Clicking RIS from Google Scholar creates new collections for each item
I keep getting the following
"Do you want to import the file "scholar.ris"?
Items will be added to a new collection."
Hi, I've used Zotero Firefox extension extensively in the past, and have come back to download Standalone. However, my usual one-click add-in use seems to have disappeared. Every click creates a new collection, and there doesn't seem to be a single option to default them to existing collections. Immensely confusing/frustrating.
"Do you want to import the file "scholar.ris"?
Items will be added to a new collection."
Hi, I've used Zotero Firefox extension extensively in the past, and have come back to download Standalone. However, my usual one-click add-in use seems to have disappeared. Every click creates a new collection, and there doesn't seem to be a single option to default them to existing collections. Immensely confusing/frustrating.
Also, where are you clicking "RIS" on google scholar? I can get it to show "Import to Endnote" and "Import to RefMan", both of which do also import into Zotero. If you actually import an RIS file, yes, that does go into a new collection and always has.
"Cited by 12 - Related articles - All 2 versions - Import into RefMan"
Clicking on "Import into RefMan" saves these as RIS files, and then when I open them with Zotero from the folder (or from Chrome) they open as new collections.
A manual import of those files will create a new collection, but the connector makes it possible to get one-click imports from Chrome while using Standalone.
@article{thompson1981life,
title={Life histories and the analysis of social change},
author={Thompson, P.},
journal={Biography and Society},
pages={289--306},
year={1981},
publisher={Beverly Hills: Sage Publications}
}
)
rather than to a file. I'll have to get used to the operation using the URL bar, which is far more unwieldy on a page with dozens of results and the results look like this (with up to 100 titles)
[BOOK] The voice of the past: Oral history (books.google.com)
[CITATION] Life histories and the analysis of social change (Beverly Hills: Sage Publications)
[CITATION] The Voice ofthe Past: Oral History (New York: Oxford University, 1988)
The second issue is that I didn't understand that the connector was required, as "save items to Zotero from within browsers other than Firefox" suggested operation like Zotero classic. Perhaps replacing 'from within' with 'using' would make this clearer.
I don't mean to complain. It's still easier than using Endnote, but not incredibly so (and now slightly more complex).
It may be enough to have the "Use Zotero to open RIS/Refer..." preferences box checked in Zotero. If not, playing with the Chrome file handling should do the trick.
SA is beta software and the documentation isn't done - I don't think the difference between "using" and "from within" changes anything for most users. Note that the connectors are called "Standalone Connectors" - i.e. pieces of software that connect a browser with Standalone.
In Standalone with the connectors, the same code gets executed when you click a link to an RIS file from a webpage or double-click an RIS file on the drive, which is why that dialog appears, and why the item(s) import into a new collection. We could potentially provide a preference to suppress the dialog and/or not make a new collection on import.
I think that would be highly desirable. I don't think it's that crucial for sites with an existing translator, but for sites with RIS export but no Zotero translator that option would be _very_ useful.
When I am on the SpringerLink site and use the address bar icon the download does not include the article abstract and includes both the print and electronic ISSNs.
If I download the same record using the Springer "Export Citation" process, I can get the abstract and only one ISSN. However, that requires many clicks:
1. EXPORT CITATION
2. select citation + abstract
3. export citation (again)
4. FF OPEN WITH dialog box select Zotero SA OK
5. Citation.ris may be an application....Are you sure... OPEN
6. Import file Do you want to import the file "citation.ris"? Items will be added to a new collection. OK
7. Move item into the desired collection
8. Delete item from collection
9. Delete the now empty collection
As it stands now, when I download records from SpringerLink using the address bar icon; I download each item individually, copy and paste-in the abstract, and remove one of the ISSNs. (I don't want two ISSNs for Springer journals and only one for those of other publishers.) A further problem with this method is that my university library's proxy connection to SpringerLink is always slow and usually painfully slow.
The "download into a new connection" has been the only Zotero design issue that causes me extra work. I've followed several forum issues where this has been discussed. It has been my only real frustration. If there is a philosophical need to keep items that were downloaded from a publisher's RIS separate from those downloaded using the address bar icon, please allow these to be downloaded into one single collection for that purpose instead of creating a new collection with each download. This could be a new standard collection similar to the recently added standard collections: "outside" and "duplicate items".
Thank you for considering changes to the way Zotero handles ris imports.
This has nothing to do with keeping publisher RIS separate - with the automatically import RIS/Refer option, Zotero does download into the selected collection.
The issue with Springer is that
a) the Zotero translator doesn't work (and Springer link uses DOI instead)
b) the RIS file is served with the wrong mime type, so that the automatically import option doesn't work, even in Zotero FF.
Springer would save work by simply modifying how their website serves article metadata instead of setting up several hundred XML feeds to my FTP system. Because I will be selectively including articles from their journals instead of indexing them cover-to-cover, I would prefer to gather metadata from the SpringerLink website using Zotero instead of getting metadata fed to my pre-production FTP site.
this is what Zotero looks for:
this.addHandler("application/x-endnote-refer", _importHandler, true);
this.addHandler("application/x-research-info-systems", _importHandler, true);
// Add ISI
this.addHandler("application/x-inst-for-Scientific-info", _importHandler, true);
//
// And some non-standard ones
//
this.addHandler("text/x-research-info-systems", _importHandler, true);
// Nature uses this one
this.addHandler("text/application/x-research-info-systems", _importHandler, true);
// Cell uses this one
this.addHandler("text/ris", _importHandler, true);
// Not even trying
this.addHandler("ris", _importHandler, true);
}
(the mime types are in quotation marks - this is from xpcom/mimeTypeHandler.js)
For translator purposes, they could either include metadata in their pages - either in a standard format that Zotero can ready
http://www.zotero.org/support/dev/exposing_metadata
or at least in some other standard way so that it'd be easy to write a translator.
Or they could make the RIS more accessible - if I remember correctly, ajlyon has just thrown the towel on this one, as the RIS is not actually at a determinable URL, but but rather served by a mysterious "wizard" (I'm not kidding: "wizard exportCitation")
http://www.nanoscalereslett.com/content/6/1/529/abstract
An option to add the reference in an existing collection would be great. Or an easy way to merge large numbers of collections.
Thanks for a great tool.
For me it's often easier just to export to RIS and then drag and top that file onto Zotero standalone. If it weren't for the new collection, I'd have little to do.
Btw. on a completely different note - DWL, did you see we have been able to fix the SpringerLink translator?
*if you set aside the more general issue of Avram's and Simon's endless patience in pointing out my beginner's js errors over the last year.
Love Zotero though and have found solutions to most problems on these forums - this is only a minor annoyance. All the best.
Edit: if you do start a thread, please include as much detail as possible: software versions, URLs, etc.