Closed jpeddicord closed 6 years ago
Never mind -- I was testing this on our own source headers and it's actually not an 0.8 spdx match for the header; it's 0.79. askalono is correct.
There's a bigger issue here in collecting different kinds of license headers. I'm not sure this project should be the curator of that. That's a bit out-of-scope though.
For some reason, --optimize on a low-confidence-overall file prints "Error: Confidence threshold not high enough for any known license", even though it clearly found a license header shown in
RUST_LOG=debug
. This is probably just a print/format issue.A unit/integration test to prevent this from regressing again would be awesome.