git-csv-timesheet v0.1.6

git-csv-timesheet

Logging hours? No thanks. Let git-csv-timesheet estimate time spent and print out a csv.

Install

yarn global add git-csv-timesheet

# or with npm

npm install -g git-csv-timesheet

Basic usage

$ timesheet --authors me@example.com --since thismonth
date;hours
2020-07-01;7.5
2020-07-02;8
2020-07-03;10

$ timesheet  \
  --authors me@example.com \
  --since thismonth \
  --repositories /usr/repos/frontend,/usr/repos/backend
date;hours
2020-07-03;/usr/repos/frontend;1.4
2020-07-05;/usr/repos/backend;1.7
2020-07-06;/usr/repos/backend;2.7
2020-07-09;/usr/repos/frontend;1.2
2020-07-13;/usr/repos/frontend;3.2
2020-07-16;/usr/repos/frontend;2.4
2020-07-16;/usr/repos/backend;3.9
2020-07-27;/usr/repos/frontend;2.0
2020-07-28;/usr/repos/frontend;4.1

# Write to file
$ timesheet > report.csv

git-csv-timesheet will by default print out time spent this month in the current repository.

• For more advanced use, see Advanced usage.
• To override month, print for a week, year or specified date range, see time range options.
• To gather data from multiple repositories at once, see timesheetrc config.
• To gather data on individual tasks, see task tracking.

git-csv-timesheet guesses the time spent on individual repositories based on timestamps of git commits. Read more about how it works and configuring assumptions.

The generated output might not be accurate enough to use for billing.

How it works

The algorithm for estimating hours is quite simple. For each author in the commit history, do the following:

Go through all commits and compare the difference between them in time.

If the difference is smaller or equal then a given threshold, group the commits to a same coding session.

If the difference is bigger than a given threshold, the coding session is finished.

To compensate the first commit whose work is unknown, we add extra hours to the coding session.

Continue until we have determined all coding sessions and sum the hours made by individual authors.

The algorithm comes from @kimmobrunfeldt/git-hours and is about 30 lines of code.

Options

Advanced usage

Usage: timesheet [options]

Options:
  -V, --version                         output the version number
  -a, --authors [email@gmail.com]       Only care about commits from these
                                        emails.
  -d, --max-commit-diff [minutes]       max minutes between commits counted as
                                        one session.
                                        [default: 180]
  -f, --first-commit-add [minutes]      how many minutes first commit of
                                        session should add to total.
                                        [default: 60]
  -s, --since [date]                    Analyze data since date (including).
                                        [today|lastweek|thismonth|yyyy-mm-dd]
                                        [default: always]
  -u, --until [date]                    Analyze data until date (excluding).
                                        [today|lastweek|thismonth|yyyy-mm-dd]
                                        [default: always]
  -r, --repositories [path,other-path]  Git repositories to analyze.
                                        [default: .]
  -e, --email [emailOther=emailMain]    Group person by email.
  -m, --merge-request [false|true]      Include merge requests into
                                        calculation.
                                        [default: true]
  -i, --ignore-timesheetrc              Ignores .timesheetrc from home
                                        directory.
                                        [default: false]
  -j, --json                            Reports in JSON format.
                                        [default: false]
  -v --verbose                          Prints extra stats
                                        [default: false]
  -D --debug                            Prints debug information
                                        [default: false]
  -h, --help                            display help for command

  Examples:

  - Estimate hours of project

   $ timesheet

  - Estimate hours by me@example.com

   $ timesheet -a me@example.com

  - Estimate hours where developers commit seldom

   $ timesheet --max-commit-diff 240

  - Estimate hours in when working 5 hours before first commit of day

   $ timesheet --first-commit-add 300

  - Estimate hours work this month

   $ timesheet --since thismonth

  - Estimate hours work until 2020-01-01

   $ timesheet --until 2020-01-01

Config

.timesheetrc config

By default, the repository parameter will check the current git repository. You can also summarize multiple repositories by adding a config file to your home folder.

The config has the following structure:

{
  "repositories": [
    { "project": "Personal blog", "path": "/Users/tomfa/repos/notes" },
    { "project": "Personal blog", "path": "/Users/tomfa/repos/notes-backend" },
    "/Users/tomfa/repos/random-project",
    {
      "project": "Client 1",
      "path": "/Users/tomfa/repos/app",
      "trackTasks": true
    },
    {
      "project": "Client 1",
      "path": "/Users/tomfa/repos/backend",
      "trackTasks": true,
      "countMerges": false
    }
  ],
  "maxCommitDiffInMinutes": 120,
  "firstCommitAdditionInMinutes": 60,
  "countMerges": true,
  "authors": ["me@companymail.com"],
  "emailAliases": {
    "me@gmail.com": "me@companymail.com",
    "me@oldworkplace.com": "me@companymail.com"
  }
}

The config above will:

• track commits by author "me@companymail.com"
• count commits made by "me@gmail.com" and "me@oldworkplace.com" towards the author "me@companymail.com"
• add 60 minutes before first commits (for a day)
• "glue together" commits that are less than 2 hours between.
• count merges as your commit (TODO: #16)except for /Users/tomfa/repos/backend, where it's overriden)
• TODO: #17 count 1 repo for a "Unspecified" project (/Users/tomfa/repos/random-project)
• TODO: #17 count 2 repos each for the two projects Client 1" and Personal blog.
• TODO: #16 #10 Client 1 repos: it will split up the work into tasks specified in commits (see below.)

With the above config located at ~/.timesheetrc, the script will run from anywhere with the same output:

➜ timesheet --week
date;project;repository;hours
2020-07-27;Personal blog;@tomfa/notes;3.5
2020-07-27;Personal blog;@tomfa/notes-frontend;1
2020-07-27;Client 1;@client/dashboard;3
2020-07-28;Client 1;@client/app;8
2020-07-28;Client 1;client/backend;2
# etc

If you have a config in ~/.timesheetrc, and you do not wish to use it, pass the -i flag to ignore the config file.

Task tracking

TODO: #10 This feature is not yet implemented.

If you need to specify what you've worked on (I'm sorry), git-csv-timesheet can look for # in your commits to categorise work based on individual tasks.

> timesheet --week --config ~/.timesheetrc --tasks
date;project;repository;task;hours
2020-07-27;Personal blog;@tomfa/notes;#14;1.5
2020-07-27;Personal blog;@tomfa/notes;#13;2.5
2020-07-27;Personal blog;@tomfa/notes-frontend;#12;1
2020-07-27;Client 1;@client/dashboard;#152;2
2020-07-27;Client 1;@client/dashboard;;2
2020-07-28;Client 1;@client/app;#81;4
2020-07-28;Client 1;@client/app;#84;2
2020-07-28;Client 1;@client/app;#86;1
2020-07-28;Client 1;@client/app;;1
2020-07-28;Client 1;@client/backend;#421;2
# etc

This requires that your commits contain a task reference, with git commits ala:

Fix bug with login form

#TASK-123

The commit above would add its time to the task #TASK-123. The script includes everything after the first #, up to a space or line shift.

Commits without a task reference are added to a separate line where the task column is blank