/osfclient

A python library and command-line client for file storage on OSF

Primary LanguagePythonBSD 3-Clause "New" or "Revised" LicenseBSD-3-Clause

osfclient

osfclient

travisbadge

The osfclient is a python library and a command-line client for up- and downloading files to and from your Open Science Framework projects. The Open Science Framework (OSF) is an open source project which facilitates the open collaboration of researchers on the web, by sharing data and other research outputs.

As such the OSF hosts large data sets, associated with papers or scientific projects, that can be freely downloaded. The osfclient allows people to store and retrieve large datasets associated to their scientific projects and papers on the OSF via the command line interface. If you are completely new to the OSF you can read their introductory materials

This is a very new project, it has some rough edges.

Installing

To use osfclient install it via pip:

$ pip install osfclient

For details on participating in the development of osfclient check out the Contributing section.

Usage

This project provides two things: a python library and a command-line program for interacting with files stored in the OSF.

The python library forms the basis for the command-line program. If you want programmatic access to your files use the library, otherwise try out the command-line program.

Read the full documentation: https://osfclient.readthedocs.io/en/latest/

Below are some examples on how to use it:

# get help and see available commands, get help on a specific command
$ osf -h
$ osf <command> -h

# setup a local folder for an existing project
$ osf init

# list all files for the project
$ osf ls

# fetch all files for the project
$ osf clone

# fetch an individual file from a project
$ osf fetch remote/path.txt local/file.txt

# get web view url for an individual file from a project
$ osf geturl remote/path.txt

# add a new file
$ osf upload local/file.txt remote/path.txt

# add a new directory
$ osf upload -r local/directory/ remote/directory

If the project is private you will need to provide authentication details. You can provide either username & password credentials or a Personal Access Token (PAT). You can provide these by setting either the OSF_USERNAME and OSF_PASSWORD environment variables or by setting the OSF_TOKEN environment variable. The password will be retrieved from the OSF_PASSWORD environment variable or you will be asked directly by the tool when you run it.

You can set default values for the username and project by using a configuration file in the current directory. This is what osf init does for you. To set the username and project ID create .osfcli.config:

[osf]
username = yourOSFaccount@example.com
project = 9zpcy

To avoid having to provide credentials on each use, you can provide either your password or a PAT in your config with the following keys:

# basic auth (username/password)
password = this-password-is-fake

# token auth
token = kej2R9IU6Gr2uThsswSNdP1cd0cu9eaCerVXjVf7zNwfXHyT0QzMZtX0PGTYmp9Fzaixwq

After which you can simply run osf ls to list the contents of the project.

JSON-LD Support

You can find some json-ld support for project metadata. To create a project in OSF with JSON-LD, you can use the following command.

from osfclient import OSF
osf = OSF()
osf.login(token="XYZ")
osf.create_project_jsonld(jsonld)

The jsonld object have to have the following structure. Title and category are mandatory, description and keywords should be given by the user, otherwise it is empty.

{
    "https://schema.org/category":"project",
    "https://schema.org/description":"this is a test for preprint citations",
    "https://schema.org/dateModified":"2017-03-17T16:11:35.721000",
    "https://schema.org/title":"Preprint Citations Test",
    "https://schema.org/dateCreated":"2017-03-17T16:09:14.864000",
    "https://schema.org/publicAccess":true,
    "https://schema.org/keywords":[
        "qatest"
    ],
    "https://schema.org/identifier":"f3szh",
    "https://schema.org/url":"https://api.osf.io/v2/nodes/f3szh/",
    "https://schema.org/downloadUrl":"https://api.osf.io/v2/nodes/f3szh/files/"
}

All other keys are optional or cannot be set by the user, but will be printed out, when you call want to get the metadata.

from osfclient import OSF
osf = OSF()
osf.login(token="XYZ")
osf.project("f3szh").metadata(jsonld=True)

Contributing

Contributions from everyone and anyone are welcome. Fork this repository, make your changes, add a test to cover them and create a Pull Request. Then one of the maintainers will review your changes. When all comments have been addressed and all tests pass your changes will be merged.

To setup a development version:

$ git clone https://github.com/YOURNAMEHERE/osfclient
$ git remote add upstream https://github.com/osfclient/osfclient
$ cd osfclient
$ pip install -r devRequirements.txt -c constraints.txt
$ pip install -e . -c constraints.txt

There are a few secret keys relevant to this project, like passwords to pypi.org, test.pypi.org, and the osfclient email account. We store these in an encrypted git repo on Keybase. If you need access to this repo, contact any of the following maintainters on Keybase:

  • Tim Head (@betatim)
  • Ben Lindsay (@benlindsay)
  • Fitz Elliott (@felliott)
  • Longze Chen (@cslzchen)

For more details and instructions: CONTRIBUTING.md