Datasette plugin for authenticating access using passwords
Install this plugin in the same environment as Datasette.
datasette install datasette-auth-passwords
A demo of this plugin is running at https://datasette-auth-passwords-demo.datasette.io/
The demo is configured to show the public.db
database to everyone, but the private.db
database only to logged in users.
You can log in at https://datasette-auth-passwords-demo.datasette.io/-/login with username root
and password password!
.
This plugin works based on a list of username/password accounts that are hard-coded into the plugin configuration.
First, you'll need to create a password hash. There are three ways to do that:
- Install the plugin, then use the interactive tool located at
/-/password-tool
- Use the hosted version of that tool at https://datasette-auth-passwords-demo.datasette.io/-/password-tool
- Use the
datasette hash-password
command, described below
Now add the following to your metadata.json
:
{
"plugins": {
"datasette-auth-passwords": {
"someusername_password_hash": {
"$env": "PASSWORD_HASH_1"
}
}
}
}
The password hash can now be specified in an environment variable when you run Datasette. You can do that like so:
PASSWORD_HASH_1='pbkdf2_sha256$...' \
datasette -m metadata.json
Be sure to use single quotes here otherwise the $
symbols in the password hash may be incorrectly interpreted by your shell.
You will now be able to log in to your instance using the form at /-/login
with someusername
as the username and the password that you used to create your hash as the password.
You can include as many accounts as you like in the configuration, each with different usernames.
The plugin exposes a new CLI command, datasette hash-password
. You can run this without arguments to interactively create a new password hash:
datasette hash-password
Password:
Repeat for confirmation:
pbkdf2_sha256$260000$1513...
Or if you want to use it as part of a script, you can add the --no-confirm
option to generate a hash directly from a value passed to standard input:
echo 'my password' | datasette hash-password --no-confirm
pbkdf2_sha256$260000$daa...
By default, a logged in user will result in an actor block that just contains their username:
{
"id": "someusername"
}
You can customize the actor that will be used for a username by including an "actors"
configuration block, like this:
{
"plugins": {
"datasette-auth-passwords": {
"someusername_password_hash": {
"$env": "PASSWORD_HASH_1"
},
"actors": {
"someusername": {
"id": "someusername",
"name": "Some user"
}
}
}
}
}
This plugin defaults to implementing login using an HTML form that sets a signed authentication cookie.
You can alternatively configure it to use HTTP Basic authentication instead.
Do this by adding "http_basic_auth": true
to the datasette-auth-passwords
block in your plugin configuration.
This option introduces the following behaviour:
- Account usernames and passwords are configured in the same way as form-based authentication
- Every page within Datasette - even pages that normally do not use authentication, such as static assets - will display a browser login prompt
- Users will be unable to log out without closing their browser entirely
There is a demo of this mode at https://datasette-auth-passwords-http-basic-demo.datasette.io/ - sign in with username root
and password password!
If you are publishing data using a datasette publish command you can use the --plugin-secret
option to securely configure your password hashes (see secret configuration values).
You would run the command something like this:
datasette publish cloudrun mydatabase.db \
--install datasette-auth-passwords \
--plugin-secret datasette-auth-passwords root_password_hash 'pbkdf2_sha256$...' \
--service datasette-auth-passwords-demo
This will allow you to log in as username root
using the password that you used to create the hash.
To set up this plugin locally, first checkout the code. Then create a new virtual environment:
cd datasette-auth-passwords
python3 -mvenv venv
source venv/bin/activate
Or if you are using pipenv
:
pipenv shell
Now install the dependencies and tests:
pip install -e '.[test]'
To run the tests:
pytest