ttscoff/doing

Can I use autotagging but for tags

krzysiek-b opened this issue · 12 comments

krzysiek-b commented

Hi,

I enter those entries each day
doing now Implementing some functionality @dev @proj-A-0
doing now Implementing some functionality @dev @proj-B-0
doing now Analysing requirements @meeting @proj-B-0
doing now Daily @meeting

So some are task related some don't. Now what I would like to see when I did
doing today --totals

--- Tag Totals ---
dev: 00:00:00
meeting: 00:00:00
dev-proj-A: 00:00:00
proj-A-0: 00:00:00
dev-proj-B: 00:00:00
proj-B-0: 00:00:00
meeting-proj-B: 00:00:00

So I can see how much time i spent on each.
Do you see any existing option that can make this happen automatically without me entering those additional tags by hand?

ttscoff commented
krzysiek-b commented

I expected it to be not implemented nor easy ;) I understand I can add those manually but and I try to do so but this seems to be error prone. What I was thinking we can do is extend autotag so I can to so

autotag:
patters:
-(\w+)-\d+:$1

Which should result in adding another tag from the regex capture group. Seems this sould be easy to implement and enables me to do what I need. It would be also nice if I could decide if I would like to add a new tag or replace the matched one.

What do you think?

ttscoff commented
krzysiek-b commented

Sure. I would like to enter (\w+)-\d+:$1 which gets for input

... @dev @projA-1
... @dev @projA-2
... @dev @projB-1

and outputs

... @dev @proja @projA-1
... @dev @proja @projA-2
... @dev @ProJB @projB-1

ttscoff commented

I don't know how useful it will be to the general public, but I added this to v1.0.35, which is now available for gem update doing. It uses a subarray under autotag: called transform:. Not extensively tested, so see how it works for you.

Example config:

autotag:
  transform:
  - (\w+)-\d+:$1

Takes the following input:

Working on @project-12

And adds the tag @project to the recorded entry.

Working on @project-12 @project

See how it goes for you.

krzysiek-b commented

Great news, Thank you!

I've experimented with it a bit and found this. What I tried to achieve is:

  • I do dev or meetings
  • I need to track how much dev or metting I do
  • I need to track how much dev or meeting I do per project

What i defined is this

autotag:
  whitelist: []
  synonyms: {}
  transform:
  - meeting\s+.(\w+)-\d+:$1
  - daily:meeting
  - (\w+)-\d+:$1

What I entered is this

doing now Speaking about X @meeting @proja-0
doing now With A and B @daily
doing now Implementing feature z @dev @projb-2
doing now Implementing feature z @dev @projb-1
doing now Implementing feature x @dev @proja-0
doing now Implementing feature y @dev @proja-1

What I got is this

Currently:
	- 2020-11-14 10:48 | Speaking about X @meeting @proja-0 @proja
	- 2020-11-14 10:34 | With A and B @daily @@
	- 2020-11-14 10:31 | Implementing feature z @dev @projb-2 @projb
	- 2020-11-14 10:28 | Implementing feature z @dev @projb-1 @projb
	- 2020-11-14 10:28 | Implementing feature x @dev @proja-0 @proja
	- 2020-11-14 10:28 | Implementing feature y @dev @proja-1 @proja

So seems 2 of 3 work as expected.

I've added another rule meeting\s+.(\w+)-\d+:meeting-$1 to see per project meeting time but seemy puting any constant text after : is not included.

Anyway having those two

  - meeting\s+.(\w+)-\d+:meeting-$1
  - meeting\s+.(\w+)-\d+:$1

when entering doing now Speaking about X @meeting @proja-0 I expect those two patterns to be matched to procude this 2020-11-14 10:48 | Speaking about X @meeting @proja-0 @meeting-proja @proja

So this makes ma life easier because I just need to say what I do (meeting or dev) and which tak it is. With those rules all other is calculated as expected.

What do you think?

krzysiek-b commented

I've created a pull request for this one too #117

I've tested it on this

text = '@mytags @jira-0 @abc'
transform = Array['simpletag', '(\w+)-\d+:$1', '(\w+)-\d+:prefix-$1', 'abc:def']
transform.each {|tag|
	if tag =~ /\S+:\S+/
		rx, r = tag.split(/:/)
		r.gsub!(/\$/,'\\')
		rx.sub!(/^@/,'')
		regex = Regexp.new('@' + rx + '\b')

		matches = text.scan(regex)
		matches.each {|m|
			puts "rx=" + rx + ", r=" + r
			new_tag = r
			if m.kind_of?(Array)
    			index = 1
    			m.each {|p|
    				puts " p=" + p + index.to_s
    				new_tag = new_tag.sub('\\' + index.to_s, p)
    				index = index + 1
    			}
    		end
			puts "=>" +new_tag
		} if matches
	end
}
ttscoff commented

Will be manually incorporating some of this, not ignoring your pull request.

krzysiek-b commented

Feature Merged.

krzysiek-b commented

Found one issue when I use pattern like this (deploy|merge)\s+@(\w+)-\d+:$1-$2 @dev @dev-$2 seems the next $2 is not replaced as execed. Instead the reteg capture value I see \2. Will have to debug this :)

ttscoff commented
krzysiek-b commented

Works perfectly now, Thank you!