pdf-codec 1.3.0 → 1.4.0

package/dist/images-read.cjs

@@ -0,0 +1,250 @@
+Object.defineProperty(exports, Symbol.toStringTag, { value: "Module" });
+const require_objects = require("./objects.cjs");
+const require_filters = require("./filters.cjs");
+const require_image_png_encode = require("./image/png-encode.cjs");
+const require_image_jpeg_info = require("./image/jpeg-info.cjs");
+//#region src/images-read.ts
+function componentsOf(cs) {
+	if (cs.kind === "gray") return 1;
+	if (cs.kind === "cmyk") return 4;
+	return 3;
+}
+function resolveColorSpace(csObj, resolver, sink) {
+	const resolved = resolver.resolve(csObj);
+	const directName = resolved?.kind === "name" ? resolved.name : void 0;
+	if (directName !== void 0) {
+		if (directName === "DeviceGray" || directName === "CalGray" || directName === "G") return { kind: "gray" };
+		if (directName === "DeviceRGB" || directName === "CalRGB" || directName === "RGB") return { kind: "rgb" };
+		if (directName === "DeviceCMYK" || directName === "CMYK") return { kind: "cmyk" };
+		return {
+			kind: "unsupported",
+			name: directName
+		};
+	}
+	const arr = resolved?.kind === "array" ? resolved.items : void 0;
+	if (arr === void 0) return {
+		kind: "unsupported",
+		name: "(missing or invalid /ColorSpace)"
+	};
+	const family = require_objects.asName(arr[0]);
+	if (family === "ICCBased") {
+		const streamObj = resolver.resolve(arr[1]);
+		const n = streamObj?.kind === "stream" ? require_objects.asNumber(require_objects.dictGet(streamObj.dict, "N")) : void 0;
+		if (n === 1) return { kind: "gray" };
+		if (n === 4) return { kind: "cmyk" };
+		return { kind: "rgb" };
+	}
+	if (family === "Indexed") {
+		const base = resolveColorSpace(arr[1], resolver, sink);
+		const lookupObj = resolver.resolve(arr[3]);
+		let lookup;
+		if (lookupObj?.kind === "stream") lookup = require_filters.decodeStream(lookupObj.raw, lookupObj.dict, sink).bytes;
+		else if (lookupObj?.kind === "string") lookup = lookupObj.bytes;
+		else lookup = /* @__PURE__ */ new Uint8Array(0);
+		return {
+			kind: "indexed",
+			base,
+			lookup
+		};
+	}
+	if (family === "CalGray") return { kind: "gray" };
+	if (family === "CalRGB") return { kind: "rgb" };
+	return {
+		kind: "unsupported",
+		name: family ?? "(unrecognised array colour space)"
+	};
+}
+function unpackSamples(data, width, height, componentsPerPixel, bitsPerComponent) {
+	if (bitsPerComponent === 8) return Array.from(data.subarray(0, width * height * componentsPerPixel));
+	const samplesPerRow = width * componentsPerPixel;
+	const bytesPerRow = Math.ceil(samplesPerRow * bitsPerComponent / 8);
+	const out = [];
+	for (let row = 0; row < height; row++) {
+		const rowStart = row * bytesPerRow;
+		let bitPos = 0;
+		for (let s = 0; s < samplesPerRow; s++) {
+			let value = 0;
+			for (let b = 0; b < bitsPerComponent; b++) {
+				const byteIndex = rowStart + Math.floor(bitPos / 8);
+				const bitIndex = 7 - bitPos % 8;
+				const bit = (data[byteIndex] ?? 0) >> bitIndex & 1;
+				value = value << 1 | bit;
+				bitPos++;
+			}
+			out.push(value);
+		}
+	}
+	return out;
+}
+function scaleToByte(value, maxValue, inverted) {
+	const v = inverted ? maxValue - value : value;
+	return maxValue === 255 ? v : Math.round(v * 255 / maxValue);
+}
+function cmykToRgbByte(c, m, y, k) {
+	return {
