add options to make reading ZIP archives less strict

aternosorg · Dec 10, 2024 · f95bf80 · f95bf80
1 parent 4ed35d0
commit f95bf80
Show file tree

Hide file tree

Showing 13 changed files with 130 additions and 34 deletions.
diff --git a/README.md b/README.md
@@ -58,11 +58,13 @@ await archive.init();
 The ReadArchive constructor optionally accepts an [ReadArchiveOptions](src/Options/ReadArchiveOptions.js) object with
 the following properties:
 
-| Name                         | Type                                        | Description                                                                                                                                                                  |
-|------------------------------|---------------------------------------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
-| `centralDirectoryBufferSize` | number                                      | Buffer size used when reading central directory contents.<br/>Larger buffer sizes may improve performance, but also increase RAM usage.                                      |
-| `createEntryIndex`           | boolean                                     | Whether an index of all central directory entries should be created the first time they are read.<br/>Massively increases performance when using `findEntry` multiple times. |
-| `entryOptions`               | [EntryOptions](src/Options/EntryOptions.js) | Options passed to each created Entry object.                                                                                                                                 |
+| Name                             | Type                                        | Description                                                                                                                                                                  |
+|----------------------------------|---------------------------------------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
+| `centralDirectoryBufferSize`     | number                                      | Buffer size used when reading central directory contents.<br/>Larger buffer sizes may improve performance, but also increase RAM usage.                                      |
+| `createEntryIndex`               | boolean                                     | Whether an index of all central directory entries should be created the first time they are read.<br/>Massively increases performance when using `findEntry` multiple times. |
+| `entryOptions`                   | [EntryOptions](src/Options/EntryOptions.js) | Options passed to each created Entry object.                                                                                                                                 |
+| `ignoreMultiDiskErrors`          | boolean                                     | Simply ignore information about multiple disks instead of throwing an error when encountering a multi disk archive                                                           |
+| `allowTruncatedCentralDirectory` | boolean                                     | Do not throw an error if the central directory does not contain the expected numbber of entries                                                                              |
 
 [EntryOptions](src/Options/EntryOptions.js) can have the following properties:
 
@@ -133,6 +135,14 @@ file. Since this data is decompressed, the size of the returned chunk might diff
 Also note that an empty chunk returned from `EntryDataReader.read` does not necessarily indicate that all data has been read.
 After all data was read, `null` will be returned instead.
 
+Both `getDataReader` and `getData` optionally accept an [EntryDataReaderOptions](src/Options/EntryDataReaderOptions.js) object with
+the following properties:
+
+| Name                            | Type    | Description                                                                     |
+|---------------------------------|---------|---------------------------------------------------------------------------------|
+| `ignoreInvalidChecksums`        | boolean | Do not throw an error if the uncompressed data does not match the checksum      |
+| `ignoreInvalidUncompressedSize` | boolean | Do not throw an error if the uncompressed data does not match the expected size |
+
 ### Writing archives
 
 New archives can be created using a [WriteArchive](src/Archive/WriteArchive.js) object.

diff --git a/index-browser.js b/index-browser.js
@@ -34,11 +34,13 @@ export { default as UnicodeExtraField } from "./src/Archive/Structure/ExtraField
 
 export { default as CP437 } from "./src/Util/CP437.js";
 export { default as MsDosTime } from "./src/Util/MsDosTime.js";
+export { default as Resize } from "./src/Util/Resize.js";
 
 export { default as Options } from "./src/Options/Options.js";
 export { default as EntrySourceOptions } from "./src/Options/EntrySourceOptions.js";
 export { default as ReadArchiveOptions } from "./src/Options/ReadArchiveOptions.js";
 export { default as WriteArchiveOptions } from "./src/Options/WriteArchiveOptions.js";
+export { default as EntryDataReaderOptions } from "./src/Options/EntryDataReaderOptions.js";
 
 export { default as ArmariusError } from "./src/Error/ArmariusError.js";
 export { default as ChecksumError } from "./src/Error/ChecksumError.js";

diff --git a/index.js b/index.js
@@ -34,11 +34,13 @@ export { default as UnicodeExtraField } from "./src/Archive/Structure/ExtraField
 
 export { default as CP437 } from "./src/Util/CP437.js";
 export { default as MsDosTime } from "./src/Util/MsDosTime.js";
+export { default as Resize } from "./src/Util/Resize.js";
 
 export { default as Options } from "./src/Options/Options.js";
 export { default as EntrySourceOptions } from "./src/Options/EntrySourceOptions.js";
 export { default as ReadArchiveOptions } from "./src/Options/ReadArchiveOptions.js";
 export { default as WriteArchiveOptions } from "./src/Options/WriteArchiveOptions.js";
+export { default as EntryDataReaderOptions } from "./src/Options/EntryDataReaderOptions.js";
 
 export { default as ArmariusError } from "./src/Error/ArmariusError.js";
 export { default as ChecksumError } from "./src/Error/ChecksumError.js";

diff --git a/package-lock.json b/package-lock.json
diff --git a/package.json b/package.json
@@ -1,6 +1,6 @@
 {
   "name": "armarius",
-  "version": "2.1.2",
+  "version": "2.2.0",
   "description": "A JavaScript library to read, write, and merge ZIP archives in web browsers.",
   "repository": "github:aternosorg/armarius",
   "type": "module",

diff --git a/src/Archive/Entry/ArchiveEntry.js b/src/Archive/Entry/ArchiveEntry.js
@@ -10,6 +10,8 @@ import EntryOptions from '../../Options/EntryOptions.js';
 import {BigInt, CRC32} from 'armarius-io';
 import FeatureError from '../../Error/FeatureError.js';
 import ArmariusError from '../../Error/ArmariusError.js';
+import ZipError from '../../Error/ZipError.js';
+import Resize from '../../Util/Resize.js';
 
 const decoder = new TextDecoder();
 
@@ -230,35 +232,50 @@ export default class ArchiveEntry {
     }
 
     /**
+     * @param {EntryDataReaderOptions|EntryDataReaderOptionsObject} options
      * @returns {Promise<EntryDataReader>}
      */
-    async getDataReader() {
+    async getDataReader(options = {}) {
         if (this.isDirectory()) {
             throw new ArmariusError(`Cannot create data reader: ${this.getFileNameString()} is a directory`);
         }
         await this.readLocalFileHeader();
         return new EntryDataReader(
             await this.getDataProcessor(),
             this.centralDirectoryFileHeader.crc32,
-            Number(this.getUncompressedSize())
+            Number(this.getUncompressedSize()),
+            options
         );
     }
 
     /**
      * @param {number} chunkSize
+     * @param {EntryDataReaderOptions|EntryDataReaderOptionsObject} options
      * @returns {Promise<Uint8Array>}
      */
-    async getData(chunkSize = 1024 * 64) {
+    async getData(chunkSize = 1024 * 64, options = {}) {
         let res = new Uint8Array(Number(this.getUncompressedSize()));
         let offset = 0;
-        let reader = await this.getDataReader();
+        let reader = await this.getDataReader(options);
 
         let chunk;
         while ((chunk = await reader.read(chunkSize)) !== null) {
+            if (offset + chunk.byteLength > res.byteLength) {
+                if (!options.ignoreInvalidUncompressedSize) {
+                    throw new ZipError(`Data size exceeds expected uncompressed size.`);
+                }
+                res = Resize.resizeBuffer(res, res.byteLength + chunkSize * 8);
+            }
             res.set(chunk, offset);
             offset += chunk.byteLength;
         }
 
