Bring jrnl journals and taskwarrior tasks together in a single document.
Find a file
Marty Oehme 417536f1b5
Switching subprocess commands to recommended ones
Switched out all `subprocess.Popen` statements for `subprocess.run`
since they are the more recommended and less low-level versions.
2021-12-14 17:55:48 +01:00
.gitignore Prepare for remote upload 2021-10-31 19:20:05 +01:00
.gitlab-ci.yml Prepare for remote upload 2021-10-31 19:20:05 +01:00
jrnlwarrior.py Switching subprocess commands to recommended ones 2021-12-14 17:55:48 +01:00
LICENSE Prepare for remote upload 2021-10-31 19:20:05 +01:00
options.py Refactor options 2021-11-01 12:00:48 +01:00
README.md Add idea parsing for jrnl file 2021-11-01 10:50:49 +01:00

jrnlwarrior - taskwarrior and jrnl cross-pollination

This little script allows simple interaction between taskwarrior and a jrnl file. It parses the jrnl file and transfers accomplished and unfinished tasks to taskwarrior, as well as adding today's tasks as an entry to the jrnl file.

More specifically, it:

  • transfers completed tasks to taskwarrior by logging them and removes the lines from this file
    (marked with [x] or x at the beginning of the line)
  • transfers incomplete tasks to taskwarrior by adding them as new tasks scheduled at the date of the entry they appear under
    (marked with an empty [ ] at the beginning of their line)
  • adds an entry for today's date if none exists yet and populates it with due/overdue tasks from taskwarrior

To accomplish this it borrows a little from the todo.txt syntax --- namely the idea of (A) (B) (C) prioritization and x task done syntax (i.e. starting a line with x or [x] means it represents an accomplished task).

All three of these operations only operate on entries which are marked with a special title (todotxt by default), though this can be changed through a regex option. That means, it is also entirely possible to have other entries (with the same date) which will simply be ignored by the script.

Lastly, it transfers what I call 'ideas' to taskwarrior. These are small tidbits and notions I jot down during the day (hence, 'ideas') and marked as such by starting the line either with idea: or the taskwarrior equivalent +idea . I want them transferred to taskwarrior so they don't get lost overall, but they should be a) not bound to a specific date, even when written in a specific jrnl entry, and b) specifically marked as vague ideas in taskwarrior. For now, they will be applied the fixed tags +idea +maybe in taskwarrior and simply transferred date-less.

An example jrnl file:

[2021-10-30 10:16] todotxt
This is just a test entry. It shows how tasks in todo blocks are handled.
[ ] This is an incomplete task - it will be transferred directly to tw.
[x] This is a completed task - it will be logged as completed in tw.
[ ] (A) This is an incomplete task which will get assigned high priority in taskwarrior.
x This is also a completed task, as long as the x is the first thing on the line with a space behind.
x All four of these tasks will be transferred to tw, and their lines deleted from this file.

idea: This is an idea and it will be transferred to tw with corresponding tags, then deleted from this file.

This is not a task but a normal line and will remain in the file.
xAs will this,
[ ]As will this.

[ ] This will be transferred again, however.

[2021-10-29 10:16] Another entry
This is a regular jrnl entry, since it is missing the correct title.
By default it looks for `todotxt` titles, 
but the title can be freely set for the script (`-b` option).
jrnlwarrior will not look for incomplete or completed tasks within this block at all.
[x] This is a completed task, but it will not be transferred to taskwarrior.
[ ] Neither will this incomplete task.
+idea This idea however *will* be transferred, since ideas are looked for in *all* entries.

Usage

Point the file to your jrnl file (ideally it needs to be a single file) and set the syntax which declares a todo entry within it.

./jrnlwarrior.py -f ~/.local/share/jrnl/journal.txt -b 'todotxt'

The command above sets the script to work on a specific jrnl file and only work on entries which have exactly todotxt as their title. The settings above are also the default settings of the script.

BE AWARE that this script actively changes your jrnl file during normal operations. If you are afraid of destructive operations, you have two options: Create a backup of your jrnl file before running this script to be able to diff the two afterwards. Or, to see what the program would do without actually implementing any changes, invoke dry run mode:

./jrnlwarrior.py -n

It will spit out a list of tasks that are added to taskwarrior, as well as its own options, lines to be removed from the file and any to-do entries added to the file. It will not change your jrnl file in any way.

If you want to switch off one of the three ways this script connects the two, you can use the -L, -I, -T options which turn off logging, adding, or creating entries respectively.

./jrnlwarrior.py -DIT

The above invocation will only log completed tasks to taskwarrior. Adding new to-dos and creating today entries is turned off.

./jrnlwarrior.py -T

This will create a one-way connection to taskwarrior, by transferring both new and completed tasks and ideas to the program. However, tasks to be done today will not be transferred back to this file.

Scope

This was a fun weekend project and it shines through --- the code got increasingly spaghetti towards the end (especially entry-adding to jrnl file was one Sunday night hour), it's not packaged particularly well and there are no tests of any kind implemented.

I am happily using the script in my daily workflow, however some assumptions are made and edge cases will happen at some point. Please don't run it willy-nilly on a long-treasured jrnl file without having proper backups. Otherwise, there is no clear road-map --- mostly, I think, the code should be cleaned up and a way found that handles taskwarrior task duplication better (with uuid or description comparisons or similar means).

I will probably not be able to devote much more time to this in the foreseeable future, but if you find an issue tell me nonetheless or, even better, see if you can spot how to fix it and I will gladly merge the changes in!