reamkit 1.33.2 → 1.34.1

package/dist/esm/pdf-reader/text.js

@@ -1,4 +1,5 @@
-import { PDF_NULL, PdfName, PdfStream } from "../pdf/objects.js";
+import { PDF_NULL, PdfHexString, PdfName, PdfStream } from "../pdf/objects.js";
+import { pdfTextString } from "./encodings.js";
 import { buildColorSpaceMap, buildShadingMap } from "./shading.js";
 import { IDENTITY, interpretContent, multiply } from "./content.js";
 import { textMarkupOf } from "./annot-draw.js";
@@ -275,8 +276,29 @@ function spacesOf(file, resources) {
 	spaceCache.set(resources, made);
 	return made;
 }
+/**
+* §14.9.4 — the replacement text of each property list the page's
+* `/Properties` names: the content says `/Span /P1 BDC` and the `/ActualText`
+* is in the resources, not in the stream.
+*
+* @param file      The owning file.
+* @param resources The resource dictionary in force.
+* @returns Name → the text, for the lists that hold one.
+*/
+function actualTextsOf(file, resources) {
+	const out = /* @__PURE__ */ new Map();
+	const props = resources ? file.get(resources, "Properties") : void 0;
+	if (!(props instanceof Map)) return out;
+	for (const [name, value] of props) {
+		const list = file.resolve(value);
+		const text = list instanceof Map ? file.get(list, "ActualText") : void 0;
+		if (text instanceof PdfHexString) out.set(name, pdfTextString(text.bytes));
+		else if (typeof text === "string") out.set(name, pdfTextString(Uint8Array.from(text, (c) => c.charCodeAt(0) & 255)));
+	}
+	return out;
+}
 function collectRuns(file, resources, content, baseCtm, depth, visiting, out, shown) {
-	const result = interpretContent(content, buildFonts(file, resources), baseCtm, shadingsOf(file, resources), void 0, spacesOf(file, resources), hiddenProperties(file, resources));
+	const result = interpretContent(content, buildFonts(file, resources), baseCtm, shadingsOf(file, resources), void 0, spacesOf(file, resources), hiddenProperties(file, resources), actualTextsOf(file, resources));
 	out.push(...result.texts.map((r) => withPatternColour(file, resources, r, visiting)));
 	if (shown) addShown(shown, result.shown);
 	if (depth >= MAX_FORM_DEPTH) return;
@@ -411,4 +433,4 @@ function inRect(x, y, r) {
 	return x >= r[0] - 1 && x <= r[2] + 1 && y >= r[1] - 2 && y <= r[3] + 2;
 }
 //#endregion
-export { buildFonts, extractPageText };
+export { actualTextsOf, buildFonts, extractPageText };

package/dist/esm/pdf-reader/vector.js

@@ -4,7 +4,7 @@ import { buildAlphaMap, buildColorSpaceMap, buildShadingMap, gradientShading, sh
 import { IDENTITY, interpretContent, multiply } from "./content.js";
 import { appearanceContent, collectPageAppearances } from "./annots.js";
 import { hiddenProperties, hiddenXObject } from "./optional-content.js";
-import { buildFonts } from "./text.js";
+import { actualTextsOf, buildFonts } from "./text.js";
 import { patternTint, tintedHex } from "./pattern-tint.js";
 //#region src/pdf-reader/vector.ts
 function paintedVectors(file, page) {
@@ -53,7 +53,7 @@ function paintedVectors(file, page) {
 		const xobjects = resources ? file.get(resources, "XObject") : PDF_NULL;
 		const xobjDict = xobjects instanceof Map ? xobjects : void 0;
 		const maps = mapsOf(resources);
