An instance of LLM prompting for programming

Just lately I watched a captivating name on Zoom. Xu Hao, Thoughtworks’s
Head of Expertise in China, defined his explorations in utilizing ChatGPT to
assist construct Self Testing Code. He
went by means of the sort of interplay that works for him.

He begins with a immediate that units the context for the appliance and the way
he needs the code to be structured.

The present system is a web based whiteboard system. Tech stack:
typescript, react, redux, konvajs and react-konva. And vitest, react
testing library for mannequin, view mannequin and associated hooks, cypress element
assessments for view.

All codes must be written within the tech stack talked about above.
Necessities must be carried out as react parts within the MVVM
structure sample.

There are 2 varieties
of view mannequin within the system.

1. Shared view mannequin. View mannequin that represents states shared amongst
native and distant customers.

2. Native view mannequin. View mannequin that represents states solely relevant
to native person

Listed here are the frequent implementation technique:

1. Shared view mannequin is carried out as Redux retailer slice. Examined in
vitest.

2. Native view mannequin is carried out as React element props or states(by
useState hook), until for international native view mannequin, which can also be
carried out as Redux retailer slice. Examined in vitest.

3. Hooks are used as the main view helpers to retrieve knowledge from shared
view mannequin. For many the case, it would use ‘createSelector’ and
‘useSelector’ for memorization. Examined in vitest and react testing
library.

4. Don’t dispatch motion instantly to vary the states of shared view
mannequin, use an encapsulated view mannequin interface as an alternative. Within the interface,
every redux motion is mapped to a technique. Examined in vitest.

5. View is include konva shapes, and carried out as react element by way of
react-konva. Examined in cypress element assessments 

Listed here are sure patterns must be adopted when implement and take a look at
the element

1. When write take a look at, use `describe` as an alternative of `take a look at`

2. Information-driven assessments are most popular.

3. When take a look at the view element, faux view mannequin by way of the view mannequin
interface

Consciousness Layer 

Requirement: 

Show different customers’ consciousness data(cursor, title and on-line
info) on the whiteboard. 

AC1: Don’t show native person

AC2: When distant person modifications cursor location, show the change in
animation. 

Present an total answer following the steering talked about above.
Trace, maintain all consciousness info in a Konva layer, and an consciousness
data element to render cursor, and title. Don’t generate code. Describe
the answer, and breaking the answer down as a activity checklist primarily based on the
steering talked about above. And we'll refer this activity checklist as our grasp
plan.

There’s so much happening with this immediate, so he highlighted just a few
factors.

He is utilizing a generic utility instance in right here: one factor to be
cautious of when interacting with ChatGPT and the like is that we must always by no means
put something which may be confidential into the immediate, as that may be a
safety threat. Enterprise guidelines, any code from an actual undertaking – all these should
not enter the interplay with ChatGPT.

Many of the immediate is setting out the design pointers that he needs
ChatGPT’s generated code to observe. He refined this by placing prompts into
ChatGPT, trying on the consequence, and crafting the immediate to make it generate
within the type he needed. As soon as he has context working, he can paste it into
each session he has with ChatGPT, and share it with the remainder of the
staff.

Xu Hao begins his session by prompting the
ChatGPT LLM with context in addition to the instruction to generate a plan

Regardless that the ultimate aim of this session is to generate code and its
assessments, at this stage he particularly mentioned Don’t generate code.
At this level he needs ChatGPT to stipulate its strategy earlier than we glance to any
code era. This can be a immediate method known as Generated Data.

A standard means of interacting with an LLM is to present it an issue and ask
it to generate some last product, whether or not than be prose or (on this case)
code. With Generated Data we cut up our interplay into separate steps:
first ask the LLM to generate some helpful details about the issue, and
then we feed that info again into the LLM to generate the ultimate
product.

Essentially the most fascinating a part of the immediate is the “implementation technique”
part which mixes two types of immediate:

• Instruction prompting offers directions for the LLM to observe
• Chain of Thought prompting encourages the LLM to elucidate its
  reasoning.

Within the Implementation Technique part, Xu Hao described the specified
structure sample as an anticipated “Chain of Thought directions” for
ChatGPT to observe. Then he instructed ChatGPT to construct a activity checklist (the
generated information) primarily based on this chain of thought.

