arsam
e8afef3d36
Import the existing local project snapshot. This commit organizes existing files; it does not represent newly developed functionality.
58 lines
1.9 KiB
Python
58 lines
1.9 KiB
Python
from fastapi import FastAPI
|
|
from fastapi.testclient import TestClient
|
|
|
|
|
|
def test_crawl_returns_pages_and_links(monkeypatch):
|
|
async def fake_fetch(url, timeout=10, retries=2):
|
|
return "<html></html>"
|
|
|
|
async def fake_extract(html_text, include_images=False):
|
|
return "content", "Title"
|
|
|
|
async def fake_links(url, html_text, limit=100):
|
|
if url.endswith("/start"):
|
|
return ["https://example.com/next"]
|
|
return []
|
|
|
|
monkeypatch.setattr("axon.routers.crawl.fetch_html", fake_fetch)
|
|
monkeypatch.setattr("axon.routers.crawl.extract_html", fake_extract)
|
|
monkeypatch.setattr("axon.routers.crawl.discover_links", fake_links)
|
|
|
|
from axon.routers.crawl import router
|
|
|
|
app = FastAPI()
|
|
app.include_router(router)
|
|
client = TestClient(app)
|
|
|
|
resp = client.post("/crawl", json={"url": "https://example.com/start", "max_depth": 1})
|
|
assert resp.status_code == 200
|
|
data = resp.json()
|
|
assert len(data["results"]) == 2
|
|
assert data["results"][0]["links"] == ["https://example.com/next"]
|
|
|
|
|
|
def test_map_returns_discovered_urls(monkeypatch):
|
|
async def fake_fetch(url, timeout=10, retries=2):
|
|
return "<html></html>"
|
|
|
|
async def fake_links(url, html_text, limit=100):
|
|
if url.endswith("/start"):
|
|
return ["https://example.com/a", "https://other.com/b"]
|
|
return []
|
|
|
|
monkeypatch.setattr("axon.routers.map.fetch_html", fake_fetch)
|
|
monkeypatch.setattr("axon.routers.map.discover_links", fake_links)
|
|
|
|
from axon.routers.map import router
|
|
|
|
app = FastAPI()
|
|
app.include_router(router)
|
|
client = TestClient(app)
|
|
|
|
resp = client.post("/map", json={"url": "https://example.com/start", "max_depth": 1})
|
|
assert resp.status_code == 200
|
|
data = resp.json()
|
|
assert "https://example.com/start" in data["urls"]
|
|
assert "https://example.com/a" in data["urls"]
|
|
assert "https://other.com/b" not in data["urls"]
|