Spaces:
Running
Running
Commit
•
1bb14f7
1
Parent(s):
2abb453
fix
Browse files
searcher/sementic_search.py
CHANGED
|
@@ -31,10 +31,15 @@ def extract(text, type):
|
|
| 31 |
return text
|
| 32 |
else:
|
| 33 |
return ""
|
| 34 |
-
|
| 35 |
def download(url):
|
| 36 |
try:
|
| 37 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
if response.status_code == 200:
|
| 39 |
return response.content
|
| 40 |
else:
|
|
|
|
| 31 |
return text
|
| 32 |
else:
|
| 33 |
return ""
|
| 34 |
+
|
| 35 |
def download(url):
|
| 36 |
try:
|
| 37 |
+
headers = {
|
| 38 |
+
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) '
|
| 39 |
+
'AppleWebKit/537.36 (KHTML, like Gecko) '
|
| 40 |
+
'Chrome/87.0.4280.88 Safari/537.36'
|
| 41 |
+
} # Mimic a common browser's User-Agent
|
| 42 |
+
response = requests.get(url,headers=headers,timeout=120)
|
| 43 |
if response.status_code == 200:
|
| 44 |
return response.content
|
| 45 |
else:
|