txt2stix 1.1.7__tar.gz → 1.1.8__tar.gz

{txt2stix-1.1.7 → txt2stix-1.1.8}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: txt2stix
-Version: 1.1.7
+Version: 1.1.8
 Summary: txt2stix is a Python script that is designed to identify and extract IoCs and TTPs from text files, identify the relationships between them, convert them to STIX 2.1 objects, and output as a STIX 2.1 bundle.
 Project-URL: Homepage, https://github.com/muchdogesec/txt2stix
 Project-URL: Issues, https://github.com/muchdogesec/txt2stix/issues
@@ -14,6 +14,7 @@ Classifier: Programming Language :: Python :: 3
 Requires-Python: >=3.9
 Requires-Dist: base58>=2.1.1
 Requires-Dist: beautifulsoup4>=4.12.3
+Requires-Dist: json-repair
 Requires-Dist: llama-index-core>=0.12.42
 Requires-Dist: llama-index-llms-openai>=0.4.5
 Requires-Dist: mistune>=3.0.2

{txt2stix-1.1.7 → txt2stix-1.1.8}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "txt2stix"
-version = "1.1.7"
+version = "1.1.8"
 authors = [{ name = "dogesec" }]
 maintainers = [{ name = "dogesec" }]
 description = "txt2stix is a Python script that is designed to identify and extract IoCs and TTPs from text files, identify the relationships between them, convert them to STIX 2.1 objects, and output as a STIX 2.1 bundle."
@@ -33,6 +33,7 @@ dependencies = [
   'mistune>=3.0.2',
   'beautifulsoup4>=4.12.3',
   'stix2-validator',
+  "json-repair",
 ]
 
 [tool.hatch.metadata]

{txt2stix-1.1.7 → txt2stix-1.1.8}/requirements.txt

@@ -20,8 +20,13 @@ anyio==4.10.0
     # via
     #   httpx
     #   openai
+arrow==1.4.0
+    # via isoduration
 attrs==25.3.0
-    # via aiohttp
+    # via
+    #   aiohttp
+    #   jsonschema
+    #   referencing
 banks==2.2.0
     # via llama-index-core
 base58==2.1.1
@@ -38,7 +43,11 @@ charset-normalizer==3.4.3
 click==8.2.1
     # via nltk
 colorama==0.4.6
-    # via griffe
+    # via
+    #   griffe
+    #   stix2-validator
+cpe==1.3.1
+    # via stix2-validator
 dataclasses-json==0.6.7
     # via llama-index-core
 deprecated==1.2.18
@@ -54,6 +63,8 @@ filelock==3.19.1
     # via tldextract
 filetype==1.2.0
     # via llama-index-core
+fqdn==1.5.1
+    # via jsonschema
 frozenlist==1.7.0
     # via
     #   aiohttp
@@ -76,15 +87,28 @@ idna==3.10
     # via
     #   anyio
     #   httpx
+    #   jsonschema
     #   requests
     #   tldextract
     #   yarl
+isoduration==20.11.0
+    # via jsonschema
 jinja2==3.1.6
     # via banks
 jiter==0.10.0
     # via openai
 joblib==1.5.1
     # via nltk
+json-repair==0.52.4
+    # via txt2stix (pyproject.toml)
+jsonpointer==3.0.0
+    # via jsonschema
+jsonschema[format-nongpl]==4.25.1
+    # via stix2-validator
+jsonschema-specifications==2025.9.1
+    # via jsonschema
+lark==1.3.1
+    # via rfc3987-syntax
 llama-index-core==0.13.3
     # via
     #   llama-index-llms-openai
@@ -144,12 +168,20 @@ pydantic==2.11.7
     #   openai
 pydantic-core==2.33.2
     # via pydantic
+python-dateutil==2.9.0.post0
+    # via
+    #   arrow
+    #   stix2-validator
 python-dotenv==1.1.1
     # via txt2stix (pyproject.toml)
 pytz==2025.2
     # via stix2
 pyyaml==6.0.2
     # via llama-index-core
+referencing==0.37.0
+    # via
+    #   jsonschema
+    #   jsonschema-specifications
 regex==2025.7.34
     # via
     #   nltk
@@ -159,20 +191,36 @@ requests==2.32.5
     #   llama-index-core
     #   requests-file
     #   stix2
+    #   stix2-validator
     #   stix2extensions
     #   tiktoken
     #   tldextract
     #   txt2stix (pyproject.toml)
 requests-file==2.1.0
     # via tldextract
+rfc3339-validator==0.1.4
+    # via jsonschema
+rfc3986-validator==0.1.1
+    # via jsonschema
+rfc3987-syntax==1.1.0
+    # via jsonschema
+rpds-py==0.28.0
+    # via
+    #   jsonschema
+    #   referencing
 rstr==3.2.2
     # via schwifty
 schwifty==2025.7.0
     # via txt2stix (pyproject.toml)
 simplejson==3.20.1
-    # via stix2
+    # via
+    #   stix2
+    #   stix2-validator
 six==1.17.0
-    # via stix2-patterns
+    # via
+    #   python-dateutil
+    #   rfc3339-validator
+    #   stix2-patterns
 sniffio==1.3.1
     # via
     #   anyio
@@ -184,7 +232,11 @@ sqlalchemy[asyncio]==2.0.43
 stix2==3.0.1
     # via stix2extensions
 stix2-patterns==2.0.0
-    # via stix2
+    # via
+    #   stix2
+    #   stix2-validator
+stix2-validator==3.2.0
+    # via txt2stix (pyproject.toml)
 stix2extensions==1.2.0
     # via txt2stix (pyproject.toml)
 tenacity==9.1.2
@@ -218,10 +270,16 @@ typing-inspect==0.9.0
     #   llama-index-core
 typing-inspection==0.4.1
     # via pydantic
+tzdata==2025.2
+    # via arrow
+uri-template==1.3.0
+    # via jsonschema
 urllib3==2.5.0
     # via requests
 validators==0.35.0
     # via txt2stix (pyproject.toml)
+webcolors==25.10.0
+    # via jsonschema
 wrapt==1.17.3
     # via
     #   deprecated

