Overview

ro-crate-py is a Python library for creating and consuming Research Object Crates (RO-Crate).

https://doi.org/10.5281/zenodo.3956493

ro-crate-py is a Python library to create and consume Research Object Crates. It currently supports the RO-Crate 1.1 specification.

Goal

The goal is to create a page like the one shown below.

https://nakamura196.github.io/rocrate_demo/crate/test/data/ro-crate-preview.html

Dataset Page

Individual Item Page

JSON Data

We will create JSON data like the following.

https://nakamura196.github.io/rocrate_demo/crate/test/data/ro-crate-metadata.json

For the item ID, we use the following OAI-PMH record.

https://da.dl.itc.u-tokyo.ac.jp/portal/oai?verb=GetRecord&metadataPrefix=dcndl_simple&identifier=oai:da.dl.itc.u-tokyo.ac.jp:fbd0479b-dbb4-4eaa-95b8-f27e1c423e4b

For the item creator, we specify an ORCID ID (as a dummy value).

https://orcid.org/0000-0001-8245-7925

For the data publisher, we specify the University of Tokyo’s Research Organization Registry (ROR).

https://ror.org/057zh3y96

Installing Libraries

bagit is not required by rocrate, but we use it here to output the final result in BagIt format.

pip install rocrate
pip install bagit

from rocrate.rocrate import ROCrate
from rocrate.model.person import Person
from rocrate.model.contextentity import ContextEntity
import os
import bagit
import shutil
import json

Data

dataset_name = "百鬼夜行図コレクション"
dataset_description = "百鬼夜行図(ひやつきやぎうず) 蔭山源広迢写 百鬼夜行は今昔物語などの説話にでてくる言葉で、京の大路を夜な夜な化け物たちが練り歩く様子を表している。"
dataset_license = "https://www.lib.u-tokyo.ac.jp/ja/library/contents/archives-top/reuse"

item_id = "https://da.dl.itc.u-tokyo.ac.jp/portal/oai?verb=GetRecord&metadataPrefix=dcndl_simple&identifier=oai:da.dl.itc.u-tokyo.ac.jp:fbd0479b-dbb4-4eaa-95b8-f27e1c423e4b"
item_name = "百鬼夜行図"
item_description = "OAI-PMH（Open Archives Initiative Protocol for Metadata Harvesting）"
item_license = "https://www.lib.u-tokyo.ac.jp/ja/library/contents/archives-top/reuse"

person_id = "https://orcid.org/0000-0001-8245-7925"
person_name = "Satoru Nakamura"

org_id = "https://ror.org/057zh3y96"
org_name = "The University of Tokyo"

Creating an ROCrate Instance

By setting gen_preview=True, a preview.html file is automatically created when saving.

crate = ROCrate(gen_preview=True)

Creating root_dataset Metadata

root_dataset = crate.root_dataset
root_dataset["name"] = dataset_name
root_dataset["description"] = dataset_description
root_dataset["license"] = dataset_license

Creating an Item

Here, we add remote entities.

https://github.com/ResearchObject/ro-crate-py?tab=readme-ov-file#adding-remote-entities

item = crate.add_file(item_id, properties={
    "name": item_name,
    "description": item_description,
    "license": item_license
})

Adding a Creator

person = Person(crate, person_id, properties={
    "name": person_name})
crate.add(person)

Add the Person as the item’s author.

item["author"] = person

Adding a Publishing Organization

class Organization(ContextEntity):

    def __init__(self, crate, identifier=None, properties=None):
        super(Organization, self).__init__(crate, identifier, properties)

    def _empty(self):
        val = {
            "@id": self.id,
            "@type": 'Organization'
        }
        return val

org = Organization(crate, org_id, properties={
    "name": org_name})

crate.add(org)

root_dataset["publisher"] = org

Output

Here, we set the output directory to docs/crate/test.

output_dir = f"docs/crate/test"

if os.path.exists(output_dir):
    shutil.rmtree(output_dir)

crate.write(output_dir)

You can also save as a compressed file using write_zip.

crate.write_zip(output_dir)

Japanese Character Support

By default, Japanese characters are escaped, so set ensure_ascii to False.

output_path = f"{output_dir}/ro-crate-metadata.json"

# Load the escaped JSON file
with open(output_path, 'r', encoding='utf-8') as f:
    data = json.load(f)

# Write back as unescaped JSON
with open(output_path, 'w', encoding='utf-8') as f:
    json.dump(data, f, ensure_ascii=False, indent=4)

Creating a BagIt Archive

bag = bagit.make_bag(output_dir, {"Contact-Name": org_name})
shutil.make_archive(f'docs/bagit/test', format='zip', root_dir=output_dir)

Additional Notes

While ROCrate(gen_preview=True) can generate a preview.html, using the following module allows you to create a preview.html like the one shown at the beginning of this article.

npm install ro-crate-html-js
node node_modules/ro-crate-html-js/roc-html.js {output_path}

Summary

We hope this serves as a useful reference for working with RO-Crate.