Using aws-sdk-go, I have been able to sucessfully copy the objects in my s3 bucket when the key contains normal alphanumeric and few special characters like (-,_). But when a key contains an arabic character, the golang aws-sdk throws an error.
NoSuchKey: The specified key does not exist.
status code: 404, request id: 438DC6xxxxxx, host id: Xp+xxxxxxxxxx
The key in the bucket looks like this:
public/10009/img__١٣٤١١١-1600x1200.jpg
The code is pretty straigh-forward as well:
func copyObject(existingKey, key string, svc *s3.S3) {
copyObjectInput := &s3.CopyObjectInput{
Bucket: aws.String("dummy-bucket"),
CopySource: aws.String(existingKey),
Key: aws.String(key),
}
result, err := svc.CopyObject(copyObjectInput)
if err != nil {
log.Fatal("Copy failed due to: ", err) // logs the above error here
}
spew.Dump(result)
}
I am also printing out the key, just in case:
dummy-bucket/public/10009/img__١٣٤١١١-1600x1200.jpg
I was also able to successfully downloaded the image using the aws-sdk-go
, with the same key.
According to the documentation, CopySource must be URL-encoded.
https://docs.aws.amazon.com/sdk-for-go/api/service/s3/#CopyObjectInput
// The name of the source bucket and key name of the source object, separated
// by a slash (/). Must be URL-encoded.
//
// CopySource is a required field
CopySource *string `location:"header" locationName:"x-amz-copy-source" type:"string" required:"true"`
Try this,
import "net/url"
func copyObject(existingKey, key string, svc *s3.S3) {
// existingKey is source bucket and key name separated by "/"
e := url.QueryEscape(existingKey)
copyObjectInput := &s3.CopyObjectInput{
Bucket: aws.String("dummy-bucket"),
CopySource: aws.String(e),
Key: aws.String(key),
}
result, err := svc.CopyObject(copyObjectInput)
if err != nil {
log.Fatal("Copy failed due to: ", err) // logs the above error here
}
spew.Dump(result)
}