json-repair 0.39.1__py3-none-any.whl → 0.41.0__py3-none-any.whl

json_repair/json_parser.py

@@ -2,6 +2,7 @@ from typing import Any, Dict, List, Literal, Optional, TextIO, Tuple, Union
 
 from .json_context import ContextValues, JsonContext
 from .string_file_wrapper import StringFileWrapper
+from .object_comparer import ObjectComparer
 
 JSONReturnType = Union[Dict[str, Any], List[Any], str, float, int, bool, None]
 
@@ -54,6 +55,9 @@ class JSONParser:
             while self.index < len(self.json_str):
                 j = self.parse_json()
                 if j != "":
+                    if ObjectComparer.is_same_object(json[-1], j):
+                        # replace the last entry with the new one since the new one seems an update
+                        json.pop()
                     json.append(j)
                 if self.index == last_index:
                     self.index += 1
@@ -617,7 +621,14 @@ class JSONParser:
                             string_acc += str(char)
                             self.index += 1
                             char = self.get_char_at()
-
+                        elif self.context.current == ContextValues.OBJECT_KEY:
+                            # In this case we just ignore this and move on
+                            self.log(
+                                "While parsing a string in Object Key context, we detected a quoted section that would have closed the string but has a different meaning here, ignoring it",
+                            )
+                            string_acc += str(char)
+                            self.index += 1
+                            char = self.get_char_at()
         if (
             char
             and missing_quotes
@@ -698,51 +709,6 @@ class JSONParser:
         self.index = starting_index
         return ""
 
-    def get_char_at(self, count: int = 0) -> Union[str, Literal[False]]:
-        # Why not use something simpler? Because try/except in python is a faster alternative to an "if" statement that is often True
-        try:
-            return self.json_str[self.index + count]
-        except IndexError:
-            return False
-
-    def skip_whitespaces_at(self, idx: int = 0, move_main_index=True) -> int:
-        """
-        This function quickly iterates on whitespaces, syntactic sugar to make the code more concise
-        """
-        try:
-            char = self.json_str[self.index + idx]
-        except IndexError:
-            return idx
-        while char.isspace():
-            if move_main_index:
-                self.index += 1
-            else:
-                idx += 1
-            try:
-                char = self.json_str[self.index + idx]
-            except IndexError:
-                return idx
-        return idx
-
-    def skip_to_character(self, character: str, idx: int = 0) -> int:
-        """
-        This function quickly iterates to find a character, syntactic sugar to make the code more concise
-        """
-        try:
-            char = self.json_str[self.index + idx]
-        except IndexError:
-            return idx
-        while char != character:
-            idx += 1
-            try:
-                char = self.json_str[self.index + idx]
-            except IndexError:
-                return idx
-        if self.index + idx > 0 and self.json_str[self.index + idx - 1] == "\\":
-            # Ah this is an escaped character, try again
-            return self.skip_to_character(character=character, idx=idx + 1)
-        return idx
-
     def parse_comment(self) -> str:
         """
         Parse code-like comments:
@@ -813,6 +779,51 @@ class JSONParser:
             self.index += 1
             return ""
 
+    def get_char_at(self, count: int = 0) -> Union[str, Literal[False]]:
+        # Why not use something simpler? Because try/except in python is a faster alternative to an "if" statement that is often True
+        try:
+            return self.json_str[self.index + count]
+        except IndexError:
+            return False
+
+    def skip_whitespaces_at(self, idx: int = 0, move_main_index=True) -> int:
+        """
+        This function quickly iterates on whitespaces, syntactic sugar to make the code more concise
+        """
+        try:
+            char = self.json_str[self.index + idx]
+        except IndexError:
+            return idx
+        while char.isspace():
+            if move_main_index:
+                self.index += 1
+            else:
+                idx += 1
+            try:
+                char = self.json_str[self.index + idx]
+            except IndexError:
+                return idx
+        return idx
+
+    def skip_to_character(self, character: str, idx: int = 0) -> int:
+        """
+        This function quickly iterates to find a character, syntactic sugar to make the code more concise
+        """
+        try:
+            char = self.json_str[self.index + idx]
+        except IndexError:
+            return idx
+        while char != character:
+            idx += 1
+            try:
+                char = self.json_str[self.index + idx]
+            except IndexError:
+                return idx
+        if self.index + idx > 0 and self.json_str[self.index + idx - 1] == "\\":
+            # Ah this is an escaped character, try again
+            return self.skip_to_character(character=character, idx=idx + 1)
+        return idx
+
     def _log(self, text: str) -> None:
         window: int = 10
         start: int = max(self.index - window, 0)

