ooxml.js 4.2.0 → 4.3.0

package/README.md

@@ -144,7 +144,7 @@ const pkg = buildXlsxPackageFromContent(content); // a fresh Package built from
 
 Both docx readers read the block-scoped fidelity constructs — structured document tags, fields, bookmarks, and tracked changes. In the flat form they are `document-schema.js` construct-boundary markers bracketing the blocks each one spans, one matched `constructStart`/`constructEnd` pair per construct, nesting as brackets do; in the tree they are promoted to a group over exactly that region, carrying the same descriptor. Both docx writers write them back.
 
-A construct whose extent is a sub-sequence of one paragraph's runs — a field inside a sentence, a bookmark over three words, a few inserted words in an otherwise untouched paragraph — has no marker encoding: the marker pair brackets whole blocks, so those occurrences keep their existing treatment (the text still comes through; the construct itself does not). See the gotchas below.
+A bookmark whose extent is a sub-sequence of one paragraph's runs — the bookmark over three words inside a sentence — is read onto that paragraph's own `constructs` field (`document-schema.js`'s run-level construct extents: a descriptor plus a half-open run range, so the paragraph is never split to host a wrapper), and the docx writer puts its `w:bookmarkStart`/`w:bookmarkEnd` halves back between the runs the range names. Two such bookmarks whose extents cross both survive — run ranges are data, not brackets. The other run-level constructs — a field inside a sentence, a run-level content control, a few inserted words in an otherwise untouched paragraph, and a bookmark whose halves sit in two different paragraphs — still have no encoding here: the text comes through, the construct does not. See the gotchas below.
 
 A construct with format-specific specifics the harmonised vocabulary does not name carries them in the descriptor's quarantined `source` residue (`document-schema.js`'s residue channel): an SDT whose `w:docPartObj` gallery is not the Table of Contents degrades to a `richText` control with the whole `w:docPartObj` element carried verbatim, and the docx writer re-emits that element from the residue in place of the default `w:richText` type element — the restorable tier in action, the gallery name no longer lost to the round trip.
 
@@ -296,7 +296,7 @@ The package layers a lossless core outward to lossy convenience views:
 ## Gotchas and quirks
 
