FastGedcom

A lightweight tool to easily parse, browse and edit gedcom files.

Install FastGedcom using pip from the PyPI repository:

pip install fastgedcom

To install the Ansel codec use the following command. It enables the use of the Ansel text encoding often used for gedcom files.

pip install fastgedcom[ansel]

Highlights of FastGedcom

• FastGedcom is easy to write.
• FastGedcom has type annotations.
• FastGedcom has a documentation.
• FastGedcom has code examples.
• FastGedcom has unit tests.
• FastGedcom has less methods than the alternatives, which make it easy to learn.
• FastGedcom is concise thanks to operator overloads. (optional)
• FastGedcom has a linear syntax, if/else and try/except blocks are less needed.
• Last but not least, FastGedcom is fast. Go to benchmarks.

Comparison:

Gedcom file	FastGedcom	python-gedcom
0 HEAD 1 FILE my-file.ged 0 @I1@ INDI 1 NAME John Doe 1 BIRT 2 DATE 1 Jan 1970 1 DEAT 2 DATE 2 Feb 2081 0 TRLR	from fastgedcom.parser import strict_parse document = strict_parse("my-file.ged") person = document["@I1@"] # use ">" to get a sub-line death = person > "DEAT" # use ">=" to get a sub-line value date = death >= "DATE" print(date) # Prints "" if the field is missing	from gedcom.parser import Parser document = Parser() document.parse_file("my-file.ged") records = document.get_element_dictionary() person = records["@I1@"] death_data = person.get_death_data() # data is (date, place, sources) date = death_data[0] print(date)

Features

Multi-encoding support

It supports a broad set of encoding for gedcom files such as UTF-8 (with and without BOM), UTF-16 (also named UNICODE), ANSI, and ANSEL.

Kept closed from gedcom with free choice of formatting

There is a lot of genealogy software out there, and every one of them have its own tags and formats to write information. With the FastGedcom approach, you can easily adapt your code to your gedcom files. You have to choose how do you want to parse and format the values. You can use non-standard field, for example the "_AKA" field (standing for Also Known As).

from fastgedcom.parser import strict_parse
from fastgedcom.helpers import extract_name_parts

document = strict_parse("gedcom_file.ged")

person = document["@I1@"]
name = person >= "NAME"
print(name)  # Unformatted string such as "John /Doe/"

given_name, surname = extract_name_parts(name)
print(f"{given_name.capitalize()} {surname.upper()}")  # Would be "John DOE"

alias = person > "NAME" >= "_AKA"
print(f"a.k.a: {alias}")  # Could be "Johnny" or ""

The Option paradigm replaces the if blocks:

If a field is missing, you will get a FakeLine containing an empty string. This helps reduce the boilerplate code massively. And, you can differentiate a TrueLine from a FakeLine with a simple boolean check.

indi = document["@I13@"]

# You can access the date of death, whether the person is deceased or not.
date = (indi > "DEAT") >= "DATE"

# The date of death or an empty string
print("Death date:", date)

Another example:

for record in document:
    line = record > "_UID"
    if line:  # Check if field _UID exists to avoid ValueError in list.remove()
        record.sub_lines.remove(line)

# Get the Document as a gedcom string to write it into a file
gedcom_without_uids = document.get_source()

with open("./gedcom_without_uids.ged", "w", encoding="utf-8-sig") as f:
    f.write(gedcom_without_uids)

Typehints for salvation!

Autocompletion and type checking make development so much easier.

• There are only 3 main classes: Document, TrueLine, and FakeLine.
• There are type aliases for code clarity and code documentation: Record, XRef, IndiRef, FamRef, and more.

from fastgedcom.base import Record, FakeLine
from fastgedcom.family_link import FamilyLink

# For fast and easy family lookups
families = FamilyLink(document)


def ancestral_generation_count(indi: Record | FakeLine) -> int:
    """Return the number of generation registered as ancestors of the given person."""
    if not indi:
        return 1
    father, mother = families.get_parents(indi.tag)
    return 1 + max(
        ancestral_generation_count(father),
        ancestral_generation_count(mother),
    )


root = document["@I1@"]
number_generations_above_root = ancestral_generation_count(root)

Why it is called FastGedcom?

FastGedcom's aim is to keep the code close to your gedcom files. So, you don't have to learn what FastGedcom does. The data you have is the data you get. The content of the gedcom file is unchanged and there is no abstraction. Hence, the learning curve of the library is faster than the alternatives. The data processing is optional to best suit your needs. FastGedcom is more of a starting point for your data processing than a feature-rich library.

The name FastGedcom doesn't just come from its ease of use. Parsing is the fastest among Python libraries. Especially for parsing and getting the relatives of a person, the FamilyLink class is build for this purpose. Here are the benchmarks.

Documentation and examples

Want to see more of FastGedcom? Here are some examples

The documentation of FastGedcom is available on ReadTheDocs.

Feedback

Comments and contributions are welcomed, and they will be greatly appreciated!

If you like this project, consider putting a star on GitHub. Thank you!

For any feedback or questions, please feel free to contact me by email at [email protected] or via GitHub issues.