utils: restore TLS settings when a pcapng reader closes

[KernelClint]

When a pcapng capture carries a Decryption Secrets Block, Scapy loads the keys and switches on TLS
session processing so the encrypted traffic in that file can be read
(scapy/utils.py:1994-2016).
Both are process-global settings on conf.

Closing the reader restores neither. rdpcap() closes it (scapy/utils.py:1303-1314), but the
inherited close only closes the file. The keys and the enable flag outlive the capture that supplied
them, so a later unrelated capture is read with the first file's keys still installed. A
key-carrying capture with no packets in it, read first, was enough to make a second encrypted
capture decrypt.

The change remembers the previous values the first time a reader installs keys, and puts them back
when it closes:

+        self._tls_state = None  # type: Optional[Tuple[Dict[str, bytes], bool]]
...
+                    if self._tls_state is None:
+                        self._tls_state = (
+                            conf.tls_nss_keys,
+                            conf.tls_session_enable,
+                        )
...
+    def close(self):
+        # type: () -> None
+        if self._tls_state is not None:
+            conf.tls_nss_keys, conf.tls_session_enable = self._tls_state
+            self._tls_state = None
+        RawPcapReader.close(self)

Settings the operator configured themselves are preserved: only what this reader changed is undone.
Decryption within the capture that carries the keys is unaffected.

The added regression reads a key-carrying capture, closes it, and asserts conf.tls_nss_keys and
conf.tls_session_enable are back to what they were. Without the source change it fails.

Performance was measured on one computer, before and after the fix: reading a capture took
1,668.8 ns before and 1,808.2 ns after. Repeat runs of this test moved by about 7%, which is wide
enough that only a large change would show — this is not one.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 80.64%. Comparing base (b3bbcc8) to head (880e6cf).

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #5137   +/-   ##
=======================================
  Coverage   80.63%   80.64%           
=======================================
  Files         390      390           
  Lines       96936    96944    +8     
=======================================
+ Hits        78168    78180   +12     
+ Misses      18768    18764    -4     

Files with missing lines	Coverage Δ
scapy/utils.py	72.79% <100.00%> (+0.20%)	⬆️

... and 2 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.