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 »

Hammerite: Mar 9, 2007; And you don't remember what I said here, either, but it was pompous and stupid.; Jade Ear Joe

Ithaqua posted:

There's nothing confusing about it, it's just less readable, and the second you start stacking them it turns into a mess. I submit for your consideration: http://stackoverflow.com/questions/9101719/im-looking-for-a-free-tool-stand-alone-or-add-in-that-can-decompose-ternary-ex/

I think other people here are defending the practice of nesting a further conditional expression as the third operand (if-false-value) of a conditional expression, not embedding further conditional expressions as operands of a conditional expression in full generality. Although it would be possible to write reasonable code doing so, if you don't take it to an insane extreme, and you guide the reader using visual cues like indentation.

code:

var x = some_test() ?
            some_other_test() ?
                value_true_true :
                value_true_false :
            value_false;

shrughes posted:

It's because some people are afraid of functional programming.

I don't understand what this has to do with functional programming.

# ? Sep 24, 2012 20:19

Adbot: ADBOT LOVES YOU

# ? May 27, 2024 07:34

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

Hammerite posted:

code:

var x = some_test() ?
            some_other_test() ?
                value_true_true :
                value_true_false :
            value_false;

See, I hate that. That takes me way longer to parse than if it were just written out with ifs and elses. I don't think it's a horror to do it that way, I just don't like it personally, and I'd probably change it if I ran across it in code I was working on.

# ? Sep 24, 2012 20:27

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

Hammerite posted:

code:

var x = some_test() ?
            some_other_test() ?
                value_true_true :
                value_true_false :
            value_false;

That's different. I hate the nested case. The original case is not nested at all.

# ? Sep 24, 2012 20:43

Hammerite: Mar 9, 2007; And you don't remember what I said here, either, but it was pompous and stupid.; Jade Ear Joe

Suspicious Dish posted:

That's different. I hate the nested case. The original case is not nested at all.

I don't understand what you are talking about when you refer to "the nested case" and "the original case". What do you hate?

# ? Sep 24, 2012 20:52

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

Hammerite posted:

I don't understand what you are talking about when you refer to "the nested case" and "the original case". What do you hate?

See, we're having trouble even telling the different cases apart! :v:

# ? Sep 24, 2012 21:01

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

Hammerite posted:

I don't understand what you are talking about when you refer to "the nested case" and "the original case". What do you hate?

I originally had a bigger example, but removed it because I thought you guys would just be annoyed by the noise.

JavaScript code:

function ternary() {
    return test_one ? value_one :
           test_two ? value_two :
           test_three ? value_three :
           value_default
}

function ifstmt() {
    if (test_one)
        return value_one
    else if (test_two)
        return value_two
    else if (test_three)
        return value_three
    else
        return value_default
}

JavaScript code:

function ternary() {
    return test_one ?
           test_two ?
               value_one_true_two_true :
               value_one_true_two_false :
           value_one_false

}

function ifstmt() {
    if (test_one)
        if (test_two)
            return value_one_true_two_true
        else
            return value_one_true_two_false
    else
        return value_one_false
}

Can you identify which one is the nested case, and which one is the original case?

# ? Sep 24, 2012 21:12

darthbob88: Oct 13, 2011; YOSPOS

Hammerite posted:

I don't understand what you are talking about when you refer to "the nested case" and "the original case". What do you hate?

Edit: Removed for redundant, Suspicious Dish did it better.

darthbob88 fucked around with this message at 21:16 on Sep 24, 2012

# ? Sep 24, 2012 21:13

raminasi: Jan 25, 2005; a last drink with no ice

I did this once :v:

C++ code:

bool surface_pair::contributes_to_envelope() const {
    return
        !is_self() &&
        !are_perpendicular() &&
        (!are_parallel() || // if they're parallel...
            (opposite_senses() && other_is_above_base() && areas_intersect())) &&
        (are_parallel() || // if they're not parallel...
            (other_in_correct_halfspace() && drape_hits_other_plane()));
}

I won't defend it.

# ? Sep 24, 2012 21:16

McGlockenshire: Dec 16, 2005; GOLLOCKS!

