Prevent `DictDumper.object_hook` from calling `.to_dict()` on native `dict` objects (avoids `AttributeError`)

[neomosaic]

Summary

Run into an issue with PyPCAPKit v1.3.5.post17, where its internal DictDumper (from the dictdumper project) erroneously calls .to_dict() on native Python dict objects during JSON dumping, leading to an AttributeError when using external engines like PyShark . Despite PyPCAPKit’s support for third-party engines (Scapy, DPKT, PyShark) via pcapkit.foundation.engines , the dumper lacks a guard for dict instances.

---

Environment

• Library: PyPCAPKit v1.3.5.post17 (released May 17, 2025)
• Python: 3.10 (any 3.6+ supported)
• OS: macOS
• External Engine: PyShark via pyshark.FileCapture (native dict packets)
• Dumper: dictdumper.JSON from the DictDumper project

---

Description

When calling:

from pcapkit import extract

extract(
    fin          = 'capture.pcap',
    fout         = 'out.json',
    format       = 'json',
    engine       = 'pyshark',
    store        = True,
    files        = True,
)

the internal dumper’s object_hook attempts:

if isinstance(o, (Info, Schema)):
    return o.to_dict()

on an object o that is already a plain dict, resulting in:

AttributeError: 'dict' object has no attribute 'to_dict'

This prevents any JSON file from being produced, even though packet decoding and trace collection run correctly under the hood.

---

Steps to Reproduce

1. Install PyPCAPKit:

   pip install pypcapkit==1.3.5.post17
   

   ([PyPI][1])

2. Install PyShark:

   pip install pyshark
   

3. Run the above extract(...) call on any PCAP file

4. Observe the AttributeError from dictdumper/common.py when dumping JSON.

---

Expected Behavior

• Complete extraction, trace logging, and JSON output of all packets and trace events.
• dictdumper should serialize Python dict objects directly (no .to_dict()), or wrap them appropriately.

---

Actual Behavior

• extract() fails during the dumping phase with:

  AttributeError: 'dict' object has no attribute 'to_dict'
  

• No JSON or trace files are created, despite successful parsing and tracing

---

Impact

Users cannot leverage PyPCAPKit’s full trace and per-engine dumping capabilities when using JSON format with external engines, effectively breaking workflows.

---

Proposed Fix

In pcapkit/dumpkit/common.py, modify DictDumper.object_hook to short-circuit on dicts:

 def object_hook(self, o):
     if isinstance(o, (tuple, set, frozenset)):
         return self.make_object(o, list(o))
+    if isinstance(o, dict):
+        return o
     if isinstance(o, (Info, Schema)):
         return o.to_dict()
     if isinstance(o, (ipaddress.IPv4Address, ipaddress.IPv6Address)):
         return str(o)

This ensures that native dict instances are passed through unmodified, preventing the .to_dict() call on plain dicts.