Handled multiple database tables in dataframe operator. #325
Conversation
Codecov Report
@@ Coverage Diff @@
## main #325 +/- ##
==========================================
+ Coverage 93.08% 93.48% +0.40%
==========================================
Files 35 35
Lines 1532 1535 +3
Branches 268 271 +3
==========================================
+ Hits 1426 1435 +9
+ Misses 74 69 -5
+ Partials 32 31 -1
Continue to review full report at Codecov.
|
utkarsharma2
changed the title
There was a problem hiding this comment.
@utkarsharma2 it seems there is something strange because there is a commit that is part of this PR that is different from the commit in the main related to the 0.8.1 release(a0d0ddf).
Also, I suggest the commit message have the word Fix since it fixes a live bug
| conn_id_set = { | ||
| t.conn_id for t in self.parameters.values() if isinstance(t, Table) | ||
| } | ||
| if param_tables and len(conn_id_set) == 1: |
There was a problem hiding this comment.
@utkarsharma2 as we spoke, since this was a bug found in an Astro release, we need a regression test to make sure the same bug doesn't appear again - particularly given the refactoring we're doing.
There was a problem hiding this comment.
It should be possible to unit test _set_variables_from_first_table
There was a problem hiding this comment.
I have added a UnitTest case checking all possible conditions.
|
|
||
|
|
||
| def test__set_variables_from_first_table_with_different_db_tables_in_parameters(): | ||
| """Test _set_variables_from_first_table() when the tables passed are with same tables in parameters""" |
There was a problem hiding this comment.
"with same tables" -> "have different databases"
Closes: #323
Within method - _set_variables_from_first_table
we try to fill in all DB params from the first table passes as a param to the dataframe operator, there was a corner case not handled there.
Check to see if all tables belong to same conn_id. Otherwise, we this can go wrong for cases