 - **A `Content` suffix means the flat form, not a lesser one.** `readDocx` and `readDocxContent` are the same read; the suffix says which shape comes back (`DocumentPackage` versus the flat `DocxDocument`/`ContentDocument`), never which is more faithful. The pair that genuinely differs in what it reads is `readXlsxWorkbook`, whose name says so by naming its own return type rather than taking a suffix.
-- **A construct marker brackets whole blocks, never a sub-sequence of runs.** `document-schema.js`'s flat form has no run-level extent, so an inline field, a mid-paragraph bookmark, a run-level content control, or a few inserted words inside an otherwise untouched paragraph are read exactly as before — the text survives, the construct does not. So does an extent that crosses another, or that straddles a section break: neither is expressible as balanced brackets, and both are dropped rather than emitted at a position that would decode to a different nesting.
+- **A construct marker brackets whole blocks, never a sub-sequence of runs.** A mid-paragraph bookmark is the one run-level construct with an encoding here — it lands on the paragraph's `constructs` field as a run range (see the fidelity-constructs section above). An inline field, a run-level content control, or a few inserted words inside an otherwise untouched paragraph are still read exactly as before — the text survives, the construct does not — as is a bookmark whose two halves sit in different paragraphs. So does a block extent that crosses another, or that straddles a section break or table-cell boundary: neither is expressible as balanced brackets (nor as a tree group, which is what the marker pair promotes to), and both are dropped — a drop `document-schema.js`'s extent-scope note ratifies — rather than emitted at a position that would decode to a different nesting.
 - **The docx pair round-trips `sections`, not a whole `DocxDocument`.** Numbering definitions, cell border styling/shading, and `w:themeColor` (without `themeShade`/`themeTint`) are read; images read into `ContentImageBlock` (floating `wp:anchor` position not recorded); `PAGE`/`NUMPAGES` fields resolve to Word's cached text. An inline `w:object`/`o:OLEObject` whose payload part is itself a ZIP archive — a modern producer's embedded xlsx/docx/pptx, detected by magic bytes through `archive-codec` — or a classic OLE compound-file `.bin` whose root storage carries the embedded file as an OLE-packaged `Package` stream, read through `archive-codec`'s bounded CFB reader, is decoded into a nested content document and emitted as a `ContentEmbeddedObjectBlock` beside its containing paragraph's own block, sized from `w:dxaOrig`/`w:dyaOrig`; the object's VML preview picture (`v:imagedata`) and a `w:object` inside a header/footer/footnote (recovery walks the document body's block flow only) are not read, a compound file holding native legacy streams (BIFF and friends — no `Package` stream, or a packaged file that is not a ZIP) stays opaque, and a payload that does not decode as one of the three OOXML flavours (a plain archive, corrupt zip or compound-file data, a bodyless docx) degrades to no embedded block rather than failing the host read. The writer emits such a block back out as the inverse `w:object`/`o:OLEObject`: the nested document is re-serialised through its own format's builder into a fresh `word/embeddings/oleObject<N>.<docx|xlsx>` ZIP part carrying its relationship and content-type override, and `w:dxaOrig`/`w:dyaOrig` are derived from the block's frame (twips). An embedded presentation document has no serialiser here — pptx is read-only in this package — so a `ContentDocument` carrying one is refused with a thrown error rather than a silent drop, and no VML preview picture is regenerated (the reader never read one into the model, and real previews are WMF/EMF this ecosystem has no writer for; Word shows a blank until activated). The writer emits no `styles.xml`, `numbering.xml`, comments, footnotes, or headers/footers, so the `DocxDocument` fields outside `sections` do not come back — a paragraph's `styleId` is still written as a `w:pStyle` reference, resolving to nothing, since every property the style would have contributed is already spelled as direct formatting by then.
 - **pptx is read-only.** Connector shapes (`p:cxnSp`) are skipped; shape rotation composes through groups; a chart graphic frame reads its chart part's cached series/category model into a table block (header row = series names, one row per category index); a SmartArt graphic frame reads its diagram data model's node text as paragraphs in diagram order (depth-first over `parOf` connections, siblings by `srcOrd`); an OLE graphic frame reads the fallback picture its `mc:Fallback` carries (or, with none reachable, a paragraph naming the `p:oleObj`'s `progId`), and when the payload part is itself a ZIP archive — a modern producer's embedded xlsx/docx/pptx, detected by magic bytes through `archive-codec` — or a classic OLE compound-file `.bin` whose root storage carries the embedded file as an OLE-packaged `Package` stream, read through `archive-codec`'s bounded CFB reader, also decodes it into a nested content document and emits it as an `embeddedObject` block beside the picture, sized to the frame's own geometry; a compound file holding native legacy streams (no `Package` stream, or a packaged file that is not a ZIP) stays opaque external-application data, and a payload that does not decode as one of the three OOXML flavours degrades to no embedded block rather than failing the slide read.
 - **xlsx has no native percentage/currency/date/time cell type.** Both directions are closed via the number-format engine: reading classifies style → format code → kind; writing emits interned `numFmt` codes, fed back through the classifier in tests. `displayText` is the typed-value spelling, not the producer's rendered string.

package/dist/index.cjs

@@ -15,7 +15,7 @@ const require_typed_util = require("./typed/util.cjs");
 const require_typed_shared_color = require("./typed/shared/color.cjs");
 const require_typed_shared_metadata = require("./typed/shared/metadata.cjs");
 const require_image_sniff = require("./image/sniff.cjs");
-const require_read = require("./read-C3AaxcT9.cjs");
+const require_read = require("./read-Di5UfiSh.cjs");
 const require_typed_xlsx_content = require("./typed/xlsx/content.cjs");
 const require_typed_docx_numbering = require("./typed/docx/numbering.cjs");
 const require_typed_xlsx_build = require("./typed/xlsx/build.cjs");

package/dist/index.js

@@ -14,7 +14,7 @@ import { attr, childrenWithTag, decodeEntities, elementsWithTag, resolveRelation
 import { applyColorTransforms } from "./typed/shared/color.js";
 import { DocumentMetadataSchema } from "./typed/shared/metadata.js";
 import { sniffImageFormat } from "./image/sniff.js";
-import { c as readPptxContent, i as readDocxContent, n as DocxDocumentSchema, r as FootnoteSchema, s as PptxDocumentSchema, t as CommentSchema } from "./read-fuJpC4BH.js";
+import { c as readPptxContent, i as readDocxContent, n as DocxDocumentSchema, r as FootnoteSchema, s as PptxDocumentSchema, t as CommentSchema } from "./read-BQoFiqsb.js";
 import { readXlsxContent } from "./typed/xlsx/content.js";
 import { NumberingDefinitionSchema, NumberingLevelSchema, readNumberingDefinitions } from "./typed/docx/numbering.js";
 import { buildXlsxPackageFromContent } from "./typed/xlsx/build.js";

