实现 ClinDB 第一批清洗组件

tests/test_html_table_double_escape.py

@@ -0,0 +1,59 @@
+from __future__ import annotations
+
+import pytest
+
+from mdpolish import Pipeline, RunStatus
+from mdpolish.components import HtmlTableDoubleEscapeComponent
+
+
+def transform(markdown: str):  # type: ignore[no-untyped-def]
+    return Pipeline([HtmlTableDoubleEscapeComponent()]).transform(markdown)
+
+
+def test_unescapes_one_layer_only_in_strict_cell_text() -> None:
+    markdown = (
+        '<table data-note="&amp;lt;"><tr><td>&amp;lt;5</td><td>&amp;gt;2 &amp;amp; x &lt;</td></tr></table>'
+        "\noutside &amp;lt;"
+    )
+    result = transform(markdown)
+
+    assert result.status is RunStatus.SUCCESS
+    assert result.output_markdown == (
+        '<table data-note="&amp;lt;"><tr><td>&lt;5</td><td>&gt;2 &amp; x &lt;</td></tr></table>'
+        "\noutside &amp;lt;"
+    )
+    assert len(result.changes) == 3
+
+
+def test_multiple_tables_and_cells_report_source_order() -> None:
+    markdown = "<table><tr><td>&amp;gt;</td></tr></table> x <table><tr><th>&amp;lt;</th></tr></table>"
+    result = transform(markdown)
+
+    assert result.output_markdown == "<table><tr><td>&gt;</td></tr></table> x <table><tr><th>&lt;</th></tr></table>"
+    assert [change.span.start for change in result.changes] == sorted(change.span.start for change in result.changes)
+
+
+@pytest.mark.parametrize(
+    "markdown",
+    [
+        "outside &amp;lt;",
+        "<table><tr><td>&amp;lt;</tr></table>",
+        "<table><tr><td><em>&amp;lt;</em></td></tr></table>",
+        "<table><tr><td><table><tr><td>&amp;lt;</td></tr></table></td></tr></table>",
+        "<table><tbody><tr><td>&amp;lt;</td></tr></tbody></table>",
+    ],
+)
+def test_non_strict_or_outside_content_is_preserved(markdown: str) -> None:
+    assert transform(markdown).output_markdown == markdown
+
+
+def test_fenced_table_is_not_protected() -> None:
+    markdown = "```html\n<table><tr><td>&amp;lt;</td></tr></table>\n```"
+    assert transform(markdown).output_markdown == "```html\n<table><tr><td>&lt;</td></tr></table>\n```"
+
+
+def test_successful_output_is_idempotent() -> None:
+    pipeline = Pipeline([HtmlTableDoubleEscapeComponent()])
+    first = pipeline.transform("<table><tr><td>&amp;lt;</td></tr></table>")
+    assert first.output_markdown is not None
+    assert pipeline.transform(first.output_markdown).changes == ()