-		const result = interpretContent(content, buildFonts(file, resources), baseCtm, maps.shadings, maps.alphas, maps.spaces, hiddenProperties(file, resources));
+		const result = interpretContent(content, buildFonts(file, resources), baseCtm, maps.shadings, maps.alphas, maps.spaces, hiddenProperties(file, resources), actualTextsOf(file, resources));
 		for (const paint of result.shadings) {
 			if (out.length >= MAX_VECTORS) break;
 			const dict = resources ? file.get(resources, "Shading") : PDF_NULL;

package/dist/esm/pptx/embedded-fonts.js

@@ -1,4 +1,5 @@
 import { FontRegistry } from "../core/font/font-registry.js";
+import "../core/font/index.js";
 import { poAttr, poChildren, poIs } from "../core/po-helpers.js";
 //#region src/pptx/embedded-fonts.ts
 var REL_FONT = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/font";

package/dist/esm/pptx/sp-helpers.js

@@ -85,7 +85,10 @@ function rPrToRunProps(rPr, colors, themeFonts) {
 		...u !== void 0 ? { underline: u === "none" ? "none" : "single" } : {},
 		...sz !== void 0 ? { fontSizePt: pt(sz / 100) } : {},
 		...colorHex ? { colorHex } : {},
-		...typeface ? { fontFamily: { ascii: typeface } } : {},
+		...typeface ? { fontFamily: {
+			ascii: typeface,
+			hAnsi: typeface
+		} } : {},
 		...capsOf(poAttr(rPr, "cap"))
 	};
 }

package/dist/esm/svg/svg-text.js

@@ -1,7 +1,7 @@
 import { svgPathData } from "../core/vector.js";
 import { reverseByCodePoint } from "../core/bidi/index.js";
 import { FAUX_ITALIC_SHEAR, computeJustifyExtra, lineVisualOrder } from "../layout/line-paint.js";
-import { warpGlyphMatrix } from "../core/drawingml/text-warp.js";
+import { warpGlyphMatrix, warpGlyphRun } from "../core/drawingml/text-warp.js";
 import { sanitizeHref } from "../core/links.js";
 import { cffOutlineSource, openTypeCff } from "../pdf-reader/cff-outline.js";
 import { outlineSource } from "../pdf-reader/glyf-outline.js";
@@ -128,10 +128,16 @@ function emitTextToken(out, tok, x, baseline, ctx) {
 	const scale = (tok.synthetic?.widthScale ?? 1) * (tok.resolvedRun.widthScale ?? 1);
 	const slant = tok.synthetic?.italic === true ? FAUX_ITALIC_SHEAR : 0;
 	const rise = tok.risePt ?? 0;
-	const text = tok.bidiLevel % 2 === 1 ? reverseByCodePoint(tok.text) : tok.text;
+	const reversed = tok.bidiLevel % 2 === 1;
+	const text = reversed ? reverseByCodePoint(tok.text) : tok.text;
 	const run = tok.font.measure.glyphRun(text);
 	const upem = tok.font.parsed.unitsPerEm;
 	const spacingEm = (tok.resolvedRun.letterSpacingPt ?? 0) / (size * scale);
+	const outline = tok.resolvedRun.textOutline;
+	const strokeWidth = outline !== void 0 ? outline.widthPt : tok.synthetic?.bold === true ? size * .03 : 0;
+	const colours = reversed ? tok.charColorsHex?.slice().reverse() : tok.charColorsHex;
+	const sources = colours ? tok.font.measure.glyphSources(text) : void 0;
+	const paint = (hex) => hex === void 0 || hex === tok.resolvedRun.colorHex ? "" : ` fill="#${hex}"${strokeWidth > 0 && outline === void 0 ? ` stroke="#${hex}"` : ""}`;
 	const uses = [];
 	let pen = 0;
 	for (let g = 0; g < run.gids.length; g++) {
@@ -140,12 +146,16 @@ function emitTextToken(out, tok, x, baseline, ctx) {
 			emitAsText(out, tok, x, baseline);
 			return;
 		}
-		if (id !== null) uses.push(`<use href="#${id}"${pen !== 0 ? ` x="${fmtEm(pen)}"` : ""}/>`);
