pdf-codec 5.7.0 → 5.7.1

package/README.md

@@ -283,7 +283,7 @@ The package is layered from generic primitives outward to the codec itself:
   - **Embedded text faces**: `embedded-font.ts` (parses one TrueType-outline face's metrics and `GPOS` pair kerning, and `encodeForShowEmbedded` — the single code path both measurement and text-showing go through so encoding and measuring cannot disagree). Every geometry field is converted into PDF's 1000-units-per-em glyph space. `embedded-font-write.ts` builds the `/Type0`/`/CIDFontType2`/`/FontDescriptor`/`/FontFile2`/ToUnicode group, with `/CIDToGIDMap /Identity` written explicitly and `/Length1` set to the **uncompressed** subset length. Its subset tag is a CRC32 over the face's PostScript name and glyph-ID list, so identical input yields byte-identical output.
   - **ToUnicode CMaps**: `tounicode.ts`, shared by both embedded-font writers — a character code → Unicode code point mapping written as a bfchar CMap (9.10.3), with supplementary-plane code points encoded as UTF-16BE surrogate pairs and entries emitted in blocks of at most 100.
   - **CFF reading**: `cff.ts` (shared `INDEX`/`DICT` container structures), `cff-bounds.ts` (a Type 2 charstring interpreter computing each glyph's tight ink bounding box by tracking the current point through every path operator and solving each cubic's real extrema from the roots of its derivative — a path walker, not a rasteriser; verified against the vendored STIX Two Math font's whole 5,543-glyph repertoire, matching fontTools' `BoundsPen` to within 0.01 design units). `cff-probe.ts` reads a bare CFF program's header/Name INDEX/Top DICT to detect the `ROS` operator defining a CID-keyed font — the guard a future source-embedded-font phase needs before it can trust CID == GID against an arbitrary caller-supplied font.
-  - **Read**: `lexer.ts`/`parse.ts` (byte tokenizer and tokens → `PdfObject`), `filters.ts`/`predictors.ts` (Flate/LZW/ASCII85/ASCIIHex/RunLength/CCITTFax, TIFF/PNG predictors), `xref.ts`/`document.ts` (classic and cross-reference-stream resolution, object streams, `/Prev` chains, linear-scan recovery, the page tree with attribute inheritance), `encrypt.ts` (standard security handler: `/Encrypt` parsing, empty-user-password key derivation and `/U` verification, per-object keys, transparent string/stream decryption), `content-read.ts`/`interpret.ts` (content-stream tokenizer and graphics/text state machine, form-XObject recursion, general vector-path tracking), `cmap.ts`/`font-style.ts`/`font-read.ts` (`/ToUnicode` CMaps, font-dictionary resolution), `builtin-encoding.ts` (a font's own built-in encoding, read out of the embedded program itself — see [Text extraction and font encodings](#text-extraction-and-font-encodings)), `images-read.ts` (Image XObjects → PNG/JPEG bytes), `names.ts` (the document-level name-tree walker: one flattening pass for every `/Names` tenant), `navigation.ts` (named destinations from `/Dests` and `/Names` `/Dests`, reader-minted entries for direct destination arrays, and the `/Outlines` bookmark tree), `attachments.ts` (embedded files from the name tree, `/FileAttachment` filespecs, and `/AF`), `optional-content.ts` (`/OCProperties` groups and default-configuration visibility, plus the `/OC`-to-name resolution `interpret.ts` stamps onto span items), `annotations.ts` (sticky notes, FreeText, the `/QuadPoints` markup family, and residue for the opaque kinds), `form.ts` (the AcroForm field tree), `structure.ts` (the tagged-PDF `/StructTreeRoot` element tree with `/RoleMap` resolution and `/ClassMap` attribute merging, plus the `/ParentTree` walk that resolves each `(page, MCID)` pair to the owning element `read.ts` stamps onto an item), `xmp.ts` (a bounded Dublin Core extractor for the `/Metadata` packet), `pdf-text.ts` (PDF string/date scalar decoding), `read.ts` (`readPdf`, assembling all of the above into a `LayoutDocument`).
+  - **Read**: `lexer.ts`/`parse.ts` (byte tokenizer and tokens → `PdfObject`), `filters.ts`/`predictors.ts` (Flate/LZW/ASCII85/ASCIIHex/RunLength/CCITTFax, TIFF/PNG predictors), `xref.ts`/`document.ts` (classic and cross-reference-stream resolution, object streams, `/Prev` chains, linear-scan recovery, the page tree with attribute inheritance), `encrypt.ts` (standard security handler: `/Encrypt` parsing, empty-user-password key derivation and `/U` verification, per-object keys, transparent string/stream decryption), `content-read.ts`/`interpret.ts`/`interpret-state.ts` (content-stream tokenizer and graphics/text state machine, with the state records and the per-glyph displacement arithmetic in the last, form-XObject recursion, general vector-path tracking), `cmap.ts`/`font-style.ts`/`font-read.ts` (`/ToUnicode` CMaps, font-dictionary resolution), `builtin-encoding.ts` (a font's own built-in encoding, read out of the embedded program itself — see [Text extraction and font encodings](#text-extraction-and-font-encodings)), `images-read.ts` (Image XObjects → PNG/JPEG bytes), `names.ts` (the document-level name-tree walker: one flattening pass for every `/Names` tenant), `navigation.ts` (named destinations from `/Dests` and `/Names` `/Dests`, reader-minted entries for direct destination arrays, and the `/Outlines` bookmark tree), `attachments.ts` (embedded files from the name tree, `/FileAttachment` filespecs, and `/AF`), `optional-content.ts` (`/OCProperties` groups and default-configuration visibility, plus the `/OC`-to-name resolution `interpret.ts` stamps onto span items), `annotations.ts` (sticky notes, FreeText, the `/QuadPoints` markup family, and residue for the opaque kinds), `form.ts` (the AcroForm field tree), `structure.ts` (the tagged-PDF `/StructTreeRoot` element tree with `/RoleMap` resolution and `/ClassMap` attribute merging, plus the `/ParentTree` walk that resolves each `(page, MCID)` pair to the owning element `read.ts` stamps onto an item), `xmp.ts` (a bounded Dublin Core extractor for the `/Metadata` packet), `pdf-text.ts` (PDF string/date scalar decoding), `read.ts` (`readPdf`, assembling all of the above into a `LayoutDocument`).