+		r: Math.round(255 * (1 - c) * (1 - k)),
+		g: Math.round(255 * (1 - m) * (1 - k)),
+		b: Math.round(255 * (1 - y) * (1 - k))
+	};
+}
+function sampleFromPalette(base, lookup, offset) {
+	if (base.kind === "gray") {
+		const g = lookup[offset] ?? 0;
+		return {
+			r: g,
+			g,
+			b: g
+		};
+	}
+	if (base.kind === "cmyk") return cmykToRgbByte((lookup[offset] ?? 0) / 255, (lookup[offset + 1] ?? 0) / 255, (lookup[offset + 2] ?? 0) / 255, (lookup[offset + 3] ?? 0) / 255);
+	return {
+		r: lookup[offset] ?? 0,
+		g: lookup[offset + 1] ?? 0,
+		b: lookup[offset + 2] ?? 0
+	};
+}
+function buildRawImage(data, width, height, bitsPerComponent, colorSpace, inverted) {
+	const maxValue = (1 << bitsPerComponent) - 1;
+	if (colorSpace.kind === "indexed") {
+		const indices = unpackSamples(data, width, height, 1, bitsPerComponent);
+		const baseComponents = componentsOf(colorSpace.base);
+		const out = new Uint8Array(width * height * (colorSpace.base.kind === "gray" ? 1 : 3));
+		let outIdx = 0;
+		for (const index of indices) {
+			const rgb = sampleFromPalette(colorSpace.base, colorSpace.lookup, index * baseComponents);
+			if (colorSpace.base.kind === "gray") out[outIdx++] = rgb.r;
+			else {
+				out[outIdx++] = rgb.r;
+				out[outIdx++] = rgb.g;
+				out[outIdx++] = rgb.b;
+			}
+		}
+		return {
+			width,
+			height,
+			channels: colorSpace.base.kind === "gray" ? 1 : 3,
+			data: out
+		};
+	}
+	if (colorSpace.kind === "gray") {
+		const samples = unpackSamples(data, width, height, 1, bitsPerComponent);
+		return {
+			width,
+			height,
+			channels: 1,
+			data: Uint8Array.from(samples, (v) => scaleToByte(v, maxValue, inverted))
+		};
+	}
+	if (colorSpace.kind === "rgb") {
+		const samples = unpackSamples(data, width, height, 3, bitsPerComponent);
+		return {
+			width,
+			height,
+			channels: 3,
+			data: Uint8Array.from(samples, (v) => scaleToByte(v, maxValue, inverted))
+		};
+	}
+	const samples = unpackSamples(data, width, height, 4, bitsPerComponent);
+	const out = new Uint8Array(width * height * 3);
+	for (let px = 0; px < width * height; px++) {
+		const rgb = cmykToRgbByte(scaleToByte(samples[px * 4] ?? 0, maxValue, inverted) / 255, scaleToByte(samples[px * 4 + 1] ?? 0, maxValue, inverted) / 255, scaleToByte(samples[px * 4 + 2] ?? 0, maxValue, inverted) / 255, scaleToByte(samples[px * 4 + 3] ?? 0, maxValue, inverted) / 255);
+		out[px * 3] = rgb.r;
+		out[px * 3 + 1] = rgb.g;
+		out[px * 3 + 2] = rgb.b;
+	}
+	return {
+		width,
+		height,
+		channels: 3,
+		data: out
+	};
+}
+function readSoftMaskAlpha(dict, width, height, resolver, sink) {
+	const smaskObj = resolver.resolve(require_objects.dictGet(dict, "SMask"));
+	if (smaskObj?.kind !== "stream") return;
+	const decoded = require_filters.decodeStream(smaskObj.raw, smaskObj.dict, sink);
+	if (decoded.remainingFilter !== void 0) return;
+	const smaskWidth = require_objects.asNumber(require_objects.dictGet(smaskObj.dict, "Width")) ?? width;
+	const smaskHeight = require_objects.asNumber(require_objects.dictGet(smaskObj.dict, "Height")) ?? height;
+	const smaskBpc = require_objects.asNumber(require_objects.dictGet(smaskObj.dict, "BitsPerComponent")) ?? 8;
+	if (smaskWidth !== width || smaskHeight !== height || smaskBpc !== 8) return;
+	return decoded.bytes.subarray(0, width * height);
+}
+function isTrue(obj) {
+	return obj?.kind === "bool" && obj.value;
+}
+function readImageXObject(dict, raw, resolver, sink) {
+	if (isTrue(require_objects.dictGet(dict, "ImageMask") ?? require_objects.dictGet(dict, "IM"))) {
+		sink({
+			code: "image/mask-unsupported",
+			severity: "info",
+			message: "an /ImageMask stencil paints with the current fill colour rather than standing alone as an image; skipping"
+		});
+		return;
+	}
+	const decoded = require_filters.decodeStream(raw, dict, sink);
+	if (decoded.remainingFilter === "DCTDecode") {
+		let info;
+		try {
+			info = require_image_jpeg_info.readJpegInfo(decoded.bytes);
+		} catch {
+			sink({
+				code: "image/undecodable",
+				severity: "warning",
+				message: "DCTDecode image bytes did not look like a valid JPEG (no SOF marker found); skipping this image"
+			});
+			return;
+		}
+		return {
+			format: "jpeg",
+			bytes: decoded.bytes,
+			widthPx: info.width,
+			heightPx: info.height
+		};
+	}
+	if (decoded.remainingFilter !== void 0) return;
+	const width = require_objects.asNumber(require_objects.dictGet(dict, "Width") ?? require_objects.dictGet(dict, "W"));
+	const height = require_objects.asNumber(require_objects.dictGet(dict, "Height") ?? require_objects.dictGet(dict, "H"));
+	if (width === void 0 || height === void 0 || width <= 0 || height <= 0) {
+		sink({
+			code: "image/undecodable",
+			severity: "warning",
+			message: "image XObject is missing a valid /Width or /Height; skipping"
+		});
+		return;
+	}
+	const bitsPerComponent = require_objects.asNumber(require_objects.dictGet(dict, "BitsPerComponent") ?? require_objects.dictGet(dict, "BPC")) ?? 8;
+	if (bitsPerComponent !== 1 && bitsPerComponent !== 2 && bitsPerComponent !== 4 && bitsPerComponent !== 8) {
+		sink({
+			code: "image/unsupported-bit-depth",
+			severity: "warning",
+			message: `image has an unsupported /BitsPerComponent (${String(bitsPerComponent)}); skipping`
+		});
+		return;
+	}
+	const colorSpace = resolveColorSpace(require_objects.dictGet(dict, "ColorSpace") ?? require_objects.dictGet(dict, "CS"), resolver, sink);
+	if (colorSpace.kind === "unsupported") {
+		sink({
+			code: "image/unsupported-colorspace",
+			severity: "warning",
+			message: `image has an unsupported colour space (${colorSpace.name}); skipping`
+		});
+		return;
+	}
+	const decodeArr = require_objects.asArray(require_objects.dictGet(dict, "Decode"));
+	const inverted = decodeArr !== void 0 && require_objects.asNumber(decodeArr[0]) === 1 && require_objects.asNumber(decodeArr[1]) === 0;
+	const rawImage = buildRawImage(decoded.bytes, width, height, bitsPerComponent, colorSpace, inverted);
+	const alpha = readSoftMaskAlpha(dict, width, height, resolver, sink);
+	const withAlpha = alpha !== void 0 ? {
+		...rawImage,
+		alpha
+	} : rawImage;
+	return {
+		format: "png",
+		bytes: require_image_png_encode.encodePng(withAlpha),
+		widthPx: width,
+		heightPx: height
+	};
+}
+//#endregion
+exports.readImageXObject = readImageXObject;

package/dist/images-read.d.cts

@@ -0,0 +1,13 @@
+import { PdfDiagnosticSink } from "./diagnostics.cjs";
+import { PdfDict } from "./objects.cjs";
+import { PdfObjectResolver } from "./interpret.cjs";
+//#region src/images-read.d.ts
+interface ExtractedPdfImage {
+  readonly format: 'png' | 'jpeg';
+  readonly bytes: Uint8Array<ArrayBuffer>;
+  readonly widthPx: number;
+  readonly heightPx: number;
+}
+declare function readImageXObject(dict: PdfDict, raw: Uint8Array<ArrayBuffer>, resolver: PdfObjectResolver, sink: PdfDiagnosticSink): ExtractedPdfImage | undefined;
+//#endregion
+export { ExtractedPdfImage, readImageXObject };

package/dist/images-read.d.ts

@@ -0,0 +1,13 @@
+import { PdfDiagnosticSink } from "./diagnostics.js";
+import { PdfDict } from "./objects.js";
+import { PdfObjectResolver } from "./interpret.js";
+//#region src/images-read.d.ts
+interface ExtractedPdfImage {
+  readonly format: 'png' | 'jpeg';
+  readonly bytes: Uint8Array<ArrayBuffer>;
+  readonly widthPx: number;
+  readonly heightPx: number;
+}
+declare function readImageXObject(dict: PdfDict, raw: Uint8Array<ArrayBuffer>, resolver: PdfObjectResolver, sink: PdfDiagnosticSink): ExtractedPdfImage | undefined;
+//#endregion
+export { ExtractedPdfImage, readImageXObject };

package/dist/images-read.js

@@ -0,0 +1,249 @@
+import { asArray, asName, asNumber, dictGet } from "./objects.js";
+import { decodeStream } from "./filters.js";
+import { encodePng } from "./image/png-encode.js";
+import { readJpegInfo } from "./image/jpeg-info.js";
+//#region src/images-read.ts
+function componentsOf(cs) {
+	if (cs.kind === "gray") return 1;
+	if (cs.kind === "cmyk") return 4;
+	return 3;
+}
+function resolveColorSpace(csObj, resolver, sink) {
+	const resolved = resolver.resolve(csObj);
+	const directName = resolved?.kind === "name" ? resolved.name : void 0;
+	if (directName !== void 0) {
+		if (directName === "DeviceGray" || directName === "CalGray" || directName === "G") return { kind: "gray" };
+		if (directName === "DeviceRGB" || directName === "CalRGB" || directName === "RGB") return { kind: "rgb" };
+		if (directName === "DeviceCMYK" || directName === "CMYK") return { kind: "cmyk" };
+		return {
+			kind: "unsupported",
+			name: directName
+		};
+	}
+	const arr = resolved?.kind === "array" ? resolved.items : void 0;
+	if (arr === void 0) return {
+		kind: "unsupported",
+		name: "(missing or invalid /ColorSpace)"
+	};
+	const family = asName(arr[0]);
+	if (family === "ICCBased") {
+		const streamObj = resolver.resolve(arr[1]);
+		const n = streamObj?.kind === "stream" ? asNumber(dictGet(streamObj.dict, "N")) : void 0;
+		if (n === 1) return { kind: "gray" };
+		if (n === 4) return { kind: "cmyk" };
+		return { kind: "rgb" };
+	}
+	if (family === "Indexed") {
+		const base = resolveColorSpace(arr[1], resolver, sink);
+		const lookupObj = resolver.resolve(arr[3]);
+		let lookup;
+		if (lookupObj?.kind === "stream") lookup = decodeStream(lookupObj.raw, lookupObj.dict, sink).bytes;
+		else if (lookupObj?.kind === "string") lookup = lookupObj.bytes;
+		else lookup = /* @__PURE__ */ new Uint8Array(0);
+		return {
+			kind: "indexed",
+			base,
+			lookup
+		};
+	}
+	if (family === "CalGray") return { kind: "gray" };
+	if (family === "CalRGB") return { kind: "rgb" };
+	return {
+		kind: "unsupported",
+		name: family ?? "(unrecognised array colour space)"
+	};
+}
+function unpackSamples(data, width, height, componentsPerPixel, bitsPerComponent) {
+	if (bitsPerComponent === 8) return Array.from(data.subarray(0, width * height * componentsPerPixel));
+	const samplesPerRow = width * componentsPerPixel;
+	const bytesPerRow = Math.ceil(samplesPerRow * bitsPerComponent / 8);
+	const out = [];
+	for (let row = 0; row < height; row++) {
+		const rowStart = row * bytesPerRow;
+		let bitPos = 0;
+		for (let s = 0; s < samplesPerRow; s++) {
+			let value = 0;
+			for (let b = 0; b < bitsPerComponent; b++) {
+				const byteIndex = rowStart + Math.floor(bitPos / 8);
+				const bitIndex = 7 - bitPos % 8;
+				const bit = (data[byteIndex] ?? 0) >> bitIndex & 1;
+				value = value << 1 | bit;
+				bitPos++;
+			}
+			out.push(value);
+		}
+	}
+	return out;
+}
+function scaleToByte(value, maxValue, inverted) {
+	const v = inverted ? maxValue - value : value;
+	return maxValue === 255 ? v : Math.round(v * 255 / maxValue);
+}
+function cmykToRgbByte(c, m, y, k) {
+	return {
+		r: Math.round(255 * (1 - c) * (1 - k)),
+		g: Math.round(255 * (1 - m) * (1 - k)),
+		b: Math.round(255 * (1 - y) * (1 - k))
+	};
+}
+function sampleFromPalette(base, lookup, offset) {
+	if (base.kind === "gray") {
+		const g = lookup[offset] ?? 0;
+		return {
+			r: g,
+			g,
+			b: g
+		};
+	}
+	if (base.kind === "cmyk") return cmykToRgbByte((lookup[offset] ?? 0) / 255, (lookup[offset + 1] ?? 0) / 255, (lookup[offset + 2] ?? 0) / 255, (lookup[offset + 3] ?? 0) / 255);
+	return {
+		r: lookup[offset] ?? 0,
+		g: lookup[offset + 1] ?? 0,
+		b: lookup[offset + 2] ?? 0
+	};
+}
+function buildRawImage(data, width, height, bitsPerComponent, colorSpace, inverted) {
+	const maxValue = (1 << bitsPerComponent) - 1;
+	if (colorSpace.kind === "indexed") {
+		const indices = unpackSamples(data, width, height, 1, bitsPerComponent);
+		const baseComponents = componentsOf(colorSpace.base);
+		const out = new Uint8Array(width * height * (colorSpace.base.kind === "gray" ? 1 : 3));
+		let outIdx = 0;
+		for (const index of indices) {
+			const rgb = sampleFromPalette(colorSpace.base, colorSpace.lookup, index * baseComponents);
+			if (colorSpace.base.kind === "gray") out[outIdx++] = rgb.r;
+			else {
+				out[outIdx++] = rgb.r;
+				out[outIdx++] = rgb.g;
+				out[outIdx++] = rgb.b;
+			}
+		}
+		return {
+			width,
+			height,
+			channels: colorSpace.base.kind === "gray" ? 1 : 3,
+			data: out
+		};
+	}
+	if (colorSpace.kind === "gray") {
+		const samples = unpackSamples(data, width, height, 1, bitsPerComponent);
+		return {
+			width,
+			height,
+			channels: 1,
+			data: Uint8Array.from(samples, (v) => scaleToByte(v, maxValue, inverted))
+		};
+	}
+	if (colorSpace.kind === "rgb") {
+		const samples = unpackSamples(data, width, height, 3, bitsPerComponent);
+		return {
+			width,
+			height,
+			channels: 3,
+			data: Uint8Array.from(samples, (v) => scaleToByte(v, maxValue, inverted))
+		};
+	}
+	const samples = unpackSamples(data, width, height, 4, bitsPerComponent);
+	const out = new Uint8Array(width * height * 3);
+	for (let px = 0; px < width * height; px++) {
+		const rgb = cmykToRgbByte(scaleToByte(samples[px * 4] ?? 0, maxValue, inverted) / 255, scaleToByte(samples[px * 4 + 1] ?? 0, maxValue, inverted) / 255, scaleToByte(samples[px * 4 + 2] ?? 0, maxValue, inverted) / 255, scaleToByte(samples[px * 4 + 3] ?? 0, maxValue, inverted) / 255);
+		out[px * 3] = rgb.r;
+		out[px * 3 + 1] = rgb.g;
+		out[px * 3 + 2] = rgb.b;
+	}
+	return {
+		width,
+		height,
+		channels: 3,
+		data: out
