fix(site-import): extract 清洗未闭合块/style 守卫/logo 日期钉位

评审收口三项：clean_html 配对块清理后兜底删 EOF 未闭合块（>500KB
截断防 script/style 泄漏误判 is_thin，顺序钉死配对先行）；style 属性
正则补 \s 守卫（data-style 不参与配色计数）；FIXTURE_WP 旧日期 logo
前置使日期优先真被甄别（首现即胜的伪绿不再成立）。

Co-Authored-By: Claude Code <noreply@anthropic.com>

app/platform/site_import/extract.py

@@ -19,12 +19,18 @@ _RE_BLOCK = re.compile(
     r"<(script|style|nav|footer|header)\b[^>]*>.*?</\1\s*>",
     re.IGNORECASE | re.DOTALL,
 )
+# EOF 未闭合块（>500KB 截断的 fetch 可能切在 script/style 中间）——配对清理后兜底
+_RE_BLOCK_UNCLOSED = re.compile(
+    r"<(script|style|nav|footer|header)\b[^>]*>.*$",
+    re.IGNORECASE | re.DOTALL,
+)
 _RE_COMMENT = re.compile(r"<!--.*?-->", re.DOTALL)
 _RE_TAG = re.compile(r"<[^>]+>")
 _RE_WS = re.compile(r"\s+")
 
 # —— extract_color：只看内联 style 属性 + 外部 css_text，不扫全文（避免误收 URL 等） ——
-_RE_STYLE_ATTR = re.compile(r"""style\s*=\s*["']([^"']*)["']""", re.IGNORECASE)
+# \sstyle 防 data-style 等复合属性名误命中（与 \ssrc/\shref 同款守卫）
+_RE_STYLE_ATTR = re.compile(r"""\sstyle\s*=\s*["']([^"']*)["']""", re.IGNORECASE)
 # 6 位优先、3 位展开；负向前瞻防截取更长 hex（如 #aabbccdd）的前缀
 _RE_HEX = re.compile(r"#([0-9a-fA-F]{6}|[0-9a-fA-F]{3})(?![0-9a-fA-F])")
 
@@ -61,8 +67,13 @@ _RESIDUE_PATTERNS: list[tuple[re.Pattern[str], str]] = [
 
 
 def clean_html(html_text: str) -> str:
-    """去 script/style/nav/footer/header/注释，剩余标签转文本，空白折叠，>8192 截断。"""
+    """去 script/style/nav/footer/header/注释，剩余标签转文本，空白折叠，>8192 截断。
+
+    先删配对块、再兜底删 EOF 未闭合块——顺序不可倒（未闭合模式会吞掉
+    配对块之后的全部文本）。
+    """
     t = _RE_BLOCK.sub(" ", html_text)
+    t = _RE_BLOCK_UNCLOSED.sub(" ", t)
     t = _RE_COMMENT.sub(" ", t)
     t = _RE_TAG.sub(" ", t)
     t = _html.unescape(t)

tests/test_site_import_extract.py

@@ -29,6 +29,7 @@ FIXTURE_WP = """<!DOCTYPE html>
   times on every order of building products.</p>
   <div class="hero" style="background:#ffffff">
     <img src="https://ex.com/wp-content/uploads/2026/01/hero-banner.jpg" alt="hero">
+    <img src="https://ex.com/wp-content/uploads/2020/02/logo-old.png" alt="old logo">
     <a href="/"><img src="https://ex.com/wp-content/uploads/2026/03/logo-new.png" alt="logo"></a>
   </div>
   <section style="background-color:#ffffff">
@@ -41,7 +42,6 @@ FIXTURE_WP = """<!DOCTYPE html>
     <a href="https://wa.me/8618603858592">WhatsApp us</a> |
     <a href="mailto:sales@hesuu.com">Email sales</a> |
     <a href="https://blog.other.com/building-tips">External blog</a>
-    <img src="https://ex.com/wp-content/uploads/2020/02/logo-old.png" alt="old logo">
   </section>
   <section class="cards" style="background:#ffffff">
     <div class="card" style="background:#ffffff;border-top:4px solid #FCCA03">
@@ -96,6 +96,9 @@ FIXTURE_CROSS = """<html><body>
 <a href="javascript:void(0)">JS</a>
 </body></html>"""
 
+# >500KB 抓取截断可能切在 script 中间——EOF 未闭合块不得泄漏进清洗文本
+FIXTURE_UNCLOSED = '<html><body><p>keep this text</p><script>var j = "JAVAJUNK";'
+
 
 class TestCleanHtml:
     def test_去script_style_nav_footer_header_comment(self):
@@ -108,6 +111,11 @@ class TestCleanHtml:
         t = clean_html(FIXTURE_9K)
         assert len(t) <= 8192
 
+    def test_EOF未闭合script不泄漏(self):
+        t = clean_html(FIXTURE_UNCLOSED)
+        assert "JAVAJUNK" not in t
+        assert "keep this text" in t
+
 
 class TestExtractColor:
     def test_排除白黑灰取最高频(self):
@@ -125,6 +133,11 @@ class TestExtractColor:
         html = '<span style="color:#fc3">x</span><b style="color:#fc3">y</b>'
         assert extract_color(html, None) == "#ffcc33"
 
+    def test_data_style不参与计数(self):
+        # #123456 为可配色——若 data-style 被误收则返回它而非 None
+        html = '<div data-style="color:#123456">x</div>'
+        assert extract_color(html, None) is None
+
 
 class TestExtractLogoUrl:
     def test_取日期新者(self):