TODOs:

• CRUD, API endpoints and React Components

  • birdCalls
    • get all
    • search by id
    • get specific (level, class)
    • Get mnemonics, work on formatting and how best to display
    • Open database of bird images (possible hint)
    • mnemonic, scientific name, other metadata and info fields
    • birdCall page (locked/unlocked level etc, link to lesson, similar, search by metadata)
    • popout modal for each birdcall
    • synonyms, different localities and variants
    • don't double up level ups/downs in same session
  • users
    • get all
    • search by id
    • get specific (username, email)
    • update (e.g. level)
    • delete
    • level up
    • levelData field to append level started dates
  • userLevel
    • get all
    • search by id
    • get specific (level number)
  • item
    • get all
    • search by id
    • get specific (userId, birdCallId, level)
    • create (all items for new level for user)
    • level up
    • update (level, reviewdate etc)
      • do if lesson/review complete
    • delete
      • remake lesson if deleted/reset
    • count how many due for review now, in next hour, next 24 hours.
  • itemLevel
    • get all
    • get by id
    • get specific (num, name)
  • lesson
    • in UI, [display audio, spectrogram, metadata]
    • make for all current level items when levelup
    • move item from level 0 to 1 when done and delete lesson
    • create, get all, get by id, delete
    • make lesson page
    • lesson unlock
    • recreate if item reset
    • don't create if pair already exists
  • reviewSession (combines all reviews ready for review)
    • UI [display audio, spectrogram, metadata]
    • text box for answer
    • disable next review button until answered
    • prev review, disable if first etc
    • multiple choice (hint)
    • create new review whether correct or incorrect with associated times etc.
    • only do at end of session (pick small block size e.g. 10)
    • spectrogram, audio, scientific name could be could be different testing grounds, like reading + meaning in WK
    • stats for session (temporary or store)
  • ungradedSession (select some items to review, no change to item level/activity)
    • pass items to reviewSession page
  • classroom (connects teacher and students)
    • Lesson/course creator (teacher, admin) probably just permissions based
    • Assign students to teacher
    • Add classroom ID to user in create and addLearnerToClass
    • Teacher view (correct/incorrect items)
    • group stats
    • class tests (create, push/assign, store results)
  • general components
    • Navigation tabs/buttons
    • Connect to AWS to load sounds
    • make sure user can't go over max level, same with item. need to have way to make date blank
    • Modularize common scripts
    • Modern styling
    • carry User ID through session
    • login/logout functionality
    • (loading) notifications
    • logo
  • error handling
    • users with no lessons, items with empty activity, etc
    • don't levelup if lesson exists for that item
  • Activity/history field of Item
    • name: [started/lesson-complete, level-up/correct level-down/incorrect, complete, reset], date.
    • calculate stats
    • plot levelups/downs, lessons completed
  • Helper document with example API calls
  • faker for creating users, lessons, progress etc
    • need to make a random array of activity with levelups, downs, resets and iteravely call API, check actual item level against expected level
  • tokenized (maybe JWT) api calls for security
  • table styling e.g. colour item by level

• Stats:

  • upcoming reviews by hour next 24 hours
  • show activity (last 24 hours, average per hour in lifetime)
    • review count past 24 hours
  • overall num attempts, num correct/incorrect, % correct
  • when different review types, accuracy by type
  • critical items (pick those with accuracy below a certain threshold)
  • Level progression timing
  • projections for finishing level
  • error handling e.g. with empty activity
  • start date, time spent on app
  • streak/consistency (current, longest)
  • retention metric(s)
  • last year activity (like GitHub contribution map)
  • num items at max level
  • coverage of irish birds, bto pipeline, xenocanto etc (charts by level, items ordering)
  • current level, level progress (plot and proportion)
    • how many items on current level are >= item level 3

• Testing

  • https://enzymejs.github.io/enzyme/
  • Backend (Node.js/Express API)
    • Unit Tests:
      • Use Jest for testing individual functions and logic.
      • Create mocks and stubs for external services and databases to isolate unit tests.
    • Route/Integration Tests:
      • Use Supertest with Jest to test your Express routes.
      • Write tests to make HTTP requests to your API endpoints and assert the responses.
    • Database Seed Scripts:
      • Create scripts to seed your database with dummy data before tests.
      • Write cleanup scripts to remove test data after testing.
    • Mocking and Stubs:
      • Utilize Jest to mock external modules and services.
      • Create stubs for database interactions to avoid hitting the actual database in tests.
    • Code Coverage:
      • Use Istanbul (nyc) with Jest to generate code coverage reports.
      • Aim for high code coverage while ensuring tests are meaningful.
  • Frontend Testing (React Components)
    • Unit and Component Tests:
      • Use Jest and React Testing Library to test individual components.
      • Test component rendering and user interactions.
    • Mocking Hooks and Contexts:
      • Mock custom hooks and context providers if necessary to isolate components for testing.
    • Snapshot Testing:
      • Optionally use Jest's snapshot testing to ensure UI does not change unexpectedly.
  • End-to-End Testing
    • Cypress or Puppeteer:
      • Set up Cypress or Puppeteer for end-to-end testing.
      • Write tests that simulate real user interactions from the frontend through to the backend.
  • Continuous Integration and Continuous Deployment (CI/CD)
    • Setup CI/CD Pipeline:
      • Use tools like GitHub Actions, GitLab CI/CD, or CircleCI.
      • Configure the pipeline to run tests on every push or pull request.
      • Automate deployment on successful builds/tests to staging or production environments.

• AI Components:

  • RAG (Retrieval Augmented Generation) application - building LLM skills
  • noise addition/reduction
  • class prediction as hint/second opinion
  • mnemonic image generation
  • clustering to identify audibly/visually similar
  • bird syllable recognition