   - **Raster**: `raster.ts` — the `PageRasteriser` port and `renderPdfPage` (#1198): re-walks one page through `interpretContentStream` (the same interpreter `readPdf` runs), recomputes the page's crop/rotation geometry (test-pinned to agree with `readPdf` bit for bit, since the `clipPt` contract rides on it), and drives a caller-supplied rasteriser with device-space draw ops — rect fills, filled/stroked vector paths, affine-placed images, and text as filled glyph-outline paths resolved from the page's own embedded sfnt programs (`glyf-contours.ts`'s decoded contours through the run's text matrices, per-glyph advances corrected against the run's exact end matrix). Optional-content layers the default configuration leaves OFF are not drawn; faces with no sfnt outlines are refused by name. No canvas dependency, no backend, no OCR/vision/network — the port is pure data plus callbacks, so `pdf-raster-cpu` (or a browser canvas wrapper) implements it elsewhere.
   - `codec.ts` — `pdfCodec`, a `z.codec()` pair over `readPdf`/`writePdf`, plus a standalone local copy of the `%PDF-` header check.
 - **`src/test-support/`** — hand-built PDF fixtures (`pdf.ts`) built by literal byte/string concatenation and deliberately importing NOTHING from this package's own writer (a fixture built by `writePdf` would let a writer bug hide from the corresponding reader test). `encrypted-pdfs.ts` applies the same principle: real PDFs encrypted by [qpdf](https://qpdf.sourceforge.io/), embedded as base64, so a bug in key derivation cannot cancel out between write and read halves. `fonts.ts` holds the real vendored Carlito and Caladea faces as raw sfnt bytes, and asserts values read out of the `.ttf` files by a standalone script with a bare `DataView`, not by this package's own parsers — external cross-checks rather than a parser's output compared against itself.
@@ -334,7 +334,7 @@ Reading a program's built-in encoding is what stops a symbol-encoded subset —
 
 ## Grouping text runs into lines and words
 
-`readPdf` reports one positioned run per text-showing operator, and a PDF says nothing about lines or words: a line of prose can arrive as one run, as one run per word, or as one run per glyph, entirely at the producer's discretion. `groupPdfTextRuns` turns those runs back into lines of words, and `pdf-codec/text-group` serves it without the write half or the vendored fonts, alongside `pdf-codec/read`.
+`readPdf` reports one positioned run per text-showing operator, split wherever the operator itself opens a gap at least a word space wide between two of its glyphs (a `TJ` adjustment, a character spacing `Tc`, or the two summed, since a producer can set one against the other), and a PDF says nothing about lines or words: a line of prose can arrive as one run, as one run per word, or as one run per glyph, entirely at the producer's discretion. `groupPdfTextRuns` turns those runs back into lines of words, and `pdf-codec/text-group` serves it without the write half or the vendored fonts, alongside `pdf-codec/read`.
 
 ```ts
 import { readPdf } from "pdf-codec/read";

package/dist/index.cjs

@@ -21,7 +21,7 @@ const require_write_skipped_surfaces = require("./write-skipped-surfaces.cjs");
 const require_write = require("./write.cjs");
 const require_codec = require("./codec.cjs");
 const require_font_face = require("./font-face.cjs");
-const require_raster = require("./raster-B8i3Tgjj.cjs");
+const require_raster = require("./raster-De_zVqCu.cjs");
 exports.DEFAULT_BASELINE_TOLERANCE_EM = require_text_group.DEFAULT_BASELINE_TOLERANCE_EM;
 exports.DEFAULT_COLUMN_GAP_EM = require_text_group.DEFAULT_COLUMN_GAP_EM;
 exports.DEFAULT_VERTICAL_METRIC_POLICY = require_measure.DEFAULT_VERTICAL_METRIC_POLICY;

package/dist/index.js

@@ -20,6 +20,6 @@ import { PdfSkippedSurfacesError } from "./write-skipped-surfaces.js";
 import { PdfPartialDocumentError, writePdf } from "./write.js";
 import { PdfBytesSchema, pdfCodec } from "./codec.js";
 import { FontFaceParseError, readFontFace } from "./font-face.js";
-import { r as renderPdfPage } from "./raster-CIc6ot36.js";
+import { r as renderPdfPage } from "./raster-DCDcD7no.js";
 export * from "byte-codec";
 export { DEFAULT_BASELINE_TOLERANCE_EM, DEFAULT_COLUMN_GAP_EM, DEFAULT_VERTICAL_METRIC_POLICY, DEFAULT_WORD_GAP_EM, FontFaceParseError, Jbig2ParseError, Jbig2UnsupportedError, Jpeg2000ParseError, Jpeg2000UnsupportedError, LAYOUT_FORMAT_VERSION, LayoutAnnotationQuadSchema, LayoutAnnotationSchema, LayoutAttachmentSchema, LayoutDestinationSchema, LayoutDestinationTargetSchema, LayoutDocumentSchema, LayoutDocumentSurfaceSchema, LayoutEllipseSchema, LayoutFormFieldSchema, LayoutFormWidgetSchema, LayoutImageAssetSchema, LayoutImageSchema, LayoutInternalLinkSchema, LayoutItemSchema, LayoutLayerSchema, LayoutLineSchema, LayoutLinkSchema, LayoutOutlineItemSchema, LayoutPageSchema, LayoutPathSchema, LayoutPathSegmentSchema, LayoutRectSchema, LayoutStructureElementSchema, LayoutSubpathSchema, LayoutTextSchema, NOOP_DIAGNOSTIC_SINK, PdfBytesSchema, PdfEncryptedError, PdfEncryptionError, PdfPageSelectionError, PdfParseError, PdfPartialDocumentError, PdfPasswordRequiredError, PdfSkippedSurfacesError, STANDARD_METRICS, assembleStretchyGlyph, attachScriptRuns, createFontMeasurer, createFontRegistry, createStandardFontMeasurer, decodeCcittFax, decodeJbig2Embedded, decodeJpeg2000, filterScanlines, groupPdfTextRuns, loadMathFont, looksLikeBareCodestream, parseJp2Container, pdfCodec, readFontFace, readJpeg2000Metadata, readPdf, renderPdfPage, resolveFaceWithRegistry, resolveStandardFont, runGapPt, runsShareBaseline, scaleMathStretchConstruction, scriptPositionOf, sourcePageIndex, unfilterScanlines, writePdf };

