TrackMeCore

Main is an experimental branch with more analysis and some BE features, that are not yet on stable Back-end functionality to collect personal data and run initial analysis this is the back-end part of the app to track different aspects of the personal life and identify some useful patterns, correlations and trends to better adjust habits, routines and such

Local setup

Local DATABASE

if you have PostgresDB running on your host machine, make sure you have trackme database created and given privileges to root:

CREATE DATABASE trackme;
CREATE USER root;
GRANT ALL PRIVELEGES ON DATABASE trackme TO root;

Adjust .env file (rename .env.example to .env or create it), use .env.example as a reference for needed values
Create a configuration file for topics and attributes to track in the app. You can use configuration_example.json as a reference or read wiki for default setup (no configuration file supplied)

Docker start

(MacOS and Linux) run make build
in case you want to supply your own structure, run make docker-migrate-own-config PATH=absolute_path_to_the_host_file
in case you want to run default structure, run make docker-migrate.default
to start the backend app run make run and navigate to localhost:5000/docs

Local start

you will need python 3.9+ and poetry installed
to setup dependencies for local run or development: make local-setup
to start the backend app run make local-run and navigate to localhost:5000/docs

Container setup

TBD

Development

within active poetry environment (after running make local-setup run poetry shell to activate python environment)

run linters: make linter
to generate openapi client for this server:

- save newest opeanpi.json to openapiclient folder
- run: make generate-client laguage=<deisred_language>

make sure you have openapi-generator installed. Preferred way of installation is a package with OS specific package manager.

Tests

!! tests are run on default data seeding !! run: make test

Motivation behind the project

This program is designed to combine several approaches from psychological research as well as ease of tracing. Several information classes are predefined for an easier and more granular analysis. Although this split can cerate an information bias, where people have to choose slightly wrong form for their thoughts. I believe, that with categories held as abstract as possible and a possibility to customize this information, the bias will be insufficient.

More documentation about ideas and default structure can be found here: wiki

Alisa-lisa / TrackMeCore