[feat] - Support S3 Source Resumption

[ahrav]

Description:

This PR adds resumption to the S3 source.

Checklist:

• Tests passing (make test-community)?
• Lint passing (make lint this requires golangci-lint)?

[rgmz]

This PR adds resumption to the S3 source.

This would be a beneficial capability for other sources. e.g., resuming a large GitHub org scan.

[rosecodym]

This seems pretty straightforward, although I would like to see a test case for (and handling of) the case where an in-progress bucket is ignored while the scan is stopped (described in an inline comment)

[rosecodym]

This seems drastic - what do you think of instead restarting the scan from the beginning when we can't get a resume point? I feel like we should err on the side of scanning too much rather than scanning too little.

[rosecodym]

Why are we re-using the existing resume info? I expected to see something from the progress tracker here.

[ahrav]

I was essentially saying, "Don’t modify EncodeResumeInfo—it’s already been updated by progressTracker, so just use it as-is." Since progressTracker and s.Progress reference the same underlying object, would it be clearer if we explicitly used s.progressTracker.Progress instead?

[rosecodym]

This doesn't reset entirely, right? It just resets for a new page? I wish I'd been well enough to leave that comment on #3568 :(

[ahrav]

Yea, this was my mistake. I'm going to fix the tracker so it's named more accurately. here
It's no longer a progress tracker, it's pretty much just a checkpointer.

[ahrav]

This seems pretty straightforward, although I would like to see a test case for (and handling of) the case where an in-progress bucket is ignored while the scan is stopped (described in an inline comment)

done