//! EBML (Extensible Binary Meta Language) write primitives for Matroska. //! //! EBML uses variable-length integers for element IDs and sizes. //! This module provides low-level writers for constructing MKV files. use std::io::{self, Read, Seek, SeekFrom, Write}; /// Write an EBML element ID (1-4 bytes, already encoded). /// Element IDs are predefined constants — we write them verbatim. pub fn write_id(w: &mut impl Write, id: u32) -> io::Result<()> { if id <= 0xFF { w.write_all(&[id as u8]) } else if id <= 0xFFFF { w.write_all(&[(id >> 8) as u8, id as u8]) } else if id <= 0xFF_FFFF { w.write_all(&[(id >> 16) as u8, (id >> 8) as u8, id as u8]) } else { w.write_all(&[ (id >> 24) as u8, (id >> 16) as u8, (id >> 8) as u8, id as u8, ]) } } /// Write an EBML variable-length size (1-8 bytes). /// Uses the EBML VINT encoding: leading bits indicate width. pub fn write_size(w: &mut impl Write, size: u64) -> io::Result<()> { if size < 0x7F { w.write_all(&[(size as u8) | 0x80]) } else if size < 0x3FFF { w.write_all(&[((size >> 8) as u8) | 0x40, size as u8]) } else if size < 0x1F_FFFF { w.write_all(&[((size >> 16) as u8) | 0x20, (size >> 8) as u8, size as u8]) } else if size < 0x0FFF_FFFF { w.write_all(&[ ((size >> 24) as u8) | 0x10, (size >> 16) as u8, (size >> 8) as u8, size as u8, ]) } else if size >= 0x00FF_FFFF_FFFF_FFFF { // 0x00FF_FFFF_FFFF_FFFF (max 56-bit) encodes byte-for-byte // identical to write_unknown_size (the EBML all-ones // "unknown/open-ended" sentinel), and anything larger doesn't fit // the 7-byte payload. Reject so a finite size can never be emitted // as the unknown-size marker. Err(crate::error::Error::MkvInvalid.into()) } else { // 8-byte size for large elements w.write_all(&[ 0x01, (size >> 48) as u8, (size >> 40) as u8, (size >> 32) as u8, (size >> 24) as u8, (size >> 16) as u8, (size >> 8) as u8, size as u8, ]) } } /// Write an EBML "unknown size" marker (all 1s in VINT, 8 bytes). /// Used for the Segment element when total size isn't known upfront. pub fn write_unknown_size(w: &mut impl Write) -> io::Result<()> { w.write_all(&[0x01, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF]) } /// Write a complete EBML unsigned integer element. pub fn write_uint(w: &mut impl Write, id: u32, val: u64) -> io::Result<()> { write_id(w, id)?; if val <= 0xFF { write_size(w, 1)?; w.write_all(&[val as u8]) } else if val <= 0xFFFF { write_size(w, 2)?; w.write_all(&[(val >> 8) as u8, val as u8]) } else if val <= 0xFF_FFFF { write_size(w, 3)?; w.write_all(&[(val >> 16) as u8, (val >> 8) as u8, val as u8]) } else if val <= 0xFFFF_FFFF { write_size(w, 4)?; w.write_all(&[ (val >> 24) as u8, (val >> 16) as u8, (val >> 8) as u8, val as u8, ]) } else { write_size(w, 8)?; w.write_all(&val.to_be_bytes()) } } /// Write a complete EBML float element (8-byte double). pub fn write_float(w: &mut impl Write, id: u32, val: f64) -> io::Result<()> { write_id(w, id)?; write_size(w, 8)?; w.write_all(&val.to_be_bytes()) } /// Write a complete EBML UTF-8 string element. pub fn write_string(w: &mut impl Write, id: u32, val: &str) -> io::Result<()> { write_id(w, id)?; write_size(w, val.len() as u64)?; w.write_all(val.as_bytes()) } /// Write a complete EBML binary element. pub fn write_binary(w: &mut impl Write, id: u32, data: &[u8]) -> io::Result<()> { write_id(w, id)?; write_size(w, data.len() as u64)?; w.write_all(data) } /// Start a master element: write ID + placeholder size. /// Returns the file offset of the size field for later fixup. pub fn start_master(w: &mut W, id: u32) -> io::Result { write_id(w, id)?; let size_pos = w.stream_position()?; // 8-byte size placeholder (will be overwritten by end_master) w.write_all(&[0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00])?; Ok(size_pos) } /// End a master element: seek back and write the actual size. /// /// `size_pos` must be the offset returned by [`start_master`], which always /// writes the 8-byte size placeholder before any body bytes. Therefore /// `end_pos >= size_pos + 8` always holds, and the resulting `data_size` /// fits the 7-byte VINT payload (a single MKV element exceeding 2^56 bytes /// is not representable and never produced here). pub fn end_master(w: &mut W, size_pos: u64) -> io::Result<()> { let end_pos = w.stream_position()?; debug_assert!( end_pos >= size_pos + 8, "end_master: end_pos {end_pos} < size_pos {size_pos} + 8 (placeholder not written?)" ); let data_size = end_pos - size_pos - 8; // subtract the 8-byte size field itself debug_assert!