Usage

List archive contents: lha l archive.lzh

The most common operation. The output lists every member of the archive with size, ratio, and timestamp:

$ lha l README.lzh
PERM  UID  GID    SIZE  RATIO     STAMP           NAME
[generic]                  1234  60.0%  Jul 14 20:30  README.md
[generic]                  5678  72.5%  Jul 14 20:30  src/main.c
---------- -------------------- ------- ----------
                      2 files,    6912 bytes (62.5%)

Add v for verbose (path permissions + CRC):

$ lha lv README.lzh

Use v twice for ultra-verbose (per-block info):

$ lha lvv README.lzh

Extract: lha x archive.lzh

The default extract mode. Files are written relative to the current directory, with the path as stored in the archive:

$ lha x archive.lzh

Common options:

• q (quiet) — skip per-file output: lha xq archive.lzh
• w=DIR — extract into DIR (creates if missing): lha xw=output archive.lzh
• f — force overwrite without prompting: lha xf archive.lzh
• i — ignore directory entries (extract files but skip recreating the directory hierarchy): lha xi archive.lzh

Filter to specific members using glob patterns:

$ lha x archive.lzh "*.c"        # only *.c members
$ lha x archive.lzh src/*        # only files under src/

Verify CRC: lha t archive.lzh

Walks the archive, recomputes each member's CRC16, compares against the stored value. Exit code 0 = all members OK; non-zero = a member failed CRC (archive is corrupt or partially written):

$ lha t README.lzh
test README.lzh            : OK

Use this in CI artifact verification — e.g. after downloading an .lzh from an upstream mirror, sanity-check it with lha t before extracting.

Print to stdout: lha p archive.lzh file

Print a single archive member to stdout, decompressed. Equivalent to tar -xOf archive.lzh file:

$ lha p README.lzh src/main.c | head -5
/* main.c — entry point */

With no member filter, lha p prints every member in sequence. Useful for piping into grep / less / another archiving tool:

$ lha p archive.lzh | grep TODO

What lhasa does not do

Lhasa is decoder-only. It cannot:

• Create a new .lzh archive. Use the original lha binary built from upstream jca02266/lha (LHa for UNIX 1.14i) for that.
• Write a new .lzs or .pma archive. Same story — lhasa's encoder path was never developed on purpose.

For reads (the 99% use case in 2026), lhasa is the canonical free tool.