I am new on Zotero and I have the same problem on Zotero 6. I really see no point in copying over my documents in `zotero/storage`. If anything it makes searching for them outside of Zotero very hard.
I have deleted all of the documents in the collection, deleted the zotero database, and set my *File Handling* settings to * X Automatically take snapshots... * [] Automatically attach associated... * X Automatically retrieve metadata * [] Automatically rename attachment...
Still, when importing from online Mendeley, my .pdfs are moved to subfolders within `zotero/storage` and are renamed.
First, those settings don't have anything to do with importing from Mendeley. Unless you never want Zotero to save PDFs for you, you should leave that option enabled.
The thread you posted to was about the original Mendeley importer that imported directly from a (pre-encryption) Mendeley Desktop 1.18 database, and it was about files in Mendeley Desktop's Downloaded folder, which were equivalent to stored files in Zotero, so the behavior described there was correct. I've moved this to a new thread.
The Mendeley online library — and Mendeley Reference Manager, which is basically just a wrapper around the web library — has no concept of files on your computer, and even in the Mendeley Desktop era Mendeley has never synced local file paths. So all an online import can do is download files from Mendeley servers and save them as stored files.
If anything it makes searching for them outside of Zotero very hard.
Read through Stored Files and Linked Files for an explanation of how to easily use stored files from outside Zotero.
Files imported from Mendeley also shouldn't be renamed, though, and I'm not seeing that.
I realized it's pulling from the Mendeley online database of *partially* uploaded pdfs only a few minutes ago when I counted the files downloaded and they're 280. When actually the local files are more like 2000. So Zotero always downloads the pdfs uploaded to Mendeley into the storage folder, and that takes more of the sync time.
It still renames them, despite my options, but it's not relevant really.
The reason I opted to do a Mendeley online sync is that it's the only way to preserve the Folder structure. I know I could try the local, pre v19 method, but I don't dare, to be honest.
Exporting the Mendeley library into a .bib file DOES preserve the filepaths, but not the folders. I'm pretty sure by importing the .bib file I managed to have Zotero link (or at least attach the files).
Both things are important to me, so I'll try to do both. First import the Mendeley online database and then the .bib file to get the pdf filepaths. I will end up with a bunch of duplicates, most likely, which I hope to solve using this: https://github.com/frangoud/ZoteroDuplicatesMerger
We do strongly recommend using stored files, so make sure you've read through that section and fully understand the implications of using linked files going forward if you choose to go that route.
But yes, Zotero can only download what you've uploaded to the online library. We have no way of accessing local Mendeley data or files in current versions of Mendeley.
I acknowledge that Zotero won't be able to sync and keep the files in a place it manages and ensures cohesion. My files are 10GB, backed up by Dropbox and at this point I'm not that much in love with Zotero to shill an additional Dropbox-equivalent subscription to it.
I won't be able to download them in other devices either. That's fine by me.
I set the root files path in the settings, just for good measure.
I don't see any other unexpected downside to the process.
Importing a .bib file which was referring to a local file and asking to link the files successfuly works. The .bib files look like:
""" @inproceedings{Frenkel2012, author = {Frenkel, Jens and Kunze, G{\"{u}}nter and Fritzson, Peter}, doi = {10.3384/ecp12076433}, file = {:_path_/Textbooks/Papers/Survey of appropriate matching algorithms for large scale systems of differential algebraic equations - Frenkel, Kunze, Fritzson - 2012.pdf:pdf}, keywords = {blt,form,index reduction,index reduction and,matching,modelimark,necessary information to apply,providing all,so called perfect matching,to block lower triangular,transforming the system}, month = {nov}, pages = {433--442}, title = {{Survey of appropriate matching algorithms for large scale systems of differential algebraic equations}}, url = {http://www.ep.liu.se/ecp/article.asp?issue=076{\%}26article=45}, year = {2012} } """
However, if the file path has special characters it does not import nor link successfully:
""" @article{Unger1995, author = {Unger, J. and Kr{\"{o}}ner, A. and Marquardt, W.}, doi = {10.1016/0098-1354(94)00094-5}, file = {:_path_/Textbooks/Papers/Structural analysis of differential-algebraic equation systems—theory and applications - Unger, Kr{\"{o}}ner, Marquardt - 1995.pdf:pdf}, issn = {00981354}, journal = {Computers {\&} Chemical Engineering}, month = {aug}, number = {8}, pages = {867--882}, title = {{Structural analysis of differential-algebraic equation systems—theory and applications}}, url = {http://www.sciencedirect.com/science/article/pii/0098135494000945http://linkinghub.elsevier.com/retrieve/pii/0098135494000945}, volume = {19}, year = {1995} } """
This is what the file is called in my Ubuntu system: "Unger, Kröner, Marquardt - 1995 - Structural analysis of differential-algebraic equation systems—theory and applications.pdf"
I hope I won't have too many of these in my library.
