Spaces:
Sleeping
Sleeping
Update src/utils/search_utility.py
Browse files- src/utils/search_utility.py +66 -66
src/utils/search_utility.py
CHANGED
|
@@ -1,67 +1,67 @@
|
|
| 1 |
-
from typing import List, Dict, Any
|
| 2 |
-
import json
|
| 3 |
-
import os
|
| 4 |
-
from components.search_engine import SearchEngine
|
| 5 |
-
|
| 6 |
-
class SearchUtility:
|
| 7 |
-
"""Utility class for performing web searches"""
|
| 8 |
-
|
| 9 |
-
def __init__(self, config_path: str = "config/google_config.json"):
|
| 10 |
-
self.search_engine = None
|
| 11 |
-
self._initialize_search_engine(config_path)
|
| 12 |
-
|
| 13 |
-
def _initialize_search_engine(self, config_path: str):
|
| 14 |
-
"""Initialize the search engine with configuration"""
|
| 15 |
-
try:
|
| 16 |
-
# Load config if exists
|
| 17 |
-
if os.path.exists(config_path):
|
| 18 |
-
with open(config_path, 'r') as f:
|
| 19 |
-
config = json.load(f)
|
| 20 |
-
|
| 21 |
-
# Set environment variables
|
| 22 |
-
os.environ['GOOGLE_API_KEY'] = config['google_search']['api_key']
|
| 23 |
-
os.environ['GOOGLE_CUSTOM_SEARCH_ID'] = config['google_search']['custom_search_id']
|
| 24 |
-
|
| 25 |
-
self.search_engine = SearchEngine()
|
| 26 |
-
|
| 27 |
-
except Exception as e:
|
| 28 |
-
print(f"Failed to initialize search engine: {str(e)}")
|
| 29 |
-
self.search_engine = None
|
| 30 |
-
|
| 31 |
-
async def search(self, query: str, max_results: int = 3) -> str:
|
| 32 |
-
"""
|
| 33 |
-
Perform a web search and return formatted results
|
| 34 |
-
|
| 35 |
-
Args:
|
| 36 |
-
query (str): The search query
|
| 37 |
-
max_results (int): Maximum number of results to return
|
| 38 |
-
|
| 39 |
-
Returns:
|
| 40 |
-
str: Formatted search results with sources
|
| 41 |
-
"""
|
| 42 |
-
if not self.search_engine:
|
| 43 |
-
return "Search functionality is not available. Please check configuration."
|
| 44 |
-
|
| 45 |
-
try:
|
| 46 |
-
return await self.search_engine.get_knowledge(query, max_results)
|
| 47 |
-
except Exception as e:
|
| 48 |
-
return f"Search failed: {str(e)}"
|
| 49 |
-
|
| 50 |
-
async def get_raw_results(self, query: str, num_results: int = 5) -> List[Dict[str, Any]]:
|
| 51 |
-
"""
|
| 52 |
-
Get raw search results for further processing
|
| 53 |
-
|
| 54 |
-
Args:
|
| 55 |
-
query (str): The search query
|
| 56 |
-
num_results (int): Number of results to return
|
| 57 |
-
|
| 58 |
-
Returns:
|
| 59 |
-
List[Dict]: List of search results with title, link, and snippet
|
| 60 |
-
"""
|
| 61 |
-
if not self.search_engine:
|
| 62 |
-
return []
|
| 63 |
-
|
| 64 |
-
try:
|
| 65 |
-
return await self.search_engine.search(query, num_results)
|
| 66 |
-
except Exception:
|
| 67 |
return []
|
|
|
|
| 1 |
+
from typing import List, Dict, Any
|
| 2 |
+
import json
|
| 3 |
+
import os
|
| 4 |
+
from ..components.search_engine import SearchEngine
|
| 5 |
+
|
| 6 |
+
class SearchUtility:
|
| 7 |
+
"""Utility class for performing web searches"""
|
| 8 |
+
|
| 9 |
+
def __init__(self, config_path: str = "config/google_config.json"):
|
| 10 |
+
self.search_engine = None
|
| 11 |
+
self._initialize_search_engine(config_path)
|
| 12 |
+
|
| 13 |
+
def _initialize_search_engine(self, config_path: str):
|
| 14 |
+
"""Initialize the search engine with configuration"""
|
| 15 |
+
try:
|
| 16 |
+
# Load config if exists
|
| 17 |
+
if os.path.exists(config_path):
|
| 18 |
+
with open(config_path, 'r') as f:
|
| 19 |
+
config = json.load(f)
|
| 20 |
+
|
| 21 |
+
# Set environment variables
|
| 22 |
+
os.environ['GOOGLE_API_KEY'] = config['google_search']['api_key']
|
| 23 |
+
os.environ['GOOGLE_CUSTOM_SEARCH_ID'] = config['google_search']['custom_search_id']
|
| 24 |
+
|
| 25 |
+
self.search_engine = SearchEngine()
|
| 26 |
+
|
| 27 |
+
except Exception as e:
|
| 28 |
+
print(f"Failed to initialize search engine: {str(e)}")
|
| 29 |
+
self.search_engine = None
|
| 30 |
+
|
| 31 |
+
async def search(self, query: str, max_results: int = 3) -> str:
|
| 32 |
+
"""
|
| 33 |
+
Perform a web search and return formatted results
|
| 34 |
+
|
| 35 |
+
Args:
|
| 36 |
+
query (str): The search query
|
| 37 |
+
max_results (int): Maximum number of results to return
|
| 38 |
+
|
| 39 |
+
Returns:
|
| 40 |
+
str: Formatted search results with sources
|
| 41 |
+
"""
|
| 42 |
+
if not self.search_engine:
|
| 43 |
+
return "Search functionality is not available. Please check configuration."
|
| 44 |
+
|
| 45 |
+
try:
|
| 46 |
+
return await self.search_engine.get_knowledge(query, max_results)
|
| 47 |
+
except Exception as e:
|
| 48 |
+
return f"Search failed: {str(e)}"
|
| 49 |
+
|
| 50 |
+
async def get_raw_results(self, query: str, num_results: int = 5) -> List[Dict[str, Any]]:
|
| 51 |
+
"""
|
| 52 |
+
Get raw search results for further processing
|
| 53 |
+
|
| 54 |
+
Args:
|
| 55 |
+
query (str): The search query
|
| 56 |
+
num_results (int): Number of results to return
|
| 57 |
+
|
| 58 |
+
Returns:
|
| 59 |
+
List[Dict]: List of search results with title, link, and snippet
|
| 60 |
+
"""
|
| 61 |
+
if not self.search_engine:
|
| 62 |
+
return []
|
| 63 |
+
|
| 64 |
+
try:
|
| 65 |
+
return await self.search_engine.search(query, num_results)
|
| 66 |
+
except Exception:
|
| 67 |
return []
|