json-repair 0.12.3__tar.gz → 0.13.1__tar.gz

{json_repair-0.12.3/src/json_repair.egg-info → json_repair-0.13.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: json_repair
-Version: 0.12.3
+Version: 0.13.1
 Summary: A package to repair broken json strings
 Author-email: Stefano Baccianella <4247706+mangiucugna@users.noreply.github.com>
 License: MIT License
@@ -65,6 +65,33 @@ or just
 
     decoded_object = json_repair.repair_json(json_string, return_objects=True)
 
+### Read json from a file or file descriptor
+
+JSON repair provides also a drop-in replacement for `json.load()`:
+
+    import json_repair
+
+    try:
+        file_descriptor = open(fname, 'rb')
+    except OSError:
+        ...
+
+    with file_descriptor:
+        decoded_object = json_repair.load(file_descriptor)
+
+and another method to read from a file:
+
+    import json_repair
+
+    try:
+        decoded_object = json_repair.from_file(json_file)
+    except OSError:
+        ...
+    except IOError:
+        ...
+
+Keep in mind that the library will not catch any IO-related exception and those will need to be managed by you
+
 ### Performance considerations
 If you find this library too slow because is using `json.loads()` you can skip that by passing `skip_json_loads=True` to `repair_json`. Like:
 

{json_repair-0.12.3 → json_repair-0.13.1}/README.md

@@ -28,6 +28,33 @@ or just
 
     decoded_object = json_repair.repair_json(json_string, return_objects=True)
 
+### Read json from a file or file descriptor
+
+JSON repair provides also a drop-in replacement for `json.load()`:
+
+    import json_repair
+
+    try:
+        file_descriptor = open(fname, 'rb')
+    except OSError:
+        ...
+
+    with file_descriptor:
+        decoded_object = json_repair.load(file_descriptor)
+
+and another method to read from a file:
+
+    import json_repair
+
+    try:
+        decoded_object = json_repair.from_file(json_file)
+    except OSError:
+        ...
+    except IOError:
+        ...
+
+Keep in mind that the library will not catch any IO-related exception and those will need to be managed by you
+
 ### Performance considerations
 If you find this library too slow because is using `json.loads()` you can skip that by passing `skip_json_loads=True` to `repair_json`. Like:
 

{json_repair-0.12.3 → json_repair-0.13.1}/pyproject.toml

@@ -3,7 +3,7 @@ requires = ["setuptools>=61.0"]
 build-backend = "setuptools.build_meta"
 [project]
 name = "json_repair"
-version = "0.12.3"
+version = "0.13.1"
 license = {file = "LICENSE"}
 authors = [
   { name="Stefano Baccianella", email="4247706+mangiucugna@users.noreply.github.com" },

{json_repair-0.12.3 → json_repair-0.13.1}/src/json_repair/__init__.py

@@ -1,2 +1,4 @@
 from .json_repair import repair_json as repair_json
 from .json_repair import loads as loads
+from .json_repair import loads as load
+from .json_repair import loads as from_file

{json_repair-0.12.3 → json_repair-0.13.1}/src/json_repair/json_repair.py

@@ -23,7 +23,7 @@ All supported use cases are in the unit tests
 """
 
 import json
-from typing import Any, Dict, List, Union
+from typing import Any, Dict, List, Union, TextIO
 
 
 class JSONParser:
@@ -34,6 +34,7 @@ class JSONParser:
         self.index = 0
         # This is used in the object member parsing to manage the special cases of missing quotes in key or value
         self.context = ""
+        self.context_stack = []
 
     def parse(self) -> Union[Dict[str, Any], List[Any], str, float, int, bool, None]:
         return self.parse_json()
@@ -97,7 +98,7 @@ class JSONParser:
 
             # We are now searching for they string key
             # Context is used in the string parser to manage the lack of quotes
-            self.context = "object_key"
+            self.update_context("object_key")
 
             self.skip_whitespaces_at()
 
@@ -119,12 +120,13 @@ class JSONParser:
             if (self.get_char_at() or "") != ":":
                 self.insert_char_at(":")
             self.index += 1
-            self.context = "object_value"
+            self.update_context("")
+            self.update_context("object_value")
             # The value can be any valid json
             value = self.parse_json()
 
             # Reset context since our job is done
-            self.context = ""
+            self.update_context("")
             obj[key] = value
 
             if (self.get_char_at() or "") in [",", "'", '"']:
@@ -157,6 +159,9 @@ 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.context == "object_value" and char == "}":
+                break
 
         # Especially at the end of an LLM generated json you might miss the last "]"
         char = self.get_char_at()
@@ -166,6 +171,7 @@ class JSONParser:
                 # Remove trailing "," before adding the "]"
                 self.remove_char_at()
             self.insert_char_at("]")
+            self.index -= 1
 
         self.index += 1
         return arr
@@ -319,6 +325,19 @@ class JSONParser:
             except IndexError:
                 return
 
