reamkit 1.33.2 → 1.34.1

package/README.md

@@ -31,8 +31,8 @@ format is sniffed from the bytes; no fonts to wire up — an open
 metric-compatible substitute font (Arimo for sans, Tinos for serif, Cousine for
 monospace, plus Carlito/Caladea for Calibri/Cambria — the same families LibreOffice
 substitutes) is fetched automatically based on the document's referenced fonts,
-with a Noto face for the Japanese, Korean, Chinese, Arabic, Hebrew or Thai text
-the document holds:
+with a Noto face for the Japanese, Korean, Chinese, Arabic, Hebrew, Thai or
+Indic (Hindi, Bengali, Tamil and six more scripts) text the document holds:
 
 ```ts
 import { Ream } from 'reamkit';
@@ -196,11 +196,12 @@ bars, icon sets, and `expression` rules evaluated by a ~140-function formula
 engine), **sparklines** and **Excel tables**, DrawingML shapes and
 charts, OMML math, Type0+CIDFontType2 embedding with subsetting, Knuth-Plass
 line breaking, Liang hyphenation, OpenType ligatures/kerning + Arabic cursive
-joining, BiDi (UAX #9), hyperlinks (PDF link annotations + HTML anchors,
-scheme-allowlisted), tagged PDF, PDF/A-1/2/3 (a/b/u), PDF/UA-1, AES-256
-encryption, digital signatures (PKCS#7/ECDSA/PAdES/RFC 3161), SVG page
-preview, flowed HTML export, and **docx + xlsx output** (write WordprocessingML
-/ SpreadsheetML back out, incl. round-trips). Reads OOXML Transitional and Strict.
+joining + Indic shaping (GSUB/GPOS, as HarfBuzz shapes), BiDi (UAX #9),
+hyperlinks (PDF link annotations + HTML anchors, scheme-allowlisted), tagged
+PDF, PDF/A-1/2/3 (a/b/u), PDF/UA-1, AES-256 encryption, digital signatures
+(PKCS#7/ECDSA/PAdES/RFC 3161), SVG page preview, flowed HTML export, and
+**docx + xlsx output** (write WordprocessingML / SpreadsheetML back out, incl.
+round-trips). Reads OOXML Transitional and Strict.
 
 A Word document is laid out by Word's own rules, each measured in Word: a line
 as tall as Word sets its faces, two paragraphs standing the larger of their
@@ -274,9 +275,11 @@ hyperlinks — then converts onward to PDF, SVG, HTML, Markdown or DOCX like any
 
 **Reads legacy `.doc`, `.xls` and `.ppt`, too.** The binary Word / Excel /
 PowerPoint 97–2003 formats (OLE2/CFB) parse through a shared container reader: a
-`.doc` yields its text with run and paragraph formatting, tables (with cell
-borders, vertical merges and background shading), inline images, fields,
-headers/footers and lists (numbered or bulleted, in their number format);
+`.doc` yields its text with run and paragraph formatting (through its style sheet
+and font table), page, column and line breaks, sections with their own page
+size, margins and headers/footers, page numbers, tables (with cell
+borders, vertical merges and background shading), inline images, fields
+and lists (numbered or bulleted, in their number format);
 an `.xls` yields the grid with styling, embedded
 images, charts, drawing shapes, cell hyperlinks, the page-setup print model,
 defined names (named ranges, print area, repeated titles), cell comments, data

package/dist/esm/core/converter/facade.js

@@ -1,5 +1,6 @@
 import { fetchFontSet } from "../fonts/remote-fonts.js";
 import { FontRegistry } from "../font/font-registry.js";
+import "../font/index.js";
 import { FEATURES } from "../ir/features.js";
 import { ConversionLossError } from "../ir/loss.js";
 import { layoutStyledDocument } from "../../layout/styled-layout.js";

package/dist/esm/core/converter/ream.d.ts

@@ -47,11 +47,11 @@ export interface ReamParseOptions {
 }
 /**
  * Options for {@link Ream.convert} and {@link Ream.convertWithReport}. Extends
- * the low-level {@link StyledRenderOptions} (minus the font `registry` and
- * `styles`, which Ream builds itself) with font resolution and source-touching
- * conveniences.
+ * the low-level {@link StyledRenderOptions} (minus the font `registry`, the
+ * `styles` and the `losses` sink, which Ream builds itself) with font
+ * resolution and source-touching conveniences.
