Merge pull request #4679 from remotion-dev/4668-media-parser-ebml-id-…

…0x536e-is-not-supported

packages/convert/app/components/Probe.tsx

@@ -112,12 +112,10 @@ export const Probe: React.FC<{
 							)}
 						</CardTitle>
 						{error ? (
-							<CardDescription className="!mt-0">
-								<p className="text-red-500">
-									Failed to parse media:
-									<br />
-									{error.message}
-								</p>
+							<CardDescription className="!mt-0 text-red-500">
+								Failed to parse media:
+								<br />
+								{error.message}
 							</CardDescription>
 						) : null}
 						<CardDescription

packages/media-parser/src/boxes/iso-base-media/ftyp.ts

@@ -17,13 +17,13 @@ export const parseFtyp = ({
 	size: number;
 	offset: number;
 }): FtypBox => {
-	const majorBrand = iterator.getByteString(4);
+	const majorBrand = iterator.getByteString(4, false);
 	const minorVersion = iterator.getFourByteNumber();
 
 	const types = (size - iterator.counter.getOffset()) / 4;
 	const compatibleBrands: string[] = [];
 	for (let i = 0; i < types; i++) {
-		compatibleBrands.push(iterator.getByteString(4).trim());
+		compatibleBrands.push(iterator.getByteString(4, false).trim());
 	}
 
 	const offsetAtEnd = iterator.counter.getOffset();

packages/media-parser/src/boxes/iso-base-media/meta/hdlr.ts

@@ -26,7 +26,7 @@ export const parseHdlr = ({
 	// predefined
 	iterator.discard(4);
 	// type
-	const hdlrType = iterator.getByteString(4);
+	const hdlrType = iterator.getByteString(4, false);
 	// component manufactor
 	iterator.discard(4);
 	// component flags

packages/media-parser/src/boxes/iso-base-media/meta/ilst.ts

@@ -31,7 +31,7 @@ const parseFromWellKnownType = (
 	size: number,
 ): Value => {
 	if (wellKnownType === 1) {
-		const value = iterator.getByteString(size);
+		const value = iterator.getByteString(size, false);
 		return {type: 'text', value};
 	}
 

packages/media-parser/src/boxes/iso-base-media/parse-icc-profile.ts

@@ -44,28 +44,28 @@ export const parseIccProfile = (data: Uint8Array): IccProfile => {
 		throw new Error('Invalid ICC profile size');
 	}
 
-	const preferredCMMType = iterator.getByteString(4);
-	const profileVersion = iterator.getByteString(4);
-	const profileDeviceClass = iterator.getByteString(4);
-	const colorSpace = iterator.getByteString(4);
-	const pcs = iterator.getByteString(4);
+	const preferredCMMType = iterator.getByteString(4, false);
+	const profileVersion = iterator.getByteString(4, false);
+	const profileDeviceClass = iterator.getByteString(4, false);
+	const colorSpace = iterator.getByteString(4, false);
+	const pcs = iterator.getByteString(4, false);
 	const dateTime = iterator.getSlice(12);
-	const signature = iterator.getByteString(4);
+	const signature = iterator.getByteString(4, false);
 	if (signature !== 'acsp') {
 		throw new Error('Invalid ICC profile signature');
 	}
 
-	const primaryPlatform = iterator.getByteString(4);
+	const primaryPlatform = iterator.getByteString(4, false);
 	const profileFlags = iterator.getUint32();
-	const deviceManufacturer = iterator.getByteString(4);
-	const deviceModel = iterator.getByteString(4);
+	const deviceManufacturer = iterator.getByteString(4, false);
+	const deviceModel = iterator.getByteString(4, false);
 	const deviceAttributes = iterator.getUint64();
 	const renderingIntent = iterator.getUint32();
 	const pcsIlluminant1 = iterator.getUint32();
 	const pcsIlluminant2 = iterator.getUint32();
 	const pcsIlluminant3 = iterator.getUint32();
-	const profileCreator = iterator.getByteString(4);
-	const profileId = iterator.getByteString(16);
+	const profileCreator = iterator.getByteString(4, false);
+	const profileId = iterator.getByteString(16, false);
 
