feat: 完成 DOCX 通用结构映射与严格收口

2026-07-31 15:19:34 +08:00
parent 6e40374200
commit f3847f3e4b
31 changed files with 3175 additions and 64 deletions
@@ -0,0 +1,147 @@
+import { describe, expect, it } from "vitest";
+import {
+  WORD_NAMESPACE,
+  appendElement,
+  parseXmlPart,
+  serializeXmlPart,
+  validateWordprocessingElementOrder
+} from "../src/ooxml.js";
+
+const encoder = new TextEncoder();
+const decoder = new TextDecoder();
+
+function paragraphPropertiesDocument(children = "") {
+  return parseXmlPart(
+    encoder.encode(
+      `<w:document xmlns:w="${WORD_NAMESPACE}">` +
+        `<w:body><w:p><w:pPr>${children}</w:pPr></w:p></w:body>` +
+        "</w:document>"
+    ),
+    "word/document.xml"
+  );
+}
+
+describe("OOXML 子节点顺序", () => {
+  it("按 WordprocessingML 顺序插入新增属性", () => {
+    const document = paragraphPropertiesDocument(
+      "<w:pStyle w:val=\"Normal\"/><w:jc w:val=\"left\"/>"
+    );
+    const properties = document.getElementsByTagNameNS(
+      WORD_NAMESPACE,
+      "pPr"
+    )[0]!;
+
+    appendElement(properties, WORD_NAMESPACE, "w:pBdr");
+    appendElement(properties, WORD_NAMESPACE, "w:keepNext");
+    appendElement(properties, WORD_NAMESPACE, "w:spacing");
+
+    const xml = decoder.decode(serializeXmlPart(document));
+    expect(xml.indexOf("<w:keepNext")).toBeLessThan(
+      xml.indexOf("<w:pBdr")
+    );
+    expect(xml.indexOf("<w:pBdr")).toBeLessThan(
+      xml.indexOf("<w:spacing")
+    );
+    expect(xml.indexOf("<w:spacing")).toBeLessThan(
+      xml.indexOf("<w:jc")
+    );
+    expect(() =>
+      validateWordprocessingElementOrder(
+        document,
+        "word/document.xml"
+      )
+    ).not.toThrow();
+  });
+
+  it("在序列化时规范化已有乱序属性", () => {
+    const document = paragraphPropertiesDocument(
+      "<w:pStyle w:val=\"Normal\"/>" +
+        "<w:pBdr/><w:keepLines/><w:keepNext/>"
+    );
+
+    const serialized = serializeXmlPart(document);
+    const reparsed = parseXmlPart(
+      serialized,
+      "word/document.xml"
+    );
+    expect(() =>
+      validateWordprocessingElementOrder(
+        reparsed,
+        "word/document.xml"
+      )
+    ).not.toThrow();
+
+    const xml = decoder.decode(serialized);
+    expect(xml.indexOf("<w:keepNext")).toBeLessThan(
+      xml.indexOf("<w:keepLines")
+    );
+    expect(xml.indexOf("<w:keepLines")).toBeLessThan(
+      xml.indexOf("<w:pBdr")
+    );
+  });
+
+  it("拒绝未经规范化的乱序属性", () => {
+    const document = paragraphPropertiesDocument(
+      "<w:pStyle w:val=\"Normal\"/>" +
+        "<w:jc w:val=\"left\"/><w:spacing/>"
+    );
+
+    expect(() =>
+      validateWordprocessingElementOrder(
+        document,
+        "word/document.xml"
+      )
+    ).toThrow(
+      "word/document.xml 的 w:pPr 子节点顺序无效"
+    );
+  });
+
+  it("规范化表格、单元格和边框顺序", () => {
+    const document = parseXmlPart(
+      encoder.encode(
+        `<w:document xmlns:w="${WORD_NAMESPACE}"><w:body>` +
+          "<w:tbl><w:tblPr>" +
+          "<w:tblLook/><w:tblW/><w:tblLayout/>" +
+          "</w:tblPr><w:tblGrid/><w:tr><w:tc><w:tcPr>" +
+          "<w:gridSpan/><w:shd/><w:tcW/>" +
+          "</w:tcPr><w:p/></w:tc></w:tr></w:tbl>" +
+          "<w:sectPr><w:docGrid/><w:vAlign/></w:sectPr>" +
+          "</w:body></w:document>"
+      ),
+      "word/document.xml"
+    );
+
+    const serialized = serializeXmlPart(document);
+    const reparsed = parseXmlPart(
+      serialized,
+      "word/document.xml"
+    );
+    expect(() =>
+      validateWordprocessingElementOrder(
+        reparsed,
+        "word/document.xml"
+      )
+    ).not.toThrow();
+  });
+
+  it("规范化表格条件样式的属性顺序", () => {
+    const document = parseXmlPart(
+      encoder.encode(
+        `<w:styles xmlns:w="${WORD_NAMESPACE}">` +
+          '<w:style w:type="table" w:styleId="Table">' +
+          '<w:tblStylePr w:type="firstRow">' +
+          "<w:tcPr/><w:rPr/><w:pPr/>" +
+          "</w:tblStylePr></w:style></w:styles>"
+      ),
+      "word/styles.xml"
+    );
+
+    const xml = decoder.decode(serializeXmlPart(document));
+    expect(xml.indexOf("<w:pPr")).toBeLessThan(
+      xml.indexOf("<w:rPr")
+    );
+    expect(xml.indexOf("<w:rPr")).toBeLessThan(
+      xml.indexOf("<w:tcPr")
+    );
+  });
+});