Metadata

[MaartenUreel]

Is it possible to configure metadata when uploading / creating a file this way?

[liormizr]

Hi @MaartenUreel Do you mean configure metadata specifically per file? Or per every call to upload / create a file?

[MaartenUreel]

I use S3Path in a lambda function to create one file, so it doesn't matter for me. Any implementation is fine :)

Specifically for that file would be the cleanest I guess.

I need to add 'ContentType' => 'text/html',

[liormizr]

Personally I would have done something like this:

from s3path import S3Path, register_configuration_parameter

# All files (Keys) will be created or updated with ContentType='text/html'
register_configuration_parameter(S3Path('/'), parameters={'ContentType': 'text/html'})

def lambda_handler(event, context):
    ...

Is that good?

[liormizr]

@MaartenUreel did my comment helped you?

[MaartenUreel]

Hello

I ended up writing the file with just boto3. But your approach would have helped probably indeed too, altough it's a bit counter-intuitive to set this as default instead of on file-level...

Maybe a nice addition for a future version :)

[cooperlees]

Howdy - Seems bandersnatch has hit this writing to S3. Is there any plans to support correct Content-Type setting for known files that need it or do you want to keep it the consumer/library user sets it? It shouldn't be a big change to bandersnatch to supply the ContentType on files here.

As someone who has hardly used S3, can an existing file's ContentType be checked and changed? I feel we'll need to create a bandersnatch run to go and check and set the content type on all .html and .json files bandersnatch has uploaded to s3 correctly due to us having this bug ... How would you recommend doing that?