Add support for reading from a WebStreams

- Stream Blob via a WebStreams, instead of buffering the full content
- Update strtok3 to v7.0.0

Author: Borewit

.github/workflows/main.yml

@@ -10,8 +10,8 @@ jobs:
       fail-fast: false
       matrix:
         node-version:
+          - 22
           - 20
-          - 18
     steps:
       - uses: actions/checkout@v4
       - uses: actions/setup-node@v4

browser.js

@@ -1,10 +1,8 @@
-import {ReadableWebToNodeStream} from 'readable-web-to-node-stream';
-import {fileTypeFromStream as coreFileTypeFromStream} from './core.js';
+import {fileTypeFromWebStream} from './core.js';
 
-export async function fileTypeFromStream(stream) {
-	const readableWebToNodeStream = new ReadableWebToNodeStream(stream);
-	const fileType = await coreFileTypeFromStream(readableWebToNodeStream);
-	await readableWebToNodeStream.close();
+export async function fileTypeFromStream(webStream) {
+	const fileType = await fileTypeFromWebStream(webStream);
+	await webStream.close();
 	return fileType;
 }
 

core.js

@@ -10,6 +10,10 @@ import {extensions, mimeTypes} from './supported.js';
 
 const minimumBytes = 4100; // A fair amount of file-types are detectable within this range.
 
+export async function fileTypeFromWebStream(webStream) {
+	return new FileTypeParser().fromWebStream(webStream);
+}
+
 export async function fileTypeFromStream(stream) {
 	return new FileTypeParser().fromStream(stream);
 }
@@ -88,8 +92,7 @@ export class FileTypeParser {
 	}
 
 	async fromBlob(blob) {
-		const buffer = await blob.arrayBuffer();
-		return this.fromBuffer(new Uint8Array(buffer));
+		return this.fromWebStream(blob.stream());
 	}
 
 	async fromStream(stream) {
@@ -101,6 +104,15 @@ export class FileTypeParser {
 		}
 	}
 
+	async fromWebStream(webStream) {
+		const tokenizer = await strtok3.fromWebStream(webStream);
+		try {
+			return await this.fromTokenizer(tokenizer);
+		} finally {
+			await tokenizer.close();
+		}
+	}
+
 	async toDetectionStream(readableStream, options = {}) {
 		const {default: stream} = await import('node:stream');
 		const {sampleSize = minimumBytes} = options;
@@ -576,7 +588,7 @@ export class FileTypeParser {
 		) {
 			// They all can have MIME `video/mp4` except `application/mp4` special-case which is hard to detect.
 			// For some cases, we're specific, everything else falls to `video/mp4` with `mp4` extension.
-			const brandMajor = this.buffer.toString('binary', 8, 12).replace('\0', ' ').trim();
+			const brandMajor = this.buffer.toString('latin1', 8, 12).replace('\0', ' ').trim();
 			switch (brandMajor) {
 				case 'avif':
 				case 'avis':
@@ -1059,7 +1071,7 @@ export class FileTypeParser {
 		}
 
 		if (this.checkString('AC')) {
-			const version = this.buffer.toString('binary', 2, 6);
+			const version = this.buffer.toString('latin1', 2, 6);
 			if (version.match('^d*') && version >= 1000 && version <= 1050) {
 				return {
 					ext: 'dwg',
@@ -1126,7 +1138,7 @@ export class FileTypeParser {
 			async function readChunkHeader() {
 				return {
 					length: await tokenizer.readToken(Token.INT32_BE),
-					type: await tokenizer.readToken(new Token.StringType(4, 'binary')),
+					type: await tokenizer.readToken(new Token.StringType(4, 'latin1')),
 				};
 			}
 

package.json

@@ -210,9 +210,8 @@
 		"fbx"
 	],
 	"dependencies": {
-		"readable-web-to-node-stream": "^3.0.2",
-		"strtok3": "^7.0.0",
-		"token-types": "^5.0.1"
+		"strtok3": "^7.1.0",
+		"token-types": "^6.0.0"
 	},
 	"devDependencies": {
 		"@tokenizer/token": "^0.3.0",