package/dist/{read-fuJpC4BH.js → read-BQoFiqsb.js}

@@ -12,7 +12,7 @@ import { readDiagramText } from "./typed/pptx/diagram.js";
 import { readXlsxContent } from "./typed/xlsx/content.js";
 import { resolveParagraphProperties, resolveRunProperties } from "./typed/docx/styles.js";
 import { NumberingDefinitionSchema, readNumberingDefinitions } from "./typed/docx/numbering.js";
-import { PROVENANCE_CHANGE_BY_TAG, bookmarkAnchorDescriptor, contentBearingChildren, fieldCharType, indexParagraphContent, insertConstructMarkers, isDeletedChange, readContentControlDescriptor, readProvenanceDescriptor, runInstructionText } from "./typed/docx/constructs.js";
+import { PROVENANCE_CHANGE_BY_TAG, bookmarkAnchorDescriptor, contentBearingChildren, fieldCharType, indexParagraphContent, insertConstructMarkers, isDeletedChange, readContentControlDescriptor, readProvenanceDescriptor, runBookmarkExtents, runInstructionText } from "./typed/docx/constructs.js";
 import { z } from "zod";
 import { COLOR_BLACK, ContentSectionSchema, ContentSlideSchema, PAGE_SIZE_LETTER, SLIDE_SIZE_WIDESCREEN, clampHeadingLevel, rgbHexToColor } from "document-schema.js";
 import { isCompoundFile, isZipArchive, readCompoundFile, readOlePackage, walkArchive } from "archive-codec";
@@ -662,7 +662,7 @@ function readRun(run, paragraph, context) {
 		color: props.color
 	};
 }
-function readParagraphRuns(paragraph, ctx, carryDeletions) {
+function readParagraphRuns(paragraph, ctx, carryDeletions, halves) {
 	const runs = [];
 	let fieldState = "none";
 	function walk(nodes, hyperlinkTarget) {
@@ -693,6 +693,18 @@ function readParagraphRuns(paragraph, ctx, carryDeletions) {
 			} else if (node.tag === "w:sdt") {
 				const sdtContent = childrenWithTag(node, "w:sdtContent")[0];
 				if (sdtContent !== void 0) walk(sdtContent.children, hyperlinkTarget);
+			} else if (node.tag === "w:bookmarkStart" || node.tag === "w:bookmarkEnd") {
+				const id = attr(node, "w:id");
+				if (id !== void 0) {
+					const name = attr(node, "w:name");
+					halves.push({
+						element: node,
+						id,
+						name: node.tag === "w:bookmarkStart" && name !== void 0 ? decodeEntities(name) : void 0,
+						kind: node.tag === "w:bookmarkStart" ? "start" : "end",
+						runPosition: runs.length
+					});
+				}
 			}
 		}
 	}
@@ -703,9 +715,13 @@ function readParagraph(paragraph, ctx, carryDeletions) {
 	const pPr = childrenWithTag(paragraph, "w:pPr")[0];
 	const pStyleEl = pPr === void 0 ? void 0 : childrenWithTag(pPr, "w:pStyle")[0];
 	const props = resolveParagraphProperties(paragraph, ctx.styles);
+	const halves = [];
+	const runs = readParagraphRuns(paragraph, ctx, carryDeletions, halves);
+	const constructs = runBookmarkExtents(halves, indexParagraphContent(paragraph));
 	return {
 		kind: "paragraph",
-		runs: readParagraphRuns(paragraph, ctx, carryDeletions),
+		runs,
+		...constructs.length > 0 ? { constructs } : {},
 		styleId: pStyleEl === void 0 ? void 0 : attr(pStyleEl, "w:val"),
 		headingLevel: props.outlineLvl === void 0 ? void 0 : clampHeadingLevel(props.outlineLvl + 1),
 		alignment: props.alignment,

