Plugins > datasette-publish-vercel

datasette-publish-vercel by simonw

1,202 downloads this week        Star

README source code

datasette-publish-vercel

PyPI Changelog Tests License

Datasette plugin for publishing data using Vercel.

Installation

Install this plugin in the same environment as Datasette.

$ datasette install datasette-publish-vercel

Usage

First, install the Vercel CLI tool by following their instructions.

Run vercel login to login to (or create) an account.

Now you can use datasette publish vercel to publish your data:

datasette publish vercel my-database.db --project=my-database

The --project argument is required - it specifies the project name that should be used for your deployment. This will be used as part of the deployment's URL.

Other options

  • --no-prod deploys to the project without updating the "production" URL alias to point to that new deployment. Without that option all deploys go directly to production.
  • --debug enables the Vercel CLI debug output.
  • --token allows you to pass a Now authentication token, rather than needing to first run now login to configure the tool. Tokens can be created in the Vercel web dashboard under Account Settings -> Tokens.
  • --public runs vercel --public to publish the application source code at /_src e.g. https://datasette-public.now.sh/_src and make recent logs visible at /_logs e.g. https://datasette-public.now.sh/_logs
  • --generate-dir - by default this tool generates a new Vercel app in a temporary directory, deploys it and then deletes the directory. Use --generate-dir=my-app to output the generated application files to a new directory of your choice instead. You can then deploy it by running vercel in that directory.
  • --setting default_page_size 10 - use this to set Datasette settings, as described in the documentation. This is a replacement for the unsupported --extra-options option.

Full help

Warning: Some of these options are not yet implemented by this plugin. In particular, the following do not yet work:

  • --extra-options - use --setting described above instead.
  • --plugin-secret
  • --version-note
$ datasette publish vercel --help

Usage: datasette publish vercel [OPTIONS] [FILES]...

  Publish to https://vercel.com/

Options:
  -m, --metadata FILENAME         Path to JSON/YAML file containing metadata to publish
  --extra-options TEXT            Extra options to pass to datasette serve
  --branch TEXT                   Install datasette from a GitHub branch e.g. main
  --template-dir DIRECTORY        Path to directory containing custom templates
  --plugins-dir DIRECTORY         Path to directory containing custom plugins
  --static MOUNT:DIRECTORY        Serve static files from this directory at /MOUNT/...
  --install TEXT                  Additional packages (e.g. plugins) to install
  --plugin-secret <TEXT TEXT TEXT>...
                                  Secrets to pass to plugins, e.g. --plugin-secret
                                  datasette-auth-github client_id xxx
  --version-note TEXT             Additional note to show on /-/versions
  --secret TEXT                   Secret used for signing secure values, such as signed
                                  cookies
  --title TEXT                    Title for metadata
  --license TEXT                  License label for metadata
  --license_url TEXT              License URL for metadata
  --source TEXT                   Source label for metadata
  --source_url TEXT               Source URL for metadata
  --about TEXT                    About label for metadata
  --about_url TEXT                About URL for metadata
  --token TEXT                    Auth token to use for deploy
  --project PROJECT               Vercel project name to use  [required]
  --scope TEXT                    Optional Vercel scope (e.g. a team name)
  --no-prod                       Don't deploy directly to production
  --debug                         Enable Vercel CLI debug output
  --public                        Publish source with Vercel CLI --public
  --generate-dir DIRECTORY        Output generated application files and stop without
                                  deploying
  --generate-vercel-json          Output generated vercel.json file and stop without
                                  deploying
  --vercel-json FILENAME          Custom vercel.json file to use instead of generating
                                  one
  --setting SETTING...            Setting, see docs.datasette.io/en/stable/settings.html
  --crossdb                       Enable cross-database SQL queries
  --help                          Show this message and exit.

Using a custom vercel.json file

If you want to add additional redirects or similar to your Vercel configuration you may want to provide a custom vercel.json file.

To do this, first generate a configuration file (without running a deploy) using the --generate-vercel-json option:

datasette publish vercel my-database.db \
  --project=my-database \
  --generate-vercel-json > vercel.json

You can now edit the vercel.json file that this creates to add your custom options.

Then run the deploy using:

datasette publish vercel my-database.db \
  --project=my-database \
  --vercel-json=vercel.json

Setting a DATASETTE_SECRET

Datasette uses a secret string for purposes such as signing authentication cookies. This secret is reset when the server restarts, which will sign out any users who are authenticated using a signed cookie.

You can avoid this by generating a DATASETTE_SECRET secret string and setting that as a Vercel environment variable. If you do this the secret will stay consistent and your users will not be signed out.

Using this with GitHub Actions

This plugin can be used together with GitHub Actions to deploy Datasette instances automatically on new pushes to a repo, or on a schedule.

The GitHub Actions runners already have the Vercel deployment tool installed. You'll need to create an API token for your account at vercel.com/account/tokens, and store that as a secret in your GitHub repository called VERCEL_TOKEN.

Make sure your workflow has installed datasette and datasette-publish-vercel using pip, then add the following step to your GitHub Actions workflow:

    - name: Deploy Datasette using Vercel
      env:
        VERCEL_TOKEN: ${{ secrets.VERCEL_TOKEN }}
      run: |-
        datasette publish vercel mydb.db \
          --token $VERCEL_TOKEN \
          --project my-vercel-project

You can see a full example of a workflow that uses Vercel in this way in the simonw/til repository.