cleaning

Author: crisely09

python/mlcroissant/mlcroissant/_src/core/optional.py

@@ -91,11 +91,6 @@ def librosa(cls) -> types.ModuleType:  # pylint: disable=invalid-name
         """Cached librosa module."""
         return _try_import("librosa", package_name="librosa")
 
-    @cached_class_property
-    def orjson(cls) -> types.ModuleType:  # pylint: disable=invalid-name
-        """Cached orjson module."""
-        return _try_import("orjson", package_name="orjson")
-
     @cached_class_property
     def scipy(cls) -> types.ModuleType:  # pylint: disable=invalid-name
         """Cached scipy module."""

python/mlcroissant/mlcroissant/_src/operation_graph/operations/parse_json.py

@@ -9,11 +9,6 @@
 from mlcroissant._src.structure_graph.nodes.field import Field
 from mlcroissant._src.structure_graph.nodes.source import FileProperty
 
-try:
-    orjson = deps.orjson
-except ModuleNotFoundError:
-    orjson = None
-
 
 def _unwrap_single_item(value: Any) -> Any:
     """Unwraps a single-item list to its value, or returns the value as is."""
@@ -99,7 +94,7 @@ def parse(self, fh: TextIO) -> pd.DataFrame:
         """
         # Load entire JSON file (could be a list or a single dict).
         raw = fh.read()
-        data = orjson.loads(raw) if orjson else json.loads(raw)
+        data = json.loads(raw)
 
         # Always treat as list of records.
         records = data if isinstance(data, list) else [data]
@@ -137,7 +132,7 @@ def raw(self, fh: TextIO) -> pd.DataFrame:
         """
         # Raw JSON fallback: one‐cell DataFrame.
         raw = fh.read()
-        content = orjson.loads(raw) if orjson else json.loads(raw)
+        content = json.loads(raw)
         return pd.DataFrame({FileProperty.content: [content]})
 
 
@@ -183,9 +178,9 @@ def __init__(self, fields, validate_fhir: bool = False):
 
         # Add FHIR validator if needed
         if validate_fhir:
-            from mlcroissant._src.operation_graph.operations.fhir_validator import FhirValidator
+            from mlcroissant._src.operation_graph.operations import fhir_validator
 
-            self.fhir_validator = FhirValidator(validate_fhir=True)
+            self.fhir_validator = fhir_validator.FhirValidator(validate_fhir=True)
         else:
             self.fhir_validator = None
 
@@ -211,7 +206,7 @@ def parse(self, fh):
             line = line.strip()
             if not line:
                 continue
-            rec = orjson.loads(line) if orjson else json.loads(line)
+            rec = json.loads(line)
             # Optional FHIR validation
             if self.fhir_validator:
                 rec = self.fhir_validator.validate_resource(rec)

python/mlcroissant/mlcroissant/_src/operation_graph/operations/parse_json_test.py

@@ -56,16 +56,14 @@ def test_jsonreader_parse():
 
 
 def test_jsonreader_parse_deep():
-    import orjson
-
     # Test nested JSONPath ($.level1.level2[*].value)
     field = create_test_field(
         source=Source(extract=Extract(json_path="$.level1.level2[*].value"))
     )
     fields = (field,)
     json_obj = {"level1": {"level2": [{"value": 100}, {"value": 200}]}}
     expected_df = pd.DataFrame({"$.level1.level2[*].value": [[100, 200]]})
-    raw_str = orjson.dumps(json_obj).decode("utf-8")
+    raw_str = json.dumps(json_obj)
     fh = io.StringIO(raw_str)
     reader = JsonReader(fields=fields)
     df = reader.parse(fh)

python/mlcroissant/mlcroissant/_src/operation_graph/operations/read.py

@@ -188,9 +188,7 @@ def _read_file_content(
                         and os.getenv("CROISSANT_VALIDATE_FHIR", "").lower()
                         in ("1", "true")
                     )
-                    reader = JsonlReader(
-                        self.fields, validate_fhir=validate_fhir
-                    )
+                    reader = JsonlReader(self.fields, validate_fhir=validate_fhir)
                     if reading_method == ReadingMethod.JSON:
                         return reader.parse(file)
                     return reader.raw(file)

python/mlcroissant/mlcroissant/_src/structure_graph/nodes/source.py

@@ -6,9 +6,9 @@
 from rdflib.namespace import SDO
 
 from mlcroissant._src.core import constants
-from mlcroissant._src.core.optional import deps
 from mlcroissant._src.core import dataclasses as mlc_dataclasses
 from mlcroissant._src.core.context import CroissantVersion
+from mlcroissant._src.core.optional import deps
 from mlcroissant._src.core.uuid import formatted_uuid_to_json
 from mlcroissant._src.core.uuid import uuid_from_jsonld
 from mlcroissant._src.structure_graph.base_node import Node

python/mlcroissant/pyproject.toml

@@ -28,7 +28,6 @@ dependencies = [
   "pandas",
   "pandas-stubs",
   "python-dateutil",
-  "orjson",
   "rdflib",
   "requests",
   "scipy",
@@ -104,7 +103,6 @@ module = [
   "jsonpath_rw",
   "librosa",
   "networkx",
-  "orjson",
   "pandas",
   "pillow",
   "scipy",