@usejunior/docx-core 0.9.1 → 0.11.0

package/dist/integration/generation-probes.js

@@ -0,0 +1,84 @@
+/**
+ * LibreOffice probes for generated full-package documents.
+ *
+ * Deliberately separate from runLibreOfficeOracle: the oracle's contract is
+ * "main XML string per accept/reject/identity job over a bare document.xml or
+ * .odt", driven by an injected Basic macro because the resolve-all commands
+ * are dispatch-only. These probes have a different contract — a complete
+ * generated package goes in, and the *saved package* (or rendered PDF) comes
+ * back out — and need no macro: `--convert-to` performs exactly the
+ * load→save (or load→render) cycle that exposes recovery-dialog failures,
+ * because a package LibreOffice cannot load cleanly produces no output.
+ *
+ * Local-only, like the oracle: callers skip when `resolveSoffice()` is null
+ * (CI does not install LibreOffice; the structural checks are the
+ * CI-enforceable proxy).
+ */
+import { execFile } from 'node:child_process';
+import { existsSync, mkdirSync, mkdtempSync, readFileSync, rmSync, writeFileSync } from 'node:fs';
+import os from 'node:os';
+import path from 'node:path';
+import { pathToFileURL } from 'node:url';
+import { promisify } from 'node:util';
+import { resolveSoffice } from './libreoffice-oracle.js';
+const execFileAsync = promisify(execFile);
+async function runConvert(input, convertTo, outExt, soffice) {
+    const work = mkdtempSync(path.join(os.tmpdir(), 'sdx-gen-probe-'));
+    try {
+        const inDir = path.join(work, 'in');
+        const outDir = path.join(work, 'out');
+        const profile = path.join(work, 'profile');
+        mkdirSync(inDir, { recursive: true });
+        mkdirSync(outDir, { recursive: true });
+        mkdirSync(profile, { recursive: true });
+        const inPath = path.join(inDir, 'probe.docx');
+        writeFileSync(inPath, new Uint8Array(input));
+        const args = [
+            '--headless',
+            '--norestore',
+            '--nologo',
+            `-env:UserInstallation=${pathToFileURL(profile).href}`,
+            '--convert-to',
+            convertTo,
+            '--outdir',
+            outDir,
+            inPath,
+        ];
+        let diag = '';
+        try {
+            const r = await execFileAsync(soffice, args, { timeout: 45_000, killSignal: 'SIGKILL', maxBuffer: 8 * 1024 * 1024 });
+            diag = String(r.stderr ?? '') || String(r.stdout ?? '');
+        }
+        catch (err) {
+            const e = err;
+            diag = String(e.stderr ?? e.stdout ?? e.message ?? '');
+        }
+        const outPath = path.join(outDir, `probe.${outExt}`);
+        if (!existsSync(outPath)) {
+            throw new Error(`LibreOffice could not convert the generated package to ${outExt} — likely a load failure ` +
+                `(the recovery-dialog class of bug).\nsoffice output:\n${diag.trim() || '(no output)'}`);
+        }
+        return readFileSync(outPath);
+    }
+    finally {
+        rmSync(work, { recursive: true, force: true });
+    }
+}
+/**
+ * Load→save a complete generated package through LibreOffice's DOCX filter.
+ * Throws when LibreOffice cannot load the package (no output is produced).
+ */
+export async function probeDocxIdentity(generated, soffice = resolveSoffice()) {
+    if (!soffice)
+        throw new Error('probeDocxIdentity: no soffice binary (call resolveSoffice() and skip)');
+    const savedPackage = await runConvert(generated, 'docx:MS Word 2007 XML', 'docx', soffice);
+    return { savedPackage };
+}
+/** Render a generated package to PDF headlessly — the headless-renderer probe. */
+export async function probeDocxToPdf(generated, soffice = resolveSoffice()) {
+    if (!soffice)
+        throw new Error('probeDocxToPdf: no soffice binary (call resolveSoffice() and skip)');
+    const pdf = await runConvert(generated, 'pdf', 'pdf', soffice);
+    return { pdf };
+}
+//# sourceMappingURL=generation-probes.js.map

package/dist/integration/generation-probes.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"generation-probes.js","sourceRoot":"","sources":["../../src/integration/generation-probes.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AACH,OAAO,EAAE,QAAQ,EAAE,MAAM,oBAAoB,CAAC;AAC9C,OAAO,EAAE,UAAU,EAAE,SAAS,EAAE,WAAW,EAAE,YAAY,EAAE,MAAM,EAAE,aAAa,EAAE,MAAM,SAAS,CAAC;AAClG,OAAO,EAAE,MAAM,SAAS,CAAC;AACzB,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,aAAa,EAAE,MAAM,UAAU,CAAC;AACzC,OAAO,EAAE,SAAS,EAAE,MAAM,WAAW,CAAC;AACtC,OAAO,EAAE,cAAc,EAAE,MAAM,yBAAyB,CAAC;AAEzD,MAAM,aAAa,GAAG,SAAS,CAAC,QAAQ,CAAC,CAAC;AAW1C,KAAK,UAAU,UAAU,CAAC,KAAa,EAAE,SAAiB,EAAE,MAAc,EAAE,OAAe;IACzF,MAAM,IAAI,GAAG,WAAW,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,MAAM,EAAE,EAAE,gBAAgB,CAAC,CAAC,CAAC;IACnE,IAAI,CAAC;QACH,MAAM,KAAK,GAAG,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC;QACpC,MAAM,MAAM,GAAG,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,KAAK,CAAC,CAAC;QACtC,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,SAAS,CAAC,CAAC;QAC3C,SAAS,CAAC,KAAK,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;QACtC,SAAS,CAAC,MAAM,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;QACvC,SAAS,CAAC,OAAO,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;QACxC,MAAM,MAAM,GAAG,IAAI,CAAC,IAAI,CAAC,KAAK,EAAE,YAAY,CAAC,CAAC;QAC9C,aAAa,CAAC,MAAM,EAAE,IAAI,UAAU,CAAC,KAAK,CAAC,CAAC,CAAC;QAE7C,MAAM,IAAI,GAAG;YACX,YAAY;YACZ,aAAa;YACb,UAAU;YACV,yBAAyB,aAAa,CAAC,OAAO,CAAC,CAAC,IAAI,EAAE;YACtD,cAAc;YACd,SAAS;YACT,UAAU;YACV,MAAM;YACN,MAAM;SACP,CAAC;QACF,IAAI,IAAI,GAAG,EAAE,CAAC;QACd,IAAI,CAAC;YACH,MAAM,CAAC,GAAG,MAAM,aAAa,CAAC,OAAO,EAAE,IAAI,EAAE,EAAE,OAAO,EAAE,MAAM,EAAE,UAAU,EAAE,SAAS,EAAE,SAAS,EAAE,CAAC,GAAG,IAAI,GAAG,IAAI,EAAE,CAAC,CAAC;YACrH,IAAI,GAAG,MAAM,CAAC,CAAC,CAAC,MAAM,IAAI,EAAE,CAAC,IAAI,MAAM,CAAC,CAAC,CAAC,MAAM,IAAI,EAAE,CAAC,CAAC;QAC1D,CAAC;QAAC,OAAO,GAAG,EAAE,CAAC;YACb,MAAM,CAAC,GAAG,GAA+D,CAAC;YAC1E,IAAI,GAAG,MAAM,CAAC,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,OAAO,IAAI,EAAE,CAAC,CAAC;QACzD,CAAC;QAED,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,SAAS,MAAM,EAAE,CAAC,CAAC;QACrD,IAAI,CAAC,UAAU,CAAC,OAAO,CAAC,EAAE,CAAC;YACzB,MAAM,IAAI,KAAK,CACb,0DAA0D,MAAM,2BAA2B;gBACzF,yDAAyD,IAAI,CAAC,IAAI,EAAE,IAAI,aAAa,EAAE,CAC1F,CAAC;QACJ,CAAC;QACD,OAAO,YAAY,CAAC,OAAO,CAAC,CAAC;IAC/B,CAAC;YAAS,CAAC;QACT,MAAM,CAAC,IAAI,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC;IACjD,CAAC;AACH,CAAC;AAED;;;GAGG;AACH,MAAM,CAAC,KAAK,UAAU,iBAAiB,CACrC,SAAiB,EACjB,UAAyB,cAAc,EAAE;IAEzC,IAAI,CAAC,OAAO;QAAE,MAAM,IAAI,KAAK,CAAC,uEAAuE,CAAC,CAAC;IACvG,MAAM,YAAY,GAAG,MAAM,UAAU,CAAC,SAAS,EAAE,uBAAuB,EAAE,MAAM,EAAE,OAAO,CAAC,CAAC;IAC3F,OAAO,EAAE,YAAY,EAAE,CAAC;AAC1B,CAAC;AAED,kFAAkF;AAClF,MAAM,CAAC,KAAK,UAAU,cAAc,CAClC,SAAiB,EACjB,UAAyB,cAAc,EAAE;IAEzC,IAAI,CAAC,OAAO;QAAE,MAAM,IAAI,KAAK,CAAC,oEAAoE,CAAC,CAAC;IACpG,MAAM,GAAG,GAAG,MAAM,UAAU,CAAC,SAAS,EAAE,KAAK,EAAE,KAAK,EAAE,OAAO,CAAC,CAAC;IAC/D,OAAO,EAAE,GAAG,EAAE,CAAC;AACjB,CAAC"}

package/dist/integration/libreoffice-oracle.d.ts

@@ -0,0 +1,49 @@
+/** Resolve a LibreOffice binary, or null if none is available (callers skip the oracle). */
+export declare function resolveSoffice(): string | null;
+/** Pack a bare `word/document.xml` string into a minimal valid .docx package. */
+export declare function packMinimalDocx(documentXml: string): Promise<Buffer>;
+/** Read `word/document.xml` back out of a .docx package. */
+export declare function extractDocumentXml(docx: Buffer): Promise<string>;
+/**
+ * `accept` / `reject` dispatch the corresponding resolve-all command before saving — the oracle's
+ * normal voting mode. `identity` loads and saves WITHOUT any dispatch, so unresolved tracked
+ * changes flow through LibreOffice's import/export: it exists to characterize the oracle's
+ * trust boundary (LibreOffice's save round-trip mangles some unresolved revision shapes — see
+ * libreoffice-oracle-trust-boundary.test.ts), NOT to vote on engine behavior.
+ *
+ * DOCX jobs carry a bare `word/document.xml` (packed into a minimal package and read back out);
+ * ODT jobs carry a complete `.odt` package buffer (ODF packaging — mimetype-first STORED — is
+ * the caller's concern) and return its post-op `content.xml`.
+ *
+ * CONVERSION jobs (`docx` + `saveAs: 'odt'`) carry a complete `.docx` package buffer and save
+ * through LibreOffice's `writer8` filter, returning the converted `content.xml` — the reference
+ * path for differential-testing odf-core's native DOCX→ODT converter (issue #331).
+ */
+type OracleOp = 'accept' | 'reject' | 'identity';
+export type OracleJob = {
+    op: OracleOp;
+    documentXml: string;
+} | {
+    op: OracleOp;
+    odt: Buffer;
+} | {
+    op: OracleOp;
+    docx: Buffer;
+    saveAs: 'odt';
+};
+/**
+ * Run LibreOffice over a batch of jobs in ONE headless launch and return each job's resulting
+ * main XML part — `word/document.xml` for DOCX jobs, `content.xml` for ODT jobs. Throws if the
+ * binary is missing or the macro did not run.
+ */
+export declare function runLibreOfficeOracle(jobs: OracleJob[], soffice?: string | null): Promise<string[]>;
+/**
+ * Structural projection of a `word/document.xml`: one entry per top-level body paragraph,
+ * recording whether it carries visible text (a non-whitespace `w:t` descendant). This captures
+ * the paragraph-collapse claim the oracle is authoritative for — how many paragraphs survive and
+ * which collapsed to empty — without depending on revision-markup or formatting details (which
+ * LibreOffice rewrites). `[]`-length is the paragraph count.
+ */
+export declare function paragraphShape(documentXml: string): boolean[];
+export {};
+//# sourceMappingURL=libreoffice-oracle.d.ts.map

package/dist/integration/libreoffice-oracle.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"libreoffice-oracle.d.ts","sourceRoot":"","sources":["../../src/integration/libreoffice-oracle.ts"],"names":[],"mappings":"AAoEA,4FAA4F;AAC5F,wBAAgB,cAAc,IAAI,MAAM,GAAG,IAAI,CAU9C;AAeD,iFAAiF;AACjF,wBAAsB,eAAe,CAAC,WAAW,EAAE,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC,CAO1E;AAED,4DAA4D;AAC5D,wBAAsB,kBAAkB,CAAC,IAAI,EAAE,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC,CAItE;AA+DD;;;;;;;;;;;;;;GAcG;AACH,KAAK,QAAQ,GAAG,QAAQ,GAAG,QAAQ,GAAG,UAAU,CAAC;AACjD,MAAM,MAAM,SAAS,GACjB;IAAE,EAAE,EAAE,QAAQ,CAAC;IAAC,WAAW,EAAE,MAAM,CAAA;CAAE,GACrC;IAAE,EAAE,EAAE,QAAQ,CAAC;IAAC,GAAG,EAAE,MAAM,CAAA;CAAE,GAC7B;IAAE,EAAE,EAAE,QAAQ,CAAC;IAAC,IAAI,EAAE,MAAM,CAAC;IAAC,MAAM,EAAE,KAAK,CAAA;CAAE,CAAC;AAUlD;;;;GAIG;AACH,wBAAsB,oBAAoB,CAAC,IAAI,EAAE,SAAS,EAAE,EAAE,OAAO,gBAAmB,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,CA4F3G;AAED;;;;;;GAMG;AACH,wBAAgB,cAAc,CAAC,WAAW,EAAE,MAAM,GAAG,OAAO,EAAE,CAkB7D"}

package/dist/integration/libreoffice-oracle.js

@@ -0,0 +1,290 @@
+/**
+ * LibreOffice accept/reject oracle — a committed, reproducible reference voter.
