Spaces:
Sleeping
Sleeping
cyberosa
commited on
Commit
·
98e00e9
1
Parent(s):
51751a6
extending WebSearchTool
Browse files
tools.py
CHANGED
|
@@ -1,10 +1,10 @@
|
|
| 1 |
-
from smolagents import
|
| 2 |
import requests
|
| 3 |
import time
|
| 4 |
from typing import List, Dict, Any, Optional
|
| 5 |
|
| 6 |
|
| 7 |
-
class MySearchTool(
|
| 8 |
name = "Search tool to avoid rate limit"
|
| 9 |
description = "It receives a question and returns a list of search results from various sources. It uses DuckDuckGo as the primary source"
|
| 10 |
inputs = {
|
|
@@ -15,11 +15,11 @@ class MySearchTool(Tool):
|
|
| 15 |
}
|
| 16 |
output_type = "string"
|
| 17 |
|
| 18 |
-
def __init__(self,
|
| 19 |
-
|
| 20 |
-
self.
|
| 21 |
|
| 22 |
-
def
|
| 23 |
"""
|
| 24 |
Free search using DuckDuckGo Instant Answer API
|
| 25 |
|
|
@@ -34,7 +34,7 @@ class MySearchTool(Tool):
|
|
| 34 |
# DuckDuckGo Instant Answer API (free)
|
| 35 |
url = "https://api.duckduckgo.com/"
|
| 36 |
params = {
|
| 37 |
-
"q":
|
| 38 |
"format": "json",
|
| 39 |
"no_html": "1",
|
| 40 |
"skip_disambig": "1",
|
|
@@ -42,8 +42,9 @@ class MySearchTool(Tool):
|
|
| 42 |
|
| 43 |
response = requests.get(url, params=params, timeout=10)
|
| 44 |
response.raise_for_status()
|
| 45 |
-
time.sleep(2) # 2-second delay
|
| 46 |
data = response.json()
|
|
|
|
|
|
|
| 47 |
results = []
|
| 48 |
|
| 49 |
# Process abstract
|
|
@@ -58,7 +59,9 @@ class MySearchTool(Tool):
|
|
| 58 |
)
|
| 59 |
|
| 60 |
# Process related topics
|
| 61 |
-
for topic in data.get("RelatedTopics", [])[
|
|
|
|
|
|
|
| 62 |
if isinstance(topic, dict) and "Text" in topic:
|
| 63 |
results.append(
|
| 64 |
{
|
|
@@ -69,7 +72,7 @@ class MySearchTool(Tool):
|
|
| 69 |
}
|
| 70 |
)
|
| 71 |
|
| 72 |
-
return results[:max_results]
|
| 73 |
|
| 74 |
except Exception as e:
|
| 75 |
print(f"DuckDuckGo search failed: {str(e)}")
|
|
|
|
| 1 |
+
from smolagents import WebSearchTool
|
| 2 |
import requests
|
| 3 |
import time
|
| 4 |
from typing import List, Dict, Any, Optional
|
| 5 |
|
| 6 |
|
| 7 |
+
class MySearchTool(WebSearchTool):
|
| 8 |
name = "Search tool to avoid rate limit"
|
| 9 |
description = "It receives a question and returns a list of search results from various sources. It uses DuckDuckGo as the primary source"
|
| 10 |
inputs = {
|
|
|
|
| 15 |
}
|
| 16 |
output_type = "string"
|
| 17 |
|
| 18 |
+
def __init__(self, max_results=5):
|
| 19 |
+
super().__init__()
|
| 20 |
+
self.max_results = max_results
|
| 21 |
|
| 22 |
+
def search_duckduckgo(self, query: str) -> list:
|
| 23 |
"""
|
| 24 |
Free search using DuckDuckGo Instant Answer API
|
| 25 |
|
|
|
|
| 34 |
# DuckDuckGo Instant Answer API (free)
|
| 35 |
url = "https://api.duckduckgo.com/"
|
| 36 |
params = {
|
| 37 |
+
"q": query,
|
| 38 |
"format": "json",
|
| 39 |
"no_html": "1",
|
| 40 |
"skip_disambig": "1",
|
|
|
|
| 42 |
|
| 43 |
response = requests.get(url, params=params, timeout=10)
|
| 44 |
response.raise_for_status()
|
|
|
|
| 45 |
data = response.json()
|
| 46 |
+
time.sleep(2) # 2-second delay
|
| 47 |
+
|
| 48 |
results = []
|
| 49 |
|
| 50 |
# Process abstract
|
|
|
|
| 59 |
)
|
| 60 |
|
| 61 |
# Process related topics
|
| 62 |
+
for topic in data.get("RelatedTopics", [])[
|
| 63 |
+
: self.max_results - len(results)
|
| 64 |
+
]:
|
| 65 |
if isinstance(topic, dict) and "Text" in topic:
|
| 66 |
results.append(
|
| 67 |
{
|
|
|
|
| 72 |
}
|
| 73 |
)
|
| 74 |
|
| 75 |
+
return results[: self.max_results]
|
| 76 |
|
| 77 |
except Exception as e:
|
| 78 |
print(f"DuckDuckGo search failed: {str(e)}")
|