json-repair 0.23.1__tar.gz → 0.25.0__tar.gz

{json_repair-0.23.1/src/json_repair.egg-info → json_repair-0.25.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: json_repair
-Version: 0.23.1
+Version: 0.25.0
 Summary: A package to repair broken json strings
 Author-email: Stefano Baccianella <4247706+mangiucugna@users.noreply.github.com>
 License: MIT License
@@ -45,6 +45,11 @@ If you find this library useful, you can help me by donating toward my monthly b
 
 ---
 
+# Demo
+If you are unsure if this library will fix your specific problem, or simply want your json validated online, you can visit the demo site on GitHub pages: https://mangiucugna.github.io/json_repair/
+
+---
+
 # Motivation
 Some LLMs are a bit iffy when it comes to returning well formed JSON data, sometimes they skip a parentheses and sometimes they add some words in it, because that's what an LLM does.
 Luckily, the mistakes LLMs make are simple enough to be fixed without destroying the content.
@@ -160,6 +165,7 @@ You will need owner access to this repository
 # Repair JSON in other programming languages
 - Typescript: https://github.com/josdejong/jsonrepair
 - Go: https://github.com/RealAlexandreAI/json-repair
+- Ruby: https://github.com/sashazykov/json-repair-rb
 ---
 ## Star History
 

{json_repair-0.23.1 → json_repair-0.25.0}/README.md

@@ -8,6 +8,11 @@ If you find this library useful, you can help me by donating toward my monthly b
 
 ---
 
+# Demo
+If you are unsure if this library will fix your specific problem, or simply want your json validated online, you can visit the demo site on GitHub pages: https://mangiucugna.github.io/json_repair/
+
+---
+
 # Motivation
 Some LLMs are a bit iffy when it comes to returning well formed JSON data, sometimes they skip a parentheses and sometimes they add some words in it, because that's what an LLM does.
 Luckily, the mistakes LLMs make are simple enough to be fixed without destroying the content.
@@ -123,6 +128,7 @@ You will need owner access to this repository
 # Repair JSON in other programming languages
 - Typescript: https://github.com/josdejong/jsonrepair
 - Go: https://github.com/RealAlexandreAI/json-repair
+- Ruby: https://github.com/sashazykov/json-repair-rb
 ---
 ## Star History
 

{json_repair-0.23.1 → json_repair-0.25.0}/pyproject.toml

@@ -3,7 +3,7 @@ requires = ["setuptools>=61.0"]
 build-backend = "setuptools.build_meta"
 [project]
 name = "json_repair"
-version = "0.23.1"
+version = "0.25.0"
 license = {file = "LICENSE"}
 authors = [
   { name="Stefano Baccianella", email="4247706+mangiucugna@users.noreply.github.com" },

{json_repair-0.23.1 → json_repair-0.25.0}/src/json_repair/json_repair.py

@@ -91,6 +91,10 @@ class JSONParser:
     ) -> Union[JSONReturnType, Tuple[JSONReturnType, List[Dict[str, str]]]]:
         json = self.parse_json()
         if self.index < len(self.json_str):
+            self.log(
+                "The parser returned early, checking if there's more json elements",
+                "info",
+            )
             json = [json]
             last_index = self.index
             while self.index < len(self.json_str):
@@ -100,10 +104,13 @@ class JSONParser:
                 if self.index == last_index:
                     self.index += 1
                 last_index = self.index
+            # If nothing extra was found, don't return an array
             if len(json) == 1:
+                self.log(
+                    "There were no more elements, returning the element without the array",
+                    "info",
+                )
                 json = json[0]
-            elif len(json) == 0:
-                json = ""
         if self.logger.log_level == "none":
             return json
         else:
@@ -363,9 +370,34 @@ class JSONParser:
                 if self.get_context() == "object_key" and (
                     char == ":" or char.isspace()
                 ):
+                    self.log(
+                        "While parsing a string missing the left delimiter in object key context, we found a :, stopping here",
+                        "info",
+                    )
                     break
                 elif self.get_context() == "object_value" and char in [",", "}"]:
-                    break
+                    rstring_delimiter_missing = True
+                    # check if this is a case in which the closing comma is NOT missing instead
+                    i = 1
+                    next_c = self.get_char_at(i)
+                    while next_c and next_c != rstring_delimiter:
+                        i += 1
+                        next_c = self.get_char_at(i)
+                    if next_c:
+                        i += 1
+                        next_c = self.get_char_at(i)
+                        # found a delimiter, now we need to check that is followed strictly by a comma or brace
+                        while next_c and next_c.isspace():
+                            i += 1
+                            next_c = self.get_char_at(i)
+                        if next_c and next_c in [",", "}"]:
+                            rstring_delimiter_missing = False
+                    if rstring_delimiter_missing:
+                        self.log(
+                            "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",
+                            "info",
+                        )
+                        break
             string_acc += char
             self.index += 1
             char = self.get_char_at()
@@ -386,6 +418,33 @@ class JSONParser:
                         "While parsing a string, we found a doubled quote, ignoring it",
                         "info",
                     )
+                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
+                    next_c = self.get_char_at(i)
+                    while next_c and next_c not in [
+                        rstring_delimiter,
+                        lstring_delimiter,
+                    ]:
+                        i += 1
+                        next_c = self.get_char_at(i)
+                    if next_c:
+                        # We found a quote, now let's make sure there's a ":" following
+                        i += 1
+                        next_c = self.get_char_at(i)
+                        # found a delimiter, now we need to check that is followed strictly by a comma or brace
+                        while next_c and next_c.isspace():
+                            i += 1
+                            next_c = self.get_char_at(i)
+                        if next_c and next_c == ":":
+                            # Reset the cursor
+                            self.index -= 1
+                            char = self.get_char_at()
+                            self.log(
+                                "In a string with missing quotes and object value context, I found a delimeter but it turns out it was the beginning on the next key. Stopping here.",
+                                "info",
+                            )
+                            break
                 else:
                     # Check if eventually there is a rstring delimiter, otherwise we bail
                     i = 1
@@ -496,7 +555,8 @@ class JSONParser:
         number_str = ""
         number_chars = set("0123456789-.eE/,")
         char = self.get_char_at()
-        while char and char in number_chars:
+        is_array = self.get_context() == "array"
+        while char and char in number_chars and (char != "," or not is_array):
             number_str += char
             self.index += 1
             char = self.get_char_at()

