Input dataset files for the provenance index database

This directory contains input files used to generate provenance DB.
Each file is a CSV file with 3 columns:

  revision hash, author date,[directory]

The third columnm (root directory IDs) may not be populated.

Revisions should be sorted by commit date (needed for an efficient provenance database creation).

Provided datasets are:

| name            | N   | first date                | last date                 | description
| ese2005-begin2m | 2M  | 1970-01-01 00:00:01       | 1988-06-30 08:13:36       | first 2 million revisions of the ESE2005 dataset
| ese2005-end2m   | 2M  | 2004-09-27 20:56:03       | 2005-01-01 00:00:00       | last 2 million revisions of the ESE2005 dataset
| ese2005         | 38M | 1970-01-01 00:00:01       | 2005-01-01 00:00:00       | ESE2005 dataset: same as ese2018 but limited to revisions which date is between EPOCH and 2005-01-01
| ese2018         | 1B  | 1787-08-06 17:00:00       | 6912-01-03 15:38:42+00:00 | ESE2018 dataset; all the revisions of the Software Heritage Archive at date 2018-02-13
| sample_10k      | 10k | 2008-10-06 16:02:56+00:00 | 2016-06-22 06:13:17+00:00 | a small sample of 10k revisions, useful for testing