Parse timelogs for upload

A script for converting basic timelogs to the formats for timetracking import.

Initially for Pomodoro Prompt to Harvest and Business Tracker

Pmodoro Prompt is extremely simplistic, and only has a description field and automatically saves the date. The time unit for each entry is half an hour.

Installation

Prerequisites

Install venv and pip- and python, too!

sudo apt install python3-venv python3-pip

This will also install python3 if it isn't already.

We don't use venv in these instructions but you can if you want to sort of sandbox this project.

Pip is needed.

Make Python 3 the default

sudo su
update-alternatives --install /usr/bin/python python /usr/bin/python3 1
exit

If you don't do the above, substitute python3 for python in the following.

Install

mkdir -p ~/Projects/agaric/python                                                                              
git clone git@gitlab.com:agaric/python/parse-timelogs-for-upload.git
cd parse-timelogs-for-upload
python -m pip install --user -r requirements.txt

Create local environments file

In a .env file, put your Harvest account ID and access token, both of which you can get at https://id.getharvest.com/

HARVEST_ACCESS_TOKEN=12345.pt.6W7wKRJEsG73NaNwBWBhv_5rQz1YkiC7_0U-OuYNnYZlMh4xP-HvmloBlrFcpJ5ZbT666HJOhNo3tXispFz4wk
HARVEST_ACCOUNT_ID=123456

Usage

python pomodoro_to_harvest.py

Background notes

To import into a timetracking system of any sophistication, we need to parse our description and

Harvest allows CSV import, with a bunch of annoying fields.

If project doesn't exist it will create a new project.

Learning from this script and continuing development

Rather than having to type out all 40 plus lines of data processing, you can also run the whole script in the interactive shell and play with it:

After typing python to get the interactive Python shell in this directory, you can do this line:

exec(open('pomodoro_to_harvest.py').read())

And now you can interact with the resulting timelog DataFrame:

timelog.query("time>30").loc[:100,["description","time","orig_desc"]].tail(50)

Or the slightly more processed tl DataFrame, for example to get the hours worked per project:

tl.groupby("project").agg({"time": "sum"})["time"]/60

And yeah you can just sort of tack on the column you want to mess with and do an operation like that!

List tasks that had no properly defined project

tl[tl.project == ""]

See which tasks we ended up with

pd.unique(harvest.Task)