xiaoxia-saas/tests/unit/test_xml_parser.py

"""xml_parser CDATA 剥离单元测试。"""

from packages.application.viral_video.xml_parser import find_all, text_of

XML = """<script>
<copy_display_markdown><![CDATA[# 标题

这是第一段，含**加粗**和[链接](https://a.com)。

第二行，保留换行。]]></copy_display_markdown>
<voiceover>口播不带 CDATA，保持原样。</voiceover>
<visual><![CDATA[画面：产品特写，光线柔和]]></visual>
<action_details><![CDATA[未闭合标签里的 CDATA 也要剥离]]></action_details>
</script>"""


def test_text_of_strips_cdata_with_markdown_newlines():
    text = text_of(XML, "copy_display_markdown")
    assert not text.startswith("<![CDATA[")
    assert not text.endswith("]]>")
    assert "# 标题" in text
    assert "**加粗**" in text
    assert "[链接](https://a.com)" in text
    # markdown 换行被保留
    assert "\n\n第二行" in text


def test_plain_text_unchanged():
    assert text_of(XML, "voiceover") == "口播不带 CDATA，保持原样。"


def test_other_cdata_fields_stripped():
    assert text_of(XML, "visual") == "画面：产品特写，光线柔和"


def test_unclosed_tag_cdata_stripped():
    # action_details 没有闭合标签，走未闭合兜底分支
    node = find_all(XML, "action_details")[0]
    assert node["text"] == "未闭合标签里的 CDATA 也要剥离"


def test_no_cdata_returns_original():
    xml = "<copy_display_markdown>普通内容]]> 残留结尾</copy_display_markdown>"
    # 非完整 CDATA 包裹不应被误剥离
    assert text_of(xml, "copy_display_markdown") == "普通内容]]> 残留结尾"


def test_missing_tag_default():
    assert text_of(XML, "nope", default="缺省") == "缺省"