  */
-export interface ReamConvertOptions extends Omit<StyledRenderOptions, 'registry' | 'styles'> {
+export interface ReamConvertOptions extends Omit<StyledRenderOptions, 'registry' | 'styles' | 'losses'> {
     /** Explicit font bytes per variant (regular/bold/italic/bold-italic). */
     readonly fonts?: FontBytesByVariant;
     /** Shorthand for supplying a single regular-variant font as raw bytes. */
@@ -197,8 +197,8 @@ export declare class Ream {
     private resolveFonts;
     /**
      * Fetch one face per writing system the document holds text in — the curated
-     * families are Latin, and Han, Kana, Hangul, Arabic and the geometric symbols
-     * are a notdef box in every one of them.
+     * families are Latin, and Han, Kana, Hangul, Arabic, the scripts of India and
+     * the geometric symbols are a notdef box in every one of them.
      *
      * Only the regular weight is fetched: Noto Sans SC is ten megabytes, and a
      * bold run in it is better stroked (see `SyntheticFace`) than downloaded four

package/dist/esm/core/converter/ream.js

@@ -2,6 +2,7 @@ import { fetchFontSet, fetchScriptFont, resolveFamilyKey } from "../fonts/remote
 import { scriptsInFlow } from "../fonts/scripts.js";
 import { FontRegistry } from "../font/font-registry.js";
 import { createFontMeasure } from "../font/measure.js";
+import "../font/index.js";
 import { ConversionLossError } from "../ir/loss.js";
 import { layoutStyledDocument } from "../../layout/styled-layout.js";
 import { signPdf } from "../../pdf/signature.js";
@@ -218,10 +219,14 @@ var Ream = class Ream {
 				registry,
 				...registriesByFamily ? { registriesByFamily } : {}
 			};
-			const laid = this.sheet ? layoutSheetImages(this.sheet, faces, sheetOptions, renderOptions) : layoutStyledDocument(flow.body, {
+			const laid = this.sheet ? layoutSheetImages(this.sheet, faces, sheetOptions, {
+				...renderOptions,
+				losses
+			}) : layoutStyledDocument(flow.body, {
 				...faces,
 				...flowRenderOptions(flow),
-				...renderOptions
+				...renderOptions,
+				losses
 			});
 			const svg = writeSvg(laid);
 			losses.push(...svg.losses);
@@ -251,7 +256,8 @@ var Ream = class Ream {
 			...info ? { info } : {},
 			...attachments.length > 0 ? { attachments } : {},
 			...signature ? { signaturePlaceholder: signature } : {},
-			...renderOptions
+			...renderOptions,
+			losses
 		};
 		let pdf = styled.encrypt ? await renderStyledPdfEncrypted(paginated.body, styled) : renderStyledPdf(paginated.body, styled);
 		if (signature) pdf = await signPdf(pdf, signature);
@@ -306,8 +312,8 @@ var Ream = class Ream {
 	}
 	/**
 	* Fetch one face per writing system the document holds text in — the curated
-	* families are Latin, and Han, Kana, Hangul, Arabic and the geometric symbols
-	* are a notdef box in every one of them.
+	* families are Latin, and Han, Kana, Hangul, Arabic, the scripts of India and
+	* the geometric symbols are a notdef box in every one of them.
