No major performance impact, more of a learning example for myself (had +3000 join order badness).
Initial tuple counts
```
Evaluated recursive predicate SqlAlchemy::SqlAlchemy::Connection::ConnectionConstruction#45e716e0@594cfx2g in 1ms on iteration 1 (delta size: 4).
Evaluated relational algebra for predicate SqlAlchemy::SqlAlchemy::Connection::ConnectionConstruction#45e716e0@594cfx2g on iteration 1 running pipeline base with tuple counts:
37793 ~0% {3} r1 = JOIN `ApiGraphs::API::Node.getACall/0#dispred#312deb92_10#join_rhs` WITH DataFlowPublic::CallCfgNode#b8ddbf81 ON FIRST 1 OUTPUT Lhs.1, Lhs.0, Rhs.1
0 ~0% {2} | JOIN WITH `SqlAlchemy::SqlAlchemy::Connection::classRef/0#565fc3ad` ON FIRST 1 OUTPUT Lhs.1, Lhs.2
30 ~0% {5} r2 = JOIN DataFlowPublic::CallCfgNode#b8ddbf81 WITH `DataFlowPublic::MethodCallNode.calls/2#dispred#1dd1e0f4#ffb` ON FIRST 1 OUTPUT Lhs.0, Lhs.1, Rhs.1, Rhs.2, _
{4} | REWRITE WITH NOT [NOT [Tmp.4 := "begin", TEST InOut.3 = Tmp.4], NOT [Tmp.4 := "connect", TEST InOut.3 = Tmp.4]] KEEPING 4
21 ~0% {3} | SCAN OUTPUT In.2, In.0, In.1
4 ~0% {2} | JOIN WITH `SqlAlchemy::SqlAlchemy::Engine::instance/0#1828baef` ON FIRST 1 OUTPUT Lhs.1, Lhs.2
4 ~0% {2} r3 = r1 UNION r2
return r3
```
which is fixed by the only_bind_out
```
Evaluated recursive predicate SqlAlchemy::SqlAlchemy::Connection::ConnectionConstruction#45e716e0@49effxtg in 0ms on iteration 1 (delta size: 0).
Evaluated relational algebra for predicate SqlAlchemy::SqlAlchemy::Connection::ConnectionConstruction#45e716e0@49effxtg on iteration 1 running pipeline base with tuple counts:
0 ~0% {1} r1 = JOIN `SqlAlchemy::SqlAlchemy::Connection::classRef/0#565fc3ad` WITH `ApiGraphs::API::Node.getACall/0#dispred#312deb92` ON FIRST 1 OUTPUT Rhs.1
0 ~0% {2} | JOIN WITH DataFlowPublic::CallCfgNode#b8ddbf81 ON FIRST 1 OUTPUT Lhs.0, Rhs.1
return r1
```
We also had this initial problem
```
Evaluated recursive predicate SqlAlchemy::SqlAlchemy::Connection::ConnectionConstruction#45e716e0@594cfx2g in 1ms on iteration 4 (delta size: 0).
