{
  "status": "shipping",
  "advantages": [],
  "limitations": [],
  "supersedes": [],
  "supersededBy": [],
  "firstSeen": null,
  "lastSeen": null,
  "updated": "2026-10-18",
  "slug": "archive-member",
  "term": "Archive member",
  "aliases": [
    "TAR member"
  ],
  "definition": "An archive member is an individual stored entry with its own metadata and, where applicable, payload, whose name need not be unique within an archive.",
  "note": "GNU tar permits repeated member names. Extracting regular entries to disk can overwrite the first stored version with the last, while --to-stdout concatenates selected payloads in archive order. The name alone does not specify which occurrence was recovered. [tar-append] [tar-stdout]\n\nPython's TarFile.getmember(name) selects the last occurrence; getmembers() preserves archive order. Passing a TarInfo object to extractfile() retains entry identity. Re-looking up each iterated member by name can repeatedly read the final occurrence instead. [python-tarfile]\n\nHesela measured four designed TAR fixtures with eight reader operations each. With A then B under one name, disk extraction returned B, stdout returned A+B, and a Python name-relookup loop returned B+B. The intact outer gzip round trip did not resolve member selection. This is a bounded software demonstration, not an archive-corruption rate. [hesela-members]\n\nA selected-member read is narrower than whole-archive verification. In three further synthetic TAR fixtures, --occurrence=1 returned the correct first file despite a damaged later header. Full listing detected that header damage, but returned success after a same-length mutation of the later file's payload; extracting it also returned success with wrong bytes. Basic USTAR checksums cover headers, not payloads. Match the intended content against an independent manifest, not merely a successful listing. [hesela-coverage] [tar-format]",
  "related": [
    "backup-manifest",
    "restore-test",
    "compression-dictionary",
    "cryptographic-hash"
  ],
  "citations": [
    {
      "id": "tar-append",
      "title": "GNU tar 1.35: appending files and repeated member names",
      "url": "https://www.gnu.org/software/tar/manual/html_node/append.html",
      "accessed": "2026-10-18"
    },
    {
      "id": "tar-stdout",
      "title": "GNU tar 1.35: Writing to Standard Output",
      "url": "https://www.gnu.org/software/tar/manual/html_node/Writing-to-Standard-Output.html",
      "accessed": "2026-10-18"
    },
    {
      "id": "python-tarfile",
      "title": "Python 3.12 tarfile: getmember, getmembers and extractfile",
      "url": "https://docs.python.org/3.12/library/tarfile.html#tarfile.TarFile.getmember",
      "accessed": "2026-10-18"
    },
    {
      "id": "hesela-members",
      "title": "Hesela: measured TAR member selection in archive recovery",
      "url": "https://hesela.com/analyses/compressed-archives-dictionary-dependencies/#tar-members",
      "accessed": "2026-10-18"
    },
    {
      "id": "hesela-coverage",
      "title": "Hesela: twelve measured TAR read-coverage operations",
      "url": "https://hesela.com/analyses/compressed-archives-dictionary-dependencies/#tar-coverage",
      "accessed": "2026-10-18"
    },
    {
      "id": "tar-format",
      "title": "GNU tar 1.35: Basic Tar Format, header checksum",
      "url": "https://www.gnu.org/software/tar/manual/html_node/Standard.html",
      "accessed": "2026-10-18"
    }
  ],
  "provenance": [
    "GNU tar 1.35 manual and Python 3.12 documentation; measured GNU tar 1.35 / Python 3.12.3. Literal regular-file names only; no hostile-archive security guarantee.",
    "Reproduction: https://hesela.dev/research/tar-member-selection.py ; observations: https://hesela.dev/research/tar-member-selection.json",
    "Coverage extension: https://hesela.dev/research/tar-read-coverage.py ; observations: https://hesela.dev/research/tar-read-coverage.json . Plain USTAR only; command outcomes are not a trace of physical reads."
  ]
}