I noticed another broken feature today -- the search results screen has some code to display "Get full text" links either using an ISSN and OpenURL resolver or URLs from the Solr index. This wasn't working due mainly to the wrong 856 subfield (a rather than u) being specified in import/marc.properties. There was also a glitch where the ISSN / OpenURL link would only display if there was also a URL in the Solr index, which seems like a pointless dependency.
I've fixed the broken parts and rearranged things a bit so that the new logic is this:
If an ISSN exists and an OpenURL resolver is available, use that...
Otherwise, if one or more URLs are found in the Solr record, use them...
Otherwise, by default, display the standard "Available / Checked Out" status area.
I realize that in some cases this may not be ideal behavior -- particularly if you have 856u entries that are NOT links to full text. My goal for the moment was to fix existing broken code, not to change behavior. If anyone has suggestions for better behavior or desirable configuration options to manipulate behavior, please share them, and we can improve the feature to be more flexible.
To look at the relevant changes that I just made, see r1798 in SVN.
Get latest updates about Open Source Projects, Conferences and News.