You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The DFG edges for tuples used as variables used when iterating over a list/set/dict comprehension are incorrect. Currently, the variables/references which hold the values flow into the tuple but it should be the opposite direction. Example:
[foo(k, v) for (k, v) inx]
I'd expect a flow from x to the tuple (InitializerListExpression) and then, an indexed flow to k and v respectively and from there to the References k and v used as arguments of foo.
This issue is a bit related to #1957 but even when fixing #1957, this issue would remain.
The text was updated successfully, but these errors were encountered:
That's my goal. However, I think we should figure out if there are more places and/or a generalized way where we have to turn around the DFG edges between a node n and its children based on n.astParent and also in which cases checking a single layer of the astParent won't be sufficient.
The DFG edges for tuples used as variables used when iterating over a list/set/dict comprehension are incorrect. Currently, the variables/references which hold the values flow into the tuple but it should be the opposite direction. Example:
I'd expect a flow from
x
to the tuple (InitializerListExpression
) and then, an indexed flow tok
andv
respectively and from there to theReference
sk
andv
used as arguments offoo
.This issue is a bit related to #1957 but even when fixing #1957, this issue would remain.
The text was updated successfully, but these errors were encountered: