Revisit EPMC licence detection

richard-jones commented 9 years ago

[x] use the url prefixes to detect the licence, so we don't need to know about all the variations on CC
[x] if there is a licence, but it is not one we recognise, record it as "non-standard-licence" and do not send on to OAG

richard-jones commented 9 years ago

We also should probably look at normalising the text when we compare licence type strings, so that CC BY matches "cc-by".

emanuil-tolev commented 9 years ago

if there is a licence, but it is not one we recognise, record it as "non-standard-licence" and do not send on to OAG

Er, ok, but if we do that we're not gonna get what's an author manuscript and what's not. Because author manuscripts mostly don't have an open license. So by not sending closed articles to OAG we also can't see if they're author manuscripts or not. Maybe this matters, maybe it doesn't (e.g. if being an author manuscript is one of Wellcome's conditions for Openness, it doesn't matter because the license isn't even open to begin with). See #16 too.

richard-jones commented 9 years ago

In response to the above: any item with a PMCID for which we haven't yet been able to determine the AAM status will be sent to OAG even if we have already detected a licence (closed or open), so this shouldn't be an issue.

Otherwise, this issue is now complete.

CottageLabs / oacwellcome

Revisit EPMC licence detection #40