Just chiming in to say that I also rely on the ODT Scan to make Scrivener and Zotero be friends. My only option now is downgrading to Zotero 6 which is obviously not ideal. Otherwise I'm at a loss - and in the middle of a big writing project (with lots of scannable cites already included). I've been a huge fan of Zotero since the beginning, and I added Scrivener to my workflow relatively recently, only after carefully making sure it could be done (albeit awkwardly through ODT Scan, but there was a way). Thank you!
It has instructions on how to use BetterBibTex and DocDown.
Make sure to set your Scrivener compile for MultiMarkdown, and, in the compile preferences, check the box to Convert rich text to markdown.
The workflow becomes: write with quickcite in Scrivener, compile to markdown, use DocDown to convert to docx, open in Word with live Zotero citations for formatting etc.
Any updates on this? I'm another years-long Scrivener user who has depended on ODF/RTF scan for my academic writing... I've been using Zotero 6 to keep my workflow functional but am considering switching workflows to the one mentioned above to upgrade to Zotero 7.
I have tried @aschock 's workflow, but it does not seem to work for me. I still get only formatted plain text, not interactive Zotero links in my final document. Maybe I am missing something (I hope so), but automatic Bibliographies or quick switches between citation styles are impossible this way.
I have chosen to use Zotero 6 instead. You CAN have two versions of Zotero on a Mac at the same time. Download Zotero 6 from the 'other versions' dropdown, but rename the app before dragging it into your applications folder, so id does not overwrite Zotero 7.
Make sure to close Zotero 7 before opening 6, they CANNOT run at the same time. 6 will automatically use the same database/library/plugins as 7.
That also means, that ODF scan will be disabled at first. Simply remove it from your plugin list and install the .xpi again.
After extensive searching, my sense is that this post describes the best way to replace ODF/RTF functionality in Scrivener. In many ways, it's an improvement:
The one change I made is to use a "Better CSL JSON" library when exporting my library with BetterBibTex, which then requires that you slightly edit the "arguments" line so that the correct library path is provided, e.g.:
At this point, rather than wait for RTF/ODF to be updated, which seems unlikely, it seems like the best option would be to convert every "Scannable Cite" to "Better BibTeX Citation Keys."
I would do this by hand, but I literally have thousands of scannable cites in my manuscripts. Does anyone know whether a script already exists to do this? Or how hard it would be to implement?
If a script like this were available, I think we could say this issue was resolved.
I would really love to have an update on this. I have received a message several times telling me that it is "strongly recommended" to upgrade to Zotero 7, but then again have not heard anything about RTF/ODF scan becoming available again.
I am wondering: why is it "strongly recommended"? Just becuase it is "better"? What happens if I don't move on to Zotero 7? Will Zotero 6 suddenly become unusable at some point? I feel stuck. I spent some time reading about the alternative suggested above by charles petersen, and it seems so complicated to me. It just feels overwhelming.
Upgrading is recommended because Zotero won't troubleshoot issues anymore, there won't be any security updates, and, in the medium term, things will start breaking (e.g., Zotero will make changes to translators and sync that will not work with Zotero 6 breaking both import and sync). I'd guess that won't be in the next 6 months, but no guarantees, certainly not beyond that. It very likely won't suddenly stop working anymore, though especially if you're on a Mac and your system updates, Zotero also won't make sure older versions work with newer versions of OS and Word for Mac.
Thank you very much for this clarification, even if it's not reassuring.
When I look at the Scrivener forum, I see :
" There are THREE main workflows to use Zotero with Scrivener: ① the built-in RTF Scan — RTF Scan ② using BetterBibTex — Better BibTeX ③ RTF/ODF-Scan — RTF/ODF-scan Zotero 6 only "
Does this mean that this "built-in RTF Scan" still exist under Zotero 7?
I don't really understand the difference between workflows 1 and 3. If this RTF scan still exists under Zotero 7, wouldn't this be the simplest workflow?
RTF Scan still exists in Zotero 7, but it's less robust than ODF scan and doesn't convert to 'live' citations, which is why Frank originally developed ODF scan.
It works exactly the same way in Zotero 6 and 7, so you can test it out in Zotero 6 and see if its sufficient for your purposes before upgrading.
Just to clarify how nonstatic citekeys impact the workflow. If I insert a citekey in Scrivener, then later update the bibliographic info in scrivener, and then compile from scrivener--will the citekey be broken (if I don't go back in scrivener and re-insert it for each citation)? Or will it only break for the special item types you mention?
I did sometimes have citations break like this when using RTF-ODT, but hopefully will be able to get set up with Pandoc now (never used it before!).
@slorsak Yes, that's correct. Because the citekeys are formatted [@lastnameTitleDate] any changes to those info fields for any item type in Zotero will alter the citekey and break the link. I referenced legislation and archival ephemera in the guide just based off my own personal experience where I often add peculiar items incorrectly and only later determine what the proper citation format is, which often involves re-arranging those fields.
