Working with Inline Data tables in Python

Updated May 16, 2026

Inline data handling for tables embedded directly in resource definitions.

Installation

pip install fairspec

Getting Started

The Inline plugin provides:

load_inline_table - Load tables from inline data
InlinePlugin - Plugin for framework integration

For example:

from fairspec import load_inline_table, Resource

table = load_inline_table(Resource(data=[
    {"id": 1, "name": "Alice"},
    {"id": 2, "name": "Bob"},
]))

Basic Usage

Object Format Data

The most common format is an array of objects:

from fairspec import load_inline_table, Resource

table = load_inline_table(Resource(data=[
    {"id": 1, "name": "english", "native": "English"},
    {"id": 2, "name": "chinese", "native": "中文"},
    {"id": 3, "name": "spanish", "native": "Español"},
]))

Array Format Data

You can also use array-of-arrays format with the first row as headers:

from fairspec import load_inline_table, Resource

table = load_inline_table(Resource(data=[
    ["id", "name", "native"],
    [1, "english", "English"],
    [2, "chinese", "中文"],
    [3, "spanish", "Español"],
]))

Advanced Features

With Table Schema

Provide a Table Schema for type validation and conversion:

from fairspec import load_inline_table, Resource
from fairspec_metadata import TableSchema, IntegerColumnProperty, StringColumnProperty, BooleanColumnProperty

table = load_inline_table(Resource(
    data=[
        {"id": 1, "name": "english", "active": True},
        {"id": 2, "name": "chinese", "active": False},
    ],
    tableSchema=TableSchema(properties={
        "id": IntegerColumnProperty(),
        "name": StringColumnProperty(),
        "active": BooleanColumnProperty(),
    }),
))

Mixed with File Data

Inline data can be used alongside file-based resources in datasets:

from fairspec import load_inline_table, load_csv_table, Resource

# Load inline reference data
languages = load_inline_table(Resource(
    name="languages",
    data=[
        {"id": 1, "name": "english"},
        {"id": 2, "name": "chinese"},
    ],
))

# Load main data from file
users = load_csv_table(Resource(name="users", data="users.csv"))

Resource Metadata

You can include metadata with inline data resources:

from fairspec import load_inline_table, Resource
from fairspec_metadata import TableSchema, StringColumnProperty

table = load_inline_table(Resource(
    name="countries",
    title="Country Reference Data",
    description="ISO country codes and names",
    data=[
        {"code": "US", "name": "United States"},
        {"code": "CN", "name": "China"},
        {"code": "ES", "name": "Spain"},
    ],
    tableSchema=TableSchema(properties={
        "code": StringColumnProperty(),
        "name": StringColumnProperty(),
    }),
))

Search