fix(exceptions): rename InvalidDatasetError.format to data_format

.gitignore

@@ -240,3 +240,6 @@ tiktoken-docs.md
 config.yaml
 config.groq.yaml
 config.*.yaml
+
+# Generated eval report artifacts
+reports/

openagent_eval/datasets/csv_loader.py

@@ -65,7 +65,7 @@ def load(self, path: Path) -> Dataset:
             raise InvalidDatasetError(
                 message=f"Invalid CSV format: {e}",
                 dataset_path=str(path),
-                format="csv",
+                data_format="csv",
             ) from e
 
         return self._parse_rows(rows, path)

openagent_eval/datasets/factory.py

@@ -80,7 +80,7 @@ def _get_loader(config: DatasetConfig) -> BaseDatasetLoader:
         raise InvalidDatasetError(
             message=f"Unsupported dataset format: '{config.format}'. "
             f"Supported formats: {list(_FORMAT_MAP.keys())}",
-            format=config.format,
+            data_format=config.format,
         )
 
     # Guard against directory paths, which have no usable extension and

openagent_eval/datasets/hf_loader.py

@@ -69,7 +69,7 @@ def load(self, path: Path) -> Dataset:
                         "Install it with: pip install openagent-eval[datasets]"
                     ),
                     dataset_path=str(path),
-                    format="hf",
+                    data_format="hf",
                 ) from e
 
             try:
@@ -78,7 +78,7 @@ def load(self, path: Path) -> Dataset:
                 raise InvalidDatasetError(
                     message=f"Failed to load HuggingFace dataset: {e}",
                     dataset_path=str(path),
-                    format="hf",
+                    data_format="hf",
                 ) from e
 
             return self._parse_hf_dataset(hf_dataset, path)
@@ -153,15 +153,15 @@ def load_from_hub(
                     "The 'datasets' package is required for HuggingFace dataset loading. "
                     "Install it with: pip install openagent-eval[datasets]"
                 ),
-                format="hf",
+                data_format="hf",
             ) from e
 
         try:
             hf_dataset = load_dataset(dataset_name, split=split, **kwargs)
         except Exception as e:
             raise InvalidDatasetError(
                 message=f"Failed to load dataset '{dataset_name}': {e}",
-                format="hf",
+                data_format="hf",
             ) from e
 
         # Convert to list and apply limit

openagent_eval/datasets/json_loader.py

@@ -63,7 +63,7 @@ def load(self, path: Path) -> Dataset:
             raise InvalidDatasetError(
                 message=f"Invalid JSON in file: {e}",
                 dataset_path=str(path),
-                format="json",
+                data_format="json",
                 line_number=e.lineno,
             ) from e
 

openagent_eval/datasets/jsonl_loader.py

@@ -120,7 +120,7 @@ def _parse_lines(self, lines: list[str], path: Path) -> Dataset:
                 raise InvalidDatasetError(
                     message=f"Invalid JSON on line {line_num}: {e}",
                     dataset_path=str(path),
-                    format="jsonl",
+                    data_format="jsonl",
                     line_number=line_num,
                 ) from e
 

openagent_eval/datasets/pdf_loader.py

@@ -88,7 +88,7 @@ def load(self, path: Path) -> Dataset:
                     "Install it with: pip install openagent-eval[pdf]"
                 ),
                 dataset_path=str(path),
-                format="pdf",
+                data_format="pdf",
             ) from e
 
         try:
@@ -97,7 +97,7 @@ def load(self, path: Path) -> Dataset:
             raise InvalidDatasetError(
                 message=f"Failed to read PDF: {e}",
                 dataset_path=str(path),
-                format="pdf",
+                data_format="pdf",
             ) from e
 
         raw_items = self._extract_raw(reader, path)

openagent_eval/exceptions/dataset.py

@@ -57,15 +57,15 @@ class InvalidDatasetError(DatasetError):
     """Raised when a dataset has invalid format or structure.
 
     Attributes:
-        format: The detected or expected format of the dataset.
+        data_format: The detected or expected format of the dataset.
         line_number: The line number where the error occurred (if applicable).
     """
 
     def __init__(
         self,
         message: str,
         dataset_path: str | None = None,
-        format: str | None = None,
+        data_format: str | None = None,
         line_number: int | None = None,
         details: dict | None = None,
     ) -> None:
@@ -74,18 +74,18 @@ def __init__(
         Args:
             message: Human-readable error message.
             dataset_path: Path to the invalid dataset.
-            format: The detected or expected format.
+            data_format: The detected or expected format.
             line_number: The line number where the error occurred.
             details: Additional context about the error.
         """
         error_details = details or {}
-        if format:
-            error_details["format"] = format
+        if data_format:
+            error_details["format"] = data_format
         if line_number:
             error_details["line_number"] = line_number
 
         super().__init__(message=message, dataset_path=dataset_path, details=error_details)
-        self.format = format
+        self.data_format = data_format
         self.line_number = line_number
 
 

tests/unit/test_exceptions.py

@@ -85,8 +85,8 @@ def test_not_found_error(self) -> None:
 
     def test_invalid_dataset_error(self) -> None:
         """Test invalid dataset error."""
-        error = InvalidDatasetError("Invalid format", format="json", line_number=10)
-        assert error.format == "json"
+        error = InvalidDatasetError("Invalid format", data_format="json", line_number=10)
+        assert error.data_format == "json"
         assert error.line_number == 10
 
     def test_validation_error(self) -> None: