[Question] Does --stdout stream the entire download process?

[jolespin]

I'd like to use Kingfisher to download some large files but I get charged for I/O. Was wondering if using --stdout does the following operation:

• Download SRA
• Pipe SRA to Fastq converter (Interleaved)
• Simultaneously write gzipped paired fastq files (e.g., bbsuite's reformat.sh in=stdin.fastq out1=sample_1.fastq.gz out2=sample_2.fastq.gz)

Or if it just write the final output to stdout.

[wwood]

Hi @jolespin ,

Good question. The way it works is to download the .sra file to disk, and then stream from there. Your pipeline sounds concept reasonable.

So in effect, you are right. The main use-case for this is to source them in the cloud where they are locally stored as .sra format.

One possibility (not implemented atm) is to stream the reads from an ENA download. These would already be split since ENA keeps them this way (as fastq.gz). I probably won't get around to implementing this but would certainly be happy to review a PR.

ben