Upgrade Storage
I really see no point in copying over my documents in `zotero/storage`. If anything it makes searching for them outside of Zotero very hard.
I have deleted all of the documents in the collection, deleted the zotero database, and set my *File Handling* settings to
* X Automatically take snapshots...
* [] Automatically attach associated...
* X Automatically retrieve metadata
* [] Automatically rename attachment...
Still, when importing from online Mendeley, my .pdfs are moved to subfolders within `zotero/storage` and are renamed.
This is my only gripe so far.
The thread you posted to was about the original Mendeley importer that imported directly from a (pre-encryption) Mendeley Desktop 1.18 database, and it was about files in Mendeley Desktop's Downloaded folder, which were equivalent to stored files in Zotero, so the behavior described there was correct. I've moved this to a new thread.
The Mendeley online library — and Mendeley Reference Manager, which is basically just a wrapper around the web library — has no concept of files on your computer, and even in the Mendeley Desktop era Mendeley has never synced local file paths. So all an online import can do is download files from Mendeley servers and save them as stored files. Read through Stored Files and Linked Files for an explanation of how to easily use stored files from outside Zotero.
Files imported from Mendeley also shouldn't be renamed, though, and I'm not seeing that.
I realized it's pulling from the Mendeley online database of *partially* uploaded pdfs only a few minutes ago when I counted the files downloaded and they're 280. When actually the local files are more like 2000. So Zotero always downloads the pdfs uploaded to Mendeley into the storage folder, and that takes more of the sync time.
It still renames them, despite my options, but it's not relevant really.
The reason I opted to do a Mendeley online sync is that it's the only way to preserve the Folder structure. I know I could try the local, pre v19 method, but I don't dare, to be honest.
Exporting the Mendeley library into a .bib file DOES preserve the filepaths, but not the folders. I'm pretty sure by importing the .bib file I managed to have Zotero link (or at least attach the files).
Both things are important to me, so I'll try to do both. First import the Mendeley online database and then the .bib file to get the pdf filepaths. I will end up with a bunch of duplicates, most likely, which I hope to solve using this: https://github.com/frangoud/ZoteroDuplicatesMerger
Wish me luck!
But yes, Zotero can only download what you've uploaded to the online library. We have no way of accessing local Mendeley data or files in current versions of Mendeley.
I acknowledge that Zotero won't be able to sync and keep the files in a place it manages and ensures cohesion. My files are 10GB, backed up by Dropbox and at this point I'm not that much in love with Zotero to shill an additional Dropbox-equivalent subscription to it.
I won't be able to download them in other devices either. That's fine by me.
I set the root files path in the settings, just for good measure.
I don't see any other unexpected downside to the process.
Importing a .bib file which was referring to a local file and asking to link the files successfuly works. The .bib files look like:
"""
@inproceedings{Frenkel2012,
author = {Frenkel, Jens and Kunze, G{\"{u}}nter and Fritzson, Peter},
doi = {10.3384/ecp12076433},
file = {:_path_/Textbooks/Papers/Survey of appropriate matching algorithms for large scale systems of differential algebraic equations - Frenkel, Kunze, Fritzson - 2012.pdf:pdf},
keywords = {blt,form,index reduction,index reduction and,matching,modelimark,necessary information to apply,providing all,so called perfect matching,to block lower triangular,transforming the system},
month = {nov},
pages = {433--442},
title = {{Survey of appropriate matching algorithms for large scale systems of differential algebraic equations}},
url = {http://www.ep.liu.se/ecp/article.asp?issue=076{\%}26article=45},
year = {2012}
}
"""
However, if the file path has special characters it does not import nor link successfully:
"""
@article{Unger1995,
author = {Unger, J. and Kr{\"{o}}ner, A. and Marquardt, W.},
doi = {10.1016/0098-1354(94)00094-5},
file = {:_path_/Textbooks/Papers/Structural analysis of differential-algebraic equation systems—theory and applications - Unger, Kr{\"{o}}ner, Marquardt - 1995.pdf:pdf},
issn = {00981354},
journal = {Computers {\&} Chemical Engineering},
month = {aug},
number = {8},
pages = {867--882},
title = {{Structural analysis of differential-algebraic equation systems—theory and applications}},
url = {http://www.sciencedirect.com/science/article/pii/0098135494000945 http://linkinghub.elsevier.com/retrieve/pii/0098135494000945},
volume = {19},
year = {1995}
}
"""
This is what the file is called in my Ubuntu system: "Unger, Kröner, Marquardt - 1995 - Structural analysis of differential-algebraic equation systems—theory and applications.pdf"
I hope I won't have too many of these in my library.