pdf-parser.js 3.0.8 → 4.5.1

package/dist/xmp.js

@@ -0,0 +1,45 @@
+//#region src/xmp.ts
+function readXmpMetadata(packet) {
+	return {
+		...field(xmpValue(packet, "dc:title"), "title"),
+		...field(xmpValue(packet, "dc:creator"), "author"),
+		...field(xmpValue(packet, "dc:description"), "subject"),
+		...field(xmpValue(packet, "xmp:CreatorTool"), "creator"),
+		...field(xmpValue(packet, "pdf:Producer"), "producer"),
+		...field(xmpValue(packet, "xmp:CreateDate"), "createdIso"),
+		...field(xmpValue(packet, "xmp:ModifyDate"), "modifiedIso"),
+		...keywordsOf(packet)
+	};
+}
+function xmpValue(packet, element) {
+	const match = new RegExp(`<${element}(?:\\s[^>]*)?>([\\s\\S]*?)</${element}>`).exec(packet);
+	if (match === null) return;
+	const inner = match[1] ?? "";
+	const items = listItems(inner);
+	if (items.length > 0) return items;
+	const text = inner.trim();
+	return text.length > 0 ? text : void 0;
+}
+function listItems(inner) {
+	const items = [];
+	const pattern = /<rdf:li(?:\s[^>]*)?>([\s\S]*?)<\/rdf:li>/g;
+	let match;
+	while ((match = pattern.exec(inner)) !== null) {
+		const text = (match[1] ?? "").trim();
+		if (text.length > 0) items.push(text);
+	}
+	return items;
+}
+function field(value, key) {
+	if (typeof value === "string") return { [key]: value };
+	if (Array.isArray(value) && value.length > 0) return { [key]: value.join(", ") };
+	return {};
+}
+function keywordsOf(packet) {
+	const match = (/* @__PURE__ */ new RegExp("<dc:subject(?:\\s[^>]*)?>([\\s\\S]*?)</dc:subject>")).exec(packet);
+	if (match === null) return {};
+	const items = listItems(match[1] ?? "");
+	return items.length > 0 ? { keywords: items } : {};
+}
+//#endregion
+export { readXmpMetadata };

package/dist/xref.cjs

@@ -1,8 +1,8 @@
 Object.defineProperty(exports, Symbol.toStringTag, { value: "Module" });
-const require_bytes_reader = require("./bytes/reader.cjs");
-const require_lexer = require("./lexer.cjs");
 const require_objects = require("./objects.cjs");
+const require_bytes_reader = require("./bytes/reader.cjs");
 const require_filters = require("./filters.cjs");
+const require_lexer = require("./lexer.cjs");
 const require_parse = require("./parse.cjs");
 //#region src/xref.ts
 const STARTXREF_BYTES = new TextEncoder().encode("startxref");

package/dist/xref.d.cts

@@ -1,13 +1,13 @@
 import { PdfDiagnosticSink } from "./diagnostics.cjs";
-import { PdfDict } from "./objects.cjs";
+import { n as PdfDict } from "./objects-B7HdMeWS.cjs";
 //#region src/xref.d.ts
 interface XrefOffsetEntry {
-  readonly type: 'offset';
+  readonly type: "offset";
   readonly offset: number;
   readonly gen: number;
 }
 interface XrefCompressedEntry {
-  readonly type: 'compressed';
+  readonly type: "compressed";
   readonly streamObjNum: number;
   readonly indexInStream: number;
 }

package/dist/xref.d.ts

@@ -1,13 +1,13 @@
 import { PdfDiagnosticSink } from "./diagnostics.js";
-import { PdfDict } from "./objects.js";
+import { n as PdfDict } from "./objects-B7HdMeWS.js";
 //#region src/xref.d.ts
 interface XrefOffsetEntry {
-  readonly type: 'offset';
+  readonly type: "offset";
   readonly offset: number;
   readonly gen: number;
 }
 interface XrefCompressedEntry {
-  readonly type: 'compressed';
+  readonly type: "compressed";
   readonly streamObjNum: number;
   readonly indexInStream: number;
 }

