dakota
v1.0.0
Published
Monorepo for Dakota ecosystem
Downloads
1
Readme
Dakota
This is a monorepo for Dakota Development. It contains the following packages:
dakota-ui
- Next.js frontend replacement of MRC - runs on http://localhost:2022validation-ui
- Next.js frontend to support import validation - runs on http://localhost:3000dakota-service
- Nest.js REST API - runs on http://localhost:1985@bfc-software/dakota-schema
- (CommonJS) Open API schema definition for the Dakota API@bfc-software/infrastructure
- (ESM) Infrastructure tooling for deploys, application retrieval for the IBM, etc
Table of Contents
- Dakota
Workspace Level Commands
Start with "Environment Configuration" if this is your first time running Dakota.
Ensure the latest dependencies are installed:
npm install
To Work on the MRC replacement app
Then run the development servers:
npm run frontend
npm run backend
To work on the Validation UI
npm run validation-ui
Libraries
Local Development
The packages contained in the @bfc-software
directory represent npm packages that are available in our private registry. The monorepo approach allows making local changes that can be tested in the consuming entities prior to official publish.
To incorporate changes you've made to a package in the @bfc-software
directory:
Bump the version in the
package.json
for the package. Your version bump should represent the change type - see the semver website for further details on semantic versioning. In broad strokes the change types are- [major]: Represents a significant change to the package approach or architecture, may contain breaking changes.
- [minor]: Represents the addition of a new feature or deliverable with no breaking changes.
- [patch]: Represents a small change such as a bug fix or amendment with no breaking changes.
Build the package using
npm run build
to generate processed files for consumption.Bump the version in the
package.json
for any consuming packages.Run
npm install
at the root directory level to ensure the installation is picked up by all packages.
Debugging
Libraries make use of the debug
library to provide feedback during the development process. Logs are enabled by default in the application dev process but you can enable them in the browser using the following method:
localStorage.debug = 'pallet'
Current Namespaces:
- pallet
Publishing
Packages that live under the @bfc-software
directory get published on NPM using a tool called Lerna.
Private packages (which is all of our packages to date) will have an entry in their package.json
that points to our private repository ensuring they get published correctly:
// package.json
"publishConfig": {
"registry": "https://www.npmjs.com/package/@bfc-software"
},
Start by running the following to authenticate with npm:
npm login
Then run:
npm run publish
Changelog
This repo uses Lerna to generate an automated changelog of all library changes. Each PR is validated using the validate-release-labels
workflow to ensure a valid tag is added when a change is made to any of the files in the packages/@bfc-software directories.
If a change is present and a tag has not been added you will be asked to add one.
Tag types:
- breaking
- enhancement
- bug
- documentation
- internal
You can generate the changelog using:
npm run generate-changelog
Environment Configuration
Setting up ODBC
ODBC is the preferred method of connecting to Db2 on i.
You will need unixODBC
as your ODBC driver manager. unixODBC
is available on homebrew, and can be installed running the following command:
$ brew install unixodbc
Set up your DSN (datasource name)
ODBC uses a connection string with keywords to create a database connection. Because these strings can be cumbersome to maintain, a more ergonomic format called a DSN (datasource name) is available.
When using unixODBC, DSNs are defined in odbc.ini
(available to all users) and .odbc.ini
(user scoped). You can find the files using the following command:
$ odbcinst -j
Open ~/.odbc.ini
in your preferred editor and add the following configuration for accessing DEV5:
[DEV5]
Description = The IBM i System
Driver = IBM i Access ODBC Driver
System = 10.150.0.37
Deployment
When a pull request is merged to the main
branch, the GitHub deploy workflow is triggered and production assets get pushed to S3. Each deploy creates a new object where the name is the Git SHA.
IBM Setup
Installing the bfc
executable
Use the Code for IBM i VSCode plugin to access the PASE shell (or whatever means you usually use).
First Install
Add the following to ~/.npmrc
, replace AUTH_TOKEN
with a token generated on www.npmjs.com that can read from our private registry:
N.B. This will overwrite the existing ~/.npmrc file.
cat <<EOL > ~/.npmrc
//registry.npmjs.org/:_authToken=AUTH_TOKEN
@bfc-software:registry=https://registry.npmjs.org/
EOL
Add the package for the executable:
npm install -g @bfc-software/infrastructure
Ensure it is present by running bfc
without any args:
bfc
Update to latest package version
npm update -g @bfc-software/infrastructure
If the executable isn't immediately available run:
source ~/.profile
Retrieve a release
You can retrieve a release using:
bfc fetch-assets
This will prompt you for the AWS credentials and then present a series of Git SHAs. The SHA is logged during the build-and-deploy workflow in case there is any question.
Start the Frontend
cd
down to the dakota-ui
directory that was unpacked by bfc
.
Run npm run start:prod
.
It should start the app on http://localhost:2022.
Start the Backend
cd
down to the dakota-service
directory that was unpacked by bfc
.
It should start the app on http://localhost:1985.