
9
  import torch
10
+ import re
11
+ from urllib.parse import urlparse, unquote, parse_qs, urlencode
12
 
13
  model: RobertaForSequenceClassification = None
14
  tokenizer: RobertaTokenizer = None
15
  device: str = None
16
 
17
+ # Remove spaces query params from query
18
+ regex = r"__theme=(.+)"
19
+
20
+
21
  def log(*args):
22
  print(f"[{os.environ.get('RANK', '')}]", *args, file=sys.stderr)
23
 
 
49
  self.wfile.write(json.dumps({"error": str(e)}).encode('utf-8'))
50
 
51
  def do_GET(self):
52
+ query = urlparse(self.path).query
53
+ query = re.sub(regex, "", query, 0, re.MULTILINE)
54
+ query = unquote(query)
55
 
56
+ if not query:
57
  self.begin_content('text/html')
58
 
59
  html = os.path.join(os.path.dirname(__file__), 'index.html')
 
62
 
63
  self.begin_content('application/json;charset=UTF-8')
64
 
65
+ all_tokens, used_tokens, fake, real = self.infer(query)
66
 
67
  self.wfile.write(json.dumps(dict(
68
  all_tokens=all_tokens,