-		pen += (run.advances[g] ?? 0) / upem + spacingEm;
+		const gx = pen + (run.xOffsets?.[g] ?? 0) / upem;
+		const gy = (run.yOffsets?.[g] ?? 0) / upem;
+		if (id !== null) {
+			const fill = paint(sources ? colours?.[sources[g]] : void 0);
+			uses.push(`<use href="#${id}"${gx !== 0 ? ` x="${fmtEm(gx)}"` : ""}${gy !== 0 ? ` y="${fmtEm(gy)}"` : ""}${fill}/>`);
+		}
+		const spaced = run.clusterEnds ? run.clusterEnds[g] === true : true;
+		pen += (run.advances[g] ?? 0) / upem + (spaced ? spacingEm : 0);
 	}
 	if (uses.length === 0) return;
-	const outline = tok.resolvedRun.textOutline;
-	const strokeWidth = outline !== void 0 ? outline.widthPt : tok.synthetic?.bold === true ? size * .03 : 0;
 	const stroke = strokeWidth > 0 ? ` stroke="#${outline?.colorHex ?? tok.resolvedRun.colorHex}" stroke-width="${fmtEm(strokeWidth / size)}"` : "";
 	const group = `<g transform="matrix(${fmt(size * scale)} 0 ${fmt(slant * size)} ${fmt(-size)} ${fmt(x)} ${fmt(baseline - rise)})" fill="#${tok.resolvedRun.colorHex}"${stroke}>${uses.join("")}</g>`;
 	const href = tok.href !== void 0 ? sanitizeHref(tok.href) : void 0;
@@ -282,6 +292,19 @@ function emitWarped(out, item, ctx) {
 			continue;
 		}
 		const size = tok.fontSizePt;
+		const group = (m, uses, hex = tok.resolvedRun.colorHex) => `<g transform="matrix(${fmt(m[0] * size)} ${fmt(m[1] * size)} ${fmt(m[2] * size)} ${fmt(m[3] * size)} ${fmt(m[4])} ${fmt(m[5])})" fill="#${hex}">${uses}</g>`;
+		if (tok.font.measure.clustered(tok.text)) {
+			const glyphs = warpGlyphRun(warp, tok.font.measure.glyphRun(tok.text), cursor, tok.widthPt, size / tok.font.parsed.unitsPerEm, item.baselineY);
+			const colours = tok.charColorsHex;
+			const sources = colours ? tok.font.measure.glyphSources(tok.text) : void 0;
+			for (const g of glyphs) {
+				const id = ctx.glyphs.use(tok.font, g.gid);
+				const hex = sources ? colours?.[sources[g.index]] : void 0;
+				if (id) out.push(group(g.matrix, `<use href="#${id}"/>`, hex));
+			}
+			cursor += tok.widthPt;
+			continue;
+		}
 		const chars = [...tok.text];
 		const raw = chars.map((c) => tok.font.measure.textWidthPt(c, size));
 		const total = raw.reduce((a, b) => a + b, 0);
@@ -301,7 +324,7 @@ function emitWarped(out, item, ctx) {
 				pen += (run.advances[g] ?? 0) / upem;
 			}
 			if (uses.length === 0) continue;
-			out.push(`<g transform="matrix(${fmt(m[0] * size)} ${fmt(m[1] * size)} ${fmt(m[2] * size)} ${fmt(m[3] * size)} ${fmt(m[4])} ${fmt(m[5])})" fill="#${tok.resolvedRun.colorHex}">${uses.join("")}</g>`);
+			out.push(group(m, uses.join("")));
 		}
 	}
 }

package/dist/esm/word/doc/doc-reader.js

@@ -27,6 +27,11 @@ var TABLE_BORDERS = {
 	insideH: TABLE_BORDER,
 	insideV: TABLE_BORDER
 };
+var LVL_FOLLOW = [
+	"	",
+	" ",
+	""
+];
 function createListMarkers(tables) {
 	const counters = /* @__PURE__ */ new Map();
 	const resolve = (ilfo, ilvl) => {
@@ -43,7 +48,8 @@ function createListMarkers(tables) {
 		const r = resolve(ilfo, ilvl);
 		if (!r) return void 0;
 		const { lsid, lvl } = r;
-		if (lvl.nfc === 23 || lvl.nfc === 255 || !lvl.xst.some((cu) => cu <= 8)) return lvl.xst.map((cu) => String.fromCharCode(cu)).join("");
+		const follow = LVL_FOLLOW[lvl.follow] ?? "	";
