[minor] Using `in { }` rather than `in [ ]` in dot11.py

[nrathaus]

Brief description

In the latest HEAD (git) dot11.py has:
lambda pkt: (pkt.type not in {1, 3} or

While this works, it doesn't correctly function when you call fuzz(..) on dot11, while having it written like this:
lambda pkt: (pkt.type not in [1, 3] or

Doesn't manifest this issue (calling fuzz(..) failing)

Same bug exists also in this line:
ConditionalField(LEShortField("SC", 0), lambda pkt: pkt.type not in {1, 3}),

I can provide a PR - but this is a minor bug / patch that I think a full PR is a bit too much :)

Scapy version

5e13d41

Python version

3.11

Operating system

Linux

Additional environment information

No response

How to reproduce

UTS style reproduction is ok?

= Dot11 - ConditionalField conditions must not require hashing a volatile value
~ dot11

# Regression test.
#
# Dot11's 'addr2' and 'SC' fields are ConditionalFields whose condition
# tests set membership: `lambda pkt: pkt.type not in {1, 3}`. Scapy supports
# assigning a RandNum-family volatile object directly as a field's value -
# the same documented idiom as IP(ttl=RandByte()) or IP(id=RandShort()) -
# and VolatileValue.__hash__ is intentionally disabled (VolatileValue
# instances are unhashable). A set membership test needs to hash its
# operand, so as soon as 'type' is such a live volatile value, evaluating
# the condition raises TypeError, even though the exact same check written
# with a list instead of a set works fine (list membership uses __eq__,
# which VolatileValue does implement).
#
# This never surfaces through fuzz(Dot11()) alone, because Dot11 also
# defines a MultipleTypeField (FCfield): fuzz() freezes every other
# volatile value into a concrete one as soon as a packet has any
# MultipleTypeField, which incidentally hides this bug. Assigning the
# volatile value directly, as below, bypasses that and reproduces it
# directly against upstream Scapy.

pkt = Dot11(type=RandNum(0, 3))
assert expect_exception(TypeError, lambda: pkt.show(dump=True))

Actual result

No response

Expected result

No response

Related resources

No response

[nrathaus]

I used Claude to write the UTS as I don't know how to write UTS... I normally use Pytest