jcowey
commented
2 years ago I was wrong. Please do BOTH DCLP and HGV files.
Here are the relevant comments from Mike in
https://github.com/papyri/idp.data/pull/280
A few things to note:
-
Changes have been made only to the APIS files; HGV/DCLP have been left untouched. Perhaps Carmen can cook up something to transfer the new links to those databases, too.
-
Princeton has a habit of housing multiple fragments within the same glazing. Occasionally (but not always) a separate page has been created in the catalogue for each of the individual pieces, but where that is not true I've resorted to using the single hyperlink for the multiple papyri under the same glass. Where that happens, the catalogue entry refers to another papyrus even though the image also depicts the papyrus in question. I acknowledge that will be potentially confusing for some users, though the intent is to provide access to an image of the papyrus.
Edelweiss
With the following commit:
https://github.com/papyri/idp.data/commit/8b532008e66968f7a3c16f6bab586b4d65c4c803
Mike Sampson uploaded new image links for Princeton images. The digital library there had two rounds of creating URN / URL s. Unfortunately the URN uploaded well over a year ago: http://arks.princeton.edu/ark:/88435/s7526g009 has been replaced with https://dpul.princeton.edu/papyri/catalog/v405sd89 which also has an ark: ark:/88435/v405sd890
Unfortunately moreover the old ark does not seem to be
ark:/88435/s7526g009is not pushed through to the new oneark:/88435/v405sd890, as far as I can see.Would you please crawl
https://github.com/papyri/idp.data/tree/master/APIS/princeton/xml
for
dpul.princeton.edu/papyri/catalogand using the relevant TM / HGV idno replace the old arks in the corresponding files inhttps://github.com/papyri/idp.data/tree/master/HGV_meta_EpiDoc
as far as I remember Mike Sampson dealt with the DCLP files accordingly, so you do not need to do them.
I dislike overwriting URNs but it seems the best thing to do here, I think. It is possible that the HGV meta files will in this way get "new" links (the ones we failed to pick up years ago)