 	*
 	* Only the regular weight is fetched: Noto Sans SC is ten megabytes, and a
 	* bold run in it is better stroked (see `SyntheticFace`) than downloaded four

package/dist/esm/core/document-model/index.d.ts

@@ -5,4 +5,4 @@
  *
  * @packageDocumentation
  */
-export type { AbstractNumbering, Alignment, Border, BorderStyle, BodyElement, Comment, CellBorders, CellMerge, DocumentInfo, CellMargins, CellProperties, CellShading, CellDataBar, CellIcon, CellNoteFlag, CellIconShape, CellSparkline, Chart, ChartBlock, ChartDataLabels, ChartDataPoint, ChartLabelPosition, ChartLineStyle, ChartMarker, ChartMarkerSymbol, ChartSeries, ChartSurface, ChartTextStyle, ChartTextStyles, ChartType, CustomGeometry, CustomPathCmd, DocumentModel, FontFamilyMap, HeaderFooterReference, HeaderFooterType, ImageBlock, ImageCrop, InlineImage, MathAccent, MathBar, MathDelimiter, MathEqArray, MathFraction, MathFunc, MathGroupChr, MathLimit, MathMatrix, MathNary, MathNode, MathRadical, MathRow, MathRun, MathScript, Numbering, NumberingFormat, NumberingInstance, NumberingLevel, PictureBullet, PictureOutline, NumberingReference, PageMargins, PageSize, Paragraph, ParagraphProperties, FrameProperties, RelativeSize, TabStop, Run, RunProperties, TextOutline, Ligatures, RowConditionalFormat, RowProperties, Section, FloatAnchor, SectionColumns, SectionProperties, ShapeBlock, ShapeShadow, ShapeDash, ShapeFill, ShapeFillKind, ShapeGeometry, ShapeGroupChild, ShapeLine, LineEnd, ShapeTextBody, ShapeTransform, Style, StyleSheet, StyleType, Table, TableCell, TableLook, TableOverlay, TableProperties, TableStyleCondition, TableStyleConditionType, TableStyleLayer, TableRow, UnderlineStyle, VerticalAlign, } from './types.js';
+export type { AbstractNumbering, Alignment, Border, BorderStyle, BodyElement, Comment, CellBorders, CellMerge, DocumentInfo, CellMargins, CellProperties, CellShading, CellDataBar, CellIcon, CellNoteFlag, CellIconShape, CellSparkline, Chart, ChartBlock, ChartDataLabels, ChartDataPoint, ChartLabelPosition, ChartLineStyle, ChartMarker, ChartMarkerSymbol, ChartSeries, ChartSurface, ChartTextStyle, ChartTextStyles, ChartType, CustomGeometry, CustomPathCmd, DocumentModel, FontFamilyMap, HeaderFooterReference, HeaderFooterType, ImageBlock, ImageCrop, InlineImage, MathAccent, MathBar, MathDelimiter, MathEqArray, MathFraction, MathFunc, MathGroupChr, MathLimit, MathMatrix, MathNary, MathNode, MathRadical, MathRow, MathRun, MathScript, Numbering, NumberingFormat, NumberingInstance, NumberingLevel, NumberingSuffix, PictureBullet, PictureOutline, NumberingReference, PageMargins, PageSize, Paragraph, ParagraphProperties, FrameProperties, RelativeSize, TabStop, Run, RunProperties, TextOutline, Ligatures, RowConditionalFormat, RowProperties, Section, FloatAnchor, SectionColumns, SectionProperties, ShapeBlock, ShapeShadow, ShapeDash, ShapeFill, ShapeFillKind, ShapeGeometry, ShapeGroupChild, ShapeLine, LineEnd, ShapeTextBody, ShapeTransform, Style, StyleSheet, StyleType, Table, TableCell, TableLook, TableOverlay, TableProperties, TableStyleCondition, TableStyleConditionType, TableStyleLayer, TableRow, UnderlineStyle, VerticalAlign, } from './types.js';

package/dist/esm/core/document-model/types.d.ts

@@ -495,6 +495,13 @@ export interface Run {
      * `"•"`). Tagged PDF wraps its glyphs in a Lbl structure element (§14.8.4.3.3).
      */
     readonly listMarker?: true;
+    /**
+     * §17.9.29 `w:suff="space"` — the run (a `listMarker` one, its text a single
+     * space) is the space a list level puts after its marker. Word sets it at a
+     * fixed width for the marker's size, whatever the marker's face, and never
+     * stretches it in a justified line.
+     */
+    readonly listMarkerSpace?: true;
     /** When set, the run renders this image inline in the line; `text` is ignored. */
     readonly inlineImage?: InlineImage;
     /** When set, the run is an inline OfficeMath object; `text` is ignored. */
@@ -589,9 +596,17 @@ export interface NumberingLevel {
      * numbering: "Sect I.01" becomes "Sect 1.01".
      */
     readonly isLegal?: boolean;
+    /**
+     * §17.9.29 `w:suff` — what separates the marker from the paragraph's text:
+     * a tab (the default, and what an absent value means), one space, or
+     * nothing at all.
+     */
+    readonly suffix?: NumberingSuffix;
     readonly paragraphProperties: ParagraphProperties;
     readonly runProperties: RunProperties;
 }
+/** §17.18.53 `ST_LevelSuffix` — the character between a list marker and its text. */
+export type NumberingSuffix = 'tab' | 'space' | 'nothing';
 /** §17.9.21 `w:numPicBullet` — the image a level uses in place of a bullet. */
 export interface PictureBullet {
     /** Content-addressed bytes. A bullet whose picture does not resolve is not one. */

package/dist/esm/core/drawingml/text-warp.d.ts

@@ -46,6 +46,40 @@ export interface WarpFrame {
  *          draw — the pinched end of a fade, or a preset this file cannot bend.
