+1 Just went through and had to update 50+ references for a paper and having some way of automating this would have saved me hours of my time. Maybe it would be possible to select multiple entries and have an "update meta data" option under right click which then brings up a dialogue window that shows you what you have, and what Zotero pulls down and you can go through, reference by reference, and choose what to keep, as well as have a "accept all changes" button that throws a warning "This will overwrite all of the existing metadata. Do you want to continue"
Not only for pre-prints, also when I am quickly adding some references but some sources have more complete and accurate metadata than others, so being able afterwards to update and complete and/or correct the metadata for an item would be awesome.
I like the way and interface Calibre uses to achieve this for ebooks, if it's any help
+1 as well. Seems this has been a long requested feature, fingers crossed for 2021!
One of the biggest issues I have with the workaround of adding a new entry then merging, is that it often results in doubling up of the PDF files. Fine, so you delete the duplicated pdf file in Zotero but this doesn't remove the file from the Zotero storage folder and there seems to be no simple way to remove unused attachments from the zotero storage folder. Mendely does all this easily, but I am dying to ditch Elsevier!
a workflow that could be very handy: - user sees errors in the meta data for a PDF that was "auto-detected" - user enters correct information in select fields (this might include updating DOI, or perhaps entering a single correct word in Title, and so on). The point here is that the user enters a **minimal** set of distinguishing features. - user invokes command to Rebuild Metadata for Selected Reference. Zotero uses currently populated information to do a search. It also uses the data it can extract from the PDF itself. - The top 5 'hits' are shown to the user. User selects one and metadata populates with the corresponding information.
I'd love to have this feature. I switched from Mendeley and overall like Zotero much better. The lack of this feature, however, makes collaborating with others (I frequently import bib-lists from multiple users) difficult, as each entry is formatted differently. Re-fetching the data would be of massive help in making my lists consistent.
+1 I am often getting articles where the metadata is not embedded correctly in the PDF, or I have imported directly from a mobile device (where metadata import is lagging far behind still) and am not able to update these entries without significant intervention (i.e. constructing a new entry manually). If a PubMed ID or DOI is present, or even using a title-based search, it should be straightforward to query a database to offer updated metadata. I am not yet familiar enough with Zotero to write such a plug-in myself, but I hope someone out there is.
+1 For example, when a new conference is finish, the article is new published and the metadata may not be updated so timely in crossref. We can only add the pdf manually. So, a metadata update feature is very need for update and complete the article information later. More importantly, an update feature can active the library. Though we can modify the metadata manually, the method is inefficient, resulting the library can't have the timely change.
Please trust the experts (adamsmith, bwiernik) and lead developer (dstillman) here that this will be implemented as soon as it is feasible to do so. They have provided the best way to handle updates [re-download and merge] given the current status of available metadata from CrossRef, PubMed, journal publishers, etc.
This is not straightforward to accomplish in a fully automated way. I know because I am responsible for an online database and we have tried several simple and elaborate ways to accomplish this. If someone really wants me to go into great detail about why our efforts to automate this have failed, I can; but simply some publishers send updated metadata to CrossRef and the NCBI PubMed that contains nothing new. If you have hand-edited your record and you blindly accept the new metadata you risk overwriting good information. Thus, the expert recommendations to download again and merge - keeping the changes that you want to keep.
Elsevier, Sage, and Taylor and Francis publishers each will update their ahead-or-print metadata (and thus the date last updated at CrossRef and PubMed) without making meaningful changes. Some T&F journals (e.g. Journal of interpersonal violence) have a backlog of epub articles that wait four or more years before they are assigned to an issue -- the epub year will often be updated one of more times while waiting for final release. Elsevier has both ahead-of-print status and also assigned to an issue-in-progress status. However, in-progress articles have been known to move from the initially assigned volume to a subsequent volume (usually but not always keeping the assigned article number).
I could go on and on and on about the difficulties.
If overwriting is the difficulty here I would suggest a two-step update. One could:
1. Check for changes, resulting in little "!" indicating which fields are "outdated" (with a tooltip indicating the new value. 2. Force overwrite, if the user wants to.
Honestly I think that this is much safer than creating a new record and deleting the old.
You never want to delete the old article, as that will mess up earlier citations of it. Merging accomplishes the same outcome and will keep the citations working.
Coming from Mendeley (I am evaluating Zotero for my group currently), the lack of this feature might be the reason for not moving over. Otherwise, Zotero seems feasible.
Mendeley's system does this successfully only sometimes, at best. The updated metadata is often completely wrong -- even that from Elsevier journals. Often the Mendeley updated metadata differs from metadata listed with the Scopus database (I think that Mendeley and Scopus are affiliated.) I don't know if this is still (or ever was) true but there were rumors that Mendeley used crowd-sourced metadata for the update process. Mendeley is said to sometimes even change an accurate DOI to one that is incorrect!! This is support for the idea that Mendeley uses user-supplied metadata for this utility. How else can one explain replacing good information with bad.