+		if (lvl.nfc === 23 || lvl.nfc === 255 || !lvl.xst.some((cu) => cu <= 8)) return lvl.xst.map((cu) => String.fromCharCode(cu)).join("") + follow;
 		const col = counters.get(lsid) ?? [];
 		col[ilvl] = (col[ilvl] ?? lvl.iStartAt - 1) + 1;
 		for (let l = ilvl + 1; l < col.length; l++) col[l] = void 0;
@@ -53,7 +59,7 @@ function createListMarkers(tables) {
 			const other = resolve(ilfo, cu);
 			out += formatListNumber(other?.lvl.nfc ?? 0, col[cu] ?? other?.lvl.iStartAt ?? 1);
 		} else out += String.fromCharCode(cu);
-		return out;
+		return out + follow;
 	} };
 }
 function formatListNumber(nfc, n) {
@@ -128,54 +134,79 @@ var DOC_ENCRYPTED_LOSS = {
 function readDoc(bytes) {
 	const content = extractDocContent(bytes);
 	const resources = new ResourceStore();
-	const body = buildBody(content, resources);
+	const { body, sectionEnds } = buildBody(content, resources);
 	const headersFooters = /* @__PURE__ */ new Map();
-	const headers = [];
-	const footers = [];
-	const addStory = (story, kind, type) => {
-		if (!story) return;
-		const id = `${kind}-${type}`;
-		const els = resolveBodyStyles(story.flatMap((p) => mapParagraph(p, resources)), EMPTY_STYLE_SHEET);
-		headersFooters.set(id, els);
-		(kind === "header" ? headers : footers).push({
-			type,
-			relationshipId: id
-		});
+	const bandIds = /* @__PURE__ */ new Map();
+	const bands = (index) => {
+		const headers = [];
+		const footers = [];
+		const hf = content.headerFooters?.[index];
+		const add = (story, kind, type) => {
+			if (!story) return;
+			let id = bandIds.get(story);
+			if (id === void 0) {
+				id = `${kind}-${type}-${index}`;
+				bandIds.set(story, id);
+				headersFooters.set(id, resolveBodyStyles(story.flatMap((p) => mapParagraph(p, resources)), EMPTY_STYLE_SHEET));
+			}
+			(kind === "header" ? headers : footers).push({
+				type,
+				relationshipId: id
+			});
+		};
+		add(hf?.defaultHeader, "header", "default");
+		add(hf?.firstHeader, "header", "first");
+		add(hf?.evenHeader, "header", "even");
+		add(hf?.defaultFooter, "footer", "default");
+		add(hf?.firstFooter, "footer", "first");
+		add(hf?.evenFooter, "footer", "even");
+		return {
+			headers,
+			footers
+		};
 	};
-	const hf = content.headerFooters;
-	if (hf) {
-		addStory(hf.defaultHeader, "header", "default");
-		addStory(hf.firstHeader, "header", "first");
-		addStory(hf.evenHeader, "header", "even");
-		addStory(hf.defaultFooter, "footer", "default");
-		addStory(hf.firstFooter, "footer", "first");
-		addStory(hf.evenFooter, "footer", "even");
-	}
-	const ps = content.pageSize;
-	const pageSize = ps ? {
-		width: pt(ps.widthPt),
-		height: pt(ps.heightPt),
-		orientation: ps.widthPt > ps.heightPt ? "landscape" : "portrait"
-	} : {
-		width: pt(612),
-		height: pt(792)
+	const sectionProperties = (sec, index) => {
+		const ps = sec?.pageSize;
+		const sectionStart = sec?.bkc !== void 0 ? sectionStartOf(sec.bkc) : void 0;
+		const pageNumberFormat = sec?.pageNumberNfc !== void 0 ? pageNumberFormatOf(sec.pageNumberNfc) : void 0;
+		return {
+			pageSize: ps ? {
+				width: pt(ps.widthPt),
+				height: pt(ps.heightPt),
+				orientation: ps.widthPt > ps.heightPt ? "landscape" : "portrait"
+			} : {
+				width: pt(612),
+				height: pt(792)
+			},
+			margins: toMargins(sec?.margins),
