Enquiry about Unformatted and Formatted Citation One-click Transfer

A helpful suggestion for writing articles in any preferred software (e.g., a markdown editor) would be to incorporate a feature similar to Endnote's into Zotero. This feature would allow users to convert formatted citations to unformatted citations with just one click. Specifically, unformatted citations refer to plain text enclosed in {}, e.g., {Fuhrman, 2020 #104}. This functionality would greatly simplify the writing process in platforms like Bear, Obsidian, or other editors. When authors are ready to finalize their articles, they can easily copy the entire plain text into a word processor. With a simple click in Word, they can then 'Convert to formatted citation'.
  • edited February 4, 2024

    There are also various other options for plain-text writing using plugins.
  • edited February 4, 2024
    Thank you Dstillman! But I am not that sure about 'Whenever you wish to create a citation, write it in one of the following formats' in https://www.zotero.org/support/rtf_scan

    What are the rules? When I use quick copy by "Command + shift + C". It copied, e.g.,

    {Abobo et al., "The effect of Co2 willingness to communicate and speaking proficiency", 2022}

    which is too long...

    In Endnote, by easily 'command + C', the unformatted citation can be directly copied.
  • What are the rules?
    The rules are exactly what the page says the rules are.
    When I use quick copy by "Command + shift + C". It copied, e.g., […] which is too long...
    From the page:
    You can also install the RTF Scan citation style into Zotero and use Quick Copy to easily copy citations in the expected format into your document without typing.
    Similarly, the third-party ODF-Scan plugin provides a Scannable Cite translator that can be used in the same way.

    Again, though, there are various other plugins for writing in plain-text editors, including a bunch that let you use the normal Zotero integration window.
  • Thanks Dstillman. You are heart-heated. Though I have to clarify that I have used RTF Scan and the plain text citation is still very long, i.e.,

    {Chen et al., "Virtual reality in problem-based learning contexts: Effects on the problem-solving performance, vocabulary acquisition and motivation of English language learners", 2021}

    Only in this citation can it be scanned as a citation by RTF. What do you think
  • Hi, unfortunately I fear the problem is much more serious for plain text writers and will get worse with the release of Zotero 7.

    The developer of RTF/ODF Scan has made it very clear that he will not update the plug-in for version 7. Rather, he hopes that the functions will be integrated into Zotero: https://forums.zotero.org/discussion/comment/454348/#Comment_454348

    Even though the native RTF scan in Zotero is reasonable usable, it is unfortunately rudimentary and can only be used for simple documents. Is there a more robust solution in sight from Zotero with the final release of version 7?

    Thank you.
  • To be clear -- if I want to make very clear that I won't update an add-on, I will say plainly "I won't update the plugin for Z7," which I did not. What I said is that it's tough to find time to do it.

    Also, for actual plain text writing (as opposed to rich text writers like Scrivener) robust solutions for Zotero exist using the BetterBibTeX add-on, markdown, and pandoc. ODF-scans principal use case is Scrivener afaik.
  • @adamsmith: I apologise, I didn't mean to be rude and certainly didn't mean to imply anything wrong. On the contrary, I really appreciate your voluntary work, from which I and many others benefit. In my world, "hard to find time" means that the intended work will never be realised - hence my misunderstanding.

    Unfortunately, it doesn't change the fact that people like me get nervous when it comes to a robust solution for an RTF/ODF placeholder scan in Z7.

    Anyway, thanks for the tip about BetterBibTeX add-on, markdown, and pandoc. I didn't know that, I'll have a look.
  • I'm totally lost. Just tried to get BetterBibTeX add-on, markdown, and pandoc working for plain text. Total failure. I have to admit: I am a pure app user, no technical knowledge. Install app, open app, click - that's my way. I didn't even know what a terminal was until just now. pandoc - sorry, no alternative.

    I would be grateful for any advice or hint on how plain-text writing and Zotero 7 can be used together (for non-techy end users).
  • edited February 7, 2024
    A more streamlined, Endnote-like version of unformatted citations fully integrated into Zotero's standard workflows would be great (without the need for the current extra layers of user processes) ...

    - inbuilt item citation keys (like Endnote's record number), that would then allow short, totally unambiguous unformatted plain text citations (removing the need for ambiguity resolution). Those would be exposed to the user, for easy use (eg as an optional main pane column in the apps).
    - one-button conversion of unformatted citations to/from live formatted citations within the word processor plugins (where style choice already exists).
    - selection of the desired item for unformatted citation (ie plain text not field code) within the word processor plugins (for those who don't want to type them manually).

    While working with unsupported word processors is the promoted use case for the current RTF Scan, unformatted citations also help avoid other often-difficult problems (for users and support staff):
    - less error-prone circulation of draft papers amongst colleagues, that can be loaded into any word processor (because unformatted citations do not need to be converted to live/field-code citations until the penultimate version)
    - safer use of Track Changes in Word, any text copy/paste, merging documents
    - faster means of citing on platforms that are sometimes unworkably slow with conventional citations via the Zotero word processor plugins (especially with large documents).
    - no need to have Zotero open when citing in the word processor if you don't need it.

    And just faster citations all round ... for those who don't really need or want to deal with the word processor plugins often. I always worked with manually-typed unformatted citations in Endnote (it's the only thing about Endnote I miss).
  • inbuilt item citation keys

    These are coming

    (like Endnote's record number), that would then allow short, totally unambiguous unformatted plain text citations

    That's what RTF/ODF-scan does. The citations may seem long, but of

    {See | <span class="hljs-type">Smith</span>, (<span class="hljs-number">2012</span>) |<span class="hljs-type">p</span>. <span class="hljs-number">45</span> | <span class="hljs-type">for</span> an example |<span class="hljs-type">zu</span>:<span class="hljs-number">2433</span>:WQVBH98K}

    the Smith, (2012) part is for your convenience, you can put there whatever you want, including nothing. A minimal citation would be {||||zu:2433:WQVBH98K}. You can drag-and-drop these into your document.

    (removing the need for ambiguity resolution). Those would be exposed to the user, for easy use (eg as an optional main pane column in the apps).

    A plugin could easily do that. Maybe this can be added to RTF/ODF scan

  • edited February 7, 2024
    Thanks @emilianoheyns. I think both *small* and *unambiguous* are equally important for unformatted citations. And I should add 'read friendly'. If a grad student is circulating their work to their supervisors for review (who can load work with unformatted citations into any word processor, or use Track Changes in Word, or just cut/paste/move text around, all without the risk of lost live citations), those readers still need to know which references the student is citing.

    Zotero unformatted citation examples:
    {Smith et al., 2020} - current minimalist form when typing manually; but requires the extra layer of Zotero final processing on an RTF-format file in the app (away from the word processor), because the library item it refers to could be ambiguous.
    {Smith et al., "the_long_verbose_item_title", 2020} - current long-form version inserted by Zotero; which is unambiguous but fills one's writing with long citations (the sort of thing that annoys grad student supervisors no end ;) ). And Zotero still requires it to go through the ambiguity checking layer in the app anyway.
    { | Smith et al., 2020 | | |zu:5906489:K8NDNXXN} - RTF/ODF Scan/Scannable Cite standard long-form Quick Copy drag version
    { | | | |zu:5906489:K8NDNXXN} - RTF/ODF Scan minimalist version

    Endnote equivalent example:
    {Smith et al, 2020 #117} - converted to live citations by one-button click in the word processor (at least in MS Word; for others Endnote still requires RTF scan AFAIK)

    As also noted above, the future of the RTF/ODF Scan plugin is uncertain. And of course it requires LibreOffice at least for citation processing. Its developer has suggested that a more flexible Zotero-based alternative would be preferable once citekeys are available.
  • edited February 7, 2024
    To be clear - the problem with the Endnote version is that it wouldn't work with a synced library. Obviously #117 isn't going to be unique even per library let alone more generally, so Zotero always had a harder problem here.

    And the reason the ODFScan syntax is so complex/weird is that we wanted 100% reliable transformation _including any affixes and locators_ not sure how Endnote handles those, but wouldn't be simple with the basic syntax.

    I think using citekeys and pandoc citation syntax Zotero could build something that's both human-read-and-writeable and very reliable.
    And just to re-iterate that that option is already available (thanks to a significant extent to Emiliano) for anyone writing in actual plain text, be it markdown or LaTeX. In some editors, e.g. RStudio writing in either R-Markdown or Quarto -- it even integrates natively.
  • I appreciate the powerful workflows that can be implemented with additional, currently available tools. But their complexity is beyond the needs and/or abilities of most users. The Endnote style approach just "works" for a basic user. As does the current RTF Scan to some extent, albeit with extra layers that puts off some users.

