For example:
hash("HelloWorld") = 1234567
Is there any built-in function could do this ?
Thanks.
-
2Not Go specific but interesting and related: stackoverflow.com/a/107657/532430thwd– thwd2012-11-27 21:11:34 +00:00Commented Nov 27, 2012 at 21:11
-
1that's known as prehash, theoretically.Roy Lee– Roy Lee2016-05-15 10:12:55 +00:00Commented May 15, 2016 at 10:12
Add a comment
|
2 Answers
The hash package is helpful for this. Note it's an abstraction over specific hash implementations. Some ready made are found in the package subdirectories.
Example:
package main
import (
"fmt"
"hash/fnv"
)
func hash(s string) uint32 {
h := fnv.New32a()
h.Write([]byte(s))
return h.Sum32()
}
func main() {
fmt.Println(hash("HelloWorld"))
fmt.Println(hash("HelloWorld."))
}
(Also here)
Output:
926844193
107706013
7 Comments
Ivan Black
Also, it's possible to use single instance and
Write/Reset, instead of creating New32a every time. I don't know how much it's cost. For example: play.golang.org/p/aUeNBo755g
Gokul Raj Kumar
Are these hashes unique? can i use it as a unique identifier?
Tomer
These hashes are not very unique. The chances of two different strings matching the same uint32 are approximately 1 to 4 billion. I would not use it for uniqueness. Here is an example which may prove more useful: every task has a unique "string", and I would like to uniformly distribute the tasks between 50 queues. I would do (with the above function) hash("HelloWorld") % 50.
Backgammon
@iwind Every hash function into a smaller space (like arbitrary-length string into uint32 here) is guaranteed to have collisions so "I found one specific collision" is not an argument for whether a method is good or bad. If you analyzed the method and found a high collision rate or nonuniform distribution or some other meaningful flaw that would be a compelling argument, but I imagine such analyses already exist for FNV.
Adé
@IvanBlack I wonder about the concurrency and race condition issues when using that same New32a instance in each call to getHash. In my mind, two simultaneous getHash function calls could interleave in a way that results in one of them Resetting the other's Write before it does a Sum call, giving a result you wouldn't want.
|
Here is a function you could use to generate a hash number:
// FNV32a hashes using fnv32a algorithm
func FNV32a(text string) uint32 {
algorithm := fnv.New32a()
algorithm.Write([]byte(text))
return algorithm.Sum32()
}
I put together a group of those utility hash functions here: https://github.com/shomali11/util
You will find FNV32, FNV32a, FNV64, FNV64a, MD5, SHA1, SHA256 and SHA512
Comments
