Extending Karel’s Vocabulary This PPT originated with Dr. Judy Hankins Modifications have been...

Extending Karel’s Vocabulary

This PPT originated with Dr. Judy HankinsModifications have been done by Dr. Untch

& Dr. Cripps

Why should we extend Karel’s vocabulary?

• Karel’s vocabulary is limited.

• For example, Karel does not understand a TurnRight(); command.

He can only turn right by executing three TurnLeft(); commands.

• As a second example, suppose we need to program Karel to travel over vast distances.

• Assume that the robot must move east one mile (a mile is eight blocks long), pick up a beeper, and then move another one mile north.

• Karel understands how to move a block, but not a mile.

• Conversion from miles to blocks is straightforward, but results in a very long and unreadable program. How many Move(); instructions would be needed to move Karel one mile?

Karel can “learn” new commands via our instructions

• Our programs can furnish him with a dictionary of useful instruction names and their definitions.

• Each definition must be built from simpler instructions that Karel already understands.

• Given this ability to extend Karel’s vocabulary, we can solve our problems using instructions natural to our way of thinking.

How do we instruct Karel to learn a new command?

• For example, suppose we want Karel to learn the new command TurnRight, then how do tell Karel to remember this information?

• We know that if we instruct Karel to perform three TurnLeft instructions, then we can accomplish a TurnRight.

Defining a new function

void identifier(){

statement(s) that define the function

}

1 2 3 4 5 6

1

2

3

4

5

6

Suppose we want Karel to pickup the beeper shown in the situation file. Then what are Karel’s instructions?

#include <karel.h>int main(){ TurnOn(); TurnLeft(); TurnLeft(); TurnLeft(); Move(); PickBeeper(); TurnOff();}

1 2 3 4 5 6

1

2

3

4

5

6

The instructions for Karel to pickup the beeper shown in the situation file.

#include <karel.h>int main(){ TurnOn(); TurnLeft(); TurnLeft(); TurnLeft(); Move(); PickBeeper(); TurnOff();}

1 2 3 4 5 6

1

2

3

4

5

6

Now, lets modify Karel’s instructions so as to learn a new command, TurnRight.

#include <karel.h>void TurnRight(){ TurnLeft(); TurnLeft(); TurnLeft();}int main(){ TurnOn(); TurnRight(); Move(); PickBeeper(); TurnOff();}

1 2 3 4 5 6

1

2

3

4

5

6

Now, lets modify Karel’s instructions so as to learn a new command, TurnRight.

Defining a new function/command

void identifier(){

statement(s) that define the function

}

• Using this function definition mechanism to create a kind of dictionary for Karel, new commands can be defined to extend Karel’s vocabulary by writing functions.

• For example, although Karel does not have a predefined TurnRight(); command, we can user-define this instruction by writing the function shown in the previous slide and on the next slide.

void TurnRight(){ TurnLeft(); TurnLeft(); TurnLeft();}

• Similarly, we can define MoveAMile(); to mean eight Move(); instructions.

void MoveAMile(){ Move(); Move(); Move(); Move(); Move(); Move(); Move(); Move();}

Benefits of extending Karel’s vocabulary

• We can use instructions “natural” to us.

• As in the move a mile problem, the size of our programs can be significantly reduced.

• The smaller program is much easier to read and understand.

#include <karel.h>

int main()

{

TurnOn();

Move();

PickBeeper();

Move();

TurnLeft();

Move();

Move();

PutBeeper();

Move();

TurnLeft();

TurnLeft();

TurnLeft();

Move();

TurnLeft();

TurnLeft();

TurnLeft();

Move();

TurnLeft();

TurnLeft();

TurnLeft();

TurnOff();

}

#include <karel.h>

int main()

{

TurnOn();

Move();

PickBeeper();

Move();

TurnLeft();

Move();

Move();

PutBeeper();

Move();

TurnLeft();

TurnLeft();

TurnLeft();

Move();

TurnLeft();

TurnLeft();

TurnLeft();

Move();

TurnLeft();

TurnLeft();

TurnLeft();

TurnOff();

}

TurnRight();

TurnRight();

TurnRight();

#include <karel.h>void TurnRight(){

TurnLeft();TurnLeft();TurnLeft();

}

int main()

{

TurnOn();

Move();

PickBeeper();

Move();

TurnLeft();

Move();

Move();

PutBeeper();

Move();

TurnRight();

Move();

TurnRight();

Move();

TurnRight();

TurnOff();

}

A Stair Cleaning Task

• Karel is supposed to climb stairs and pick up the beepers on each step. When he is done, he should be standing on the top step, facing East.

#include <karel.h>…. int main() { TurnOn(); TurnLeft(); Move(); TurnRight(); Move(); PickBeeper(); TurnLeft(); Move(); TurnRight();…

If we invent a new instruction called ClimbAStep();

then the program can be written as:

#include <karel.h> int main() { TurnOn(); ClimbAStep(); PickBeeper(); ClimbAStep(); PickBeeper(); ClimbAStep(); PickBeeper(); TurnOff(); }

Now we must write the function ClimbAStep();

void ClimbAStep(){ TurnLeft(); Move(); TurnRight(); Move();}

Notice that this function depends on TurnRight(); , an instruction that we wrote previously.

Recapping how to dofunction definitions

• The reserved word void starts a new function followed by the name of the function.

• The name of the function specifies what the procedure is intended to do.

• The statements that perform the task are enclosed in matching curly braces { } . These statements specify how the new instruction does what the name implies.

• The two must match exactly – otherwise one or both need to be changed.

• To demonstrate this, there are no restrictions prohibiting the following instructions definition:

void TurnRight() { TurnLeft(); TurnLeft(); }

• This is a perfectly legal definition, but it is wrong because it does not accomplish what it should.