@hiraokahypertools/pst-extractor 0.1.0-alpha.1

package/dist/SingleAsyncProvider.js

@@ -0,0 +1,30 @@
+"use strict";
+var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
+    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
+    return new (P || (P = Promise))(function (resolve, reject) {
+        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
+        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
+        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
+        step((generator = generator.apply(thisArg, _arguments || [])).next());
+    });
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.SingleAsyncProvider = void 0;
+class SingleAsyncProvider {
+    constructor() {
+        this._ready = false;
+    }
+    getOrCreate(provider) {
+        return __awaiter(this, void 0, void 0, function* () {
+            if (!this._ready) {
+                this._value = yield provider();
+                this._ready = true;
+            }
+            if (this._value === undefined) {
+                throw new Error("provider must provide non-undefined value");
+            }
+            return this._value;
+        });
+    }
+}
+exports.SingleAsyncProvider = SingleAsyncProvider;

package/dist/TableContext.d.ts

@@ -0,0 +1,4 @@
+import { TableRow } from "./TableRow";
+export interface TableContext {
+    rows(): Promise<TableRow[]>;
+}

package/dist/TableContext.js

@@ -0,0 +1,2 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });

package/dist/TableContextUtil.d.ts

@@ -0,0 +1,4 @@
+import { PHNodeHeap } from "./PHNodeHeap";
+import { PropertyValueResolver } from "./PropertyValueResolver";
+import { TableContext } from "./TableContext";
+export declare function getTableContext(heap: PHNodeHeap, resolver: PropertyValueResolver): Promise<TableContext>;

package/dist/TableContextUtil.js

@@ -0,0 +1,147 @@
+"use strict";
+var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
+    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
+    return new (P || (P = Promise))(function (resolve, reject) {
+        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
+        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
+        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
+        step((generator = generator.apply(thisArg, _arguments || [])).next());
+    });
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.getTableContext = void 0;
+const PLMisc_1 = require("./PLMisc");
+const bTypeTC = 0x7c;
+function copy(rows) {
+    return rows.map(it => Object.assign({}, it));
+}
+function getTableContext(heap, resolver) {
+    return __awaiter(this, void 0, void 0, function* () {
+        if (heap.bClientSig != bTypeTC) {
+            throw new Error("expected type bTypeTC");
+        }
+        const reader = heap.getReader();
+        const headerData = yield reader.getHeapBuffers(heap.userRootHnid);
+        if (headerData.length !== 1) {
+            throw new Error("table context header buffer must be single");
+        }
+        const headerView = new DataView(headerData[0]);
+        const tcSig = headerView.getUint8(0);
+        const numCols = headerView.getUint8(1);
+        const cebOffset = headerView.getUint16(6, true);
+        const numRowBytes = headerView.getUint16(8, true);
+        const hidRowIndex = headerView.getUint32(10, true);
+        const hnidRows = headerView.getUint32(14, true);
+        var tColDesc = headerData[0].slice(22);
+        const schema = Array.from((0, PLMisc_1.splitPer)(tColDesc, 8))
+            .map(it => {
+            const view = new DataView(it);
+            return {
+                type: view.getUint16(0, true),
+                key: view.getUint16(2, true),
+                ibData: view.getUint16(4, true),
+                cbData: view.getUint8(6),
+                iBit: view.getUint8(7),
+            };
+        });
+        if (numCols !== schema.length) {
+            throw new Error("schema length not matched");
+        }
+        if (tcSig !== 0x7c) {
+            throw new Error("tcSig is not 0x7c");
