@adeu/core 1.14.0 → 1.15.1

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@adeu/core",
-  "version": "1.14.0",
+  "version": "1.15.1",
   "description": "",
   "main": "./dist/index.js",
   "types": "./dist/index.d.ts",

package/src/comment_dedup.test.ts

@@ -0,0 +1,269 @@
+// FILE: node/packages/core/src/comment_dedup.test.ts
+import { describe, it, expect } from "vitest";
+import { zipSync, strToU8 } from "fflate";
+import { extractTextFromBuffer } from "./ingest.js";
+import { DocumentObject } from "./docx/bridge.js";
+import { DocumentMapper } from "./mapper.js";
+
+// ---------------------------------------------------------------------------
+// Helpers
+// ---------------------------------------------------------------------------
+
+const NS_W =
+  'xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main"';
+const NS_W14 =
+  'xmlns:w14="http://schemas.microsoft.com/office/word/2010/wordml"';
+const NS_R =
+  'xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships"';
+
+/**
+ * Builds a minimal in-memory DOCX buffer from raw document body XML and a
+ * raw <w:comment> element list. We hand-assemble the OPC package to control
+ * exactly how runs are split inside comment ranges — python-docx-style
+ * builders would coalesce runs and hide the bug.
+ */
+function buildDocx(bodyXml: string, commentsListXml: string = ""): Buffer {
+  const documentXml = `<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<w:document ${NS_W} ${NS_W14} ${NS_R}>
+  <w:body>${bodyXml}</w:body>
+</w:document>`;
+
+  const commentsXml = `<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<w:comments ${NS_W} ${NS_W14}>${commentsListXml}</w:comments>`;
+
+  const hasComments = commentsListXml.length > 0;
+
+  const contentTypesXml = `<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<Types xmlns="http://schemas.openxmlformats.org/package/2006/content-types">
+  <Default Extension="rels" ContentType="application/vnd.openxmlformats-package.relationships+xml"/>
+  <Default Extension="xml" ContentType="application/xml"/>
+  <Override PartName="/word/document.xml" ContentType="application/vnd.openxmlformats-officedocument.wordprocessingml.document.main+xml"/>
+  ${hasComments ? '<Override PartName="/word/comments.xml" ContentType="application/vnd.openxmlformats-officedocument.wordprocessingml.comments+xml"/>' : ""}
+</Types>`;
+
+  const rootRelsXml = `<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">
+  <Relationship Id="rId1" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument" Target="word/document.xml"/>
+</Relationships>`;
+
+  const docRelsXml = `<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">
+  ${hasComments ? '<Relationship Id="rId1" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/comments" Target="comments.xml"/>' : ""}
+</Relationships>`;
+
+  const files: Record<string, Uint8Array> = {
+    "[Content_Types].xml": strToU8(contentTypesXml),
+    "_rels/.rels": strToU8(rootRelsXml),
+    "word/document.xml": strToU8(documentXml),
+    "word/_rels/document.xml.rels": strToU8(docRelsXml),
+  };
+  if (hasComments) {
+    files["word/comments.xml"] = strToU8(commentsXml);
+  }
+
+  return Buffer.from(zipSync(files));
+}
+
+function commentXml(
+  id: string,
+  author: string,
+  date: string,
+  text: string,
+): string {
+  return `<w:comment w:id="${id}" w:author="${author}" w:date="${date}" w:initials="X">
+    <w:p w14:paraId="${id.padStart(8, "0")}" w14:textId="77777777">
+      <w:r><w:t>${text}</w:t></w:r>
+    </w:p>
+  </w:comment>`;
+}
+
+// ---------------------------------------------------------------------------
+// Regression tests for the multi-run comment duplication bug
+// ---------------------------------------------------------------------------
+
+describe("Multi-run comment deduplication (BUG: comment text repeated per run)", () => {
+  it("emits a single comment block when one comment spans many runs (the reported bug)", async () => {
+    // 8 runs (Word naturally fragments around numbers, punctuation, formatting).
+    // Before the fix this produced 8 duplicate {>>...<<} blocks.