{txt2stix-1.1.7 → txt2stix-1.1.8}/tests/src/test_retriever.py

@@ -48,7 +48,10 @@ def f():
         (
             "ctibutler-mitre-attack-mobile-name",
             "Impair Defenses",
-            ["attack-pattern--20b0931a-8952-42ca-975f-775bad295f1a"],
+            [
+                "attack-pattern--20b0931a-8952-42ca-975f-775bad295f1a",
+                "x-mitre-detection-strategy--5560747b-ad67-478e-b3f2-14e55864e532",
+            ],
         ),
         (
             "ctibutler-mitre-attack-mobile-aliases",
@@ -58,7 +61,10 @@ def f():
         (
             "ctibutler-mitre-attack-ics-name",
             "Program Upload",
-            ["attack-pattern--3067b85e-271e-4bc5-81ad-ab1a81d411e3"],
+            [
+                "attack-pattern--3067b85e-271e-4bc5-81ad-ab1a81d411e3",
+                "x-mitre-detection-strategy--6f921aa8-deb3-4286-8101-26a7cbe80c0e",
+            ],
         ),
         (
             "ctibutler-mitre-attack-ics-aliases",
@@ -81,9 +87,7 @@ def f():
         (
             "ctibutler-location",
             "NG",
-            (
-                "location--6dbe266a-c149-5ba3-8b39-74f1b5063312",
-            ),
+            ("location--6dbe266a-c149-5ba3-8b39-74f1b5063312",),
         ),
         (
             "ctibutler-mitre-atlas-id",

{txt2stix-1.1.7 → txt2stix-1.1.8}/txt2stix/ai_extractor/utils.py

@@ -5,6 +5,8 @@ import logging
 import dotenv
 import textwrap
 
+import json_repair
+
 from ..extractions import Extractor
 
 from pydantic import BaseModel, Field, RootModel
@@ -63,7 +65,8 @@ class ParserWithLogging(PydanticOutputParser):
         print(text, file=f)
         print("=================close=================" + "\n"*5, file=f)
         logging.debug(f.getvalue())
-        return super().parse(text)
+        repaired_json = json_repair.repair_json(text)
+        return super().parse(repaired_json)
 
 def get_extractors_str(extractors):
     extractor: Extractor = None