+        }
+        //const min_size = schema.reduce((accum, it) => accum + it.cbData, 0);
+        //const rowIndex = getBTHeapReaderFrom(reader, hidRowIndex);
+        const rows_pages = (hnidRows !== 0)
+            ? yield reader.getHeapBuffers(hnidRows)
+            : [];
+        //const data2 = await reader.getHeapBuffers(offset2);
+        const rows_per_page = (rows_pages.length !== 0)
+            ? Math.floor(rows_pages[0].byteLength / numRowBytes)
+            : 0;
+        function get_record(record_index) {
+            const page_index = (record_index / rows_per_page) | 0;
+            const heap_index = (record_index % rows_per_page) | 0;
+            const buffer = rows_pages[page_index].slice(numRowBytes * (heap_index + 0), numRowBytes * (heap_index + 1));
+            if (buffer.byteLength !== numRowBytes) {
+                throw new Error(`get_record(${record_index}) is reaching EOF while reading record. The heap is ${reader}`);
+            }
+            const ceb = [];
+            {
+                const rgCEB = new Uint8Array(buffer.slice(cebOffset));
+                for (let x = 0; x < numCols; x++) {
+                    ceb.push((rgCEB[(x / 8) | 0] & (1 << (7 - (x % 8)))) != 0);
+                }
+            }
+            return {
+                buffer,
+                ceb,
+            };
+        }
+        const count = rows_pages.reduce((accum, it) => (accum + it.byteLength / numRowBytes) | 0, 0);
+        function listRaw(index) {
+            return __awaiter(this, void 0, void 0, function* () {
+                const record = get_record(index);
+                const list = [];
+                for (let column of schema) {
+                    list.push({
+                        key: column.key,
+                        type: column.type,
+                        value: record.ceb[column.iBit]
+                            ? record.buffer.slice(column.ibData, column.ibData + column.cbData)
+                            : new ArrayBuffer(0),
+                    });
+                }
+                return list;
+            });
+        }
+        function list(record) {
+            return __awaiter(this, void 0, void 0, function* () {
+                const rawProps = yield listRaw(record);
+                const list = [];
+                for (let rawProp of rawProps) {
+                    try {
+                        list.push({
+                            key: rawProp.key,
+                            type: rawProp.type,
+                            value: (rawProp.value.byteLength !== 0)
+                                ? yield resolver.resolveValueOf(rawProp.key, rawProp.type, rawProp.value, reader)
+                                : undefined
+                        });
+                    }
+                    catch (ex) {
+                        throw new Error(`getTableContext.list(rowIndex=${record}) resolving property`
+                            + ` key=0x${rawProp.key.toString(16).padStart(4, '0')}`
+                            + ` type=0x${rawProp.type.toString(16).padStart(4, '0')}`
+                            + ` of ${heap} failure`
+                            + ` --> ${ex}`);
+                    }
+                }
+                return list;
+            });
+        }
+        const rows = [];
+        for (let x = 0; x < count; x++) {
+            rows.push({
+                listRaw() {
+                    return __awaiter(this, void 0, void 0, function* () {
+                        return yield listRaw(x);
+                    });
+                },
+                list() {
+                    return __awaiter(this, void 0, void 0, function* () {
+                        return yield list(x);
+                    });
+                },
+            });
+        }
+        return {
+            rows() {
+                return __awaiter(this, void 0, void 0, function* () {
+                    return copy(rows);
+                });
+            },
+        };
+    });
+}
+exports.getTableContext = getTableContext;