+        if (offset < res.byteLength) {
+            if (!options.ignoreInvalidUncompressedSize) {
+                throw new ZipError(`Data size is less than expected uncompressed size.`);
+            }
+            return new Uint8Array(res.buffer, res.byteOffset, offset);
+        }
         return res;
     }
 

diff --git a/src/Archive/Entry/EntryDataReader.js b/src/Archive/Entry/EntryDataReader.js
@@ -1,23 +1,27 @@
 import ChecksumError from '../../Error/ChecksumError.js';
 import {DataStream} from 'armarius-io';
+import EntryDataReaderOptions from '../../Options/EntryDataReaderOptions.js';
 
 export default class EntryDataReader extends DataStream {
     /** @type {import("armarius-io").DataProcessor} */ dataProcessor;
     /** @type {import("armarius-io").CRC32} */ crc32;
     /** @type {number} */ expectedCrc32;
     /** @type {number} */ expectedSize;
     /** @type {number} */ offset = 0;
+    /** @type {EntryDataReaderOptions} */ options;
 
     /**
      * @param {import("armarius-io").DataProcessor} dataProcessor
      * @param {number} expectedCrc32
      * @param {number} expectedSize
+     * @param {EntryDataReaderOptions|EntryDataReaderOptionsObject} options
      */
-    constructor(dataProcessor, expectedCrc32, expectedSize) {
+    constructor(dataProcessor, expectedCrc32, expectedSize, options) {
         super();
         this.dataProcessor = dataProcessor;
         this.expectedCrc32 = expectedCrc32;
         this.expectedSize = expectedSize;
+        this.options = EntryDataReaderOptions.from(options);
     }
 
     /**
@@ -44,7 +48,9 @@ export default class EntryDataReader extends DataStream {
 
         if (this.dataProcessor.getPostCrc()) {
             if (eof && this.dataProcessor.getPostCrc().finish() !== this.expectedCrc32) {
-                throw new ChecksumError('CRC32 checksum does not match expected value');
+                if (!this.options.ignoreInvalidChecksums) {
+                    throw new ChecksumError(`CRC32 checksum does not match expected value. Expected ${this.expectedCrc32}, got ${this.dataProcessor.getPostCrc().finish()}`);
+                }
             }
         }
 

diff --git a/src/Archive/Entry/EntryIterator.js b/src/Archive/Entry/EntryIterator.js
@@ -1,10 +1,12 @@
 import ArchiveIndex from "../../Index/ArchiveIndex.js";
 import ArchiveEntry from "./ArchiveEntry.js";
 import {BigInt} from 'armarius-io';
+import ZipError from '../../Error/ZipError.js';
 
 export default class EntryIterator {
     /** @type {ReadArchive} */ archive;
     /** @type {BigInt} */ entryCount;
+    /** @type {BigInt} */ size;
     /** @type {number} */ startOffset;
     /** @type {BigInt} */ currentEntry;
     /** @type {import("armarius-io").IO} */ io;
@@ -22,6 +24,7 @@ export default class EntryIterator {
         this.createIndex = createIndex;
         this.startOffset = io.offset;
         this.entryCount = archive.getCentralDirectoryEntryCount();
+        this.size = archive.getCentralDirectorySize();
         this.reset();
     }
 
@@ -53,6 +56,12 @@ export default class EntryIterator {
         if (this.currentEntry >= this.entryCount) {
             return null;
         }
+        if (this.io.offset >= BigInt(this.startOffset) + this.size) {
+            if (!this.archive.options.allowTruncatedCentralDirectory) {
+                throw new ZipError("Reached end of central directory data before all entries were read");
+            }
+            return null;
+        }
         let offset = this.io.offset;
         let entry = await ArchiveEntry.load(this.archive, this.io, offset);
         this.currentEntry++;

diff --git a/src/Archive/ReadArchive.js b/src/Archive/ReadArchive.js
@@ -55,7 +55,9 @@ export default class ReadArchive {
         this.prependedDataLength = 0;
 
         if(this.endOfCentralDirectoryRecord.diskNumber !== 0 || this.endOfCentralDirectoryRecord.centralDirectoryDiskNumber !== 0) {
-            throw new FeatureError('Multi disk archives are not supported');
+            if (!this.options.ignoreMultiDiskErrors) {
+                throw new FeatureError('Multi disk archives are not supported');
+            }
         }
 