( data_size < 0x0100_0000_0000_0000, "end_master: data_size {data_size} exceeds the 7-byte VINT payload" ); w.seek(SeekFrom::Start(size_pos))?; // Write as 8-byte VINT: 0x01 followed by 7 bytes of size w.write_all(&[ 0x01, (data_size >> 48) as u8, (data_size >> 40) as u8, (data_size >> 32) as u8, (data_size >> 24) as u8, (data_size >> 16) as u8, (data_size >> 8) as u8, data_size as u8, ])?; w.seek(SeekFrom::Start(end_pos))?; Ok(()) } // ============================================================ // EBML Read primitives // ============================================================ /// Read an EBML element ID. Returns (id, bytes_consumed). pub fn read_id(r: &mut impl Read) -> io::Result<(u32, usize)> { let mut first = [0u8; 1]; r.read_exact(&mut first)?; let b0 = first[0]; if b0 & 0x80 != 0 { Ok((b0 as u32, 1)) } else if b0 & 0x40 != 0 { let mut b = [0u8; 1]; r.read_exact(&mut b)?; Ok((((b0 as u32) << 8) | b[0] as u32, 2)) } else if b0 & 0x20 != 0 { let mut b = [0u8; 2]; r.read_exact(&mut b)?; Ok((((b0 as u32) << 16) | (b[0] as u32) << 8 | b[1] as u32, 3)) } else if b0 & 0x10 != 0 { let mut b = [0u8; 3]; r.read_exact(&mut b)?; Ok(( ((b0 as u32) << 24) | (b[0] as u32) << 16 | (b[1] as u32) << 8 | b[2] as u32, 4, )) } else { Err(crate::error::Error::MkvInvalid.into()) } } /// Read an EBML variable-length size. Returns (size, bytes_consumed). /// Size of u64::MAX means "unknown size". pub fn read_size(r: &mut impl Read) -> io::Result<(u64, usize)> { let mut first = [0u8; 1]; r.read_exact(&mut first)?; let b0 = first[0]; if b0 & 0x80 != 0 { let val = (b0 & 0x7F) as u64; if val == 0x7F { return Ok((u64::MAX, 1)); } // unknown Ok((val, 1)) } else if b0 & 0x40 != 0 { let mut b = [0u8; 1]; r.read_exact(&mut b)?; let val = (((b0 & 0x3F) as u64) << 8) | b[0] as u64; if val == 0x3FFF { return Ok((u64::MAX, 2)); } Ok((val, 2)) } else if b0 & 0x20 != 0 { let mut b = [0u8; 2]; r.read_exact(&mut b)?; let val = (((b0 & 0x1F) as u64) << 16) | (b[0] as u64) << 8 | b[1] as u64; if val == 0x1F_FFFF { return Ok((u64::MAX, 3)); } Ok((val, 3)) } else if b0 & 0x10 != 0 { let mut b = [0u8; 3]; r.read_exact(&mut b)?; let val = (((b0 & 0x0F) as u64) << 24) | (b[0] as u64) << 16 | (b[1] as u64) << 8 | b[2] as u64; if val == 0x0FFF_FFFF { return Ok((u64::MAX, 4)); } Ok((val, 4)) } else if b0 & 0x08 != 0 { let mut b = [0u8; 4]; r.read_exact(&mut b)?; let val = (((b0 & 0x07) as u64) << 32) | (b[0] as u64) << 24 | (b[1] as u64) << 16 | (b[2] as u64) << 8 | b[3] as u64; if val == 0x07_FFFF_FFFF { return Ok((u64::MAX, 5)); } Ok((val, 5)) } else if b0 & 0x04 != 0 { let mut b = [0u8; 5]; r.read_exact(&mut b)?; let val = (((b0 & 0x03) as u64) << 40) | (b[0] as u64) << 32 | (b[1] as u64) << 24 | (b[2] as u64) << 16 | (b[3] as u64) << 8 | b[4] as u64; if val == 0x3FF_FFFF_FFFF { return Ok((u64::MAX, 6)); } Ok((val, 6)) } else if b0 & 0x02 != 0 { let mut b = [0u8; 6]; r.read_exact(&mut b)?; let val = (((b0 & 0x01) as u64) << 48) | (b[0] as u64) << 40 | (b[1] as u64) << 32 | (b[2] as u64) << 24 | (b[3] as u64) << 16 | (b[4] as u64) << 8 | b[5] as u64; if val == 0x01_FFFF_FFFF_FFFF { return Ok((u64::MAX, 7)); } Ok((val, 7)) } else if b0 & 0x01 != 0 { let mut b = [0u8; 7]; r.read_exact(&mut b)?; let val = (b[0] as u64) << 48 | (b[1] as u64) << 40 | (b[2] as u64) << 32 | (b[3] as u64) << 24 | (b[4] as u64) << 16 | (b[5] as u64) << 8 | b[6] as u64; if val == 0x00FF_FFFF_FFFF_FFFF { return Ok((u64::MAX, 8)); } Ok((val, 8)) } else { // b0 == 0x00: no length marker in the first byte. A VINT wider than // 8 bytes is not representable by Matroska's size encoding, so this // is a malformed/over-long size field rather than a valid 8-byte // length. Reject it instead of silently building a size from the // following 7 bytes (which would desync the parse). Err(crate::error::Error::MkvInvalid.into()) } } /// Read an EBML element header (ID + size). Returns (id, data_size, header_bytes). pub fn read_element_header(r: &mut impl Read) -> io::Result<(u32, u64, usize)> { let (id, id_len) = read_id(r)?; let (size, size_len) = read_size(r)?; Ok((id, size, id_len + size_len)) } /// Read an unsigned integer value of `len` bytes. pub fn read_uint_val(r: &mut impl Read, len: usize) -> io::Result { // An EBML unsigned integer is at most 8 bytes. A malformed element // claiming `len > 8` would index past this stack buffer and panic // (DoS on untrusted input) — reject it at the source so every caller // is safe, not just the ones that pre-check. if len > 8 { return Err(crate::error::Error::MkvInvalid.into()); } let mut buf = [0u8; 8]; r.read_exact(&mut buf[..len])?; let mut val = 0u64; for &b in &buf[..len] { val = (val << 8) | b as u64; } Ok(val) } /// Read a float value. EBML floats are exactly 0, 4, or 8 bytes; any other /// length is rejected as [`Error::MkvInvalid`] and exactly the float width is /// consumed (so a malformed element never under- or over-reads and desyncs the /// rest of the parent element). pub fn read_float_val(r: &mut impl Read, len: usize) -> io::Result { match len { 0 => Ok(0.0), 4 => { let mut buf = [0u8; 4]; r.read_exact(&mut buf)?; Ok(f32::from_be_bytes(buf) as f64) } 8 => { let mut buf = [0u8; 8]; r.read_exact(&mut buf)?; Ok(f64::from_be_bytes(buf)) } _ => Err(crate::error::Error::MkvInvalid.into()), } } /// Read a UTF-8 string value of `len` bytes. pub fn read_string_val(r: &mut impl Read, len: usize) -> io::Result { let mut buf = read_exact_bounded(r, len)?; // Strip trailing nulls while buf.last() == Some(&0) { buf.pop(); } // Library rule: errors are numeric variants, never English strings. // A non-UTF-8 string element is malformed input → MkvInvalid. String::from_utf8(buf).map_err(|_| crate::error::Error::MkvInvalid.into()) } /// Read binary data of `len` bytes. pub fn read_binary_val(r: &mut impl Read, len: usize) -> io::Result> { read_exact_bounded(r, len) } /// Read exactly `len` bytes WITHOUT trusting `len` to size the allocation. /// /// `vec![0u8; len]` on an attacker-controlled EBML size would allocate /// gigabytes before the read fails. Instead we cap the reader to `len` /// and grow the buffer as bytes actually arrive: a malformed element that /// claims a huge length but supplies few bytes allocates only what it /// delivers, then errors on the short read. fn read_exact_bounded(r: &mut impl Read, len: usize) -> io::Result> { let mut buf = Vec::new(); let got = r.take(len as u64).read_to_end(&mut buf)?; if got != len { // A truncated element is malformed input. Use the typed crate error // so callers matching on Error::MkvInvalid catch short reads rather // than a bare io::ErrorKind that bypasses the numeric-code identity. return Err(crate::error::Error::MkvInvalid.into()); } Ok(buf) } // ============================================================ // Matroska Element IDs // ============================================================ // EBML Header pub const EBML: u32 = 0x1A45_DFA3; pub const EBML_VERSION: u32 = 0x4286; pub const EBML_READ_VERSION: u32 = 0x42F7; pub const EBML_MAX_ID_LENGTH: u32 = 0x42F2; pub const EBML_MAX_SIZE_LENGTH: u32 = 0x42F3; pub const EBML_DOC_TYPE: u32 = 0x4282; pub const EBML_DOC_TYPE_VERSION: u32 = 0x4287; pub const EBML_DOC_TYPE_READ_VERSION: u32 = 0x4285; // Segment pub const SEGMENT: u32 = 0x1853_8067; // SeekHead pub const SEEK_HEAD: u32 = 0x114D_9B74; pub const SEEK: u32 = 0x4DBB; /// Void — RFC 9559 (Matroska) §EBML global element 0xEC. Used to neutralise a /// reserved-but-unused region (e.g. the CUES SeekHead entry when no Cues element /// is written) so it carries no meaning to a parser. pub const VOID: u32 = 0xEC; pub const SEEK_ID: u32 = 0x53AB; pub const SEEK_POSITION: u32 = 0x53AC; // Segment Info pub const INFO: u32 = 0x1549_A966; pub const TIMESTAMP_SCALE: u32 = 0x2A_D7B1; pub const DURATION: u32 = 0x4489; pub const MUXING_APP: u32 = 0x4D80; pub const WRITING_APP: u32 = 0x5741; pub const TITLE: u32 = 0x7BA9; // Tracks pub const TRACKS: u32 = 0x1654_AE6B; pub const TRACK_ENTRY: u32 = 0xAE; pub const TRACK_NUMBER: u32 = 0xD7; pub const TRACK_UID: u32 = 0x73C5; pub const TRACK_TYPE: u32 = 0x83; pub const FLAG_LACING: u32 = 0x9C; pub const FLAG_DEFAULT: u32 = 0x88; pub const FLAG_FORCED: u32 = 0x55AA; pub const LANGUAGE: u32 = 0x22_B59C; pub const CODEC_ID: u32 = 0x86; pub const CODEC_PRIVATE: u32 = 0x63A2; pub const TRACK_NAME: u32 = 0x536E; pub const DEFAULT_DURATION: u32 = 0x23_E383; /// DefaultDecodedFieldDuration — nanoseconds per FIELD (half a frame for /// interlaced content). Emitting it on an interlaced track tells a reader the /// field rate so it stops halving the frame rate (Windows shell shows 12.5 fps /// for a 25 fps 576i stream without it). RFC 9559 / Matroska v4. pub const DEFAULT_DECODED_FIELD_DURATION: u32 = 0x23_4E7A; // Video pub const VIDEO: u32 = 0xE0; pub const PIXEL_WIDTH: u32 = 0xB0; pub const PIXEL_HEIGHT: u32 = 0xBA; // Scan type (children of Video). pub const FLAG_INTERLACED: u32 = 0x9A; pub const FIELD_ORDER: u32 = 0x9D; // FlagInterlaced values: 1 = interlaced, 2 = progressive (0 = undetermined). pub const INTERLACED_INTERLACED: u64 = 1; pub const INTERLACED_PROGRESSIVE: u64 = 2; // FieldOrder values (Matroska / RFC 9559, element 0x9D): 1 = top-field-first, // 6 = bottom-field-first, 0 = progressive. The muxer derives TFF vs BFF from the // bitstream's measured top_field_first when available, falling back to TFF for // interlaced content (NTSC 480i / PAL 576i / HD 1080i are overwhelmingly TFF). // 0xFF is our sentinel for "undetermined / omit the element". pub const FIELD_ORDER_TFF: u8 = 1; /// Bottom-field-first (RFC 9559 element 0x9D = 6). Emitted when the bitstream's /// measured top_field_first is false. pub const FIELD_ORDER_BFF: u8 = 6; pub const FIELD_ORDER_UNDETERMINED: u8 = 0xFF; pub const DISPLAY_WIDTH: u32 = 0x54B0; pub const DISPLAY_HEIGHT: u32 = 0x54BA; pub const COLOUR: u32 = 0x55B0; pub const TRANSFER_CHARACTERISTICS: u32 = 0x55BA; pub const MATRIX_COEFFICIENTS: u32 = 0x55B1; pub const PRIMARIES: u32 = 0x55BB; pub const RANGE: u32 = 0x55B9; // Dolby Vision — BlockAdditionMapping carries the DOVIDecoderConfigurationRecord // (dvcC) so players / mediainfo recognise the track as Dolby Vision. pub const BLOCK_ADDITION_MAPPING: u32 = 0x41E4; pub const BLOCK_ADD_ID_TYPE: u32 = 0x41E7; pub const BLOCK_ADD_ID_EXTRA_DATA: u32 = 0x41ED; // Audio pub const AUDIO: u32 = 0xE1; pub const SAMPLING_FREQUENCY: u32 = 0xB5; pub const CHANNELS: u32 = 0x9F; pub const BIT_DEPTH: u32 = 0x6264; // Cluster pub const CLUSTER: u32 = 0x1F43_B675; pub const CLUSTER_TIMESTAMP: u32 = 0xE7; pub const SIMPLE_BLOCK: u32 = 0xA3; pub const BLOCK_GROUP: u32 = 0xA0; pub const BLOCK: u32 = 0xA1; pub const BLOCK_DURATION: u32 = 0x9B; // Cues pub const CUES: u32 = 0x1C53_BB6B; pub const CUE_POINT: u32 = 0xBB; pub const CUE_TIME: u32 = 0xB3; pub const CUE_TRACK_POSITIONS: u32 = 0xB7; pub const CUE_TRACK: u32 = 0xF7; pub const CUE_CLUSTER_POSITION: u32 = 0xF1; // Tags — per-track statistics tags. mkvmerge convention: a `BPS` SimpleTag // per track carries the bits-per-second so readers (Windows Explorer's MKV // property handler) that read the container tag rather than computing from // stream size show a bitrate for every track, not just CBR audio. pub const TAGS: u32 = 0x1254_C367; pub const TAG: u32 = 0x7373; pub const TARGETS: u32 = 0x63C0; pub const TAG_TRACK_UID: u32 = 0x63C5; pub const SIMPLE_TAG: u32 = 0x67C8; pub const TAG_NAME: u32 = 0x45A3; pub const TAG_STRING: u32 = 0x4487; // Chapters pub const CHAPTERS: u32 = 0x1043_A770; pub const EDITION_ENTRY: u32 = 0x45B9; pub const CHAPTER_ATOM: u32 = 0xB6; pub const CHAPTER_UID: u32 = 0x73C4; pub const CHAPTER_TIME_START: u32 = 0x91; pub const CHAPTER_DISPLAY: u32 = 0x80; pub const CHAP_STRING: u32 = 0x85; pub const CHAP_LANGUAGE: u32 = 0x437C; // Track types pub const TRACK_TYPE_VIDEO: u64 = 1; pub const TRACK_TYPE_AUDIO: u64 = 2; pub const TRACK_TYPE_SUBTITLE: u64 = 17; // Matroska CodecID strings (the `CodecID` element value per the Matroska codec // registry). Single source of truth for both the muxer (Codec -> string) and // the demuxer (string -> Codec), so the two can never drift. pub const CODEC_HEVC: &str = "V_MPEGH/ISO/HEVC"; pub const CODEC_H264: &str = "V_MPEG4/ISO/AVC"; pub const CODEC_VC1: &str = "V_MS/VFW/FOURCC"; pub const CODEC_MPEG2: &str = "V_MPEG2"; pub const CODEC_AC3: &str = "A_AC3"; pub const CODEC_EAC3: &str = "A_EAC3"; pub const CODEC_TRUEHD: &str = "A_TRUEHD"; pub const CODEC_DTS: &str = "A_DTS"; pub const CODEC_PCM_BE: &str = "A_PCM/INT/BIG"; pub const CODEC_PGS: &str = "S_HDMV/PGS"; pub const CODEC_VOBSUB: &str = "S_VOBSUB"; #[cfg(test)] mod tests { use super::*; use std::io::Cursor; #[test] fn test_write_size() { let mut buf = Vec::new(); write_size(&mut buf, 0).unwrap(); assert_eq!(buf, [0x80]); buf.clear(); write_size(&mut buf, 127).unwrap(); assert_eq!(buf, [0x40, 127]); // 127 >= 0x7F, uses 2 bytes: (0>>8)|0x40, 127 buf.clear(); write_size(&mut buf, 126).unwrap(); assert_eq!(buf, [126 | 0x80]); // 126 < 0x7F, uses 1 byte } #[test] fn write_size_rejects_unknown_size_sentinel() { // 0x00FF_FFFF_FFFF_FFFF would encode byte-for-byte identical to the // EBML unknown-size marker; it must be rejected, not silently emitted. let mut buf = Vec::new(); let e = write_size(&mut buf, 0x00FF_FFFF_FFFF_FFFF).unwrap_err(); assert_eq!(e.kind(), io::ErrorKind::InvalidData); assert!(buf.is_empty(), "no bytes should be written on rejection"); // One below the boundary still encodes as a normal 8-byte size whose // payload is NOT all-ones, so read_size yields the finite value back. buf.clear(); let v = 0x00FF_FFFF_FFFF_FFFE; write_size(&mut buf, v).unwrap(); let (back, consumed) = read_size(&mut Cursor::new(&buf)).unwrap(); assert_eq!(consumed, 8); assert_eq!(back, v); } #[test] fn test_write_uint() { let mut buf = Vec::new(); write_uint(&mut buf, 0x4286, 1).unwrap(); // EBML_VERSION = 1 // ID: 42 86, Size: 81 (1 byte), Data: 01 assert_eq!(buf, [0x42, 0x86, 0x81, 0x01]); } #[test] fn test_write_string() { let mut buf = Vec::new(); write_string(&mut buf, 0x4282, "matroska").unwrap(); // ID: 42 82, Size: 88 (8 bytes), Data: "matroska" assert_eq!(&buf[0..2], &[0x42, 0x82]); assert_eq!(buf[2], 0x88); // size = 8 assert_eq!(&buf[3..], b"matroska"); } #[test] fn test_master_element() { let mut buf = Cursor::new(Vec::new()); let pos = start_master(&mut buf, EBML).unwrap(); write_uint(&mut buf, EBML_VERSION, 1).unwrap(); end_master(&mut buf, pos).unwrap(); let data = buf.into_inner(); // EBML header: 1A 45 DF A3, then 8-byte size, then content assert_eq!(&data[0..4], &[0x1A, 0x45, 0xDF, 0xA3]); } #[test] fn write_read_id_roundtrip() { // 1-byte IDs have high bit set (0x80..=0xFF) for &id in &[0x80u32, 0xA3, 0xFF] { let mut buf = Vec::new(); write_id(&mut buf, id).unwrap(); assert_eq!(buf.len(), 1); let mut cursor = Cursor::new(&buf); let (read_back, consumed) = read_id(&mut cursor).unwrap(); assert_eq!(read_back, id, "1-byte ID roundtrip failed for 0x{:X}", id); assert_eq!(consumed, 1); } // 2-byte IDs (0x4000..=0x7FFF) for &id in &[0x4286u32, 0x4282, 0x7FFF] { let mut buf = Vec::new(); write_id(&mut buf, id).unwrap(); assert_eq!(buf.len(), 2); let mut cursor = Cursor::new(&buf); let (read_back, consumed) = read_id(&mut cursor).unwrap(); assert_eq!(read_back, id, "2-byte ID roundtrip failed for 0x{:X}", id); assert_eq!(consumed, 2); } // 3-byte IDs (0x200000..=0x3FFFFF) for &id in &[0x22B59Cu32, 0x23E383] { let mut buf = Vec::new(); write_id(&mut buf, id).unwrap(); assert_eq!(buf.len(), 3); let mut cursor = Cursor::new(&buf); let (read_back, consumed) = read_id(&mut cursor).unwrap(); assert_eq!(read_back, id, "3-byte ID roundtrip failed for 0x{:X}", id); assert_eq!(consumed, 3); } // 4-byte IDs (0x10000000..=0x1FFFFFFF) for &id in &[EBML, SEGMENT, TRACKS, CLUSTER] { let mut buf = Vec::new(); write_id(&mut buf, id).unwrap(); assert_eq!(buf.len(), 4); let mut cursor = Cursor::new(&buf); let (read_back, consumed) = read_id(&mut cursor).unwrap(); assert_eq!(read_back, id, "4-byte ID roundtrip failed for 0x{:X}", id); assert_eq!