""" 公共用例通过启动注入的 Module 能力解释站点协议,保留旧配置、链接和插件入口。 """ import base64 import hashlib import hmac import json from importlib import import_module from types import SimpleNamespace from unittest.mock import Mock import pytest from Crypto.Cipher import AES from Crypto.Util.Padding import unpad import app.application.rss as rss_module import app.modules.indexer as indexer import app.modules.indexer.download as download_protocol from app.application.rss import RssHelper, _require_rss_site_port, configure_rss_ports from app.application.search.session import torrent_snapshot from app.application.site.protocol import ( SiteConnectionResult, get_site_protocol_port, is_expiring_download_url, site_domain, torrent_resource_identity, ) from app.domain import site as site_rules from app.domain import torrent as torrent_rules from app.domain.context import TorrentInfo from app.sdk.security import nexusphp_encrypt from app.sdk.string import StringUtils from app.startup.composition import site as composition @pytest.fixture def protocol_boundary(monkeypatch): """ 用真实调度器执行索引模块,替换目录边界并禁止访问插件目录。 """ module = object.__new__(indexer.IndexerModule) calls = [] def running(method): """ 记录内部能力分发并只提供当前索引模块。 """ calls.append(method) return [module] monkeypatch.setattr(composition._SiteModuleCatalog, "get_running_modules", staticmethod(running)) plugins = Mock(side_effect=AssertionError("站点内部能力不能访问插件目录")) monkeypatch.setattr(composition._SiteModuleCatalog, "get_plugin_modules", plugins) composition.configure_site_protocol_composition() for name in ("_rss_http_port", "_rss_browser_port", "_rss_parser_port"): monkeypatch.setattr(rss_module, name, getattr(rss_module, name)) return SimpleNamespace(module=module, calls=calls, plugins=plugins) @pytest.mark.parametrize("url,expected", [ (None, ""), ("", ""), ("https://tracker.example.com/rss", "example.com"), ("https://u2.dmhy.org/torrents.php", "u2.dmhy.org"), ("https://www.U2.dmhy.org:8443/", "u2.dmhy.org"), (".u2.dmhy.org", "u2.dmhy.org"), ("https://pt.ecust.pp.ua/", "pt.ecust.pp.ua"), (".pt.gtkpw.xyz", "pt.gtkpw.xyz"), ("https://pt.gtk.pw/", "pt.gtk.pw"), ]) def test_domain_exceptions_are_dispatched_only_by_module(protocol_boundary, url, expected): """ 多级域名例外和普通域名都通过 Module 运行,插件和宿主消费相同结果。 """ assert site_domain(url) == expected assert StringUtils.get_url_domain(url) == expected assert protocol_boundary.calls == ["get_site_domain", "get_site_domain"] protocol_boundary.plugins.assert_not_called() def test_domain_matching_is_host_scoped_and_domain_stays_pure(protocol_boundary): """ 站点名出现在查询串或其他主机中不能命中例外,Domain 不加载模块规则。 """ assert site_rules.extract_domain("https://u2.dmhy.org/") == "dmhy.org" url = "https://other.example/?redirect=https://u2.dmhy.org/" assert site_domain(url) == "other.example" assert site_domain("https://u2.dmhy.org.evil.test/") != "u2.dmhy.org" def test_legacy_site_and_cookiecloud_use_module_domain_rules(protocol_boundary): """ 精确 Compat 路由保留旧插件域名函数和无参 CookieCloud 的归组行为。 """ legacy_site = import_module("app.utils.site") legacy_cookiecloud = import_module("app.helper.cookiecloud") assert legacy_site.extract_domain("https://u2.dmhy.org/") == "u2.dmhy.org" helper = legacy_cookiecloud.CookieCloudHelper() assert helper._domain_resolver(".u2.dmhy.org") == "u2.dmhy.org" @pytest.mark.parametrize("page_url,enclosure,expected", [ ("https://totheglory.im/t/17/", "https://totheglory.im/dl/17/old", ("7", "id=17")), ("https://totheglory.im/t/17/", "https://totheglory.im/download?id=23", ("7", "id=17")), ("https://kp.m-team.cc/detail/17", "https://tracker.example/download?id=23", ("7", "id=17")), ("https://tracker.example/detail/17", "[old]https://tracker.example/api", ("7", "id=17")), ("https://tracker.example/torrents.php?id=12&torrentid=17", None, ("7", "id=17")), ("https://tracker.example/t/17/", "https://tracker.example/dl/17/old", None), (None, "https://[invalid", None), ]) def test_module_identity_preserves_url_precedence_and_legacy_paths(protocol_boundary, page_url, enclosure, expected): """ 详情页身份优先于下载地址,迁移不能把临时票据或同形的其他站点路径当成 ID。 """ torrent = TorrentInfo(site=7, page_url=page_url, enclosure=enclosure) assert torrent_resource_identity(torrent) == expected assert torrent.page_url == page_url and torrent.enclosure == enclosure def test_old_checkpoint_restoration_does_not_persist_tickets(protocol_boundary): """ 旧 TTG 检查点按公开详情恢复身份,没有下载地址时继续交给原页重取。 """ torrent = TorrentInfo(site=7, page_url="https://totheglory.im/t/17/", enclosure="https://totheglory.im/dl/17/old") record = torrent_snapshot(torrent) assert record["enclosure"] is None restored = protocol_boundary.module.restore_search_torrent( {"id": 7, "cookie": "current-cookie"}, record, allow_missing=True, ) assert restored is not None and restored.enclosure is None assert torrent_resource_identity(restored) == ("7", "id=17") assert restored.site_cookie == "current-cookie" assert "current-cookie" not in json.dumps(record) def test_yema_checkpoint_keeps_existing_page_recovery(protocol_boundary): """ 旧 Yema fragment 和换票配方不进入检查点,迁移不假定可以按缺失的 ID 重建下载。 """ record = torrent_snapshot(TorrentInfo(site=7, page_url="https://www.yemapt.org/#/torrent/detail/17/", enclosure="[old-credentials]https://www.yemapt.org/openApi/download")) restored = protocol_boundary.module.restore_search_torrent({"parser": "Yema", "apikey": "current"}, record, allow_missing=True) assert restored is not None and restored.enclosure is None assert torrent_resource_identity(restored) is None assert "old-credentials" not in json.dumps(record) @pytest.mark.parametrize("url,expected", [ ("https://totheglory.im/dl/17/old", True), ("https://www.totheglory.im/dl/17/old/", True), ("https://tracker.example/download?id=17&t=123&sign=abc", True), ("https://tracker.example/dl/17/old", False), ("https://totheglory.im/rssdd.php?par=encoded", False), ("https://totheglory.im/dl/17/", False), (None, False), ]) def test_module_signature_rules_preserve_plain_and_indirect_links(protocol_boundary, url, expected): """ 专站签名解释通过 Module 分发,Domain 只保留通用查询签名。 """ assert is_expiring_download_url(url) is expected if url == "https://totheglory.im/dl/17/old": assert torrent_rules.is_expiring_download_url(url) is False @pytest.mark.parametrize("domain,path,xpath,params", [ ("hares.top", "getrss.php", "//*[@id='layui-layer100001']/div[2]/div/p[4]/a/@href", {"showrows": 50}), ("et8.org", "getrss.php", "//*[@id='outer']/table/tbody/tr/td/table/tbody/tr/td/a[2]/@href", {"showrows": 50}), ("pttime.org", "getrss.php", "//*[@id='outer']/table/tbody/tr/td/table/tbody/tr/td/text()[5]", {"showrows": 10}), ("ourbits.club", "getrss.php", "//a[@class='gen_rsslink']/@href", {"search_mode": 1}), ("totheglory.im", "rsstools.php?c51=51", "//textarea/text()", {"showrows": 50}), ("monikadesign.uk", "rss", "//a/@href", {"showrows": 50}), ("zhuque.in", "user/rss", "//a/@href", {"showrows": 50}), ("hdwing.org", "getrss.php", None, {"rsscart": 0}), ("audiences.me", "getrss.php", None, {"torrent_type": 1, "exp": 180}), ("shadowflow.org", "getrss.php", None, {"paid": 0, "search_mode": 0, "showrows": 30}), ("hddolby.com", "getrss.php", None, {"exp": 180}), ("hdhome.org", "getrss.php", None, {"exp": 180}), ("pthome.net", "getrss.php", None, {"exp": 180}), ("ptsbao.club", "getrss.php", None, {"size": 0}), ("leaves.red", "getrss.php", None, {"search_mode": 0, "paid": 2}), ("hdtime.org", "getrss.php", None, {"search_mode": 0}), ("m-team.io", "getrss.php", None, {"https": 1}), ("u2.dmhy.org", "getrss.php", None, {"inclautochecked": 1, "trackerssl": 1}), ]) def test_all_rss_site_recipes_are_dispatched_and_isolated(protocol_boundary, domain, path, xpath, params): """ 原有十八个专站 RSS 配方迁移后保留页面、选择器、参数和浏览器选择。 """ port = _require_rss_site_port() recipe = port.get_site_rss_recipe(f"https://{domain}/") assert recipe.url.startswith(f"https://{domain}/{path}") assert recipe.xpath == (xpath or "//a[@class='faqlink']/@href") assert all(recipe.params[key] == value for key, value in params.items()) assert recipe.render is (domain == "zhuque.in") recipe.params["local-change"] = "not-shared" assert "local-change" not in port.get_site_rss_recipe(f"https://{domain}/").params assert protocol_boundary.calls == ["get_site_rss_recipe", "get_site_rss_recipe"] @pytest.mark.parametrize("text,expected", [ ("RSS 链接已过期, 您需要获得一个新的!", None), ("RSS Link has expired, You need to get a new one!", None), ("RSS Link has expired, You need to get new!", None), ("ordinary error", False), ]) def test_rss_expiration_is_interpreted_by_module(protocol_boundary, text, expected): """ Application 只消费 Module 的过期事实,普通内容错误不触发 RSS 续期。 """ http = Mock(get=Mock(return_value=SimpleNamespace(status_code=200, text=text, content=b""))) configure_rss_ports(http=http, browser=Mock(), parser=Mock()) assert RssHelper().parse("https://tracker.example/rss") is expected assert "is_site_rss_expired" in protocol_boundary.calls def test_encryption_sdk_keeps_original_site_payload(protocol_boundary, monkeypatch): """ 旧 SDK 加密入口经过 Module 后仍保留随机 IV、MAC、字段和多字节文本口径。 """ monkeypatch.setattr(download_protocol.os, "urandom", lambda _length: b"x" * 16) key = b"k" * 32 payload = json.loads(base64.b64decode(nexusphp_encrypt("中文协议", key))) assert payload.keys() == {"iv", "value", "mac", "tag"} assert payload["tag"] == "" assert payload["mac"] == hmac.new(key, (payload["iv"] + payload["value"]).encode(), hashlib.sha256).hexdigest() cipher = AES.new(key, AES.MODE_CBC, base64.b64decode(payload["iv"])) assert unpad(cipher.decrypt(base64.b64decode(payload["value"])), AES.block_size).decode() == "中文协议" assert protocol_boundary.calls == ["encrypt_site_payload"] def test_unavailable_module_is_not_a_false_protocol_fact(protocol_boundary, monkeypatch): """ Module 缺失不能被布尔转换成普通链接或无过期响应,从而绕过下载安全策略。 """ monkeypatch.setattr(composition._SiteModuleCatalog, "get_running_modules", staticmethod(lambda _method: [])) with pytest.raises(RuntimeError, match="Module 不可用"): site_domain("https://u2.dmhy.org/") with pytest.raises(RuntimeError, match="Module 不可用"): is_expiring_download_url("https://totheglory.im/dl/17/old") with pytest.raises(RuntimeError, match="Module 不可用"): get_site_protocol_port().is_torrent_download_confirmation(b"page") with pytest.raises(RuntimeError, match="Module 不可用"): get_site_protocol_port().has_site_subtitle_api({"parser": "mTorrent"}) with pytest.raises(RuntimeError, match="Module 不可用"): _require_rss_site_port().is_site_rss_expired("expired") with pytest.raises(RuntimeError, match="Module 不可用"): _require_rss_site_port().get_site_rss_recipe("https://u2.dmhy.org/") def test_connection_unsupported_is_an_explicit_receipt(protocol_boundary, monkeypatch): """ 不支持检测与已知适配器能力缺失有不同回执,公共层不需要复制站点路由。 """ assert protocol_boundary.module.test_site({"domain": "https://ordinary.example/"}) == SiteConnectionResult(supported=False) monkeypatch.delitem(indexer.SPIDER_PARSER_CLASSES, "Piggo") receipt = protocol_boundary.module.test_site({"domain": "https://piggo.me/", "schema": "NexusPhp"}) assert receipt.supported and not receipt.success assert receipt.message == "站点适配器未提供连接检测能力" @pytest.mark.parametrize("resource", [{"schema": "MTorrent"}, {"parser": "mTorrent"}, {"schema": "TNode"}, {"parser": "TNodeSpider"}]) def test_missing_declared_protocol_is_not_generic_connectivity(protocol_boundary, monkeypatch, resource): """ 兼容站已经声明认证协议时,缺失该适配器也不能回退成网页成功。 """ for parser in ("mTorrent", "TNodeSpider"): monkeypatch.delitem(indexer.SPIDER_PARSER_CLASSES, parser) receipt = protocol_boundary.module.test_site({"domain": "https://mirror.example/", **resource}) assert receipt.supported and not receipt.success assert receipt.message == "站点适配器未提供连接检测能力"