microblog/app/utils/opengraph.py

149 lines
4.1 KiB
Python

import asyncio
import mimetypes
import re
from typing import Any
from urllib.parse import urlparse
import httpx
from bs4 import BeautifulSoup # type: ignore
from loguru import logger
from pydantic import BaseModel
from app import ap_object
from app import config
from app.actor import LOCAL_ACTOR
from app.actor import fetch_actor
from app.database import AsyncSession
from app.models import InboxObject
from app.models import OutboxObject
from app.utils.url import is_url_valid
from app.utils.url import make_abs
class OpenGraphMeta(BaseModel):
url: str
title: str
image: str | None
description: str | None
site_name: str
def _scrap_og_meta(url: str, html: str) -> OpenGraphMeta | None:
soup = BeautifulSoup(html, "html5lib")
ogs = {
og.attrs["property"]: og.attrs.get("content")
for og in soup.html.head.findAll(property=re.compile(r"^og"))
}
# FIXME some page have no <title>
raw = {
"url": url,
"title": soup.find("title").text.strip(),
"image": None,
"description": None,
"site_name": urlparse(url).hostname,
}
for field in OpenGraphMeta.__fields__.keys():
og_field = f"og:{field}"
if ogs.get(og_field):
raw[field] = ogs.get(og_field, None)
if "title" not in raw:
return None
for maybe_rel in {"url", "image"}:
if u := raw.get(maybe_rel):
raw[maybe_rel] = make_abs(u, url)
return OpenGraphMeta.parse_obj(raw)
async def external_urls(
db_session: AsyncSession,
ro: ap_object.RemoteObject | OutboxObject | InboxObject,
) -> set[str]:
note_host = urlparse(ro.ap_id).hostname
tags_hrefs = set()
for tag in ro.tags:
if tag_href := tag.get("href"):
tags_hrefs.add(tag_href)
if tag.get("type") == "Mention":
if tag["href"] != LOCAL_ACTOR.ap_id:
mentioned_actor = await fetch_actor(db_session, tag["href"])
tags_hrefs.add(mentioned_actor.url)
tags_hrefs.add(mentioned_actor.ap_id)
else:
tags_hrefs.add(LOCAL_ACTOR.ap_id)
tags_hrefs.add(LOCAL_ACTOR.url)
urls = set()
if ro.content:
soup = BeautifulSoup(ro.content, "html5lib")
for link in soup.find_all("a"):
h = link.get("href")
ph = urlparse(h)
mimetype, _ = mimetypes.guess_type(h)
if (
ph.scheme in {"http", "https"}
and ph.hostname != note_host
and is_url_valid(h)
and (
not mimetype
or mimetype.split("/")[0] not in ["image", "video", "audio"]
)
):
urls.add(h)
return urls - tags_hrefs
async def _og_meta_from_url(url: str) -> OpenGraphMeta | None:
async with httpx.AsyncClient() as client:
resp = await client.get(
url,
headers={
"User-Agent": config.USER_AGENT,
},
follow_redirects=True,
)
resp.raise_for_status()
if not (ct := resp.headers.get("content-type")) or not ct.startswith("text/html"):
return None
try:
return _scrap_og_meta(url, resp.text)
except Exception:
logger.info(f"Failed to scrap OG meta for {url}")
return None
async def og_meta_from_note(
db_session: AsyncSession,
ro: ap_object.RemoteObject,
) -> list[dict[str, Any]]:
og_meta = []
urls = await external_urls(db_session, ro)
logger.debug(f"Lookig OG metadata in {urls=}")
for url in urls:
logger.debug(f"Processing {url}")
try:
maybe_og_meta = None
try:
maybe_og_meta = await asyncio.wait_for(
_og_meta_from_url(url),
timeout=5,
)
except asyncio.TimeoutError:
logger.info(f"Timing out fetching {url}")
except Exception:
logger.exception(f"Failed scrap OG meta for {url}")
if maybe_og_meta:
og_meta.append(maybe_og_meta.dict())
except httpx.HTTPError:
pass
return og_meta