package/dist/TableRow.d.ts

@@ -0,0 +1,6 @@
+import { Property } from "./Property";
+import { RawProperty } from "./RawProperty";
+export interface TableRow {
+    listRaw(): Promise<RawProperty[]>;
+    list(): Promise<Property[]>;
+}

package/dist/TableRow.js

@@ -0,0 +1,2 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });

package/dist/index.d.ts

@@ -0,0 +1,15 @@
+export { PSTFile } from './PSTFile.class';
+export { PSTTask } from './PSTTask.class';
+export { PSTRecipient } from './PSTRecipient.class';
+export { PSTAttachment } from './PSTAttachment.class';
+export { PSTMessage } from './PSTMessage.class';
+export { PSTFolder } from './PSTFolder.class';
+export { openPstFile, openPst, } from './openPstFile';
+export { PSTOpts } from './PSTOpts';
+export { PSTContact } from './PSTContact.class';
+export { PSTAppointment } from './PSTAppointment.class';
+export { PSTMessageStore } from './PSTMessageStore.class';
+export { PSTObject } from './PSTObject.class';
+export { Property } from './Property';
+export { PropertyTypeObject } from './PropertyTypeObject';
+export { ReadFileApi, ReadFile, Close, } from './PLUtil';

package/dist/index.js

@@ -0,0 +1,28 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.PropertyTypeObject = exports.PSTObject = exports.PSTMessageStore = exports.PSTAppointment = exports.PSTContact = exports.openPst = exports.openPstFile = exports.PSTFolder = exports.PSTMessage = exports.PSTAttachment = exports.PSTRecipient = exports.PSTTask = exports.PSTFile = void 0;
+var PSTFile_class_1 = require("./PSTFile.class");
+Object.defineProperty(exports, "PSTFile", { enumerable: true, get: function () { return PSTFile_class_1.PSTFile; } });
+var PSTTask_class_1 = require("./PSTTask.class");
+Object.defineProperty(exports, "PSTTask", { enumerable: true, get: function () { return PSTTask_class_1.PSTTask; } });
+var PSTRecipient_class_1 = require("./PSTRecipient.class");
+Object.defineProperty(exports, "PSTRecipient", { enumerable: true, get: function () { return PSTRecipient_class_1.PSTRecipient; } });
+var PSTAttachment_class_1 = require("./PSTAttachment.class");
+Object.defineProperty(exports, "PSTAttachment", { enumerable: true, get: function () { return PSTAttachment_class_1.PSTAttachment; } });
+var PSTMessage_class_1 = require("./PSTMessage.class");
+Object.defineProperty(exports, "PSTMessage", { enumerable: true, get: function () { return PSTMessage_class_1.PSTMessage; } });
+var PSTFolder_class_1 = require("./PSTFolder.class");
+Object.defineProperty(exports, "PSTFolder", { enumerable: true, get: function () { return PSTFolder_class_1.PSTFolder; } });
+var openPstFile_1 = require("./openPstFile");
+Object.defineProperty(exports, "openPstFile", { enumerable: true, get: function () { return openPstFile_1.openPstFile; } });
+Object.defineProperty(exports, "openPst", { enumerable: true, get: function () { return openPstFile_1.openPst; } });
+var PSTContact_class_1 = require("./PSTContact.class");
+Object.defineProperty(exports, "PSTContact", { enumerable: true, get: function () { return PSTContact_class_1.PSTContact; } });
+var PSTAppointment_class_1 = require("./PSTAppointment.class");
+Object.defineProperty(exports, "PSTAppointment", { enumerable: true, get: function () { return PSTAppointment_class_1.PSTAppointment; } });
+var PSTMessageStore_class_1 = require("./PSTMessageStore.class");
+Object.defineProperty(exports, "PSTMessageStore", { enumerable: true, get: function () { return PSTMessageStore_class_1.PSTMessageStore; } });
+var PSTObject_class_1 = require("./PSTObject.class");
+Object.defineProperty(exports, "PSTObject", { enumerable: true, get: function () { return PSTObject_class_1.PSTObject; } });
+var PropertyTypeObject_1 = require("./PropertyTypeObject");
+Object.defineProperty(exports, "PropertyTypeObject", { enumerable: true, get: function () { return PropertyTypeObject_1.PropertyTypeObject; } });

package/dist/msftUuidStringify.d.ts

@@ -0,0 +1,17 @@
+/**
+ * byte to lower case hex string
+ *
+ * @internal
+ */
+export declare function toHex1(value: number): string;
+/**
+ * Variant 2 UUIDs, historically used in Microsoft's COM/OLE libraries,
+ * use a mixed-endian format, whereby the first three components of the UUID are little-endian,
+ * and the last two are big-endian.
+ * For example, `00112233-4455-6677-8899-aabbccddeeff` is encoded as the bytes
+ * `33 22 11 00 55 44 77 66 88 99 aa bb cc dd ee ff`.
+ *
+ * @see https://en.wikipedia.org/wiki/Universally_unique_identifier
+ * @internal
+ */
+export declare function msftUuidStringify(array: ArrayLike<number>, offset: number): string;

