Merge pull request #18301 from joefarebrother/python-model-missing-builtins

Python: Add models for builtins `map`, `filter`, `zip`, and `enumerate`.
This commit is contained in:
Joe Farebrother
2025-01-20 16:39:37 +00:00
committed by GitHub
4 changed files with 366 additions and 1 deletions

View File

@@ -0,0 +1,4 @@
---
category: minorAnalysis
---
* Additional data flow models for the builtin functions `map`, `filter`, `zip`, and `enumerate` have been added.

View File

@@ -4523,6 +4523,124 @@ module StdlibPrivate {
}
}
/** A flow summary for `map`. */
class MapSummary extends SummarizedCallable {
MapSummary() { this = "builtins.map" }
override DataFlow::CallCfgNode getACall() { result = API::builtin("map").getACall() }
override DataFlow::ArgumentNode getACallback() {
result = API::builtin("map").getAValueReachableFromSource()
}
override predicate propagatesFlow(string input, string output, boolean preservesValue) {
exists(int i | exists(any(Call c).getArg(i)) |
(
input = "Argument[" + (i + 1).toString() + "].ListElement"
or
input = "Argument[" + (i + 1).toString() + "].SetElement"
or
// We reduce generality slightly by not tracking tuple contents on list arguments beyond the first, for performance.
// TODO: Once we have TupleElementAny, this generality can be increased.
i = 0 and
exists(DataFlow::TupleElementContent tc, int j | j = tc.getIndex() |
input = "Argument[1].TupleElement[" + j.toString() + "]"
)
// TODO: Once we have DictKeyContent, we need to transform that into ListElementContent
) and
output = "Argument[0].Parameter[" + i.toString() + "]" and
preservesValue = true
)
or
input = "Argument[0].ReturnValue" and
output = "ReturnValue.ListElement" and
preservesValue = true
}
}
/** A flow summary for `filter`. */
class FilterSummary extends SummarizedCallable {
FilterSummary() { this = "builtins.filter" }
override DataFlow::CallCfgNode getACall() { result = API::builtin("filter").getACall() }
override DataFlow::ArgumentNode getACallback() {
result = API::builtin("filter").getAValueReachableFromSource()
}
override predicate propagatesFlow(string input, string output, boolean preservesValue) {
(
input = "Argument[1].ListElement"
or
input = "Argument[1].SetElement"
or
exists(DataFlow::TupleElementContent tc, int i | i = tc.getIndex() |
input = "Argument[1].TupleElement[" + i.toString() + "]"
)
// TODO: Once we have DictKeyContent, we need to transform that into ListElementContent
) and
(output = "Argument[0].Parameter[0]" or output = "ReturnValue.ListElement") and
preservesValue = true
}
}
/**A summary for `enumerate`. */
class EnumerateSummary extends SummarizedCallable {
EnumerateSummary() { this = "builtins.enumerate" }
override DataFlow::CallCfgNode getACall() { result = API::builtin("enumerate").getACall() }
override DataFlow::ArgumentNode getACallback() {
result = API::builtin("enumerate").getAValueReachableFromSource()
}
override predicate propagatesFlow(string input, string output, boolean preservesValue) {
(
input = "Argument[0].ListElement"
or
input = "Argument[0].SetElement"
or
exists(DataFlow::TupleElementContent tc, int i | i = tc.getIndex() |
input = "Argument[0].TupleElement[" + i.toString() + "]"
)
// TODO: Once we have DictKeyContent, we need to transform that into ListElementContent
) and
output = "ReturnValue.ListElement.TupleElement[1]" and
preservesValue = true
}
}
/** A flow summary for `zip`. */
class ZipSummary extends SummarizedCallable {
ZipSummary() { this = "builtins.zip" }
override DataFlow::CallCfgNode getACall() { result = API::builtin("zip").getACall() }
override DataFlow::ArgumentNode getACallback() {
result = API::builtin("zip").getAValueReachableFromSource()
}
override predicate propagatesFlow(string input, string output, boolean preservesValue) {
exists(int i | exists(any(Call c).getArg(i)) |
(
input = "Argument[" + i.toString() + "].ListElement"
or
input = "Argument[" + i.toString() + "].SetElement"
or
// We reduce generality slightly by not tracking tuple contents on arguments beyond the first two, for performance.
// TODO: Once we have TupleElementAny, this generality can be increased.
i in [0 .. 1] and
exists(DataFlow::TupleElementContent tc, int j | j = tc.getIndex() |
input = "Argument[" + i.toString() + "].TupleElement[" + j.toString() + "]"
)
// TODO: Once we have DictKeyContent, we need to transform that into ListElementContent
) and
output = "ReturnValue.ListElement.TupleElement[" + i.toString() + "]" and
preservesValue = true
)
}
}
// ---------------------------------------------------------------------------
// Flow summaries for container methods
// ---------------------------------------------------------------------------

