Skip to content

Latest commit

 

History

History
288 lines (200 loc) · 14.5 KB

README.md

File metadata and controls

288 lines (200 loc) · 14.5 KB
Raw

tdr-transfer-frontend

Repository for TDR transfer code

Running locally

There are two ways to develop this project:

  • Frontend development only, using the AWS integration environment for everything else. This is the default
  • Full stack local development, using a local dev copy of the API, Keycloak, etc

Prerequisites

Regardless of how you set up the development environment, you will need:

  • IntelliJ with the Scala plugin (or equivalent Scala dev environment)
  • Docker

Frontend development only

Follow these instructions if you want to make changes to the frontend application without needing to set up a full development environment for the other TDR services.

  • Run Redis using Docker: 
    docker run -d --name redis -p 6379:6379 redis
  • If you don't already have an admin user account for the Integration Keycloak site, ask another member of the TDR development team to create one for you
  • Look up the integration environment variables in the AWS Integration account:
    • Keycloak client secret
      • In the AWS console:
        • Go the Systems Manager service
        • Go the Parameter Store in the left-hand menu
        • Find the /intg/keycloak/client/secret parameter
        • Copy the parameter's value
      • With the AWS CLI:
        • Run: 
          aws ssm get-parameter --name "/intg/keycloak/client/secret" --with-decryption
        • Copy the Value from the object returned
      • get arn for the slack notifications SNS topic for integration (find in AWS console)
  • In IntelliJ, create a new sbt run configuration:
    • Set the Tasks parameter to run
    • Configure the environment variables:
      • AUTH_URL=https://auth.tdr-integration.nationalarchives.gov.uk
      • AUTH_SECRET=<the secret for the Keycloak client that you copied above>
      • READ_AUTH_SECRET=<follow the steps above for obtaining AUTH_SECRET, but using parameter key /intg/keycloak/user_read_client/secret>
      • NOTIFICATION_SNS_TOPIC_ARN=<the arn for sns topic for slack notifications you obtained above>
      • AWS_PROFILE=
  • Follow the Static Assets steps below to build the CSS and JS
  • Run the project from IntelliJ
  • Visit http://localhost:9000

When you log into the site, you will need to log in as a user from the Integration environment.

Troubleshooting

java.io.IOException

If you encounter the following error when trying to run in Intellij:

java.io.IOException: Cannot run program "npm" (in directory "/home/{username}/Tdr/tdr-transfer-frontend/npm"): error=2, No such file or directory

This is because NVM installs npm in a non-standard location.

To resolve this, you need to create a sym link with the install version of npm. Run the following commands:

  • sudo ln -s /home/{username}/.nvm/versions/node/{version}/bin/npm /usr/bin/npm; and
  • sudo ln -s /home/{username}/.nvm/versions/node/{version}/bin/node /usr/bin/node

Incognito mode

If you are using the AWS integration environment, uploads to S3 go through upload.tdr-integration.nationalarchives.gov.uk If you try to upload using incognito/private browsing, the upload will fail. This is because the browser blocks third party cookies by default in incognito mode. To allow the cookies, you can add an exception for the url upload.tdr-integration.nationalarchives.gov.uk

Instructions for Chrome

Instructions for firefox

Firefox state partitioning

You may get this error when running locally in Firefox

Partitioned cookie or storage access was provided to "https://auth..." because it is loaded in the third party context and dynamic state partitioning is enabled.

Because the domains are different, Firefox blocks the cookie request. The request is the way Keycloak JS works so we can't stop it. You can disable the setting for these sites though.

Go to about:config in the Firefox URL bar. Select the privacy.restrict3rdpartystorage.skip_list option and change the setting to http://localhost:9000,https://auth.tdr-integration.nationalarchives.gov.uk

This should allow the Keycloak cookies on localhost.

Full stack local development

NOTE: These instructions are out of date. Ask for assistance from a colleague if require setting up full stack development

Follow these instructions if you want to make changes to the API, database and/or auth system at the same time as updating the frontend.