package/dist/msftUuidStringify.js

@@ -0,0 +1,48 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.msftUuidStringify = exports.toHex1 = void 0;
+const hex = "0123456789abcdef";
+/**
+ * byte to lower case hex string
+ *
+ * @internal
+ */
+function toHex1(value) {
+    return hex[(value >> 4) & 15]
+        + hex[(value) & 15];
+}
+exports.toHex1 = toHex1;
+/**
+ * Variant 2 UUIDs, historically used in Microsoft's COM/OLE libraries,
+ * use a mixed-endian format, whereby the first three components of the UUID are little-endian,
+ * and the last two are big-endian.
+ * For example, `00112233-4455-6677-8899-aabbccddeeff` is encoded as the bytes
+ * `33 22 11 00 55 44 77 66 88 99 aa bb cc dd ee ff`.
+ *
+ * @see https://en.wikipedia.org/wiki/Universally_unique_identifier
+ * @internal
+ */
+function msftUuidStringify(array, offset) {
+    return ""
+        + toHex1(array[offset + 3])
+        + toHex1(array[offset + 2])
+        + toHex1(array[offset + 1])
+        + toHex1(array[offset + 0])
+        + "-"
+        + toHex1(array[offset + 5])
+        + toHex1(array[offset + 4])
+        + "-"
+        + toHex1(array[offset + 7])
+        + toHex1(array[offset + 6])
+        + "-"
+        + toHex1(array[offset + 8])
+        + toHex1(array[offset + 9])
+        + "-"
+        + toHex1(array[offset + 10])
+        + toHex1(array[offset + 11])
+        + toHex1(array[offset + 12])
+        + toHex1(array[offset + 13])
+        + toHex1(array[offset + 14])
+        + toHex1(array[offset + 15]);
+}
+exports.msftUuidStringify = msftUuidStringify;

package/dist/openPstFile.d.ts

@@ -0,0 +1,19 @@
+import { PSTFile } from './PSTFile.class';
+import { ReadFileApi } from './PLUtil';
+import { PSTOpts } from './PSTOpts';
+/**
+ * Open pst/ost file from os file path.
+ *
+ * @param path os file path.
+ * @param opts options
+ * @returns
+ */
+export declare function openPstFile(path: string, opts?: PSTOpts): Promise<PSTFile>;
+/**
+ * Open pst/ost file using user defined callback.
+ *
+ * @param api reader callback
+ * @param opts options
+ * @returns
+ */
+export declare function openPst(api: ReadFileApi, opts?: PSTOpts): Promise<PSTFile>;

package/dist/openPstFile.js

@@ -0,0 +1,67 @@
+"use strict";
+var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
+    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
+    return new (P || (P = Promise))(function (resolve, reject) {
+        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
+        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
+        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
+        step((generator = generator.apply(thisArg, _arguments || [])).next());
+    });
+};
+var __importDefault = (this && this.__importDefault) || function (mod) {
+    return (mod && mod.__esModule) ? mod : { "default": mod };
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.openPst = exports.openPstFile = void 0;
+const PSTFile_class_1 = require("./PSTFile.class");
+const fs_1 = __importDefault(require("fs"));
+const PLUtil_1 = require("./PLUtil");
+const PropertyValueResolverV1_1 = require("./PropertyValueResolverV1");
+const iconv_lite_1 = __importDefault(require("iconv-lite"));
+const PAUtil_1 = require("./PAUtil");
+/**
+ * Open pst/ost file from os file path.
+ *
+ * @param path os file path.
+ * @param opts options
+ * @returns
+ */
+function openPstFile(path, opts) {
+    return __awaiter(this, void 0, void 0, function* () {
+        const file = yield fs_1.default.promises.open(path, "r");
+        function readFile(buffer, offset, length, position) {
+            return __awaiter(this, void 0, void 0, function* () {
+                const view = new Uint8Array(buffer);
+                const { bytesRead } = yield file.read(view, offset, length, position);
+                return bytesRead;
+            });
+        }
+        return yield openPst({
+            readFile,
+            close: () => __awaiter(this, void 0, void 0, function* () {
+                yield file.close();
+            }),
+        }, opts);
+    });
+}
+exports.openPstFile = openPstFile;
+/**
+ * Open pst/ost file using user defined callback.
+ *
+ * @param api reader callback
+ * @param opts options
+ * @returns
+ */
+function openPst(api, opts) {
+    return __awaiter(this, void 0, void 0, function* () {
+        const lowPst = yield (0, PLUtil_1.openLowPst)(api);
+        const convertAnsiString = (opts && opts.convertAnsiString)
+            || ((array) => __awaiter(this, void 0, void 0, function* () {
+                return iconv_lite_1.default.decode(Buffer.from(array), (opts && opts.ansiEncoding) || "latin1");
+            }));
+        const resolver = new PropertyValueResolverV1_1.PropertyValueResolverV1(convertAnsiString);
+        const nodeMap = yield (0, PAUtil_1.processNameToIDMap)(yield lowPst.getOneNodeByOrError(97), resolver);
+        return new PSTFile_class_1.PSTFile(lowPst, nodeMap, opts);
+    });
+}
+exports.openPst = openPst;