+    const body = `
+      <w:p>
+        <w:commentRangeStart w:id="1"/>
+        <w:r><w:t xml:space="preserve">Party A shall pay </w:t></w:r>
+        <w:r><w:t>100</w:t></w:r>
+        <w:r><w:t>%</w:t></w:r>
+        <w:r><w:t xml:space="preserve"> of the total</w:t></w:r>
+        <w:r><w:t xml:space="preserve"> amount</w:t></w:r>
+        <w:r><w:t xml:space="preserve"> on </w:t></w:r>
+        <w:r><w:t>time</w:t></w:r>
+        <w:r><w:t>.</w:t></w:r>
+        <w:commentRangeEnd w:id="1"/>
+        <w:r><w:rPr><w:rStyle w:val="CommentReference"/></w:rPr><w:commentReference w:id="1"/></w:r>
+      </w:p>`;
+
+    const comments = commentXml(
+      "1",
+      "Reviewer",
+      "2026-06-15T10:00:00Z",
+      "Risk note: please confirm the payment percentage and timing.",
+    );
+
+    const buf = buildDocx(body, comments);
+    const md = await extractTextFromBuffer(buf, false);
+
+    // The full highlighted range must be a single contiguous {==...==} block.
+    expect(md).toContain(
+      "{==Party A shall pay 100% of the total amount on time.==}",
+    );
+
+    // Exactly one comment meta block.
+    expect((md.match(/\[Com:1\]/g) || []).length).toBe(1);
+    expect((md.match(/\{==/g) || []).length).toBe(1);
+    expect((md.match(/\{>>/g) || []).length).toBe(1);
+
+    // And the comment payload appears exactly once.
+    const payload =
+      "Risk note: please confirm the payment percentage and timing.";
+    expect(md.split(payload).length - 1).toBe(1);
+  });
+
+  it("keeps two separate comments on adjacent runs independent (must NOT merge)", async () => {
+    // Comment 1 wraps "First part", comment 2 wraps "second part".
+    // These are distinct ranges with no overlap and must stay distinct in output.
+    const body = `
+      <w:p>
+        <w:commentRangeStart w:id="1"/>
+        <w:r><w:t xml:space="preserve">First </w:t></w:r>
+        <w:r><w:t>part</w:t></w:r>
+        <w:commentRangeEnd w:id="1"/>
+        <w:r><w:rPr><w:rStyle w:val="CommentReference"/></w:rPr><w:commentReference w:id="1"/></w:r>
+        <w:r><w:t xml:space="preserve"> middle </w:t></w:r>
+        <w:commentRangeStart w:id="2"/>
+        <w:r><w:t xml:space="preserve">second </w:t></w:r>
+        <w:r><w:t>part</w:t></w:r>
+        <w:commentRangeEnd w:id="2"/>
+        <w:r><w:rPr><w:rStyle w:val="CommentReference"/></w:rPr><w:commentReference w:id="2"/></w:r>
+      </w:p>`;
+
+    const comments =
+      commentXml("1", "A", "2026-06-15T10:00:00Z", "First annotation") +
+      commentXml("2", "B", "2026-06-15T11:00:00Z", "Second annotation");
+
+    const buf = buildDocx(body, comments);
+    const md = await extractTextFromBuffer(buf, false);
+
+    // Each comment appears exactly once, and they are distinct blocks.
+    expect((md.match(/\[Com:1\]/g) || []).length).toBe(1);
+    expect((md.match(/\[Com:2\]/g) || []).length).toBe(1);
+    expect((md.match(/\{>>/g) || []).length).toBe(2);
+    expect((md.match(/\{==/g) || []).length).toBe(2);
+
+    // Highlighted ranges must contain the right text on each side of the middle.
+    expect(md).toContain("{==First part==}");
+    expect(md).toContain("{==second part==}");
+    // The un-commented "middle" text must NOT be inside a highlight wrapper.
+    expect(md).toContain(" middle ");
+  });
+
+  it("merges overlapping comments (same range, two comment IDs) into a single meta block", async () => {
+    // Both comment 1 and comment 2 span the same exact text. Per the existing
+    // state-machine semantics, both IDs are active simultaneously, so the meta
+    // block should list both — but only once.
