json-repair 0.27.1__py3-none-any.whl → 0.28.0__py3-none-any.whl

json_repair/json_repair.py

@@ -51,9 +51,6 @@ class StringFileWrapper:
             self.fd.seek(current_position)
         return self.length
 
-    def __setitem__(self) -> None:
-        raise Exception("This is read-only!")
-
 
 class LoggerConfig:
     # This is a type class to simplify the declaration
@@ -179,21 +176,12 @@ class JSONParser:
 
             # <member> starts with a <string>
             key = ""
-            while key == "" and self.get_char_at():
-                current_index = self.index
+            while self.get_char_at():
                 key = self.parse_string()
 
-                # This can happen sometimes like { "": "value" }
-                if key == "" and self.get_char_at() == ":":
-                    key = "empty_placeholder"
-                    self.log(
-                        "While parsing an object we found an empty key, replacing with empty_placeholder",
-                        "info",
-                    )
+                if key != "" or (key == "" and self.get_char_at() == ":"):
+                    # If the string is empty but there is a object divider, we are done here
                     break
-                elif key == "" and self.index == current_index:
-                    # Sometimes the string search might not move the index at all, that might lead us to an infinite loop
-                    self.index += 1
 
             self.skip_whitespaces_at()
 
@@ -226,13 +214,6 @@ class JSONParser:
             # Remove trailing spaces
             self.skip_whitespaces_at()
 
-        # Especially at the end of an LLM generated json you might miss the last "}"
-        if (self.get_char_at() or "}") != "}":
-            self.log(
-                "While parsing an object, we couldn't find the closing }, ignoring",
-                "info",
-            )
-
         self.index += 1
         return obj
 
@@ -261,13 +242,6 @@ class JSONParser:
             while char and (char.isspace() or char == ","):
                 self.index += 1
                 char = self.get_char_at()
-            # If this is the right value of an object and we are closing the object, it means the array is over
-            if self.get_context() == "object_value" and char == "}":
-                self.log(
-                    "While parsing an array inside an object, we got to the end without finding a ]. Stopped parsing",
-                    "info",
-                )
-                break
 
         # Especially at the end of an LLM generated json you might miss the last "]"
         char = self.get_char_at()
@@ -275,14 +249,6 @@ class JSONParser:
             self.log(
                 "While parsing an array we missed the closing ], adding it back", "info"
             )
-            # Sometimes when you fix a missing "]" you'll have a trailing "," there that makes the JSON invalid
-            if char == ",":
-                # Remove trailing "," before adding the "]"
-                self.log(
-                    "While parsing an array, found a trailing , before adding ]",
-                    "info",
-                )
-
             self.index -= 1
 
         self.index += 1
@@ -337,7 +303,11 @@ class JSONParser:
 
         # There is sometimes a weird case of doubled quotes, we manage this also later in the while loop
         if self.get_char_at() == lstring_delimiter:
-            # This is a valid exception only if it's closed by a double delimiter again
+            # If it's an empty key, this was easy
+            if self.get_context() == "object_key" and self.get_char_at(1) == ":":
+                self.index += 1
+                return ""
+            # Find the next delimiter
             i = 1
             next_c = self.get_char_at(i)
             while next_c and next_c != rstring_delimiter:
@@ -429,6 +399,7 @@ class JSONParser:
                         "While parsing a string, we found a doubled quote, ignoring it",
                         "info",
                     )
+                    self.index += 1
                 elif missing_quotes and self.get_context() == "object_value":
                     # In case of missing starting quote I need to check if the delimeter is the end or the beginning of a key
                     i = 1
@@ -467,7 +438,7 @@ class JSONParser:
                     ]:
                         # This is a bit of a weird workaround, essentially in object_value context we don't always break on commas
                         # This is because the routine after will make sure to correct any bad guess and this solves a corner case
-                        if next_c.isalpha():
+                        if check_comma_in_object_value and next_c.isalpha():
                             check_comma_in_object_value = False
                         # If we are in an object context, let's check for the right delimiters
                         if (
@@ -575,22 +546,18 @@ class JSONParser:
             # The number ends with a non valid character for a number/currency, rolling back one
             number_str = number_str[:-1]
             self.index -= 1
-        if number_str:
-            try:
-                if "," in number_str:
-                    return str(number_str)
-                if "." in number_str or "e" in number_str or "E" in number_str:
-                    return float(number_str)
-                elif number_str == "-":
-                    # If there is a stray "-" this will throw an exception, throw away this character
-                    return self.parse_json()
-                else:
-                    return int(number_str)
-            except ValueError:
-                return number_str
-        else:
-            # If nothing works, let's skip and keep parsing
-            return self.parse_json()
+        try:
+            if "," in number_str:
+                return str(number_str)
+            if "." in number_str or "e" in number_str or "E" in number_str:
+                return float(number_str)
+            elif number_str == "-":
+                # If there is a stray "-" this will throw an exception, throw away this character
+                return self.parse_json()
+            else:
+                return int(number_str)
+        except ValueError:
+            return number_str
 
     def parse_boolean_or_null(self) -> Union[bool, str, None]:
         # <boolean> is one of the literal strings 'true', 'false', or 'null' (unquoted)
@@ -644,16 +611,10 @@ class JSONParser:
             self.context.append(value)
 
     def reset_context(self) -> None:
-        try:
-            self.context.pop()
-        except Exception:
-            return
+        self.context.pop()
 
     def get_context(self) -> str:
-        try:
-            return self.context[-1]
-        except Exception:
-            return ""
+        return self.context[-1]
 
     def log(self, text: str, level: str) -> None:
         if level == self.logger.log_level:

{json_repair-0.27.1.dist-info → json_repair-0.28.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: json_repair
-Version: 0.27.1
+Version: 0.28.0
 Summary: A package to repair broken json strings
 Author-email: Stefano Baccianella <4247706+mangiucugna@users.noreply.github.com>
 License: MIT License
@@ -27,6 +27,7 @@ License: MIT License
         
 Project-URL: Homepage, https://github.com/mangiucugna/json_repair/
 Project-URL: Bug Tracker, https://github.com/mangiucugna/json_repair/issues
+Project-URL: Live demo, https://mangiucugna.github.io/json_repair/
 Keywords: JSON,REPAIR,LLM,PARSER
 Classifier: Programming Language :: Python :: 3
 Classifier: License :: OSI Approved :: MIT License
@@ -38,7 +39,7 @@ License-File: LICENSE
 [![PyPI](https://img.shields.io/pypi/v/json-repair)](https://pypi.org/project/json-repair/)
 ![Python version](https://img.shields.io/badge/python-3.8+-important)
 [![PyPI downloads](https://img.shields.io/pypi/dm/json-repair)](https://pypi.org/project/json-repair/)
-
+[![Github Sponsors](https://img.shields.io/github/sponsors/mangiucugna)](https://github.com/sponsors/mangiucugna)
 
 This simple package can be used to fix an invalid json string. To know all cases in which this package will work, check out the unit test.
 

json_repair-0.28.0.dist-info/RECORD

@@ -0,0 +1,8 @@
+json_repair/__init__.py,sha256=IIzSm1DsCRrr8seF3UeMZXwxcq-tE3j-8d1WBxvEJvE,178
+json_repair/json_repair.py,sha256=5WjVoNO7Grdq9wfnbSjggiVlDv2XVcvWbCIQVmREP38,30109
+json_repair/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+json_repair-0.28.0.dist-info/LICENSE,sha256=wrjQo8MhNrNCicXtMe3MHmS-fx8AmQk1ue8AQwiiFV8,1076
+json_repair-0.28.0.dist-info/METADATA,sha256=rPrWno-My7ZQt7EJIR0BodmpLQ84fr0vjJwBM4XYmYU,8043
+json_repair-0.28.0.dist-info/WHEEL,sha256=HiCZjzuy6Dw0hdX5R3LCFPDmFS4BWl8H-8W39XfmgX4,91
+json_repair-0.28.0.dist-info/top_level.txt,sha256=7-VZwZN2CgB_n0NlSLk-rEUFh8ug21lESbsblOYuZqw,12
+json_repair-0.28.0.dist-info/RECORD,,

{json_repair-0.27.1.dist-info → json_repair-0.28.0.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (72.1.0)
+Generator: setuptools (72.2.0)
 Root-Is-Purelib: true
 Tag: py3-none-any
 

json_repair-0.27.1.dist-info/RECORD

@@ -1,7 +0,0 @@
-json_repair/__init__.py,sha256=IIzSm1DsCRrr8seF3UeMZXwxcq-tE3j-8d1WBxvEJvE,178
-json_repair/json_repair.py,sha256=54sOesew2z21WAd4UVaEYOLI8oUkSqF8Z4HbkYBdCj8,31792
-json_repair-0.27.1.dist-info/LICENSE,sha256=wrjQo8MhNrNCicXtMe3MHmS-fx8AmQk1ue8AQwiiFV8,1076
-json_repair-0.27.1.dist-info/METADATA,sha256=z1CIA9xV6Cnd7Ko49bZ1snzdwJHK-ev8vKCfhV62vzo,7863
-json_repair-0.27.1.dist-info/WHEEL,sha256=R0nc6qTxuoLk7ShA2_Y-UWkN8ZdfDBG2B6Eqpz2WXbs,91
-json_repair-0.27.1.dist-info/top_level.txt,sha256=7-VZwZN2CgB_n0NlSLk-rEUFh8ug21lESbsblOYuZqw,12
-json_repair-0.27.1.dist-info/RECORD,,