(consumed, 4); } } #[test] fn write_read_size_roundtrip() { let test_sizes: &[u64] = &[ 0, 1, 0x7E, 127, 128, 0x3FFE, 16383, 16384, 0x1FFFFE, 0x0FFFFFFE, 0x1_0000_0000, ]; for &size in test_sizes { let mut buf = Vec::new(); write_size(&mut buf, size).unwrap(); let mut cursor = Cursor::new(&buf); let (read_back, _consumed) = read_size(&mut cursor).unwrap(); assert_eq!(read_back, size, "size roundtrip failed for {}", size); } } #[test] fn write_read_uint_roundtrip() { let test_vals: &[u64] = &[ 0, 1, 127, 255, 256, 0xFFFF, 0xFF_FFFF, 0xFFFF_FFFF, 1_000_000_000_000, ]; let test_id = EBML_VERSION; for &val in test_vals { let mut buf = Vec::new(); write_uint(&mut buf, test_id, val).unwrap(); let mut cursor = Cursor::new(&buf); let (id, _id_len) = read_id(&mut cursor).unwrap(); assert_eq!(id, test_id); let (size, _) = read_size(&mut cursor).unwrap(); let read_val = read_uint_val(&mut cursor, size as usize).unwrap(); assert_eq!(read_val, val, "uint roundtrip failed for {}", val); } } #[test] fn write_read_string_roundtrip() { let test_strings = &[ "", "matroska", "freemkv", "Hello, World!", "unicode: \u{1F600}", ]; let test_id = EBML_DOC_TYPE; for &s in test_strings { let mut buf = Vec::new(); write_string(&mut buf, test_id, s).unwrap(); let mut cursor = Cursor::new(&buf); let (id, _) = read_id(&mut cursor).unwrap(); assert_eq!(id, test_id); let (size, _) = read_size(&mut cursor).unwrap(); let read_s = read_string_val(&mut cursor, size as usize).unwrap(); assert_eq!(read_s, s, "string roundtrip failed for {:?}", s); } } #[test] fn write_read_float_roundtrip() { let test_vals: &[f64] = &[ 0.0, 1.0, -1.0, std::f64::consts::PI, 48000.0, 7200000.0, f64::MIN, f64::MAX, ]; let test_id = DURATION; for &val in test_vals { let mut buf = Vec::new(); write_float(&mut buf, test_id, val).unwrap(); let mut cursor = Cursor::new(&buf); let (id, _) = read_id(&mut cursor).unwrap(); assert_eq!(id, test_id); let (size, _) = read_size(&mut cursor).unwrap(); assert_eq!(size, 8); let read_val = read_float_val(&mut cursor, size as usize).unwrap(); assert_eq!( read_val.to_bits(), val.to_bits(), "float roundtrip failed for {}", val ); } } #[test] fn read_size_unknown_sentinel_all_widths() { // The all-ones VINT of each width is the EBML "unknown size" marker // and must read back as u64::MAX. write_size never emits the 5/6/7-byte // widths, so these are hand-crafted. Each entry is (bytes, expected_len). let cases: &[(&[u8], usize)] = &[ // 1-byte: 0x80 | 0x7F (&[0xFF], 1), // 2-byte: 0x40 marker, value bits all 1 (&[0x7F, 0xFF], 2), // 3-byte (&[0x3F, 0xFF, 0xFF], 3), // 4-byte (&[0x1F, 0xFF, 0xFF, 0xFF], 4), // 5-byte (0x08 marker) (&[0x0F, 0xFF, 0xFF, 0xFF, 0xFF], 5), // 6-byte (0x04 marker) (&[0x07, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF], 6), // 7-byte (0x02 marker) (&[0x03, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF], 7), // 8-byte (0x01 marker) (&[0x01, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF], 8), ]; for (bytes, expected_len) in cases { let mut cursor = Cursor::new(*bytes); let (size, consumed) = read_size(&mut cursor).unwrap(); assert_eq!( size, u64::MAX, "all-ones {}-byte VINT should be unknown-size", expected_len ); assert_eq!(consumed, *expected_len); } } #[test] fn read_size_concrete_5_6_7_byte_values() { // A non-sentinel 5/6/7-byte size must read back as its concrete value, // not be mistaken for unknown-size. // 5-byte: marker 0x08, value 0x01 (0x0800000001 with width bit only). let mut c = Cursor::new(&[0x08u8, 0x00, 0x00, 0x00, 0x01]); assert_eq!(read_size(&mut c).unwrap(), (1, 5)); // 6-byte let mut c = Cursor::new(&[0x04u8, 0x00, 0x00, 0x00, 0x00, 0x05]); assert_eq!(read_size(&mut c).unwrap(), (5, 6)); // 7-byte let mut c = Cursor::new(&[0x02u8, 0x00, 0x00, 0x00, 0x00, 0x00, 0x09]); assert_eq!(read_size(&mut c).unwrap(), (9, 7)); } #[test] fn read_size_rejects_zero_first_byte() { // b0 == 0x00 has no width marker — an over-long/invalid VINT. It must // be rejected, not silently treated as an 8-byte size. let mut c = Cursor::new(&[0x00u8, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF]); let e = read_size(&mut c).unwrap_err(); assert_eq!(e.kind(), io::ErrorKind::InvalidData); } #[test] fn write_size_rejects_at_or_above_2_56() { // 2^56 cannot be encoded in the 7-payload-byte 8-byte VINT and must // error rather than silently truncate. let mut buf = Vec::new(); let e = write_size(&mut buf, 0x0100_0000_0000_0000).unwrap_err(); assert_eq!(e.kind(), io::ErrorKind::InvalidData); // The largest encodable size still succeeds. let mut buf = Vec::new(); write_size(&mut buf, 0x00FF_FFFF_FFFF_FFFE).unwrap(); assert_eq!(buf.len(), 8); } #[test] fn unknown_size() { let mut buf = Vec::new(); write_unknown_size(&mut buf).unwrap(); assert_eq!(buf.len(), 8); assert_eq!(buf[0], 0x01); for &b in &buf[1..] { assert_eq!( b, 0xFF, "unknown size bytes should all be 0xFF after first byte" ); } // Reading it back should yield u64::MAX let mut cursor = Cursor::new(&buf); let (size, consumed) = read_size(&mut cursor).unwrap(); assert_eq!(size, u64::MAX); assert_eq!