Search through 2500+ mind maps and curated study plans with Alfred

[nikivi]

I have published an Alfred Workflow that allows you to search various curated mind maps I have made.

 

It also has an option to just search for study guide learning mind maps that provide the best resources one can take to learn the subject in a guided manner. In future, this will cover every topic one can learn and if you want to improve and change it, you can. 

 

Here is the GitHub repository of the workflow and here is the direct download of it. I hope you enjoy it.

 

The cool thing is that if you dislike something or disagree with the way they are structured, you can propose the changes you want to make and thus everyone benefits.

Edited April 11, 2017 by nikivi

[nikivi]

I never can submit workflows on Packal.

 

Always get this error : 

 

 

 

Don't really know what is wrong.

[nikivi]

I have updated the workflow to search with scope. It's incredibly fast and the quality of mind maps is improving with each day. I have also started working on an 'alfred' study guide.

 

Would love to hear your thoughts on what would you say should be added there. The goal is to create a step by step guide for getting oneself acquainted with Alfred both for using and for creating awesome workflows.

 

Here is the download for the updated workflow. Would love to hear your thoughts on how I can improve it. 

 

Emoji meanings are outlined here.

Edited April 25, 2017 by nikivi

[nikivi]

I made a little screencast of the potential power of this workflow : http://quick.as/b1GwSneaO 

 

Alfred is amazing. I want to replicate the same and make a web search engine that does something similar.

 

If anyone wants to help make it, I would love to work together. ?

[nikivi]

I have updated the workflow and now every mind map is indexable through one list filter. All that is needed is one hotkey or keyword trigger.  

 

Here is the download.

 

@vitor I distribute my workflow via dropbox so I just have a link to the download and I replace the workflow every time I update to it. For me this is the fastest way to update it. However I am not sure if your bash script can in any way inform the user that I have issued a new workflow in this way, right? Is there anything I can do to inform users that I have issued a workflow and let them download the update?

Edited May 3, 2017 by nikivi

[nikivi]

I also updated the workflow's readme with more information on the changes (https://github.com/nikitavoloboev/alfred-knowledge-map)

Edited May 3, 2017 by nikivi

[vitor]

11 hours ago, nikivi said:

@vitor

 

You didn’t actually tag me, there. It needs to become purple for it to work. Like this → @nikivi. It’s a coincidence I opened this post and noticed that.

 

11 hours ago, nikivi said:

However I am not sure if your bash script can in any way inform the user that I have issued a new workflow in this way, right?

 

What bash script? If you’re referring to OneUpdater, you can have the download URL be whatever you like, as long as there’s a version of the info.plist online. But why are you updating it in Dropbox over and over with different links? Why not just include it in the Github repo and save yourself the headache?

[nikivi]

Yeah OneUpdater. I will try to move the workflow to the directory itself. Well updating with dropbox for me takes 2 seconds. Export workflow to Desktop, swap the workflow in ~/Dropbox/links/workflows/.. with the new workflow and this link  points at the updated version. The only problem is that user does not know there is an updated version. I will try to set it up as you say now, hope that works.  

 

I am not sure why but Alfred prompts bug out for me. Perhaps it is because I have too big of a list filter or there is something wrong in the markdown parser but when I prompt on a list filter, it just dismisses the result and doesn't present any links as it should. For example I activate the list filter with all these results that it can do : 

 

 

 

I activate on a result like this one : 

 

 

 

And nothing happens. This doesn't happen often and not for all entries but sometimes it does and it is extremely annoying. The debugger shows nothing when that happens. I don't really know what to do as I really wish to have this workflow be 100 % reliable.  

I made a screencast to show better what I mean : http://quick.as/m19vSzY5g 

 

Edited May 4, 2017 by nikivi

[nikivi]

Another odd thing is that it only starts to misbehave really badly as soon as I try to open the debugger. I did not get a single successful result when the debugger was open. I don't think there should be any correlation though since debugger doesn't affect the running of the workflows but still really strange. 

[nikivi]

I am really confused but the debugger actually does prevent running my workflow. http://quick.as/b1GwSneAJ 

 

With debugger, none of the workflow prompts work 100 % of times. Without it I get a 60/40 of 70/30 chance of it being successful. Success meaning I actually see my list of results.

 

Also certain prompts are more prone to having this 'dismiss bug' than others. Certain prompts like 'github' for example have nearly 100 % chance of not working. I don't know why.

Edited May 4, 2017 by nikivi

[deanishe]

I refer you, once again, to the stickied thread on reporting problems with workflows.

 

Quote

If you're having a problem with a workflow you're writing, post the actual, non-working workflow. Upload it somewhere (Dropbox?) and post a link.

 

Screenshots and code snippets are not enough. Without the actual workflow, we can only really guess at what may be wrong.

 

You aren't going to get any meaningful help with screenshots or animations or vague descriptions of your issue. You're wasting your time and ours.

 

How many more times do you need to be told this?

[nikivi]

13 hours ago, deanishe said:

I refer you, once again, to the stickied thread on reporting problems with workflows.