As long as you catch those on export it shouldn't be much of a hassle, though. Step 4D has instructions on adding a log file on export which will include any broken citekeys, which can then easily be corrected with find/replace in Scrivener.
Got it! I also end up correcting info in Zotero at the end of the writing process.
Looking at the BBT instructions, I'm thinking I might 'pin' those citekeys if/when I edit an item. Looks like that would prevent the need to correct the citekeys in Scrivener later...
Upgrade Storage
You can use this page!
https://github.com/lowercasename/docdown
It has instructions on how to use BetterBibTex and DocDown.
Make sure to set your Scrivener compile for MultiMarkdown, and, in the compile preferences, check the box to Convert rich text to markdown.
The workflow becomes: write with quickcite in Scrivener, compile to markdown, use DocDown to convert to docx, open in Word with live Zotero citations for formatting etc.
That sounds very promising. Thank you for the suggestion!
I have tried @aschock 's workflow, but it does not seem to work for me. I still get only formatted plain text, not interactive Zotero links in my final document. Maybe I am missing something (I hope so), but automatic Bibliographies or quick switches between citation styles are impossible this way.
I have chosen to use Zotero 6 instead. You CAN have two versions of Zotero on a Mac at the same time. Download Zotero 6 from the 'other versions' dropdown, but rename the app before dragging it into your applications folder, so id does not overwrite Zotero 7.
Make sure to close Zotero 7 before opening 6, they CANNOT run at the same time. 6 will automatically use the same database/library/plugins as 7.
That also means, that ODF scan will be disabled at first. Simply remove it from your plugin list and install the .xpi again.
Done. You can now use the scan like always.
https://forum.literatureandlatte.com/t/zotero-integration-plugin-workflow-at-least/146591/16
The one change I made is to use a "Better CSL JSON" library when exporting my library with BetterBibTex, which then requires that you slightly edit the "arguments" line so that the correct library path is provided, e.g.:
--lua-filter=zotero.lua --citeproc --output <$outputname>.docx --bibliography /Users/[username]/.local/share/pandoc/myibrary.json <$inputfile>
I would do this by hand, but I literally have thousands of scannable cites in my manuscripts. Does anyone know whether a script already exists to do this? Or how hard it would be to implement?
If a script like this were available, I think we could say this issue was resolved.
I am wondering: why is it "strongly recommended"? Just becuase it is "better"?
What happens if I don't move on to Zotero 7?
Will Zotero 6 suddenly become unusable at some point?
I feel stuck. I spent some time reading about the alternative suggested above by charles petersen, and it seems so complicated to me. It just feels overwhelming.
Upgrading is recommended because Zotero won't troubleshoot issues anymore, there won't be any security updates, and, in the medium term, things will start breaking (e.g., Zotero will make changes to translators and sync that will not work with Zotero 6 breaking both import and sync). I'd guess that won't be in the next 6 months, but no guarantees, certainly not beyond that. It very likely won't suddenly stop working anymore, though especially if you're on a Mac and your system updates, Zotero also won't make sure older versions work with newer versions of OS and Word for Mac.
When I look at the Scrivener forum, I see :
"
There are THREE main workflows to use Zotero with Scrivener:
① the built-in RTF Scan — RTF Scan
② using BetterBibTex — Better BibTeX
③ RTF/ODF-Scan — RTF/ODF-scan Zotero 6 only
"
Number 1 does not indicate that it is impossible with Zotero 7, and it links to here:
https://www.zotero.org/support/rtf_scan
Does this mean that this "built-in RTF Scan" still exist under Zotero 7?
I don't really understand the difference between workflows 1 and 3.
If this RTF scan still exists under Zotero 7, wouldn't this be the simplest workflow?
It works exactly the same way in Zotero 6 and 7, so you can test it out in Zotero 6 and see if its sufficient for your purposes before upgrading.
Just to clarify how nonstatic citekeys impact the workflow. If I insert a citekey in Scrivener, then later update the bibliographic info in scrivener, and then compile from scrivener--will the citekey be broken (if I don't go back in scrivener and re-insert it for each citation)? Or will it only break for the special item types you mention?
I did sometimes have citations break like this when using RTF-ODT, but hopefully will be able to get set up with Pandoc now (never used it before!).
As long as you catch those on export it shouldn't be much of a hassle, though. Step 4D has instructions on adding a log file on export which will include any broken citekeys, which can then easily be corrected with find/replace in Scrivener.
Got it! I also end up correcting info in Zotero at the end of the writing process.
Looking at the BBT instructions, I'm thinking I might 'pin' those citekeys if/when I edit an item. Looks like that would prevent the need to correct the citekeys in Scrivener later...
Thanks so much for writing up your process!