A simple Python API that can interact with datalab instances.
The idea here is to provide a set of utility functions and models for manipulating samples, cells, inventory, files, users and metadata associated with datalab entries in an automated way.
This API may not expose all the functionality available in a given datalab instance, and the idea would be that this package can support multiple versions of the underlying datalab REST API.
This means that the API is primarily functional in nature, as opposed to object-oriented, since datalab instances are free to use their own custom data models.
The available schemas are reported as instance metadata and in the future object-oriented models may be able to be genereated directly in the client (so e.g., the returned data would be Python objects like Sample
rather than JSON data).
The issue tracker here will be used to track development of features, as well as bug reports. If you have any suggestions or feedback, please post it there.
The API can be used by installing this repository with pip
, ideally in a fresh Python 3.9+ environment (created using e.g., conda, virtualenv or other related tools -- if you're not sure about this, ask).
Either from PyPI, for the latest released version:
pip install datalab-api
or for the latest development version from GitHub:
git clone git@github.com:datalab-org/datalab-api
cd datalab-api
pip install .
Example usage as a Jupyter notebook can be found in the examples
directory or
in the online documentation, as
well as the full API
documentation.
Currently the only supported authentication method is via an API key.
You can generate one for your account for a given datalab instance by visiting the /get-api-key
endpoint of your chosen instance, or, if using a recent version of datalab, by visiting your account settings in the browser.
This API key can be set via the environment variable DATALAB_API_KEY
.
To suport the use case of needing to interact with multiple datalab instances, the client will also check prefixed environment variables that use the IDENTIFIER_PREFIX
of the chosen datalab instance, e.g., GREY_DATALAB_API_KEY
or PUBLIC_DATALAB_API_KEY
.
Only keys that match will be read (e.g., other environment variables starting with PUBLIC_
will be ignored, when connecting to the public demo datalab).
This package implements basic functionality for displaying and manipulating entries:
from datalab_api import DatalabClient
with DatalabClient("https://public.api.odbx.science") as client:
# List all items of a given type
items = client.get_items()
# Get more info on a particular item
item = client.get_item(item_id="test")
# Upload a file to an item
file_response = client.upload_file(filepath="my_echem_data.mpr", item_id="test")
There is also an EXPERIMENTAL CLI, exposed via the datalab
command.
This CLI has many subcommands, which can be listed with datalab help
or by entering the shell with just datalab
.
For now, the CLI does not expose all of the functionality of the underlying API, and is mostly oriented on GET-focused operations.
$ datalab
โญโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ
โ โ
โ oooo o8 o888 oooo โ
โ ooooo888 ooooooo o888oo ooooooo 888 ooooooo 888ooooo โ
โ 888 888 ooooo888 888 ooooo888 888 ooooo888 888 888 โ
โ 888 888 888 888 888 888 888 888 888 888 888 888 โ
โ 88ooo888o 88ooo88 8o 888o 88ooo88 8o o888o 88ooo88 8o o888ooo88 โ
โ โ
โฐโ Copyright (c) 2020-2024 Matthew Evans, Joshua Bocarsly & the Datalab Development Team. โโฏ
datalab > help
Documented commands (type help <topic>):
========================================
authenticate get info
Undocumented commands:
======================
exit help quit
datalab > help get
Usage: [INSTANCE_URL] get [OPTIONS] ITEM_TYPE [INSTANCE_URL]
Get a table of items of the given type.
โญโ Arguments โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ
โ * item_type TEXT [default: None] [required] โ
โ instance_url [INSTANCE_URL] [default: None] โ
โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
โญโ Options โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฎ
โ --page-limit INTEGER [default: 10] โ
โ --api-key TEXT [default: None] โ
โ --log-level TEXT [default: WARNING] โ
โ --help Show this message and exit. โ
โฐโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฏ
datalab > get samples api.odbx.science --page-limit 2
/samples/
โโโโโโโโณโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโณโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโ
โ type โ ID โ refcode โ name โ nblocks โ collections โ creators โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ C โ test_api โ grey:WHQFXG โ โ 0 โ โ Matthew Evans โ
โโโโโโโโผโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโผโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโค
โ S โ test12331231312 โ grey:GFSUQM โ โ 1 โ โ Matthew Evans โ
โโโโโโโโดโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโดโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโ