json-repair 0.39.0__tar.gz → 0.40.0__tar.gz

{json_repair-0.39.0/src/json_repair.egg-info → json_repair-0.40.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: json_repair
-Version: 0.39.0
+Version: 0.40.0
 Summary: A package to repair broken json strings
 Author-email: Stefano Baccianella <4247706+mangiucugna@users.noreply.github.com>
 License: MIT License
@@ -196,12 +196,12 @@ pipx install json-repair
 to know all options available:
 ```
 $ json_repair -h
-usage: json_repair [-h] [-i] [-o TARGET] [--ensure_ascii] [--indent INDENT] filename
+usage: json_repair [-h] [-i] [-o TARGET] [--ensure_ascii] [--indent INDENT] [filename]
 
 Repair and parse JSON files.
 
 positional arguments:
-  filename              The JSON file to repair
+  filename              The JSON file to repair (if omitted, reads from stdin)
 
 options:
   -h, --help            show this help message and exit
@@ -226,13 +226,13 @@ In this example, any version that starts with `0.` will be acceptable, allowing
 # How to cite
 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_-_2024,
-        author = {Baccianella, Stefano},
-        month = aug,
-        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 = {0.28.3},
-        year = {2024}
+    @software{Baccianella_JSON_Repair_-_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-0.39.0 → json_repair-0.40.0}/README.md

@@ -158,12 +158,12 @@ pipx install json-repair
 to know all options available:
 ```
 $ json_repair -h
-usage: json_repair [-h] [-i] [-o TARGET] [--ensure_ascii] [--indent INDENT] filename
+usage: json_repair [-h] [-i] [-o TARGET] [--ensure_ascii] [--indent INDENT] [filename]
 
 Repair and parse JSON files.
 
 positional arguments:
-  filename              The JSON file to repair
+  filename              The JSON file to repair (if omitted, reads from stdin)
 
 options:
   -h, --help            show this help message and exit
@@ -188,13 +188,13 @@ In this example, any version that starts with `0.` will be acceptable, allowing
 # How to cite
 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_-_2024,
-        author = {Baccianella, Stefano},
-        month = aug,
-        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 = {0.28.3},
-        year = {2024}
+    @software{Baccianella_JSON_Repair_-_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-0.39.0 → json_repair-0.40.0}/pyproject.toml

@@ -3,7 +3,7 @@ requires = ["setuptools>=61.0"]
 build-backend = "setuptools.build_meta"
 [project]
 name = "json_repair"
-version = "0.39.0"
+version = "0.40.0"
 license = {file = "LICENSE"}
 authors = [
   { name="Stefano Baccianella", email="4247706+mangiucugna@users.noreply.github.com" },

{json_repair-0.39.0 → json_repair-0.40.0}/src/json_repair/json_parser.py

@@ -9,6 +9,7 @@ JSONReturnType = Union[Dict[str, Any], List[Any], str, float, int, bool, None]
 class JSONParser:
     # Constants
     STRING_DELIMITERS = ['"', "'", "“", "”"]
+    NUMBER_CHARS = set("0123456789-.eE/,")
 
     def __init__(
         self,
@@ -129,8 +130,6 @@ class JSONParser:
             # Context is used in the string parser to manage the lack of quotes
             self.context.set(ContextValues.OBJECT_KEY)
 
-            self.skip_whitespaces_at()
-
             # Save this index in case we need find a duplicate key
             rollback_index = self.index
 
@@ -219,18 +218,13 @@ class JSONParser:
                 char = self.get_char_at()
 
         # Especially at the end of an LLM generated json you might miss the last "]"
-        char = self.get_char_at()
         if char and char != "]":
             self.log(
-                "While parsing an array we missed the closing ], adding it back",
-            )
-            self.index -= 1
-            # Add the missing closing bracket
-            self.json_str = (
-                self.json_str[: self.index + 1] + "]" + self.json_str[self.index + 1 :]
+                "While parsing an array we missed the closing ], ignoring it",
             )
 
         self.index += 1
+
         self.context.reset()
         return arr
 
@@ -275,15 +269,11 @@ class JSONParser:
             self.log(
                 "While parsing a string, we found a literal instead of a quote",
             )
-            self.log(
-                "While parsing a string, we found no starting quote. Will add the quote back",
-            )
             missing_quotes = True
 