         /*
@@ -134,7 +136,7 @@ export default class ReadArchive {
         );
         this.endOfCentralDirectoryLocator64 = await EndOfCentralDirectoryLocator64.fromIO(endOfDirectoryLocatorReader);
 
-        if(this.endOfCentralDirectoryLocator64.disks > 1) {
+        if(this.endOfCentralDirectoryLocator64.disks > 1 && !this.options.ignoreMultiDiskErrors) {
             throw new FeatureError('Multi disk archives are not supported');
         }
 
@@ -167,7 +169,10 @@ export default class ReadArchive {
 
         if(this.endOfCentralDirectoryRecord64.diskNumber !== 0 ||
             this.endOfCentralDirectoryRecord64.centralDirectoryDiskNumber !== 0) {
-            throw new FeatureError('Multi disk archives are not supported');
+
+            if (!this.options.ignoreMultiDiskErrors) {
+                throw new FeatureError('Multi disk archives are not supported');
+            }
         }
 
         this.centralDirectoryByteLength = Number(this.endOfCentralDirectoryRecord64.centralDirectorySize);

diff --git a/src/Options/EntryDataReaderOptions.js b/src/Options/EntryDataReaderOptions.js
@@ -0,0 +1,13 @@
+import Options from './Options.js';
+
+/**
+ * @typedef {Object} EntryDataReaderOptionsObject
+ * @property {boolean} [ignoreInvalidChecksums]
+ * @property {boolean} [ignoreInvalidUncompressedSize]
+ */
+
+export default class EntryDataReaderOptions extends Options {
+    /** @type {boolean} */ ignoreInvalidChecksums = false;
+    /** @type {boolean} */ ignoreInvalidUncompressedSize = false;
+}
+
diff --git a/src/Options/EntryOptions.js b/src/Options/EntryOptions.js
@@ -8,10 +8,10 @@ defaultDataProcessors.set(Constants.COMPRESSION_METHOD_DEFLATE, DefaultInflateDa
 
 /**
  * @typedef {Object} EntryOptionsObject
- * @property {Map<number, typeof DataProcessor>} [dataProcessors]
+ * @property {Map<number, typeof import("armarius-io").DataProcessor>} [dataProcessors]
  */
 
 export default class EntryOptions extends Options {
-    /** @type {Map<number, typeof DataProcessor>} */ dataProcessors = defaultDataProcessors;
+    /** @type {Map<number, typeof import("armarius-io").DataProcessor>} */ dataProcessors = defaultDataProcessors;
 }
 
diff --git a/src/Options/ReadArchiveOptions.js b/src/Options/ReadArchiveOptions.js
@@ -6,12 +6,16 @@ import Constants from '../Constants.js';
  * @property {number} [centralDirectoryBufferSize]
  * @property {boolean} [createEntryIndex]
  * @property {EntryOptions|EntryOptionsObject} [entryOptions]
+ * @property {boolean} [ignoreMultiDiskErrors]
+ * @property {boolean} [allowTruncatedCentralDirectory]
  */
 
 
-export default class ReadArchiveOptions extends Options{
+export default class ReadArchiveOptions extends Options {
     /** @type {number} */ centralDirectoryBufferSize = Constants.DEFAULT_CHUNK_SIZE;
     /** @type {boolean} */ createEntryIndex = true;
     /** @type {EntryOptions|EntryOptionsObject} */ entryOptions = {};
+    /** @type {boolean} */ ignoreMultiDiskErrors = false;
+    /** @type {boolean} */ allowTruncatedCentralDirectory = false;
 }