+			...bands(index),
+			...sec?.titlePage ? { titlePg: true } : {},
+			...content.facingPages ? { evenAndOddHeaders: true } : {},
+			...sectionStart ? { sectionStart } : {},
+			...sec?.pageNumberStart !== void 0 ? { pageNumberStart: sec.pageNumberStart } : {},
+			...pageNumberFormat ? { pageNumberFormat } : {}
+		};
 	};
+	const docSections = content.sections ?? [];
+	const sections = [];
+	for (const end of sectionEnds) {
+		const last = sections[sections.length - 1]?.endIndex ?? 0;
+		if (end.endIndex <= last || end.endIndex >= body.length) continue;
+		sections.push({
+			properties: sectionProperties(docSections[end.section], end.section),
+			endIndex: end.endIndex
+		});
+	}
+	if (sections.length > 0) sections.push({
+		properties: sectionProperties(docSections[docSections.length - 1], docSections.length - 1),
+		endIndex: body.length
+	});
 	return {
 		doc: {
 			kind: "flow",
 			body: resolveBodyStyles(body, EMPTY_STYLE_SHEET),
-			sections: [],
-			section: {
-				pageSize,
-				margins: {
-					top: pt(72),
-					right: pt(72),
-					bottom: pt(72),
-					left: pt(72)
-				},
-				headers,
-				footers
-			},
+			sections,
+			section: sectionProperties(docSections[0], 0),
 			styles: EMPTY_STYLE_SHEET,
 			resources,
 			...headersFooters.size > 0 ? { headersFooters } : {},
@@ -186,12 +217,16 @@ function readDoc(bytes) {
 }
 function buildBody(content, resources) {
 	const markers = createListMarkers(content.listTables);
-	const body = groupTables(content.paragraphs, resources, markers);
+	const sectionEnds = [];
+	const body = groupTables(content.paragraphs, resources, markers, sectionEnds);
 	while (body.length > 1 && isEmptyParagraph(body[body.length - 1])) body.pop();
 	if (body.length === 0) body.push(emptyParagraph());
-	return body;
+	return {
+		body,
+		sectionEnds
+	};
 }
-function groupTables(paras, resources, markers) {
+function groupTables(paras, resources, markers, sectionEnds) {
 	const out = [];
 	let i = 0;
 	while (i < paras.length) if (paras[i].props.inTable) {
@@ -201,7 +236,14 @@ function groupTables(paras, resources, markers) {
 			kind: "table",
 			table: buildTable(group, resources, markers)
 		});
-	} else out.push(...mapParagraph(paras[i++], resources, markers));
+	} else {
+		const p = paras[i++];
+		out.push(...mapParagraph(p, resources, markers));
+		if (p.sectionEnd !== void 0) sectionEnds.push({
+			section: p.sectionEnd,
+			endIndex: out.length
+		});
+	}
 	return out;
 }
 function buildTable(group, resources, markers) {
@@ -331,16 +373,24 @@ function mapParagraph(p, resources, markers) {
 	const textRuns = p.runs.filter((r) => r.picture === void 0);
 	const ilvl = p.props.listIlvl ?? 0;
 	const isList = (p.props.listIlfo ?? 0) > 0;
-	if (textRuns.some((r) => r.text.length > 0) || isList) {
+	if (textRuns.some((r) => r.text.length > 0 || r.pageBreak || r.field) || isList) {
 		const runs = textRuns.map((r) => ({
 			text: r.text,
-			properties: toRunProperties(r.props)
+			properties: toRunProperties(r.props),
+			...r.pageBreak ? { pageBreak: true } : {},
+			...r.columnBreak ? { columnBreak: true } : {},
+			...r.field ? { field: r.field } : {}
 		}));
+		const last = textRuns[textRuns.length - 1];
+		if (last?.pageBreak || last?.columnBreak) runs.push({
