From c4abb263085f5a824d7507a5c0353188b73ec3b5 Mon Sep 17 00:00:00 2001
From: Alexander Bilz <mail@alexbilz.com>
Date: Sun, 30 Jun 2024 12:58:40 +0200
Subject: [PATCH] fix: JSON decode error (#76)

* fix: JSON decode error
---
 src/forensicsim/backend.py |  4 +---
 src/forensicsim/parser.py  | 28 ++++++++++++++++++----------
 2 files changed, 19 insertions(+), 13 deletions(-)

diff --git a/src/forensicsim/backend.py b/src/forensicsim/backend.py
index eef8ea6..70c0906 100644
--- a/src/forensicsim/backend.py
+++ b/src/forensicsim/backend.py
@@ -209,8 +209,6 @@ def write_results_to_json(data: list[dict[str, Any]], outputpath: Path) -> None:
     # Dump messages into a json file
     try:
         with open(outputpath, "w", encoding="utf-8") as f:
-            json.dump(
-                data, f, indent=4, sort_keys=True, default=str, ensure_ascii=False
-            )
+            json.dump(data, f, indent=4, default=str, ensure_ascii=False)
     except OSError as e:
         print(e)
diff --git a/src/forensicsim/parser.py b/src/forensicsim/parser.py
index 3ec3084..f7fadad 100644
--- a/src/forensicsim/parser.py
+++ b/src/forensicsim/parser.py
@@ -2,6 +2,7 @@
 import warnings
 from dataclasses import dataclass, field
 from datetime import datetime
+from json import JSONDecodeError
 from pathlib import Path
 from typing import Any, Optional, Union
 
@@ -26,15 +27,22 @@ def strip_html_tags(value: str) -> str:
 
 
 def decode_dict(properties: Union[bytes, str, dict]) -> dict[str, Any]:
-    if isinstance(properties, bytes):
-        soup = BeautifulSoup(properties, features="html.parser")
-        properties = properties.decode(soup.original_encoding)
-    if isinstance(properties, dict):
-        # handle case where nested childs are dicts or list but provided with "" but have to be expanded.
-        for key, value in properties.items():
-            if isinstance(value, str) and value.startswith(("[", "{")):
-                properties[key] = json.loads(value, strict=False)
-        return properties
+    try:
+        if isinstance(properties, bytes):
+            soup = BeautifulSoup(properties, features="html.parser")
+            properties = properties.decode(
+                encoding=soup.original_encoding, errors="ignore"
+            )
+        if isinstance(properties, dict):
+            # handle case where nested childs are dicts or list but provided with "" but have to be expanded.
+            for key, value in properties.items():
+                if isinstance(value, str) and value.startswith(("[", "{")):
+                    properties[key] = json.loads(value, strict=False)
+            return properties
+    except JSONDecodeError as e:
+        print(e)
+        print("Couldn't decode dictionary ", properties)
+        return {}
 
     return json.loads(properties, strict=False)
 
@@ -292,7 +300,7 @@ def _parse_reply_chains(reply_chains: list[dict], version: str) -> set[Message]:
                 elif version == "v2":
                     rc |= {"cached_deduplication_key": md.get("dedupeKey")}
                     rc |= {"clientmessageid": md.get("clientMessageId")}
-                    # set to clientArrivalTime as compose Time is no longer present
+                    # set to clientArrivalTime as compose time is no longer present
                     rc |= {"composetime": md.get("clientArrivalTime")}
                     rc |= {"contenttype": md.get("contentType")}
                     # set to clientArrivalTime as created time is no longer present