This PR combines 2 unrelated things, because git wouldn't let me rebase with a dirty worktree and I didn't want to undo the changes to Makefile.NEQ while it was actually running:
Add some SQL and shell scripts to enable EQ.tsv to be loaded into a table called EQ in an SQLite3 database (bineq.sqlite). Some reports and interactive exploration is much easier to express in SQL (than with Unix pipelines using cut, etc., or Excel). The table gets an extra run_id column, to make it possible to easily compare results between different runs.
This PR combines 2 unrelated things, because git wouldn't let me rebase with a dirty worktree and I didn't want to undo the changes to
Makefile.NEQwhile it was actually running:EQ.tsvto be loaded into a table calledEQin an SQLite3 database (bineq.sqlite). Some reports and interactive exploration is much easier to express in SQL (than with Unix pipelines usingcut, etc., or Excel). The table gets an extrarun_idcolumn, to make it possible to easily compare results between different runs.make. Ideally, typingmakewould do everything at https://github.com/binaryeq/jcompile/pull/93#issuecomment-2207376698, but we're not there yet.