Overview
ISO/IEC 11558:1992 defines the DCLZ algorithm - a lossless, dictionary-based compression method (Data Compression according to Lempel and Ziv) for 8‑bit byte streams. Intended primarily for use with interchangeable recording media (magnetic tape, cartridges, optical disks), DCLZ reduces the number of bits required to represent input bytes while allowing exact restoration by a complementary decompression algorithm. The algorithm is registered with identifier 32 and is structured to support sequential record boundaries and efficient recreation of the dictionary on decompression.
Key Topics and Requirements
- Lossless compression: original data must be exactly reconstructible.
- Dictionary-based adaptive coding: a dictionary of up to 3 832 entries stores repeated byte pairs/strings; entries are assigned Code Values.
- Code Values (0–4095):
- 0–7: Control codes (0 = Dictionary Frozen, 1 = Dictionary Reset, 2 = Increment Codeword Size, 3 = End of Record; values 4–7 unused)
- 8–263: Encoded Bytes (single byte values encoded as byte + 8)
- 264–4095: Dictionary Codes (references to stored strings)
- Codewords and bit packing: Codeword sizes begin at 9 bits and can increase (9–12 bits supported) via the Increment Codeword Size control code. Codewords are packed into 8‑bit bytes by entering bits least-significant first into bytes.
- String limits and dictionary behavior:
- Maximum string length for dictionary entries: 128 bytes
- Dictionary states: empty, frozen (no further entries), and reset (allowed anytime after pending bytes expressed as Codewords)
- Strategies for freezing/resetting the dictionary are implementation-dependent (standard specifies behavior but not policy)
- Record boundaries: EOR control code ensures record boundaries in the input stream are preserved in the output.
Applications and Who Should Use It
- Implementers of compression in storage systems, archival software, device firmware for tape/optical media, and file transfer utilities that prioritize interoperability and lossless recovery.
- Hardware and firmware engineers building compression-capable drives or controllers where compact interchangeable medium formats are required.
- Software developers integrating standard-compliant compression/decompression libraries for archival, backup, or interchange formats.
- Standards bodies and product teams seeking a registered, well-documented Lempel–Ziv style algorithm (DCLZ) for compatibility across systems.
Related Standards
- ISO/IEC 11576 - Procedure for registration of lossless compression algorithms (relevant for algorithm identifier and registration).
- The DCLZ standard was prepared from ECMA-151 (reference to the originating ECMA document).
Keywords: ISO/IEC 11558:1992, DCLZ, lossless compression, dictionary-based compression, Lempel–Ziv, adaptive coding, embedded dictionary, codeword, End of Record (EOR).