package/dist/{read-C3AaxcT9.cjs → read-Di5UfiSh.cjs}

@@ -662,7 +662,7 @@ function readRun(run, paragraph, context) {
 		color: props.color
 	};
 }
-function readParagraphRuns(paragraph, ctx, carryDeletions) {
+function readParagraphRuns(paragraph, ctx, carryDeletions, halves) {
 	const runs = [];
 	let fieldState = "none";
 	function walk(nodes, hyperlinkTarget) {
@@ -693,6 +693,18 @@ function readParagraphRuns(paragraph, ctx, carryDeletions) {
 			} else if (node.tag === "w:sdt") {
 				const sdtContent = require_typed_util.childrenWithTag(node, "w:sdtContent")[0];
 				if (sdtContent !== void 0) walk(sdtContent.children, hyperlinkTarget);
+			} else if (node.tag === "w:bookmarkStart" || node.tag === "w:bookmarkEnd") {
+				const id = require_typed_util.attr(node, "w:id");
+				if (id !== void 0) {
+					const name = require_typed_util.attr(node, "w:name");
+					halves.push({
+						element: node,
+						id,
+						name: node.tag === "w:bookmarkStart" && name !== void 0 ? require_typed_util.decodeEntities(name) : void 0,
+						kind: node.tag === "w:bookmarkStart" ? "start" : "end",
+						runPosition: runs.length
+					});
+				}
 			}
 		}
 	}
@@ -703,9 +715,13 @@ function readParagraph(paragraph, ctx, carryDeletions) {
 	const pPr = require_typed_util.childrenWithTag(paragraph, "w:pPr")[0];
 	const pStyleEl = pPr === void 0 ? void 0 : require_typed_util.childrenWithTag(pPr, "w:pStyle")[0];
 	const props = require_typed_docx_styles.resolveParagraphProperties(paragraph, ctx.styles);
+	const halves = [];
+	const runs = readParagraphRuns(paragraph, ctx, carryDeletions, halves);
+	const constructs = require_typed_docx_constructs.runBookmarkExtents(halves, require_typed_docx_constructs.indexParagraphContent(paragraph));
 	return {
 		kind: "paragraph",
-		runs: readParagraphRuns(paragraph, ctx, carryDeletions),
+		runs,
+		...constructs.length > 0 ? { constructs } : {},
 		styleId: pStyleEl === void 0 ? void 0 : require_typed_util.attr(pStyleEl, "w:val"),
 		headingLevel: props.outlineLvl === void 0 ? void 0 : (0, document_schema_js.clampHeadingLevel)(props.outlineLvl + 1),
 		alignment: props.alignment,

package/dist/typed/document-package.cjs

@@ -1,5 +1,5 @@
 Object.defineProperty(exports, Symbol.toStringTag, { value: "Module" });
-const require_read = require("../read-C3AaxcT9.cjs");
+const require_read = require("../read-Di5UfiSh.cjs");
 const require_typed_xlsx_content = require("./xlsx/content.cjs");
 const require_typed_xlsx_build = require("./xlsx/build.cjs");
 const require_typed_docx_write = require("./docx/write.cjs");

package/dist/typed/document-package.js

@@ -1,4 +1,4 @@
-import { c as readPptxContent, i as readDocxContent } from "../read-fuJpC4BH.js";
+import { c as readPptxContent, i as readDocxContent } from "../read-BQoFiqsb.js";
 import { readXlsxContent } from "./xlsx/content.js";
 import { buildXlsxPackageFromContent } from "./xlsx/build.js";
 import { buildDocxPackageFromContent } from "./docx/write.js";

package/dist/typed/docx/constructs.cjs

@@ -86,6 +86,37 @@ function insertConstructMarkers(blocks, extents) {
 	}
 	return out;
 }
+function isBlockScopedHalf(half, index) {
+	const position = index.elements.indexOf(half.element);
+	if (position === -1) return false;
+	const leading = index.firstContentIndex === -1 || position < index.firstContentIndex;
+	const trailing = index.lastContentIndex === -1 || position > index.lastContentIndex;
+	return leading || trailing;
+}
+function runBookmarkExtents(halves, index) {
+	const byId = /* @__PURE__ */ new Map();
+	for (const half of halves) {
+		const existing = byId.get(half.id);
+		if (existing === void 0) byId.set(half.id, [half]);
+		else existing.push(half);
+	}
+	const extents = [];
+	for (const pair of byId.values()) {
+		const starts = pair.filter((half) => half.kind === "start");
+		const ends = pair.filter((half) => half.kind === "end");
+		const open = starts[0];
+		const close = ends[0];
+		if (starts.length !== 1 || ends.length !== 1 || open === void 0 || close === void 0) continue;
+		if (open.name === void 0 || close.runPosition < open.runPosition) continue;
+		if (isBlockScopedHalf(open, index) && isBlockScopedHalf(close, index)) continue;
+		extents.push({
+			descriptor: bookmarkAnchorDescriptor(open.name),
+			startRun: open.runPosition,
+			endRun: close.runPosition
+		});
+	}
+	return extents;
+}
 const CONTROL_TYPE_BY_TAG = /* @__PURE__ */ new Map([
 	["w:richText", "richText"],
 	["w:text", "plainText"],
@@ -230,4 +261,5 @@ exports.insertConstructMarkers = insertConstructMarkers;
 exports.isDeletedChange = isDeletedChange;
 exports.readContentControlDescriptor = readContentControlDescriptor;
 exports.readProvenanceDescriptor = readProvenanceDescriptor;
+exports.runBookmarkExtents = runBookmarkExtents;
 exports.runInstructionText = runInstructionText;