 	// reserved
 	iterator.discard(28);
@@ -74,7 +74,7 @@ export const parseIccProfile = (data: Uint8Array): IccProfile => {
 	const entries: Entry[] = [];
 	for (let i = 0; i < tagCount; i++) {
 		const entry: Entry = {
-			tag: iterator.getByteString(4),
+			tag: iterator.getByteString(4, false),
 			offset: iterator.getUint32(),
 			size: iterator.getUint32(),
 		};

packages/media-parser/src/boxes/iso-base-media/process-box.ts

@@ -192,7 +192,7 @@ export const processBox = async ({
 		};
 	}
 
-	const boxType = iterator.getByteString(4);
+	const boxType = iterator.getByteString(4, false);
 
 	const boxSize = boxSizeRaw === 1 ? iterator.getEightByteNumber() : boxSizeRaw;
 

packages/media-parser/src/boxes/iso-base-media/stsd/colr.ts

@@ -27,7 +27,7 @@ export const parseColorParameterBox = ({
 	iterator: BufferIterator;
 	size: number;
 }): ColorParameterBox => {
-	const byteString = iterator.getByteString(4);
+	const byteString = iterator.getByteString(4, false);
 
 	if (byteString === 'nclx') {
 		const primaries = iterator.getUint16();

packages/media-parser/src/boxes/iso-base-media/stsd/keys.ts

@@ -35,7 +35,7 @@ export const parseKeys = ({
 		// key_size
 		const keySize = iterator.getUint32();
 		const namespace = iterator.getAtom();
-		const value = iterator.getByteString(keySize - 8);
+		const value = iterator.getByteString(keySize - 8, false);
 		// data
 		const entry: KeysEntry = {
 			keySize,

packages/media-parser/src/boxes/riff/expect-riff-box.ts

@@ -36,7 +36,7 @@ export const expectRiffBox = async ({
 		};
 	}
 
-	const ckId = iterator.getByteString(4);
+	const ckId = iterator.getByteString(4, false);
 	const ckSize = iterator.getUint32Le();
 
 	if (isMoviAtom(iterator, ckId)) {

packages/media-parser/src/boxes/riff/is-movi.ts

@@ -5,7 +5,7 @@ export const isMoviAtom = (iterator: BufferIterator, ckId: string): boolean => {
 		return false;
 	}
 
-	const listType = iterator.getByteString(4);
+	const listType = iterator.getByteString(4, false);
 	iterator.counter.decrement(4);
 	return listType === 'movi';
 };

packages/media-parser/src/boxes/riff/parse-box.ts

@@ -184,7 +184,7 @@ export const parseRiff = ({
 	state: ParserState;
 	fields: Options<ParseMediaFields>;
 }): Promise<ParseResult> => {
-	const riff = iterator.getByteString(4);
+	const riff = iterator.getByteString(4, false);
 	if (riff !== 'RIFF') {
 		throw new Error('Not a RIFF file');
 	}
@@ -195,7 +195,7 @@ export const parseRiff = ({
 	}
 
 	const size = iterator.getUint32Le();
-	const fileType = iterator.getByteString(4);
+	const fileType = iterator.getByteString(4, false);
 	if (fileType !== 'WAVE' && fileType !== 'AVI') {
 		throw new Error(`File type ${fileType} not supported`);
 	}

packages/media-parser/src/boxes/riff/parse-isft.ts

@@ -9,7 +9,7 @@ export const parseIsft = ({
 	size: number;
 }): IsftBox => {
 	const {expectNoMoreBytes} = iterator.startBox(size);
-	const software = iterator.getByteString(size - 1);
+	const software = iterator.getByteString(size - 1, false);
 	const last = iterator.getUint8();
 	if (last !== 0) {
 		throw new Error(`Expected 0 byte, got ${last}`);

packages/media-parser/src/boxes/riff/parse-list-box.ts

@@ -14,7 +14,7 @@ export const parseListBox = async ({
 	state: ParserState;
 }): Promise<RiffBox> => {
 	const counter = iterator.counter.getOffset();
-	const listType = iterator.getByteString(4);
+	const listType = iterator.getByteString(4, false);
 
 	if (listType === 'movi') {
 		throw new Error('should not be handled here');

packages/media-parser/src/boxes/riff/parse-movi.ts

@@ -148,7 +148,7 @@ export const parseMovi = async ({
 			};
 		}
 
-		const ckId = iterator.getByteString(4);
+		const ckId = iterator.getByteString(4, false);
 		const ckSize = iterator.getUint32Le();
 
 		if (

packages/media-parser/src/boxes/riff/parse-strf.ts

@@ -44,7 +44,7 @@ const parseStrfVideo = ({
 	const height = iterator.getInt32Le();
 	const planes = iterator.getUint16Le();
 	const bitCount = iterator.getUint16Le();
-	const compression = iterator.getByteString(4);
+	const compression = iterator.getByteString(4, false);
 	const sizeImage = iterator.getUint32Le();
 	const xPelsPerMeter = iterator.getInt32Le();
 	const yPelsPerMeter = iterator.getInt32Le();

packages/media-parser/src/boxes/riff/parse-strh.ts

@@ -9,13 +9,15 @@ export const parseStrh = ({
 	size: number;
 }): RiffBox => {
 	const box = iterator.startBox(size);
-	const fccType = iterator.getByteString(4);
+	const fccType = iterator.getByteString(4, false);
 	if (fccType !== 'vids' && fccType !== 'auds') {
 		throw new Error('Expected AVI handler to be vids / auds');
 	}
 
