License expressions of system packages (as in Debian) are too complex

[pombredanne]

(lgpl-2.1-plus AND gpl-3.0-plus AND (lgpl-3.0-plus AND gpl-2.0-plus)) AND lgpl-2.1-plus AND gpl-3.0-plus AND gfdl-1.1-plus AND gfdl-1.3-plus AND (lgpl-2.0-plus AND lgpl-3.0 AND gpl-1.0-plus AND gpl-3.0 AND gfdl-1.3) AND lgpl-2.1-plus AND (lgpl-3.0-plus OR gpl-2.0-plus) AND lgpl-2.0-plus AND gpl-3.0 AND (bsd-new OR gpl-1.0-plus) AND bsd-simplified AND public-domain AND isc AND cc0-1.0 AND mit AND mit AND agpl-3.0 AND agpl-3.0 AND agpl-3.0 AND (lgpl-3.0-plus AND gpl-2.0-plus) AND apache-2.0 AND apache-2.0 AND apache-2.0 AND apache-2.0 AND mit AND mit AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus AND gpl-2.0-plus) AND mit AND mit AND lgpl-2.1-plus AND (lgpl-3.0-plus AND gpl-2.0-plus) AND (lgpl-3.0-plus OR gpl-2.0-plus) AND gpl-3.0-plus AND gpl-3.0-plus AND gpl-3.0-plus

could be deduped to:

'(lgpl-2.1-plus AND gpl-3.0-plus AND (lgpl-3.0-plus AND gpl-2.0-plus)) AND lgpl-2.1-plus AND gpl-3.0-plus AND gfdl-1.1-plus AND gfdl-1.3-plus AND (lgpl-2.0-plus AND lgpl-3.0 AND gpl-1.0-plus AND gpl-3.0 AND gfdl-1.3) AND (lgpl-3.0-plus OR gpl-2.0-plus) AND lgpl-2.0-plus AND gpl-3.0 AND (bsd-new OR gpl-1.0-plus) AND bsd-simplified AND public-domain AND isc AND cc0-1.0 AND mit AND agpl-3.0 AND (lgpl-3.0-plus AND gpl-2.0-plus) AND apache-2.0'

or simplified to:

'agpl-3.0 AND apache-2.0 AND bsd-simplified AND cc0-1.0 AND gfdl-1.1-plus AND gfdl-1.3 AND gfdl-1.3-plus AND gpl-1.0-plus AND gpl-2.0-plus AND gpl-3.0 AND gpl-3.0-plus AND isc AND lgpl-2.0-plus AND lgpl-2.1-plus AND lgpl-3.0 AND lgpl-3.0-plus AND mit AND public-domain'

And this without even applying any transform.

We should return a simplified version of every license expression.

[AyanSinhaMahapatra]

This is aboutcode-org/license-expression#67

[mjherzog]

I am in favor of presenting a streamlined license expression, but the "hairy" version is also useful as a digest of the number of license detections so I do not think that we should lose that data completely. We need to show the raw detection data and somehow mark that we applied streamlining (summarization) to the license expression.

[chinyeungli]

The PR at aboutcode-org/license-expression#128 returns the above "hairy" expression to

lgpl-2.1-plus AND gpl-3.0-plus AND lgpl-3.0-plus AND gpl-2.0-plus AND gfdl-1.1-plus AND gfdl-1.3-plus AND lgpl-2.0-plus AND lgpl-3.0 AND gpl-1.0-plus AND gpl-3.0 AND gfdl-1.3 AND (lgpl-3.0-plus OR gpl-2.0-plus) AND (bsd-new OR gpl-1.0-plus) AND bsd-simplified AND public-domain AND isc AND cc0-1.0 AND mit AND agpl-3.0 AND apache-2.0