package/dist/{interpret-2_N_lh0h.cjs → interpret-BkDxq0QA.cjs}

@@ -2,7 +2,8 @@ const require_objects = require("./objects.cjs");
 const require_matrix = require("./matrix.cjs");
 const require_filters = require("./filters.cjs");
 const require_content_read = require("./content-read.cjs");
-require("./text-group.cjs");
+const require_text_group = require("./text-group.cjs");
+const require_interpret_state = require("./interpret-state.cjs");
 const require_interpret_shapes = require("./interpret-shapes.cjs");
 const require_interpret_marked = require("./interpret-marked.cjs");
 let document_schema_js = require("document-schema.js");
@@ -44,38 +45,7 @@ function genericColor(operands) {
 //#endregion
 //#region src/interpret.ts
 const MAX_FORM_XOBJECT_DEPTH = 12;
-const FALLBACK_GLYPH_WIDTH_PER_1000 = 500;
-const GLYPH_SPACE_PER_TEXT_SPACE_UNIT = 1e3;
-const ASCII_SPACE = 32;
 const DEFAULT_LINE_WIDTH_PT = 1;
-function initialTextParameters() {
-	return {
-		fontResourceName: void 0,
-		fontSizePt: 0,
-		charSpace: 0,
-		wordSpace: 0,
-		horizScale: 1,
-		leading: 0,
-		rise: 0
-	};
-}
-function defaultTextObjectState() {
-	return {
-		tm: require_matrix.IDENTITY_MATRIX,
-		tlm: require_matrix.IDENTITY_MATRIX
-	};
-}
-function computeTrm(gs, text) {
-	const fontMatrix = [
-		gs.fontSizePt * gs.horizScale,
-		0,
-		0,
-		gs.fontSizePt,
-		0,
-		gs.rise
-	];
-	return require_matrix.multiplyMatrices(require_matrix.multiplyMatrices(fontMatrix, text.tm), gs.ctm);
-}
 function numAt(operands, index) {
 	return require_objects.asNumber(operands[index]) ?? 0;
 }
@@ -104,7 +74,7 @@ function interpretContentStream(bytes, resources, context) {
 		strokeColor: document_schema_js.COLOR_BLACK,
 		lineWidth: DEFAULT_LINE_WIDTH_PT,
 		dashArray: [],
-		...initialTextParameters()
+		...require_interpret_state.initialTextParameters()
 	}, context, 0, new require_interpret_shapes.MarkedContentStack());
 }
 function lastPointOf(subpath) {
@@ -122,7 +92,7 @@ function runContentStream(bytes, resources, initialState, context, depth, marked
 	const operations = require_content_read.readContentStream(bytes, context.sink);
 	const gsStack = [];
 	let gs = initialState;
-	let text = defaultTextObjectState();
+	let text = require_interpret_state.defaultTextObjectState();
 	let pathSubpaths = [];
 	let currentSubpath;
 	const inScope = (key) => markedContent.inScope(key);
@@ -236,25 +206,16 @@ function runContentStream(bytes, resources, initialState, context, depth, marked
 		}
 		resetPath();
 	};
-	const advanceThroughString = (codes) => {
-		const fontResourceName = gs.fontResourceName;
-		if (fontResourceName === void 0) return;
-		let offset = 0;
-		while (offset < codes.length) {
-			const glyph = context.fontMetrics.glyphAdvance(fontResourceName, resources, codes, offset);
-			if (glyph === void 0) context.sink({
-				code: "pdf/font-not-resolved",
-				severity: "warning",
-				message: `could not resolve font resource /${fontResourceName} to compute a glyph advance; assuming a fallback width`
-			});
-			const widthPer1000 = glyph?.widthPer1000 ?? FALLBACK_GLYPH_WIDTH_PER_1000;
-			const byteLength = glyph?.byteLengthConsumed ?? 1;
-			const isSingleByteSpace = byteLength === 1 && codes[offset] === ASCII_SPACE;
-			const spacing = gs.charSpace + (isSingleByteSpace ? gs.wordSpace : 0);
-			const vertical = glyph?.vertical;
-			text.tm = vertical === void 0 ? require_matrix.multiplyMatrices(require_matrix.translationMatrix((widthPer1000 / GLYPH_SPACE_PER_TEXT_SPACE_UNIT * gs.fontSizePt + spacing) * gs.horizScale, 0), text.tm) : require_matrix.multiplyMatrices(require_matrix.translationMatrix(0, vertical.displacementPer1000 / GLYPH_SPACE_PER_TEXT_SPACE_UNIT * gs.fontSizePt + spacing), text.tm);
-			offset += byteLength;
-		}
+	const advanceThroughGlyph = (fontResourceName, codes, offset) => {
+		const glyph = context.fontMetrics.glyphAdvance(fontResourceName, resources, codes, offset);
+		if (glyph === void 0) context.sink({
+			code: "pdf/font-not-resolved",
+			severity: "warning",
+			message: `could not resolve font resource /${fontResourceName} to compute a glyph advance; assuming a fallback width`
+		});
+		const step = require_interpret_state.stepPastGlyph(gs, text.tm, glyph, codes, offset);
+		text.tm = step.tm;
+		return step;
 	};
 	const positionVectorOf = (codes) => {
 		const fontResourceName = gs.fontResourceName;
@@ -262,18 +223,22 @@ function runContentStream(bytes, resources, initialState, context, depth, marked
 		const vertical = context.fontMetrics.glyphAdvance(fontResourceName, resources, codes, 0)?.vertical;
 		if (vertical === void 0) return;
 		return {
-			x: vertical.positionXPer1000 / GLYPH_SPACE_PER_TEXT_SPACE_UNIT,
-			y: vertical.positionYPer1000 / GLYPH_SPACE_PER_TEXT_SPACE_UNIT
+			x: vertical.positionXPer1000 / require_interpret_state.GLYPH_SPACE_PER_TEXT_SPACE_UNIT,
+			y: vertical.positionYPer1000 / require_interpret_state.GLYPH_SPACE_PER_TEXT_SPACE_UNIT
 		};
 	};
 	const showTextArray = (elements) => {
 		const fontResourceName = gs.fontResourceName;
 		if (fontResourceName === void 0) return;
 		const vertical = context.fontMetrics.isVertical(fontResourceName, resources);
-		let startMatrix = computeTrm(gs, text);
+		const wordGap = require_text_group.DEFAULT_WORD_GAP_EM * gs.fontSizePt;
+		let startMatrix = require_interpret_state.computeTrm(gs, text);
 		let chunks = [];
 		let totalLength = 0;
 		let firstGlyphPosition;
+		let lastWidthEndTm;
+		let gapSinceLastGlyph = 0;
+		const gapIsWordWide = () => !vertical && totalLength > 0 && gapSinceLastGlyph >= wordGap;
 		const flushRun = () => {
 			if (totalLength === 0) return;
 			const combined = new Uint8Array(totalLength);
@@ -282,7 +247,10 @@ function runContentStream(bytes, resources, initialState, context, depth, marked
 				combined.set(chunk, at);
 				at += chunk.length;
 			}
-			const endMatrix = computeTrm(gs, text);
+			const endMatrix = require_interpret_state.computeTrm(gs, gapIsWordWide() && lastWidthEndTm !== void 0 ? {
+				...text,
+				tm: lastWidthEndTm
+			} : text);
 			const shift = firstGlyphPosition === void 0 ? void 0 : require_matrix.translationMatrix(-firstGlyphPosition.x, -firstGlyphPosition.y);
 			pushItem({
 				kind: "text",
@@ -300,15 +268,22 @@ function runContentStream(bytes, resources, initialState, context, depth, marked
 			firstGlyphPosition = void 0;
 		};
 		for (const el of elements) if (el.kind === "string") {
-			firstGlyphPosition ??= vertical ? positionVectorOf(el.bytes) : void 0;
-			chunks.push(el.bytes);
-			totalLength += el.bytes.length;
-			advanceThroughString(el.bytes);
+			let offset = 0;
+			while (offset < el.bytes.length) {
+				if (gapIsWordWide()) flushRun();
+				if (totalLength === 0) startMatrix = require_interpret_state.computeTrm(gs, text);
+				firstGlyphPosition ??= vertical ? positionVectorOf(el.bytes.subarray(offset)) : void 0;
+				const glyph = advanceThroughGlyph(fontResourceName, el.bytes, offset);
+				chunks.push(el.bytes.subarray(offset, offset + glyph.byteLength));
+				totalLength += glyph.byteLength;
+				offset += glyph.byteLength;
+				lastWidthEndTm = glyph.widthEndTm;
+				gapSinceLastGlyph = glyph.charSpaceAfter;
+			}
 		} else if (el.kind === "number") {
-			const adjustment = -(el.value / GLYPH_SPACE_PER_TEXT_SPACE_UNIT) * gs.fontSizePt;
-			if (!vertical && totalLength > 0 && adjustment >= .125 * gs.fontSizePt) flushRun();
+			const adjustment = -(el.value / require_interpret_state.GLYPH_SPACE_PER_TEXT_SPACE_UNIT) * gs.fontSizePt;
+			gapSinceLastGlyph += adjustment;
 			text.tm = require_matrix.multiplyMatrices(vertical ? require_matrix.translationMatrix(0, adjustment) : require_matrix.translationMatrix(adjustment * gs.horizScale, 0), text.tm);
-			if (totalLength === 0) startMatrix = computeTrm(gs, text);
 		}
 		flushRun();
 	};
@@ -573,7 +548,7 @@ function runContentStream(bytes, resources, initialState, context, depth, marked
 				emitPaint(operator);
 				break;
 			case "BT":
-				text = defaultTextObjectState();
+				text = require_interpret_state.defaultTextObjectState();
 				break;
 			case "Tf":
 				gs = {

package/dist/{interpret-BeMKZGBm.js → interpret-BtzX-LWm.js}

@@ -2,7 +2,8 @@ import { asArray, asName, asNumber, dictGet } from "./objects.js";
 import { IDENTITY_MATRIX, applyMatrix, multiplyMatrices, translationMatrix } from "./matrix.js";
 import { decodeStream } from "./filters.js";
 import { readContentStream } from "./content-read.js";
-import "./text-group.js";
+import { DEFAULT_WORD_GAP_EM } from "./text-group.js";
+import { GLYPH_SPACE_PER_TEXT_SPACE_UNIT, computeTrm, defaultTextObjectState, initialTextParameters, stepPastGlyph } from "./interpret-state.js";
 import { MarkedContentStack, classifyShape, strokeStyleFromDashArray } from "./interpret-shapes.js";
 import { markedContentProperties, markedContentScopeProps } from "./interpret-marked.js";
 import { COLOR_BLACK } from "document-schema.js";
@@ -44,38 +45,7 @@ function genericColor(operands) {
 //#endregion
 //#region src/interpret.ts
 const MAX_FORM_XOBJECT_DEPTH = 12;
-const FALLBACK_GLYPH_WIDTH_PER_1000 = 500;
-const GLYPH_SPACE_PER_TEXT_SPACE_UNIT = 1e3;
-const ASCII_SPACE = 32;
 const DEFAULT_LINE_WIDTH_PT = 1;
-function initialTextParameters() {
-	return {
-		fontResourceName: void 0,
-		fontSizePt: 0,
-		charSpace: 0,
-		wordSpace: 0,
-		horizScale: 1,
-		leading: 0,
-		rise: 0
-	};
-}
-function defaultTextObjectState() {
-	return {
-		tm: IDENTITY_MATRIX,
-		tlm: IDENTITY_MATRIX
-	};
-}
-function computeTrm(gs, text) {
-	const fontMatrix = [
-		gs.fontSizePt * gs.horizScale,
-		0,
-		0,
-		gs.fontSizePt,
-		0,
-		gs.rise
-	];
-	return multiplyMatrices(multiplyMatrices(fontMatrix, text.tm), gs.ctm);
-}
 function numAt(operands, index) {
 	return asNumber(operands[index]) ?? 0;
 }
@@ -236,25 +206,16 @@ function runContentStream(bytes, resources, initialState, context, depth, marked
 		}
 		resetPath();
 	};
-	const advanceThroughString = (codes) => {
-		const fontResourceName = gs.fontResourceName;
-		if (fontResourceName === void 0) return;
-		let offset = 0;
-		while (offset < codes.length) {
-			const glyph = context.fontMetrics.glyphAdvance(fontResourceName, resources, codes, offset);
-			if (glyph === void 0) context.sink({
-				code: "pdf/font-not-resolved",
-				severity: "warning",
-				message: `could not resolve font resource /${fontResourceName} to compute a glyph advance; assuming a fallback width`
-			});
-			const widthPer1000 = glyph?.widthPer1000 ?? FALLBACK_GLYPH_WIDTH_PER_1000;
-			const byteLength = glyph?.byteLengthConsumed ?? 1;
