fix(indexer): 修复 NexusPHP 用户流量详情表格兜底解析 (#7008)

app/modules/indexer/parser/__init__.py

@@ -98,9 +98,9 @@ class SiteParserBase(metaclass=ABCMeta):
         self.join_at = None
         self.bonus = 0.0
 
-        # 流量信息
-        self.upload = 0.0
-        self.download = 0.0
+        # 上传/下载量按整数字节计，分享率为浮点数
+        self.upload = 0
+        self.download = 0
         self.ratio = 0.0
 
         # 做种信息

app/modules/indexer/parser/nexus_php.py

@@ -650,9 +650,42 @@ class NexusPhpSiteUserInfo(SiteParserBase):
         return message_head_text, message_date_text, message_content_text
 
     def _fixup_traffic_info(self, html):
-        """从详情页补充首页缺失的魔力值。"""
+        """从详情表格补读为零的流量、分享率和魔力值，非零值保持不变。"""
+        if not self.upload:
+            upload_text = self._traffic_table_value(html, r"上[传傳]量?")
+            if upload_text is not None:
+                self.upload = self.num_filesize(upload_text)
+        if not self.download:
+            download_text = self._traffic_table_value(html, r"下[载載]量?")
+            if download_text is not None:
+                self.download = self.num_filesize(download_text)
+        if not self.ratio:
+            ratio_text = self._traffic_table_value(html, "分享率")
+            if ratio_text is not None:
+                ratio = text_tools.parse_float(ratio_text)
+            else:
+                ratio = round(self.upload / self.download, 3) if self.download else 0.0
+            self.ratio = ratio
+
         # fixup bonus
         if not self.bonus:
             bonus_text = html.xpath('//tr/td[text()="魔力值" or text()="猫粮"]/following-sibling::td[1]/text()')
             if bonus_text:
                 self.bonus = text_tools.parse_float(bonus_text[0].strip())
+
+    @staticmethod
+    def _traffic_table_value(html: Any, field: str) -> Optional[str]:
+        """按单元格字段名读取数值，支持同格或相邻列布局，排除“实际”流量。"""
+        for cell in html.xpath('//tr/td[not(.//td)]'):
+            cell_text = " ".join(cell.xpath("string(.)").split())
+            field_match = re.fullmatch(rf"{field}\s*[:：]?\s*(.*)", cell_text)
+            if not field_match:
+                continue
+            value = field_match.group(1).strip()
+            if not value:
+                siblings = cell.xpath('following-sibling::td[1]')
+                value = " ".join(siblings[0].xpath("string(.)").split()) if siblings else ""
+            value_match = re.fullmatch(r"([\d,.]+(?:\s*[KMGTPI]i?B)?)(?:\s*[（(].*)?", value, re.I)
+            if value_match:
+                return value_match.group(1)
+        return None

tests/fixtures/depthstudio_userdetails.html

@@ -0,0 +1,47 @@
+<!DOCTYPE html>
+<html>
+<body>
+<!-- 从用户提供的详情页摘取流量结构，省略身份、链接和其他个人信息。 -->
+<div class="info-card-section info-card-stats info-card-stats-count-6">
+    <span class="info-stat info-stat-primary info-stat-ratio">
+        <span class="info-stat-copy">
+            <span class="info-stat-label color_ratio">分享率</span>
+            <span class="info-stat-value">4.393</span>
+        </span>
+    </span>
+    <span class="info-stat info-stat-primary info-stat-uploaded">
+        <span class="info-stat-copy">
+            <span class="info-stat-label color_uploaded">上传量</span>
+            <span class="info-stat-value">
+                323.04 <span class="info-stat-unit">GB</span>
+            </span>
+        </span>
+    </span>
+    <span class="info-stat info-stat-primary info-stat-downloaded">
+        <span class="info-stat-copy">
+            <span class="info-stat-label color_downloaded">下载量</span>
+            <span class="info-stat-value">
+                73.54 <span class="info-stat-unit">GB</span>
+            </span>
+        </span>
+    </span>
+</div>
+<table border="0" cellspacing="0" cellpadding="0">
+    <tr>
+        <td class="embedded">
+            <strong>分享率</strong>: <font color="">4.392</font>
+            （<strong>实际分享率</strong>：0.045）
+        </td>
+    </tr>
+    <tr>
+        <td class="embedded"><strong>上传量</strong>: 323.04 GB</td>
+        <td class="embedded">&nbsp;&nbsp;<strong>下载量</strong>: 73.54 GB</td>
+    </tr>
+    <tr>
+        <td class="embedded"><strong>实际上传量</strong>: 317.04 GB</td>
+        <td class="embedded">&nbsp;&nbsp;<strong>实际下载量</strong>: 6.838 TB</td>
+        <td class="embedded text-muted">实际上传/下载量 (仅用于记录, 不参与分享率计算)</td>
+    </tr>
+</table>
+</body>
+</html>