+ *
+ * Drives LibreOffice headless as a track-changes accept/reject implementation so the
+ * production engine's paragraph-collapse behavior (the G3/G4/G5 differential cases) can be
+ * validated against a real word processor, not just Lean↔TS self-consistency. LibreOffice is
+ * the native engine for the .uno:AcceptAllTrackedChanges / .uno:RejectAllTrackedChanges
+ * dispatches, so its paragraph-structure output is authoritative ground truth for the
+ * mark-based rule (an untracked paragraph mark is kept on accept/reject; a PPR-INS/PPR-DEL mark
+ * is dropped).
+ *
+ * Mechanism (macOS-portable; also works on Linux): pyuno from a terminal is blocked on macOS by
+ * Launch Constraints, so this injects a Basic macro into a throwaway user profile and invokes it
+ * with a bare `macro:///` URL. The order matters — a fresh profile regenerates the Standard Basic
+ * library on first launch, clobbering any hand-placed Module1.xba — so we (1) init the profile via
+ * a throwaway convert, (2) THEN overwrite Module1.xba, (3) THEN run the macro. See
+ * reference_libreoffice_macos_oracle.
+ *
+ * This module is gated by callers: when `resolveSoffice()` returns null the oracle is skipped.
+ * CI does not install LibreOffice, so the oracle voter is a local developer check.
+ */
+import { execFile } from 'node:child_process';
+import { existsSync, mkdirSync, mkdtempSync, readFileSync, rmSync, writeFileSync } from 'node:fs';
+import os from 'node:os';
+import path from 'node:path';
+import { pathToFileURL } from 'node:url';
+import { promisify } from 'node:util';
+import { createZipBuffer, readZipText } from '../primitives/zip.js';
+import { parseXml } from '../primitives/xml.js';
+const execFileAsync = promisify(execFile);
+const W_NS = 'http://schemas.openxmlformats.org/wordprocessingml/2006/main';
+const sleep = (ms) => new Promise((resolve) => setTimeout(resolve, ms));
+/** Run soffice, capturing stdout/stderr; never throws (the macro terminates the desktop, so a
+ *  nonzero exit is expected). */
+async function runSoffice(soffice, args, timeout) {
+    try {
+        const r = await execFileAsync(soffice, args, { timeout, killSignal: 'SIGKILL', maxBuffer: 8 * 1024 * 1024 });
+        return { stdout: String(r.stdout ?? ''), stderr: String(r.stderr ?? '') };
+    }
+    catch (err) {
+        const e = err;
+        return { stdout: String(e.stdout ?? ''), stderr: String(e.stderr ?? e.message ?? '') };
+    }
+}
+/**
+ * LibreOffice's single-instance model forwards a new `soffice` invocation to an existing instance
+ * that shares the same UserInstallation — so the init-convert process MUST be fully gone before the
+ * macro launch, or the `macro:///` command is delivered to the dying init instance (which has no
+ * Module1 yet) and never runs. The instance lock is `<profile>/.lock`. LibreOffice often leaves a
+ * STALE lock after `--convert-to` exits, so we wait a short bounded window for it to self-clear and
+ * then remove it unconditionally; the macro launch's own retry backstops a slow init exit.
+ */
+async function settleProfile(profile, timeoutMs = 1_500) {
+    const lock = path.join(profile, '.lock');
+    const deadline = Date.now() + timeoutMs;
+    while (existsSync(lock) && Date.now() < deadline)
+        await sleep(150);
+    if (existsSync(lock)) {
+        try {
+            rmSync(lock, { force: true });
+        }
+        catch { /* best effort — fresh launch will relock */ }
+    }
+}
+/** Resolve a LibreOffice binary, or null if none is available (callers skip the oracle). */
+export function resolveSoffice() {
+    const candidates = [
+        process.env.SAFE_DOCX_SOFFICE_BIN,
+        process.env.ODF_SOFFICE_BIN,
+        '/opt/homebrew/bin/soffice',
+        '/usr/bin/soffice',
+        '/usr/local/bin/soffice',
+        '/Applications/LibreOffice.app/Contents/MacOS/soffice',
+    ].filter(Boolean);
+    return candidates.find((c) => existsSync(c)) ?? null;
+}
+const CONTENT_TYPES = `<?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"/>
+</Types>`;
+const RELS = `<?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 DOC_RELS = `<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships"/>`;
+/** Pack a bare `word/document.xml` string into a minimal valid .docx package. */
+export async function packMinimalDocx(documentXml) {
+    return createZipBuffer({
+        '[Content_Types].xml': CONTENT_TYPES,
+        '_rels/.rels': RELS,
+        'word/_rels/document.xml.rels': DOC_RELS,
+        'word/document.xml': documentXml,
+    });
+}
+/** Read `word/document.xml` back out of a .docx package. */
+export async function extractDocumentXml(docx) {
+    const xml = await readZipText(docx, 'word/document.xml');
+    if (xml == null)
+        throw new Error('word/document.xml not found in oracle output');
+    return xml;
+}
+const SCRIPT_XLC = `<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE library:libraries PUBLIC "-//OpenOffice.org//DTD OfficeDocument 1.0//EN" "libraries.dtd">
+<library:libraries xmlns:library="http://openoffice.org/2000/library" xmlns:xlink="http://www.w3.org/1999/xlink">
+ <library:library library:name="Standard" xlink:href="$(USER)/basic/Standard/script.xlb/" xlink:type="simple" library:link="false"/>
+</library:libraries>`;
+const SCRIPT_XLB = `<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE library:library PUBLIC "-//OpenOffice.org//DTD OfficeDocument 1.0//EN" "library.dtd">
+<library:library xmlns:library="http://openoffice.org/2000/library" library:name="Standard" library:readonly="false" library:passwordprotected="false">
+ <library:element library:name="Module1"/>
+</library:library>`;
+const REGMOD = `<?xml version="1.0" encoding="UTF-8"?>
+<oor:items xmlns:oor="http://openoffice.org/2001/registry" xmlns:xs="http://www.w3.org/2001/XMLSchema" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
+ <item oor:path="/org.openoffice.Office.Common/Security/Scripting"><prop oor:name="MacroSecurityLevel" oor:op="fuse"><value>0</value></prop></item>
+ <item oor:path="/org.openoffice.Office.Common/Misc"><prop oor:name="FirstRun" oor:op="fuse"><value>false</value></prop></item>
+</oor:items>`;
+/** The Basic macro: load each doc Hidden, dispatch accept/reject-all (or, for `identity`,
+ *  dispatch nothing at all), save with the job's filter. */
+function module1Xba(jobsPath, markerPath) {
+    return `<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE script:module PUBLIC "-//OpenOffice.org//DTD OfficeDocument 1.0//EN" "module.dtd">
+<script:module xmlns:script="http://openoffice.org/2000/script" script:name="Module1" script:language="StarBasic">
+Sub RunOracle
+  Dim iFile As Integer, sLine As String, parts() As String, m As Integer
+  m = FreeFile : Open "${markerPath}" For Output As #m : Print #m, "started" : Close #m
+  iFile = FreeFile
+  Open "${jobsPath}" For Input As #iFile
+  Do While Not EOF(iFile)
+    Line Input #iFile, sLine
+    If Len(Trim(sLine)) &gt; 0 Then
+      parts = Split(sLine, "|")
+      ProcessOne(parts(0), ConvertToURL(parts(1)), ConvertToURL(parts(2)), parts(3))
+    End If
+  Loop
+  Close #iFile
+  StarDesktop.terminate()
+End Sub
+
+Sub ProcessOne(op As String, inUrl As String, outUrl As String, filterName As String)
+  Dim oDoc As Object, oFrame As Object, oDisp As Object
+  Dim loadArgs(0) As New com.sun.star.beans.PropertyValue
+  loadArgs(0).Name = "Hidden" : loadArgs(0).Value = True
+  oDoc = StarDesktop.loadComponentFromURL(inUrl, "_blank", 0, loadArgs())
+  oFrame = oDoc.getCurrentController().getFrame()
+  oDisp = createUnoService("com.sun.star.frame.DispatchHelper")
+  Dim noArgs()
+  If op = "accept" Then
+    oDisp.executeDispatch(oFrame, ".uno:AcceptAllTrackedChanges", "", 0, noArgs())
+  ElseIf op = "reject" Then
+    oDisp.executeDispatch(oFrame, ".uno:RejectAllTrackedChanges", "", 0, noArgs())
+  End If
+  ' op = "identity": no dispatch — a plain load-&gt;save, exposing LibreOffice's own DOCX
+  ' import/export of UNRESOLVED tracked changes (the oracle trust-boundary check).
