sqlalchemy-bot

**Michael Bayer ([@zzzeek](https://github.com/zzzeek)) wrote:** well there is no "correct" order if the tables have a cycle. The constraints have to be added with ALTER after the fact.

**Michael Bayer ([@zzzeek](https://github.com/zzzeek)) wrote:** see also #146

**Michael Bayer ([@zzzeek](https://github.com/zzzeek)) wrote:** We should be able to leverage [sort_tables_and_constraints](https://bitbucket.org/zzzeek/sqlalchemy/src/7eb34baf99179eec966ddd8b3607a6d8cfdfba21/lib/sqlalchemy/sql/ddl.py?at=master&fileviewer=file-view-default#ddl.py-996) in order to break out CreateTable ops vs. individual create constraint ops; however those constraints would need to be...

**Michael Bayer ([@zzzeek](https://github.com/zzzeek)) wrote:** Issue #387 was marked as a duplicate of this issue.

**tomkcook ([@tomkcook](https://github.com/tomkcook)) wrote:** Moving discussion here from #387. I've had a go at implementing a simple version of this, by just ignoring foreign keys while comparing tables (nobbling `_render_foreign_keys`) and...

**tomkcook ([@tomkcook](https://github.com/tomkcook)) wrote:** See https://bitbucket.org/tomkcook/alembic. 16 tests fail at present; I'm not at all familiar with the tests, but as far as I can tell the failing cases all work...

**tomkcook ([@tomkcook](https://github.com/tomkcook)) wrote:** I've got a vague feeling that this will do odd things if you drop a table that has foreign key constraints - it might well attempt to...

**Michael Bayer ([@zzzeek](https://github.com/zzzeek)) wrote:** the mutually-dependent constraints problem also has to be addressed on the drop side, yes. per my comment at https://bitbucket.org/zzzeek/alembic/issues/326/take-use_alter-true-other-fk-foreign-key#comment-21818371, I had envisioned the problem here being...

**Michael Bayer ([@zzzeek](https://github.com/zzzeek)) wrote:** Issue #450 was marked as a duplicate of this issue.

**Changes by Michael Bayer ([@zzzeek](https://github.com/zzzeek)):** * removed labels: **bug** * added labels: **feature**