Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
|
@@ -3,6 +3,7 @@ from functools import lru_cache
|
|
| 3 |
import random
|
| 4 |
import requests
|
| 5 |
import logging
|
|
|
|
| 6 |
import config
|
| 7 |
import plotly.graph_objects as go
|
| 8 |
from typing import Dict
|
|
@@ -25,15 +26,32 @@ from release_notes import get_release_notes_html
|
|
| 25 |
|
| 26 |
# Update the logging format to redact URLs
|
| 27 |
logging.basicConfig(
|
| 28 |
-
level=logging.
|
| 29 |
format='%(asctime)s - %(levelname)s - %(message)s'
|
| 30 |
)
|
| 31 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 32 |
class RedactURLsFilter(logging.Filter):
|
| 33 |
def filter(self, record):
|
| 34 |
-
# Redact
|
| 35 |
-
|
| 36 |
-
record.msg =
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
return True
|
| 38 |
|
| 39 |
# Apply the filter to all handlers
|
|
@@ -100,7 +118,6 @@ def call_ollama_api(model, prompt):
|
|
| 100 |
]
|
| 101 |
|
| 102 |
# Extract thinking part and main content using regex
|
| 103 |
-
import re
|
| 104 |
thinking_match = re.search(r'<think>(.*?)</think>', content, flags=re.DOTALL)
|
| 105 |
|
| 106 |
if thinking_match:
|
|
|
|
| 3 |
import random
|
| 4 |
import requests
|
| 5 |
import logging
|
| 6 |
+
import re
|
| 7 |
import config
|
| 8 |
import plotly.graph_objects as go
|
| 9 |
from typing import Dict
|
|
|
|
| 26 |
|
| 27 |
# Update the logging format to redact URLs
|
| 28 |
logging.basicConfig(
|
| 29 |
+
level=logging.WARNING, # Only show warnings and errors
|
| 30 |
format='%(asctime)s - %(levelname)s - %(message)s'
|
| 31 |
)
|
| 32 |
|
| 33 |
+
# Suppress verbose HTTP request logging
|
| 34 |
+
logging.getLogger("urllib3").setLevel(logging.CRITICAL)
|
| 35 |
+
logging.getLogger("httpx").setLevel(logging.CRITICAL)
|
| 36 |
+
logging.getLogger("openai").setLevel(logging.CRITICAL)
|
| 37 |
+
|
| 38 |
class RedactURLsFilter(logging.Filter):
|
| 39 |
def filter(self, record):
|
| 40 |
+
# Redact all URLs using regex pattern
|
| 41 |
+
url_pattern = r'http[s]?://(?:[a-zA-Z]|[0-9]|[$-_@.&+]|[!*\\(\\),]|(?:%[0-9a-fA-F][0-9a-fA-F]))+'
|
| 42 |
+
record.msg = re.sub(url_pattern, '[REDACTED_URL]', str(record.msg))
|
| 43 |
+
|
| 44 |
+
# Remove HTTP status codes
|
| 45 |
+
record.msg = re.sub(r'HTTP/\d\.\d \d+ \w+', '', record.msg)
|
| 46 |
+
|
| 47 |
+
# Remove sensitive API references
|
| 48 |
+
record.msg = record.msg.replace(config.API_URL, '[API]')
|
| 49 |
+
record.msg = record.msg.replace(config.NEXTCLOUD_URL, '[CLOUD]')
|
| 50 |
+
|
| 51 |
+
# Clean up residual artifacts
|
| 52 |
+
record.msg = re.sub(r'\s+', ' ', record.msg).strip()
|
| 53 |
+
record.msg = re.sub(r'("?) \1', '', record.msg) # Remove empty quotes
|
| 54 |
+
|
| 55 |
return True
|
| 56 |
|
| 57 |
# Apply the filter to all handlers
|
|
|
|
| 118 |
]
|
| 119 |
|
| 120 |
# Extract thinking part and main content using regex
|
|
|
|
| 121 |
thinking_match = re.search(r'<think>(.*?)</think>', content, flags=re.DOTALL)
|
| 122 |
|
| 123 |
if thinking_match:
|