gedcom-lite 0.1.0

A lightweight, fidelity-preserving GEDCOM parser, writer, and CLI for genealogy files (5.5.1, 5.5.5, and FamilySearch GEDCOM 7).

gedcom-lite

A lightweight, fidelity-preserving GEDCOM parser, writer, and command-line toolkit for genealogy files. Supports GEDCOM 5.5.1, 5.5.5, and the current FamilySearch GEDCOM 7.0+ standard.

The "lite" in the name is a deliberate scope choice: pure Python, no runtime dependencies, no schema validation, no genealogy domain modelling beyond what the file format itself defines. The library models GEDCOM as a tree of Structure records and exposes search and mutation primitives that round-trip every file in the official sample suite byte-for-byte.

Why use it

• Round-trip fidelity. Parse a file and write it back unchanged: BOM, line endings, CONC/CONT choice, sibling order, and unknown extension tags are all preserved.
• Targeted edits. Mutation primitives mark only the structures you change as dirty; every other line emits byte-identical to the source. An update that touches a single NAME produces a one-line diff.
• Encoding coverage. Detects UTF-8 (with or without BOM), UTF-16 LE/BE (with BOM), and ANSEL (legacy 5.5.1) — the ANSEL codec ships in-package.
• No dependencies. Pure Python 3.11+ standard library.
• Three CLI tools. gedcom-read, gedcom-search, gedcom-update cover the common verbs without scripting.

Install

# from PyPI
pip install gedcom-lite

# from git
pip install "git+https://github.com/vaelen/gedcom-lite@v0.1.0"

# editable, for hacking
git clone https://github.com/vaelen/gedcom-lite
cd gedcom-lite
pip install -e ".[dev]"

Library usage

from gedcom_lite import GedcomDocument

doc = GedcomDocument.parse("tree.ged")

print(doc.version)               # '7.0'
print(doc.encoding)              # 'utf-8-sig'
print(doc.record_counts())       # {'INDI': 142, 'FAM': 47, ...}

indi = doc.resolve("@I1@")
name = indi.find("NAME")
name.set_payload("Jane /Doe/")

doc.write("tree-edited.ged")

The full surface is documented in src/gedcom_lite/__init__.py. Common helpers:

from gedcom_lite import (
    GedcomDocument, Structure,
    parse_date_value, DateValue,
    parents_of, children_of, ancestors_of, descendants_of,
    document_to_dict, structure_to_dict,
)

CLI usage

Three console scripts are installed:

gedcom-read   FILE [...]
gedcom-search FILE [filters...]
gedcom-update FILE [-o OUT | --in-place] SUBCOMMAND ...

Examples:

# overview of a file
gedcom-read tree.ged

# list every individual
gedcom-read tree.ged list INDI

# show one record with pointers resolved
gedcom-read tree.ged record @I1@

# JSON dump for piping into jq
gedcom-read tree.ged --json | jq '.records | length'

# find people named Smith
gedcom-search tree.ged --person Smith

# people born between 1900 and 1910
gedcom-search tree.ged --born-between 1900 1910

# children, parents, ancestors, descendants
gedcom-search tree.ged --children-of @I1@
gedcom-search tree.ged --ancestors-of @I1@ --depth 4

# edit a name (default writes to stdout)
gedcom-update tree.ged set-payload @I1@ NAME "Jane /Doe/" > new.ged

# add a NOTE to a record (write to a file)
gedcom-update tree.ged -o new.ged add-substructure @I1@ "" NOTE "Met at reunion"

# delete a record, voiding inbound pointers
gedcom-update tree.ged -o new.ged delete-record @I7@ --orphan-pointers void

# overwrite the source (be sure)
gedcom-update tree.ged --in-place set-payload @I1@ NAME "Jane /Doe/"

gedcom-update never modifies the input file unless --in-place is given.

Run any tool with --help for the full flag list.

Tests

pip install -e ".[dev]"
pytest

The suite covers:

• Tokenizer edge cases (xref slot, missing payload, malformed lines)
• Encoding detection (BOM-first, then 1 CHAR, then ANSEL)
• ANSEL codec round-trips, including combining-mark ordering
• Date parsing across ABT, BET … AND …, FROM … TO …, calendar escapes
• Round-trip of every fixture in examples/ (the gating test)
• Mutation primitives, including the "only-touched-lines-change" property
• Ancestry/descendancy traversal
• CLI smoke tests for gedcom-read, gedcom-search, gedcom-update

Releasing

Maintainer notes — releases publish to PyPI from .github/workflows/release.yml via OIDC trusted publishing (no stored tokens).

1. Bump the version in pyproject.toml, src/gedcom_lite/__init__.py, and CHANGELOG.md. Push the changes to the main branch.
2. (Optional) Rehearse to TestPyPI: gh workflow run release.yml.
3. Cut the release: gh release create vX.Y.Z --generate-notes. The workflow builds and uploads to PyPI on the release: published event.

The PyPI / TestPyPI trusted-publisher entries and the matching pypi / testpypi GitHub Environments are one-time setup, documented at the top of release.yml. CI runs pytest on push and PR across Python 3.11/3.12/3.13.

Versions in scope

• GEDCOM 5.5.1 (1999) — the de-facto interchange format used by virtually every genealogy app shipped before 2022.
• GEDCOM 5.5.5 (2019) — cleanup release of the 5.5.x line.
• FamilySearch GEDCOM 7.0+ (2021–) — current, actively maintained. UTF-8 only, real extension mechanism, GEDZIP packaging, no CONC.

GEDCOM 7 is a breaking change from 5.5.x. gedcom-lite reads any of the three; on write it never silently promotes a file to a different version.

License

MIT — see LICENSE.

References

• GEDCOM 5.5.5 specification
• FamilySearch GEDCOM 7 specification
• FamilySearch GEDCOM Compatibility Guide