package/dist/typed/docx/constructs.d.cts

@@ -1,5 +1,5 @@
 import { l as XmlElement } from "../../node-CkBWRjNR.cjs";
-import { AnchorDescriptor, ConstructDescriptor, ContentBlock, ContentControlDescriptor, ProvenanceChange, ProvenanceDescriptor } from "document-schema.js";
+import { AnchorDescriptor, ConstructDescriptor, ContentBlock, ContentControlDescriptor, ProvenanceChange, ProvenanceDescriptor, RunConstructExtent } from "document-schema.js";
 //#region src/typed/docx/constructs.d.ts
 interface ParagraphContentIndex {
   readonly elements: readonly XmlElement[];
@@ -15,6 +15,14 @@ interface ConstructExtent {
   readonly descriptor: ConstructDescriptor;
 }
 declare function insertConstructMarkers(blocks: readonly ContentBlock[], extents: readonly ConstructExtent[]): ContentBlock[];
+interface ParagraphBookmarkHalf {
+  readonly element: XmlElement;
+  readonly id: string;
+  readonly name: string | undefined;
+  readonly kind: 'start' | 'end';
+  readonly runPosition: number;
+}
+declare function runBookmarkExtents(halves: readonly ParagraphBookmarkHalf[], index: ParagraphContentIndex): RunConstructExtent[];
 declare const TABLE_OF_CONTENTS_GALLERY = "Table of Contents";
 declare function readContentControlDescriptor(sdt: XmlElement): ContentControlDescriptor;
 declare const PROVENANCE_CHANGE_BY_TAG: ReadonlyMap<string, ProvenanceChange>;
@@ -24,4 +32,4 @@ declare function bookmarkAnchorDescriptor(name: string): AnchorDescriptor;
 declare function runInstructionText(run: XmlElement): string;
 declare function fieldCharType(run: XmlElement): string | undefined;
 //#endregion
-export { ConstructExtent, PROVENANCE_CHANGE_BY_TAG, ParagraphContentIndex, TABLE_OF_CONTENTS_GALLERY, bookmarkAnchorDescriptor, contentBearingChildren, fieldCharType, indexParagraphContent, insertConstructMarkers, isDeletedChange, readContentControlDescriptor, readProvenanceDescriptor, runInstructionText };
+export { ConstructExtent, PROVENANCE_CHANGE_BY_TAG, ParagraphBookmarkHalf, ParagraphContentIndex, TABLE_OF_CONTENTS_GALLERY, bookmarkAnchorDescriptor, contentBearingChildren, fieldCharType, indexParagraphContent, insertConstructMarkers, isDeletedChange, readContentControlDescriptor, readProvenanceDescriptor, runBookmarkExtents, runInstructionText };

package/dist/typed/docx/constructs.d.ts

@@ -1,5 +1,5 @@
 import { l as XmlElement } from "../../node-CkBWRjNR.js";
-import { AnchorDescriptor, ConstructDescriptor, ContentBlock, ContentControlDescriptor, ProvenanceChange, ProvenanceDescriptor } from "document-schema.js";
+import { AnchorDescriptor, ConstructDescriptor, ContentBlock, ContentControlDescriptor, ProvenanceChange, ProvenanceDescriptor, RunConstructExtent } from "document-schema.js";
 //#region src/typed/docx/constructs.d.ts
 interface ParagraphContentIndex {
   readonly elements: readonly XmlElement[];
@@ -15,6 +15,14 @@ interface ConstructExtent {
   readonly descriptor: ConstructDescriptor;
 }
 declare function insertConstructMarkers(blocks: readonly ContentBlock[], extents: readonly ConstructExtent[]): ContentBlock[];
+interface ParagraphBookmarkHalf {
+  readonly element: XmlElement;
+  readonly id: string;
+  readonly name: string | undefined;
+  readonly kind: 'start' | 'end';
+  readonly runPosition: number;
+}
+declare function runBookmarkExtents(halves: readonly ParagraphBookmarkHalf[], index: ParagraphContentIndex): RunConstructExtent[];
 declare const TABLE_OF_CONTENTS_GALLERY = "Table of Contents";
 declare function readContentControlDescriptor(sdt: XmlElement): ContentControlDescriptor;
 declare const PROVENANCE_CHANGE_BY_TAG: ReadonlyMap<string, ProvenanceChange>;
@@ -24,4 +32,4 @@ declare function bookmarkAnchorDescriptor(name: string): AnchorDescriptor;
 declare function runInstructionText(run: XmlElement): string;
 declare function fieldCharType(run: XmlElement): string | undefined;
 //#endregion
-export { ConstructExtent, PROVENANCE_CHANGE_BY_TAG, ParagraphContentIndex, TABLE_OF_CONTENTS_GALLERY, bookmarkAnchorDescriptor, contentBearingChildren, fieldCharType, indexParagraphContent, insertConstructMarkers, isDeletedChange, readContentControlDescriptor, readProvenanceDescriptor, runInstructionText };
+export { ConstructExtent, PROVENANCE_CHANGE_BY_TAG, ParagraphBookmarkHalf, ParagraphContentIndex, TABLE_OF_CONTENTS_GALLERY, bookmarkAnchorDescriptor, contentBearingChildren, fieldCharType, indexParagraphContent, insertConstructMarkers, isDeletedChange, readContentControlDescriptor, readProvenanceDescriptor, runBookmarkExtents, runInstructionText };

package/dist/typed/docx/constructs.js

@@ -85,6 +85,37 @@ function insertConstructMarkers(blocks, extents) {
 	}
 	return out;
 }
+function isBlockScopedHalf(half, index) {
+	const position = index.elements.indexOf(half.element);
+	if (position === -1) return false;
+	const leading = index.firstContentIndex === -1 || position < index.firstContentIndex;
+	const trailing = index.lastContentIndex === -1 || position > index.lastContentIndex;
+	return leading || trailing;
+}
+function runBookmarkExtents(halves, index) {
+	const byId = /* @__PURE__ */ new Map();
+	for (const half of halves) {
+		const existing = byId.get(half.id);
+		if (existing === void 0) byId.set(half.id, [half]);
+		else existing.push(half);
+	}
+	const extents = [];
+	for (const pair of byId.values()) {
+		const starts = pair.filter((half) => half.kind === "start");
+		const ends = pair.filter((half) => half.kind === "end");
+		const open = starts[0];
+		const close = ends[0];
+		if (starts.length !== 1 || ends.length !== 1 || open === void 0 || close === void 0) continue;
+		if (open.name === void 0 || close.runPosition < open.runPosition) continue;
+		if (isBlockScopedHalf(open, index) && isBlockScopedHalf(close, index)) continue;
+		extents.push({
+			descriptor: bookmarkAnchorDescriptor(open.name),
+			startRun: open.runPosition,
+			endRun: close.runPosition
+		});
+	}
+	return extents;
+}
 const CONTROL_TYPE_BY_TAG = /* @__PURE__ */ new Map([
 	["w:richText", "richText"],
 	["w:text", "plainText"],
@@ -219,4 +250,4 @@ function fieldCharType(run) {
 	return fldChar === void 0 ? void 0 : attr(fldChar, "w:fldCharType");
 }
 //#endregion
-export { PROVENANCE_CHANGE_BY_TAG, TABLE_OF_CONTENTS_GALLERY, bookmarkAnchorDescriptor, contentBearingChildren, fieldCharType, indexParagraphContent, insertConstructMarkers, isDeletedChange, readContentControlDescriptor, readProvenanceDescriptor, runInstructionText };
+export { PROVENANCE_CHANGE_BY_TAG, TABLE_OF_CONTENTS_GALLERY, bookmarkAnchorDescriptor, contentBearingChildren, fieldCharType, indexParagraphContent, insertConstructMarkers, isDeletedChange, readContentControlDescriptor, readProvenanceDescriptor, runBookmarkExtents, runInstructionText };

package/dist/typed/docx/read.cjs

@@ -1,5 +1,5 @@
 Object.defineProperty(exports, Symbol.toStringTag, { value: "Module" });
-const require_read = require("../../read-C3AaxcT9.cjs");
+const require_read = require("../../read-Di5UfiSh.cjs");
 exports.CommentSchema = require_read.CommentSchema;
 exports.DocxDocumentSchema = require_read.DocxDocumentSchema;
 exports.FootnoteSchema = require_read.FootnoteSchema;

package/dist/typed/docx/read.js

@@ -1,2 +1,2 @@
-import { i as readDocxContent, n as DocxDocumentSchema, r as FootnoteSchema, t as CommentSchema } from "../../read-fuJpC4BH.js";
+import { i as readDocxContent, n as DocxDocumentSchema, r as FootnoteSchema, t as CommentSchema } from "../../read-BQoFiqsb.js";
 export { CommentSchema, DocxDocumentSchema, FootnoteSchema, readDocxContent };

