Coding Horrors: You can gather all your technical debt into one easy framework!

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Cavern of COBOL > Coding Horrors: You can gather all your technical debt into one easy framework!

«‹›1503 »

Spatial: Nov 15, 2007

Java code:

public class Stack<E> {

	E[] elements;
	int maxSize;
	int top;


	
	public int push(E elem) {
		if (top == maxSize)
			throw new RuntimeException();

		elements[++top] = elem;

		return 1;
	}



	public int pop(E elem) {
		if (top == 0)
			throw new RuntimeException();

		elem = elements[top--];

		return 1;
	}
	
	

	public Stack( int size ) {
		top      = -1;
		maxSize  = size;
		elements = (E[]) new Object[ size ];
}

Written by my data structures lecturer.

# ? Mar 9, 2013 17:39

Adbot: ADBOT LOVES YOU

# ? May 13, 2024 00:12

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

What's wrong with it, other than having it be a bounded size?

EDIT: whoops, didn't read pop() enough carefully. It seems this is a C guy who was forced into Java.

# ? Mar 9, 2013 17:57

kitten smoothie: Dec 29, 2001

My first thought: "What's wrong with an array implementation? It's even in the CLR book."
My second thought: "Oh."

# ? Mar 9, 2013 17:58

HappyHippo: Nov 19, 2003; Do you have an Air Miles Card?

It seems like if you push one thing and then try to pop it you'll get an exception?

Edit: Also, returning 1. I'm guessing that was originally error handling but now it's just pointless

# ? Mar 9, 2013 17:58

carry on then: Jul 10, 2010; by VideoGames
(and can't post for 10 years!)

Let me guess, they also think Java is pass-by-reference for objects.

# ? Mar 9, 2013 18:02

Volmarias: Dec 31, 2002; EMAIL... THE INTERNET... SEARCH ENGINES...

If you instantiate, push, pop, you get an exception because he starts at -1 instead of 0.

You must push a sacrificial element to this implementation. I would recommend pushing it's own reference.

E;f,b...

# ? Mar 9, 2013 18:11

Sedro: Dec 31, 2008

HappyHippo posted:

It seems like if you push one thing and then try to pop it you'll get an exception?

Volmarias posted:

If you instantiate, push, pop, you get an exception because he starts at -1 instead of 0.

You must push a sacrificial element to this implementation. I would recommend pushing it's own reference.

E;f,b...

No, it works

Java code:

top = -1;
elements[++top] = elem; // top = 0, elements[0] = elem
elem = elements[top--]; // elem = elements[0], top = -1

Edit: oh, nevermind

Java code:

if (top == 0)
    throw new RuntimeException();

# ? Mar 9, 2013 18:37

Freakus: Oct 21, 2000

Also, memory leak: references to popped elements stay around until that spot is re-used.

# ? Mar 9, 2013 19:11

Zombywuf: Mar 29, 2008

Xerophyte posted:

_{Note: consider not trying to learn Agda.}

I've almost managed to write a modulus function on nats that passes the termination checker.

# ? Mar 9, 2013 19:24

Volmarias: Dec 31, 2002; EMAIL... THE INTERNET... SEARCH ENGINES...

Freakus posted:

Also, memory leak: references to popped elements stay around until that spot is re-used.

Aha, good catch. Obviously not a leak in C but he doesn't know Java at all.

# ? Mar 9, 2013 21:38

PrBacterio: Jul 19, 2000

Also nobody has commented on the fact that, why the hell do the push and pop functions return an int? And why do they always return the value 1?

_{_{vvv I knew that, it's a rhetorical question to bring attention to the fact that this, too, is a horror vvv}}

PrBacterio fucked around with this message at 04:35 on Mar 10, 2013

# ? Mar 10, 2013 04:24

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

Because it originally looked like this:

C++ code:

int stack_push (Stack *stack, void *elem);
int stack_pop (Stack *stack, void *elem_out);

# ? Mar 10, 2013 04:32

ninjeff: Jan 19, 2004

Spatial posted:

Java code:

(E[]) new Object[ size ]

Arrays ~~don't~~shouldn't work that way!

# ? Mar 10, 2013 05:55

Malloc Voidstar: May 7, 2007; Fuck the cowboys. Unf. Fuck em hard.

It's Java, that's how they do work. Yay type erasure. Though JDK code casts elements, not arrays, from what I remember.

# ? Mar 10, 2013 06:25

ninjeff: Jan 19, 2004

Aleksei Vasiliev posted:

It's Java, that's how they do work. Yay type erasure. Though JDK code casts elements, not arrays, from what I remember.

It's array covariance, not type erasure - C# and VB have the same problem despite having real generics :eng99:

# ? Mar 10, 2013 06:29

Flobbster: Feb 17, 2005; "Cadet Kirk, after the way you cheated on the Kobayashi Maru test I oughta punch you in tha face!"

ninjeff posted:

Arrays ~~don't~~shouldn't work that way!

Every semester when I teach my data structures course, I spend a good 10 minutes ranting about Java's horrible design decisions when it comes to generics, especially about how there's no legitimate reason to NOT allow new E[...] since E is always going to be an object type so the compiler should just allocate an Object array and implicitly do the cast for me.

And then the cast isn't even enough, I have to @SuppressWarnings to completely shut it up. :suicide:

# ? Mar 10, 2013 17:35

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

PrBacterio posted:

vvv I knew that, it's a rhetorical question to bring attention to the fact that this, too, is a horror vvv

I guess I don't understand what's wrong with it in C.

# ? Mar 10, 2013 17:39

PrBacterio: Jul 19, 2000

Suspicious Dish posted:

I guess I don't understand what's wrong with it in C.

There's nothing wrong with it in C and I never said it was, but that was written in Java and it served absolutely no purpose there? :confused:

# ? Mar 10, 2013 18:04

ExcessBLarg!: Sep 1, 2001

Suspicious Dish posted:

Because it originally looked like this:
C++ code:
int stack_push (Stack *stack, void *elem);
int stack_pop (Stack *stack, void *elem_out);

This is an amusingly perfect description of what was obviously the original C (or C++ I guess) version. Because it's exactly as useless as the Java one.

# ? Mar 10, 2013 18:19

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

Why?

# ? Mar 10, 2013 18:25

ExcessBLarg!: Sep 1, 2001

There's no useful way to get the element out of "pop" (or "stack_pop") since the assignment is local to that method only. It actually makes the data structure completely useless since it's "input only".

In other words:

carry on then posted:

Let me guess, they also think Java is pass-by-reference for objects.

# ? Mar 10, 2013 18:31

Zemyla: Aug 6, 2008; I'll take her off your hands. Pleasure doing business with you!

Shugojin posted:

The entirety of Numerical Recipes in C and how they didn't completely port it from Numerical Recipes in FORTRAN so every single loving bit of array math starts at 1 instead of 0 so you need to constantly shift so that it accesses what you want

Oh god yes. I used it as a pedagogical text in high school, and then when I got to college I found out about the GNU Scientific Library. Seriously, screw 1-based arrays in C.

FAKE EDIT: Also, I'm pretty sure ptr - 1 isn't even valid Standard C, since it can have undefined behavior.

# ? Mar 10, 2013 18:32

carry on then: Jul 10, 2010; by VideoGames
(and can't post for 10 years!)

ExcessBLarg! posted:

There's no useful way to get the element out of "pop" (or "stack_pop") since the assignment is local to that method only. It actually makes the data structure completely useless since it's "input only".

Yeah, but how is the C one just as useless?

# ? Mar 10, 2013 18:38

ExcessBLarg!: Sep 1, 2001

Zemyla posted:

FAKE EDIT: Also, I'm pretty sure ptr - 1 isn't even valid Standard C, since it can have undefined behavior.

(ptr-1) is valid for any complete (not void *) pointer type.

# ? Mar 10, 2013 18:40

ExcessBLarg!: Sep 1, 2001

carry on then posted:

Yeah, but how is the C one just as useless?

Although not obligated, presumably the implementation of the function is:

code:

...
elem_out = elem;
...

And just like in the Java version, the assignemnt to elem_out is meaningless since it doesn't influence any program state outside the "stack_pop" method.

