From 92a8d61e6dec3ba1e9eb835cd2f7f603d01bbfae Mon Sep 17 00:00:00 2001 From: Harold <8866033@gmail.com> Date: Mon, 20 Apr 2026 11:00:03 +0800 Subject: [PATCH] feat: add zxzj detail parsing --- py/tests/test_在线之家.py | 66 ++++++++++++++++ py/在线之家.py | 154 ++++++++++++++++++++++++++++++++++++++ 2 files changed, 220 insertions(+) diff --git a/py/tests/test_在线之家.py b/py/tests/test_在线之家.py index cf02424..fe4031a 100644 --- a/py/tests/test_在线之家.py +++ b/py/tests/test_在线之家.py @@ -98,6 +98,72 @@ class TestZXZJSpider(unittest.TestCase): self.assertEqual(result["list"][0]["vod_name"], "搜索命中") self.assertNotIn("pagecount", result) + def test_detect_pan_type_prefers_share_domain(self): + self.assertEqual(self.spider._detect_pan_type("百度网盘", "https://pan.quark.cn/s/demo"), "quark") + self.assertEqual(self.spider._detect_pan_type("资源链接", "https://drive.uc.cn/s/demo"), "uc") + self.assertEqual(self.spider._detect_pan_type("阿里网盘", "https://www.alipan.com/s/demo"), "aliyun") + + @patch.object(Spider, "_request_html") + def test_detail_content_merges_zxzj_and_netdisk_groups(self, mock_request_html): + mock_request_html.side_effect = [ + """ +
+
+

示例剧 2025 日本 剧情

+

导演：导演甲

+

主演：演员甲演员乙

+
+
一段剧情简介
+

在线播放

+ +

夸克网盘

+ +

百度云

+ + """, + '', + '', + '', + ] + result = self.spider.detailContent(["voddetail/999.html"]) + vod = result["list"][0] + self.assertEqual(vod["vod_id"], "voddetail/999.html") + self.assertEqual(vod["vod_name"], "示例剧") + self.assertEqual(vod["vod_pic"], "https://www.zxzjhd.com/poster.jpg") + self.assertEqual(vod["vod_year"], "2025") + self.assertEqual(vod["vod_area"], "日本") + self.assertEqual(vod["vod_class"], "剧情") + self.assertEqual(vod["vod_director"], "导演甲") + self.assertEqual(vod["vod_actor"], "演员甲,演员乙") + self.assertEqual(vod["vod_content"], "一段剧情简介") + self.assertEqual(vod["vod_play_from"], "zxzj$$$quark$$$baidu") + self.assertEqual( + vod["vod_play_url"], + "第1集$vodplay/999-1-1.html#第2集$vodplay/999-1-2.html$$$夸克一$https://pan.quark.cn/s/q-demo$$$百度合集$https://pan.baidu.com/s/b-demo", + ) + + @patch.object(Spider, "_request_html") + def test_extract_pan_groups_skips_invalid_or_unknown_links(self, mock_request_html): + mock_request_html.side_effect = [ + '', + 'empty', + ] + groups = self.spider._extract_pan_groups( + [ + {"name": "无效一", "url": "https://www.zxzjhd.com/vodplay/x-1.html", "tab_name": "网盘资源"}, + {"name": "无效二", "url": "https://www.zxzjhd.com/vodplay/x-2.html", "tab_name": "网盘资源"}, + ] + ) + self.assertEqual(groups, []) + if __name__ == "__main__": unittest.main() diff --git a/py/在线之家.py b/py/在线之家.py index 533568c..f14b8cf 100644 --- a/py/在线之家.py +++ b/py/在线之家.py @@ -225,3 +225,157 @@ class Spider(BaseSpider): url = self._build_url("/vodsearch/{0}-------------.html".format(quote(str(key or "").strip()))) items = self._parse_cards(self._request_html(url)) return {"list": items, "page": page, "limit": len(items), "total": len(items)} + + def _extract_actor_like_field(self, html, label): + matched = re.search(rf"{label}：([\s\S]*?)

