I'm using the react-s3-uploader node package, which takes in a signingUrlfor obtaining a signedUrl for storing an object into S3.
Currently I've configured a lambda function (with an API Gateway endpoint) to generate this signedUrl. After some tinkering, I've got it to work, but noticed that I have to define in my lambda function the content-type, which looks like this:
var AWS = require('aws-sdk');
const S3 = new AWS.S3()
AWS.config.update({
region: 'us-west-2'
})
exports.handler = function(event, context) {
console.log('context, ', context)
console.log('event, ', event)
var params = {
Bucket: 'video-bucket',
Key: 'videoname.mp4',
Expires: 120,
ACL: 'public-read',
ContentType:'video/mp4'
};
S3.getSignedUrl('putObject', params, function (err, url) {
console.log('The URL is', url);
context.done(null, {signedUrl: url})
});
}
The issue is that I want this signed url to be able to accept multiple types of video files, and I've tried setting ContentType to video/*, which doesn't work. Also, because this lambda endpoint isn't what actually takes the upload, I can't pass in the filetype to this function beforehand.
You'll have to find a way to discover the file type and pass it to the Lambda function as an argument. There isn't an alternative, here, with a pre-signed PUT.
The request signing process for PUT has no provision for wildcards or multiple/alternative values.
In case anyone else is looking for a working answer, I eventually found out that react-s3-uploader does pass the content-type and filename over to the getSignin url (except I had forgotten to pass the query through in API Gateway earlier), so I was able to extract it as event.params.querystring.contentType in lambda.
Then in the params, I simply set {ContentType: event.params.querystring.contentType} and now it accepts all file formats.