The response from ChatGPT is non-deterministic, so this result’s simply
one run. Should you (or I) put this immediate in once more, we’ll get a distinct response.

The important thing component of the response is an ordered activity checklist, with numbers for
every step. We will use these numbers to refer to those duties for the remainder of
the session.

The LLM responds with a plan.

Xu Hao can now have a look at the grasp plan and see if it is smart. If
there are issues with it, he can ask ChatGPT to re-write a number of the duties
or your entire grasp plan by offering extra info.

For instance, the present grasp plan was not grouped by differing types
of parts. One potential immediate is perhaps:

Rewrite the grasp plan, and group duties as mannequin, view mannequin and look at.

This might be helpful to confirm the alignment between the structure
imaginative and prescient and grasp plan.

These tweaks might be added to the preliminary immediate to get an total
higher response.For this explicit immediate, the next instruction was
not within the preliminary immediate, however had been added in later iteration:

4. Don’t dispatch motion instantly to vary the states of shared view
    mannequin, use an encapsulated view mannequin interface as an alternative. Within the interface,
    every redux motion is mapped to a technique. Examined in vitest.

As soon as the grasp plan seems to be cheap, Xu Hao asks to regenerate it with
names for this system components.

Rewrite the grasp plan, this time with detailed element names,
strategies, and props

At this level, Xu Hao is able to ask ChatGPT to offer code for considered one of
the steps

present me instance assessments and implementation for the notice slice
and AwarenessViewModel. Check first.

Positive, let's begin with writing assessments for the notice slice and AwarenessViewModel:

Check for Consciousness Slice:

describe('consciousness slice', () => {
  let retailer: MockStore;
  beforeEach(() => {
    retailer = configureMockStore()(initialState);
  });

  it('ought to set the cursor location', () => {
    retailer.dispatch(setCursorLocation({ x: 100, y: 200 }));
    const state = retailer.getState().consciousness;
    count on(state.cursorLocation).toEqual({ x: 100, y: 200 });
  });

  it('ought to set the title', () => {
    retailer.dispatch(setName('John Doe'));
    const state = retailer.getState().consciousness;
    count on(state.title).toBe('John Doe');
  });

  it('ought to set the net standing', () => {
    retailer.dispatch(setOnline(true));
    const state = retailer.getState().consciousness;
    count on(state.isOnline).toBe(true);
  });
});
Check for AwarenessViewModel:

describe('AwarenessViewModel', () => {
  let retailer: MockStore;
  let awarenessViewModel: AwarenessViewModel;
  beforeEach(() => {
    retailer = configureMockStore()(initialState);
    awarenessViewModel = new AwarenessViewModel(retailer);
  });

  it('ought to set the cursor location', () => {
    awarenessViewModel.setCursorLocation({ x: 100, y: 200 });
    const state = retailer.getState().consciousness;
    count on(state.cursorLocation).toEqual({ x: 100, y: 200 });
  });

  it('ought to set the title', () => {
    awarenessViewModel.setName('John Doe');
    const state = retailer.getState().consciousness;
    count on(state.title).toBe('John Doe');
  });

  it('ought to set the net standing', () => {
    awarenessViewModel.setOnline(true);
    const state = retailer.getState().consciousness;
    count on(state.isOnline).toBe(true);
  });
});
Now, let's implement the Consciousness slice and AwarenessViewModel:

Consciousness slice:

interface AwarenessState {
  cursorLocation: { x: quantity; y: quantity };
  title: string;
  isOnline: boolean;
}

const initialState: AwarenessState = {
  cursorLocation: { x: 0, y: 0 },
  title: '',
  isOnline: false,
};

const awarenessSlice = createSlice({
  title: 'consciousness',
  initialState,
  reducers: {
    setCursorLocation: (state, motion: PayloadAction) => {
      state.cursorLocation = motion.payload;
    },
    setName: (state, motion: PayloadAction) => {
      state.title = motion.payload;
    },
    setOnline: (state, motion: PayloadAction) => {
      state.isOnline = motion.payload;
    },
  },
});

export const { setCursorLocation, setName, setOnline } = awarenessSlice.actions;

export default awarenessSlice.reducer;
AwarenessViewModel:

class AwarenessViewModel {
  personal retailer: Retailer;

  constructor(retailer: Retailer) {
    this.retailer = retailer;
  }

  setCursorLocation(cursorLocation: { x: quantity; y: quantity }) {
    this.retailer.dispatch(setCursorLocation(cursorLocation));
  }

  setName(title: string) {
    this.retailer.dispatch(setName(title));
  }

  setOnline(isOnline: boolean) {
    this.retailer.dispatch(setOnline(isOnline));
  }
}