-			const isSingleByteSpace = byteLength === 1 && codes[offset] === ASCII_SPACE;
-			const spacing = gs.charSpace + (isSingleByteSpace ? gs.wordSpace : 0);
-			const vertical = glyph?.vertical;
-			text.tm = vertical === void 0 ? multiplyMatrices(translationMatrix((widthPer1000 / GLYPH_SPACE_PER_TEXT_SPACE_UNIT * gs.fontSizePt + spacing) * gs.horizScale, 0), text.tm) : multiplyMatrices(translationMatrix(0, vertical.displacementPer1000 / GLYPH_SPACE_PER_TEXT_SPACE_UNIT * gs.fontSizePt + spacing), text.tm);
-			offset += byteLength;
-		}
+	const advanceThroughGlyph = (fontResourceName, codes, offset) => {
+		const glyph = context.fontMetrics.glyphAdvance(fontResourceName, resources, codes, offset);
+		if (glyph === void 0) context.sink({
+			code: "pdf/font-not-resolved",
+			severity: "warning",
+			message: `could not resolve font resource /${fontResourceName} to compute a glyph advance; assuming a fallback width`
+		});
+		const step = stepPastGlyph(gs, text.tm, glyph, codes, offset);
+		text.tm = step.tm;
+		return step;
 	};
 	const positionVectorOf = (codes) => {
 		const fontResourceName = gs.fontResourceName;
@@ -270,10 +231,14 @@ function runContentStream(bytes, resources, initialState, context, depth, marked
 		const fontResourceName = gs.fontResourceName;
 		if (fontResourceName === void 0) return;
 		const vertical = context.fontMetrics.isVertical(fontResourceName, resources);
+		const wordGap = DEFAULT_WORD_GAP_EM * gs.fontSizePt;
 		let startMatrix = computeTrm(gs, text);
 		let chunks = [];
 		let totalLength = 0;
 		let firstGlyphPosition;
+		let lastWidthEndTm;
+		let gapSinceLastGlyph = 0;
+		const gapIsWordWide = () => !vertical && totalLength > 0 && gapSinceLastGlyph >= wordGap;
 		const flushRun = () => {
 			if (totalLength === 0) return;
 			const combined = new Uint8Array(totalLength);
@@ -282,7 +247,10 @@ function runContentStream(bytes, resources, initialState, context, depth, marked
 				combined.set(chunk, at);
 				at += chunk.length;
 			}
-			const endMatrix = computeTrm(gs, text);
+			const endMatrix = computeTrm(gs, gapIsWordWide() && lastWidthEndTm !== void 0 ? {
+				...text,
+				tm: lastWidthEndTm
+			} : text);
 			const shift = firstGlyphPosition === void 0 ? void 0 : translationMatrix(-firstGlyphPosition.x, -firstGlyphPosition.y);
 			pushItem({
 				kind: "text",
@@ -300,15 +268,22 @@ function runContentStream(bytes, resources, initialState, context, depth, marked
 			firstGlyphPosition = void 0;
 		};
 		for (const el of elements) if (el.kind === "string") {
-			firstGlyphPosition ??= vertical ? positionVectorOf(el.bytes) : void 0;
-			chunks.push(el.bytes);
-			totalLength += el.bytes.length;
-			advanceThroughString(el.bytes);
+			let offset = 0;
+			while (offset < el.bytes.length) {
+				if (gapIsWordWide()) flushRun();
+				if (totalLength === 0) startMatrix = computeTrm(gs, text);
+				firstGlyphPosition ??= vertical ? positionVectorOf(el.bytes.subarray(offset)) : void 0;
+				const glyph = advanceThroughGlyph(fontResourceName, el.bytes, offset);
+				chunks.push(el.bytes.subarray(offset, offset + glyph.byteLength));
+				totalLength += glyph.byteLength;
+				offset += glyph.byteLength;
+				lastWidthEndTm = glyph.widthEndTm;
+				gapSinceLastGlyph = glyph.charSpaceAfter;
+			}
 		} else if (el.kind === "number") {
 			const adjustment = -(el.value / GLYPH_SPACE_PER_TEXT_SPACE_UNIT) * gs.fontSizePt;
-			if (!vertical && totalLength > 0 && adjustment >= .125 * gs.fontSizePt) flushRun();
+			gapSinceLastGlyph += adjustment;
 			text.tm = multiplyMatrices(vertical ? translationMatrix(0, adjustment) : translationMatrix(adjustment * gs.horizScale, 0), text.tm);
-			if (totalLength === 0) startMatrix = computeTrm(gs, text);
 		}
 		flushRun();
 	};

package/dist/interpret-colors.cjs

@@ -1,5 +1,5 @@
 Object.defineProperty(exports, Symbol.toStringTag, { value: "Module" });
-const require_interpret = require("./interpret-2_N_lh0h.cjs");
+const require_interpret = require("./interpret-BkDxq0QA.cjs");
 exports.cmykColor = require_interpret.cmykColor;
 exports.genericColor = require_interpret.genericColor;
 exports.grayColor = require_interpret.grayColor;

package/dist/interpret-colors.js

@@ -1,2 +1,2 @@
-import { a as grayColor, i as genericColor, o as rgbColor, r as cmykColor } from "./interpret-BeMKZGBm.js";
+import { a as grayColor, i as genericColor, o as rgbColor, r as cmykColor } from "./interpret-BtzX-LWm.js";
 export { cmykColor, genericColor, grayColor, rgbColor };

package/dist/interpret-state.cjs

@@ -0,0 +1,66 @@
+Object.defineProperty(exports, Symbol.toStringTag, { value: "Module" });
+const require_matrix = require("./matrix.cjs");
+//#region src/interpret-state.ts
+const FALLBACK_GLYPH_WIDTH_PER_1000 = 500;
+const GLYPH_SPACE_PER_TEXT_SPACE_UNIT = 1e3;
+const ASCII_SPACE = 32;
+function initialTextParameters() {
+	return {
+		fontResourceName: void 0,
+		fontSizePt: 0,
+		charSpace: 0,
+		wordSpace: 0,
+		horizScale: 1,
+		leading: 0,
+		rise: 0
+	};
+}
+function defaultTextObjectState() {
+	return {
+		tm: require_matrix.IDENTITY_MATRIX,
+		tlm: require_matrix.IDENTITY_MATRIX
+	};
+}
+function computeTrm(gs, text) {
+	const fontMatrix = [
+		gs.fontSizePt * gs.horizScale,
+		0,
+		0,
+		gs.fontSizePt,
+		0,
+		gs.rise
+	];
+	return require_matrix.multiplyMatrices(require_matrix.multiplyMatrices(fontMatrix, text.tm), gs.ctm);
+}
+/**
+* Moves a text matrix past one glyph.
