feat: 完成 DOCX R4 元素级视觉门禁

建立 Chromium、Word 与 WPS 的可编辑内容、段落几何、页眉页脚、封面、媒体和逐页视觉比较链路。支持封面独立分节、正文页码重启、主题页边距与横向纸张，并将自然分页差异降为诊断项。修正代码块内边距和折叠表格单元格边框映射，14 套主题生产矩阵与六组布局视觉矩阵通过。
2026-08-02 03:27:11 +08:00
parent f9f5fccfc9
commit 58f87cc19f
100 changed files with 10409 additions and 386 deletions
@@ -1,9 +1,12 @@
 import { describe, expect, it } from "vitest";

 import {
+  calculateContentSimilarity,
  comparePdfSnapshotsBasic,
  createPdfDocumentSnapshot,
  type PdfDocumentSnapshot,
+  type PdfTextLineSnapshot,
+  type VisualPageSemanticExpectation,
 } from "../src/index.js";
 import { createMinimalPdf } from "./pdf-fixture.js";

@@ -19,7 +22,105 @@ async function snapshot(
  });
 }

+function textLine(
+  text: string,
+  y: number,
+  role: PdfTextLineSnapshot["role"] = "content",
+): PdfTextLineSnapshot {
+  return {
+    text,
+    normalizedText: text,
+    bounds: { x: 72, y, width: 200, height: 12 },
+    baselineY: y + 10,
+    role,
+    items: [],
+  };
+}
+
+function withLines(
+  source: PdfDocumentSnapshot,
+  lines: PdfTextLineSnapshot[],
+): PdfDocumentSnapshot {
+  const page = source.pages[0];
+  if (!page) {
+    throw new Error("测试快照缺少页面");
+  }
+  return {
+    ...source,
+    contentText: lines.map((line) => line.normalizedText).join("\n"),
+    pages: [{ ...page, lines }],
+  };
+}
+
+const pageSemantics: VisualPageSemanticExpectation[] = [
+  {
+    physicalPageNumber: 1,
+    kind: "body-first",
+    logicalPageNumber: 1,
+    logicalPageCount: 1,
+    headerVisible: true,
+    headerSlots: [{ alignment: "left", text: "页眉左栏" }],
+    footerVisible: true,
+    footerAlignment: "center",
+    pageNumberText: "1 / 1",
+  },
+];
+
 describe("PDF 基础视觉门禁", () => {
+  it("将 PDF 字体映射产生的等价部首字形规范为正文汉字", () => {
+    expect(
+      calculateContentSimilarity(
+        "示例市人⺠政府办公室",
+        "示例市人民政府办公室",
+      ),
+    ).toBe(1);
+    expect(calculateContentSimilarity("项目⻔户", "项目门户")).toBe(1);
+  });
+
+  it("以 DOCX 可编辑正文为语义基准并容许 Chromium 媒体内部文本", async () => {
+    const baseline = withLines(await snapshot("baseline"), [
+      textLine("页眉左栏", 10),
+      textLine("正文甲", 100),
+      textLine("媒体内部标签", 150),
+      textLine("正文乙", 200),
+      textLine("1 / 1", 820, "page-number"),
+    ]);
+    const candidate = withLines(await snapshot("candidate"), [
+      textLine("页眉左栏", 10),
+      textLine("正文甲☒", 100),
+      textLine("正文乙", 200),
+      textLine("1 / 1", 820, "page-number"),
+    ]);
+    const result = comparePdfSnapshotsBasic(baseline, candidate, {}, {
+      expectedEditableText: "正文甲正文乙",
+      pageSemantics,
+    });
+    expect(result.status).toBe("passed");
+    expect(result.baselineEditableCoverage).toBe(1);
+    expect(result.candidateEditableSimilarity).toBe(1);
+    expect(result.candidateEditableExact).toBe(true);
+  });
+
+  it("候选缺少任一可编辑正文字符时严格失败", async () => {
+    const baseline = withLines(await snapshot("baseline"), [
+      textLine("正文甲正文乙", 100),
+    ]);
+    const candidate = withLines(await snapshot("candidate"), [
+      textLine("正文甲正文", 100),
+    ]);
+    const result = comparePdfSnapshotsBasic(baseline, candidate, {}, {
+      expectedEditableText: "正文甲正文乙",
+    });
+    expect(result.status).toBe("failed");
+    expect(result.baselineEditableCoverage).toBe(1);
+    expect(result.candidateEditableExact).toBe(false);
+    expect(result.issues).toEqual(
+      expect.arrayContaining([
+        expect.objectContaining({ code: "CONTENT_MISMATCH" }),
+      ]),
+    );
+  });
+
  it("通过纸张和正文一致的文档", async () => {
    const baseline = await snapshot("baseline");
    const candidate = await snapshot("candidate");