1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
|
"""URL scraping tool that fetches a page and extracts readable text."""
from __future__ import annotations
from bs4 import BeautifulSoup
from ..config import settings
from .helper.safe_http import UnsafeURL, fetch_public_bytes
def scrape_url(url: str, max_chars: int | None = None) -> str:
"""Fetch a web page and return its main text content.
Scripts, styles and other non-content elements are removed before
collapsing whitespace. The returned text is truncated to ``max_chars``.
Args:
url: The absolute URL of the page to scrape.
max_chars: Maximum number of characters to return. Defaults to the value
in settings.
Returns:
The extracted plain text of the page, truncated to ``max_chars``. If the
request fails a short error description is returned instead.
"""
if not url or not url.strip():
return "Invalid URL."
limit = max_chars if max_chars is not None and max_chars > 0 else settings.scrape_max_chars
try:
data, final_url, content_type = fetch_public_bytes(url, timeout=settings.scrape_timeout)
encoding = "utf-8"
if "charset=" in content_type:
encoding = content_type.rsplit("charset=", 1)[-1].split(";", 1)[0].strip()
html = data.decode(encoding or "utf-8", errors="replace")
url = final_url
except UnsafeURL as exc:
return f"Blocked URL: {exc}"
except Exception as exc: # noqa: BLE001
return f"Failed to fetch {url}: {exc}"
soup = BeautifulSoup(html, "html.parser")
# Remove non-content tags.
for tag in soup(["script", "style", "noscript", "template", "head", "nav", "footer", "aside"]):
tag.decompose()
# Prefer article/main content when available.
container = soup.find("article") or soup.find("main") or soup
text = container.get_text(separator="\n")
# Collapse blank lines and trailing whitespace per line.
lines = [ln.strip() for ln in text.splitlines()]
cleaned = "\n".join(ln for ln in lines if ln)
if len(cleaned) > limit:
cleaned = cleaned[:limit].rsplit(" ", 1)[0] + "\n…[truncated]"
return cleaned or "No readable text content found."
|