package/package.json

@@ -0,0 +1,105 @@
+{
+  "name": "@hiraokahypertools/pst-extractor",
+  "version": "0.1.0-alpha.1",
+  "description": "Extract objects from MS Outlook/Exchange PST files",
+  "main": "dist/index.js",
+  "types": "dist/index.d.ts",
+  "engines": {
+    "node": ">=10"
+  },
+  "scripts": {
+    "prebuild": "rimraf dist",
+    "build": "tsc",
+    "test": "run-s test:unit test:build",
+    "test:build": "run-s build",
+    "test:unit": "cross-env CI=1 jest --coverage --reporters=default --reporters=jest-junit"
+  },
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/HiraokaHyperTools/pst-extractor.git"
+  },
+  "author": "kenjiuno",
+  "contributors": [
+    {
+      "name": "Ed Pfromer",
+      "email": "epfromer@gmail.com"
+    }
+  ],
+  "license": "MIT",
+  "devDependencies": {
+    "@babel/core": "^7.17.8",
+    "@babel/plugin-proposal-class-properties": "^7.16.7",
+    "@babel/preset-env": "^7.16.11",
+    "@babel/preset-typescript": "^7.16.7",
+    "@types/debug": "4.1.7",
+    "@types/jest": "^27.4.1",
+    "@types/long": "^4.0.1",
+    "@types/node": "^17.0.23",
+    "@typescript-eslint/eslint-plugin": "^5.18.0",
+    "@typescript-eslint/parser": "^5.18.0",
+    "babel-eslint": "^10.1.0",
+    "babel-jest": "^27.5.1",
+    "cross-env": "^7.0.3",
+    "debug": "^4.3.4",
+    "eslint": "^8.12.0",
+    "eslint-config-prettier": "^8.5.0",
+    "eslint-config-standard": "^16.0.3",
+    "eslint-plugin-import": "^2.26.0",
+    "eslint-plugin-node": "^11.1.0",
+    "eslint-plugin-prettier": "^4.0.0",
+    "eslint-plugin-promise": "^6.0.0",
+    "eslint-plugin-standard": "^5.0.0",
+    "jest": "^27.5.1",
+    "jest-junit": "^13.1.0",
+    "npm-run-all": "^4.1.5",
+    "prettier": "^2.6.2",
+    "rimraf": "^3.0.2",
+    "source-map-support": "^0.5.21",
+    "ts-node": "^10.7.0",
+    "typedoc": "^0.23.10",
+    "typescript": "^4.6.3"
+  },
+  "dependencies": {
+    "iconv-lite": "^0.6.3",
+    "long": "^5.2.0"
+  },
+  "eslintConfig": {
+    "extends": [
+      "plugin:@typescript-eslint/recommended",
+      "prettier/@typescript-eslint",
+      "plugin:prettier/recommended"
+    ],
+    "parserOptions": {
+      "ecmaVersion": 2018,
+      "sourceType": "module"
+    }
+  },
+  "prettier": {
+    "semi": false,
+    "singleQuote": true,
+    "endOfLine": "auto"
+  },
+  "babel": {
+    "presets": [
+      "@babel/preset-env",
+      "@babel/preset-typescript"
+    ],
+    "plugins": [
+      "@babel/plugin-proposal-class-properties"
+    ]
+  },
+  "jest": {
+    "testMatch": [
+      "<rootDir>/src/__tests__/**/*.spec.ts"
+    ],
+    "collectCoverageFrom": [
+      "<rootDir>/src/**/*.ts",
+      "!<rootDir>/src/__tests__/**/*.*"
+    ],
+    "coverageReporters": [
+      "text",
+      "cobertura"
+    ]
+  },
+  "files": ["dist"]
+}

package/readme.md

@@ -0,0 +1,16 @@
+# pst-extractor
+
+Extract objects from MS Outlook/Exchange PST files
+
+This is a private fork of https://github.com/epfromer/pst-extractor repository.
+
+Links: [typedoc](https://hiraokahypertools.github.io/pst-extractor/typedoc/)
+
+## Authors
+
+- kenjiuno
+- Ed Pfromer (epfromer@gmail.com)
+
+## License
+
+MIT © [epfromer](https://github.com/epfromer), kenjiuno