Searching within collections by keywords, dates, and custom meta-data

I buried this question in another thread, and I realize it might be more helpful to expand it and make it a thread on how to search within collections, rather than try to glom onto another conversation.

  1. Let’s say I’ve created a new collection from the results of searching for a particular LABEL. Is it possible to further search within that collection by searching for a KEYWORD?

In other words, I have for example 500 documents that share a particular label and which are all now a part of COLLECTION A.

Within COLLECTION A, I want to search for all 36 documents that share a keyword. Is that a possible search sequence?

  1. Within that same COLLECTION A, is it possible to search for a particular custom meta-data that’s a Checkbox selection? So that within that collection I can see all the files that share a check box selection?

  2. Finally, within COLLECTION A, if 20% of the files are labeled with metadata of a specific day and 20% are labeled with metadata of a particular year, is it possible to filter them in any way by their chronology? So that I can search for all items in that collection that fall in a given day, or a given year?

Hopefully that all makes sense. Thanks in advance for anyone that can share some of their filtering and searching knowledge.

1 Like

We can’t even search within selected documents or existing search results. We’d need that before we could hope to search within a Collection. Same for your other hoped-for nested searches.

On the contrary, one of the Project Search options is “Search Binder Selection only.”

To search a collection, I merely display it, select the files within it, then select “Navigate->Reveal in Binder.” Now the collection files are selected in the Binder. I then search using Project search, using the “Search Binder Selection only” option. This technique can be used on the Search Results collection as well, enabling me to search within searches quite easily (I just tested it to be certain I’m not leading anyone wrong, here. ;))

To address the original questions in detail:

Base step: Click on Collection A in the sidebar. Then click on one of the filenames displayed in the collection. Then use cmd-A to select all the files in the collection. Finally, use Navigate->Reveal in Binder to select the collection’s files in the Binder.

Keyword search: Complete the Base Step. In the Project Search text box, type the desired keyword. From the Project Search options menu select both “Keyword” and “Search in Binder Selection only”. Result: a list of files in Search Results that are both in Collection A and have the desired keyword.

Metadata checkbox search: Complete the Base Step. From the Project Search options menu select both the name of the checkbox column and “Search in Binder Selection Only”. To find ticked items, type 1 (numeral one) in the search text box. To find unticked items, type for 0 (numeral zero). Result: a list of files in that have the checkbox either ticked or unticked, as you specified in your search string, and are in Collection A.

Date search: Complete the Base Step. From the Project Search options menu select both the name of the date column and “Search in Binder Selection Only”. Proceed to search as if the date field were a text field. Result: a list of files that have the date as you typed it, that are also in Collection A.

For any search: To search inside a project search. click on one of the files in the Search Results list. Then use cmd-A to select all the files in results list. Finally, use Navigate->Reveal in Binder to select the search result files in the Binder. Proceed to use Project search with “Search in Binder Selection only” turned on.

This is simply extraordinary. The ‘reveal in binder’! Of course! I wish I figured it out but I’m so very glad you shared it. Thank you very much! I also had no idea about the checkbox and date search sorting methods. Incredibly helpful. Thanks again.

Glad to be of service!

Wow, it does seem like I get it wrong half the time these days.

It’s too bad you can’t save the result as a dynamic, nested search.

Yes, it is. You can make the base collection dynamic, but not the subsidiary searches. A pity.

With certain searches, I’ve been able to combine them as a single search and save them. (Example: Turn on both label search and keyword search. Search for both a label and a keyword, using the “all words” option. If there’s no overlap between labels and keywords, it can work.) But it’s not an option for many searches.

I long for full conditional searches (as in a programming language) but figure it will be a cold day in Hades when they arrive, :smiley:

Silverdragon, Very cool of you to share these detailed tips, which my quick & dirty experiments confirm also work in Windows v3. :smiley:

Best,
Jim

Gentlefolk, as I said, I’m glad to be of service. :smiley: Even I didn’t know about the checkbox and date searches before I started, as that’s an experiment I ran to be able to answer the questions fully.

If there’s a wider lesson to be learned here, it’s that it seldom hurts to just try something. I keep a much-abused copy of the tutorial project around just to test things I want to do (or answer intriguing forum questions!) I’ve also experimented on a copy of my main Work In Procrastination from time to time.

Try it! Who knows, it might work!

Have you tried Obsidian? I found the app after reading about your experiences with Bear. Lots of plugins that bring added functionality… such as iframes inside the app (see the screen grab from Obsidian below), a dataview that lets you search and manipulate data, etc, etc, etc. Can also work in conjunction with Scrivener.

blacksmithgu.github.io/obsidian-dataview/#/

github.com/blacksmithgu/obsidia … /README.md

help.obsidian.md/Index

[attachment=0]obsidian iframe.png[/attachment]

Merx

Bryan Jenks has a useful guide here:

youtu.be/2234DXKbNgM?t=540

Merx

Thanks for your effort to be helpful, but I’m not interested in moving note databases again at this time. The only thing that would interest me is handwriting indexing within images, the one feature of Evernote I regret leaving behind.

Understood. I am not using Obsidian (at the moment) for notes, but using Scrivener’s “sync with external folder” feature to open my Scrivener projects in Obsidian. Finding it useful to see and manipulate my work in a different format and to use dataview to help me make connections that I have not been able to see in Scrivener. Database-centred Bear can’t sync in that way (and it doesn’t have the tools that Obsidian does anyway), but as Scrivener can sync in markdown, and as Obsidian is a markdown playground that can read any folder of markdown files, the two apps can work in tandem. Haven’t run into any issues as yet.

I have no plans at the moment to replace Scrivener as an app, but as the pace of development and the community support is so strong, I think there is a possibility that the power of Obsidian might see it take the place of Scrivener in the future, especially as my ADHD mind prefers the WYSIWYM interface offered by markdown (hidden or not), rather than the WYSIWYG leanings of RTF. Not for everyone, of course.

There are some writers already using Obsidian for their work. It is, I think, far more than a note-taking app.

In the meantime, I’m enjoying Bear’s alpha editor (Panda), with hidden markdown. Clutter-free power.

[attachment=0]panda.gif[/attachment]

All the best

Merx

Duplicate post. See below.

I too am ADHD and prefer Markdown to RTF for similar reasons. But Panda…

I’m glad it’s just a beta of the editor only. Hidden Markdown drives me almost as nuts as RTF. I look forward to the full release in Bear when I will have the option to turn the Markdown hiding off.

As for Scrivener v. Obsidian, I’m already using a Markdown Editor That Must Not Be Named (METMNBN) with Scrivener’s External Folder Sync. Actually I have several METMNBNs for different needs. In conjunction with Aeon Timeline, these serve my current purposes.

Like hiding the markdown when proofing. Good option (for me) to have.

The toggle is available in macOS Panda, under the format menu, though you probably know that already.

Get distracted by RTF / WYSIWYG / word-processor design-and-layout tools and ribbons while actually writing. Markdown gets me closer to a simple notebook or sheet of paper. Want to write; not think about layout or output.

I have used iA Writer in the past, and even TextEdit at times, with Scrivener files. Have also been looking at Typora. Think Obsidian (with some Scrivener) might be a settling point for a few years. It fits my needs and am enjoying how I am having to rethink things from the ground up.

Merx

Proof happens when I compile, so Markdown hiding is irrelevant to the purpose.

As for the rest, if you’d like to continue this discussion in the Software & Development forum, I’m game, but I’m growing increasingly uncomfortable with discussing all these editing alternatives in the Technical Support forum.

Ah, okay. Respect. Another day.

Merx