View File

@@ -366,3 +366,246 @@ def test_next_dict():
i = iter(d)
n = next(i)
SINK(n) #$ MISSING: flow="SOURCE, l:-3 -> n"
### map
@expects(4)
def test_map_list():
l1 = [SOURCE]
l2 = [NONSOURCE]
def f(p1,p2):
SINK(p1) #$ flow="SOURCE, l:-4 -> p1"
SINK_F(p2)
return p1,p2
rl = list(map(f, l1, l2))
SINK(rl[0][0]) #$ flow="SOURCE, l:-10 -> rl[0][0]"
SINK_F(rl[0][1])
@expects(4)
def test_map_set():
s1 = {SOURCE}
s2 = {NONSOURCE}
def f(p1,p2):
SINK(p1) #$ flow="SOURCE, l:-4 -> p1"
SINK_F(p2)
return p1,p2
rl = list(map(f, s1, s2))
SINK(rl[0][0]) #$ flow="SOURCE, l:-10 -> rl[0][0]"
SINK_F(rl[0][1])
@expects(4)
def test_map_tuple():
t1 = (SOURCE,)
t2 = (NONSOURCE,)
def f(p1,p2):
SINK(p1) #$ flow="SOURCE, l:-4 -> p1"
SINK_F(p2)
return p1,p2
rl = list(map(f, t1, t2))
SINK(rl[0][0]) #$ flow="SOURCE, l:-10 -> rl[0][0]"
SINK_F(rl[0][1])
@expects(4)
def test_map_dict():
d1 = {SOURCE: "v1"}
d2 = {NONSOURCE: "v2"}
def f(p1,p2):
SINK(p1) #$ MISSING: flow="SOURCE, l:-4 -> p1"
SINK_F(p2)
return p1,p2
rl = list(map(f, d1, d2))
SINK(rl[0][0]) #$ MISSING: flow="SOURCE, l:-10 -> rl[0][0]"
SINK_F(rl[0][1])
@expects(4)
def test_map_multi_list():
l1 = [SOURCE]
l2 = [SOURCE]
def f(p1,p2):
SINK(p1) #$ flow="SOURCE, l:-4 -> p1"
SINK(p2) #$ flow="SOURCE, l:-4 -> p2"
return p1,p2
rl = list(map(f, l1, l2))
SINK(rl[0][0]) #$ flow="SOURCE, l:-9 -> rl[0][0]"
SINK(rl[0][1]) #$ flow="SOURCE, l:-9 -> rl[0][1]"
@expects(4)
def test_map_multi_tuple():
l1 = (SOURCE,)
l2 = (SOURCE,)
def f(p1,p2):
SINK(p1) #$ flow="SOURCE, l:-4 -> p1"
SINK(p2) #$ MISSING: flow="SOURCE, l:-4 -> p2" # Tuples are not tracked beyond the first list argument for performance.
return p1,p2
rl = list(map(f, l1, l2))
SINK(rl[0][0]) #$ flow="SOURCE, l:-9 -> rl[0][0]"
SINK(rl[0][1]) #$ MISSING: flow="SOURCE, l:-9 -> rl[0][1]"
### filter
@expects(2)
def test_filter_list():
l = [SOURCE]
def f(p):
SINK(p) #$ flow="SOURCE, l:-3 -> p"
return True
rl = list(filter(f,l))
SINK(rl[0]) #$ flow="SOURCE, l:-7 -> rl[0]"
@expects(2)
def test_filter_set():
s = {SOURCE}
def f(p):
SINK(p) #$ flow="SOURCE, l:-3 -> p"
return True
rl = list(filter(f,s))
SINK(rl[0]) #$ flow="SOURCE, l:-7 -> rl[0]"
@expects(2)
def test_filter_tuple():
t = (SOURCE,)
def f(p):
SINK(p) #$ flow="SOURCE, l:-3 -> p"
return True
rl = list(filter(f,t))
SINK(rl[0]) #$ flow="SOURCE, l:-7 -> rl[0]"
@expects(2)
def test_filter_dict():
d = {SOURCE: "v"}
def f(p):
SINK(p) #$ MISSING: flow="SOURCE, l:-3 -> p"
return True
rl = list(filter(f,d))