+  Dim saveArgs(0) As New com.sun.star.beans.PropertyValue
+  saveArgs(0).Name = "FilterName" : saveArgs(0).Value = filterName
+  oDoc.storeToURL(outUrl, saveArgs())
+  oDoc.close(False)
+End Sub
+</script:module>`;
+}
+function isOdtJob(job) {
+    return 'odt' in job;
+}
+function isConvertJob(job) {
+    return 'docx' in job;
+}
+/**
+ * Run LibreOffice over a batch of jobs in ONE headless launch and return each job's resulting
+ * main XML part — `word/document.xml` for DOCX jobs, `content.xml` for ODT jobs. Throws if the
+ * binary is missing or the macro did not run.
+ */
+export async function runLibreOfficeOracle(jobs, soffice = resolveSoffice()) {
+    if (!soffice)
+        throw new Error('runLibreOfficeOracle: no soffice binary (call resolveSoffice() and skip)');
+    if (jobs.length === 0)
+        return [];
+    const work = mkdtempSync(path.join(os.tmpdir(), 'lo-oracle-'));
+    const profile = path.join(work, 'profile');
+    const userDir = path.join(profile, 'user');
+    const basicDir = path.join(userDir, 'basic', 'Standard');
+    const inDir = path.join(work, 'in');
+    const outDir = path.join(work, 'out');
+    const marker = path.join(work, 'macro_ran.txt');
+    const jobsPath = path.join(work, 'jobs.txt');
+    const profileUrl = pathToFileURL(profile).href;
+    let keepWork = false;
+    try {
+        for (const d of [userDir, basicDir, inDir, outDir])
+            mkdirSync(d, { recursive: true });
+        // Write each job's input package and build the jobs file (op|inPath|outPath|filter).
+        const outPaths = [];
+        const jobLines = [];
+        for (let i = 0; i < jobs.length; i++) {
+            const job = jobs[i];
+            const inExt = isOdtJob(job) ? 'odt' : 'docx';
+            const outExt = isOdtJob(job) || isConvertJob(job) ? 'odt' : 'docx';
+            const filter = isOdtJob(job) || isConvertJob(job) ? 'writer8' : 'MS Word 2007 XML';
+            const inPath = path.join(inDir, `job${i}.${inExt}`);
+            const outPath = path.join(outDir, `job${i}.${outExt}`);
+            writeFileSync(inPath, isOdtJob(job)
+                ? new Uint8Array(job.odt)
+                : isConvertJob(job)
+                    ? new Uint8Array(job.docx)
+                    : new Uint8Array(await packMinimalDocx(job.documentXml)));
+            outPaths.push(outPath);
+            jobLines.push(`${job.op}|${inPath}|${outPath}|${filter}`);
+        }
+        writeFileSync(jobsPath, jobLines.join('\n') + '\n');
+        // Macro security level 0 so the Standard-library macro runs headless.
+        writeFileSync(path.join(userDir, 'registrymodifications.xcu'), REGMOD);
+        const baseArgs = ['--headless', '--norestore', '--nologo', `-env:UserInstallation=${profileUrl}`];
+        const diag = [];
+        // (1) INIT the profile: a throwaway convert makes soffice populate user/basic/Standard
+        // (which would otherwise clobber a pre-placed Module1.xba on first real launch).
+        const init = await runSoffice(soffice, [...baseArgs, '--convert-to', 'txt', '--outdir', path.join(work, 'init'), path.join(inDir, 'job0.docx')], 20_000);
+        diag.push(`[init] ${(init.stderr || init.stdout || '(no output)').trim()}`);
+        // (2) Overwrite the Basic library with our macro.
+        mkdirSync(basicDir, { recursive: true });
+        writeFileSync(path.join(userDir, 'basic', 'script.xlc'), SCRIPT_XLC);
+        writeFileSync(path.join(basicDir, 'script.xlb'), SCRIPT_XLB);
+        writeFileSync(path.join(basicDir, 'Module1.xba'), module1Xba(jobsPath, marker));
+        // (3) Run the macro via a bare macro:/// URL — but only once the init instance has fully
+        // released the profile, so LibreOffice's single-instance model can't forward our macro command
+        // to the dying init process. Retry once: the first macro launch can still race a slow init exit.
+        for (let attempt = 1; attempt <= 2 && !existsSync(marker); attempt++) {
+            await settleProfile(profile);
+            const run = await runSoffice(soffice, [...baseArgs, 'macro:///Standard.Module1.RunOracle'], 45_000);
+            diag.push(`[macro attempt ${attempt}] ${(run.stderr || run.stdout || '(no output)').trim()}`);
+            if (!existsSync(marker))
+                await sleep(400);
+        }
+        if (!existsSync(marker)) {
+            keepWork = Boolean(process.env.SAFE_DOCX_ORACLE_DEBUG);
+            throw new Error('LibreOffice oracle macro did not run (no marker file) after 2 attempts.\nsoffice output:\n' +
+                diag.join('\n') +
+                (keepWork ? `\n(profile kept for debugging at ${work})` : ' (set SAFE_DOCX_ORACLE_DEBUG=1 to keep the profile)'));
+        }
+        return Promise.all(outPaths.map(async (p, i) => {
+            if (!existsSync(p))
+                throw new Error(`LibreOffice oracle produced no output for ${path.basename(p)}`);
+            if (isOdtJob(jobs[i]) || isConvertJob(jobs[i])) {
+                const contentXml = await readZipText(readFileSync(p), 'content.xml');
+                if (contentXml == null)
+                    throw new Error(`content.xml not found in oracle output ${path.basename(p)}`);
+                return contentXml;
+            }
+            return extractDocumentXml(readFileSync(p));
+        }));
+    }
+    finally {
+        if (!keepWork)
+            rmSync(work, { recursive: true, force: true });
+    }
+}
+/**
+ * Structural projection of a `word/document.xml`: one entry per top-level body paragraph,
+ * recording whether it carries visible text (a non-whitespace `w:t` descendant). This captures
+ * the paragraph-collapse claim the oracle is authoritative for — how many paragraphs survive and
+ * which collapsed to empty — without depending on revision-markup or formatting details (which
+ * LibreOffice rewrites). `[]`-length is the paragraph count.
