Avoid attach pdf from arxiv/ prioritize other sources
188BET靠谱吗Sometimes*, when adding a new item by identifier with its publication doi, and being its pdf open to download, instead zotero downloads automatically the arxiv version pdf.
For example, it just happened with 10.1029/2021MS002575.
Is there any way to prioritize the published version or to avoid arxiv pdf download if not explicitely asked for an arxiv preprint?
188BET靠谱吗* When dealing with many items, this is a big quantity of papers, and it rises a big job of removing each of these pdf preprints, downloading the published versions (open or through sci-hub), attaching them to zotero and removing the temporary downloads.
For example, it just happened with 10.1029/2021MS002575.
Is there any way to prioritize the published version or to avoid arxiv pdf download if not explicitely asked for an arxiv preprint?
188BET靠谱吗* When dealing with many items, this is a big quantity of papers, and it rises a big job of removing each of these pdf preprints, downloading the published versions (open or through sci-hub), attaching them to zotero and removing the temporary downloads.
If it's not working somewhere you, we'd want to see aDebug ID188BET靠谱吗from Zotero for a save attempt after reloading the page.If you delete the PDF item and use Find Available PDF, it will follow the DOI and download the OA published version.
I'm actually not sure if we intended that.We'll need to think about whether Add Item by Identifier should use the same process as Find Available PDF.
It would be great prioritizing this way over Unpaywall or giving the user the option to choose the preferent method.
1) It would result in a load of the publisher page, potentially a number of subsequent requests, and potentially a first-party PDF download for every identifier pasted into Add Item by Identifier, even if there were dozens or hundreds of identifiers.Add Item by Identifier currently mostly uses APIs that are designed for automated access, but publisher pages aren't, and depending on the speed of access, it could result in people being blocked.
We do support this for Find Available PDF, but given the risk of blocking, Find Available PDF runs more slowly and tries to space out requests to given domains.If we supported this for Add Item by Identifier, file downloading for multiple items would need to operate in the same slower way.
188BET靠谱吗2) As with Find Available PDF, this would only help for people with direct (on-campus or VPN-based) access to PDFs on publisher pages — not access via web-based proxies that the Zotero Connector can use — or for OA PDFs.And the latter of course should be in Unpaywall.Your OA example is only a problem because Unpaywall doesn't yet have a download link for the published version.(I agree that this example is confusing and counterintuitive, though.)