tests/test_discuz_parser.py

@@ -19,10 +19,13 @@ def parser() -> Iterator[DiscuzUserInfo]:
     instance.close()
 
 
-def test_discuz_traffic_defaults_are_floats(parser: DiscuzUserInfo) -> None:
-    for value in (parser.upload, parser.download, parser.ratio):
-        assert value == 0.0
-        assert isinstance(value, float)
+def test_discuz_traffic_defaults_use_integer_bytes(parser: DiscuzUserInfo) -> None:
+    """上传/下载量使用整数字节，分享率保持浮点数。"""
+    for value in (parser.upload, parser.download):
+        assert value == 0
+        assert type(value) is int
+    assert parser.ratio == 0.0
+    assert isinstance(parser.ratio, float)
 
 
 @pytest.mark.parametrize("include_menu", [True, False])

tests/test_nexus_php_parser.py

@@ -1,6 +1,8 @@
 # -*- coding: utf-8 -*-
+from pathlib import Path
 from urllib.parse import parse_qs, urlsplit
 
+import pytest
 from lxml import etree
 
 from app.modules.indexer.parser.nexus_php import NexusPhpSiteUserInfo
@@ -19,6 +21,121 @@ def _build_parser() -> NexusPhpSiteUserInfo:
     )
 
 
+def test_nexus_php_depthstudio_traffic_falls_back_to_details():
+    """首页解析失败后从详情表格补齐流量，不依赖卡片类名或实际流量。"""
+    html_text = (Path(__file__).parent / "fixtures" / "depthstudio_userdetails.html").read_text(encoding="utf-8")
+    html = etree.HTML(html_text)
+    homepage = etree.tostring(html.xpath("//div")[0], encoding="unicode")
+    homepage = homepage.replace("info-stat", "other-style")
+    parser = _build_parser()
+
+    parser._parse_user_base_info(homepage)
+
+    assert parser.upload == 0
+    assert parser.download == 0
+
+    parser._parse_user_detail_info(html_text)
+
+    assert parser.upload == round(323.04 * 1024 ** 3)
+    assert parser.download == round(73.54 * 1024 ** 3)
+    assert parser.ratio == 4.393
+
+
+def test_nexus_php_detail_traffic_preserves_nonzero_values():
+    """详情兜底不能覆盖已有的非零流量和分享率。"""
+    parser = _build_parser()
+    parser._parse_user_base_info(
+        "<html><body>上传量: 1 GB 下载量: 2 GB 分享率: 0.5</body></html>"
+    )
+    original = parser.upload, parser.download, parser.ratio
+
+    parser._parse_user_detail_info(
+        '<table><tr><td><strong>上传量</strong>: 323.04 GB</td>'
+        '<td><strong>下载量</strong>: 73.54 GB</td></tr>'
+        '<tr><td><strong>分享率</strong>: 4.392</td></tr></table>'
+    )
+
+    assert (parser.upload, parser.download, parser.ratio) == original
+
+
+@pytest.mark.parametrize("homepage, expected_upload, expected_download, expected_ratio", [
+    ("上传量: 1 GB", 1024 ** 3, round(73.54 * 1024 ** 3), 4.392),
+    ("下载量: 0 B 分享率: 0", round(323.04 * 1024 ** 3), round(73.54 * 1024 ** 3), 4.392),
+    ("上传量: 0 B 下载量: 0 B 分享率: 0", round(323.04 * 1024 ** 3), round(73.54 * 1024 ** 3), 4.392),
+    ("上传量: 1 GB 下载量: 2 GB 分享率: 0", 1024 ** 3, 2 * 1024 ** 3, 4.392),
+])
+def test_nexus_php_detail_traffic_fills_zero_fields(
+    homepage, expected_upload, expected_download, expected_ratio