# ? Mar 10, 2013 18:42

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

ExcessBLarg! posted:

There's no useful way to get the element out of "pop" (or "stack_pop") since the assignment is local to that method only. It actually makes the data structure completely useless since it's "input only".

Not in C. Perhaps I should have put void **elem_out as the signature instead, but it's the same.

# ? Mar 10, 2013 18:43

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

ExcessBLarg! posted:

Although not obligated, presumably the implementation of the function is:
code:
...
elem_out = elem;
...

That wouldn't make any sense. The implementation would be *elem_out = elem; which would assign to the out pointer.

# ? Mar 10, 2013 18:43

ExcessBLarg!: Sep 1, 2001

Suspicious Dish posted:

Not in C. Perhaps I should have put void **elem_out as the signature instead, but it's the same.

That's exactly the point though. To make "stack_pop" useful in C, you have to pass a double-pointer and assign to the deferenced pointer. Java doesn't have "double-pointers" so there's no way to assign a method reference as they're passed by value.

Java object references aren't like C++ references. Actually they're more like implicit C/C++ pointers. All Java objects are dynamically allocated from the heap (but automatically freed by the garbage collector). The Java "." operator is equivalent to the C(/C++) "->" operator for struct(/class) types. However, there's no "*" or "&" operators, no double-pointers, no pass-by-reference, etc.

There's only ~~two~~ three ways to get an object reference "out" of a method in Java:

Return it.
(Edit:) Throw it as an exception, assuming a throwable object.
Assign it to a field in a different object that's passed into the method, possibly an entirely superfluous single-field container object used to emulate C/C++ double pointers.

Edit:

Suspicious Dish posted:

That wouldn't make any sense.

Agreed, however that's entirely equivalent to what the posted Java code actually does. Hence my original comment.

ExcessBLarg! fucked around with this message at 18:58 on Mar 10, 2013

# ? Mar 10, 2013 18:50

HappyHippo: Nov 19, 2003; Do you have an Air Miles Card?

ninjeff posted:

It's array covariance, not type erasure - C# and VB have the same problem despite having real generics

Huh?

C# code:

public class Test<T>
{
    T[] SomeArray;
    public Test(int Size)
    {
        SomeArray = new T[Size];       
    }
}

Compiles fine.

# ? Mar 10, 2013 18:51

Volmarias: Dec 31, 2002; EMAIL... THE INTERNET... SEARCH ENGINES...

Yeah, I'm assuming that he took away things like * which caused compile errors in Java. So it was probably something like

code:

int stack_pop (Stack *stack, void **elem_out) 
{
  if(top == 0) {
    return 0;
  }
  *elem_out = elements[top--];
  return 1;
}

with the expected usage of

code:

void** element;
if(stack_pop(my_stack, element)) {
  // do something with element
} else {
  // oh no!
}

(My C is so rusty ogad I'm probably causing a segfault)

# ? Mar 10, 2013 18:52

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

ExcessBLarg! posted:

That's exactly the point though. To make "stack_pop" useful in C, you have to pass a double-pointer and assign to the deferenced pointer. Java doesn't have "double-pointers" so there's no way to assign a method reference as they're passed by value.

Java object references aren't like C++ references.

Yes, we know. What we're saying is that this guy used to teach C or C++, and then got stuffed into a Java course, and naively fiddled around enough to make his lesson plans compile, but not actually work.

# ? Mar 10, 2013 19:01

Amarkov: Jun 21, 2010

Suspicious Dish posted:

Yes, we know. What we're saying is that this guy used to teach C or C++, and then got stuffed into a Java course, and naively fiddled around enough to make his lesson plans compile, but not actually work.

If nothing else, this is the only explanation for using a RuntimeException to indicate an empty stack. "Hmm, I don't know what this 'throws' clause is, I'll just find a way to not have to use it."

# ? Mar 10, 2013 19:27

HappyHippo: Nov 19, 2003; Do you have an Air Miles Card?

Not to mention the half implemented "using ints to indicated success or failure" idiom.

# ? Mar 10, 2013 19:36

Shugojin: Sep 6, 2007; THE TAIL THAT BURNS TWICE AS BRIGHT...

