FileSize is 0 bytes when curling a file from a presigned url

[VGetsov]

Hey there, I have the following issue:

I curl a file from AWS S3 using a presigned url and then I pipe it to exiftool. However, the FileSize is usually 0 bytes, sometimes it's a bit more (e.g. 66 bytes) but it doesn't seem to be equal to the real size of the file. I ran the command with verbose mode and the Content-Length header shows the real size.

Below is my code, where {source} is a presigned url, {destination} is a presigned url to write to and {tagsArg} is used to add stuff to the Keywords field if needed.

I also tried the following command in the terminal by simply executing

Code Select Expand

curl {source} | exiftool - but the FileSize is still wrong.

I'd really appreciate some help 

       

Code Select Expand

const exif = spawn(
          `curl -s '${source}' | exiftool ${tagsArg} - | curl -X PUT -H "Content-Type: application/octet-stream" '${destination}' --upload-file -`,
          { shell: true },
        )


[StarGeek]

Since you are receiving the file from a pipe redirected from cURL, it might not be possible to get a file size. It would require cURL to get the file size from the header and pass that along. But I can't find any source that indicates that cURL will do this.

I do find some answers where you could use cURL to get the header for an URL and extract the file size (Content-Length header) from that. But I also know that not all servers will pass this value. You would see this when downloading a file through a browser, and it doesn't give a percentage showing how much of the file has been downloaded.

[StarGeek]

See this previous thread about file size and piped cURL data.

[VGetsov]

Thank you for the reply and the link!

[Phil Harvey]

What version of ExifTool were you using?

Version 12.87 and later should no longer report FileSize for pipes.

- Phil