Completion percentage is wrong when IDs don't start at 1

[JasonBarnabe]

I have a large table that gets regularly archived, so the IDs don't start at 1. This makes the percentage complete LHM reports way off.

jasons-mbp:panda jason$ rake db:migrate
SESSION STORE: active_record_store
CACHE STORE: file_store
==  AuditedChanges: migrating =================================================
-- index_exists?(:audits, [:association_id, :association_type], {:name=>"association_index"})
   -> 0.0024s
I, [2015-12-14T09:43:14.193077 #6202]  INFO -- : Starting LHM run on table=lhmn_audits
I, [2015-12-14T09:43:14.195034 #6202]  INFO -- : Starting run of class=Lhm::Migrator
I, [2015-12-14T09:43:14.296403 #6202]  INFO -- : Starting run of class=Lhm::Entangler
I, [2015-12-14T09:43:14.392986 #6202]  INFO -- : Starting run of class=Lhm::Chunker
I, [2015-12-14T09:43:14.395018 #6202]  INFO -- : Starting run of class=Lhm::Throttler::Time
89.22% (332992893/373246177) completeI, [2015-12-14T09:43:16.593463 #6202]  INFO -- : Starting run of class=Lhm::Throttler::Time
89.23% (333032893/373246177) completeI, [2015-12-14T09:43:18.900360 #6202]  INFO -- : Starting run of class=Lhm::Throttler::Time
89.24% (333072893/373246177) completeI, [2015-12-14T09:43:21.189020 #6202]  INFO -- : Starting run of class=Lhm::Throttler::Time
89.25% (333112893/373246177) completeI, [2015-12-14T09:43:23.605645 #6202]  INFO -- : Starting run of class=Lhm::Throttler::Time

I realize that the percentage complete will not be perfect if there are missing IDs, but to make it closer to perfect, perhaps check what the first ID is in the table and calculate based on that rather than assuming 1?

[skorfmann]

Yep, we have the same issue on a table where the first id is 1_000_000_000

[arthurnn]

You can write your own reporter. I guess the percentage reporter is assuming ids that started with 1. Fell free to send a PR, with a fix for this.
Thanks