Local auth server

  • Log into Docker with credentials from ECR and start the auth server. This will need AWS CLI version 2 to work.
export MANAGEMENT_ACCOUNT=management_account_number
aws ecr get-login-password --region eu-west-2 --profile management | docker login --username AWS --password-stdin $MANAGEMENT_ACCOUNT.dkr.ecr.eu-west-2.amazonaws.com
docker run -d --name keycloak -p 8081:8080 -e KEYCLOAK_USER=admin -e KEYCLOAK_PASSWORD=admin -e KEYCLOAK_IMPORT=/tmp/tdr-realm.json -e CLIENT_SECRET=[some value] -e BACKEND_CHECKS_CLIENT_SECRET=[some value] -e REALM_ADMIN_CLIENT_SECRET=[some value] -e KEYCLOAK_CONFIGURATION_PROPERTIES=intg_properties.json -e USER_ADMIN_CLIENT_SECRET=[some value] -e DB_VENDOR=h2 $MANAGEMENT_ACCOUNT.dkr.ecr.eu-west-2.amazonaws.com/auth-server:intg
  • Go to http://localhost:8081/auth/admin and log in with username admin and password admin.
  • Create a transferring body user:
    • Click Users in the menu on the left
    • Click Add User
    • Set a Username (all the other fields are optional) and click Save
    • Click the Groups tab
    • In the "Available Groups" box select the Mock 1 Department sub-group, and click Join The transferring_body_user/Mock 1 Department group should now appear in the "Group Membership" box
    • In the "Available Groups" box select the relevant "user type" sub-type depending on the type of user required, and click Join The user_type/standard_user or user_type/judgment_user group should now appear in the "Group Membership" box depending on which sub-group was selected
    • Click the Credentials tab
    • Set a non-temporary password for the user
    • For full details about managing transferring body users and transferring body groups see: Tdr User Administrator Manual
  • Set AUTH_SECRET as an environment variable in IntelliJ and/or the command line (depending on how you plan to run the frontend project) with the secret as its value: 
    AUTH_SECRET=[CLIENT_SECRET value from the docker run command]

Local API

Clone and run the tdr-consignment-api project.

Local S3 emulator

Create a new empty directory that the S3 emulator will save files in.

If you are running Linux, change the owner of this directory to user 2000, to give the user in the S3 ninja Docker container permission to save files there. Do not do this on a Mac, because Docker handles file permissions differently on Linux and macOS, and this step will prevent uploads from working.

sudo chown 2000:2000 /your/new/upload/directory

Run an S3 ninja Docker container, specifying a local directory in which to save the files:

docker run -d -p 9444:9000 -v /your/new/upload/directory:/home/sirius/data --name=s3ninja scireum/s3-ninja:6.4

Visit http://localhost:9444/ui and check you can create a bucket and upload a test file through the S3 ninja UI. Check that the file appears in the folder that you mounted.

Local backend checks

Follow the instructions in tdr-local-aws to run the FakeBackendChecker application, making sure you set the environment variable for the monitored folder to the same path as the mount directory that you set in the docker run command when you started the S3 ninja container. This lets the fake backend checker detect and scan files as they are uploaded to the S3 emulator.

Frontend project

  • Start Redis locally.

    docker run -d --name redis -p 6379:6379 redis

  • Ensure you have set the AUTH_SECRET environment variable, as described above. Set it in the command line or in the IntelliJ run configuration

  • Run the frontend, specifying the local full stack configuration file:

    sbt -Dconfig.file=conf/application.local-full-stack.conf run

    or set the IntelliJ SBT run configuration to -Dconfig.file=conf/application.local-full-stack.conf run

  • Visit http://localhost:9000

Static assets

Note: The TDR static assets are used by the TDR Auth Server. When updating the static assets, including the Sass, ensure that any changes are also implemented in the tdr-auth-server repo: https://github.com/nationalarchives/tdr-auth-server

  • This includes any changes to the .stylelintrc.json

  • If npm is not installed install nvm in root directory.

  • Once nvm is installed: nvm install 14.9

  • cd into tdr-transfer-frontend/npm in terminal

  • run npm install then npm run build

If you're working on the Sass files and just want to regenerate the CSS without rebuilding all the JavaScript code, you can run npm run sass-watch to monitor and rebuild the CSS files, or run npm run build-css to run a single build.