Good, because you don't need to. It's perfectly readable and understandable, once you sit down and read it, and would be even easier to understand inside an IDE with paren matching. The descriptive function names are amazingly clear.

# ? Sep 24, 2012 21:22

Hammerite: Mar 9, 2007; And you don't remember what I said here, either, but it was pompous and stupid.; Jade Ear Joe

Suspicious Dish posted:

Can you identify which one is the nested case, and which one is the original case?

I am not sure what relevance your example has to mine, since the indentation pattern in your example is dissimilar.

Your example:

JavaScript code:

function ternary() {
    return test_one ?
           test_two ?
               value_one_true_two_true :
               value_one_true_two_false :
           value_one_false
}

My example:

JavaScript code:

var x = some_test() ?
            some_other_test() ?
                value_true_true :
                value_true_false :
            value_false;

In my example, the second and third operands of each conditional expression are each indented one step (four spaces) further than the first operand of the same expression. In your example, the indentation pattern is less consistent.

# ? Sep 24, 2012 21:31

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

Because they're the exact same code, indentation different or not. (I meant to copy your indentation style; I copy/pasted the code and changed around a few variable names without thinking about indentation). Can you still not identify which one is nested?

# ? Sep 24, 2012 21:43

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

McGlockenshire posted:

Good, because you don't need to. It's perfectly readable and understandable, once you sit down and read it, and would be even easier to understand inside an IDE with paren matching. The descriptive function names are amazingly clear.

C code:

(!are_parallel() || // if they're parallel...

umm....

# ? Sep 24, 2012 21:43

Hammerite: Mar 9, 2007; And you don't remember what I said here, either, but it was pompous and stupid.; Jade Ear Joe

GrumpyDoctor posted:

I did this once :v:

C++ code:

bool surface_pair::contributes_to_envelope() const {
    return
        !is_self() &&
        !are_perpendicular() &&
        (!are_parallel() || // if they're parallel...
            (opposite_senses() && other_is_above_base() && areas_intersect())) &&
        (are_parallel() || // if they're not parallel...
            (other_in_correct_halfspace() && drape_hits_other_plane()));
}

I won't defend it.

You are an amateur in the business of writing bad code, my friend, and I am a professional.

PHP code:

if ( ( $i == $GAME['tileindustries'][$j] or
       ( $GAME['tileindustries'][$j] == 5 and
         ( !$i or
           $i == 1
           )
         ) or
       ( $GAME['tileindustries'][$j] == 6 and
         ( !$i or
           $i == 3
           )
         ) or
       ( $GAME['tileindustries'][$j] == 7 and
         ( $i == 2 or
           $i == 3
           )
         )
       ) and
     ( ( $card <= $GAME['TopLocationCard'] and
         $carddetail == $GAME['spacetowns'][$j]
         ) or
       ( $card > $GAME['TopLocationCard'] and
         ( $GAME['HasLinkedToTown'][$GAME['PlayerToMove']][$GAME['spacetowns'][$j]] or
           $GAME['HasBuiltInTown'][$GAME['PlayerToMove']][$GAME['spacetowns'][$j]] or
           $canbuildhereusingVCs
           )
         ) or
       $card == 99 or
       !$GAME['HasBuilt'][$GAME['PlayerToMove']]
       ) and
     ( $GAME['SpaceStatus'][$j] == 9 or
       ( $GAME['SpaceTile'][$j] == $i and
         $GAME['TechLevels'][$j] < $TheTechLevel and
         ( $GAME['SpaceStatus'][$j] == 8 or
           $GAME['SpaceStatus'][$j] == $GAME['PlayerToMove'] or
           ( $i == 1 and
             !$GAME['CoalInLancs'] and
             ( ( $GAME['CurrentPlayers'] == 2 and
                 $GAME['CoalDemand'] == 6
                 ) or
               ( $GAME['CurrentPlayers'] > 2 and
                 $GAME['CoalDemand'] == 8
                 )
               )
             ) or
           ( $i == 2 and
             !$GAME['IronInLancs'] and
             ( ( $GAME['CurrentPlayers'] == 2 and
                 $GAME['IronDemand'] == 6
                 ) or
               ( $GAME['CurrentPlayers'] > 2 and
                 $GAME['IronDemand'] == 8
                 )
               )
             )
           )
         )
       ) and
     ( $GAME['SpaceAlwaysExists'][$j] or
       ( $GAME['ModularBoardParts'] & $GAME['SpaceExistenceArray'][$j] )
       )
     ) {
    ...
}

Suspicious Dish posted:

Because they're the exact same code, indentation different or not. (I meant to copy your indentation style; I copy/pasted the code and changed around a few variable names without thinking about indentation). Can you still not identify which one is nested?

I thought that your argument was that it was hard to tell them apart. I think I'm still not understanding your point. For me, it is ok to nest conditional expressions up to a very limited level of complexity, but the reader should be helped by formatting the expression to make its structure easy to perceive.

# ? Sep 24, 2012 21:53

seiken: Feb 7, 2005; hah ha ha

Suspicious Dish posted:

C code:

(!are_parallel() || // if they're parallel...

umm....

The comment is perfectly correct but yes, this weird idiosyncrasy is probably the one reason it might qualify as a horror (still pretty mild, though).

# ? Sep 24, 2012 22:53

nielsm: Jun 1, 2009

Hammerite posted:

You are an amateur in the business of writing bad code, my friend, and I am a professional.

Hammerite posted:

I thought that your argument was that it was hard to tell them apart. I think I'm still not understanding your point. For me, it is ok to nest conditional expressions up to a very limited level of complexity, but the reader should be helped by formatting the expression to make its structure easy to perceive.

A nested ternary expression as the second operand to a ternary removes the immediate consequence of a false condition much further from the condition, while when you only nest on the third operand the immediate consequences of both possibilities are easily visible.

I would certainly prefer reading this (change indentation to your preference):

C++ code:

return !x ? a : y ? b : c;

to this (again, change indentation to your preference):

C++ code:

return x ? y ? b : c : a;

# ? Sep 24, 2012 22:57

That Turkey Story: Mar 30, 2003

Hammerite posted:

I don't understand what this has to do with functional programming.

The alternatives that people provide imply some form of mutation, whereas the ternary form is a non-mutating operation. It's the reason why you often use the ternary operator to initialize a const object or when doing branching at compile-time in C or C++.

# ? Sep 24, 2012 23:03

Centripetal Horse: Nov 22, 2009; Fuck money, get GBS

This could have bought you a half a tank of gas, lmfao -
Love, gromdul

qntm posted:

This is only a horror if you try it in PHP.

Maybe this one counts as a horror:

ourDiv.removeClass().addClass(((ourIndex%3+1) % 3 ? ((ourIndex%3+1) % 2 ? "chartOptionsLeft":"chartOptionsCenter") : "chartOptionsRight")).attr("id", "chartOption" + ourIndex).attr('name', ourIndex);

I'm pretty much straight-up abusing ?:, here, because it's fun to use. I could have just used a loop counter (or a dozen other methods) to determine whether the item should be left, center, or right, but what fun is that? I also loaded it down with mod operators just because.

# ? Sep 24, 2012 23:26

raminasi: Jan 25, 2005; a last drink with no ice

Suspicious Dish posted:

C code:

(!are_parallel() || // if they're parallel...

umm....

There's no logical implication operator in C++, but fortunately you can emulate one!

# ? Sep 24, 2012 23:30

Zombywuf: Mar 29, 2008

GrumpyDoctor posted:

There's no logical implication operator in C++, but fortunately you can emulate one!

I've used this idiom a lot in SQL. I'm glad I'm not the only one who wants implies to be part of the standard set of logical operations.

# ? Sep 25, 2012 00:31

PrBacterio: Jul 19, 2000

Zombywuf posted:

I've used this idiom a lot in SQL. I'm glad I'm not the only one who wants implies to be part of the standard set of logical operations.

Now you've made me think about if it would be possible to overload the '->' operator in C++ to function as a logical implication operator :gonk:

EDIT:

Zombywuf posted:

No. It only takes its lhs as an argument.