package/dist/xref.js

@@ -1,7 +1,7 @@
-import { ByteReader, isAsciiWhitespace } from "./bytes/reader.js";
-import { nextToken } from "./lexer.js";
 import { asArray, asDict, asNumber, dictGet, isName, pdfDict, pdfRef } from "./objects.js";
+import { ByteReader, isAsciiWhitespace } from "./bytes/reader.js";
 import { decodeStream } from "./filters.js";
+import { nextToken } from "./lexer.js";
 import { parseIndirectObject, parseValue } from "./parse.js";
 //#region src/xref.ts
 const STARTXREF_BYTES = new TextEncoder().encode("startxref");

package/package.json

@@ -1,15 +1,16 @@
 {
   "name": "pdf-parser.js",
-  "version": "3.0.8",
+  "version": "4.5.1",
   "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": {
     "type": "git",
-    "url": "git+https://github.com/ExaDev/pdf-codec.git"
+    "url": "git+https://github.com/ExaDev/documents.js.git",
+    "directory": "packages/pdf-codec"
   },
-  "homepage": "https://github.com/ExaDev/pdf-codec#readme",
+  "homepage": "https://github.com/ExaDev/documents.js/tree/main/packages/pdf-codec",
   "bugs": {
-    "url": "https://github.com/ExaDev/pdf-codec/issues"
+    "url": "https://github.com/ExaDev/documents.js/issues"
   },
   "exports": {
     ".": {
@@ -20,6 +21,14 @@
       "import": "./dist/index.js",
       "require": "./dist/index.cjs"
     },
+    "./read": {
+      "types": {
+        "import": "./dist/read.d.ts",
+        "require": "./dist/read.d.cts"
+      },
+      "import": "./dist/read.js",
+      "require": "./dist/read.cjs"
+    },
     "./*": {
       "types": {
         "import": "./dist/*.d.ts",
@@ -32,6 +41,13 @@
   "main": "./dist/index.cjs",
   "module": "./dist/index.js",
   "types": "./dist/index.d.ts",
+  "typesVersions": {
+    "*": {
+      "read": [
+        "./dist/read.d.ts"
+      ]
+    }
+  },
   "files": [
     "dist"
   ],
@@ -56,42 +72,39 @@
   ],
   "license": "MIT",
   "dependencies": {
-    "byte-codec": "^1.1.8",
-    "document-schema.js": "^4.3.3",
-    "fflate": "^0.8.3",
-    "zod": "^4.4.3"
+    "byte-codec": "1.4.1",
+    "document-schema.js": "7.6.0",
+    "fflate": "0.8.3",
+    "zod": "4.4.3"
   },
   "devDependencies": {
-    "@arethetypeswrong/cli": "^0.18.5",
-    "@cloudflare/vitest-pool-workers": "^0.21.2",
-    "@commitlint/cli": "^21.2.1",
-    "@commitlint/config-conventional": "^21.2.0",
-    "@eslint/js": "^10.0.1",
-    "@exadev/eslint-config": "^2.1.0",
-    "@semantic-release/changelog": "^7.0.0",
-    "@semantic-release/git": "^11.0.1",
-    "@types/node": "^26.1.1",
-    "@vitest/coverage-v8": "^4.1.10",
-    "eslint": "^10.8.0",
-    "globals": "^17.8.0",
-    "husky": "^9.1.7",
-    "lint-staged": "^17.2.0",
-    "publint": "^0.3.21",
-    "semantic-release": "^25.0.8",
-    "tsdown": "^0.22.13",
-    "turbo": "^2.10.8",
-    "typescript": "^6.0.3",
-    "typescript-eslint": "^8.65.0",
-    "vitest": "^4.1.10"
+    "@arethetypeswrong/cli": "0.18.5",
+    "@cloudflare/vitest-pool-workers": "0.21.2",
+    "@stryker-mutator/core": "10.0.0",
+    "@stryker-mutator/typescript-checker": "10.0.0",
+    "@stryker-mutator/vitest-runner": "10.0.0",
+    "@types/node": "26.2.0",
+    "@vitest/coverage-v8": "4.1.11",
+    "eslint": "10.8.1",
+    "husky": "9.1.7",
+    "jiti": "2.7.0",
+    "publint": "0.3.22",
+    "semantic-release": "25.0.9",
+    "tsdown": "0.22.14",
+    "turbo": "2.10.8",
+    "typescript": "6.0.3",
+    "typescript-eslint": "8.66.0",
+    "vitest": "4.1.11"
   },
   "scripts": {
     "build": "turbo run _build",
     "_build": "tsdown",
     "lint": "turbo run _lint",
     "_lint": "eslint . --fix --cache --max-warnings 0",
-    "typecheck": "turbo run _typecheck _typecheck:node",
+    "typecheck": "turbo run _typecheck _typecheck:node _typecheck:attw",
     "_typecheck": "tsc -p tsconfig.json",
     "_typecheck:node": "tsc -p tsconfig.node.json",
+    "_typecheck:attw": "attw --pack",
     "test": "turbo run _test",
     "_test": "vitest run --project unit",
     "test:workers": "turbo run _test:workers",
@@ -99,10 +112,12 @@
     "test:watch": "vitest --project unit",
     "test:coverage": "turbo run _test:coverage",
     "_test:coverage": "vitest run --project unit --coverage",
+    "test:mutation": "turbo run _test:mutation",
+    "_test:mutation": "stryker run stryker.config.mjs",
     "test:corpus": "turbo run _test:corpus",
     "_test:corpus": "vitest run --project corpus",
     "test:smoke": "turbo run _test:smoke",
-    "_test:smoke": "tsdown && vitest run --project smoke",
+    "_test:smoke": "vitest run --project smoke",
     "release": "semantic-release"
   }
 }