To run the Sass linter (stylelint) before commits, run the following command: from the npm folder npx stylelint **/*.scss

Full details of stylelint are available here: https://stylelint.io/

Running unit tests

We have two types of unit tests: Spec tests for our Scala code, which are located within the test directory and Jest tests (for our Typescript code) which are located in the npm > test directory

Running Spec tests

IntelliJ

In order to run Spec tests via IntelliJ, you could either: right-click on a Spec file and select "Run {Test name}" or go into the Spec file and at the top of the test, click the green play button on the left. You can also run individual tests.

Via CLI

In order to run the tests via CLI, you just navigate to the tdr-transfer-frontend directory and:

  • to run all tests - run sbt test
  • to run all tests of one Spec file - run sbt "testOnly *{class name}"
    • e.g sbt "testOnly *SeriesDetailsControllerSpec"
  • to run a specific test of a Spec file - run sbt "testOnly *{class name} -- -t \"should {test name}\"
    • e.g. sbt "testOnly *SeriesDetailsControllerSpec -- -t \"should display errors when an invalid\""
  • to run a test or multiple tests of a Spec file that contain a word/words in their name - run sbt "testOnly *SeriesDetailsControllerSpec -- -z \"{word/words in test name}\""
    • e.g sbt "testOnly *SeriesDetailsControllerSpec -- -z \"display errors\""

See here for more information about arguments you can use

Running Jest tests

IntelliJ

In order to run Jest tests via IntelliJ, you could either: right-click on a test.ts file and select "Run {Test name}" to run all or go into the test file and click one of the green play buttons (for the test you want to run) on the left to run one.

Via CLI

In order to run the tests via CLI, you just navigate to the tdr-transfer-frontend/npm directory and:

  • to run all tests - run npm test
  • to run all tests of one test file - run npm test -- {file name and extension}"
    • e.g npm test-- clientfileextractmetadata.test.ts"
  • to run a specific test of a Spec file - run npm test -- {file name and extension} -t "{test name}"
    • e.g. npm test -- clientfileextractmetadata.test.ts -t "extract function returns list of client"
  • to run a test or multiple tests of a test file that contain a word/words in their name - run npm test -- {file name and extension} -t "{words in test names}"
    • e.g. npm test -- clientfileextractmetadata.test.ts -t "extract"
  • to run tests in multiple test files that contain a word/words in their name - run npm test -- -t "{word/words in test names}"
    • e.g. npm test -- -t "throw"

Bonus: Jest test Coverage - In order to see if how much of your code is covered by the Jest tests, run npm test -- --coverage --verbose

Pre-commit checks

We have pre-commit checks that will run Scalafmt and run the unit tests associated with the files you've edited, only on Scala/Typescript files, if you have any staged. In order to run this script, you need to install Husky; to do this, run

npx husky install

Generated GraphQL classes

There is a separate repository which contains the generated case classes needed to query the Consignment API. These classes will be needed by more than one project which is why they are in a separate project. If you need to add a new query:

  • Run git clone https://github.com/nationalarchives/tdr-generated-graphql.git
  • Add the new query to the src/main/graphql directory
  • Run sbt package publishLocal
  • Set the version for tdr-generated-graphql in this projects build.sbt to be the snapshot version.

Draft Metadata Feature: Running and Testing Locally

Uploading from localhost

When uploading a metadata csv from localhost ensure have a valid set of AWS credentials for the TDR integration environment, or other environment if connecting your localhost to a different TDR environment.

CSV UTF-8 on Linux Machines

The draft metadata feature uses the BOM to check that the uploaded metadata CSV is UTF-8 encoded.

For software such a Libre office the BOM is not added to CSV files and therefore any file uploaded from a Linux machine will fail this validation check.

See here for further details: Metadata CSV file UTF-8 Validation

To add the BOM to a CSV file run the following commands:

  printf '\xEF\xBB\xBF' > with_bom.txt
  cat {file name of csv}.csv >> with_bom.txt
  mv with_bom.txt {file name of csv}.csv

These commands need to be run each time the file is saved.

Notes

  • Each environment has its own secret for the auth server. These cannot be generated inside AWS in any way and so it's difficult to get them into the Terraform scripts. At the moment, these are stored in a parameter store variable called /${env}/auth/secret although this may change.