Zemyla posted:

Oh god yes. I used it as a pedagogical text in high school, and then when I got to college I found out about the GNU Scientific Library. Seriously, screw 1-based arrays in C.

FAKE EDIT: Also, I'm pretty sure ptr - 1 isn't even valid Standard C, since it can have undefined behavior.

I was working on something that relied upon an unholy, uncommented union of NR's Levernberg-Marquardt fitting method and GSL libraries last week and bitching about NR in here was one of my rewards for success.

As a first encounter with GSL it was... Suboptimal.

All the GSL calls and such were fine but all the logic for calling it relied on NR's absolute horseshit.

# ? Mar 10, 2013 20:14

Qwertycoatl: Dec 31, 2008

HappyHippo posted:

Huh?

C# code:

public class Test<T>
{
    T[] SomeArray;
    public Test(int Size)
    {
        SomeArray = new T[Size];       
    }
}

Compiles fine.

The problem C# has with array covariance is this:

C# code:

class Base
{
}
class Derived : Base
{
}

// This function should never throw an exception, right?
// Wrong. SetBase(new Derived[1]) compiles, and the array access throws
// ArrayTypeMismatchException
static void SetBase(Base[] arr)
{
   if (arr.Length > 0)
      arr[0] = new Base();
}

# ? Mar 10, 2013 20:21

HappyHippo: Nov 19, 2003; Do you have an Air Miles Card?

Qwertycoatl posted:

The problem C# has with array covariance is this:

C# code:

class Base
{
}
class Derived : Base
{
}

// This function should never throw an exception, right?
// Wrong. SetBase(new Derived[1]) compiles, and the array access throws
// ArrayTypeMismatchException
static void SetBase(Base[] arr)
{
   if (arr.Length > 0)
      arr[0] = new Base();
}

Right. But does that affect this case (with the stack class the prof is trying to make)?

C# code:

class Vehicle
{
}
class Car : Vehicle
{
}
class Spaceship : Vehicle
{
}

public class Test<T>
{
    T[] SomeArray;
    public Test(int Size)
    {
        SomeArray = new T[Size];       
    }

    public void AddSomeItem(T item, int index)
    {
        SomeArray[index] = item;
    }
}

...

Test<Vehicle> test = new Test<Vehicle>(10);
test.AddSomeItem(new Car(), 0);
test.AddSomeItem(new Spaceship(), 5);

Works fine. Am I missing something?

# ? Mar 10, 2013 21:44

Qwertycoatl: Dec 31, 2008

HappyHippo posted:

Right. But does that affect this case (with the stack class the prof is trying to make)?
code:
...
Works fine. Am I missing something?

No, that's fine. C# isn't as crazy as Java, it's just a bit crazy.

# ? Mar 10, 2013 22:50

ninjeff: Jan 19, 2004

Aleksei Vasiliev posted:

It's Java, that's how they do work. Yay type erasure. Though JDK code casts elements, not arrays, from what I remember.

I just read the replies above and thought about it some more, and you were right - I just assumed, coming from the .NET world, that this would be legal and that the author of the awful stack class just hadn't known about it:

Java code:

new E[size]

So it's a good thing that Java has the awful feature of array covariance to allow you to work around its awful feature of type erasure!

# ? Mar 11, 2013 06:38

Bhaal: Jul 13, 2001; I ain't going down alone; Dr. Infant, MD

Don't forget the out of bounds indexing when you pop() an empty Stack, or the out of bounds indexing when you try to push size+1 elements in, and both cases the class cheerfully makes the attempt. Man, a data structures lecturer who completely bombs array index math 101, that's a new one.

# ? Mar 12, 2013 01:15

Adbot: ADBOT LOVES YOU

# ? May 13, 2024 00:12

New Yorp New Yorp: Jul 18, 2003; Only in Kenya.; Pillbug

A co-worker stomped over a change I made to a source controlled file today.

The horror is that we're both ALM consultants and this guy's spent hundreds if not thousands of hours teaching people how to use source control.

# ? Mar 12, 2013 01:51

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Cavern of COBOL > Coding Horrors: You can gather all your technical debt into one easy framework!

«‹›1503 »