Software

Extension DP provides SDKs for Python and TypeScript/JavaScript to make it easy to publish and consume Extension Data Packages.

Python

Note

In addition to the Python SDK, we recommend using frictionless-py to manage your data packages. For example, using it you can publish your data package directory to Zenodo instead of saving it locally, as well as consume it from a remote server.

Installation

pip install extensiondp frictionless

Publication

from extensiondp import Package, Table1, Table2
import frictionless

# Create Table1 records
record1 = Table1(
    id="t1-001",
    name="First Entity",
    status="active",
    value=100.5,
    itemCount=5,
    isVerified=True,
    createdDate="2024-01-15",
    description="This is the first example entity",
)

# Create Table2 records
record2 = Table2(
    id="t2-001",
    table1Id="t1-001",
    title="Related Item",
    amount=99.99,
    priority="high",
    percentage=75.5,
    notes="This item is related to the first entity",
    isActive=True,
)

package = Package(
    {
        "$schema": "https://raw.githubusercontent.com/datisthq/extensiondp/v0.1.0/extension/profile.json",
        "resources": [
            {
                "name": "table1",
                "data": [record1],
                "schema": "https://raw.githubusercontent.com/datisthq/extensiondp/v0.1.0/extension/schemas/table1.json",
            },
            {
                "name": "table2",
                "data": [record2],
                "schema": "https://raw.githubusercontent.com/datisthq/extensiondp/v0.1.0/extension/schemas/table2.json",
            },
        ],
    }
)

frictionless.Package(package).to_json("extension.json")

Validation

import frictionless

report = frictionless.validate("extension.json")
print(report)

Consumption

import frictionless

package = frictionless.Package("extension.json")
print(package)

TypeScript

Note

In addition to the TypeScript SDK, we recommend using dpkit to manage your data packages. For example, using it you can publish your data package directory to Zenodo instead of saving it locally, as well as consume it from a remote server.

Installation

npm install extensiondp dpkit

Publication

import type { Table1, Table2, Package } from "extensiondp";
import { savePackageDescriptor } from "dpkit";

const record1: Table1 = {
  id: "t1-001",
  name: "First Entity",
  status: "active",
  value: 100.5,
  itemCount: 5,
  isVerified: true,
  createdDate: "2024-01-15",
  description: "This is the first example entity",
};

const record2: Table2 = {
  id: "t2-001",
  table1Id: "t1-001",
  title: "Related Item",
  amount: 99.99,
  priority: "high",
  percentage: 75.5,
  notes: "This item is related to the first entity",
  isActive: true,
};

const dataPackage: Package = {
  $schema:
    "https://raw.githubusercontent.com/datisthq/extensiondp/v0.1.0/extension/profile.json",
  resources: [
    {
      name: "table1",
      data: [record1],
      schema:
        "https://raw.githubusercontent.com/datisthq/extensiondp/v0.1.0/extension/schemas/table1.json",
    },
    {
      name: "table2",
      data: [record2],
      schema:
        "https://raw.githubusercontent.com/datisthq/extensiondp/v0.1.0/extension/schemas/table2.json",
    },
  ],
};

await savePackageDescriptor(dataPackage, {
  path: "extension.json",
  overwrite: true,
});

Validation

import { validatePackage } from "dpkit";

const { valid, errors } = await validatePackage("extension.json");
console.log(valid, errors);

Consumption

import { loadPackageDescriptor } from "dpkit";

const dataPackage = await loadPackageDescriptor("extension.json");
console.log(dataPackage);

Command-Line

Note

As an alternative to dpkit, you can use frictionless-py to manage your data packages in Command-Line.

Installation

curl -fsSL https://dpkit.dev/install.sh | sh

Validation

./dp package validate extension.json

Consumption

./dp table explore -p extension.json