         if not missing_quotes:
             self.index += 1
 
-        self.skip_whitespaces_at()
         # There is sometimes a weird case of doubled quotes, we manage this also later in the while loop
         if self.get_char_at() in self.STRING_DELIMITERS:
             # If the next character is the same type of quote, then we manage it as double quotes
@@ -583,6 +573,13 @@ class JSONParser:
                     elif (
                         next_c == rstring_delimiter and self.get_char_at(i - 1) != "\\"
                     ):
+                        # Check if self.index:self.index+i is only whitespaces, break if that's the case
+                        if all(
+                            str(self.get_char_at(j)).isspace()
+                            for j in range(1, i)
+                            if self.get_char_at(j)
+                        ):
+                            break
                         if self.context.current == ContextValues.OBJECT_VALUE:
                             # But this might not be it! This could be just a missing comma
                             # We found a delimiter and we need to check if this is a key
@@ -610,27 +607,24 @@ class JSONParser:
                                 self.index += 1
                                 char = self.get_char_at()
                         elif self.context.current == ContextValues.ARRAY:
-                            # In array context this could be something like "lorem "ipsum" sic"
-                            # So let's check if we find a rstring_delimiter forward otherwise end early
-                            i = self.skip_to_character(rstring_delimiter, idx=i + 1)
-                            next_c = self.get_char_at(i)
-                            if next_c and next_c == rstring_delimiter:
-                                # Ok now if I find a comma or a closing ], that can be have also an optional rstring_delimiter before them
-                                # We can consider this a misplaced quote
-                                i += 1
-                                i = self.skip_whitespaces_at(
-                                    idx=i, move_main_index=False
-                                )
-                                next_c = self.get_char_at(i)
-                                if next_c and next_c in [",", "]"]:
-                                    self.log(
-                                        "While parsing a string, we a misplaced quote that would have closed the string but has a different meaning here, ignoring it",
-                                    )
-                                    unmatched_delimiter = not unmatched_delimiter
-                                    string_acc += str(char)
-                                    self.index += 1
-                                    char = self.get_char_at()
-
+                            # If we got up to here it means that this is a situation like this:
+                            # ["bla bla bla "puppy" bla bla bla "kitty" bla bla"]
+                            # So we need to ignore this quote
+                            self.log(
+                                "While parsing a string in Array context, we detected a quoted section that would have closed the string but has a different meaning here, ignoring it",
+                            )
+                            unmatched_delimiter = not unmatched_delimiter
+                            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
@@ -663,10 +657,9 @@ class JSONParser:
     def parse_number(self) -> Union[float, int, str, JSONReturnType]:
         # <number> is a valid real number expressed in one of a number of given formats
         number_str = ""
-        number_chars = set("0123456789-.eE/,")
         char = self.get_char_at()
         is_array = self.context.current == ContextValues.ARRAY
-        while char and char in number_chars and (char != "," or not is_array):
+        while char and char in self.NUMBER_CHARS and (not is_array or char != ","):
             number_str += char
             self.index += 1
             char = self.get_char_at()
@@ -712,51 +705,6 @@ 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:
@@ -827,6 +775,51 @@ 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-0.39.0 → json_repair-0.40.0/src/json_repair.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: json_repair
-Version: 0.39.0
+Version: 0.40.0
 Summary: A package to repair broken json strings
 Author-email: Stefano Baccianella <4247706+mangiucugna@users.noreply.github.com>
 License: MIT License
@@ -196,12 +196,12 @@ pipx install json-repair
 to know all options available:
 ```
 $ json_repair -h
-usage: json_repair [-h] [-i] [-o TARGET] [--ensure_ascii] [--indent INDENT] filename
+usage: json_repair [-h] [-i] [-o TARGET] [--ensure_ascii] [--indent INDENT] [filename]
 
 Repair and parse JSON files.
 
 positional arguments:
-  filename              The JSON file to repair
+  filename              The JSON file to repair (if omitted, reads from stdin)
 
 options:
   -h, --help            show this help message and exit
@@ -226,13 +226,13 @@ In this example, any version that starts with `0.` will be acceptable, allowing
 # How to cite
 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_-_2024,
-        author = {Baccianella, Stefano},
-        month = aug,
-        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 = {0.28.3},
-        year = {2024}
+    @software{Baccianella_JSON_Repair_-_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-0.39.0 → json_repair-0.40.0}/tests/test_json_repair.py

@@ -126,6 +126,9 @@ def test_array_edge_cases():
     assert repair_json('["lorem "ipsum" sic"]') == '["lorem \\"ipsum\\" sic"]'