That's why I said I was only *thinking* about it, instead of just doing it

I agree that it does seem impossible at first glance, but with some of the really weird and out there contortions that people have managed to construct out of C++ I wouldn't be prepared to deny the possibility out of hand. Which is why I offered it up as a horror in this thread with a :gonk:

to go along with it ...

PrBacterio fucked around with this message at 00:52 on Sep 25, 2012

# ? Sep 25, 2012 00:41

Zombywuf: Mar 29, 2008

PrBacterio posted:

Now you've made me think about if it would be possible to overload the '->' operator in C++ to function as a logical implication operator

No. It only takes its lhs as an argument.

# ? Sep 25, 2012 00:45

That Turkey Story: Mar 30, 2003

Zombywuf posted:

No. It only takes its lhs as an argument.

->* is a binary operator :}

# ? Sep 25, 2012 00:47

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

=>

I forget, is that actually an operator? Or are we stuck with >=

# ? Sep 25, 2012 01:07

That Turkey Story: Mar 30, 2003

Suspicious Dish posted:

=>

I forget, is that actually an operator? Or are we stuck with >=

No, it's not. But, you could always do <= and have the implication go right to left instead of left to right.

# ? Sep 25, 2012 01:08

PrBacterio: Jul 19, 2000

Suspicious Dish posted:

=>

I forget, is that actually an operator? Or are we stuck with >=

No it's not, but as I said, with some of the weird poo poo people have come up with in C++ I still wouldn't discount the possibility of someone, somehow making it work.

EDIT: More realistically speaking though, it probably would be easiest to just use /implies/ as the operator.

PrBacterio fucked around with this message at 01:17 on Sep 25, 2012

# ? Sep 25, 2012 01:09

Hammerite: Mar 9, 2007; And you don't remember what I said here, either, but it was pompous and stupid.; Jade Ear Joe

Suspicious Dish posted:

=>

I forget, is that actually an operator? Or are we stuck with >=

=> isn't an operator.

edit: oh, you were talking specifically about c/++. This is a pointless post.

# ? Sep 25, 2012 01:09

Zombywuf: Mar 29, 2008

Oh yeah, you can't overload anything that only takes primitive types as arguments (i.e. bool) as well.

# ? Sep 25, 2012 14:45

PrBacterio: Jul 19, 2000

C++ code:

#include<iostream>

namespace SlashOperators
{

class ImpliesOperator
{

public:
	class OperatorApplication
	{
		const bool lhs;
	public:
		OperatorApplication(bool _lhs): lhs(_lhs) {}
		inline bool operator|(bool rhs)
		{
			return !lhs || rhs;
		}
	};

	ImpliesOperator() {}
};

ImpliesOperator implies;

inline ImpliesOperator::OperatorApplication operator|(bool lhs, ImpliesOperator rhs)
{
	return ImpliesOperator::OperatorApplication(lhs);
}

}

using SlashOperators::implies;

#define implies |implies|

int main()
{
	int x, y, z;
	std::cin >> x >> y >> z;
	if(x == y implies y == z)
	{
		std::cout << "x != y or y == z" << std::endl;
	}
	else
	{
		std::cout << "x == y and y != z" << std::endl;
	}
	return 0;
}

The implies operator, with the same precedence and associativity as the | or operator. You're welcome :colbert:

# ? Sep 25, 2012 16:17

Floor is lava: May 14, 2007; Fallen Rib

Whoops. Swing and a miss.

Floor is lava fucked around with this message at 17:36 on Sep 25, 2012

# ? Sep 25, 2012 17:31

Zamujasa: Oct 27, 2010; Bread Liar

code:

/* 50 lines into a <style> tag */
   ...
   position: absolute;
   top: 240px;
   left: 320px;
   width: 15px;
}
<?}}?>


      </style>
   </head>
   <body>
<script ...

It's like a little PHP UFO! :3:

# ? Sep 25, 2012 17:42

That Turkey Story: Mar 30, 2003

Zombywuf posted:

Oh yeah, you can't overload anything that only takes primitive types as arguments (i.e. bool) as well.