A manuscript with indicators of ePub ahead of print when final metadata is already available is unfortunate. A manuscript with wrong metadata can be disastrous.
edit: I know this because I work with academics and their students who write reports or submit manuscripts for publication. Those who take care with their citations and use Zotero or EndNote never have manuscripts returned with revisions required for citations [or who had hired me to review their references]. Those few who used Mendeley had manuscripts returned or rejected because their citations were unverifiable [and later sought my reference advice]. Nobody in either of "my" departments still uses Mendeley at either of the universities I'm with. I had one professor tell me that she wasn't aware but it appeared that Mendeley made metadata updates automatically and unrequested.
Zotero developers say that a reliable metadata update system is being developed. See my comment from 2021 (above) to understand some of the problems that must be overcome to properly implement metadata updates.
Upgrade Storage
Not only for pre-prints, also when I am quickly adding some references but some sources have more complete and accurate metadata than others, so being able afterwards to update and complete and/or correct the metadata for an item would be awesome.
I like the way and interface Calibre uses to achieve this for ebooks, if it's any help
One of the biggest issues I have with the workaround of adding a new entry then merging, is that it often results in doubling up of the PDF files. Fine, so you delete the duplicated pdf file in Zotero but this doesn't remove the file from the Zotero storage folder and there seems to be no simple way to remove unused attachments from the zotero storage folder. Mendely does all this easily, but I am dying to ditch Elsevier!
- user sees errors in the meta data for a PDF that was "auto-detected"
- user enters correct information in select fields (this might include updating DOI, or perhaps entering a single correct word in Title, and so on). The point here is that the user enters a **minimal** set of distinguishing features.
- user invokes command to Rebuild Metadata for Selected Reference. Zotero uses currently populated information to do a search. It also uses the data it can extract from the PDF itself.
- The top 5 'hits' are shown to the user. User selects one and metadata populates with the corresponding information.
I am often getting articles where the metadata is not embedded correctly in the PDF, or I have imported directly from a mobile device (where metadata import is lagging far behind still) and am not able to update these entries without significant intervention (i.e. constructing a new entry manually). If a PubMed ID or DOI is present, or even using a title-based search, it should be straightforward to query a database to offer updated metadata.
I am not yet familiar enough with Zotero to write such a plug-in myself, but I hope someone out there is.
For example, when a new conference is finish, the article is new published and the metadata may not be updated so timely in crossref. We can only add the pdf manually. So, a metadata update feature is very need for update and complete the article information later.
More importantly, an update feature can active the library. Though we can modify the metadata manually, the method is inefficient, resulting the library can't have the timely change.
waiting for this function for years
This is not straightforward to accomplish in a fully automated way. I know because I am responsible for an online database and we have tried several simple and elaborate ways to accomplish this. If someone really wants me to go into great detail about why our efforts to automate this have failed, I can; but simply some publishers send updated metadata to CrossRef and the NCBI PubMed that contains nothing new. If you have hand-edited your record and you blindly accept the new metadata you risk overwriting good information. Thus, the expert recommendations to download again and merge - keeping the changes that you want to keep.
Elsevier, Sage, and Taylor and Francis publishers each will update their ahead-or-print metadata (and thus the date last updated at CrossRef and PubMed) without making meaningful changes. Some T&F journals (e.g. Journal of interpersonal violence) have a backlog of epub articles that wait four or more years before they are assigned to an issue -- the epub year will often be updated one of more times while waiting for final release. Elsevier has both ahead-of-print status and also assigned to an issue-in-progress status. However, in-progress articles have been known to move from the initially assigned volume to a subsequent volume (usually but not always keeping the assigned article number).
I could go on and on and on about the difficulties.
If overwriting is the difficulty here I would suggest a two-step update. One could:
1. Check for changes, resulting in little "!" indicating which fields are "outdated" (with a tooltip indicating the new value.
2. Force overwrite, if the user wants to.
Honestly I think that this is much safer than creating a new record and deleting the old.
A manuscript with indicators of ePub ahead of print when final metadata is already available is unfortunate. A manuscript with wrong metadata can be disastrous.
edit: I know this because I work with academics and their students who write reports or submit manuscripts for publication. Those who take care with their citations and use Zotero or EndNote never have manuscripts returned with revisions required for citations [or who had hired me to review their references]. Those few who used Mendeley had manuscripts returned or rejected because their citations were unverifiable [and later sought my reference advice]. Nobody in either of "my" departments still uses Mendeley at either of the universities I'm with. I had one professor tell me that she wasn't aware but it appeared that Mendeley made metadata updates automatically and unrequested.
Zotero developers say that a reliable metadata update system is being developed. See my comment from 2021 (above) to understand some of the problems that must be overcome to properly implement metadata updates.