First fix handles the case where there's interference from a class-based
decorator on a function. In this case, _technically_ we have an instance
of the decorator class, but in practice this decorator will (hopefully)
forward all accesses to the thing it wraps.
The second fix has to do with methods that are added dynamically using
`setattr`. In this case, we cannot be sure that the relevant methods are
actually missing.
The relative simplicity of the MRO computation in `DataFlowDispatch` was
causing quite a lot of FPs to appear for the "wrong number/name of
arguments in class instantiation" queries. The easiest fix was to just
exclude cases where we know the MRO computation will be imprecise --
when there are superclasses with multiple inheritance or unknown bases.
These results are missing because we no longer (unlike the points-to
analysis) track how many elements a given tuple has. This is something
we might want to implement in the future (most likely through an
`int`-indexed type tracker).
This one is a bit more involved. Of note is the fact that it at present
only uses local flow when determining the origin of some value (whereas
the points-to version used global flow). It may be desirable to rewrite
this query to use global data-flow, but this should be done with some
care (as using "all unhashable objects" as the set of sources is
somewhat iffy with respect to performance). For that reason, I'm
sticking to mostly local flow (except for well behaved things like types
and built-ins).
For this one we actually lose a test result. However, this is kind of to
be expected since we no longer have the "precise" MRO that the points-to
analysis computes.
Honestly, I'm on the fence about even keeping this query at all. It
seems like it might be superfluous in a world with good Python type
checking.
These could arguably be moved to `Class` itself, but for now I'm
choosing to limit the changes to the `DuckTyping` module (until we
decide on a proper API).
Same comment as for the preceding commit. We lose one test result due to
the fact that we don't know what to do about `for ... in 1` (because `1`
is an instance of a built-in). I'm going to defer addressing this until
we get some modelling of built-in types.
Uses the new `DuckTyping` module to handle recognising whether a class
is a container or not. Only trivial test changes (one version uses
"class", the other "Class").
Note that the ported query has no understanding of built-in classes. At
some point we'll likely want to replace `hasUnresolvedBase` (which will
hold for any class that extends a built-in) with something that's aware
of the built-in classes.
This module (which for convenience currently resides inside
`DataFlowDispatch`, but this may change later) contains convenience
predicates for bridging the gap between the data-flow layer and the old
points-to analysis.
Removes the use of points-to for accessing various built-ins from three
of the queries. In order for this to work I had to extend the lists of
known built-ins slightly.
This join had badness 1127 on the project FiacreT/M-moire, producing ~31
million tuples in order to end up with only ~27k tuples later in the
pipeline. With the fix, we reduce this by roughly the full 31 million
(the new materialised helper predicate accounting for roughly 130k
tuples on its own).
Co-authored-by: Mathias Vorreiter Pedersen <mathiasvp@github.com>
Adds `hasOverloadDecorator` as a predicate on functions. It looks for
decorators called `overload` or `something.overload` (usually
`typing.overload` or `t.overload`). These are then filtered out in the
predicates that (approximate) resolving methods according to the MRO.
As the test introduced in the previous commit shows, this removes the
spurious resolutions we had before.
