instal the last version for windows Automatic PDF Processor 1.25

Currently this is run in a local environment The Python service and Python cli are formatted using black and flake8. Look at the function user_is_allowed in main.py for details. That is, an will grant access to all AI2 people.

There's a special case when an allowed email address in this file starts meaning all users in that domain are allowed. The format of the file is simply a list of allowed email addresses.

For local development, this file is sourced fromĪnd also projected to /users/allowed.txt in the Docker container.

Projected to /users/allowed.txt in the container.

In production, this file is sourced from the secret named.

Proxy/nf specifying that the user at the function get_user_from_header in main.py forĪuthorization is enforced by the PAWLS app. Supplemented with a hard-coded authentication header in

For local development, there is no login workflow.

Authenticated requests carry an HTTP header Users are bounced to a Google login workflow, and redirected back to the site

The production deployment of PAWLS uses Skiff.

Or not these users' requests are allowed (e.g., to view a PDFs) is considerdĪuthorization. Authentication and AuthorizationĪuthentication is simply checking that users are who they say they are.

Pawls assign skiff_files/apps/pawls/papers -all -name-file skiff_files/apps/pawls/papers/name_mapping.jsonĪnd then open up the UI locally by running docker-compose up. # Assign the development user to all the papers we've downloaded. Pawls preprocess grobid skiff_files/apps/pawls/papers # ensure that the papers are pre-processed with grobid so that they have token information.

# Fetches PDFs from semantic scholar's S3 buckets.

Instal the last version for windows Automatic PDF Processor 1.25