+	};
+}
+function readSoftMaskAlpha(dict, width, height, resolver, sink) {
+	const smaskObj = resolver.resolve(dictGet(dict, "SMask"));
+	if (smaskObj?.kind !== "stream") return;
+	const decoded = decodeStream(smaskObj.raw, smaskObj.dict, sink);
+	if (decoded.remainingFilter !== void 0) return;
+	const smaskWidth = asNumber(dictGet(smaskObj.dict, "Width")) ?? width;
+	const smaskHeight = asNumber(dictGet(smaskObj.dict, "Height")) ?? height;
+	const smaskBpc = asNumber(dictGet(smaskObj.dict, "BitsPerComponent")) ?? 8;
+	if (smaskWidth !== width || smaskHeight !== height || smaskBpc !== 8) return;
+	return decoded.bytes.subarray(0, width * height);
+}
+function isTrue(obj) {
+	return obj?.kind === "bool" && obj.value;
+}
+function readImageXObject(dict, raw, resolver, sink) {
+	if (isTrue(dictGet(dict, "ImageMask") ?? dictGet(dict, "IM"))) {
+		sink({
+			code: "image/mask-unsupported",
+			severity: "info",
+			message: "an /ImageMask stencil paints with the current fill colour rather than standing alone as an image; skipping"
+		});
+		return;
+	}
+	const decoded = decodeStream(raw, dict, sink);
+	if (decoded.remainingFilter === "DCTDecode") {
+		let info;
+		try {
+			info = readJpegInfo(decoded.bytes);
+		} catch {
+			sink({
+				code: "image/undecodable",
+				severity: "warning",
+				message: "DCTDecode image bytes did not look like a valid JPEG (no SOF marker found); skipping this image"
+			});
+			return;
+		}
+		return {
+			format: "jpeg",
+			bytes: decoded.bytes,
+			widthPx: info.width,
+			heightPx: info.height
+		};
+	}
+	if (decoded.remainingFilter !== void 0) return;
+	const width = asNumber(dictGet(dict, "Width") ?? dictGet(dict, "W"));
+	const height = asNumber(dictGet(dict, "Height") ?? dictGet(dict, "H"));
+	if (width === void 0 || height === void 0 || width <= 0 || height <= 0) {
+		sink({
+			code: "image/undecodable",
+			severity: "warning",
+			message: "image XObject is missing a valid /Width or /Height; skipping"
+		});
+		return;
+	}
+	const bitsPerComponent = asNumber(dictGet(dict, "BitsPerComponent") ?? dictGet(dict, "BPC")) ?? 8;
+	if (bitsPerComponent !== 1 && bitsPerComponent !== 2 && bitsPerComponent !== 4 && bitsPerComponent !== 8) {
+		sink({
+			code: "image/unsupported-bit-depth",
+			severity: "warning",
+			message: `image has an unsupported /BitsPerComponent (${String(bitsPerComponent)}); skipping`
+		});
+		return;
+	}
+	const colorSpace = resolveColorSpace(dictGet(dict, "ColorSpace") ?? dictGet(dict, "CS"), resolver, sink);
+	if (colorSpace.kind === "unsupported") {
+		sink({
+			code: "image/unsupported-colorspace",
+			severity: "warning",
+			message: `image has an unsupported colour space (${colorSpace.name}); skipping`
+		});
+		return;
+	}
+	const decodeArr = asArray(dictGet(dict, "Decode"));
+	const inverted = decodeArr !== void 0 && asNumber(decodeArr[0]) === 1 && asNumber(decodeArr[1]) === 0;
+	const rawImage = buildRawImage(decoded.bytes, width, height, bitsPerComponent, colorSpace, inverted);
+	const alpha = readSoftMaskAlpha(dict, width, height, resolver, sink);
+	const withAlpha = alpha !== void 0 ? {
+		...rawImage,
+		alpha
+	} : rawImage;
+	return {
+		format: "png",
+		bytes: encodePng(withAlpha),
+		widthPx: width,
+		heightPx: height
+	};
+}
+//#endregion
+export { readImageXObject };