+    const body = `
+      <w:p>
+        <w:commentRangeStart w:id="1"/>
+        <w:commentRangeStart w:id="2"/>
+        <w:r><w:t xml:space="preserve">Shared </w:t></w:r>
+        <w:r><w:t>highlighted</w:t></w:r>
+        <w:r><w:t xml:space="preserve"> text</w:t></w:r>
+        <w:commentRangeEnd w:id="1"/>
+        <w:commentRangeEnd w:id="2"/>
+        <w:r><w:rPr><w:rStyle w:val="CommentReference"/></w:rPr><w:commentReference w:id="1"/></w:r>
+        <w:r><w:rPr><w:rStyle w:val="CommentReference"/></w:rPr><w:commentReference w:id="2"/></w:r>
+      </w:p>`;
+
+    const comments =
+      commentXml("1", "A", "2026-06-15T10:00:00Z", "Comment one") +
+      commentXml("2", "B", "2026-06-15T11:00:00Z", "Comment two");
+
+    const buf = buildDocx(body, comments);
+    const md = await extractTextFromBuffer(buf, false);
+
+    // Single highlight wrapper covering the full range.
+    expect(md).toContain("{==Shared highlighted text==}");
+    expect((md.match(/\{==/g) || []).length).toBe(1);
+
+    // Single meta block that mentions both comments.
+    expect((md.match(/\{>>/g) || []).length).toBe(1);
+    expect((md.match(/\[Com:1\]/g) || []).length).toBe(1);
+    expect((md.match(/\[Com:2\]/g) || []).length).toBe(1);
+  });
+
+  it("handles a tracked insertion nested inside a comment range without duplication", async () => {
+    // Comment 1 spans 3 runs; one of those runs is wrapped in <w:ins>.
+    // The redline must produce its own {++...++} marker but the comment
+    // payload must still appear only once.
+    const body = `
+      <w:p>
+        <w:commentRangeStart w:id="1"/>
+        <w:r><w:t xml:space="preserve">Plain </w:t></w:r>
+        <w:ins w:id="100" w:author="Editor" w:date="2026-06-15T12:00:00Z">
+          <w:r><w:t>inserted</w:t></w:r>
+        </w:ins>
+        <w:r><w:t xml:space="preserve"> tail</w:t></w:r>
+        <w:commentRangeEnd w:id="1"/>
+        <w:r><w:rPr><w:rStyle w:val="CommentReference"/></w:rPr><w:commentReference w:id="1"/></w:r>
+      </w:p>`;
+
+    const comments = commentXml(
+      "1",
+      "Reviewer",
+      "2026-06-15T10:00:00Z",
+      "Mixed comment over a redline",
+    );
+
+    const buf = buildDocx(body, comments);
+    const md = await extractTextFromBuffer(buf, false);
+
+    // The insertion is still surfaced as a redline.
+    expect(md).toContain("{++inserted++}");
+
+    // The comment payload appears exactly once.
+    expect((md.match(/\[Com:1\]/g) || []).length).toBe(1);
+    expect((md.match(/Mixed comment over a redline/g) || []).length).toBe(1);
+
+    // There should be the comment highlight pieces around the plain text,
+    // plus the {++ wrapper for the insertion — but the comment text itself
+    // must not be repeated.
+  });
+
+  it("DocumentMapper.full_text mirrors the projected output (parity with ingest)", async () => {
+    // If the mapper produces a different number of comment blocks than the
+    // projection layer, the engine's find/replace anchoring breaks. This guards
+    // mapper.ts against regressing independently of ingest.ts.
+    const body = `
+      <w:p>
+        <w:commentRangeStart w:id="1"/>
+        <w:r><w:t xml:space="preserve">A </w:t></w:r>
+        <w:r><w:t>B</w:t></w:r>
+        <w:r><w:t xml:space="preserve"> C</w:t></w:r>
+        <w:commentRangeEnd w:id="1"/>
+        <w:r><w:rPr><w:rStyle w:val="CommentReference"/></w:rPr><w:commentReference w:id="1"/></w:r>
+      </w:p>`;
+
+    const comments = commentXml(
+      "1",
+      "X",
+      "2026-06-15T10:00:00Z",
+      "Mapper parity check",
+    );
+
+    const buf = buildDocx(body, comments);
+    const doc = await DocumentObject.load(buf);
+    const mapper = new DocumentMapper(doc);
+
+    expect((mapper.full_text.match(/\[Com:1\]/g) || []).length).toBe(1);
+    expect((mapper.full_text.match(/\{>>/g) || []).length).toBe(1);
+    expect((mapper.full_text.match(/\{==/g) || []).length).toBe(1);
+    expect(mapper.full_text).toContain("{==A B C==}");
+  });
+});