 	const handler =
-		fccType === 'vids' ? iterator.getByteString(4) : iterator.getUint32Le();
+		fccType === 'vids'
+			? iterator.getByteString(4, false)
+			: iterator.getUint32Le();
 	if (typeof handler === 'string' && handler !== 'H264') {
 		throw new Error(
 			`Only H264 is supported as a stream type in .avi, got ${handler}`,

packages/media-parser/src/boxes/webm/description.ts

@@ -67,7 +67,7 @@ export const getAudioDescription = (
 	// type
 	bufferIterator.getUint8();
 	// vorbis
-	const vorbis = bufferIterator.getByteString(6);
+	const vorbis = bufferIterator.getByteString(6, false);
 	if (vorbis !== 'vorbis') {
 		throw new Error('Error parsing vorbis codec private');
 	}

packages/media-parser/src/boxes/webm/parse-ebml.ts

@@ -57,7 +57,7 @@ export const parseEbml = async (
 	}
 
 	if (hasInMap.type === 'string') {
-		const value = iterator.getByteString(size);
+		const value = iterator.getByteString(size, true);
 
 		return {
 			type: hasInMap.name,

packages/media-parser/src/boxes/webm/segments/all-segments.ts

@@ -407,6 +407,11 @@ export const seekId = {
 	type: 'hex-string',
 } as const satisfies Ebml;
 
+export const _name = {
+	name: 'Name',
+	type: 'string',
+} as const satisfies Ebml;
+
 export const seekPosition = {
 	name: 'SeekPosition',
 	type: 'uint',
@@ -622,6 +627,11 @@ export const referenceBlock = {
 	type: 'uint',
 } as const satisfies Ebml;
 
+export const blockDurationSegment = {
+	name: 'BlockDuration',
+	type: 'uint',
+} as const satisfies Ebml;
+
 export const blockElement = {
 	name: 'Block',
 	type: 'uint8array',
@@ -838,6 +848,7 @@ export const ebmlMap = {
 	[matroskaElements.SeekHead]: seekHead,
 	[matroskaElements.Seek]: seek,
 	[matroskaElements.SeekID]: seekId,
+	[matroskaElements.Name]: _name,
 	[matroskaElements.SeekPosition]: seekPosition,
 	[matroskaElements.Crc32]: {
 		name: 'Crc32',
@@ -879,6 +890,7 @@ export const ebmlMap = {
 	[matroskaElements.Language]: language,
 	[matroskaElements.DefaultDuration]: defaultDuration,
 	[matroskaElements.CodecPrivate]: codecPrivate,
+	[matroskaElements.BlockDuration]: blockDurationSegment,
 	[matroskaElements.BlockAdditions]: blockAdditionsSegment,
 	[matroskaElements.MaxBlockAdditionID]: maxBlockAdditionIdSegment,
 	[matroskaElements.Audio]: audioSegment,

packages/media-parser/src/buffer-iterator.ts

@@ -596,8 +596,15 @@ export const getArrayBufferIterator = (
 				0,
 			);
 		},
-		getByteString(length: number): string {
-			const bytes = getSlice(length);
+		getByteString(length: number, trimTrailingZeroes: boolean): string {
+			let bytes = getSlice(length);
+			// This file has trailing zeroes throughout
+			// https://github.com/remotion-dev/remotion/issues/4668#issuecomment-2561904068
+			// eslint-disable-next-line no-unmodified-loop-condition
+			while (trimTrailingZeroes && bytes[bytes.length - 1] === 0) {
+				bytes = bytes.slice(0, -1);
+			}
+
 			return new TextDecoder().decode(bytes).trim();
 		},
 		getFloat64: () => {

packages/media-parser/src/test/create-matroska.test.ts

@@ -278,7 +278,8 @@ test('Can we disassemble a Matroska file and assembled it again', async () => {
 		throw new Error('Not a webm file');
 	}
 
-	expect(stringifyWebm(parsed.boxes).byteLength).toEqual(
+	// +1 because two files has leading 0x0 boxes which we strip during parsing
+	expect(stringifyWebm(parsed.boxes).byteLength + 2).toEqual(
 		new Uint8Array(bytes).byteLength,
 	);
 	process.env.KEEP_SAMPLES = 'false';

packages/media-parser/src/test/matroska.test.ts

@@ -347,7 +347,7 @@ test('Should get duration of AV1 video', async () => {
 										{
 											minVintWidth: 1,
 											type: 'TagString',
-											value: '00:00:01.000000000\u0000\u0000',
+											value: '00:00:01.000000000',
 										},
 									],
 								},