json_repair/object_comparer.py

@@ -0,0 +1,63 @@
+from typing import Any
+
+
+class ObjectComparer:
+    def __init__(self) -> None:
+        return
+
+    @staticmethod
+    def is_same_object(obj1: Any, obj2: Any, path: str = "") -> bool:
+        """
+        Recursively compares two objects and ensures that:
+        - Their types match
+        - Their keys/structure match
+        """
+        if type(obj1) is not type(obj2):
+            # Fail immediately if the types don't match
+            print(
+                f"Type mismatch at {path}: {type(obj1).__name__} vs {type(obj2).__name__}"
+            )
+            return False
+
+        if isinstance(obj1, dict) and isinstance(obj2, dict):
+            # Compare dictionary keys
+            keys1, keys2 = set(obj1.keys()), set(obj2.keys())
+            common_keys = keys1 & keys2
+            extra_keys1 = keys1 - keys2
+            extra_keys2 = keys2 - keys1
+
+            if extra_keys1:
+                print(f"Extra keys in first object at {path}: {extra_keys1}")
+                return False
+            if extra_keys2:
+                print(f"Extra keys in second object at {path}: {extra_keys2}")
+                return False
+
+            # Recursively compare the common keys
+            for key in common_keys:
+                if not ObjectComparer.is_same_object(
+                    obj1[key], obj2[key], path=f"{path}/{key}"
+                ):
+                    return False
+
+        elif isinstance(obj1, list) and isinstance(obj2, list):
+            # Compare lists
+            min_length = min(len(obj1), len(obj2))
+            if len(obj1) != len(obj2):
+                print(f"Length mismatch at {path}: {len(obj1)} vs {len(obj2)}")
+                return False
+
+            for i in range(min_length):
+                if not ObjectComparer.is_same_object(
+                    obj1[i], obj2[i], path=f"{path}[{i}]"
+                ):
+                    return False
+
+            if len(obj1) > len(obj2):
+                print(f"Extra items in first list at {path}: {obj1[min_length:]}")
+                return False
+            elif len(obj2) > len(obj1):
+                print(f"Extra items in second list at {path}: {obj2[min_length:]}")
+                return False
+
+        return True

{json_repair-0.39.1.dist-info → json_repair-0.41.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
-Metadata-Version: 2.2
+Metadata-Version: 2.4
 Name: json_repair
-Version: 0.39.1
+Version: 0.41.0
 Summary: A package to repair broken json strings
 Author-email: Stefano Baccianella <4247706+mangiucugna@users.noreply.github.com>
 License: MIT License
@@ -35,6 +35,7 @@ Classifier: Operating System :: OS Independent
 Requires-Python: >=3.9
 Description-Content-Type: text/markdown
 License-File: LICENSE
+Dynamic: license-file
 
 [![PyPI](https://img.shields.io/pypi/v/json-repair)](https://pypi.org/project/json-repair/)
 ![Python version](https://img.shields.io/badge/python-3.9+-important)
@@ -227,12 +228,12 @@ In this example, any version that starts with `0.` will be acceptable, allowing
 If you are using this library in your academic work (as I know many folks are) please find the BibTex here:
 
     @software{Baccianella_JSON_Repair_-_2025,
-        author = {Baccianella, Stefano},
-        month = feb,
-        title = {{JSON Repair - A python module to repair invalid JSON, commonly used to parse the output of LLMs}},
-        url = {https://github.com/mangiucugna/json_repair},
-        version = {0.39.0},
-        year = {2025}
+        author  = "Stefano {Baccianella}",
+        month   = "feb",
+        title   = "JSON Repair - A python module to repair invalid JSON, commonly used to parse the output of LLMs",
+        url     = "https://github.com/mangiucugna/json_repair",
+        version = "0.39.1",
+        year    = 2025
     }
 