package/dist/typed/docx/write.cjs

@@ -179,10 +179,49 @@ function buildParagraph(paragraph, state, pageBreakBefore, deleted, provenance)
 	const changeTag = provenance === void 0 ? void 0 : TRACKED_CHANGE_TAG_BY_CHANGE[provenance.change];
 	const pPr = properties === void 0 && changeTag !== void 0 ? require_xml_fragment.el("w:pPr", {}, []) : properties;
 	if (pPr !== void 0 && changeTag !== void 0 && provenance !== void 0) pPr.children.push(require_xml_fragment.el("w:rPr", {}, [require_xml_fragment.el(changeTag, trackChangeAttrs(state, provenance))]));
-	const runs = paragraph.runs.map((run) => buildRun(run, state, deleted));
+	const runs = interleaveRunConstructExtents(paragraph.runs.map((run) => buildRun(run, state, deleted)), paragraph, state);
 	const content = changeTag === void 0 || provenance === void 0 ? runs : [require_xml_fragment.el(changeTag, trackChangeAttrs(state, provenance), runs)];
 	return require_xml_fragment.el("w:p", {}, [...pPr === void 0 ? [] : [pPr], ...content]);
 }
+function interleaveRunConstructExtents(runElements, paragraph, state) {
+	if (paragraph.constructs === void 0) return [...runElements];
+	const fault = (0, document_schema_js.findRunConstructFault)(paragraph);
+	if (fault !== void 0) throw new Error(`buildDocxPackageFromContent: run-level construct extent at index ${String(fault.index)} of a paragraph does not name real runs (${fault.kind})`);
+	const bookmarks = paragraph.constructs.filter((extent) => extent.descriptor.kind === "anchor" && extent.descriptor.anchorType === "bookmark");
+	if (bookmarks.length === 0) return [...runElements];
+	const closingAt = /* @__PURE__ */ new Map();
+	const openingAt = /* @__PURE__ */ new Map();
+	const pointAt = /* @__PURE__ */ new Map();
+	const push = (map, position, element) => {
+		const existing = map.get(position);
+		if (existing === void 0) map.set(position, [element]);
+		else existing.push(element);
+	};
+	for (const bookmark of bookmarks) {
+		const id = String(state.nextMarkerId++);
+		const open = require_xml_fragment.el("w:bookmarkStart", {
+			"w:id": id,
+			"w:name": require_xml_entities.encodeXmlText(bookmark.descriptor.name)
+		});
+		const close = require_xml_fragment.el("w:bookmarkEnd", { "w:id": id });
+		if (bookmark.startRun === bookmark.endRun) {
+			push(pointAt, bookmark.startRun, open);
+			push(pointAt, bookmark.startRun, close);
+		} else {
+			push(openingAt, bookmark.startRun, open);
+			push(closingAt, bookmark.endRun, close);
+		}
+	}
+	const out = [];
+	for (let position = 0; position <= runElements.length; position++) {
+		for (const close of closingAt.get(position) ?? []) out.push(close);
+		for (const open of openingAt.get(position) ?? []) out.push(open);
+		for (const half of pointAt.get(position) ?? []) out.push(half);
+		const run = runElements[position];
+		if (run !== void 0) out.push(run);
+	}
+	return out;
+}
 function trailingEmptyRunElements(paragraph, element) {
 	const runElements = [];
 	for (const child of element.children) if (child.type === "element" && (child.tag === "w:r" || child.tag === "w:hyperlink")) runElements.push(child);