  */
 export declare function warpGlyphMatrix(frame: WarpFrame, glyphX: number, advance: number, baselineY: number): readonly [number, number, number, number, number, number] | undefined;
+/** One glyph of a shaped run, set on a warp's curve: its id and its {@link warpGlyphMatrix}. */
+export interface WarpedGlyph {
+    readonly gid: number;
+    /** Where the glyph stands in the run it came from. */
+    readonly index: number;
+    readonly matrix: readonly [number, number, number, number, number, number];
+}
+/**
+ * The glyphs of a run the Indic shaper set, each placed on the warp's curve.
+ *
+ * Such a run cannot be set character by character the way other text on a
+ * curve is: its glyphs stand in an order of their own (the short i of हि
+ * before the ha), one glyph may stand for several characters, and a mark is
+ * set off its pen position by the face's anchors. So the curve takes the
+ * shaper's glyphs. Each advances by its shaped advance, the run rescaled to
+ * the width the layout measured it at. A glyph's x offset moves it along the
+ * line before the warp maps it, and its y offset lifts it off the baseline in
+ * its own frame, so a mark rises and leans with the letter it sits on.
+ *
+ * @param frame     The two frames, from the laid-out line.
+ * @param run       The run's glyphs, advances and offsets, in font units.
+ * @param x         Where the run starts in the un-warped block.
+ * @param widthPt   The width the layout measured the run at.
+ * @param emPt      Points per font unit at the run's size: the size over the em.
+ * @param baselineY The line's baseline in the un-warped block.
+ * @returns Each glyph and its matrix, in the shaper's order; a glyph the warp
+ *          leaves no room for is left out.
+ */
+export declare function warpGlyphRun(frame: WarpFrame, run: {
+    readonly gids: ReadonlyArray<number>;
+    readonly advances: ReadonlyArray<number>;
+    readonly xOffsets?: ReadonlyArray<number>;
+    readonly yOffsets?: ReadonlyArray<number>;
+}, x: number, widthPt: number, emPt: number, baselineY: number): Array<WarpedGlyph>;
 /**
  * The two edges of a warp at one position across its box.
  *

package/dist/esm/core/drawingml/text-warp.js

@@ -252,6 +252,54 @@ function warpGlyphMatrix(frame, glyphX, advance, baselineY) {
 	if (ring) return ringMatrix(frame, ring, u, v, glyphX);
 	return envelopeMatrix(frame, u, v, glyphX, advance);
 }
+/**
+* The glyphs of a run the Indic shaper set, each placed on the warp's curve.
+*
+* Such a run cannot be set character by character the way other text on a
+* curve is: its glyphs stand in an order of their own (the short i of हि
+* before the ha), one glyph may stand for several characters, and a mark is
+* set off its pen position by the face's anchors. So the curve takes the
+* shaper's glyphs. Each advances by its shaped advance, the run rescaled to
+* the width the layout measured it at. A glyph's x offset moves it along the
+* line before the warp maps it, and its y offset lifts it off the baseline in
+* its own frame, so a mark rises and leans with the letter it sits on.
+*
+* @param frame     The two frames, from the laid-out line.
+* @param run       The run's glyphs, advances and offsets, in font units.
+* @param x         Where the run starts in the un-warped block.
+* @param widthPt   The width the layout measured the run at.
+* @param emPt      Points per font unit at the run's size: the size over the em.
+* @param baselineY The line's baseline in the un-warped block.
+* @returns Each glyph and its matrix, in the shaper's order; a glyph the warp
+*          leaves no room for is left out.