+    def update_context(self, value: str) -> None:
+        # If a value is provided update the context variable and save in stack
+        if value:
+            if self.context:
+                self.context_stack.append(self.context)
+            self.context = value
+        # Otherwise pop and update the context, or empty if the stack is empty
+        else:
+            try:
+                self.context = self.context_stack.pop()
+            except Exception:
+                self.context = ""
+
 
 def repair_json(
     json_str: str, return_objects: bool = False, skip_json_loads: bool = False
@@ -351,3 +370,17 @@ def loads(
     It is a wrapper around the `repair_json()` function with `return_objects=True`.
     """
     return repair_json(json_str, True)
+
+
+def load(fp: TextIO) -> Union[Dict[str, Any], List[Any], str, float, int, bool, None]:
+    return loads(fp.read())
+
+
+def from_file(
+    filename: str,
+) -> Union[Dict[str, Any], List[Any], str, float, int, bool, None]:
+    fd = open(filename)
+    jsonobj = load(fd)
+    fd.close()
+
+    return jsonobj

{json_repair-0.12.3 → json_repair-0.13.1/src/json_repair.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: json_repair
-Version: 0.12.3
+Version: 0.13.1
 Summary: A package to repair broken json strings
 Author-email: Stefano Baccianella <4247706+mangiucugna@users.noreply.github.com>
 License: MIT License
@@ -65,6 +65,33 @@ or just
 
     decoded_object = json_repair.repair_json(json_string, return_objects=True)
 
+### Read json from a file or file descriptor
+
+JSON repair provides also a drop-in replacement for `json.load()`:
+
+    import json_repair
+
+    try:
+        file_descriptor = open(fname, 'rb')
+    except OSError:
+        ...
+
+    with file_descriptor:
+        decoded_object = json_repair.load(file_descriptor)
+
+and another method to read from a file:
+
+    import json_repair
+
+    try:
+        decoded_object = json_repair.from_file(json_file)
+    except OSError:
+        ...
+    except IOError:
+        ...
+
+Keep in mind that the library will not catch any IO-related exception and those will need to be managed by you
+
 ### Performance considerations
 If you find this library too slow because is using `json.loads()` you can skip that by passing `skip_json_loads=True` to `repair_json`. Like:
 

{json_repair-0.12.3 → json_repair-0.13.1}/tests/test_json_repair.py

@@ -1,4 +1,4 @@
-from src.json_repair.json_repair import repair_json
+from src.json_repair.json_repair import from_file, repair_json
 
 
 def test_repair_json():
@@ -169,6 +169,25 @@ def test_repair_json_with_objects():
 
     #Test markdown stupidities from ChatGPT
     assert repair_json('{ "content": "[LINK]("https://google.com")" }', True) == { "content": "[LINK](\"https://google.com\")"}
+    assert repair_json('''
+{
+  "resourceType": "Bundle",
+  "id": "1",
+  "type": "collection",
+  "entry": [
+    {
+      "resource": {
+        "resourceType": "Patient",
+        "id": "1",
+        "name": [
+          {"use": "official", "family": "Corwin", "given": ["Keisha", "Sunny"], "prefix": ["Mrs."},
+          {"use": "maiden", "family": "Goodwin", "given": ["Keisha", "Sunny"], "prefix": ["Mrs."]}
+        ]
+      }
+    }
+  ]
+}
+''', True) == {"resourceType": "Bundle", "id": "1", "type": "collection", "entry": [{"resource": {"resourceType": "Patient", "id": "1", "name": [{"use": "official", "family": "Corwin", "given": ["Keisha", "Sunny"], "prefix": ["Mrs."]}, {"use": "maiden", "family": "Goodwin", "given": ["Keisha", "Sunny"], "prefix": ["Mrs."]}]}}]}
 
 
 def test_repair_json_corner_cases_generate_by_gpt():
@@ -240,4 +259,21 @@ def test_repair_json_skip_json_loads():
     assert repair_json('{"key": true, "key2": false, "key3": null}', skip_json_loads=True) == '{"key": true, "key2": false, "key3": null}'
     assert repair_json('{"key": true, "key2": false, "key3": null}', return_objects=True, skip_json_loads=True) == {"key": True, "key2": False, "key3": None}
     assert repair_json('{"key": true, "key2": false, "key3": }', skip_json_loads=True) == '{"key": true, "key2": false, "key3": ""}'
-    assert repair_json('{"key": true, "key2": false, "key3": }', return_objects=True, skip_json_loads=True) == {"key": True, "key2": False, "key3": ""}
+    assert repair_json('{"key": true, "key2": false, "key3": }', return_objects=True, skip_json_loads=True) == {"key": True, "key2": False, "key3": ""}
+
+def test_repair_json_from_file():
+    import os
+    import tempfile
+
+    # Create a temporary file
+    temp_fd, temp_path = tempfile.mkstemp(suffix=".json")
+    try:
+        # Write content to the temporary file
+        with os.fdopen(temp_fd, 'w') as tmp:
+            tmp.write("{")
+        
+        assert(from_file(temp_path)) == {}
+        
+    finally:
+        # Clean up - delete the temporary file
+        os.remove(temp_path)

{json_repair-0.12.3 → json_repair-0.13.1}/tests/test_performance.py

@@ -580,7 +580,7 @@ def test_true_false_correct(benchmark):
   mean_time = benchmark.stats.get("median")
 
   # Define your time threshold in seconds
-  max_time = 23 * (1 / 10 ** 6)  # 23 microsecond
+  max_time = 24 * (1 / 10 ** 6)  # 24 microsecond
 
   # 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"