package/dist/typed/docx/write.js

@@ -6,7 +6,7 @@ import { encodeXmlText } from "../../xml/entities.js";
 import { ptToEighthPoints, ptToEmu, ptToHalfPoints, ptToTwips } from "../shared/units.js";
 import { TABLE_OF_CONTENTS_GALLERY, isDeletedChange } from "./constructs.js";
 import { buildXlsxPackageFromContent } from "../xlsx/build.js";
-import { colorToRgbHex, findConstructMarkerImbalance } from "document-schema.js";
+import { colorToRgbHex, findConstructMarkerImbalance, findRunConstructFault } from "document-schema.js";
 //#region src/typed/docx/write.ts
 const WML_NS = "http://schemas.openxmlformats.org/wordprocessingml/2006/main";
 const REL_NS = "http://schemas.openxmlformats.org/officeDocument/2006/relationships";
@@ -178,10 +178,49 @@ function buildParagraph(paragraph, state, pageBreakBefore, deleted, provenance)
 	const changeTag = provenance === void 0 ? void 0 : TRACKED_CHANGE_TAG_BY_CHANGE[provenance.change];
 	const pPr = properties === void 0 && changeTag !== void 0 ? el("w:pPr", {}, []) : properties;
 	if (pPr !== void 0 && changeTag !== void 0 && provenance !== void 0) pPr.children.push(el("w:rPr", {}, [el(changeTag, trackChangeAttrs(state, provenance))]));
-	const runs = paragraph.runs.map((run) => buildRun(run, state, deleted));
+	const runs = interleaveRunConstructExtents(paragraph.runs.map((run) => buildRun(run, state, deleted)), paragraph, state);
 	const content = changeTag === void 0 || provenance === void 0 ? runs : [el(changeTag, trackChangeAttrs(state, provenance), runs)];
 	return el("w:p", {}, [...pPr === void 0 ? [] : [pPr], ...content]);
 }
+function interleaveRunConstructExtents(runElements, paragraph, state) {
+	if (paragraph.constructs === void 0) return [...runElements];
+	const fault = findRunConstructFault(paragraph);
+	if (fault !== void 0) throw new Error(`buildDocxPackageFromContent: run-level construct extent at index ${String(fault.index)} of a paragraph does not name real runs (${fault.kind})`);
+	const bookmarks = paragraph.constructs.filter((extent) => extent.descriptor.kind === "anchor" && extent.descriptor.anchorType === "bookmark");
+	if (bookmarks.length === 0) return [...runElements];
+	const closingAt = /* @__PURE__ */ new Map();
+	const openingAt = /* @__PURE__ */ new Map();
+	const pointAt = /* @__PURE__ */ new Map();
+	const push = (map, position, element) => {
+		const existing = map.get(position);
+		if (existing === void 0) map.set(position, [element]);
+		else existing.push(element);
+	};
+	for (const bookmark of bookmarks) {
+		const id = String(state.nextMarkerId++);
+		const open = el("w:bookmarkStart", {
+			"w:id": id,
+			"w:name": encodeXmlText(bookmark.descriptor.name)
+		});
+		const close = el("w:bookmarkEnd", { "w:id": id });
+		if (bookmark.startRun === bookmark.endRun) {
+			push(pointAt, bookmark.startRun, open);
+			push(pointAt, bookmark.startRun, close);
+		} else {
+			push(openingAt, bookmark.startRun, open);
+			push(closingAt, bookmark.endRun, close);
+		}
+	}
+	const out = [];
+	for (let position = 0; position <= runElements.length; position++) {
+		for (const close of closingAt.get(position) ?? []) out.push(close);
+		for (const open of openingAt.get(position) ?? []) out.push(open);
+		for (const half of pointAt.get(position) ?? []) out.push(half);
+		const run = runElements[position];
+		if (run !== void 0) out.push(run);
+	}
+	return out;
+}
 function trailingEmptyRunElements(paragraph, element) {
 	const runElements = [];
 	for (const child of element.children) if (child.type === "element" && (child.tag === "w:r" || child.tag === "w:hyperlink")) runElements.push(child);

package/dist/typed/embedded.cjs

@@ -1,4 +1,4 @@
 Object.defineProperty(exports, Symbol.toStringTag, { value: "Module" });
-const require_read = require("../read-C3AaxcT9.cjs");
+const require_read = require("../read-Di5UfiSh.cjs");
 exports.readEmbeddedOoxmlPayload = require_read.readEmbeddedOoxmlPayload;
 exports.readEmbeddedPayloadPart = require_read.readEmbeddedPayloadPart;

package/dist/typed/embedded.js

@@ -1,2 +1,2 @@
-import { a as readEmbeddedOoxmlPayload, o as readEmbeddedPayloadPart } from "../read-fuJpC4BH.js";
+import { a as readEmbeddedOoxmlPayload, o as readEmbeddedPayloadPart } from "../read-BQoFiqsb.js";
 export { readEmbeddedOoxmlPayload, readEmbeddedPayloadPart };

package/dist/typed/pptx/read.cjs

@@ -1,4 +1,4 @@
 Object.defineProperty(exports, Symbol.toStringTag, { value: "Module" });
-const require_read = require("../../read-C3AaxcT9.cjs");
+const require_read = require("../../read-Di5UfiSh.cjs");
 exports.PptxDocumentSchema = require_read.PptxDocumentSchema;
 exports.readPptxContent = require_read.readPptxContent;

package/dist/typed/pptx/read.js

@@ -1,2 +1,2 @@
-import { c as readPptxContent, s as PptxDocumentSchema } from "../../read-fuJpC4BH.js";
+import { c as readPptxContent, s as PptxDocumentSchema } from "../../read-BQoFiqsb.js";
 export { PptxDocumentSchema, readPptxContent };

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "ooxml.js",
-  "version": "4.2.0",
+  "version": "4.3.0",
   "description": "Type-safe, lossless round-trip conversion between OOXML packages (docx, pptx, xlsx) and JSON, built on Zod 4 codecs.",
   "type": "module",
   "repository": {
@@ -81,7 +81,7 @@
   "packageManager": "pnpm@11.6.0",
   "dependencies": {
     "archive-codec": "^1.1.0",
-    "document-schema.js": "^4.3.7",
+    "document-schema.js": "^4.5.0",
     "fast-xml-parser": "^5.10.1",
     "fflate": "^0.8.3",
     "zod": "^4.4.3"