Zotero finds "false positives"
We are testing the summon search engine for our library catalogue. Zotero seemed to work with this search engine but the zotero-entries don't match the media - see screenshot here:
http://www.screencast.com/t/MiSfBtDX6O
Occasionally this has happened before, but only when I used an ISBN-Number and the magic-wand-icon to add a data set.
Is there anything I ca do from the user-side to fix this?
Is there anything the people from summon can do to make the search engine work with zotero?
http://www.screencast.com/t/MiSfBtDX6O
Occasionally this has happened before, but only when I used an ISBN-Number and the magic-wand-icon to add a data set.
Is there anything I ca do from the user-side to fix this?
Is there anything the people from summon can do to make the search engine work with zotero?
What does it say when you're hovering over the URL bar icon before saving?
It just says something like "Save to Zotero..." when I hover over the Zotero combo-icon.
I believe it may be the case that if you have multiple windows open that will break things for Zotero.
"CD40 is functionally expressed on human breast carcinomas: Variable inducibility by cytokines and enhancement of Fas-mediated apoptosis"
trying again without refreshing the page, I got the "Reason and Rhetoric in the Philosophy of Hobbes"
If I reload the URL or copy/paste it:
http://fhv.summon.serialssolutions.com/#!/search?ho=t&l=de-DE&q=Marketing: Grundlagen
the translator works as expected.
I am not familiar enough with the Summon 2 translator or in how Summon manages sessions to make an educated guess about what is going on.
http://www.screencast.com/t/cayOHODx4aX
The work-around of reloading the URL seems to usually work, but scrolling down to the bottom to allow more references to be added to the page by AJAX leads makes the incorrect import happen again.
Could you download the file Summon 2.js from here:
https://gist.github.com/adam3smith/8b33c35960826326783e/raw/8a92989259e97bbfd4e68a4f73fea92572ab95c8/Summon%25202.js
And place it in the "translators" folder in your Zotero data directory,
https://www.zotero.org/support/zotero_data
replacing the existing file with the same filename. Restart Firefox for good measure and try again. Does that work now?
(Technical note: for some reason doWeb doesn't seem to get the an updated URL, which breaks everything else; I'm now using doc.location.href, which seems to do this correctly. This is also why reloading or pasting works: that does get the right url do doWeb. Same for Scaffold, where the original version worked consistently)
However the quality of the data sets is not too good, and there's other weird things happening:
When I click on a book in a search result list that summon found, the page of that book is loaded. It seems to be exactly same page that would have been loaded if I had searched for that book via our library catalogue- _but_ there is an important difference when I find the book via summon. In this case Zotero does not display the Icon for the item type "book" but the icon for website which leads to a useless data set entry: http://www.screencast.com/t/kDKBmTBu
If I search for the same book via the library catalogue, Zotero displays the folder icon (although its just one medium but I think this is a problem of our library catalogue) and a useful set of data can be saved:
http://www.screencast.com/t/W8sPsnSTRK
If I check the URLs I find that the summon-URL is a lot shorter:
http://www.screencast.com/t/6vyu4HsMs67
The same happens when I search for an e-book via springerlink. If i find the book via summon, zotero displays the website-icon, if I find the book directly on the springerlink-site evrything is fine.
I guess this isn't Zoteros fault? What does Summon have to change to make their search results work properly?
Zotero is not able to match your catalog's site to a translator.
I believe your catalog uses Aleph. We have a translator for that, but the regex is overly restrictive. if you remove 'local_base=fhb01&', it would translate. So: we should add that term to the regex....
The data quality is based on Summon's JSON data, i.e. the data on which the catalog is constructed. Often the data underlying discovery layers isn't great, but if you have something specific, we can take a look (please provide search terms&specific issues for items in question; in this case that's going to be much more helpful than screenshots).
But yes, the translator fix will go up pretty soon, so at least the wrong item issue will be fixed for you.