 

 

You aren't going to get any meaningful help with screenshots or animations or vague descriptions of your issue. You're wasting your time and ours.

 

How many more times do you need to be told this?

I will use this template from now on, sorry.

 

1. Here is the workflow download.

- alfred ver number: 3.3.2

- macOS ver number: 10.11.6

- not using any library (just using list filter and one script to parse markdowns made by @vitor)

- no nodejs

- no non-system programs are used

2. I have not made any changes to macOS environment as far as I know

 

As for the issue, certain list filter items don't actually show the list of items as they should and just dismiss. I have posted a screencast above to show what I mean in case you can't replicate it on your end (http://quick.as/m19vSzY5g). 

 

Another thing is that this issue happens 100 % of times when debugger is open. Not sure why that is.

 

Thank you for your help.

Edited May 6, 2017 by nikivi

[deanishe]

First of all, delete all those Script Filters. Instead of keywords, put the paths of the Markdown files in your List Filters and connect them to one Script Filter, not dozens. Like this.

 

The way you're doing it is crazy.

 

Your screencasts are far too fast to follow. You're supposed to be showing me something, not trying to break records. I also can't really figure out which List Filters you're using because you're using Hotkeys, not keywords.

 

At any rate, as best as I can tell, the workflow works just fine for me.

 

Edited May 6, 2017 by deanishe

[nikivi]

Wow this is indeed extremely simple looking now. Didn't think I could just pass the paths to one list filter. Thank you.

 

Makes updating and maintaining it easier too as it was starting to lag a bit adding new nodes every time. It also doesn't look as intimidating now.  

 

Thank you once again for taking your time to show what a proper way to do this thing is.

Edited May 6, 2017 by nikivi

[nikivi]

Can I ask what you used to generate .csv file?

[nikivi]

I have updated the workflow and now it includes all nodes from the big mind map.

 

 

It's amazing how simple this looks now and how easy it is to update and maintain. Thank you a lot, @deanishe.

 

I only have left to set up auto updating feature and it should be all well.

Edited May 6, 2017 by nikivi

[nikivi]

I am also curious, would anyone actually get any value from this? Or am I alone in thinking it is quite awesome.

 

I am also trying to build a web search engine similar to Google for these mind maps that will have a friendlier way to find the mind map you want than this workflow although this workflow is already insanely fast.

Edited May 6, 2017 by nikivi

[nikivi]

Also with this new workflow, the bug is gone completely. ?

[deanishe]

On 06/05/2017 at 8:27 PM, nikivi said:

Can I ask what you used to generate .csv file?

 

A few lines of Python in the REPL.

 

If I were you, I'd rename the folders and markdown file to match the category names, then write a script to generate JSON files that can be fed straight into Alfred via Script Filters. 

 

No need to update the List Filter that way, and no unnecessary parsing of Markdown files on every run. 

[nikivi]

56 minutes ago, deanishe said:

If I were you, I'd rename the folders and markdown file to match the category names, then write a script to generate JSON files that can be fed straight into Alfred via Script Filters. 

 

Thank you. I will try that approach.

[nikivi]

Can someone help with how to best do this one thing. I want to improve this workflow so that users can traverse search through the mind map from Alfred.

 

For example in the big mind map there is a node 'mathematics'. And in mathematics node there are various links + more mind maps (indicated with ?).

 

Let's say user searches for 'math' in Alfred and presses on it. The contents of the mind map will be shown like this : 

 

 

 

Clicking on any of the links will go to the web page. However I want to make it so that clicking on results with ? will show the contents of that mind map. So if user was to pick this result from the list : 

 

 

 

Not the web page will be opened but the contents of that 'statistics' mind map. So this will be shown : 

 

 

 

 

I can't really think of a way to do this nicely. Is there even a way to do this painlessly? 

 

Here is updated workflow that includes markdowns in 'inside' folder. Specifically 'math.md' and 'stats.md'.

 

I would greatly appreciate any help on this as I really wish I can get this working.

Edited May 8, 2017 by nikivi

[nikivi]

On second thought I can actually do something like this : 

 

 

 

Although I can't remember how to clean the query to empty after the filter and also for regex I will probably have (if does not contain 'mindnode', follow through). This should work I think.

 

But it will get really messy I feel. As certain mind maps go many levels deep. So I will need to build out this tree of filters and checks many times. Is there any other way to do this perhaps? 

[nikivi]

Yay ?

 

I got it working (both the regex and the clean of query input). 

 

However now I don't know how to do this right at all. I hope what I am doing is not crazy. 

 

Also @vitor's markdown parser doesn't parse correctly on two files. Not really sure why. 

On this file, it doesn't give all results to Alfred. And on this one, it only shows this result : 

 

 

 

Edited May 8, 2017 by nikivi

[nikivi]

I really hope there is a better way to do this. I can't complete all of the mind map in this way, it will just get too large and slow.

 

If you want to test what is already possible to traverse, try 'tooling', then 'productivity', then 'alfred'.

Edited May 8, 2017 by nikivi

[dfay]

Recursion.