Yeah, but you can get around this by just wrapping your arguments.

_( implication_goes_here ) <= some_bool && some_other_bool

You'd want the implication_goes_here part to be a function object, though, so that you get lazy evaluation.

# ? Sep 25, 2012 18:15

Zombywuf: Mar 29, 2008

That Turkey Story posted:

Yeah, but you can get around this by just wrapping your arguments.

_( implication_goes_here ) <= some_bool && some_other_bool

You'd want the implication_goes_here part to be a function object, though, so that you get lazy evaluation.

Think I prefer PrBacterio's solution.

# ? Sep 25, 2012 18:57

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

Zombywuf posted:

Think I prefer PrBacterio's solution.

I think I prefer neither of them, and you should too.

# ? Sep 25, 2012 19:03

Golbez: Oct 9, 2002; 1 2 3!
If you want to take a shot at me get in line, line
1 2 3!
Baby, I've had all my shots and I'm fine

When a coding error becomes a coding horror: Added a line to my PHP to spawn a new instance of itself, due to inherited weirdness. But it wasn't dying right and kept spawning new processes.

In my attempt at killing them, I ran an untested command to kill all of my processes. Except I hosed it up and apparently killed everything *except* my processes, which meant I killed our production server. :saddowns:

edit: and yes I know the real horror is doing development work on the production machine.

Golbez fucked around with this message at 20:27 on Sep 25, 2012

# ? Sep 25, 2012 20:22

Hammerite: Mar 9, 2007; And you don't remember what I said here, either, but it was pompous and stupid.; Jade Ear Joe

quote:

The range of character sequences expected should be strictly enforced by the javascript on the front end and the code on the backend. Any submission with unexpected characters or the < and > and ` characters should be rejected, before it could be used in any kind of database query. Don't send the submitted data back to the browser though, just report to the user that unexpected characters were detected. Odd character sequences like '1 AND 1', 'CREATE TABLE' and 'DROP TABLE' and 'SELECT *' should be filtered against. Multiple adjacent spaces should be collapsed into one before the filtering. Numbers should be checked to make sure they are in the range expected. On some of the sites I've made, if the backend code detects any kind of SQL injection attack, it logs the ip address and automatically blocks that ip address for a few hours, in case its a script kiddy attack that is going to send a billion requests in an hour trying all kinds of attacks.

And any passwords sent should be sent through the https protocol, if you are concerned about security and people are paying money for the service. You don't really need to pay the fee for a certificate, but most browsers will pop up a warning.

And I think that mysql_real_escape_string is deprecated. I think its just mysql_escape_string now.

# ? Sep 25, 2012 22:29

Optimus Prime Ribs: Jul 25, 2007

quote:

And I think that mysql_real_escape_string is deprecated. I think its just mysql_escape_string now.

I hope, for your sake, that you don't work with the person who said that. :ughh:

# ? Sep 25, 2012 22:34

Lysidas: Jul 26, 2002; John Diefenbaker is a madman who thinks he's John Diefenbaker.; Pillbug

Poe's Law definitely seems to apply to that.

# ? Sep 25, 2012 22:44

pokeyman: Nov 26, 2006; That elephant ate my entire platoon.

Sounds quite straightforward really. All you have to do is filter against odd character sequences.

# ? Sep 25, 2012 22:50

Zamujasa: Oct 27, 2010; Bread Liar

Hammerite posted:

Stuff about filtering

My "site" actually does some filtering against typical script-kiddie attacks (usually obvious identifiers for them), and adds the requester to a block list that lasts for a few hours, just to make those scripts not actually cause much load on the server while they fire their 10-per-second requests off.

The rest of it is just pure bullshit though. Proper escaping (or just using a real database object and avoid manual queries as much as is feasible will prevent that poo poo.

# ? Sep 26, 2012 00:06

Adbot: ADBOT LOVES YOU

# ? May 27, 2024 07:34

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

Whenever I'm facing a registration form that has a bunch of arbitrary tax code validations for the password written on the side, I have visions of a garden variety IT know-it-all going into soapbox mode with more or less those same words.

# ? Sep 26, 2012 00:33

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 »