+ */
+export function paragraphShape(documentXml) {
+    const doc = parseXml(documentXml);
+    const body = doc.getElementsByTagNameNS(W_NS, 'body').item(0) ?? doc.documentElement;
+    if (!body)
+        return [];
+    const shape = [];
+    for (let i = 0; i < body.childNodes.length; i++) {
+        const node = body.childNodes[i];
+        if (node.nodeType !== 1)
+            continue;
+        const el = node;
+        if (el.namespaceURI !== W_NS || el.localName !== 'p')
+            continue; // skip w:sectPr etc.
+        const texts = el.getElementsByTagNameNS(W_NS, 't');
+        let hasText = false;
+        for (let j = 0; j < texts.length; j++) {
+            if ((texts.item(j).textContent ?? '').trim().length > 0) {
+                hasText = true;
+                break;
+            }
+        }
+        shape.push(hasText);
+    }
+    return shape;
+}
+//# sourceMappingURL=libreoffice-oracle.js.map

package/dist/integration/libreoffice-oracle.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"libreoffice-oracle.js","sourceRoot":"","sources":["../../src/integration/libreoffice-oracle.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;GAoBG;AACH,OAAO,EAAE,QAAQ,EAAE,MAAM,oBAAoB,CAAC;AAC9C,OAAO,EAAE,UAAU,EAAE,SAAS,EAAE,WAAW,EAAE,YAAY,EAAE,MAAM,EAAE,aAAa,EAAE,MAAM,SAAS,CAAC;AAClG,OAAO,EAAE,MAAM,SAAS,CAAC;AACzB,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,aAAa,EAAE,MAAM,UAAU,CAAC;AACzC,OAAO,EAAE,SAAS,EAAE,MAAM,WAAW,CAAC;AACtC,OAAO,EAAE,eAAe,EAAE,WAAW,EAAE,MAAM,sBAAsB,CAAC;AACpE,OAAO,EAAE,QAAQ,EAAE,MAAM,sBAAsB,CAAC;AAEhD,MAAM,aAAa,GAAG,SAAS,CAAC,QAAQ,CAAC,CAAC;AAC1C,MAAM,IAAI,GAAG,8DAA8D,CAAC;AAE5E,MAAM,KAAK,GAAG,CAAC,EAAU,EAAiB,EAAE,CAAC,IAAI,OAAO,CAAC,CAAC,OAAO,EAAE,EAAE,CAAC,UAAU,CAAC,OAAO,EAAE,EAAE,CAAC,CAAC,CAAC;AAE/F;iCACiC;AACjC,KAAK,UAAU,UAAU,CACvB,OAAe,EACf,IAAc,EACd,OAAe;IAEf,IAAI,CAAC;QACH,MAAM,CAAC,GAAG,MAAM,aAAa,CAAC,OAAO,EAAE,IAAI,EAAE,EAAE,OAAO,EAAE,UAAU,EAAE,SAAS,EAAE,SAAS,EAAE,CAAC,GAAG,IAAI,GAAG,IAAI,EAAE,CAAC,CAAC;QAC7G,OAAO,EAAE,MAAM,EAAE,MAAM,CAAC,CAAC,CAAC,MAAM,IAAI,EAAE,CAAC,EAAE,MAAM,EAAE,MAAM,CAAC,CAAC,CAAC,MAAM,IAAI,EAAE,CAAC,EAAE,CAAC;IAC5E,CAAC;IAAC,OAAO,GAAG,EAAE,CAAC;QACb,MAAM,CAAC,GAAG,GAA+D,CAAC;QAC1E,OAAO,EAAE,MAAM,EAAE,MAAM,CAAC,CAAC,CAAC,MAAM,IAAI,EAAE,CAAC,EAAE,MAAM,EAAE,MAAM,CAAC,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,OAAO,IAAI,EAAE,CAAC,EAAE,CAAC;IACzF,CAAC;AACH,CAAC;AAED;;;;;;;GAOG;AACH,KAAK,UAAU,aAAa,CAAC,OAAe,EAAE,SAAS,GAAG,KAAK;IAC7D,MAAM,IAAI,GAAG,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,OAAO,CAAC,CAAC;IACzC,MAAM,QAAQ,GAAG,IAAI,CAAC,GAAG,EAAE,GAAG,SAAS,CAAC;IACxC,OAAO,UAAU,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,GAAG,EAAE,GAAG,QAAQ;QAAE,MAAM,KAAK,CAAC,GAAG,CAAC,CAAC;IACnE,IAAI,UAAU,CAAC,IAAI,CAAC,EAAE,CAAC;QACrB,IAAI,CAAC;YAAC,MAAM,CAAC,IAAI,EAAE,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC;QAAC,CAAC;QAAC,MAAM,CAAC,CAAC,4CAA4C,CAAC,CAAC;IAC/F,CAAC;AACH,CAAC;AAED,4FAA4F;AAC5F,MAAM,UAAU,cAAc;IAC5B,MAAM,UAAU,GAAG;QACjB,OAAO,CAAC,GAAG,CAAC,qBAAqB;QACjC,OAAO,CAAC,GAAG,CAAC,eAAe;QAC3B,2BAA2B;QAC3B,kBAAkB;QAClB,wBAAwB;QACxB,sDAAsD;KACvD,CAAC,MAAM,CAAC,OAAO,CAAa,CAAC;IAC9B,OAAO,UAAU,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC;AACvD,CAAC;AAED,MAAM,aAAa,GAAG;;;;;SAKb,CAAC;AACV,MAAM,IAAI,GAAG;;;iBAGI,CAAC;AAClB,MAAM,QAAQ,GAAG;sFACqE,CAAC;AAEvF,iFAAiF;AACjF,MAAM,CAAC,KAAK,UAAU,eAAe,CAAC,WAAmB;IACvD,OAAO,eAAe,CAAC;QACrB,qBAAqB,EAAE,aAAa;QACpC,aAAa,EAAE,IAAI;QACnB,8BAA8B,EAAE,QAAQ;QACxC,mBAAmB,EAAE,WAAW;KACjC,CAAC,CAAC;AACL,CAAC;AAED,4DAA4D;AAC5D,MAAM,CAAC,KAAK,UAAU,kBAAkB,CAAC,IAAY;IACnD,MAAM,GAAG,GAAG,MAAM,WAAW,CAAC,IAAI,EAAE,mBAAmB,CAAC,CAAC;IACzD,IAAI,GAAG,IAAI,IAAI;QAAE,MAAM,IAAI,KAAK,CAAC,8CAA8C,CAAC,CAAC;IACjF,OAAO,GAAG,CAAC;AACb,CAAC;AAED,MAAM,UAAU,GAAG;;;;qBAIE,CAAC;AACtB,MAAM,UAAU,GAAG;;;;mBAIA,CAAC;AACpB,MAAM,MAAM,GAAG;;;;aAIF,CAAC;AAEd;4DAC4D;AAC5D,SAAS,UAAU,CAAC,QAAgB,EAAE,UAAkB;IACtD,OAAO;;;;;yBAKgB,UAAU;;UAEzB,QAAQ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;iBAgCD,CAAC;AAClB,CAAC;AAuBD,SAAS,QAAQ,CAAC,GAAc;IAC9B,OAAO,KAAK,IAAI,GAAG,CAAC;AACtB,CAAC;AAED,SAAS,YAAY,CAAC,GAAc;IAClC,OAAO,MAAM,IAAI,GAAG,CAAC;AACvB,CAAC;AAED;;;;GAIG;AACH,MAAM,CAAC,KAAK,UAAU,oBAAoB,CAAC,IAAiB,EAAE,OAAO,GAAG,cAAc,EAAE;IACtF,IAAI,CAAC,OAAO;QAAE,MAAM,IAAI,KAAK,CAAC,0EAA0E,CAAC,CAAC;IAC1G,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IAEjC,MAAM,IAAI,GAAG,WAAW,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,MAAM,EAAE,EAAE,YAAY,CAAC,CAAC,CAAC;IAC/D,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,SAAS,CAAC,CAAC;IAC3C,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC;IAC3C,MAAM,QAAQ,GAAG,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,OAAO,EAAE,UAAU,CAAC,CAAC;IACzD,MAAM,KAAK,GAAG,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC;IACpC,MAAM,MAAM,GAAG,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,KAAK,CAAC,CAAC;IACtC,MAAM,MAAM,GAAG,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,eAAe,CAAC,CAAC;IAChD,MAAM,QAAQ,GAAG,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,UAAU,CAAC,CAAC;IAC7C,MAAM,UAAU,GAAG,aAAa,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC;IAC/C,IAAI,QAAQ,GAAG,KAAK,CAAC;IAErB,IAAI,CAAC;QACH,KAAK,MAAM,CAAC,IAAI,CAAC,OAAO,EAAE,QAAQ,EAAE,KAAK,EAAE,MAAM,CAAC;YAAE,SAAS,CAAC,CAAC,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;QAEtF,qFAAqF;QACrF,MAAM,QAAQ,GAAa,EAAE,CAAC;QAC9B,MAAM,QAAQ,GAAa,EAAE,CAAC;QAC9B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YACrC,MAAM,GAAG,GAAG,IAAI,CAAC,CAAC,CAAE,CAAC;YACrB,MAAM,KAAK,GAAG,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC;YAC7C,MAAM,MAAM,GAAG,QAAQ,CAAC,GAAG,CAAC,IAAI,YAAY,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC;YACnE,MAAM,MAAM,GAAG,QAAQ,CAAC,GAAG,CAAC,IAAI,YAAY,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,kBAAkB,CAAC;YACnF,MAAM,MAAM,GAAG,IAAI,CAAC,IAAI,CAAC,KAAK,EAAE,MAAM,CAAC,IAAI,KAAK,EAAE,CAAC,CAAC;YACpD,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,MAAM,CAAC,IAAI,MAAM,EAAE,CAAC,CAAC;YACvD,aAAa,CACX,MAAM,EACN,QAAQ,CAAC,GAAG,CAAC;gBACX,CAAC,CAAC,IAAI,UAAU,CAAC,GAAG,CAAC,GAAG,CAAC;gBACzB,CAAC,CAAC,YAAY,CAAC,GAAG,CAAC;oBACjB,CAAC,CAAC,IAAI,UAAU,CAAC,GAAG,CAAC,IAAI,CAAC;oBAC1B,CAAC,CAAC,IAAI,UAAU,CAAC,MAAM,eAAe,CAAC,GAAG,CAAC,WAAW,CAAC,CAAC,CAC7D,CAAC;YACF,QAAQ,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;YACvB,QAAQ,CAAC,IAAI,CAAC,GAAG,GAAG,CAAC,EAAE,IAAI,MAAM,IAAI,OAAO,IAAI,MAAM,EAAE,CAAC,CAAC;QAC5D,CAAC;QACD,aAAa,CAAC,QAAQ,EAAE,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,GAAG,IAAI,CAAC,CAAC;QAEpD,sEAAsE;QACtE,aAAa,CAAC,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,2BAA2B,CAAC,EAAE,MAAM,CAAC,CAAC;QAEvE,MAAM,QAAQ,GAAG,CAAC,YAAY,EAAE,aAAa,EAAE,UAAU,EAAE,yBAAyB,UAAU,EAAE,CAAC,CAAC;QAClG,MAAM,IAAI,GAAa,EAAE,CAAC;QAE1B,uFAAuF;QACvF,iFAAiF;QACjF,MAAM,IAAI,GAAG,MAAM,UAAU,CAC3B,OAAO,EACP,CAAC,GAAG,QAAQ,EAAE,cAAc,EAAE,KAAK,EAAE,UAAU,EAAE,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,MAAM,CAAC,EAAE,IAAI,CAAC,IAAI,CAAC,KAAK,EAAE,WAAW,CAAC,CAAC,EACxG,MAAM,CACP,CAAC;QACF,IAAI,CAAC,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,MAAM,IAAI,IAAI,CAAC,MAAM,IAAI,aAAa,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,CAAC;QAE5E,kDAAkD;QAClD,SAAS,CAAC,QAAQ,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;QACzC,aAAa,CAAC,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,OAAO,EAAE,YAAY,CAAC,EAAE,UAAU,CAAC,CAAC;QACrE,aAAa,CAAC,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,YAAY,CAAC,EAAE,UAAU,CAAC,CAAC;QAC7D,aAAa,CAAC,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,aAAa,CAAC,EAAE,UAAU,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAC,CAAC;QAEhF,yFAAyF;QACzF,+FAA+F;QAC/F,iGAAiG;QACjG,KAAK,IAAI,OAAO,GAAG,CAAC,EAAE,OAAO,IAAI,CAAC,IAAI,CAAC,UAAU,CAAC,MAAM,CAAC,EAAE,OAAO,EAAE,EAAE,CAAC;YACrE,MAAM,aAAa,CAAC,OAAO,CAAC,CAAC;YAC7B,MAAM,GAAG,GAAG,MAAM,UAAU,CAAC,OAAO,EAAE,CAAC,GAAG,QAAQ,EAAE,qCAAqC,CAAC,EAAE,MAAM,CAAC,CAAC;YACpG,IAAI,CAAC,IAAI,CAAC,kBAAkB,OAAO,KAAK,CAAC,GAAG,CAAC,MAAM,IAAI,GAAG,CAAC,MAAM,IAAI,aAAa,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,CAAC;YAC9F,IAAI,CAAC,UAAU,CAAC,MAAM,CAAC;gBAAE,MAAM,KAAK,CAAC,GAAG,CAAC,CAAC;QAC5C,CAAC;QAED,IAAI,CAAC,UAAU,CAAC,MAAM,CAAC,EAAE,CAAC;YACxB,QAAQ,GAAG,OAAO,CAAC,OAAO,CAAC,GAAG,CAAC,sBAAsB,CAAC,CAAC;YACvD,MAAM,IAAI,KAAK,CACb,4FAA4F;gBAC1F,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC;gBACf,CAAC,QAAQ,CAAC,CAAC,CAAC,oCAAoC,IAAI,GAAG,CAAC,CAAC,CAAC,qDAAqD,CAAC,CACnH,CAAC;QACJ,CAAC;QACD,OAAO,OAAO,CAAC,GAAG,CAAC,QAAQ,CAAC,GAAG,CAAC,KAAK,EAAE,CAAC,EAAE,CAAC,EAAE,EAAE;YAC7C,