LHARK

From Just Solve the File Format Problem
Jump to: navigation, search
File Format
Name LHARK
Ontology
Extension(s) .lzh
Released ≤1995
Not to be confused with LHarc – see LHA.

LHARK is a compressed archiving utility for DOS, developed by Kerwin F. Medina. It uses LHA file format, but adds a custom compression scheme, with identifier "lh7".

Unfortunately, this conflicts with a different lh7 compression scheme supported by some of the main-line LHA software (e.g. LHA 2.66+, LHa for Unix).

Contents

Self-extracting archives

LHARK can create self-extracting archives, in DOS EXE format. Versions 0.3e through 0.4d have been evaluated.

The extraction code is compressed with LZEXE, but modified in order to run some code before LZEXE decompression takes over. An LHARK archive is embedded in the overlay segment (refer to MS-DOS EXE#Special file positions).

Refer to LHA/LHarc self-extracting archive for LHA self-extracting archives in general.

Identification

LHARK uses generalized LHA format, so start with LHA#Identification.

Distinguishing LHARK format can be problematic. An LHA member file with all of the following properties almost certainly uses LHARK compression:

  • The compression method field starting at offset 2 is "-lh7-".
  • The byte at offset 20 is 0x01 (the header level).
  • The byte at offset 24+{the value of the byte at offset 21} is 0x20. This "OS code" (an ASCII space) is reasonably, though not completely, specific to LHARK.

Note that if the first member file of an LHARK-created file was incompressible, then the standard "-lh0-" compression method ID will appear at file offset 2, instead of "-lh7-".

Software

Sample files

Personal tools
Namespaces

Variants
Actions
Navigation
Toolbox