Notice: On this implementation, I assumed that the AwarenessViewModel is used to
set the values within the retailer. If additionally it is used to retrieve the values from the
retailer, we will add corresponding strategies to the AwarenessViewModel class

(I’ve eliminated the styling from ChatGPT’s response.)

General this code is neither unhealthy nor good. It will not run because it stands.
It is lacking a configureMockStore and assumes it makes use of
redux-mock-store, which is not a part of the tech stack. Xu Hao can repair these
issues by asking the LLM to progressively tweak the code by means of a
sequence of rewrites. Here is one to get rig of redux-mock-store.

rewrite the assessments with out redux-mock-store

import { createAwarenessSlice, AwarenessState, updateAwareness } from './awarenessSlice';
import { AwarenessViewModel } from './AwarenessViewModel';

describe('AwarenessViewModel', () => {
  const mockDispatch = jest.fn();
  const initialState: AwarenessState = {
    customers: {
      'user1': {
        title: 'Consumer 1',
        cursorPosition: { x: 0, y: 0 },
        on-line: true,
      },
      'user2': {
        title: 'Consumer 2',
        cursorPosition: { x: 100, y: 100 },
        on-line: true,
      },
    },
  };
  let viewModel: AwarenessViewModel;

  beforeEach(() => {
    viewModel = new AwarenessViewModel(initialState, mockDispatch);
  });

  describe('getUserNames', () => {
    it('returns the checklist of person names', () => {
      const names = viewModel.getUserNames();
      count on(names).toEqual(['User 1', 'User 2']);
    });
  });

  describe('getCursorPosition', () => {
    it('returns the cursor place for a given person ID', () => {
      const cursorPosition = viewModel.getCursorPosition('user1');
      count on(cursorPosition).toEqual({ x: 0, y: 0 });
    });

    it('returns undefined for an unknown person ID', () => {
      const cursorPosition = viewModel.getCursorPosition('unknown');
      count on(cursorPosition).toBeUndefined();
    });
  });

  describe('updateCursorPosition', () => {
    it('dispatches the updateAwareness motion with the brand new cursor place', () => {
      viewModel.updateCursorPosition('user1', { x: 50, y: 50 });
      count on(mockDispatch).toHaveBeenCalledWith(updateAwareness({
        person: 'user1',
        cursorPosition: { x: 50, y: 50 },
      }));
    });
  });
});

import { createSlice, PayloadAction } from '@reduxjs/toolkit';

export kind CursorPosition = { x: quantity, y: quantity };

export kind UserAwareness = {
  title: string;
  cursorPosition: CursorPosition;
  on-line: boolean;
};

export kind AwarenessState = {
  customers: Report<string, UserAwareness>;
};

const initialState: AwarenessState = {
  customers: {},
};

const awarenessSlice = createSlice({
  title: 'consciousness',
  initialState,
  reducers: {
    updateAwareness: (state, motion: PayloadAction<{ person: string, cursorPosition: CursorPosition }>) => {
      const { person, cursorPosition } = motion.payload;
      state.customers[user].cursorPosition = cursorPosition;
    },
    setOnline: (state, motion: PayloadAction<{ person: string, on-line: boolean }>) => {
      const { person, on-line } = motion.payload;
      state.customers[user].on-line = on-line;
    },
  },
});

export const { updateAwareness, setOnline } = awarenessSlice.actions;
export const awarenessReducer = awarenessSlice.reducer;

Additional rewrites can match in additional changes, equivalent to asking it to
Use `configureTestingStore(‘consciousness’, initialState)` to initialize
the take a look at retailer can get it to slot in with the staff’s take a look at idioms. Then
with some path changes and a small quantity of glue for the reducers, the
code can run simply high quality.

As soon as that is working, Xu Hao can repeat the method for the remainder of the
duties within the grasp plan.