taskwarrior-data: A Shell repository from linuxcaffe

Taskwarrior Data

This repository is a template project for hosting taskwarrior data with git-sync script to deploy it on all of my working systems.

File Structure

➜  taskwarrior-data git:(master) tree .
├── README.md                            # current file
├── data                                 # taskwarrior data
│   ├── backlog.data
│   ├── completed.data
│   ├── hooks -> ../hooks
│   ├── pending.data
│   └── undo.data
├── git-sync                             # core script from `git-sync`
├── git-sync-on-inotify                  # core script from `git-sync`
└── hooks                                # taskwarrior hooks
    ├── on-exit-sync
    └── on-launch-sync

3 directories, 10 files

Deploy

Clone the current repository to local system.
```
git clone https://github.com/xingheng/taskwarrior-data.git --branch dev
```
The default master branch has some sample task for demo, so specify the dev branch to fetch a clean environment for personal usage. Recommend keeping the dev branch for upgrade in the future.
Now create a git repository code hosting service like GitHub or Bitbucket for personal data hosting, make it public or private, it’s up to you, the git-sync tool will use your local git configuration only.

Add it as a new upstream in current repository, in this example we use master branch for data sync.
```
git remote add mine git@github.com:your-name/taskwarrior-data.git
git fetch mine
git checkout --track mine/master
```
Configure it for git-sync.
```
git config --bool branch.master.sync true
git config --bool branch.master.syncNewFiles true
```
It doesn’t limit you have to use master branch for data sync, just make sure config it right and check it out once everything ready.
Configure the data storage for taskwarrior.

taskwarrior uses the ~/.task directory as default data storage as its docs said, so you need rewrite the TASKDATA environment variable to new path in ~/.taskrc file or move the current taskwarrior-data directory to the default path, migrate the existing data to taskwarrior-data/data directory if necessary.

Event Trigger

When task command started, it will trigger the on-launch-sync hook script to check current git repository’s branch configuration and file status, break the task launch progress to run if something wrong or dirty found.

When task command finished, it will trigger the on-exit-sync hook script to update the data file status, commit the task context info with a new custom message and call the external git-sync script to remote master branch’s upstream, of course save all the related operation outputs to log file git-sync.log.

Multiple Workspaces

I use taskwarrior with two workspaces, one for personal projects which track all the general task data including side projects, another one for team projects because of security data issues, so I had to create a private git repository in the our private code hosting server.

Luckily taskwarrior uses its environment variables TASKRC and TASKDATA to initialize the configuration and data storage path every time when launching task. So I use direnv to separate the workspace, here is my configuration in ~/work/team-name/.envrc:

# taskwarrior
# ------------------- BEGIN -------------------
# export TASKRC=~/.taskrc  # This is default rc.
export TASKDATA=~/work/team-name/projects/taskwarrior-data/data
# -------------------- END --------------------

That’s nice! When I work from the any subdirectories under the ~/work/team-name directory, task takes me to the team task workspace instead of global personal one. Enjoy!

References

Automated Syncing With Git

Taskwarrior - Hooks v1

Taskwarrior - Hooks v2

Taskwarrior - Hook Author's Guide

linuxcaffe/taskwarrior-data