I'm beginning to realise why I didn't do the `toString` overriding way
back when. Thankfully, now that all of our tests are in the same place,
this is actually not a terrible ordeal.
We noticed that when
- a function has more than one summary (with different charpred)
- one summary is subsumed by a subpath (or something happens around the function being extracted)
- the function is called multiple times(we needed at least three)
one of the summaries would no longer lead to flow.
We do this to remove the inconsistencies, and to be ready for a future
where type-tracking support content tracker of depth > 1.
It works because targets of loadSteps needs to be LocalSourceNodes
predicate loadStep(Node nodeFrom, LocalSourceNode nodeTo, Content content) {
(and for * patterns in match)
Since `PhaseDependentFlow` uses the following predicate, that relies on
.getScope() to be present for there to be any importTimeFlow (flow at
toplevel scope), it's important that data-flow nodes implement `.getScope`.
```
private predicate isTopLevel(Node node) { node.getScope() instanceof Module }
```
By implementing getScope, we can now rely on default implementation of
`getEnclosingCallable` in DataFlow::Node:
```
/** Gets the enclosing callable of this node. */
DataFlowCallable getEnclosingCallable() { result = getCallableScope(this.getScope()) }
```
TODO:
The member predicate `LibraryLambdaMethod::getACall` is
currently too permissive.
Ideally, we would have `libraryCallHasLambdaArg`
as in Ruby. But even a more precise
`libraryCall` predicate might be fine.
This provides variable capture in standard situations:
- nested functions
- lambdas
There are some deficiencies:
- we do not yet handle objects capturing variables.
- we do not handle variables captured via the `nonlocal` keyword.
This should be solved at the AST level, though, and then it
should "just work".
There are still inconsistencies in the case where
a `SynthesizedCaptureNode` has a comprehensions
as its enclosing callable. In this case,
`TFunction(cn.getEnclosingCallable())` is not
defined and so getEnclosingCallable does not exist
for the `CaptureNode`.
mostly removing of nodes from the graph.
One result lost:
```
check("submodule.submodule_attr", submodule.submodule_attr, "submodule_attr", globals()) #$ MISSING:prints=submodule_attr
```
