The Registry of Open Data (RODA) Browser acts as a simple, web-based visualization of the data in https://github.com/awslabs/open-data-registry.
This browser can be used to view a single data repository or multiple data repositories, depending on build-time settings. The project builds static HTML files that can be deployed on any compatible web server.
You can add logos to the src/img/logos directory for use in the detail and collab pages. Logos should be square PNGs and are currently sized down to 100px x 100px for display. The name of the logo file that the template looks for is tied to the entry in the ManagedBy data field. Spaces are replaced with '-' and special characters are removed. If the ManagedBy field is using markdown, the link text field is used, otherwise, the entire string is used to generate the logo path for testing.
/- Main datasets listing page, provides search mechanism.ex: /1000-genomes- Individual detail pages for each dataset, contains details, license, contact, documentation and example usage links and AWS resources available./usage-examples/- Lists all usage examples grouped by dataset./index.yaml- YAML formatted listing of each individual YAML file for provided datasets. [datasets.yamlis created but should be considered deprecated]/index.ndjson- NDJSON formatted listing of each individual YAML file for provided datasets. This same data is available at/roda/ndjson/index.ndjsonfor searching via tools like Amazon Athena.ex: /tag/earth-observation/- Tag-subsetted view of the main datasets listing page.ex: /tag/machine-learning/usage-examples/- Tag-subsetted list usage examples grouped by dataset.ex: /tag/astronomy/datasets.yaml- YAML for all datasets associated with a tag.ex: /datasets/awslabs-open-data-registry/datasets/1000-genomes.yaml- YAML for individual dataset, used to create the HTML pages./sitemap.txt- Sitemap listing all the HTML pages./providers.html- A simple listing of logos of data providers.
- Get this repository and the related data files with
git clone [email protected]:awslabs/open-data-registry-browser.git. npm run copy-datato copy data repositories locally, see note about using multiple repositories below.npm installto install required Node.js modules.npm run serveto develop the site with live reloading ORnpm run buildto build the site for deployment. See note about using multiple repositories below.
Tests can be run with npm test.
By default, the public data repository at https://github.com/awslabs/open-data-registry is used. If you wish to use a different or multiple repositories, you can add them via the RODA_SOURCES environment variable like
export [email protected]:awslabs/repo1.git,[email protected]:awslabs/repo2.git
Make sure this variable is set before running npm run copy-data and npm run build or npm run serve. The copy-data script runs a git clone <repo> command so the machine you're building on needs to have permissions to the repository and the URL provided must be of the form to work with git clone.
Subsequent listed repositores will take precendence over data on previous repositories. Datasets are matched across repos by their Slug, which is generated from their filename. Currently, only the Metadata dictionary is copied over between repositories. If a dataset exists only outside the primary repository, it will be created as normal.
By using the redirects map in src/config.yaml you can set up simple HTML redirects from source to target. This does not copy any content, simply creates a new HTML page. This function runs after other HTML content is generated so it can be used to create new or overwrite existing HTML pages.
Docker can be used to isolate the development environment.
Build the container with:
docker build -t roda .
Run the container with:
docker run --rm -p 3000:3000 -ti roda
Run the container with a local copy of the registry:
docker run --rm -p 3000:3000 -v pwd/open-data-registry:/app/data-sources/awslabs-open-data-registry -ti roda