fastokf

fastokf parses Open Knowledge Format (OKF) concept documents into Pydantic models.

OKF documents are Markdown files with YAML frontmatter. fastokf reads them into useful, typed Python objects.

Install

This project targets Python 3.10 and newer.

uv sync

For development, install the test group too:

uv sync --group dev

Parse a document

from fastokf import parse_string

document = parse_string("""---
type: Metric
title: Revenue
tags: [finance]
generated:
  by: human:finance-team
  at: 2026-06-20T22:53:05Z
---
# Definition

Recognized revenue for a fiscal year.
""")

print(document.frontmatter.title)  # Revenue
print(document.body)

Input sources

Parse documents from text, files, URLs, or open text streams:

from fastokf import parse_file, parse_file_object, parse_string, parse_url

parse_string(okf_text)
parse_file("knowledge/metrics/revenue.md")
with open("knowledge/metrics/revenue.md", encoding="utf-8") as file:
    parse_file_object(file)
parse_url("https://example.com/knowledge/metrics/revenue.md")

parse() also accepts any of these sources directly:

from pathlib import Path

from fastokf import parse

parse(Path("knowledge/metrics/revenue.md"))  # filesystem path
parse("https://example.com/revenue.md")      # HTTP(S) URL
parse("---\ntype: Metric\n---\n")             # document text

Development

uv run pytest