package/dist/image/png-encode.cjs

@@ -1,64 +0,0 @@
-Object.defineProperty(exports, Symbol.toStringTag, { value: "Module" });
-const require_bytes_crc32 = require("../bytes/crc32.cjs");
-const require_bytes_writer = require("../bytes/writer.cjs");
-const require_bytes_flate = require("../bytes/flate.cjs");
-const require_image_png_filter = require("./png-filter.cjs");
-//#region src/image/png-encode.ts
-const PNG_SIGNATURE = new Uint8Array([
-	137,
-	80,
-	78,
-	71,
-	13,
-	10,
-	26,
-	10
-]);
-function u32be(value) {
-	const bytes = /* @__PURE__ */ new Uint8Array(4);
-	new DataView(bytes.buffer).setUint32(0, value);
-	return bytes;
-}
-function writeChunk(writer, type, data) {
-	const typeBytes = new TextEncoder().encode(type);
-	writer.writeBytes(u32be(data.length));
-	writer.writeBytes(typeBytes);
-	writer.writeBytes(data);
-	writer.writeBytes(u32be(require_bytes_crc32.crc32(require_bytes_writer.concatBytes([typeBytes, data]))));
-}
-function colorTypeFor(image) {
-	if (image.channels === 1) return image.alpha === void 0 ? 0 : 4;
-	return image.alpha === void 0 ? 2 : 6;
-}
-function encodePng(image, options = {}) {
-	const { width, height, channels, data, alpha } = image;
-	const outChannels = alpha === void 0 ? channels : channels + 1;
-	const bytesPerRow = width * outChannels;
-	const pixelCount = width * height;
-	const interleaved = new Uint8Array(pixelCount * outChannels);
-	for (let i = 0; i < pixelCount; i++) {
-		const srcBase = i * channels;
-		const dstBase = i * outChannels;
-		for (let c = 0; c < channels; c++) interleaved[dstBase + c] = data[srcBase + c];
-		if (alpha !== void 0) interleaved[dstBase + channels] = alpha[i];
-	}
-	const filtered = require_image_png_filter.filterScanlines(interleaved, height, bytesPerRow, outChannels, options.filter ?? "adaptive");
-	const compressed = require_bytes_flate.deflate(filtered);
-	const ihdr = /* @__PURE__ */ new Uint8Array(13);
-	const ihdrView = new DataView(ihdr.buffer);
-	ihdrView.setUint32(0, width);
-	ihdrView.setUint32(4, height);
-	ihdr[8] = 8;
-	ihdr[9] = colorTypeFor(image);
-	ihdr[10] = 0;
-	ihdr[11] = 0;
-	ihdr[12] = 0;
-	const writer = new require_bytes_writer.ByteWriter();
-	writer.writeBytes(PNG_SIGNATURE);
-	writeChunk(writer, "IHDR", ihdr);
-	writeChunk(writer, "IDAT", compressed);
-	writeChunk(writer, "IEND", /* @__PURE__ */ new Uint8Array(0));
-	return writer.toBytes();
-}
-//#endregion
-exports.encodePng = encodePng;

