Matthias Gatto

I'm presently trying to find the bug, out of curiosity, which call exactly did you made (having the full URL could held debug I guess) ?

thanks for the reply, I've made a Pull Request that should improve url encoding here: https://github.com/curl/curl/pull/7600 Note that presently the PR doesn't fix your issue, but should do so, so...

just to let you know I've update the PR, would be nice if you test it, and tell us if it work with your use case.

I'm not sure where the error come from, but in "working curl version" you have `--header "kbn-xsrf: true" ` and not in the `--aws-sigv4` version. not sure what this is...

Good catch, that might be the problem. Also it seems I've also skip some "lowercase" step. I'll try to make a patch asap.

I've made a PR, note that because I've never use "AWS Elasticsearch Service" I didn't test if the auth work, I've just follow @dwagin suggestion. @poborin is it okay for...

The best would be that you build curl yourself, using my branch: https://github.com/outscale-dev/curl/tree/http_aws_sigv4_encoding Otherwise I've just make an appimage for Linux: https://github.com/outscale-mgo/curl-appimage/releases/tag/sigv4_encoding you can also build the appimage using the...

> This commit adds the X-Amz-Content-Sha256 header to all requests as defined by the AWS documentation: > > https://docs.aws.amazon.com/AmazonS3/latest/API/sig-v4-header-based-auth.html > > Note, some SigV4 server implementations appear to strictly enforce...

> Why does it have to sort the headers? the "header" have to be sort in the canonical header and the signed header. but if I'm correct, I don't change...

note that I still need to converts sequential spaces to a single space, in order to be conform to the algorithm.