", str(html or "")) + if not matched: + return "" + texts = re.findall(r">([^<>]+)<", matched.group(1)) + clean = [self._clean_text(text) for text in texts if self._clean_text(text)] + return ",".join(clean) + + def _parse_detail_meta(self, html): + root = self.html(html) + if root is None: + return {} + title_line = self._clean_text( + "".join(root.xpath("//*[contains(@class,'stui-content__detail')]//*[contains(@class,'title')][1]//text()")) + ) + title_match = re.match(r"^(.*?)(?:\s+(19\d{2}|20\d{2}))?(?:\s+([^\s]+))?(?:\s+([^\s]+))?$", title_line) + vod_name = self._clean_text(title_match.group(1) if title_match else title_line) + vod_year = self._clean_text(title_match.group(2) if title_match else "") + vod_area = self._clean_text(title_match.group(3) if title_match else "") + vod_class = self._clean_text(title_match.group(4) if title_match else "") + return { + "vod_name": vod_name, + "vod_year": vod_year, + "vod_area": vod_area, + "vod_class": vod_class, + "vod_pic": self._build_url( + ((root.xpath("//*[contains(@class,'stui-content__thumb')]//img/@data-original") or [""])[0]).strip() + ), + "vod_content": self._clean_text( + "".join( + root.xpath( + "//*[contains(concat(' ', normalize-space(@class), ' '), ' detail ')][1]//text()" + ) + ) + ), + "vod_director": self._extract_actor_like_field(html, "导演"), + "vod_actor": self._extract_actor_like_field(html, "主演"), + } + + def _extract_playlists(self, html): + root = self.html(html) + if root is None: + return {"normal": [], "pan": []} + tabs = [self._clean_text("".join(node.xpath(".//text()"))) for node in root.xpath("//*[contains(@class,'stui-vodlist__head')]//h3")] + playlists = root.xpath("//*[contains(@class,'stui-content__playlist')]") + normal = [] + pan = [] + for index, playlist in enumerate(playlists): + tab_name = tabs[index] if index < len(tabs) else "" + items = [] + for anchor in playlist.xpath(".//a[@href]"): + href = ((anchor.xpath("./@href") or [""])[0]).strip() + name = self._clean_text("".join(anchor.xpath(".//text()"))) or "正片" + play_id = re.sub(r"^/+", "", href) + if not play_id: + continue + items.append( + { + "name": name, + "url": self._build_url(play_id), + "play_id": play_id, + "tab_name": tab_name, + } + ) + if any(keyword in tab_name for keyword in ["网盘", "百度", "夸克", "UC", "阿里", "迅雷"]): + pan.extend(items) + elif items: + normal.extend([f"{item['name']}${item['play_id']}" for item in items]) + return {"normal": normal, "pan": pan} + + def _extract_pan_url_from_play_page(self, html): + matched = re.search(r"player_[a-z0-9_]+\s*=\s*(\{[\s\S]*?\})\s*;?", str(html or ""), re.I) + if not matched: + return "" + try: + payload = json.loads(matched.group(1)) + except Exception: + return "" + return str(payload.get("url") or "").strip() + + def _detect_pan_type(self, tab_name, share_url): + text = str(share_url or "") + if "pan.baidu.com" in text: + return "baidu" + if "pan.quark.cn" in text: + return "quark" + if "drive.uc.cn" in text: + return "uc" + if "alipan.com" in text or "aliyundrive.com" in text: + return "aliyun" + if "pan.xunlei.com" in text: + return "xunlei" + name = str(tab_name or "").lower() + if "百度" in str(tab_name or ""): + return "baidu" + if "夸克" in str(tab_name or ""): + return "quark" + if "uc" in name: + return "uc" + if "阿里" in str(tab_name or "") or "aliyun" in name: + return "aliyun" + if "迅雷" in str(tab_name or ""): + return "xunlei" + return "" + + def _extract_pan_groups(self, items): + grouped = {} + seen = set() + order = ["quark", "baidu", "uc", "aliyun", "xunlei"] + for item in items: + play_html = self._request_html(item["url"], referer=self.headers["Referer"]) + share_url = self._extract_pan_url_from_play_page(play_html) + pan_type = self._detect_pan_type(item.get("tab_name", ""), share_url) + if not share_url or not pan_type: + continue + key = (pan_type, share_url) + if key in seen: + continue + seen.add(key) + grouped.setdefault(pan_type, []).append(f"{item['name']}${share_url}") + return [{"from": key, "urls": "#".join(grouped[key])} for key in order if grouped.get(key)] + + def detailContent(self, ids): + raw_id = str(ids[0]).strip() + url = raw_id if raw_id.startswith("http") else self._build_url("/" + raw_id.lstrip("/")) + html = self._request_html(url) + meta = self._parse_detail_meta(html) + playlists = self._extract_playlists(html) + pan_groups = self._extract_pan_groups(playlists["pan"]) + play_from = [] + play_url = [] + if playlists["normal"]: + play_from.append("zxzj") + play_url.append("#".join(playlists["normal"])) + for group in pan_groups: + play_from.append(group["from"]) + play_url.append(group["urls"]) + vod = { + "vod_id": raw_id, + "vod_name": meta.get("vod_name", ""), + "vod_pic": meta.get("vod_pic", ""), + "vod_content": meta.get("vod_content", ""), + "vod_remarks": "", + "vod_year": meta.get("vod_year", ""), + "vod_area": meta.get("vod_area", ""), + "vod_class": meta.get("vod_class", ""), + "vod_lang": "", + "vod_director": meta.get("vod_director", ""), + "vod_actor": meta.get("vod_actor", ""), + "vod_play_from": "$$$".join(play_from), + "vod_play_url": "$$$".join(play_url), + } + return {"list": [vod]}