+			text: ZERO_WIDTH_SPACE,
+			properties: p.mark ? toRunProperties(p.mark) : {}
+		});
 		let properties = toParaProperties(p.props);
 		if (isList) {
-			const marker = markers?.marker(p.props.listIlfo, ilvl) ?? LIST_BULLETS[ilvl % LIST_BULLETS.length];
+			const marker = markers?.marker(p.props.listIlfo, ilvl) ?? `${LIST_BULLETS[ilvl % LIST_BULLETS.length]} `;
 			runs.unshift({
-				text: `${marker} `,
+				text: marker,
 				properties: {}
 			});
 			properties = {
@@ -360,7 +410,7 @@ function mapParagraph(p, resources, markers) {
 		});
 	}
 	for (const r of p.runs) if (r.picture) out.push(imageBlock(r.picture, resources, toParaProperties(p.props)));
-	if (out.length === 0) out.push(emptyParagraph());
+	if (out.length === 0) out.push(emptyParagraph(toParaProperties(p.props), p.mark));
 	return out;
 }
 function imageBlock(pic, resources, paragraphProperties) {
@@ -374,14 +424,14 @@ function imageBlock(pic, resources, paragraphProperties) {
 		}
 	};
 }
-function emptyParagraph() {
+function emptyParagraph(properties = {}, mark = void 0) {
 	return {
 		kind: "paragraph",
 		paragraph: {
-			properties: {},
+			properties,
 			runs: [{
 				text: ZERO_WIDTH_SPACE,
-				properties: {}
+				properties: mark ? toRunProperties(mark) : {}
 			}]
 		}
 	};
@@ -392,13 +442,24 @@ function isEmptyParagraph(el) {
 function toRunProperties(p) {
 	const underline = p.underlineKul ? underlineStyle(p.underlineKul) : void 0;
 	const size = p.sizeHalfPts && p.sizeHalfPts > 0 ? pt(p.sizeHalfPts / 2) : void 0;
+	const fontFamily = toFontFamily(p);
 	return {
 		...p.bold ? { bold: true } : {},
 		...p.italic ? { italic: true } : {},
 		...underline ? { underline } : {},
-		...size ? { fontSizePt: size } : {}
+		...size ? { fontSizePt: size } : {},
+		...fontFamily ? { fontFamily } : {}
 	};
 }
+function toFontFamily(p) {
+	const fonts = {
+		...p.fontAscii ? { ascii: p.fontAscii } : {},
+		...p.fontOther ? { hAnsi: p.fontOther } : {},
+		...p.fontEastAsia ? { eastAsia: p.fontEastAsia } : {},
+		...p.fontBi ? { cs: p.fontBi } : {}
+	};
+	return Object.keys(fonts).length > 0 ? fonts : void 0;
+}
 function toParaProperties(p) {
 	const alignment = p.jc !== void 0 ? alignmentFromJc(p.jc) : void 0;
 	return {
@@ -410,7 +471,8 @@ function toParaProperties(p) {
 		...p.spaceAfterTwips !== void 0 ? { spacingAfter: pt(p.spaceAfterTwips / 20) } : {},
 		...p.widowControl !== void 0 ? { widowControl: p.widowControl } : {},
 		...p.keepNext !== void 0 ? { keepNext: p.keepNext } : {},
-		...p.keepLines !== void 0 ? { keepLines: p.keepLines } : {}
+		...p.keepLines !== void 0 ? { keepLines: p.keepLines } : {},
+		...p.pageBreakBefore !== void 0 ? { pageBreakBefore: p.pageBreakBefore } : {}
 	};
 }
 function underlineStyle(kul) {
@@ -422,6 +484,35 @@ function underlineStyle(kul) {
 		default: return "single";
 	}
 }
+function toMargins(m) {
+	const side = (twips) => twipsToPt(twips ?? 1440);
+	return {
+		top: side(m?.topTwips),
+		right: side(m?.rightTwips),
+		bottom: side(m?.bottomTwips),
+		left: side(m?.leftTwips),
+		...m?.headerTwips !== void 0 ? { header: twipsToPt(m.headerTwips) } : {},
+		...m?.footerTwips !== void 0 ? { footer: twipsToPt(m.footerTwips) } : {},
+		...m?.gutterTwips ? { gutter: twipsToPt(m.gutterTwips) } : {}