Evaluated relational algebra for predicate SqlAlchemy::SqlAlchemy::Connection::ConnectionConstruction#45e716e0@594cfx2g on iteration 4 running pipeline standard with tuple counts:
48722 ~6% {2} r1 = DataFlowPublic::CallCfgNode#b8ddbf81 AND NOT SqlAlchemy::SqlAlchemy::Connection::ConnectionConstruction#45e716e0#prev(FIRST 2)
48722 ~3% {3} r2 = SCAN r1 OUTPUT In.0, _, In.1
48722 ~1% {3} | REWRITE WITH Out.1 := "connect"
16 ~0% {3} | JOIN WITH `DataFlowPublic::MethodCallNode.calls/2#dispred#1dd1e0f4#ffb_021#join_rhs` ON FIRST 2 OUTPUT Rhs.2, Lhs.0, Lhs.2
0 ~0% {2} | JOIN WITH `SqlAlchemy::SqlAlchemy::Connection::instance/0#5ed87c17#prev_delta` ON FIRST 1 OUTPUT Lhs.1, Lhs.2
48722 ~3% {3} r3 = SCAN r1 OUTPUT In.0, _, In.1
48722 ~2% {3} | REWRITE WITH Out.1 := "execution_options"
9 ~0% {3} | JOIN WITH `DataFlowPublic::MethodCallNode.calls/2#dispred#1dd1e0f4#ffb_021#join_rhs` ON FIRST 2 OUTPUT Rhs.2, Lhs.0, Lhs.2
0 ~0% {2} | JOIN WITH `SqlAlchemy::SqlAlchemy::Connection::instance/0#5ed87c17#prev_delta` ON FIRST 1 OUTPUT Lhs.1, Lhs.2
0 ~0% {2} r4 = r2 UNION r3
return r4
```
which is fixed by `connectionConstruction_helper`
```
Evaluated recursive predicate SqlAlchemy::SqlAlchemy::Connection::helper/0#62cfc178#b@4f295yef in 1ms on iteration 4 (delta size: 0).
Evaluated relational algebra for predicate SqlAlchemy::SqlAlchemy::Connection::helper/0#62cfc178#b@4f295yef on iteration 4 running pipeline standard with tuple counts:
4 ~0% {1} r1 = JOIN `SqlAlchemy::SqlAlchemy::Connection::instance/1#029b4c87#prev_delta` WITH `TypeTrackingImpl::TypeTracker::end/0#2ac2cfd4` ON FIRST 1 OUTPUT Lhs.1
16 ~0% {1} | JOIN WITH `LocalSources::Cached::hasLocalSource/2#8b3ee0ec_10#join_rhs` ON FIRST 1 OUTPUT Rhs.1
0 ~0% {3} | JOIN WITH `DataFlowPublic::MethodCallNode.calls/2#dispred#1dd1e0f4#ffb_102#join_rhs` ON FIRST 1 OUTPUT Rhs.1, Rhs.2, _
0 ~0% {2} | REWRITE WITH NOT [NOT [Tmp.2 := "connect", TEST InOut.1 = Tmp.2], NOT [Tmp.2 := "execution_options", TEST InOut.1 = Tmp.2]] KEEPING 2
0 ~0% {1} | JOIN WITH DataFlowPublic::CallCfgNode#b8ddbf81 ON FIRST 1 OUTPUT Lhs.0
0 ~0% {1} | AND NOT `SqlAlchemy::SqlAlchemy::Connection::helper/0#62cfc178#b#prev`(FIRST 1)
return r1
```
Since some sanitisers don't handle backslashes correctly, I updated the data-flow configuration to incorporate a flow state tracking whether or not backslashes have been eliminated or converted to forward slashes.
Although it might be hidden by github UI by default, it could be
interesting for a reviewer to notice the effect changes in the modeling
query has to the results in this file.
This reverts commit 62910f0cab525ca4d4901c4c27f6e6b22c3375fc.
This reverts commit 75a8197879ec47094d9b18f3dab7bcc1c1cdba28.
We don't find `kombu.serialization.pickle_load` since we respect
`__all__`. I think that was an attempt to not flood the captured
modeling with useless re-exports, but I think we've ended up doing that
anyway... we should consider to remove that restriction!
see 21d7df29c7/kombu/serialization.py (L29)
This reverts commit 1213e786519a11142746fd3a725c874181f3a42b.
By fixing a few bugs in the SubclassFinder + manually running Find.ql on the geonode DB from DCA, I found that the installed version of owslib had both: https://github.com/geopython/OWSLib/blob/0.27.2/owslib/etree.py
I fixed it in both predicates... I think we might still be able to remove
`newDirectAlias` -- but with it being better, it will allow us to better test if `newImportAlias` actually cover everything we need!
