trailpack.packing.export_service

Export service for converting UI data to Frictionless Data Package in Parquet.

Classes

DataPackageExporter

Service for exporting UI data to Frictionless Data Package in Parquet.

Module Contents

class trailpack.packing.export_service.DataPackageExporter(df: pandas.DataFrame, column_mappings: Dict[str, str], general_details: Dict[str, Any], sheet_name: str, file_name: str, suggestions_cache: Dict[str, List] = None, column_descriptions: Dict[str, str] = None, standard_version: str = '1.0.0')[source]

Service for exporting UI data to Frictionless Data Package in Parquet.

Initialize with UI session state data.

Parameters:

df – Pandas DataFrame with the actual data
column_mappings – Mapping of column names to PyST concept IDs
general_details – Metadata from the general details form
sheet_name – Name of the Excel sheet
file_name – Original file name
suggestions_cache – Cache of PyST suggestions with id and label
column_descriptions – User-provided descriptions/comments for columns
standard_version – Trailpack standard version to validate against

_find_label_for_id(concept_id: str) → str | None[source]: Find label for a PyST concept ID from suggestions cache.

_format_validation_errors(validation_result) → str[source]: Format validation errors for better readability.

_infer_field_type(series: pandas.Series) → str[source]: Infer Frictionless field type from pandas Series.

_sanitize_resource_name(name: str) → str[source]

Sanitize resource name to match the pattern ^[a-z0-9-_.]+$.

The resource name must only contain: - Lowercase letters (a-z) - Numbers (0-9) - Hyphens (-) - Underscores (_) - Dots (.)

Parameters:: name – Raw name string to sanitize
Returns:: Sanitized name matching the required pattern

_validate_dataframe_for_parquet(df: pandas.DataFrame) → None[source]

Validate DataFrame is compatible with Arrow/Parquet format.

Raises:: ValueError – If data quality issues are found (e.g., mixed types in columns)

build_fields() → List[trailpack.packing.datapackage_schema.Field][source]: Convert column mappings to Field definitions.

build_metadata(resource: trailpack.packing.datapackage_schema.Resource) → Dict[str, Any][source]: Build complete metadata using MetaDataBuilder.

build_resource(fields: List[trailpack.packing.datapackage_schema.Field]) → trailpack.packing.datapackage_schema.Resource[source]: Create Resource definition with fields.

export(output_path: str, validate_standard: bool = True) → Tuple[str, str | None, Any | None][source]

Execute full export workflow and write Parquet.

Parameters:

output_path – Path where Parquet file will be written
validate_standard – Whether to validate against Trailpack standard (default: True)

Returns:

Tuple of (output_path, quality_level, validation_result) - output_path: Path to exported Parquet file - quality_level: Validation level (“STRICT”, “STANDARD”, “BASIC”, “INVALID”) or None if validation skipped - validation_result: Full ValidationResult object for report generation, or None if validation skipped

Raises: