pdf-parser.js 5.2.30 → 5.3.0

package/dist/read-convert.cjs

@@ -0,0 +1,310 @@
+Object.defineProperty(exports, Symbol.toStringTag, { value: "Module" });
+require("./notes-annotation-author.cjs");
+const require_objects = require("./objects.cjs");
+const require_pdf_text = require("./pdf-text.cjs");
+const require_matrix = require("./matrix.cjs");
+const require_bytes_crc32 = require("./bytes/crc32.cjs");
+const require_images_read = require("./images-read.cjs");
+let byte_codec = require("byte-codec");
+//#region src/read-convert.ts
+function convertExtractedItem(item, pageMatrix, fontResolver, images, imageIdCache, resolver, sink) {
+	if (item.kind === "text") return convertText(item, pageMatrix, fontResolver);
+	if (item.kind === "rect") return convertRect(item, pageMatrix);
+	if (item.kind === "ellipse") return convertEllipse(item, pageMatrix);
+	if (item.kind === "line") return convertLine(item, pageMatrix);
+	if (item.kind === "path") return convertPath(item, pageMatrix);
+	if (item.kind === "image") return convertImage(item, pageMatrix, images, imageIdCache, resolver, sink);
+	return convertInlineImage(item, pageMatrix, images, resolver, sink);
+}
+function convertText(item, pageMatrix, fontResolver) {
+	const font = fontResolver.resolve(item.fontResourceName, item.resources);
+	const text = font?.decodeToUnicode(item.codes) ?? "";
+	if (text.length === 0) return;
+	const startTrm = require_matrix.multiplyMatrices(item.startMatrix, pageMatrix);
+	const endTrm = require_matrix.multiplyMatrices(item.endMatrix, pageMatrix);
+	const widthPt = Math.hypot(endTrm[4] - startTrm[4], endTrm[5] - startTrm[5]);
+	const sizePt = require_matrix.matrixScaleX(startTrm);
+	const rotationDeg = require_matrix.matrixRotationDegrees(startTrm);
+	const layoutFont = {
+		family: font?.family ?? "Helvetica",
+		weight: font?.bold === true ? "bold" : "normal",
+		style: font?.italic === true ? "italic" : "normal"
+	};
+	return {
+		kind: "text",
+		text,
+		xPt: startTrm[4],
+		yPt: startTrm[5],
+		font: layoutFont,
+		sizePt: sizePt > 0 ? sizePt : item.sizePt,
+		color: item.color,
+		widthPt,
+		rotationDeg: rotationDeg !== 0 ? rotationDeg : void 0,
+		...item.vertical === true ? { writingMode: "vertical" } : {},
+		...item.layerName !== void 0 ? { layer: item.layerName } : {},
+		...item.actualText !== void 0 ? { actualText: item.actualText } : {},
+		...item.alt !== void 0 ? { alt: item.alt } : {}
+	};
+}
+function paintFields(paint) {
+	return {
+		...paint.fill !== void 0 ? { fill: paint.fill } : {},
+		...paint.stroke !== void 0 ? { stroke: paint.stroke } : {}
+	};
+}
+function transformBox(item, pageMatrix) {
+	const p1 = require_matrix.applyMatrix(pageMatrix, {
+		x: item.xPt,
+		y: item.yPt
+	});
+	const p2 = require_matrix.applyMatrix(pageMatrix, {
+		x: item.xPt + item.widthPt,
+		y: item.yPt + item.heightPt
+	});
+	return {
+		xPt: Math.min(p1.x, p2.x),
+		yPt: Math.min(p1.y, p2.y),
+		widthPt: Math.abs(p2.x - p1.x),
+		heightPt: Math.abs(p2.y - p1.y)
+	};
+}
+function convertRect(item, pageMatrix) {
+	return {
+		kind: "rect",
+		...transformBox(item, pageMatrix),
+		...paintFields(item),
+		...item.layerName !== void 0 ? { layer: item.layerName } : {}
+	};
+}
+function convertEllipse(item, pageMatrix) {
+	return {
+		kind: "ellipse",
+		...transformBox(item, pageMatrix),
+		...paintFields(item),
+		...item.layerName !== void 0 ? { layer: item.layerName } : {}
+	};
+}
+function convertLine(item, pageMatrix) {
+	const p1 = require_matrix.applyMatrix(pageMatrix, {
+		x: item.x1Pt,
+		y: item.y1Pt
+	});
+	const p2 = require_matrix.applyMatrix(pageMatrix, {
+		x: item.x2Pt,
+		y: item.y2Pt
+	});
+	return {
+		kind: "line",
+		x1Pt: p1.x,
+		y1Pt: p1.y,
+		x2Pt: p2.x,
+		y2Pt: p2.y,
+		color: item.color,
+		widthPt: item.widthPt,
+		...item.style !== void 0 ? { style: item.style } : {},
+		...item.layerName !== void 0 ? { layer: item.layerName } : {}
+	};
+}
+function transformSubpath(subpath, pageMatrix) {
+	const start = require_matrix.applyMatrix(pageMatrix, {
+		x: subpath.startXPt,
+		y: subpath.startYPt
+	});
+	const segments = subpath.segments.map((segment) => {
+		if (segment.kind === "line") {
+			const p = require_matrix.applyMatrix(pageMatrix, {
+				x: segment.xPt,
+				y: segment.yPt
+			});
+			return {
+				kind: "line",
+				xPt: p.x,
+				yPt: p.y
+			};
+		}
+		const c1 = require_matrix.applyMatrix(pageMatrix, {
+			x: segment.c1xPt,
+			y: segment.c1yPt
+		});
+		const c2 = require_matrix.applyMatrix(pageMatrix, {
+			x: segment.c2xPt,
+			y: segment.c2yPt
+		});
+		const p = require_matrix.applyMatrix(pageMatrix, {
+			x: segment.xPt,
+			y: segment.yPt
+		});
+		return {
+			kind: "cubic",
+			c1xPt: c1.x,
+			c1yPt: c1.y,
+			c2xPt: c2.x,
+			c2yPt: c2.y,
+			xPt: p.x,
+			yPt: p.y
+		};
+	});
+	return {
+		startXPt: start.x,
+		startYPt: start.y,
+		segments,
+		closed: subpath.closed
+	};
+}
+function convertPath(item, pageMatrix) {
+	return {
+		kind: "path",
+		subpaths: item.subpaths.map((subpath) => transformSubpath(subpath, pageMatrix)),
+		...item.fill !== void 0 ? { fill: item.fill } : {},
+		...item.fill !== void 0 && item.fillRule === "evenodd" ? { fillRule: "evenodd" } : {},
+		...item.stroke !== void 0 ? { stroke: item.stroke } : {},
+		...item.style !== void 0 ? { style: item.style } : {},
+		...item.layerName !== void 0 ? { layer: item.layerName } : {}
+	};
+}
+function imagePlacementFrom(matrix) {
+	const rotationDeg = require_matrix.matrixRotationDegrees(matrix);
+	return {
+		xPt: matrix[4],
+		yPt: matrix[5],
+		widthPt: require_matrix.matrixScaleX(matrix),
+		heightPt: require_matrix.matrixScaleY(matrix),
+		rotationDeg: rotationDeg !== 0 ? rotationDeg : void 0
+	};
+}
+const HEX_RADIX = 16;
+function registerExtractedImage(format, bytes, widthPx, heightPx, images, original) {
+	const imageId = `img${require_bytes_crc32.crc32(bytes).toString(HEX_RADIX)}`;
+	if (!(imageId in images)) images[imageId] = {
+		format,
+		base64: (0, byte_codec.bytesToBase64)(bytes),
+		widthPx,
+		heightPx,
+		...original !== void 0 ? { original: {
+			filter: original.filter,
+			base64: (0, byte_codec.bytesToBase64)(original.bytes),
+			...original.globalsBytes !== void 0 ? { jbig2GlobalsBase64: (0, byte_codec.bytesToBase64)(original.globalsBytes) } : {}
+		} } : {}
+	};
+	return imageId;
+}
+function resolveCachedImageId(dict, raw, images, cache, resolver, sink) {
+	if (cache.has(dict)) return cache.get(dict) ?? void 0;
+	const decoded = require_images_read.readImageXObject(dict, raw, resolver, sink);
+	if (decoded === void 0) {
+		cache.set(dict, null);
+		return;
+	}
+	const imageId = registerExtractedImage(decoded.format, decoded.bytes, decoded.widthPx, decoded.heightPx, images, decoded.original);
+	cache.set(dict, imageId);
+	return imageId;
+}
+function convertImage(item, pageMatrix, images, cache, resolver, sink) {
+	const xobjects = resolver.resolveDict(require_objects.dictGet(item.resources, "XObject"));
+	const xobj = xobjects !== void 0 ? resolver.resolve(require_objects.dictGet(xobjects, item.resourceName)) : void 0;
+	if (xobj?.kind !== "stream") return;
+	const imageId = resolveCachedImageId(xobj.dict, xobj.raw, images, cache, resolver, sink);
+	if (imageId === void 0) return;
+	return {
+		kind: "image",
+		imageId,
+		...imagePlacementFrom(require_matrix.multiplyMatrices(item.matrix, pageMatrix)),
+		...item.layerName !== void 0 ? { layer: item.layerName } : {}
+	};
+}
+function convertInlineImage(item, pageMatrix, images, resolver, sink) {
+	const decoded = require_images_read.readImageXObject(item.dict, item.data, resolver, sink);
+	if (decoded === void 0) return;
+	return {
+		kind: "image",
+		imageId: registerExtractedImage(decoded.format, decoded.bytes, decoded.widthPx, decoded.heightPx, images, decoded.original),
+		...imagePlacementFrom(require_matrix.multiplyMatrices(item.matrix, pageMatrix)),
+		...item.layerName !== void 0 ? { layer: item.layerName } : {}
+	};
+}
+function readLinkAnnotations(page, pageMatrix, resolver, destinationRegistry) {
+	const annotsArr = require_objects.asArray(require_objects.dictGet(page, "Annots"));
+	if (annotsArr === void 0) return [];
+	const links = [];
+	for (const annotRef of annotsArr) {
+		const annot = resolver.resolveDict(annotRef);
+		if (annot === void 0 || require_objects.asName(require_objects.dictGet(annot, "Subtype")) !== "Link") continue;
+		const rectArr = require_objects.asArray(require_objects.dictGet(annot, "Rect"));
+		if (rectArr === void 0) continue;
+		const uri = readLinkUri(annot, resolver);
+		const destination = uri === void 0 ? readInternalDestination(annot, resolver, destinationRegistry) : void 0;
+		if (uri === void 0 && destination === void 0) continue;
+		const x1 = require_objects.asNumber(rectArr[0]) ?? 0;
+		const y1 = require_objects.asNumber(rectArr[1]) ?? 0;
+		const x2 = require_objects.asNumber(rectArr[2]) ?? 0;
+		const y2 = require_objects.asNumber(rectArr[3]) ?? 0;
+		const p1 = require_matrix.applyMatrix(pageMatrix, {
+			x: Math.min(x1, x2),
+			y: Math.min(y1, y2)
+		});
+		const p2 = require_matrix.applyMatrix(pageMatrix, {
+			x: Math.max(x1, x2),
+			y: Math.max(y1, y2)
+		});
+		const contentsObj = require_objects.dictGet(annot, "Contents");
+		const title = contentsObj?.kind === "string" ? require_pdf_text.decodePdfString(contentsObj.bytes) : void 0;
+		const box = {
+			xPt: Math.min(p1.x, p2.x),
+			yPt: Math.min(p1.y, p2.y),
+			widthPt: Math.abs(p2.x - p1.x),
+			heightPt: Math.abs(p2.y - p1.y),
+			...title !== void 0 ? { title } : {}
+		};
+		if (uri !== void 0) links.push({
+			kind: "link",
+			uri,
+			...box
+		});
+		else if (destination !== void 0) links.push({
+			kind: "internalLink",
+			destination,