    I'm not clear if I'm really missing something with "#117" not working with a synced library, as you suggest. 117 in that case is the unique record number for the item in the user's Endnote library. It is part of the item metadata and goes with the item everywhere. It never changes, nor is re-used. It can be displayed in the main item pane.

    So if Zotero were to assign an unambiguous citekey in each item's metadata, it would persist across synced devices, so would perform in the same way. It would point to the same item on all the user's devices.

    There is no real need for a citekey to be unique and unchanging beyond the library of the user using it. When we used to circulate draft word processing documents with Endnote's "#117" style unformatted citations from the original writer to colleagues, those colleagues reading/editing those documents simply knew when they saw a number like #117 that it was specific to the writer's library (just like most knew not to mess with live citations when those were used; but if they forgot or didn't know, they caused more damage than they would have with unformatted citations). If they wanted to suggest a new (unformatted) citation, they would simply type it into the document (eg "Jones et al, 2024") with any necessary supporting information for the writer to identify and later add the unformatted citation properly via their own library eg as {Jones et al, 2024 #223}. Conversely, circulating documents amongst our large group of research collaborators with *live* Endnote citations was regularly a nightmare.

    Endnote unformatted citations easily handle added prefixes or suffixes. As well as things like specifying output as in-sentence "Smith et al. (2020)" or end-of-sentence "(Smith et al., 2020)". Various library guides have examples, eg
  • I'm not clear if I'm really missing something with "#117" not working with a synced library, as you suggest. 117 in that case is the unique record number for the item in the user's Endnote library. It is part of the item metadata and goes with the item everywhere. It never changes, nor is re-used. It can be displayed in the main item pane.
    How is its uniqueness enforced? These are simple numbers -- I work on library on computer A with internet (or sync) turned off, add 100 citations, then do the same on computer B, then sync -- you're going to get a key conflict. That's why Zotero uses alphanumeric keys.

    And that doesn't even start with groups, which are an essential component of Zotero to facilitate collaboration and which Endnote doesn't properly do, hence the (imo highly unsatisfactory) set up of not actually being able to add citations collaboratively you describe. We constantly use that in our team, as, I'd assume, do many others.
  • edited February 8, 2024
    I'm not really advocating that Zotero use a simple incrementing-integer-from-1 citekey. I am advocating a reader friendly, unambiguous but suitably short, robust citekey (simpler for example than ODF Scan's default format, which was possibly designed with slightly different criteria prioritized ?). Is that not possible ? Endnote started in the 1980s on 5¼" floppy disks IIRC, so they didn't anticipate the internet. ;) It's understandable that their simple integer record number no longer covers all circumstances. But a robust read-friendly citekey would make the likelihood of duplicates due to delayed syncing very small (and any could be resolved similarly to duplicate items).

    If people really want robust citekeys for unformatted citations in Groups too then what extra barriers, if any, would make them unworkable ? I have been too scarred by our attempts at shared Endnote libraries just within our local research group years ago (nothing to do with record numbers/citekeys) to fully explore Zotero Groups. And our disparate worldwide research collaborators would never have all standardized on Zotero to make wider shared citing feasible anyway (too many hanging onto Endnote).
  • I think using citekeys and pandoc citation syntax Zotero could build something that's both human-read-and-writeable and very reliable.
    I'm just saying that Zotero, being build as a synced, collaborative tool, has a bit of a harder problem to solve here than Endnote, not that it's unsolveable.
  • I have to agree with @tim820 on all points. In my rather socio-scientific, self-proclaimed creative world, people write with all kinds of tools - sometimes they are downright proud of not using Word (or other office software, such as Libre Office, Google Docs, which in their eyes is often nothing more than a free Word).

    My university recently cancelled its campus-wide Citavi licence and asked all members to switch to Zotero. A big compliment for Zotero. Similar to Endnote, Citavi offers to insert placeholders that can later be converted into dynamic references. I'm afraid that many people here will soon realise that this is only possible in Zotero under certain difficulties.

    Now that this discussion has moved on to the usefulness of placeholders, I would like to add another use case: With the advent of AI writing tools, entirely new writing workflows have emerged. I would like to call them the copy-paste workflow. People are increasingly copying text back and forth between apps. For example, Deepl or GPT to check and adjust text snippets. Just have a look at how AI Writer, one of the pioneers of focussed writing, has responded to this trend: https://ia.net/topics/ia-writer-7. You can find this good or bad - but with literature references it is only suitable using placeholders.
Sign In or Register to comment.