     assert repair_json('{"key1": ["value1", "value2"}, "key2": ["value3", "value4"]}') == '{"key1": ["value1", "value2"], "key2": ["value3", "value4"]}'
     assert repair_json('[ "value", /* comment */ "value2" ]') == '["value", "value2"]'
+    assert repair_json('{"key": ["value" "value1" "value2"]}') == '{"key": ["value", "value1", "value2"]}'
+    assert repair_json('{"key": ["lorem "ipsum" dolor "sit" amet, "consectetur" ", "lorem "ipsum" dolor", "lorem"]}') == '{"key": ["lorem \\"ipsum\\" dolor \\"sit\\" amet, \\"consectetur\\" ", "lorem \\"ipsum\\" dolor", "lorem"]}'
+    assert repair_json('{"k"e"y": "value"}') == '{"k\\"e\\"y": "value"}'
 
 def test_escaping():
     assert repair_json("'\"'") == '""'
@@ -150,7 +153,7 @@ def test_object_edge_cases():
     assert repair_json('{"lorem": ipsum, sic, datum.",}') == '{"lorem": "ipsum, sic, datum."}'
     assert repair_json('{"lorem": sic tamet. "ipsum": sic tamet, quick brown fox. "sic": ipsum}') == '{"lorem": "sic tamet.", "ipsum": "sic tamet", "sic": "ipsum"}'
     assert repair_json('{"lorem_ipsum": "sic tamet, quick brown fox. }') == '{"lorem_ipsum": "sic tamet, quick brown fox."}'
-    assert repair_json('{"key":value, " key2":"value2" }') == '{"key": "value", "key2": "value2"}'
+    assert repair_json('{"key":value, " key2":"value2" }') == '{"key": "value", " key2": "value2"}'
     assert repair_json('{"key":value "key2":"value2" }') == '{"key": "value", "key2": "value2"}'
     assert repair_json("{'text': 'words{words in brackets}more words'}") == '{"text": "words{words in brackets}more words"}'
     assert repair_json('{text:words{words in brackets}}') == '{"text": "words{words in brackets}"}'
@@ -265,8 +268,8 @@ def test_repair_json_from_file():
         # Write content to the temporary file
         with os.fdopen(temp_fd, 'w') as tmp:
             tmp.write("{key:value}")
-        assert from_file(filename=temp_path, logging=True) == ({'key': 'value'}, [{'text': 'While parsing a string, we found a literal instead of a quote', 'context': '{key:value}'}, {'text': 'While parsing a string, we found no starting quote. Will add the quote back', 'context': '{key:value}'}, {'context': '{key:value}', 'text': 'While parsing a string missing the left delimiter in object key context, we found a :, stopping here',}, {'text': 'While parsing a string, we missed the closing quote, ignoring', 'context': '{key:value}'}, {'text': 'While parsing a string, we found a literal instead of a quote', 'context': '{key:value}'}, {'text': 'While parsing a string, we found no starting quote. Will add the quote back', 'context': '{key:value}'}, {'context': '{key:value}', 'text': 'While parsing a string missing the left delimiter in object value context, we found a , or } and we couldn\'t determine that a right delimiter was present. Stopping here'}, {'text': 'While parsing a string, we missed the closing quote, ignoring', 'context': '{key:value}'}])
-        assert from_file(filename=temp_path, logging=True, chunk_length=2) == ({'key': 'value'}, [{'text': 'While parsing a string, we found a literal instead of a quote', 'context': '{key:value}'}, {'text': 'While parsing a string, we found no starting quote. Will add the quote back', 'context': '{key:value}'}, {'context': '{key:value}', 'text': 'While parsing a string missing the left delimiter in object key context, we found a :, stopping here',}, {'text': 'While parsing a string, we missed the closing quote, ignoring', 'context': '{key:value}'}, {'text': 'While parsing a string, we found a literal instead of a quote', 'context': '{key:value}'}, {'text': 'While parsing a string, we found no starting quote. Will add the quote back', 'context': '{key:value}'}, {'context': '{key:value}', 'text': 'While parsing a string missing the left delimiter in object value context, we found a , or } and we couldn\'t determine that a right delimiter was present. Stopping here'}, {'text': 'While parsing a string, we missed the closing quote, ignoring', 'context': '{key:value}'}])
+        assert from_file(filename=temp_path, logging=True) == ({'key': 'value'}, [{'text': 'While parsing a string, we found a literal instead of a quote', 'context': '{key:value}'}, {'context': '{key:value}', 'text': 'While parsing a string missing the left delimiter in object key context, we found a :, stopping here',}, {'text': 'While parsing a string, we missed the closing quote, ignoring', 'context': '{key:value}'}, {'text': 'While parsing a string, we found a literal instead of a quote', 'context': '{key:value}'}, {'context': '{key:value}', 'text': 'While parsing a string missing the left delimiter in object value context, we found a , or } and we couldn\'t determine that a right delimiter was present. Stopping here'}, {'text': 'While parsing a string, we missed the closing quote, ignoring', 'context': '{key:value}'}])
+        assert from_file(filename=temp_path, logging=True, chunk_length=2) == ({'key': 'value'}, [{'text': 'While parsing a string, we found a literal instead of a quote', 'context': '{key:value}'}, {'context': '{key:value}', 'text': 'While parsing a string missing the left delimiter in object key context, we found a :, stopping here',}, {'text': 'While parsing a string, we missed the closing quote, ignoring', 'context': '{key:value}'}, {'text': 'While parsing a string, we found a literal instead of a quote', 'context': '{key:value}'}, {'context': '{key:value}', 'text': 'While parsing a string missing the left delimiter in object value context, we found a , or } and we couldn\'t determine that a right delimiter was present. Stopping here'}, {'text': 'While parsing a string, we missed the closing quote, ignoring', 'context': '{key:value}'}])
     finally:
         # Clean up - delete the temporary file
         os.remove(temp_path)