package/dist/image/png-encode.d.cts

@@ -1,8 +0,0 @@
-import { RawImage } from "./png-decode.cjs";
-//#region src/image/png-encode.d.ts
-interface PngEncodeOptions {
-  readonly filter?: 'none' | 'adaptive';
-}
-declare function encodePng(image: RawImage, options?: PngEncodeOptions): Uint8Array<ArrayBuffer>;
-//#endregion
-export { PngEncodeOptions, encodePng };

package/dist/image/png-encode.d.ts

@@ -1,8 +0,0 @@
-import { RawImage } from "./png-decode.js";
-//#region src/image/png-encode.d.ts
-interface PngEncodeOptions {
-  readonly filter?: 'none' | 'adaptive';
-}
-declare function encodePng(image: RawImage, options?: PngEncodeOptions): Uint8Array<ArrayBuffer>;
-//#endregion
-export { PngEncodeOptions, encodePng };

package/dist/image/png-encode.js

@@ -1,63 +0,0 @@
-import { crc32 } from "../bytes/crc32.js";
-import { ByteWriter, concatBytes } from "../bytes/writer.js";
-import { deflate } from "../bytes/flate.js";
-import { filterScanlines } from "./png-filter.js";
-//#region src/image/png-encode.ts
-const PNG_SIGNATURE = new Uint8Array([
-	137,
-	80,
-	78,
-	71,
-	13,
-	10,
-	26,
-	10
-]);
-function u32be(value) {
-	const bytes = /* @__PURE__ */ new Uint8Array(4);
-	new DataView(bytes.buffer).setUint32(0, value);
-	return bytes;
-}
-function writeChunk(writer, type, data) {
-	const typeBytes = new TextEncoder().encode(type);
-	writer.writeBytes(u32be(data.length));
-	writer.writeBytes(typeBytes);
-	writer.writeBytes(data);
-	writer.writeBytes(u32be(crc32(concatBytes([typeBytes, data]))));
-}
-function colorTypeFor(image) {
-	if (image.channels === 1) return image.alpha === void 0 ? 0 : 4;
-	return image.alpha === void 0 ? 2 : 6;
-}
-function encodePng(image, options = {}) {
-	const { width, height, channels, data, alpha } = image;
-	const outChannels = alpha === void 0 ? channels : channels + 1;
-	const bytesPerRow = width * outChannels;
-	const pixelCount = width * height;
-	const interleaved = new Uint8Array(pixelCount * outChannels);
-	for (let i = 0; i < pixelCount; i++) {
-		const srcBase = i * channels;
-		const dstBase = i * outChannels;
-		for (let c = 0; c < channels; c++) interleaved[dstBase + c] = data[srcBase + c];
-		if (alpha !== void 0) interleaved[dstBase + channels] = alpha[i];
-	}
-	const filtered = filterScanlines(interleaved, height, bytesPerRow, outChannels, options.filter ?? "adaptive");
-	const compressed = deflate(filtered);
-	const ihdr = /* @__PURE__ */ new Uint8Array(13);
-	const ihdrView = new DataView(ihdr.buffer);
-	ihdrView.setUint32(0, width);
-	ihdrView.setUint32(4, height);
-	ihdr[8] = 8;
-	ihdr[9] = colorTypeFor(image);
-	ihdr[10] = 0;
-	ihdr[11] = 0;
-	ihdr[12] = 0;
-	const writer = new ByteWriter();
-	writer.writeBytes(PNG_SIGNATURE);
-	writeChunk(writer, "IHDR", ihdr);
-	writeChunk(writer, "IDAT", compressed);
-	writeChunk(writer, "IEND", /* @__PURE__ */ new Uint8Array(0));
-	return writer.toBytes();
-}
-//#endregion
-export { encodePng };