+):
+    """零值字段允许从详情补读，分享率优先取详情表格的显示值。"""
+    parser = _build_parser()
+    parser._parse_user_base_info(f"<html><body>{homepage}</body></html>")
+    detail = (Path(__file__).parent / "fixtures" / "depthstudio_userdetails.html").read_text(encoding="utf-8")
+
+    parser._parse_user_detail_info(detail)
+
+    assert parser.upload == expected_upload
+    assert parser.download == expected_download
+    assert parser.ratio == expected_ratio
+
+
+def test_nexus_php_zero_ratio_uses_final_traffic_without_displayed_ratio():
+    """详情未显示分享率时，按补读后的上传/下载量重新计算零分享率。"""
+    parser = _build_parser()
+    parser._parse_user_base_info("<html><body>上传量: 0 B 下载量: 0 B 分享率: 0</body></html>")
+
+    parser._parse_user_detail_info(
+        '<table><tr><td><strong>上传量</strong>: 8 GB</td>'
+        '<td><strong>下载量</strong>: 2 GB</td></tr></table>'
+    )
+
+    assert parser.upload == 8 * 1024 ** 3
+    assert parser.download == 2 * 1024 ** 3
+    assert parser.ratio == 4.0
+
+
+@pytest.mark.parametrize("table, expected_upload, expected_download, expected_ratio", [
+    ('<tr><td><strong>上传量</strong>: 1.5 <span>TB</span></td>'
+     '<td><strong>下载量</strong>: 512 <span>GB</span></td></tr>',
+     round(1.5 * 1024 ** 4), 512 * 1024 ** 3, 3.0),
+    ('<tr><td>上傳量</td><td>1,024 GB</td></tr><tr><td>下載量</td><td>512 GB</td></tr>'
+     '<tr><td>分享率</td><td>2.001</td></tr>',
+     1024 ** 4, 512 * 1024 ** 3, 2.001),
+    ('<tr><td><strong>实际上传量</strong>: 1 TB</td>'
+     '<td><strong>实际下载量</strong>: 1 TB</td></tr>', 0, 0, 0.0),
+    ('<tr><td>上传量</td><td>0 B</td></tr><tr><td>下载量</td><td>0 B</td></tr>', 0, 0, 0.0),
+    ('<tr><td>上传量</td><td>未知</td></tr><tr><td>下载量</td><td>未知</td></tr>', 0, 0, 0.0),
+])
+def test_nexus_php_detail_traffic_table(table, expected_upload, expected_download, expected_ratio):
+    """按表格字段识别流量，支持嵌套单位和分列布局，并排除实际流量。"""
+    parser = _build_parser()
+    parser._parse_user_base_info("<html><body></body></html>")
+
+    parser._parse_user_detail_info(f"<html><body><table>{table}</table></body></html>")
+
+    assert parser.upload == expected_upload
+    assert parser.download == expected_download
+    assert parser.ratio == expected_ratio
+
+
+@pytest.mark.parametrize("download, ratio_text, expected_ratio", [
+    ("512 GB", "分享率: 3.001", 3.001),
+    ("512 GB", "", 3.0),
+    ("0 B", "", 0.0),
+])
+def test_nexus_php_legacy_traffic(download, ratio_text, expected_ratio):
+    """保留传统页面解析、分享率计算及零下载量行为。"""
+    parser = _build_parser()
+
+    parser._parse_user_traffic_info(
+        f"<html><body>上传量: 1.5 TB 下载量: {download} {ratio_text}</body></html>"
+    )
+
+    assert parser.upload == round(1.5 * 1024 ** 4)
+    assert parser.download == (512 * 1024 ** 3 if download == "512 GB" else 0)
+    assert parser.ratio == expected_ratio
+
+
 def test_nexus_php_seeding_next_page_stops_when_userid_missing():
     """
     userid 未识别且下一页也缺少 userid 时应停止翻页而不是抛出异常。