Closed pitman closed 12 years ago
Ah, actually it is working fine at giving you the option when the cids are in conflict, so that is good. It is just the URLs of the two options you are presented with are written wrongly by the template. This was because I forgot to update them when I change the record URL routing. However, this is updated in the latest version, which I will push to the server soon.
In the collection http://bibsoup.net/pitman/jordan_michael_i which I just refreshed, it appears that all the view record links fail, apparently due to duplication of cids in records. It looks like this dataset has each cid appearing twice. This was a mistake on my side, but it should be handled better by BibSoup. This may be a good test dataset to test handling of repeated cids. A typical view record link is
http://bibsoup.net/pitman/jordan_michael_i/yxUduqMAAAAJ:UeHWp8X0CEIC
This gives a nice error message "Unfortunately, there are multiple records in the collection with the same identifier ... " and indicates there are tow records assoc with this cid, namely
http://bibsoup.net/record/5e4b830982fc4af5899a6c1db8a9a4f4
http://bibsoup.net/record/fc89f7de6b0b4b918c449e74f7e6e397
but each of these returns "No results". Good handling of duplicate cids should allow BibServer to be used as part of a UI to assist users in deduping their datasets. I dont think it should necessarily be regarded as an error to have duplicate cids. It may rather be a feature, if such duplication is used as a method of flagging records for the same "thing" in the biblio sphere. Exactly what is meant by "same" is at present up to the user to decide. But we could provide some usage guidelines. Compare recent discussion on the list of issues of clustering and deduping of records.