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC;gBAAE,MAAM,IAAI,KAAK,CAAC,6CAA6C,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;YACrG,IAAI,QAAQ,CAAC,IAAI,CAAC,CAAC,CAAE,CAAC,IAAI,YAAY,CAAC,IAAI,CAAC,CAAC,CAAE,CAAC,EAAE,CAAC;gBACjD,MAAM,UAAU,GAAG,MAAM,WAAW,CAAC,YAAY,CAAC,CAAC,CAAC,EAAE,aAAa,CAAC,CAAC;gBACrE,IAAI,UAAU,IAAI,IAAI;oBAAE,MAAM,IAAI,KAAK,CAAC,0CAA0C,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;gBACtG,OAAO,UAAU,CAAC;YACpB,CAAC;YACD,OAAO,kBAAkB,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC,CAAC;QAC7C,CAAC,CAAC,CAAC,CAAC;IACN,CAAC;YAAS,CAAC;QACT,IAAI,CAAC,QAAQ;YAAE,MAAM,CAAC,IAAI,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC;IAChE,CAAC;AACH,CAAC;AAED;;;;;;GAMG;AACH,MAAM,UAAU,cAAc,CAAC,WAAmB;IAChD,MAAM,GAAG,GAAG,QAAQ,CAAC,WAAW,CAAC,CAAC;IAClC,MAAM,IAAI,GAAG,GAAG,CAAC,sBAAsB,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,eAAe,CAAC;IACrF,IAAI,CAAC,IAAI;QAAE,OAAO,EAAE,CAAC;IACrB,MAAM,KAAK,GAAc,EAAE,CAAC;IAC5B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,UAAU,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QAChD,MAAM,IAAI,GAAG,IAAI,CAAC,UAAU,CAAC,CAAC,CAAE,CAAC;QACjC,IAAI,IAAI,CAAC,QAAQ,KAAK,CAAC;YAAE,SAAS;QAClC,MAAM,EAAE,GAAG,IAAe,CAAC;QAC3B,IAAI,EAAE,CAAC,YAAY,KAAK,IAAI,IAAI,EAAE,CAAC,SAAS,KAAK,GAAG;YAAE,SAAS,CAAC,qBAAqB;QACrF,MAAM,KAAK,GAAG,EAAE,CAAC,sBAAsB,CAAC,IAAI,EAAE,GAAG,CAAC,CAAC;QACnD,IAAI,OAAO,GAAG,KAAK,CAAC;QACpB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YACtC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,CAAE,CAAC,WAAW,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;gBAAC,OAAO,GAAG,IAAI,CAAC;gBAAC,MAAM;YAAC,CAAC;QACtF,CAAC;QACD,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;IACtB,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC"}

package/dist/integration/synthetic-docx-fixture.d.ts

@@ -2,6 +2,8 @@ export interface SyntheticDocxOptions {
     paragraphs: string[];
     footnoteOnParagraph?: number;
     footnoteText?: string;
+    endnoteOnParagraph?: number;
+    endnoteText?: string;
     commentOnParagraph?: number;
     commentText?: string;
     commentAuthor?: string;
@@ -46,14 +48,84 @@ export interface SyntheticDocxOptions {
         name: string;
         id?: number;
     };
+    /**
+     * Multi-paragraph comment range with body-level markers: the
+     * commentRangeStart is emitted as a sibling of <w:p> before
+     * paragraphs[startBeforeParagraph] and the commentRangeEnd as a sibling
+     * after paragraphs[endAfterParagraph]. The commentReference run is
+     * appended inside paragraphs[endAfterParagraph]. This is the issue #103
+     * shape: range markers spanning whole paragraphs sit outside any <w:p>.
+     *
+     * Mutually exclusive with commentOnParagraph and commentSpanParagraphs.
+     */
+    siblingCommentRange?: {
+        startBeforeParagraph: number;
+        endAfterParagraph: number;
+    };
+    /**
+     * Pre-existing tracked move with explicit in-paragraph range markers.
+     * paragraphs[from]'s text is wrapped in <w:moveFrom> (as w:delText)
+     * bracketed by w:moveFromRangeStart/End; paragraphs[to]'s text is wrapped
+     * in <w:moveTo> bracketed by w:moveToRangeStart/End. All four markers are
+     * direct children of their <w:p>, sharing the given w:name. Callers should
+     * pass identical text for paragraphs[from] and paragraphs[to] so the shape
+     * matches what Word produces for a real tracked move. Used to verify
+     * explicit move-range marker reconstruction (issue #110).
+     */
+    trackedMove?: {
+        from: number;
+        to: number;
+        name: string;
+        author?: string;
+        firstId?: number;
+    };
+    /**
+     * Range permission spanning paragraphs. permStart opens at the start of
+     * paragraphs[start] and permEnd closes at the end of paragraphs[end].
+     * start === end yields a single-paragraph span around that paragraph's run.
+     */
+    permSpanParagraphs?: {
+        start: number;
+        end: number;
+        id?: number;
+    };
+    /**
+     * Body-level (sibling of <w:p>) permStart/permEnd pair inserted between
+     * paragraphs[index - 1] and paragraphs[index]. Used to verify scaffold
+     * perm markers do not leak into reconstructed paragraphs.
