Is it possible to specify a file name inside .zip archive compressing standard input to standard output using cli tools?

Question

I'm trying to find a way to specify a file name in a compressed.zip archive while compressing it from standard input to standard output. I want to achieve it without creating a temporal file during a process.

Currently I have an example script which creates mysqldump, passes the result as input to a zip command and outputs the stream to aws s3 command to save a result to S3.

Here is the example:

mysqldump ... | zip | aws s3 cp - s3://[bucket_name]/output.sql.zip

The problem is zip by default saves a file inside zip archive with name "-".

Maybe there is a way how to pass specific file name inside an archive using zip command or any other zip library?

Answer 1

Recent Linux distributions have a streamzip script which I wrote to handle this usecase. In the example below the -member commandline option sets the member within the streamed zipfile to data.sql

mysqldump ... | streamzip -member data.sql | aws s3 cp - s3://[bucket_name]/output.sql.zip

If the mysqldump is larger than 4G, include the -zip64 option with streamzip .

If you don't have streamzip available download from https://github.com/pmqs/IO-Compress/blob/master/bin/streamzip