+* @param gs - the graphics state whose font size, spacing and horizontal scaling apply
+* @param tm - the text matrix where the glyph starts
+* @param glyph - the font's advance for the glyph's code, or undefined when the font could not be resolved, in which case a fallback width is assumed and the code is taken as one byte
+* @param codes - the shown string
+* @param offset - where the glyph's code starts in `codes`
+* @returns the step, with the pen's new position and the end of the glyph's own width reported apart so a caller can tell the glyph's extent from the gap after it
+*/
+function stepPastGlyph(gs, tm, glyph, codes, offset) {
+	const widthPer1000 = glyph?.widthPer1000 ?? 500;
+	const byteLength = glyph?.byteLengthConsumed ?? 1;
+	const isSingleByteSpace = byteLength === 1 && codes[offset] === 32;
+	const spacing = gs.charSpace + (isSingleByteSpace ? gs.wordSpace : 0);
+	const vertical = glyph?.vertical;
+	const widthAdvance = vertical === void 0 ? widthPer1000 / GLYPH_SPACE_PER_TEXT_SPACE_UNIT * gs.fontSizePt : vertical.displacementPer1000 / GLYPH_SPACE_PER_TEXT_SPACE_UNIT * gs.fontSizePt;
+	const along = (distance) => vertical === void 0 ? require_matrix.translationMatrix(distance * gs.horizScale, 0) : require_matrix.translationMatrix(0, distance);
+	return {
+		byteLength,
+		tm: require_matrix.multiplyMatrices(along(widthAdvance + spacing), tm),
+		widthEndTm: require_matrix.multiplyMatrices(along(widthAdvance), tm),
+		charSpaceAfter: gs.charSpace
+	};
+}
+//#endregion
+exports.ASCII_SPACE = ASCII_SPACE;
+exports.FALLBACK_GLYPH_WIDTH_PER_1000 = FALLBACK_GLYPH_WIDTH_PER_1000;
+exports.GLYPH_SPACE_PER_TEXT_SPACE_UNIT = GLYPH_SPACE_PER_TEXT_SPACE_UNIT;
+exports.computeTrm = computeTrm;
+exports.defaultTextObjectState = defaultTextObjectState;
+exports.initialTextParameters = initialTextParameters;
+exports.stepPastGlyph = stepPastGlyph;

package/dist/interpret-state.d.cts

@@ -0,0 +1,51 @@
+import { r as Matrix } from "./matrix-B7_SBKQ5.cjs";
+import { GlyphAdvance } from "./interpret-types.cjs";
+import { Color } from "document-schema.js";
+//#region src/interpret-state.d.ts
+declare const FALLBACK_GLYPH_WIDTH_PER_1000 = 500;
+declare const GLYPH_SPACE_PER_TEXT_SPACE_UNIT = 1000;
+declare const ASCII_SPACE = 32;
+interface GraphicsState {
+  readonly ctm: Matrix;
+  readonly fillColor: Color;
+  readonly strokeColor: Color;
+  readonly lineWidth: number;
+  readonly dashArray: readonly number[];
+  readonly fontResourceName: string | undefined;
+  readonly fontSizePt: number;
+  readonly charSpace: number;
+  readonly wordSpace: number;
+  readonly horizScale: number;
+  readonly leading: number;
+  readonly rise: number;
+}
+interface TextObjectState {
+  tm: Matrix;
+  tlm: Matrix;
+}
+declare function initialTextParameters(): Omit<GraphicsState, "ctm" | "fillColor" | "strokeColor" | "lineWidth" | "dashArray">;
+declare function defaultTextObjectState(): TextObjectState;
+declare function computeTrm(gs: GraphicsState, text: TextObjectState): Matrix;
+/** Where one glyph leaves the text matrix, and what lies between the glyph and the next one. */
+interface GlyphStep {
+  /** How many bytes of the shown string the glyph's code took. */
+  readonly byteLength: number;
+  /** The text matrix after the glyph's width and its spacing, where the next glyph starts. */
+  readonly tm: Matrix;
+  /** The text matrix at the end of the glyph's own advance width, before any spacing. */
+  readonly widthEndTm: Matrix;
+  /** The character spacing added after the glyph's width, in unscaled text space units. The word spacing a single-byte space takes is not counted: the space character already records that word boundary in the run's text. */
+  readonly charSpaceAfter: number;
+}
+/**
+ * Moves a text matrix past one glyph.
