import contextlib import io import json import os import textwrap import time import uuid import fastapi import httpx import logging import sqlite3 import appdirs from typing import Annotated from fastapi import Query, Header, HTTPException, Request from fastapi.responses import JSONResponse from pathlib import Path from bs4 import BeautifulSoup from rich.logging import RichHandler from fastapi.middleware.cors import CORSMiddleware @contextlib.asynccontextmanager async def startup(_): if not CACHE_DIR.exists(): CACHE_DIR.mkdir(parents=True) with sqlite3.connect(CACHE_FILE) as conn: conn.execute( """ CREATE TABLE IF NOT EXISTS cache ( uuid TEXT PRIMARY KEY, url TEXT NOT NULL, ts INTEGER NOT NULL, metadata TEXT NOT NULL ) """ ) yield logging.basicConfig( level=logging.getLevelName(os.environ.get("LOG_LEVEL", "INFO").upper()), format="%(asctime)s:%(levelname)s:%(name)s:%(message)s", datefmt="%d/%m/%Y %H:%M:%S" ) app = fastapi.FastAPI( root_path=os.environ.get("PREVIEW_ROOT_PATH", ""), lifespan=startup ) app.add_middleware( CORSMiddleware, allow_origins=["*"], allow_credentials=True, allow_methods=["GET", "OPTIONS"], allow_headers=["*", "Authorization"], ) MISSING_TOKEN = JSONResponse( {"errcode": "M_MISSING_TOKEN", "error": "Missing access token"}, 401 ) INVALID_TOKEN = JSONResponse( {"errcode": "M_INVALID_TOKEN", "error": "Invalid access token"}, 401 ) VALID_OG_TAGS = [ "og:title", "og:type", "og:image", "og:url", "og:audio", "og:description", "og:determiner", "og:locale", "og:locale:alternative", "og:site_name", "og:image:url", "og:image:secure_url", "og:image:type", "og:image:width", "og:image:height", "og:image:alt", "og:video", "og:video:url", "og:video:secure_url", "og:video:type", "og:video:width", "og:video:height", "og:video:alt", "og:video:actor", "og:video:actor:role", "og:video:director", "og:video:writer", "og:video:duration", "og:video:release_date", "og:video:tag", "og:video:series" "og:audio:url", "og:audio:secure_url", "og:audio:type", "og:music:duration", "og:music:album", "og:music:album:disc", "og:music:album:track", "og:music:musician", "og:music:song", "og:music:song:disc", "og:music:song:track", "og:music:release_date", "og:music:creator", "og:article:published_time", "og:article:modified_time", "og:article:expiration_time", "og:article:author", "og:article:section", "og:article:tag", "og:book:author", "og:book:tag", "og:book:isbn", "og:book:release_date", "og:profile:first_name", "og:profile:last_name", "og:profile:username", "og:profile:gender" ] URL_OG_TAGS = [ "og:video", "og:video:url", "og:video:secure_url", "og:image", "og:image:url", "og:image:secure_url", "og:audio", "og:audio:url", "og:audio:secure_url" ] TWITTER_MAPPING = { "twitter:site": "site_name", "twitter:creator": "site_name", "twitter:image": "image", "twitter:title": "title", "twitter:image:width": "image:width", "twitter:image:height": "image:height", } if Path.cwd() == Path("/app"): logging.info("Look to be running in a docker container. Cache will be stored in /app/cache.") CACHE_DIR = Path("/app/cache") else: CACHE_DIR = Path(appdirs.user_cache_dir("matrix-url-preview")) CACHE_DIR.mkdir(parents=True, exist_ok=True) CACHE_FILE = CACHE_DIR / "db.sqlite3" CACHE_FILE.touch(exist_ok=True) logging.debug("Cache file: %r", CACHE_FILE) def upload_media(domain: str, access_token: str, file: io.BytesIO, filename: str, content_type: str): file.seek(0) logging.info( "Creating media at %r called %r with the content type %r and %d bytes", domain, filename, content_type, len(file.getvalue()) ) file.seek(0) # noinspection PyTypeChecker response = httpx.post( "%s/_matrix/media/r0/upload" % domain, headers={ "Authorization": f"Bearer {access_token}", "Content-Type": content_type }, data=file.getvalue(), params={ "filename": filename } ) if response.status_code == 200: logging.info("Media uploaded successfully") mxc_url = response.json()["content_uri"] return mxc_url else: logging.warning("Failed to upload media: HTTP %s", response.status_code) logging.debug("Response: %r", response.text) return None @app.get("/preview_url") def preview_url( req: Request, url: Annotated[str, Query(..., description="URL to preview")], ts: int = Query(None, description="The preferred point in time to return a preview for."), access_token_qs: str | None = Query(None, alias="access_token", description="Access token to use for the request."), access_token_header: str | None = Header( None, alias="Authorization", description="Access token to use for the request." ), ): if access_token_qs is not None: access_token = access_token_qs elif access_token_header and access_token_header.startswith("Bearer "): access_token = access_token_header.split("Bearer ")[1] else: return MISSING_TOKEN with sqlite3.connect(CACHE_FILE) as conn: cursor = conn.cursor() cursor.execute( "SELECT metadata,ts FROM cache WHERE url = ?", (url,) ) results = cursor.fetchall() if results: for result in results: # find the one with the closest timestamp metadata, _ts = result if ts is None or abs(ts - _ts) < 3600: logging.debug("Optimal cache hit for %r", url) return json.loads(metadata) # No close matches, get the latest one metadata, _ts = results[-1] # If the latest one is more than 3 hours old, re-fetch. Otherwise, return. if ts is None or abs(ts - _ts) < 10800: logging.debug("Cache hit for %r", url) return json.loads(metadata) domain = os.environ.get("PREVIEW_HOMESERVER", "https://" + req.url.hostname) try: with httpx.Client( headers={ "User-Agent": "Mozilla/5.0 (X11; Linux x86_64; rv:122.0) Gecko/20100101 Firefox/122.0" }, timeout=60, follow_redirects=False ) as client: response = client.get( url, headers={ "User-Agent": "Mozilla/5.0 (X11; Linux x86_64; rv:122.0) Gecko/20100101 Firefox/122.0" }, ) response.raise_for_status() while response.next_request and response.next_request.url.host == response.url.host: response = client.send(response.next_request) response.raise_for_status() except httpx.HTTPError as e: raise HTTPException(500, f"Failed to fetch URL: {e!r}") if "text/html" not in response.headers.get("content-type", ""): return {} soup = BeautifulSoup(response.text, "html.parser") og_tags = {} for tag in soup.find_all("meta"): logging.debug("Found meta tag: %r", tag) if tag.get("property", "").startswith(("og:", "twitter:")): logging.debug( "Tag %r is an OG/Twitter tag, with property: %r", tag.get("property", "N/A"), textwrap.shorten(tag.get("content", "N/A"), 100), ) tag_name = tag.get("property") if tag_name in (*VALID_OG_TAGS, *TWITTER_MAPPING.keys()): og_tags[tag_name] = tag.get("content") for tag in og_tags.keys(): if tag.startswith("twitter:"): if tag in TWITTER_MAPPING: og_tags[TWITTER_MAPPING[tag]] = og_tags.pop(tag) logging.debug("Mapped twitter tag %r to og tag %r", tag, TWITTER_MAPPING[tag]) else: logging.warning("Unrecognized Twitter tag: %r", tag) og_tags.pop(tag, None) for tag_name in URL_OG_TAGS: if tag_name in og_tags: logging.debug("Retrieving tag %r to see if it needs uploading to Matrix", tag_name) _url = og_tags[tag_name] logging.debug("%r = %r", tag_name, _url) try: # noinspection PyArgumentList with httpx.stream( url=_url, method="GET", headers={ "User-Agent": "Mozilla/5.0 (X11; Linux x86_64; rv:122.0) Gecko/20100101 Firefox/122.0" }, timeout=60, follow_redirects=True ) as response_media: if response_media.status_code not in range(200, 300): logging.warning("Failed to fetch media: %r - HTTP %s", _url, response_media.status_code) og_tags.pop(tag_name, None) elif not response_media.headers.get("content-type", "").startswith(("image/", "video/", "audio/")): logging.warning("Failed to fetch media: %r - not a media type", _url) og_tags.pop(tag_name, None) else: logging.info( "Downloading {:,} bytes of media: {!r}".format( int(response_media.headers.get("content-length", 0)), _url ) ) _file = io.BytesIO() _file.write(response_media.read()) _file.seek(0) if "content-length" in response_media.headers: _file.seek(0, os.SEEK_END) if int(response_media.headers["content-length"]) != _file.tell(): logging.warning( "Possibly failed to fetch media: {!r} - incomplete ({:,} downloaded, {:,} needed)" .format( _url, _file.tell(), int(response_media.headers["content-length"]) ) ) _file.seek(0) upload_response = upload_media( domain, access_token, _file, Path(httpx.URL(_url).path).name, response_media.headers.get("content-type", "") ) if upload_response: og_tags["original:" + tag_name.replace("og:", "")] = og_tags[tag_name] og_tags[tag_name] = upload_response if tag_name in ["og:image", "og:image:url", "og:image:secure_url"]: _file.seek(0) og_tags["matrix:image:size"] = len(_file.getvalue()) logging.info("Uploaded media: %r" % _url) else: logging.warning("Failed to upload media: %r (no returned mxc)", _url) except httpx.HTTPError as e: logging.exception("Failed to fetch url for OG tags @ %r: %r", _url, e, exc_info=True) og_tags.pop(tag_name, None) for key in og_tags.copy().keys(): if not key.startswith(("original:", "og:", "matrix:")): value = og_tags.pop(key, None) og_tags["og:" + key] = value with sqlite3.connect(CACHE_FILE) as conn: conn.execute( "INSERT INTO cache (uuid, url, ts, metadata) VALUES (?, ?, ?, ?)", (str(uuid.uuid4()), url, round(time.time()), json.dumps(og_tags)) ) return JSONResponse( og_tags, 200, headers={ "Cache-Control": "public, max-age=86400" } ) if __name__ == "__main__": import uvicorn uvicorn.run( app, host=os.getenv("PREVIEW_HOST", "0.0.0.0"), port=int(os.getenv("PREVIEW_PORT", 2226)), # If you want to enable reverse-proxy support, you must set the $FORWARDED_ALLOW_IPS environment variable. # See: https://www.uvicorn.org/settings/#http )