ReRU: Посты русскоязычного Fediverse, собранные из ботов-ретрансляторов

+ blob - /dev/null blob + a44392627831d28bdea68f8e7d2e6260fc8ea0fb (mode 644) --- /dev/null +++ data/style.css @@ -0,0 +1,7 @@ +table { + border-collapse: collapse; + width: 100%; +} +tr { + border: 1px solid; +} blob - /dev/null blob + 613715b01b1faedc454b6450c1489cfbeafebba9 (mode 644) --- /dev/null +++ reru.py @@ -0,0 +1,213 @@ +import json + +from argparse import ArgumentParser +from concurrent.futures import ThreadPoolExecutor, as_completed +from datetime import datetime +from http import HTTPStatus +from http.server import HTTPServer, BaseHTTPRequestHandler +from urllib.request import urlopen +import xml.etree.ElementTree as ET + + +_FROM = [ + ("lor.sh", "ru"), + ("mastodon.ml", "rf"), + ("mastodon.social", "russian_mastodon"), + ("lor.sh", "rur"), +] + + +def _lookup_account_id(host, name): + with urlopen(f"https://{host}/api/v1/accounts/lookup?acct={name}") as f: + return json.loads(f.read().decode("utf-8"))["id"] + + +def _get_accounts_statuses(host, account_id): + with urlopen(f"https://{host}/api/v1/accounts/{account_id}/statuses") as f: + return json.loads(f.read().decode("utf-8")) + + +def _worker(host, name): + feeds = [] + for status in _get_accounts_statuses(host, _lookup_account_id(host, name)): + reblog = status.get("reblog") + if not reblog: + continue + feeds.append(reblog) + return feeds + + +def _harvest_reblogs(executor): + already = set() + feeds = [] + + futures = [executor.submit(_worker, *args) for args in _FROM] + for f in as_completed(futures): + for reblog in f.result(): + if reblog["url"] in already: + continue + already.add(reblog["url"]) + feeds.append(reblog) + + feeds.sort(key=lambda status: datetime.fromisoformat(status["created_at"]), reverse=True) + return feeds + + +def _img(src, alt, **kwargs): + img = ET.Element("img") + img.attrib.update(src=src, alt=alt, **kwargs) + return ET.tostring(img, encoding="unicode") + + +def _make_table(feeds): + table = " \n" + for status in feeds: + avatar_static = (status.get("account") or {}).get("avatar_static") + avatar = "" + if avatar_static: + avatar = _img(src=avatar_static, alt="avatar", height="48", width="48") + user_name = "
".join([ + (status.get("account") or {}).get("display_name", ""), + (status.get("account") or {}).get("acct", ""), + ]) + table += ( + " \n" + f" \n" + f" \n" + f" \n" + f" \n" + " \n" + ) + + poll = status.get("poll") + if poll: + table += ( + " \n" + " \n" + " \n" + " \n" + " \n" + " \n" + ) + + media_attachments = status.get("media_attachments") + for attach in media_attachments or []: + url = attach.get("remote_url") or attach.get("url") + if not url: + continue + table += ( + " \n" + " \n" + " \n" + " \n" + f' \n" + " \n" + ) + table += "

{avatar}	{user_name}	↗	{status.get('content') or ''}
			\n" + " " + ) + table += "".join([ + f' {option.get("title", "")} ({option.get("votes_count", 0)}) ' + for option in (poll.get("options") or []) + ]) + table += ( + " \n" + "
			' + ) + if attach.get("type") == "image": + table += _img(src=url, alt=attach.get("description") or "", width="640") + else: + table += url + table += ( + "

\n" + return table + + +class _HTTPServer(HTTPServer): + def __init__( + self, + *args, + executor=None, + header_file_path=None, + footer_file_path=None, + icon_file_path=None, + css_file_path=None, + **kwargs + ): + super().__init__(*args, **kwargs) + self.harvester_executor = executor + with open(header_file_path, "rb") as f: + self.header_file_bytes = f.read() + with open(footer_file_path, "rb") as f: + self.footer_file_bytes = f.read() + if icon_file_path: + with open(icon_file_path, "rb") as f: + self.icon_file_bytes = f.read() + else: + self.icon_file_bytes = None + if css_file_path: + with open(css_file_path, "rb") as f: + self.css_file_bytes = f.read() + else: + self.css_file_bytes = None + + +class _RequestHandler(BaseHTTPRequestHandler): + def do_GET(self): + if self.path in {"/index.html", "/index.htm", "/index", "/"}: + self.send_response(HTTPStatus.OK) + self.send_header("Content-type", "text/html") + self.end_headers() + self.wfile.write(self.server.header_file_bytes) + self.wfile.write(_make_table(_harvest_reblogs(self.server.harvester_executor)).encode()) + self.wfile.write(self.server.footer_file_bytes) + return + + if self.path == "/favicon.ico" and self.server.icon_file_bytes: + self.send_response(HTTPStatus.OK) + self.send_header("Content-type", "image/x-icon") + self.end_headers() + self.wfile.write(self.server.icon_file_bytes) + return + + if self.path == "/style.css" and self.server.css_file_bytes: + self.send_response(HTTPStatus.OK) + self.send_header("Content-type", "text/css") + self.end_headers() + self.wfile.write(self.server.css_file_bytes) + return + + self.send_error(HTTPStatus.NOT_FOUND, "File not found") + return + + +def _main(): + parser = ArgumentParser() + parser.add_argument("--address", default="127.0.0.1", help="bind to this address (default: %(default)s)") + parser.add_argument("--port", default=8000, type=int, help="bind to this port (default: %(default)s)") + parser.add_argument("--header", required=True, help="path to `index.html.header`") + parser.add_argument("--footer", required=True, help="path to `index.html.footer`") + parser.add_argument("--icon", help="path to `favicon.ico`") + parser.add_argument("--css", help="path to `style.css`") + args = parser.parse_args() + + with ThreadPoolExecutor(max_workers=len(_FROM)) as executor: + with _HTTPServer( + (args.address, args.port), + _RequestHandler, + executor=executor, + header_file_path=args.header, + footer_file_path=args.footer, + icon_file_path=args.icon, + css_file_path=args.css, + ) as httpd: + sock_host, sock_port = httpd.socket.getsockname()[:2] + print(f"HTTP server started ({sock_host}:{sock_port})...") + try: + httpd.serve_forever() + except KeyboardInterrupt: + print("\nKeyboard interrupt received, exiting.") + + +if __name__ == '__main__': + _main()