SINK(rl[0]) #$ MISSING: flow="SOURCE, l:-7 -> rl[0]"
@expects(1)
def test_enumerate_list():
l = [SOURCE]
e = list(enumerate(l))
SINK(e[0][1]) #$ flow="SOURCE, l:-4 -> e[0][1]"
@expects(1)
def test_enumerate_set():
s = {SOURCE}
e = list(enumerate(s))
SINK(e[0][1]) #$ flow="SOURCE, l:-4 -> e[0][1]"
@expects(1)
def test_enumerate_tuple():
t = (SOURCE,)
e = list(enumerate(t))
SINK(e[0][1]) #$ flow="SOURCE, l:-4 -> e[0][1]"
@expects(2)
def test_enumerate_list_for():
l = [SOURCE]
for i, x in enumerate(l):
SINK(x) #$ flow="SOURCE, l:-3 -> x"
for t in enumerate(l):
SINK(t[1]) #$ flow="SOURCE, l:-6 -> t[1]"
@expects(1)
def test_enumerate_dict():
d = {SOURCE:"v"}
e = list(enumerate(d))
SINK(e[0][1]) # $ MISSING: flow="SOURCE, l:-4 -> e[0][1]"
@expects(8)
def test_zip_list():
l1 = [SOURCE, SOURCE]
l2 = [SOURCE, NONSOURCE]
l3 = [NONSOURCE, SOURCE]
l4 = [NONSOURCE, NONSOURCE]
z = list(zip(l1,l2,l3,l4))
SINK(z[0][0]) #$ flow="SOURCE, l:-7 -> z[0][0]"
SINK(z[0][1]) #$ flow="SOURCE, l:-7 -> z[0][1]"
SINK_F(z[0][2]) #$ SPURIOUS: flow="SOURCE, l:-7 -> z[0][2]"
SINK_F(z[0][3])
SINK(z[1][0]) #$ flow="SOURCE, l:-11 -> z[1][0]"
SINK_F(z[1][1]) #$ SPURIOUS: flow="SOURCE, l:-11 -> z[1][1]"
SINK(z[1][2]) #$ flow="SOURCE, l:-11 -> z[1][2]"
SINK_F(z[1][3])
@expects(4)
def test_zip_set():
s1 = {SOURCE}
s2 = {NONSOURCE}
s3 = {SOURCE}
s4 = {NONSOURCE}
z = list(zip(s1,s2,s3,s4))
SINK(z[0][0]) #$ flow="SOURCE, l:-7 -> z[0][0]"
SINK_F(z[0][1])
SINK(z[0][2]) #$ flow="SOURCE, l:-7 -> z[0][2]"
SINK_F(z[0][3])
@expects(8)
def test_zip_tuple():
t1 = (SOURCE, SOURCE)
t2 = (SOURCE, NONSOURCE)
t3 = (NONSOURCE, SOURCE)
t4 = (NONSOURCE, NONSOURCE)
z = list(zip(t1,t2,t3,t4))
SINK(z[0][0]) #$ flow="SOURCE, l:-7 -> z[0][0]"
SINK(z[0][1]) #$ flow="SOURCE, l:-7 -> z[0][1]"
SINK_F(z[0][2])
SINK_F(z[0][3])
SINK(z[1][0]) #$ flow="SOURCE, l:-11 -> z[1][0]"
SINK_F(z[1][1]) #$ SPURIOUS: flow="SOURCE, l:-11 -> z[1][1]"
SINK(z[1][2]) #$ MISSING: flow="SOURCE, l:-11 -> z[1][2]" # Tuple contents are not tracked beyond the first two arguments for performance.
SINK_F(z[1][3])
@expects(4)
def test_zip_dict():
d1 = {SOURCE: "v"}
d2 = {NONSOURCE: "v"}
d3 = {SOURCE: "v"}
d4 = {NONSOURCE: "v"}
z = list(zip(d1,d2,d3,d4))
SINK(z[0][0]) #$ MISSING: flow="SOURCE, l:-7 -> z[0][0]"
SINK_F(z[0][1])
SINK(z[0][2]) #$ MISSING: flow="SOURCE, l:-7 -> z[0][2]"
SINK_F(z[0][3])

View File

@@ -45,4 +45,4 @@ def test_library_call():
for x in map(set, [1]):
pass
SINK(captured["x"]) #$ MISSING: captured
SINK(captured["x"]) #$ captured