(consumed, 8); } // ============================================================ // write_id — exact width selection per EBML element-ID ranges // (Matroska/EBML spec: an element ID is written verbatim; its // declared width is implied by the position of the leading 1 bit. // write_id must pick the minimal whole-byte encoding so the ID // round-trips and parsers see the same width.) // ============================================================ #[test] fn write_id_exact_bytes_per_width() { // 1-byte ID (high bit set): emitted as a single byte verbatim. let mut b = Vec::new(); write_id(&mut b, 0xA3).unwrap(); // SimpleBlock assert_eq!(b, [0xA3]); // The boundary just above 1 byte: 0x100 must be a 2-byte ID. A // mutation that widened the 1-byte branch (id <= 0x1FF) would drop // the high byte here. let mut b = Vec::new(); write_id(&mut b, 0x0100).unwrap(); assert_eq!(b, [0x01, 0x00]); // 2-byte ID written MSB-first. let mut b = Vec::new(); write_id(&mut b, 0x4286).unwrap(); // EBMLVersion assert_eq!(b, [0x42, 0x86]); // 3-byte boundary: 0x1_0000 must be 3 bytes. let mut b = Vec::new(); write_id(&mut b, 0x01_0000).unwrap(); assert_eq!(b, [0x01, 0x00, 0x00]); // 3-byte ID (Language = 0x22B59C). let mut b = Vec::new(); write_id(&mut b, 0x22_B59C).unwrap(); assert_eq!(b, [0x22, 0xB5, 0x9C]); // 4-byte boundary: 0x100_0000 must be 4 bytes. let mut b = Vec::new(); write_id(&mut b, 0x0100_0000).unwrap(); assert_eq!(b, [0x01, 0x00, 0x00, 0x00]); // 4-byte ID (Segment = 0x18538067) MSB-first. let mut b = Vec::new(); write_id(&mut b, 0x1853_8067).unwrap(); assert_eq!(b, [0x18, 0x53, 0x80, 0x67]); } #[test] fn read_id_rejects_zero_first_byte() { // A first byte of 0x00 has no length marker in any of bits 7..4, so // read_id falls through to the else branch and must reject it (an // EBML ID wider than 4 bytes is not representable here). Otherwise the // parser would desync. let mut c = Cursor::new(&[0x00u8, 0x11, 0x22, 0x33]); let e = read_id(&mut c).unwrap_err(); assert_eq!(e.kind(), io::ErrorKind::InvalidData); } // ============================================================ // write_uint — the SIZE byte must reflect the minimal big-endian // value width (1/2/3/4/8). The Matroska spec stores unsigned ints // big-endian with no leading-zero bytes; the declared element size // is exactly that width. A boundary bug would write the wrong size // and desync every following element. // ============================================================ #[test] fn write_uint_size_byte_matches_value_width() { // (value, expected_size_byte, expected_payload) // size byte is a 1-byte VINT: 0x80 | len. let cases: &[(u64, u8, &[u8])] = &[ (0x00, 0x81, &[0x00]), // 1 byte (0xFF, 0x81, &[0xFF]), // 1 byte (boundary high) (0x0100, 0x82, &[0x01, 0x00]), // 2 bytes (just over u8) (0xFFFF, 0x82, &[0xFF, 0xFF]), // 2 bytes (boundary high) (0x01_0000, 0x83, &[0x01, 0x00, 0x00]), // 3 bytes (0xFF_FFFF, 0x83, &[0xFF, 0xFF, 0xFF]), // 3 bytes (boundary high) (0x0100_0000, 0x84, &[0x01, 0x00, 0x00, 0x00]), // 4 bytes (0xFFFF_FFFF, 0x84, &[0xFF, 0xFF, 0xFF, 0xFF]), // 4 bytes (boundary high) // Just over u32 → jumps straight to 8 bytes (no 5/6/7 path). ( 0x1_0000_0000, 0x88, &[0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00], ), ]; let id = EBML_VERSION; // 2-byte ID 0x4286 for (val, size_byte, payload) in cases { let mut buf = Vec::new(); write_uint(&mut buf, id, *val).unwrap(); assert_eq!(&buf[0..2], &[0x42, 0x86], "ID prefix for val {val:#x}"); assert_eq!(buf[2], *size_byte, "size byte for val {val:#x}"); assert_eq!(&buf[3..], *payload, "payload for val {val:#x}"); } } #[test] fn write_uint_zero_is_one_byte_not_zero_length() { // EBML stores 0 as a single 0x00 byte (size 1), NOT a zero-length // element. A muxer reader expects to consume exactly one payload byte. let mut buf = Vec::new(); write_uint(&mut buf, EBML_VERSION, 0).unwrap(); // ID(2) + size(1=0x81) + one payload byte 0x00. assert_eq!(buf, [0x42, 0x86, 0x81, 0x00]); } // ============================================================ // write_float — EBML floats here are always 8-byte IEEE-754 doubles, // big-endian (Matroska SamplingFrequency/Duration). size byte = 0x88. // ============================================================ #[test] fn write_float_is_8_byte_big_endian_double() { let mut buf = Vec::new(); write_float(&mut buf, DURATION, 48000.0).unwrap(); // ID DURATION = 0x4489 (2 bytes), size = 0x88 (8), then BE f64. assert_eq!(&buf[0..2], &[0x44, 0x89]); assert_eq!(buf[2], 0x88, "float element must declare 8-byte size"); assert_eq!(&buf[3..11], &48000.0f64.to_be_bytes()); // The reader (4-byte path) must yield an f32-promoted value, while the // 8-byte path yields the exact double. let got = read_float_val(&mut Cursor::new(&buf[3..11]), 8).unwrap(); assert_eq!(got.to_bits(), 48000.0f64.to_bits()); } // ============================================================ // write_string / write_binary — declared size must equal the byte // length (UTF-8 byte count, not char count) so the reader consumes // exactly the payload and no more. // ============================================================ #[test] fn write_string_size_is_utf8_byte_count_not_char_count() { // "é" is 2 UTF-8 bytes; the size field must be 2, not 1. let mut buf = Vec::new(); write_string(&mut buf, EBML_DOC_TYPE, "é").unwrap(); assert_eq!