cregit-mirror/persons

Unifier of identities of committers and authors of a git repo

• How to run
• SQLite
  • Schema
• License
• Dependencies
• TODO

This program extracts all the committers and authors of commits of a git repo and unifies their identities.

The output of this program are both a spreedsheet and a table (in a sql database).

How to run

java authors.jar <repodir> <spreadsheet-file> <sqlite-file>

• Warning: it will overwrite both files

SQLite

Schema

The table is called authors and has the following schema:

CREATE TABLE authors(
    recordid int PRIMARY KEY, 
    author text, 
    personid text, 
    email text, 
    name text, 
    lcemail text, 
    userid text, 
    domain tex, 
    autcount int,
    comcount int,
    dateadded text, 
    checked text, 
    notes text);

Explanation of the fields.

name	explanation	notes
recordid	An integer, unique across all records
author	full email as used in git	it should be renamed
personid	unique identifier of this record	this id is unique per individual
email	email parl of the full email
name	name part of the full email
lcemail	lowercase version of the email	useful for reviewing of records
userid	lowercase user part of the email
domain	lowercase domain part of the email
autcount	number of commits this record authored
comcount	number of commits this record committed
dateadded	date in which the record was added to the db
checked	true if this record has been manually checked
notes

License

• GPL-3+

Dependencies

jgit	https://eclipse.org/jgit/	BSD3
poi scala	https://github.com/folone/poi.scala	Apache-2.0
apache poi	https://poi.apache.org/	Apache-2.0
slick	http://slick.lightbend.com/	MIT
sqlite-jdbc		Apache-2.0
HikariCP		Apache-2.0

TODO

The following are issues that need to be addressed.

• It need to take into consideration the .mailmap file
• It needs to Support incrementally adding authors to the database. The idea behind is that once names have been vetted, those names will not be matched again.
• Ask if the files should be overwritten