SHA1/256 functions is only possible to use for String, it's impossible for byte[] in painless

[ser-sergeev]

Description

No response

[slobodanadamovic]

Thank you very much for your interest in Elasticsearch.

This appears to be a user question, and we'd like to direct these to the Elasticsearch forum. If you can stop by there, we'd appreciate it. This allows us to use GitHub for verified bug reports, feature requests, and pull requests.

There's an active community in the forum that should be able to help get an answer to your question. As such, I hope you don't mind that I close this.

[ser-sergeev]

@slobodanadamovic The feature requests also must be done on forum? Right now there is no way to calculate a hash for bytes, only for String.

[slobodanadamovic]

@ser-sergeev Apologies, I've interpreted this as a question. This is a correct place to raise an enhancement requests.
But in your case, I believe you could solve this by converting the byte[] to String and then call sha1() or sha256(). Here is what I have in mind:

byte[] bytes = doc['myByteField'].value;
StringBuilder stringBuilder = new StringBuilder();
for (int i = 0; i < bytes.length; i++) {
  stringBuilder.append((char)(bytes[i] & 0xFF));
}
return stringBuilder.toString().sha256();

[ser-sergeev]

@slobodanadamovic Pardon me, I thought, because of UTF-8 encoding, I wouldn't be able to convert from bytes to String.
Sorry for interrupting from a real work. I'm really sorry.

Anyway, big thank you for your answer and patience.

[ser-sergeev]

@slobodanadamovic No, sha256(bytes) and sha256('bytes->String') isn't equal. I compared the result of the code above and the result of sha265 with the same bytes.

[slobodanadamovic]

@ser-sergeev I see now that there can be potential issues with wrong encoding. Let me tag the responsible team for painless to get their input on this.