 Thank you for citing my work and please send me a link to the paper if you can!

json_repair-0.41.0.dist-info/RECORD

@@ -0,0 +1,14 @@
+json_repair/__init__.py,sha256=c4L2kZrHvWEKfj_ODU2naliNuvU6FlFVxtF0hbLe6s8,178
+json_repair/__main__.py,sha256=EsJb-y89uZEvGQQg1GdIDWzfDwfOMvVekKEtdguQXCM,67
+json_repair/json_context.py,sha256=mm6dOyrPJ1sDskTORZSXCW7W9-5veMlUKqXQ3Hw3EG4,971
+json_repair/json_parser.py,sha256=ff5LPEHVfaRJ7ujVxUm7dxiwlHPzcJK0cGItm_OilpU,38996
+json_repair/json_repair.py,sha256=k-5HRRlCqrxNmJi0u1KE3IUeL4HXqi1XZ7oAL-NFDLo,10314
+json_repair/object_comparer.py,sha256=oMWtBySgrHDH8q2v72HnNQm8SRmudtEsVPkaydSXckE,2210
+json_repair/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+json_repair/string_file_wrapper.py,sha256=koZmdq2-Z5K7XF1bDqX6dEbNaVMJYcMTjq-aGe6NQvA,4526
+json_repair-0.41.0.dist-info/licenses/LICENSE,sha256=wrjQo8MhNrNCicXtMe3MHmS-fx8AmQk1ue8AQwiiFV8,1076
+json_repair-0.41.0.dist-info/METADATA,sha256=9cD1PLAiAi5giNrDwS0TVerOD-wuzEd6wKrE1OeF-Jw,11860
+json_repair-0.41.0.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
+json_repair-0.41.0.dist-info/entry_points.txt,sha256=SNfge3zPSP-ASqriYU9r3NAPaXdseYr7ciPMKdV2uSw,57
+json_repair-0.41.0.dist-info/top_level.txt,sha256=7-VZwZN2CgB_n0NlSLk-rEUFh8ug21lESbsblOYuZqw,12
+json_repair-0.41.0.dist-info/RECORD,,

{json_repair-0.39.1.dist-info → json_repair-0.41.0.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (75.8.0)
+Generator: setuptools (78.1.0)
 Root-Is-Purelib: true
 Tag: py3-none-any
 

json_repair-0.39.1.dist-info/RECORD

@@ -1,13 +0,0 @@
-json_repair/__init__.py,sha256=c4L2kZrHvWEKfj_ODU2naliNuvU6FlFVxtF0hbLe6s8,178
-json_repair/__main__.py,sha256=EsJb-y89uZEvGQQg1GdIDWzfDwfOMvVekKEtdguQXCM,67
-json_repair/json_context.py,sha256=mm6dOyrPJ1sDskTORZSXCW7W9-5veMlUKqXQ3Hw3EG4,971
-json_repair/json_parser.py,sha256=kt58S7pHxCOfqktzn48iMrvd3vi7HTfK6OD02PWwWcc,38189
-json_repair/json_repair.py,sha256=k-5HRRlCqrxNmJi0u1KE3IUeL4HXqi1XZ7oAL-NFDLo,10314
-json_repair/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-json_repair/string_file_wrapper.py,sha256=koZmdq2-Z5K7XF1bDqX6dEbNaVMJYcMTjq-aGe6NQvA,4526
-json_repair-0.39.1.dist-info/LICENSE,sha256=wrjQo8MhNrNCicXtMe3MHmS-fx8AmQk1ue8AQwiiFV8,1076
-json_repair-0.39.1.dist-info/METADATA,sha256=T1k1afyqqWG-NXYHpgntQsPYpmIdkLg72eSo_iNwZZk,11827
-json_repair-0.39.1.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
-json_repair-0.39.1.dist-info/entry_points.txt,sha256=SNfge3zPSP-ASqriYU9r3NAPaXdseYr7ciPMKdV2uSw,57
-json_repair-0.39.1.dist-info/top_level.txt,sha256=7-VZwZN2CgB_n0NlSLk-rEUFh8ug21lESbsblOYuZqw,12
-json_repair-0.39.1.dist-info/RECORD,,