Gary and I had been working on this for the call today. Still a bit of draft
but we can discuss.
This repo will have multiple moderators.
Directory Layout
-----------------------
spdx/
Readme
licenseid/
testfiles.json
0BSD-id-good.[c|sh|js| whatever]
0BSD-id-bad.[c|sh|js| whatever]
...
licensetext/
testfiles.json
0BSD-canonical-good.txt
0BSD-canonical-bad.txt
...
licenseheader/
testfiles.json
0BSD-header-good.[c|sh|js| whatever]
0BSD-header-bad.[c|sh|js| whatever]
...
contrib/
Readme
testfiles.json
somefilename
...
The testfiles.json directory will contain a json file with the expected test
results regardless of the directory the file (pass/fail etc). The directory
structure above is for us humans and this file is for programmatically checking
the results.
Spdx
---------
Split the spdx and contrib directories for clarity, ease of maintenance, etc. I
don't think we should comingle the files.
The SPDX text cases include detection of the license text, license identifier
and standard header as defined by the license
list. One item to work through. There seem to be very few standard license
headers as defined by the list. For example there
are none for BSD, MIT Eclipse, etc.,. Need to work through that.
licenseid/ directory contains test files using the SPDX-LICENSE-IDENTIFIER
licensetext/ directory contains canonical versions of the license text in a
text file. No comments or other text with it.
licenseheader/ directory contains test files using the standard license header.
Contrib
----------
These are user contributions. These can be much broader test cases than the
SPDX ones.
- Other than a good and bad directory do we need anything for structure? Should
it follow spdx format?
- How to vet disputes on files and what goes in?
_______________________________________________
Spdx-tech mailing list
[email protected]
https://lists.spdx.org/mailman/listinfo/spdx-tech
