C# string literal GetHashCode() evaluates differently given context?

Question

So I have some code that is failing because two string literals have HashCodes that can evaluate to the same thing. And while I appreciate that collision can happen I wasn't quite expecting this. But whilst debugging an issue in code a colleague and I found that in an immediate window if we did

"55d02ProductAd".GetHashCode() == "55b0tProductAd".GetHashCode()

It would evalute to true. Not ideal, but not impossible. When we described this to another colleague, in his disbelief he wrote up a scratch program that did

var h1 = "55d02ProductAd".GetHashCode();
var h2 = "55b0tProductAd".GetHashCode();
Console.WriteLine(h1 == h2);

In the above, they do not evaluate to the same thing. We have our monitors next to each other and we are confused as to the different outputs. Any thoughts?

msdn.microsoft.com/en-us/library/… You should never persist or use a hash code outside the application domain in which it was created, because the same object may hash across application domains, processes, and platforms. — L.B
– L.B, Commented Sep 2, 2016 at 18:44
they do have the same hash code at least in .net 4.5 dotnetfiddle.net/O3VUtX — Daniel A. White
– Daniel A. White, Commented Sep 2, 2016 at 18:48
1. Do both on the same machine - different machines may produce different hashes. 2. The both equate to true on my machine.3. Dont ever use hash codes for equality. Hash codes are not unique. — Rick the Scapegoat
– Rick the Scapegoat, Commented Sep 2, 2016 at 18:48
@rick Even on the same machines it's allowed to produce different hashes. — Servy
– Servy, Commented Sep 2, 2016 at 18:50
Related: referencesource.microsoft.com/#mscorlib/system/string.cs,838 — Theodoros Chatzigiannakis
– Theodoros Chatzigiannakis, Commented Sep 2, 2016 at 18:53

Servy · Accepted Answer · 2016-09-02 18:45:44Z

4

Hash codes are only contractually obligated to produce the same hash for a value within the context of a single application's execution. Since you're comparing the values of GetHashCode form entirely different applications, there is no obligation for them to be equal.

answered Sep 2, 2016 at 18:45

Servy

204k27 gold badges347 silver badges470 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

Stilgar Over a year ago

He is not saying that he is comparing values from different applications. He says that one application produces different values.

Servy Over a year ago

@Stilgar No, he's not. He's saying he had an application, computed these hashes, then a coworker created a new application, hashed the same strings, and got different values. That's an entirely different application. And of course, even if it was the same application, if it was a different invocation of it, the hash would be allowed to vary. If you just print the hash of "Hello World!" and run it over and over it's allowed to give a different value.

Haney Over a year ago

Also, the birthday problem makes collisions reasonably common. Many structures that rely on hash codes (like Dictionary<TK, TV>) handle collisions elegantly.

Servy Over a year ago

@Haney Well, any that don't are buggy. Every sensible implementation needs to handle collisions.

Rick the Scapegoat Over a year ago

@Haney sensible handling is not using hashes to determine equality - but rather inequality. It's really an easy way to exclude things from matching algorithms, within the same application domain.

Collectives™ on Stack Overflow

C# string literal GetHashCode() evaluates differently given context?

1 Answer 1

5 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Related