+			...box
+		});
+	}
+	return links;
+}
+function readLinkUri(annot, resolver) {
+	const action = resolver.resolveDict(require_objects.dictGet(annot, "A"));
+	if (action === void 0 || require_objects.asName(require_objects.dictGet(action, "S")) !== "URI") return;
+	const uriObj = require_objects.dictGet(action, "URI");
+	return uriObj?.kind === "string" ? require_pdf_text.decodePdfString(uriObj.bytes) : void 0;
+}
+function readInternalDestination(annot, resolver, destinationRegistry) {
+	const dest = require_objects.dictGet(annot, "Dest");
+	if (dest !== void 0) return destinationRegistry.intern(dest);
+	const action = resolver.resolveDict(require_objects.dictGet(annot, "A"));
+	if (action !== void 0 && require_objects.asName(require_objects.dictGet(action, "S")) === "GoTo") return destinationRegistry.intern(require_objects.dictGet(action, "D"));
+}
+function readPageNotes(page, resolver) {
+	const annotsArr = require_objects.asArray(require_objects.dictGet(page, "Annots"));
+	if (annotsArr === void 0) return;
+	for (const annotRef of annotsArr) {
+		const annot = resolver.resolveDict(annotRef);
+		if (annot === void 0 || require_objects.asName(require_objects.dictGet(annot, "Subtype")) !== "Text") continue;
+		const titleObj = require_objects.dictGet(annot, "T");
+		if ((titleObj?.kind === "string" ? require_pdf_text.decodePdfString(titleObj.bytes) : void 0) !== "documents.js:notes") continue;
+		const contentsObj = require_objects.dictGet(annot, "Contents");
+		if (contentsObj?.kind === "string") return require_pdf_text.decodePdfString(contentsObj.bytes);
+	}
+}
+//#endregion
+exports.convertEllipse = convertEllipse;
+exports.convertExtractedItem = convertExtractedItem;
+exports.convertImage = convertImage;
+exports.convertLine = convertLine;
+exports.convertPath = convertPath;
+exports.convertRect = convertRect;
+exports.convertText = convertText;
+exports.imagePlacementFrom = imagePlacementFrom;
+exports.paintFields = paintFields;
+exports.readLinkAnnotations = readLinkAnnotations;
+exports.readPageNotes = readPageNotes;
+exports.registerExtractedImage = registerExtractedImage;
+exports.resolveCachedImageId = resolveCachedImageId;
+exports.transformBox = transformBox;
+exports.transformSubpath = transformSubpath;

package/dist/read-convert.d.cts

@@ -0,0 +1,58 @@
+import { i as PdfDiagnosticSink } from "./diagnostics-C9FxfZ6H.cjs";
+import { n as PdfDict } from "./objects-e2RS3-yT.cjs";
+import { r as Matrix } from "./matrix-B7_SBKQ5.cjs";
+import { ExtractedEllipse, ExtractedImage, ExtractedItem, ExtractedLine, ExtractedPaint, ExtractedPath, ExtractedRect, ExtractedSubpath, ExtractedTextRun, PdfObjectResolver } from "./interpret-types.cjs";
+import { LayoutEllipse, LayoutImageAsset, LayoutInternalLink, LayoutItem, LayoutLine, LayoutLink, LayoutPath, LayoutRect, LayoutSubpath, LayoutText } from "./layout.cjs";
+import { FontResolverService } from "./font-read.cjs";
+import { DestinationRegistry } from "./navigation.cjs";
+import { ExtractedPdfImage } from "./images-read.cjs";
+import { Color } from "document-schema.js";
+//#region src/read-convert.d.ts
