Files
codeql/python/ql/lib/semmle/python/essa
Taus 8c0725e8c6 Python: Fix bad join in ESSA getInput
Before:

```
Tuple counts for Essa::EssaEdgeRefinement::getInput#dispred#f0820431#ff/2@b84afc77 after 20.3s:
873421    ~0%     {3} r1 = JOIN Essa::TEssaEdgeDefinition#24e22a14#ffff_31#join_rhs WITH Essa::TEssaEdgeDefinition#24e22a14#ffff_30#join_rhs ON FIRST 1 OUTPUT Rhs.1, Lhs.1, Lhs.0 'this'
181627951 ~0%     {3} r2 = JOIN r1 WITH Essa::EssaDefinition::getSourceVariable#dispred#f0820431#ff_10#join_rhs ON FIRST 1 OUTPUT Rhs.1 'result', Lhs.1, Lhs.2 'this'
873418    ~0%     {2} r3 = JOIN r2 WITH Essa::EssaDefinition::reachesEndOfBlock#dispred#f0820431#ff ON FIRST 2 OUTPUT Lhs.2 'this', Lhs.0 'result'
                return r3
```
It's perhaps not immediately obvious what's going on here (because of
the `...join_rhs` indirection), but basically we're joining together
`this` and `def` and their `getSourceVariable`, and only then actually
relating `this` and `def` through `reachesEndOfBlock`.

By unbinding `var`, we prevent this early join, which now encourages the
`reachesEndOfBlock` join to happen earlier:

```
Tuple counts for Essa::EssaEdgeRefinement::getInput#dispred#f0820431#ff/2@2d63e5lb after 2s
873421  ~0%     {2} r1 = SCAN Essa::TEssaEdgeDefinition#24e22a14#ffff OUTPUT In.3 'this', In.1
873421  ~0%     {3} r2 = JOIN r1 WITH Essa::TEssaEdgeDefinition#24e22a14#ffff_30#join_rhs ON FIRST 1 OUTPUT Rhs.1, Lhs.1, Lhs.0 'this'
873421  ~0%     {3} r3 = JOIN r2 WITH Definitions::SsaSourceVariable#class#486534ab#f ON FIRST 1 OUTPUT Lhs.1, Lhs.2 'this', Lhs.0
8758877 ~0%     {3} r4 = JOIN r3 WITH Essa::EssaDefinition::reachesEndOfBlock#dispred#f0820431#ff_10#join_rhs ON FIRST 1 OUTPUT Rhs.1 'result', Lhs.2, Lhs.1 'this'
873418  ~0%     {2} r5 = JOIN r4 WITH Essa::EssaDefinition::getSourceVariable#dispred#f0820431#ff ON FIRST 2 OUTPUT Lhs.2 'this', Lhs.0 'result'
                return r5
```
2022-07-18 20:21:39 +00:00
..
2022-01-19 14:29:58 +01:00
2022-07-18 20:21:39 +00:00
2022-04-22 12:55:28 +02:00