+ * @param gs - the graphics state whose font size, spacing and horizontal scaling apply
+ * @param tm - the text matrix where the glyph starts
+ * @param glyph - the font's advance for the glyph's code, or undefined when the font could not be resolved, in which case a fallback width is assumed and the code is taken as one byte
+ * @param codes - the shown string
+ * @param offset - where the glyph's code starts in `codes`
+ * @returns the step, with the pen's new position and the end of the glyph's own width reported apart so a caller can tell the glyph's extent from the gap after it
+ */
+declare function stepPastGlyph(gs: GraphicsState, tm: Matrix, glyph: GlyphAdvance | undefined, codes: Uint8Array<ArrayBuffer>, offset: number): GlyphStep;
+//#endregion
+export { ASCII_SPACE, FALLBACK_GLYPH_WIDTH_PER_1000, GLYPH_SPACE_PER_TEXT_SPACE_UNIT, GlyphStep, GraphicsState, TextObjectState, computeTrm, defaultTextObjectState, initialTextParameters, stepPastGlyph };

package/dist/interpret-state.d.ts

@@ -0,0 +1,51 @@
+import { r as Matrix } from "./matrix-B7_SBKQ5.js";
+import { GlyphAdvance } from "./interpret-types.js";
+import { Color } from "document-schema.js";
+//#region src/interpret-state.d.ts
+declare const FALLBACK_GLYPH_WIDTH_PER_1000 = 500;
+declare const GLYPH_SPACE_PER_TEXT_SPACE_UNIT = 1000;
+declare const ASCII_SPACE = 32;
+interface GraphicsState {
+  readonly ctm: Matrix;
+  readonly fillColor: Color;
+  readonly strokeColor: Color;
+  readonly lineWidth: number;
+  readonly dashArray: readonly number[];
+  readonly fontResourceName: string | undefined;
+  readonly fontSizePt: number;
+  readonly charSpace: number;
+  readonly wordSpace: number;
+  readonly horizScale: number;
+  readonly leading: number;
+  readonly rise: number;
+}
+interface TextObjectState {
+  tm: Matrix;
+  tlm: Matrix;
+}
+declare function initialTextParameters(): Omit<GraphicsState, "ctm" | "fillColor" | "strokeColor" | "lineWidth" | "dashArray">;
+declare function defaultTextObjectState(): TextObjectState;
+declare function computeTrm(gs: GraphicsState, text: TextObjectState): Matrix;
+/** Where one glyph leaves the text matrix, and what lies between the glyph and the next one. */
+interface GlyphStep {
+  /** How many bytes of the shown string the glyph's code took. */
+  readonly byteLength: number;
+  /** The text matrix after the glyph's width and its spacing, where the next glyph starts. */
+  readonly tm: Matrix;
+  /** The text matrix at the end of the glyph's own advance width, before any spacing. */
+  readonly widthEndTm: Matrix;
+  /** The character spacing added after the glyph's width, in unscaled text space units. The word spacing a single-byte space takes is not counted: the space character already records that word boundary in the run's text. */
+  readonly charSpaceAfter: number;
+}
+/**
+ * Moves a text matrix past one glyph.
+ * @param gs - the graphics state whose font size, spacing and horizontal scaling apply
+ * @param tm - the text matrix where the glyph starts
+ * @param glyph - the font's advance for the glyph's code, or undefined when the font could not be resolved, in which case a fallback width is assumed and the code is taken as one byte
+ * @param codes - the shown string
+ * @param offset - where the glyph's code starts in `codes`
+ * @returns the step, with the pen's new position and the end of the glyph's own width reported apart so a caller can tell the glyph's extent from the gap after it
+ */
+declare function stepPastGlyph(gs: GraphicsState, tm: Matrix, glyph: GlyphAdvance | undefined, codes: Uint8Array<ArrayBuffer>, offset: number): GlyphStep;
+//#endregion
+export { ASCII_SPACE, FALLBACK_GLYPH_WIDTH_PER_1000, GLYPH_SPACE_PER_TEXT_SPACE_UNIT, GlyphStep, GraphicsState, TextObjectState, computeTrm, defaultTextObjectState, initialTextParameters, stepPastGlyph };

package/dist/interpret-state.js

@@ -0,0 +1,59 @@
+import { IDENTITY_MATRIX, multiplyMatrices, translationMatrix } from "./matrix.js";
+//#region src/interpret-state.ts
+const FALLBACK_GLYPH_WIDTH_PER_1000 = 500;
+const GLYPH_SPACE_PER_TEXT_SPACE_UNIT = 1e3;
+const ASCII_SPACE = 32;
+function initialTextParameters() {
+	return {
+		fontResourceName: void 0,
+		fontSizePt: 0,
+		charSpace: 0,
+		wordSpace: 0,
+		horizScale: 1,
+		leading: 0,
+		rise: 0
+	};
+}
+function defaultTextObjectState() {
+	return {
+		tm: IDENTITY_MATRIX,
+		tlm: IDENTITY_MATRIX
+	};
+}
+function computeTrm(gs, text) {
+	const fontMatrix = [
+		gs.fontSizePt * gs.horizScale,
+		0,
+		0,
+		gs.fontSizePt,
+		0,
+		gs.rise
+	];
+	return multiplyMatrices(multiplyMatrices(fontMatrix, text.tm), gs.ctm);
+}
+/**
+* Moves a text matrix past one glyph.