+declare function convertExtractedItem(item: ExtractedItem, pageMatrix: Matrix, fontResolver: FontResolverService, images: Record<string, LayoutImageAsset>, imageIdCache: Map<PdfDict, string | null>, resolver: Readonly<PdfObjectResolver>, sink: PdfDiagnosticSink): LayoutItem | undefined;
+declare function convertText(item: ExtractedTextRun, pageMatrix: Matrix, fontResolver: FontResolverService): LayoutText | undefined;
+declare function paintFields(paint: ExtractedPaint): {
+  fill?: Color;
+  stroke?: {
+    readonly color: Color;
+    readonly widthPt: number;
+  };
+};
+declare function transformBox(item: Readonly<{
+  xPt: number;
+  yPt: number;
+  widthPt: number;
+  heightPt: number;
+}>, pageMatrix: Matrix): {
+  xPt: number;
+  yPt: number;
+  widthPt: number;
+  heightPt: number;
+};
+declare function convertRect(item: ExtractedRect, pageMatrix: Matrix): LayoutRect;
+declare function convertEllipse(item: ExtractedEllipse, pageMatrix: Matrix): LayoutEllipse;
+declare function convertLine(item: ExtractedLine, pageMatrix: Matrix): LayoutLine;
+declare function transformSubpath(subpath: ExtractedSubpath, pageMatrix: Matrix): LayoutSubpath;
+declare function convertPath(item: ExtractedPath, pageMatrix: Matrix): LayoutPath;
+declare function imagePlacementFrom(matrix: Matrix): {
+  xPt: number;
+  yPt: number;
+  widthPt: number;
+  heightPt: number;
+  rotationDeg: number | undefined;
+};
+declare function registerExtractedImage(format: "png" | "jpeg", bytes: Uint8Array<ArrayBuffer>, widthPx: number, heightPx: number, images: Record<string, LayoutImageAsset>, original: ExtractedPdfImage["original"]): string;
+declare function resolveCachedImageId(dict: PdfDict, raw: Uint8Array<ArrayBuffer>, images: Record<string, LayoutImageAsset>, cache: Map<PdfDict, string | null>, resolver: Readonly<PdfObjectResolver>, sink: PdfDiagnosticSink): string | undefined;
+declare function convertImage(item: ExtractedImage, pageMatrix: Matrix, images: Record<string, LayoutImageAsset>, cache: Map<PdfDict, string | null>, resolver: Readonly<PdfObjectResolver>, sink: PdfDiagnosticSink): {
+  layer?: string | undefined;
+  xPt: number;
+  yPt: number;
+  widthPt: number;
+  heightPt: number;
+  rotationDeg: number | undefined;
+  kind: "image";
+  imageId: string;
+} | undefined;
+declare function readLinkAnnotations(page: PdfDict, pageMatrix: Matrix, resolver: Readonly<PdfObjectResolver>, destinationRegistry: DestinationRegistry): (LayoutLink | LayoutInternalLink)[];
+declare function readPageNotes(page: PdfDict, resolver: Readonly<PdfObjectResolver>): string | undefined;
+//#endregion
+export { convertEllipse, convertExtractedItem, convertImage, convertLine, convertPath, convertRect, convertText, imagePlacementFrom, paintFields, readLinkAnnotations, readPageNotes, registerExtractedImage, resolveCachedImageId, transformBox, transformSubpath };

package/dist/read-convert.d.ts

@@ -0,0 +1,58 @@
+import { i as PdfDiagnosticSink } from "./diagnostics-C9FxfZ6H.js";
+import { n as PdfDict } from "./objects-e2RS3-yT.js";
+import { r as Matrix } from "./matrix-B7_SBKQ5.js";
+import { ExtractedEllipse, ExtractedImage, ExtractedItem, ExtractedLine, ExtractedPaint, ExtractedPath, ExtractedRect, ExtractedSubpath, ExtractedTextRun, PdfObjectResolver } from "./interpret-types.js";
