Improve internal linking search

🌟Features
#1

I’m importing a number of outlines for speaking notes into Dynalist, and linking them up with various metadata (I put everything into a single document, Workflowy-style). My speaking notes follow this general format:

  • Title
    • Introduction
      • Something interesting
      • Something else interesting
    • Point 1
    • etc.

Naturally, I want to link to the titles. But the title is often comprised mainly of words that occur frequently within the document. When I go to add a link, I type [[ followed by the title. The trouble is, I often have to dig through the entire search results to find what I want because the top results will be items which merely contain all the words I typed.

I think the [[ search should first return exact substring matches, and then after that return the other matches. Thus, if I’m trying to link to a speech entitled “A Speech with a Common Title” and I type [[a speech with a, I should first get a search based on "a speech with a" before a search for a speech with a. This would ensure that, in nearly every case, the thing I want to link to would be the top result.

1 Like
#2

That’s strange, sounds like a bug to me. I’ll test it out a bit.

cc @Erica

#3

If you just enter a plain search term, I believe exact matches are weighed more but not that more. If you want to search for exact match, try [["a speech with a". It’s the same syntax as described in the searchbox tips.

#4

Thanks, Erica. You might want to double-check the matching weight, because it appears to me that exact matches aren’t weighed more at all. There have been many instances where I’ve seen the top several matches having the words scattered randomly throughout and had to scroll down quite some distance to find the first match which even had the words in the same order.

I can use quotes, but I still think it would be more intuitive to weigh exact matches higher, which doesn’t appear to me to be happening.

2 Likes
#5

Cool, I’ll check with @Shida to see exactly how the weighing works and how we can bump it to make the entire searching algorithm more intuitive.

1 Like
#6

i do not understand what can be made better,

too many words:

  • a gif demo,

  • or video would be 1000x better