+* @param gs - the graphics state whose font size, spacing and horizontal scaling apply
+* @param tm - the text matrix where the glyph starts
+* @param glyph - the font's advance for the glyph's code, or undefined when the font could not be resolved, in which case a fallback width is assumed and the code is taken as one byte
+* @param codes - the shown string
+* @param offset - where the glyph's code starts in `codes`
+* @returns the step, with the pen's new position and the end of the glyph's own width reported apart so a caller can tell the glyph's extent from the gap after it
+*/
+function stepPastGlyph(gs, tm, glyph, codes, offset) {
+	const widthPer1000 = glyph?.widthPer1000 ?? 500;
+	const byteLength = glyph?.byteLengthConsumed ?? 1;
+	const isSingleByteSpace = byteLength === 1 && codes[offset] === 32;
+	const spacing = gs.charSpace + (isSingleByteSpace ? gs.wordSpace : 0);
+	const vertical = glyph?.vertical;
+	const widthAdvance = vertical === void 0 ? widthPer1000 / GLYPH_SPACE_PER_TEXT_SPACE_UNIT * gs.fontSizePt : vertical.displacementPer1000 / GLYPH_SPACE_PER_TEXT_SPACE_UNIT * gs.fontSizePt;
+	const along = (distance) => vertical === void 0 ? translationMatrix(distance * gs.horizScale, 0) : translationMatrix(0, distance);
+	return {
+		byteLength,
+		tm: multiplyMatrices(along(widthAdvance + spacing), tm),
+		widthEndTm: multiplyMatrices(along(widthAdvance), tm),
+		charSpaceAfter: gs.charSpace
+	};
+}
+//#endregion
+export { ASCII_SPACE, FALLBACK_GLYPH_WIDTH_PER_1000, GLYPH_SPACE_PER_TEXT_SPACE_UNIT, computeTrm, defaultTextObjectState, initialTextParameters, stepPastGlyph };

package/dist/interpret.cjs

@@ -1,4 +1,4 @@
 Object.defineProperty(exports, Symbol.toStringTag, { value: "Module" });
-const require_interpret = require("./interpret-2_N_lh0h.cjs");
+const require_interpret = require("./interpret-BkDxq0QA.cjs");
 exports.interpretContentStream = require_interpret.interpretContentStream;
 exports.numAt = require_interpret.numAt;

package/dist/interpret.js

@@ -1,2 +1,2 @@
-import { n as numAt, t as interpretContentStream } from "./interpret-BeMKZGBm.js";
+import { n as numAt, t as interpretContentStream } from "./interpret-BtzX-LWm.js";
 export { interpretContentStream, numAt };

package/dist/{raster-CIc6ot36.js → raster-DCDcD7no.js}

@@ -11,7 +11,7 @@ import { NOOP_DIAGNOSTIC_SINK, PdfParseError } from "./diagnostics.js";
 import { openPdfDocument } from "./document.js";
 import { throwIfAborted } from "./util/abort.js";
 import { createFontResolver } from "./font-read.js";
-import { t as interpretContentStream } from "./interpret-BeMKZGBm.js";
+import { t as interpretContentStream } from "./interpret-BtzX-LWm.js";
 import { normalizeRotation, pageRotationTransform } from "./read.js";
 import { parseGlyf } from "./glyf.js";
 import { decodeGlyphOutline } from "./glyf-contours.js";

package/dist/{raster-B8i3Tgjj.cjs → raster-De_zVqCu.cjs}

@@ -11,7 +11,7 @@ const require_diagnostics = require("./diagnostics.cjs");
 const require_document = require("./document.cjs");
 const require_util_abort = require("./util/abort.cjs");
 const require_font_read = require("./font-read.cjs");
-const require_interpret = require("./interpret-2_N_lh0h.cjs");
+const require_interpret = require("./interpret-BkDxq0QA.cjs");
 const require_read = require("./read.cjs");
 const require_glyf = require("./glyf.cjs");
 const require_glyf_contours = require("./glyf-contours.cjs");

package/dist/raster-text.cjs

@@ -1,5 +1,5 @@
 Object.defineProperty(exports, Symbol.toStringTag, { value: "Module" });
-const require_raster = require("./raster-B8i3Tgjj.cjs");
+const require_raster = require("./raster-De_zVqCu.cjs");
 exports.drawGlyphOutline = require_raster.drawGlyphOutline;
 exports.drawTextRun = require_raster.drawTextRun;
 exports.glyphOutlineSubpaths = require_raster.glyphOutlineSubpaths;

package/dist/raster-text.js

@@ -1,2 +1,2 @@
-import { a as drawTextRun, i as drawGlyphOutline, o as glyphOutlineSubpaths } from "./raster-CIc6ot36.js";
+import { a as drawTextRun, i as drawGlyphOutline, o as glyphOutlineSubpaths } from "./raster-DCDcD7no.js";
 export { drawGlyphOutline, drawTextRun, glyphOutlineSubpaths };

package/dist/raster.cjs

@@ -1,5 +1,5 @@
 Object.defineProperty(exports, Symbol.toStringTag, { value: "Module" });
-const require_raster = require("./raster-B8i3Tgjj.cjs");
+const require_raster = require("./raster-De_zVqCu.cjs");
 exports.EXTENT_ZERO_EPSILON = require_raster.EXTENT_ZERO_EPSILON;
 exports.flattenCubic = require_raster.flattenCubic;
 exports.renderPdfPage = require_raster.renderPdfPage;

package/dist/raster.js

@@ -1,2 +1,2 @@
-import { n as flattenCubic, r as renderPdfPage, t as EXTENT_ZERO_EPSILON } from "./raster-CIc6ot36.js";
+import { n as flattenCubic, r as renderPdfPage, t as EXTENT_ZERO_EPSILON } from "./raster-DCDcD7no.js";
 export { EXTENT_ZERO_EPSILON, flattenCubic, renderPdfPage };

package/dist/read.cjs

@@ -17,7 +17,7 @@ const require_diagnostics = require("./diagnostics.cjs");
 const require_document = require("./document.cjs");
 const require_util_abort = require("./util/abort.cjs");
 const require_font_read = require("./font-read.cjs");
-const require_interpret = require("./interpret-2_N_lh0h.cjs");
+const require_interpret = require("./interpret-BkDxq0QA.cjs");
 const require_navigation = require("./navigation.cjs");
 //#region src/read.ts
 function skippedSurfacesOf(surfaces) {

package/dist/read.js

@@ -16,7 +16,7 @@ import { NOOP_DIAGNOSTIC_SINK, PdfPageSelectionError, PdfParseError } from "./di
 import { openPdfDocument } from "./document.js";
 import { throwIfAborted } from "./util/abort.js";
 import { createFontResolver } from "./font-read.js";
-import { t as interpretContentStream } from "./interpret-BeMKZGBm.js";
+import { t as interpretContentStream } from "./interpret-BtzX-LWm.js";
 import { createDestinationRegistry, readOutline } from "./navigation.js";
 //#region src/read.ts
 function skippedSurfacesOf(surfaces) {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "pdf-codec",
-  "version": "5.7.0",
+  "version": "5.7.1",
   "description": "Hand-written, dependency-minimal PDF codec: parses arbitrary real-world PDFs and generates new ones, built on its own codec-owned LayoutDocument item model and Zod 4 codecs.",
   "type": "module",
   "repository": {