+	};
+}
+function sectionStartOf(bkc) {
+	switch (bkc) {
+		case 0: return "continuous";
+		case 3: return "evenPage";
+		case 4: return "oddPage";
+		default: return "nextPage";
+	}
+}
+function pageNumberFormatOf(nfc) {
+	switch (nfc) {
+		case 1: return "upperRoman";
+		case 2: return "lowerRoman";
+		case 3: return "upperLetter";
+		case 4: return "lowerLetter";
+		default: return;
+	}
+}
 function alignmentFromJc(jc) {
 	switch (jc) {
 		case 1: return "center";

package/dist/esm/word/doc/doc-text.d.ts

@@ -6,6 +6,14 @@ export interface DocCharProps {
     readonly underlineKul?: number;
     /** Font size in half-points. */
     readonly sizeHalfPts?: number;
+    /** The font of ASCII characters, by name — `sprmCRgFtc0`. */
+    readonly fontAscii?: string;
+    /** The font of East Asian characters — `sprmCRgFtc1`. */
+    readonly fontEastAsia?: string;
+    /** The font of the characters neither ASCII nor East Asian — `sprmCRgFtc2`. */
+    readonly fontOther?: string;
+    /** The font of complex-script characters — `sprmCFtcBi`. */
+    readonly fontBi?: string;
     /** The character is special — `sprmCFSpec` (e.g. a picture placeholder). */
     readonly fSpec?: boolean;
     /** `sprmCPicLocation` — offset into the `Data` stream of the picture's PICF. */
@@ -80,6 +88,8 @@ export interface DocParaProps {
     readonly keepNext?: boolean;
     /** `sprmPFKeep` — every line on one page. */
     readonly keepLines?: boolean;
+    /** `sprmPFPageBreakBefore` — the paragraph starts a new page. */
+    readonly pageBreakBefore?: boolean;
 }
 /** One maximal same-formatting run of text, or an embedded picture. */
 export interface DocRun {
@@ -87,6 +97,15 @@ export interface DocRun {
     readonly props: DocCharProps;
     /** Set instead of `text` when the run is an embedded picture. */
     readonly picture?: DocPicture;
+    /** The run is a manual page break (0x0C), alone in a run with no text. */
+    readonly pageBreak?: boolean;
+    /** The run is a manual column break (0x0E), alone in a run whose text is `\n`. */
+    readonly columnBreak?: boolean;
+    /**
+     * The run is the result of a PAGE or NUMPAGES field, whose `text` is the
+     * number Word cached when it last saved.
+     */
+    readonly field?: 'PAGE' | 'NUMPAGES';
 }
 /** One paragraph: its runs plus the decoded paragraph properties. */
 export interface DocParagraph {
@@ -97,11 +116,59 @@ export interface DocParagraph {
      * closes a table cell. With `props.rowEnd` it also closes the row.
      */
     readonly cellMark?: boolean;
+    /**
+     * The paragraph was ended by the section mark of the (zero-based) section it
+     * is the last of — the index into {@link DocContent.sections}.
+     */
+    readonly sectionEnd?: number;
+    /** The paragraph mark's own character properties: what an empty line is set in. */
+    readonly mark?: DocCharProps;
+}
+/** One section of the main document (§2.8.26 PlcfSed): what its SEP says. */
+export interface DocSection {
+    /**
+     * `sprmSBkc` — where the section starts: 0 on the page in hand, 1 in the next
+     * column, 2 on a new page, 3 on an even page, 4 on an odd page. Absent when
+     * the SEP says nothing, which is a new page.