+*/
+function warpGlyphRun(frame, run, x, widthPt, emPt, baselineY) {
+	let total = 0;
+	for (const a of run.advances) total += a;
+	const along = total > 0 ? widthPt / total : emPt;
+	const out = [];
+	let pen = x;
+	for (let i = 0; i < run.gids.length; i++) {
+		const adv = (run.advances[i] ?? 0) * along;
+		const m = warpGlyphMatrix(frame, pen + (run.xOffsets?.[i] ?? 0) * along, adv, baselineY);
+		pen += adv;
+		if (!m) continue;
+		const dy = (run.yOffsets?.[i] ?? 0) * emPt;
+		out.push({
+			gid: run.gids[i],
+			index: i,
+			matrix: [
+				m[0],
+				m[1],
+				m[2],
+				m[3],
+				m[4] + m[2] * dy,
+				m[5] + m[3] * dy
+			]
+		});
+	}
+	return out;
+}
 /** The glyph matrix for a warp that hangs text between a top and a bottom edge. */
 function envelopeMatrix(frame, u, v, glyphX, advance) {
 	const edges = textWarpEdges(frame.preset, u, frame.adjust);
@@ -340,4 +388,4 @@ function textWarpEdges(preset, u, adj) {
 	return env.at(Math.min(1, Math.max(0, u)), a);
 }
 //#endregion
-export { isTextWarp, textWarpEdges, warpGlyphMatrix };
+export { isTextWarp, textWarpEdges, warpGlyphMatrix, warpGlyphRun };

package/dist/esm/core/font/index.d.ts

@@ -7,7 +7,7 @@ export { pickVariant } from './font-registry.js';
 export { parseGposKerning, parseGsubLigatures, shapeText } from './opentype-layout.js';
 export type { KerningMap, LigatureMap, ShapedRun } from './opentype-layout.js';
 export { hasSubstitutable, lettersForLigature, substituteLetters } from './ligatures.js';
-export { createFontMeasure } from './measure.js';
+export { createFontMeasure, shapeGlyphs } from './measure.js';
 export type { FontMeasure } from './measure.js';
 export { buildTrueType, editableEmbedding, MAX_BUILT_GLYPHS } from './ttf-build.js';
 export type { BuiltFace, BuiltGlyph, BuiltKernPair, BuiltLigature, GlyphSeg, } from './ttf-build.js';

package/dist/esm/core/font/index.js

@@ -0,0 +1,8 @@
+import "./opentype-layout.js";
+import { parseTtf } from "./ttf-parser.js";
+import { glyphClosure, subsetTtf } from "./ttf-subset.js";
+import { FontRegistry, pickVariant } from "./font-registry.js";
+import { hasSubstitutable, lettersForLigature, substituteLetters } from "./ligatures.js";
+import { createFontMeasure, shapeGlyphs } from "./measure.js";
+import { MAX_BUILT_GLYPHS, buildTrueType, editableEmbedding } from "./ttf-build.js";
+export { FontRegistry, MAX_BUILT_GLYPHS, buildTrueType, createFontMeasure, editableEmbedding, glyphClosure, hasSubstitutable, lettersForLigature, parseTtf, pickVariant, shapeGlyphs, subsetTtf, substituteLetters };

package/dist/esm/core/font/indic-shaper.d.ts

@@ -0,0 +1,47 @@
+import { ParsedTtf } from './ttf-parser.js';
+/**
+ * Whether a string holds a letter of one of the scripts this shapes.
+ *
+ * @param text The text.
+ * @returns `true` when one of its characters is in their blocks.
+ */
+export declare function hasIndic(text: string): boolean;
+/**
+ * Where the syllables of the scripts of India begin in a text, by the grammar
+ * the shaper cuts them with: a syllable is the unit a reph, a conjunct or a
+ * vowel sign is shaped within, and shaped in two halves it falls apart.
+ *
+ * @param text The text.
+ * @returns Per code point: whether a syllable — or anything that is not one —
+ *          begins at it.
+ */
+export declare function syllableStarts(text: string): Array<boolean>;
+/** The glyphs of a shaped run, in drawing order, with where each is drawn. */
+export interface IndicRun {
+    readonly gids: Array<number>;
+    /** Per glyph: the index, in code points, of the first character of its cluster. */
+    readonly clusters: Array<number>;
+    /**
+     * Per glyph: the indices, in code points, of the characters it stands for,
+     * least first. A conjunct stands for its letters and viramas, a reph for its
+     * Ra and virama, each part of a vowel sign drawn in two for the sign, and a
+     * dotted circle the shaper put in for nothing.
+     */
+    readonly sources: Array<ReadonlyArray<number>>;
+    /** Per glyph, in font units. */
+    readonly advances: Array<number>;
+    readonly xOffsets: Array<number>;
+    readonly yOffsets: Array<number>;
+}
+/**
+ * Shape a run of text in one face, when the face has OpenType layout for the
+ * script of India its text is in.
+ *
+ * @param parsed The face.
+ * @param text   The text.
+ * @param kern   Whether to apply the face's `kern` feature (the layout profile decides).
+ * @returns The shaped glyphs, or `undefined` when the text holds no letter of
+ *          those scripts or the face has no layout for them — the caller's
+ *          own shaping applies then.
+ */
+export declare function shapeIndic(parsed: ParsedTtf, text: string, kern?: boolean): IndicRun | undefined;