(&buf[0..2], &[0x42, 0x82]); // DocType ID assert_eq!(buf[2], 0x80 | 2, "size must be UTF-8 byte length (2)"); assert_eq!(&buf[3..], "é".as_bytes()); } #[test] fn write_binary_declares_exact_length() { let data = [0xDE, 0xAD, 0xBE, 0xEF, 0x00]; let mut buf = Vec::new(); write_binary(&mut buf, CODEC_PRIVATE, &data).unwrap(); // CODEC_PRIVATE id 0x63A2 (2 bytes), size 0x85 (len 5), then data. assert_eq!(&buf[0..2], &[0x63, 0xA2]); assert_eq!(buf[2], 0x80 | 5); assert_eq!(&buf[3..], &data); } // ============================================================ // read_string_val — Matroska strings may be null-padded; the reader // strips trailing NULs but must preserve interior content and the // payload byte-count consumed. // ============================================================ #[test] fn read_string_val_strips_only_trailing_nulls() { // "ab\0\0" → "ab"; interior content must not be touched. let raw = b"ab\0\0"; let s = read_string_val(&mut Cursor::new(raw), raw.len()).unwrap(); assert_eq!(s, "ab"); // A string that is ALL nulls collapses to empty (every byte popped). let raw = b"\0\0\0"; let s = read_string_val(&mut Cursor::new(raw), raw.len()).unwrap(); assert_eq!(s, ""); // An interior NUL is NOT a terminator for the strip loop (it only pops // from the tail), so "a\0b" keeps the interior NUL. let raw = b"a\0b"; let s = read_string_val(&mut Cursor::new(raw), raw.len()).unwrap(); assert_eq!(s.as_bytes(), b"a\0b"); } // ============================================================ // read_uint_val — big-endian assembly; an EBML uint never exceeds 8 // bytes (the reader rejects len>8 to avoid a stack OOB). // ============================================================ #[test] fn read_uint_val_big_endian_and_len_zero() { // Big-endian: 0x01 0x02 0x03 → 0x010203. let v = read_uint_val(&mut Cursor::new(&[0x01u8, 0x02, 0x03]), 3).unwrap(); assert_eq!(v, 0x01_0203); // len 0 yields 0 with no read. let v = read_uint_val(&mut Cursor::new(&[] as &[u8]), 0).unwrap(); assert_eq!(v, 0); // Full 8-byte width assembles correctly (no truncation). let bytes = [0x12u8, 0x34, 0x56, 0x78, 0x9A, 0xBC, 0xDE, 0xF0]; let v = read_uint_val(&mut Cursor::new(&bytes), 8).unwrap(); assert_eq!(v, 0x1234_5678_9ABC_DEF0); } #[test] fn read_uint_val_rejects_len_above_8() { // len 9 would index past the [0u8; 8] buffer → OOB/DoS on untrusted // input. Must be a clean MkvInvalid. let e = read_uint_val(&mut Cursor::new(&[0u8; 16]), 9).unwrap_err(); assert_eq!(e.kind(), io::ErrorKind::InvalidData); } // ============================================================ // read_float_val — exactly 0/4/8 byte widths; 4-byte is an f32 // promoted to f64, 8-byte is an exact f64. // ============================================================ #[test] fn read_float_val_4_byte_is_f32_promoted() { // 1.5 as a 32-bit float → 0x3FC00000. let bytes = 1.5f32.to_be_bytes(); let v = read_float_val(&mut Cursor::new(&bytes), 4).unwrap(); assert_eq!(v, 1.5f64); // A value with no exact f32 representation loses precision exactly as // f32→f64 would (proves the 4-byte branch uses f32, not f64). let bytes = 0.1f32.to_be_bytes(); let v = read_float_val(&mut Cursor::new(&bytes), 4).unwrap(); assert_eq!(v, 0.1f32 as f64); assert_ne!(v, 0.1f64, "4-byte path must be f32, losing f64 precision"); } #[test] fn read_float_val_rejects_odd_widths() { // Only 0/4/8 are valid; 1,2,3,5,6,7 must error (never over/under-read). for len in [1usize, 2, 3, 5, 6, 7] { let e = read_float_val(&mut Cursor::new(&[0u8; 8]), len).unwrap_err(); assert_eq!(e.kind(), io::ErrorKind::InvalidData, "len {len}"); } } // ============================================================ // read_binary_val / read_exact_bounded — a declared length that // exceeds the bytes actually present is a truncated (malformed) // element and must error without allocating the full declared size. // ============================================================ #[test] fn read_binary_val_short_read_errors() { // Declare 100 bytes but supply 4 → MkvInvalid (truncated element). let e = read_binary_val(&mut Cursor::new(&[1u8, 2, 3, 4]), 100).unwrap_err(); assert_eq!(e.kind(), io::ErrorKind::InvalidData); // Exact-length read returns the bytes verbatim. let v = read_binary_val(&mut Cursor::new(&[1u8, 2, 3, 4]), 4).unwrap(); assert_eq!(v, vec![1, 2, 3, 4]); } // ============================================================ // read_element_header — header_bytes is id_len + size_len, and a // truncated header (EOF mid-size) surfaces as an error. // ============================================================ #[test] fn read_element_header_reports_total_header_len() { // 4-byte ID (Segment) + 8-byte unknown size = 12 header bytes. let mut buf = Vec::new(); write_id(&mut buf, SEGMENT).unwrap(); write_unknown_size(&mut buf).unwrap(); let (id, size, hdr) = read_element_header(&mut Cursor::new(&buf)).unwrap(); assert_eq!(id, SEGMENT); assert_eq!(size, u64::MAX); assert_eq!(hdr, 12, "4-byte id + 8-byte size = 12 header bytes"); // 1-byte ID (SimpleBlock 0xA3) + 1-byte size = 2 header bytes. let mut buf = Vec::new(); write_id(&mut buf, SIMPLE_BLOCK).unwrap(); write_size(&mut buf, 10).unwrap(); let (id, size, hdr) = read_element_header(&mut Cursor::new(&buf)).unwrap(); assert_eq!