+import { LayoutEllipse, LayoutImageAsset, LayoutInternalLink, LayoutItem, LayoutLine, LayoutLink, LayoutPath, LayoutRect, LayoutSubpath, LayoutText } from "./layout.js";
+import { FontResolverService } from "./font-read.js";
+import { DestinationRegistry } from "./navigation.js";
+import { ExtractedPdfImage } from "./images-read.js";
+import { Color } from "document-schema.js";
+//#region src/read-convert.d.ts
+declare function convertExtractedItem(item: ExtractedItem, pageMatrix: Matrix, fontResolver: FontResolverService, images: Record<string, LayoutImageAsset>, imageIdCache: Map<PdfDict, string | null>, resolver: Readonly<PdfObjectResolver>, sink: PdfDiagnosticSink): LayoutItem | undefined;
+declare function convertText(item: ExtractedTextRun, pageMatrix: Matrix, fontResolver: FontResolverService): LayoutText | undefined;
+declare function paintFields(paint: ExtractedPaint): {
+  fill?: Color;
+  stroke?: {
+    readonly color: Color;
+    readonly widthPt: number;
+  };
+};
+declare function transformBox(item: Readonly<{
+  xPt: number;
+  yPt: number;
+  widthPt: number;
+  heightPt: number;
+}>, pageMatrix: Matrix): {
+  xPt: number;
+  yPt: number;
+  widthPt: number;
+  heightPt: number;
+};
+declare function convertRect(item: ExtractedRect, pageMatrix: Matrix): LayoutRect;
+declare function convertEllipse(item: ExtractedEllipse, pageMatrix: Matrix): LayoutEllipse;
+declare function convertLine(item: ExtractedLine, pageMatrix: Matrix): LayoutLine;
+declare function transformSubpath(subpath: ExtractedSubpath, pageMatrix: Matrix): LayoutSubpath;
+declare function convertPath(item: ExtractedPath, pageMatrix: Matrix): LayoutPath;
+declare function imagePlacementFrom(matrix: Matrix): {
+  xPt: number;
+  yPt: number;
+  widthPt: number;
+  heightPt: number;
+  rotationDeg: number | undefined;
+};
+declare function registerExtractedImage(format: "png" | "jpeg", bytes: Uint8Array<ArrayBuffer>, widthPx: number, heightPx: number, images: Record<string, LayoutImageAsset>, original: ExtractedPdfImage["original"]): string;
+declare function resolveCachedImageId(dict: PdfDict, raw: Uint8Array<ArrayBuffer>, images: Record<string, LayoutImageAsset>, cache: Map<PdfDict, string | null>, resolver: Readonly<PdfObjectResolver>, sink: PdfDiagnosticSink): string | undefined;
+declare function convertImage(item: ExtractedImage, pageMatrix: Matrix, images: Record<string, LayoutImageAsset>, cache: Map<PdfDict, string | null>, resolver: Readonly<PdfObjectResolver>, sink: PdfDiagnosticSink): {
+  layer?: string | undefined;
+  xPt: number;
+  yPt: number;
+  widthPt: number;
+  heightPt: number;
+  rotationDeg: number | undefined;
+  kind: "image";
+  imageId: string;
+} | undefined;
+declare function readLinkAnnotations(page: PdfDict, pageMatrix: Matrix, resolver: Readonly<PdfObjectResolver>, destinationRegistry: DestinationRegistry): (LayoutLink | LayoutInternalLink)[];
+declare function readPageNotes(page: PdfDict, resolver: Readonly<PdfObjectResolver>): string | undefined;
+//#endregion
+export { convertEllipse, convertExtractedItem, convertImage, convertLine, convertPath, convertRect, convertText, imagePlacementFrom, paintFields, readLinkAnnotations, readPageNotes, registerExtractedImage, resolveCachedImageId, transformBox, transformSubpath };