+     */
+    siblingPermBefore?: {
+        index: number;
+        id?: number;
+    };
 }
 export declare function buildSyntheticDocx(opts: SyntheticDocxOptions): Promise<Buffer>;
+/** Parts for {@link buildDocxFromParts}. Only the body XML is required. */
+export interface DocxPartsOptions {
+    /** Raw `<w:body>` children (paragraphs/tables), without the `<w:body>` wrapper. */
+    bodyXml: string;
+    /** Full `word/styles.xml` content. */
+    stylesXml?: string;
+    /** Full `word/numbering.xml` content. */
+    numberingXml?: string;
+    /** Extra `<Relationship …/>` entries for `word/_rels/document.xml.rels` (e.g. hyperlinks). */
+    documentRelEntries?: string[];
+}
+/**
+ * Build a loadable DOCX from raw part XML. The `testing/ooxml-fixtures.ts`
+ * `buildDocxFromBodyXml` covers the body-only case for docx-core's own tests, but it lives
+ * in the build-excluded testing tree; this builder is exported from the package root for
+ * downstream suites (odf-core's DOCX→ODT conversion tests) and additionally accepts the
+ * optional styles/numbering/relationship parts those tests exercise.
+ */
+export declare function buildDocxFromParts(opts: DocxPartsOptions): Promise<Buffer>;
 export interface SyntheticResultParts {
     documentXml: string;
     footnotesXml: string | null;
     endnotesXml: string | null;
     commentsXml: string | null;
     commentsExtendedXml: string | null;
+    commentsIdsXml: string | null;
     peopleXml: string | null;
     contentTypesXml: string | null;
     relsXml: string | null;

package/dist/integration/synthetic-docx-fixture.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"synthetic-docx-fixture.d.ts","sourceRoot":"","sources":["../../src/integration/synthetic-docx-fixture.ts"],"names":[],"mappings":"AAGA,MAAM,WAAW,oBAAoB;IACnC,UAAU,EAAE,MAAM,EAAE,CAAC;IACrB,mBAAmB,CAAC,EAAE,MAAM,CAAC;IAC7B,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,kBAAkB,CAAC,EAAE,MAAM,CAAC;IAC5B,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,aAAa,CAAC,EAAE,MAAM,CAAC;IAGvB,qBAAqB,CAAC,EAAE,OAAO,CAAC;IAChC;;;;;;;OAOG;IACH,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB;;;;;;OAMG;IACH,qBAAqB,CAAC,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,GAAG,EAAE,MAAM,CAAA;KAAE,CAAC;IACvD;;;OAGG;IACH,mBAAmB,CAAC,EAAE;QAAE,SAAS,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAC;QAAC,EAAE,CAAC,EAAE,MAAM,CAAA;KAAE,CAAC;IACvE;;;;OAIG;IACH,qBAAqB,CAAC,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAC;QAAC,EAAE,CAAC,EAAE,MAAM,CAAA;KAAE,CAAC;CACtE;AAED,wBAAsB,kBAAkB,CAAC,IAAI,EAAE,oBAAoB,GAAG,OAAO,CAAC,MAAM,CAAC,CA4MpF;AAED,MAAM,WAAW,oBAAoB;IACnC,WAAW,EAAE,MAAM,CAAC;IACpB,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;IAC5B,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,mBAAmB,EAAE,MAAM,GAAG,IAAI,CAAC;IACnC,SAAS,EAAE,MAAM,GAAG,IAAI,CAAC;IACzB,eAAe,EAAE,MAAM,GAAG,IAAI,CAAC;IAC/B,OAAO,EAAE,MAAM,GAAG,IAAI,CAAC;CACxB;AAED,wBAAsB,cAAc,CAAC,YAAY,EAAE,MAAM,GAAG,OAAO,CAAC,oBAAoB,CAAC,CAYxF"}
+{"version":3,"file":"synthetic-docx-fixture.d.ts","sourceRoot":"","sources":["../../src/integration/synthetic-docx-fixture.ts"],"names":[],"mappings":"AAGA,MAAM,WAAW,oBAAoB;IACnC,UAAU,EAAE,MAAM,EAAE,CAAC;IACrB,mBAAmB,CAAC,EAAE,MAAM,CAAC;IAC7B,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,kBAAkB,CAAC,EAAE,MAAM,CAAC;IAC5B,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,kBAAkB,CAAC,EAAE,MAAM,CAAC;IAC5B,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,aAAa,CAAC,EAAE,MAAM,CAAC;IAGvB,qBAAqB,CAAC,EAAE,OAAO,CAAC;IAChC;;;;;;;OAOG;IACH,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB;;;;;;OAMG;IACH,qBAAqB,CAAC,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,GAAG,EAAE,MAAM,CAAA;KAAE,CAAC;IACvD;;;OAGG;IACH,mBAAmB,CAAC,EAAE;QAAE,SAAS,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAC;QAAC,EAAE,CAAC,EAAE,MAAM,CAAA;KAAE,CAAC;IACvE;;;;OAIG;IACH,qBAAqB,CAAC,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAC;QAAC,EAAE,CAAC,EAAE,MAAM,CAAA;KAAE,CAAC;IACrE;;;;;;;;;OASG;IACH,mBAAmB,CAAC,EAAE;QAAE,oBAAoB,EAAE,MAAM,CAAC;QAAC,iBAAiB,EAAE,MAAM,CAAA;KAAE,CAAC;IAClF;;;;;;;;;OASG;IACH,WAAW,CAAC,EAAE;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,EAAE,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAC;QAAC,MAAM,CAAC,EAAE,MAAM,CAAC;QAAC,OAAO,CAAC,EAAE,MAAM,CAAA;KAAE,CAAC;IAC5F;;;;OAIG;IACH,kBAAkB,CAAC,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,GAAG,EAAE,MAAM,CAAC;QAAC,EAAE,CAAC,EAAE,MAAM,CAAA;KAAE,CAAC;IACjE;;;;OAIG;IACH,iBAAiB,CAAC,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,EAAE,CAAC,EAAE,MAAM,CAAA;KAAE,CAAC;CACpD;AAED,wBAAsB,kBAAkB,CAAC,IAAI,EAAE,oBAAoB,GAAG,OAAO,CAAC,MAAM,CAAC,CA+SpF;AAED,2EAA2E;AAC3E,MAAM,WAAW,gBAAgB;IAC/B,mFAAmF;IACnF,OAAO,EAAE,MAAM,CAAC;IAChB,sCAAsC;IACtC,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,yCAAyC;IACzC,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,8FAA8F;IAC9F,kBAAkB,CAAC,EAAE,MAAM,EAAE,CAAC;CAC/B;AAED;;;;;;GAMG;AACH,wBAAsB,kBAAkB,CAAC,IAAI,EAAE,gBAAgB,GAAG,OAAO,CAAC,MAAM,CAAC,CAiDhF;AAED,MAAM,WAAW,oBAAoB;IACnC,WAAW,EAAE,MAAM,CAAC;IACpB,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;IAC5B,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,mBAAmB,EAAE,MAAM,GAAG,IAAI,CAAC;IACnC,cAAc,EAAE,MAAM,GAAG,IAAI,CAAC;IAC9B,SAAS,EAAE,MAAM,GAAG,IAAI,CAAC;IACzB,eAAe,EAAE,MAAM,GAAG,IAAI,CAAC;IAC/B,OAAO,EAAE,MAAM,GAAG,IAAI,CAAC;CACxB;AAED,wBAAsB,cAAc,CAAC,YAAY,EAAE,MAAM,GAAG,OAAO,CAAC,oBAAoB,CAAC,CAaxF"}