https://github.com/multimeric/regulagity
Regulagity (reh-gew-la-git-ee): A tool for measuring the frequency on a git repo using a variety of single summary statistics
https://github.com/multimeric/regulagity
Last synced: 2 months ago
JSON representation
Regulagity (reh-gew-la-git-ee): A tool for measuring the frequency on a git repo using a variety of single summary statistics
- Host: GitHub
- URL: https://github.com/multimeric/regulagity
- Owner: multimeric
- License: lgpl-3.0
- Created: 2020-01-23T02:19:05.000Z (about 6 years ago)
- Default Branch: master
- Last Pushed: 2024-09-10T07:35:08.000Z (over 1 year ago)
- Last Synced: 2026-01-13T19:40:19.217Z (2 months ago)
- Language: Python
- Homepage:
- Size: 13.7 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Regulagity
## Introduction
Regulagity (reh-gew-la-git-ee) is a command-line tool for measuring activity in a git repo, ultimately producing a single number that summarises a repo.
For example, you can ask "how many commits per year does this repo have?" or "what proportion of months does this repo have new commits?". Both of these are good summaries of how *active* a repo is.
## Installation
Run:
```bash
pip install regulagity
```
## Usage
```bash
Usage: regulagity [OPTIONS] [LOCATION]
Calculates summary statistics for the git repository located at LOCATION.
LOCATION must be the path either to a local git repository or to a git
remote, e.g. `/home/michael/Programming/Regulagity/` or
`https://github.com/TMiguelT/Regulagity.git`
Options:
--period TEXT Period of time to summarise commits over. This
consists of an optional number followed by a
letter code, e.g. `2W` means two weeks, `3M`
means 3 months, `Y` means 1 year etc. For a full
reference on these string codes, refer to
http://pandas.pydata.org/pandas-
docs/stable/user_guide/timeseries.html#offset-
aliases
--stat [proportion|count] The way we summarise each commit over the time
period. `proportion` indicates that we should
calculate what proportion of the time period has
any commits (e.g. how many weeks on average have
any activity) whereas `count` indicates that we
should take the average number of commits in this
time period (e.g. how many commits are on average
made per week)
--help Show this message and exit.
```
## Examples
* What proportion of months have had any commit activity since this repo was created?
```
regulagity --period M --stat proportion https://github.com/TMiguelT/Regulagity.git
```
* What is the average commits per fortnight in this repo?
```
regulagity --period 2W --stat count /home/michael/Programming/Regulargity
```
## Motivation
There are numerous other tools that summarise git activity, for example:
*
*
*
*
However, when comparing repos, perhaps because you want to use the best supported and
maintained option, you really only want a single number you can use to easily compare
them. For this reason, `regulagity` only ever outputs a single number for a repo, not
a plot or a table.
The reason `regulagity` supports analyses like "what proportion of months have had any
activity" is because it avoids any biases you might get by asking "what is the average
number of commits" or "average lines changed". These can be biased by a repo that has
a million commits over a single month, because it will still has a high number of commits
per month on average, even if it's been abandoned since 2015. The same issue is
encountered for lines-of-code. Asking only if there has been *any* activity of a given
period allows you to better compare repos, and allows you to be certain that a high number
means a better amount of support. You can even use this same statistic to compare open
source repos to commercial products, because you can also ask "what proportion of months
have a new update to this software?".
## Roadmap
Planned features:
* Allow calculation of average lines-of-code, not just commit numbers
* Allow custom date-ranges for the query, e.g. "in the last year" rather than "for the
entire commit history"
* Allow specifying a branch, or assessing over all branches