fix(web): add read() via Jina Reader + skill install fallback (#238)
- WebChannel.read(): reads any URL via Jina Reader (r.jina.ai), returns Markdown - _install_skill(): add fallback from importlib.resources to Path(__file__) for editable installs - Explicit UTF-8 encoding on all file read/write operations Inspired by PR #215, implemented correctly (Jina instead of raw requests). Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
@@ -1,8 +1,11 @@
|
||||
# -*- coding: utf-8 -*-
|
||||
"""Web — any URL via Jina Reader. Always available."""
|
||||
|
||||
import urllib.request
|
||||
from .base import Channel
|
||||
|
||||
_UA = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36"
|
||||
|
||||
|
||||
class WebChannel(Channel):
|
||||
name = "web"
|
||||
@@ -15,3 +18,15 @@ class WebChannel(Channel):
|
||||
|
||||
def check(self, config=None):
|
||||
return "ok", "通过 Jina Reader 读取任意网页(curl https://r.jina.ai/URL)"
|
||||
|
||||
def read(self, url: str) -> str:
|
||||
"""通过 Jina Reader 读取网页,返回 Markdown 全文。"""
|
||||
if not url.startswith(("http://", "https://")):
|
||||
url = "https://" + url
|
||||
jina_url = f"https://r.jina.ai/{url}"
|
||||
req = urllib.request.Request(
|
||||
jina_url,
|
||||
headers={"User-Agent": _UA, "Accept": "text/plain"},
|
||||
)
|
||||
with urllib.request.urlopen(req, timeout=30) as resp:
|
||||
return resp.read().decode("utf-8")
|
||||
|
||||
Reference in New Issue
Block a user