Basic web tool to access a web page at a url.

CHANGES.txt

@@ -1,6 +1,12 @@
 Changelog
 =========
 
+0.7.2 - 
+------------------
+
+- Add a web tool, giving Ollama access to the web.
+  [ggozad] 
+
 0.7.1 - 2025-01-02
 ------------------
 

README.md

@@ -145,6 +145,7 @@ Since version `0.6.0` `oterm` supports integration with tools. Tools are special
 
 The following tools are currently supported:
 
+* `fetch_url` - allows your models access to the web, fetches a URL and provides the content as input to the model.
 * `date_time` - provides the current date and time in ISO format.
 * `current_location` - provides the current location of the user (longitude, latitude, city, region, country). Uses [ipinfo.io](https://ipinfo.io) to determine the location.
 * `current_weather` - provides the current weather in the user's location. Uses [OpenWeatherMap](https://openweathermap.org) to determine the weather. You need to provide your (free) API key in the OPEN_WEATHER_MAP_API_KEY environment variable.

src/oterm/tools/__init__.py

@@ -1,14 +1,16 @@
-from typing import Awaitable, Callable, Sequence, TypedDict
+from typing import Sequence
 
 from oterm.tools.date_time import DateTimeTool, date_time
 from oterm.tools.location import LocationTool, current_location
 from oterm.tools.shell import ShellTool, shell_command
 from oterm.tools.weather import WeatherTool, current_weather
+from oterm.tools.web import WebTool, fetch_url
 from oterm.types import ToolDefinition
 
 available: Sequence[ToolDefinition] = [
     {"tool": DateTimeTool, "callable": date_time},
     {"tool": ShellTool, "callable": shell_command},
     {"tool": LocationTool, "callable": current_location},
     {"tool": WeatherTool, "callable": current_weather},
+    {"tool": WebTool, "callable": fetch_url},
 ]

src/oterm/tools/web.py

@@ -0,0 +1,51 @@
+import json
+from html.parser import HTMLParser
+
+import httpx
+
+from oterm.types import Tool
+
+WebTool = Tool(
+    type="function",
+    function=Tool.Function(
+        name="fetch_url",
+        description="Function to return the contents of a website in text format.",
+        parameters=Tool.Function.Parameters(
+            type="object",
+            properties={
+                "url": Tool.Function.Parameters.Property(
+                    type="str", description="The URL of the website to fetch."
+                ),
+            },
+            required=["url"],
+        ),
+    ),
+)
+
+
+class HTML2Text(HTMLParser):
+    text = ""
+
+    def handle_data(self, data):
+        self.text += data
+
+
+async def fetch_url(url: str) -> str:
+    async with httpx.AsyncClient() as client:
+        try:
+            response = await client.get(url)
+
+            if response.status_code == 200:
+                html = response.text
+                parser = HTML2Text()
+                parser.feed(html)
+                return parser.text
+
+            else:
+                return json.dumps(
+                    {
+                        "error": f"Failed to fetch URL: {url}. Status code: {response.status_code}"
+                    }
+                )
+        except Exception as e:
+            return json.dumps({"error": str(e)})

tests/tools/test_web_tool.py

@@ -0,0 +1,15 @@
+import pytest
+
+from oterm.ollamaclient import OllamaLLM
+from oterm.tools.web import WebTool, fetch_url
+
+
+@pytest.mark.asyncio
+async def test_web():
+    llm = OllamaLLM(
+        tool_defs=[{"tool": WebTool, "callable": fetch_url}],
+    )
+    res = await llm.completion(
+        "What's oterm in a single phrase? oterm is hosted at https://github.com/ggozad/oterm."
+    )
+    assert "Ollama" in res