commit_extractor.sql

with raw_commit_texts as (
    select "hash", encode("content", 'escape') as info from "git.objects" where type = 1
),
commits as (
    select
    	"hash",
    	(regexp_match("info", 'tree ([^\s]+)'))[1] as tree,
        (regexp_match("info", 'parent ([^\s]+)'))[1] as parent,
    	(regexp_match("info", 'author ([^\n]+)'))[1] as author,
   		(regexp_match("info", 'committer ([^\n]+)'))[1] as committer,
    	(regexp_match("info", '\n\n(.*)'))[1] as message
    from raw_commit_texts
)

select * from commits limit 50