(id, SIMPLE_BLOCK); assert_eq!(size, 10); assert_eq!(hdr, 2); } #[test] fn read_id_truncated_after_marker_errors() { // First byte 0x40 promises a 2-byte ID but the second byte is missing. // read_exact must surface EOF, never silently produce a 1-byte ID. let e = read_id(&mut Cursor::new(&[0x40u8])).unwrap_err(); assert_eq!(e.kind(), io::ErrorKind::UnexpectedEof); } // ============================================================ // write_size — every declared width-boundary, asserting the exact // VINT bytes (length marker + payload). Grounded in the EBML VINT // spec: width W encodes 7*W payload bits, the highest value of each // width being reserved as the unknown-size sentinel. // ============================================================ #[test] fn write_size_exact_bytes_at_width_boundaries() { // Largest 1-byte value (126 = 0x7E): marker 0x80 | value. let mut b = Vec::new(); write_size(&mut b, 0x7E).unwrap(); assert_eq!(b, [0x80 | 0x7E]); // 0x7F is NOT 1-byte here (reserved sentinel region) → 2 bytes. let mut b = Vec::new(); write_size(&mut b, 0x7F).unwrap(); assert_eq!(b, [0x40, 0x7F]); // Largest 2-byte value below the 0x3FFF sentinel. let mut b = Vec::new(); write_size(&mut b, 0x3FFE).unwrap(); assert_eq!(b, [0x40 | 0x3F, 0xFE]); // First 3-byte value (0x3FFF goes 3-byte because `< 0x3FFF` is false). let mut b = Vec::new(); write_size(&mut b, 0x3FFF).unwrap(); assert_eq!(b, [0x20, 0x3F, 0xFF]); // First 4-byte value: 0x1F_FFFF is not < 0x1F_FFFF. let mut b = Vec::new(); write_size(&mut b, 0x1F_FFFF).unwrap(); assert_eq!(b, [0x10, 0x1F, 0xFF, 0xFF]); // First 8-byte value: 0x0FFF_FFFF is not < 0x0FFF_FFFF. let mut b = Vec::new(); write_size(&mut b, 0x0FFF_FFFF).unwrap(); assert_eq!(b, [0x01, 0, 0, 0, 0x0F, 0xFF, 0xFF, 0xFF]); } // ============================================================ // start_master / end_master — the size placeholder is an 8-byte VINT // (0x01 + 7 payload bytes), and end_master must back-patch the exact // body byte count (end - start - 8). This is the core of every nested // Matroska master element; a wrong subtraction silently corrupts the // declared size of EVERY master element in the file. // ============================================================ #[test] fn end_master_backpatches_exact_body_size() { let mut c = Cursor::new(Vec::new()); let pos = start_master(&mut c, SEGMENT).unwrap(); // Body: a 4-byte uint element (ID 0x4286, size 0x81, payload 0x01). write_uint(&mut c, EBML_VERSION, 1).unwrap(); end_master(&mut c, pos).unwrap(); let data = c.into_inner(); // Layout: SEGMENT id (4 bytes) | 8-byte size VINT | body (4 bytes). assert_eq!(&data[0..4], &SEGMENT.to_be_bytes()); // The size field is an 8-byte VINT; its payload must equal the body // length (4). 0x01 marker then 7 payload bytes ending in 0x04. assert_eq!(data[4], 0x01); assert_eq!(&data[5..12], &[0, 0, 0, 0, 0, 0, 4]); // Read it back: the header parser sees the exact body size. let (id, size, hdr) = read_element_header(&mut Cursor::new(&data)).unwrap(); assert_eq!(id, SEGMENT); assert_eq!(size, 4, "back-patched size must equal body byte count"); assert_eq!(hdr, 12); assert_eq!(data.len() as u64, hdr as u64 + size); } #[test] fn end_master_empty_body_is_zero_size() { // A master with no body must declare size 0 (end == start + 8). let mut c = Cursor::new(Vec::new()); let pos = start_master(&mut c, INFO).unwrap(); end_master(&mut c, pos).unwrap(); let data = c.into_inner(); let (id, size, _) = read_element_header(&mut Cursor::new(&data)).unwrap(); assert_eq!(id, INFO); assert_eq!(size, 0); } #[test] fn nested_masters_each_get_correct_size() { // Outer master containing an inner master + a sibling uint. Each // declared size must bound exactly its own body. This is the nested // sizing that mkv.rs relies on for Segment→Tracks→TrackEntry. let mut c = Cursor::new(Vec::new()); let outer = start_master(&mut c, TRACKS).unwrap(); let inner = start_master(&mut c, TRACK_ENTRY).unwrap(); write_uint(&mut c, TRACK_NUMBER, 1).unwrap(); end_master(&mut c, inner).unwrap(); write_uint(&mut c, TRACK_NUMBER, 2).unwrap(); end_master(&mut c, outer).unwrap(); let data = c.into_inner(); let mut cur = Cursor::new(&data); let (oid, osize, _) = read_element_header(&mut cur).unwrap(); assert_eq!(oid, TRACKS); let outer_body_start = cur.position(); // First child of TRACKS is TRACK_ENTRY. let (iid, isize, _) = read_element_header(&mut cur).unwrap(); assert_eq!(iid, TRACK_ENTRY); // Skip TRACK_ENTRY body; the next element must be the sibling uint. cur.set_position(cur.position() + isize); let (sid, ssize, _) = read_element_header(&mut cur).unwrap(); assert_eq!(sid, TRACK_NUMBER, "sibling after inner master"); // Skip the sibling's body too, then total bytes consumed inside the // outer master must exactly equal its declared size. cur.set_position(cur.position() + ssize); let consumed = cur.position() - outer_body_start; assert_eq!(consumed, osize, "outer size must bound both children"); // And the whole buffer is exactly the outer element. assert_eq!(data.len() as u64, outer_body_start + osize); } }