Brilliant exception handling I found in an app i had to work on
I'm offended that you'd put my code on the internet.
this post was submitted on 17 Jun 2023
101 points (95.5% liked)
Programmer Humor
32601 readers
235 users here now
Post funny things about programming here! (Or just rant about your favourite programming language.)
Rules:
- Posts must be relevant to programming, programmers, or computer science.
- No NSFW content.
- Jokes must be in good taste. No hate speech, bigotry, etc.
founded 5 years ago
MODERATORS
Seen this so many times, i weep for humanity
“I’ll come back to this”
Actually, exception rethrowing is a real thing - at least in Java. You may not always want to handle the exception at the absolute lowest level, so sometimes you will instead "bubble" the exception up the callstack. This in turn can help with centralizing exception handling, separation of concerns, and making your application more modular.
It seems counter-intuitive but it's actually legit, again at least in Java. lol
Rethrowing caught exception in C# is just throw;, not throw ex;. This will delete old stack trace, which is very punishable if someone debugs your code later and you're still around.
I am a somewhat new C# developer (2 years). Could you explain more about this?
throw ex; treats ex as a new exception, so, it starts a new stack trace for it from itself and deletes stack trace that was saved in ex.StackTrace. On the other hand, throw; takes already present exception in the scope and throws it without modifying the stack trace, preserving the original method that threw ex in the stack trace.
I feel like I wrote the same thing twice. I'm a bit bad with explaining stuff, feel free to ask more specific questions if you still don't understand the difference.
Lol what’s wrong with this if the parent function catches it
If this is C# (and it looks like it is), this leads to you losing the original stack trace up until this point.
The correct way to do this in C# is to just throw; after you're done with whatever you wanted to do in the catch.
wait what ?
So you are saying that the following code will keep throwing e but if I used throw e; it would basically be the same except for the stack trace that would be missing the important root cause ?!
try {
} catch (WhateverException e) {
// stuff, or nothing, or whatever
throw;
}
You don't catch it if that's the case
The catch is useless if it's just throwing the exception anyway
You could just not catch it and it'll get thrown up the stack anyway.
Depending on the language it either does nothing and just adds code bloat or (and this would be much worse) it will catch any exception that can be implicitly cast to type Exception and throw it as type Exception. So the next higher scope would not be able to catch e.g. a RuntimeException or w.e. to handle appropriately. It could only catch a regular Exception even if the original error was a more detailed type.
It's C# so it's just rethrowing the original exception.
It might also be messing with the stack trace though which can be a bit frustrating for future debugging. But that's only a vague recollection of something I read in the past so I could be wrong
Throwing exceptions are very costly due to the stack trace, so building the stack trace twice will cause a big performance hit
Correct me if I'm wrong, but this will actually cut the stack trace and then start another one from your try-catch block, which is an evil thing to do towards those who will actually read your stack traces. To preserve the stack trace you do throw;, not throw ex;, and I'm assuming IDE is underlining that statement exactly for this reason.
Then the parent function would catch the original exception if it was never caught in the first place. All this does is bork the stacktrace.
while (true) {
start();
}
As ugly and stupid this feels, everyone here at one point or another is guilty of this. Some "temporary" code to place a breakpoint on cause you gave up trying to understand why the five thousand line function called compareVersion keeps throwing an exception trying to compare 1.0 to 2.0.
This can't compile, can it? Only a madman commits code which doesn't compile.
Why wouldn't it? It's syntactically valid C#, with the added bonus of destroying the stack trace
Who needs stack traces anyway? Just search all of your code on the word throw until you have the right one
view more: next ›