+     */
+    readonly bkc?: number;
+    /** The page size in points (xaPage/yaPage), when present and sane. */
+    readonly pageSize?: {
+        readonly widthPt: number;
+        readonly heightPt: number;
+    };
+    /** `sprmSFTitlePage` — the first page shows the first-page header and footer. */
+    readonly titlePage?: boolean;
+    /** The margins the SEP states, in twips; each absent one is Word's default. */
+    readonly margins?: DocMargins;
+    /**
+     * `sprmSFPgnRestart` with `sprmSPgnStart97` — the number the section's first
+     * page is printed with, when the section starts its count again; absent, the
+     * count runs on from the section before.
+     */
+    readonly pageNumberStart?: number;
+    /** `sprmSNfcPgn` — the MSONFC the section's page numbers print in; absent, arabic. */
+    readonly pageNumberNfc?: number;
+}
+/**
+ * A section's margins in twips (§2.6.4 sprmSDxaLeft … sprmSDzaGutter). A top or
+ * bottom margin is stored signed, negative for one the header or footer may not
+ * push; its size is the same either way.
+ */
+export interface DocMargins {
+    readonly leftTwips?: number;
+    readonly rightTwips?: number;
+    readonly topTwips?: number;
+    readonly bottomTwips?: number;
+    readonly headerTwips?: number;
+    readonly footerTwips?: number;
+    readonly gutterTwips?: number;
 }
 /**
- * The section's header/footer stories (best-effort: the binary story ordering
- * cannot be validated against a real file, so only well-formed, non-empty stories
- * are surfaced). `default` is the odd-page (primary) header/footer.
+ * One section's header/footer stories ([MS-DOC] Plcfhdd: six per section, after
+ * the six separator stories). `default` is the odd-page (primary) one. A story
+ * the section leaves empty is the previous section's, the very same array; one
+ * that holds no text is absent.
  */
 export interface DocHeaderFooters {
     readonly defaultHeader?: ReadonlyArray<DocParagraph>;
@@ -121,6 +188,8 @@ export interface DocLvl {
     readonly nfc: number;
     readonly iStartAt: number;
     readonly xst: ReadonlyArray<number>;
+    /** `ixchFollow` — what follows the number text: 0 a tab, 1 a space, 2 nothing. */
+    readonly follow: number;
 }
 /**
  * The document's list tables (DOC-10): an LFO → lsid map (a paragraph's 1-based
@@ -136,23 +205,23 @@ export interface DocListTables {
 /** Everything extracted from a `.doc`'s `WordDocument` stream. */
 export interface DocContent {
     readonly paragraphs: ReadonlyArray<DocParagraph>;
-    readonly headerFooters?: DocHeaderFooters;
+    /** Each section's header/footer stories, in section order. */
+    readonly headerFooters?: ReadonlyArray<DocHeaderFooters>;
+    /** `Dop.fFacingPages` — even pages show the even header and footer. */
+    readonly facingPages?: boolean;
     /** The list tables resolving a paragraph's (ilfo, ilvl) to its number format. */
     readonly listTables?: DocListTables;
     /**
-     * The first section's page size in points (xaPage/yaPage from its SEP), when
-     * present and sane; absent ⇒ the reader's default (US Letter) applies.
+     * Every section of the main document, in order, when the PlcfSed is usable. A
+     * section with no page size takes the reader's default (US Letter).
      */
-    readonly pageSize?: {
-        readonly widthPt: number;
-        readonly heightPt: number;
-    };
+    readonly sections?: ReadonlyArray<DocSection>;
     /** The file is encrypted/obfuscated — text cannot be read without the key. */
     readonly encrypted: boolean;
 }
 /**
  * Extract a `.doc`'s `WordDocument` stream into its main-document paragraphs plus
- * the section's header/footer stories, list tables and first-section page size.
+ * the section's header/footer stories, list tables and sections.
  * Joins the piece table (CLX → CPs/FCs), CHPX run formatting and PAPX paragraph
  * formatting, splitting at the CR / cell marks. Every offset is bounds-checked, so
  * a structurally odd file yields no text rather than throwing.