PURL

PURL service is a URL resolver that translates a reference to a digital object (in the form of a druid), into a full content representation of that object as available in public access environment

Requirements

1. Ruby (3.4 or greater)
2. bundler gem

Installation

Clone the repository

$ git clone [email protected]:sul-dlss/purl.git

Move into the app and install dependencies

$ cd purl
$ bundle install

Start the development server

$ bin/dev

Configuring

Configuration is handled through the RailsConfig settings.yml files.

Development

The defaults in config/settings.yml should work on a locally run installation.

You can pull data from the stage stacks mount by running the fixture loader:

bin/rails runner 'FixtureLoader.load("bn208cc3480", machine: "purl-fetcher-stage")'

Or you can do this manually:

1. Create a local "stacks" folder in the root of the PURL rails app on your laptop.
2. Create druid tree folders (e.g. stacks/xk/755/gc/8675, just like they would be on stacks) and put cocina and public XML files there. You can get examples from the spec/fixtures/stacks or production Stacks. Note that the stacks folder is already in .gitignore so any content copied there will not be added to git. An easy solution would be to run ln -s spec/fixtures/stacks ..
3. Create a config/settings.local.yml file (if you don't have one already) and add the following:

stacks:
  root: stacks

Lint ERB

We use Herb to lint ERB templates. You can run it manually with:

npx @herb-tools/linter

Debugging

When the Rails server is run using bin/dev, you can start up a debugger session in a separate terminal via:

rdbg -A

To drop into the debugger, add debugger statements where needed in the code and have fun navigating the stack in your debugger session terminal.

Testing

The test suite (with RuboCop style enforcement) will be run with the default rake task (also run on travis)

$ rake

The specs can be run without RuboCop style enforcement

$ rake spec

The RuboCop style enforcement can be run without running the tests

$ rake rubocop

Update test fixtures

kinit
bin/rails update_metadata

Then run tests to see if anything needs to be updated. Commit the changes to the fixtures.

Deploying

Deployment is handled automatically via Jenkins when a release is published to GitHub.

Data

Data is stored in a pair-tree structure starting at /stacks. In each pair-tree is a file called cocina.json which is the public representation of the current version of the object. There is also a meta.json file that holds information about where this object is released to. The meta.json is not versionable data. Additionally there is an XML file called public which has a representation of the object that was derived from cocina.json.

Search engine indexing

Only items with "PURL sitemap" release tag (in meta.json) are included in the sitemap and all other items have a "noindex" meta tag.

Structured metadata in the form of schema.org markup is generated to enhance discoverability of datasets and videos. More info about schema.org markup.