{json_repair-0.23.1 → json_repair-0.25.0/src/json_repair.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: json_repair
-Version: 0.23.1
+Version: 0.25.0
 Summary: A package to repair broken json strings
 Author-email: Stefano Baccianella <4247706+mangiucugna@users.noreply.github.com>
 License: MIT License
@@ -45,6 +45,11 @@ If you find this library useful, you can help me by donating toward my monthly b
 
 ---
 
+# Demo
+If you are unsure if this library will fix your specific problem, or simply want your json validated online, you can visit the demo site on GitHub pages: https://mangiucugna.github.io/json_repair/
+
+---
+
 # Motivation
 Some LLMs are a bit iffy when it comes to returning well formed JSON data, sometimes they skip a parentheses and sometimes they add some words in it, because that's what an LLM does.
 Luckily, the mistakes LLMs make are simple enough to be fixed without destroying the content.
@@ -160,6 +165,7 @@ You will need owner access to this repository
 # Repair JSON in other programming languages
 - Typescript: https://github.com/josdejong/jsonrepair
 - Go: https://github.com/RealAlexandreAI/json-repair
+- Ruby: https://github.com/sashazykov/json-repair-rb
 ---
 ## Star History
 

{json_repair-0.23.1 → json_repair-0.25.0}/tests/test_json_repair.py

@@ -141,6 +141,8 @@ def test_object_edge_cases():
     assert repair_json('''{ "a" : "{ b": {} }" }''') == '{"a": "{ b"}'
     assert repair_json("""{"b": "xxxxx" true}""") == '{"b": "xxxxx"}'
     assert repair_json('{"key": "Lorem "ipsum" s,"}') == '{"key": "Lorem \\"ipsum\\" s,"}'
+    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"}'
 
 def test_number_edge_cases():
     assert repair_json(' - { "test_key": ["test_value", "test_value2"] }') == '{"test_key": ["test_value", "test_value2"]}'
@@ -148,7 +150,7 @@ def test_number_edge_cases():
     assert repair_json('{"key": .25}') == '{"key": 0.25}'
     assert repair_json('{"here": "now", "key": 1/3, "foo": "bar"}') == '{"here": "now", "key": "1/3", "foo": "bar"}'
     assert repair_json('{"key": 12345/67890}') == '{"key": "12345/67890"}'
-    assert repair_json('[105,12') == '["105,12"]'
+    assert repair_json('[105,12') == '[105, 12]'
     assert repair_json('{"key", 105,12,') == '{"key": "105,12"}'
     assert repair_json('{"key": 1/3, "foo": "bar"}') == '{"key": "1/3", "foo": "bar"}'
     assert repair_json('{"key": 10-20}') == '{"key": "10-20"}'
@@ -171,8 +173,8 @@ def test_leading_trailing_characters():
 def test_multiple_jsons():
     assert repair_json("[]{}") == "[[], {}]"
     assert repair_json("{}[]{}") == "[{}, [], {}]"
-    assert repair_json('{"key":"value"}[1,2,3,True]') == '[{"key": "value"}, ["1,2,3", true]]'
-    assert repair_json('lorem ```json {"key":"value"} ``` ipsum ```json [1,2,3,True] ``` 42') == '[{"key": "value"}, ["1,2,3", true]]'
+    assert repair_json('{"key":"value"}[1,2,3,True]') == '[{"key": "value"}, [1, 2, 3, true]]'
+    assert repair_json('lorem ```json {"key":"value"} ``` ipsum ```json [1,2,3,True] ``` 42') == '[{"key": "value"}, [1, 2, 3, true]]'
 
 def test_repair_json_with_objects():
     # Test with valid JSON strings
@@ -242,7 +244,7 @@ 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(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}'}, {'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}'}, {'text': 'While parsing a string, we missed the closing quote, ignoring', 'context': '{key:value}'}])
+        assert(from_file(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}'}])
     finally:
         # Clean up - delete the temporary file
         os.remove(temp_path)

{json_repair-0.23.1 → json_repair-0.25.0}/tests/test_performance.py

@@ -97,7 +97,7 @@ def test_false_false_incorrect(benchmark):
   mean_time = benchmark.stats.get("median")
 
   # Define your time threshold in seconds
-  max_time = 14 / 10 ** 4  # 1.4 millisecond
+  max_time = 15 / 10 ** 4  # 1.5 millisecond
 
   # Assert that the average time is below the threshold
   assert mean_time < max_time, f"Benchmark exceeded threshold: {mean_time:.3f}s > {max_time:.3f}s"