FAIRLinked.RDFTableConversion package

Submodules

FAIRLinked.RDFTableConversion.csv_to_jsonld_mapper module

FAIRLinked.RDFTableConversion.csv_to_jsonld_template_filler module

FAIRLinked.RDFTableConversion.csv_to_jsonld_template_filler.extract_data_from_csv(metadata_template, csv_file, row_key_cols, orcid, output_folder, prop_column_pair_dict=None, ontology_graph=None, base_uri='https://cwrusdle.bitbucket.io/mds/')[source]

Converts CSV rows into RDF graphs using a JSON-LD template and optional property mapping, writing JSON-LD files.

Parameters:

  • metadata_template (dict) – JSON-LD template with “@context” and “@graph”.

  • csv_file (str) – Path to the input CSV.

  • row_key_cols (list[str]) – Columns to uniquely identify each row.

  • orcid (str) – ORCID identifier (dashes removed automatically).

  • output_folder (str) – Directory to save JSON-LD files.

  • prop_column_pair_dict (dict or None, optional) – Maps property keys to (subject_column, object_column) column pairs. If None or empty, no properties are added.

  • ontology_graph (RDFLib Graph object or None, optional) – Ontology for property type/URI resolution. Required if prop_column_pair_dict is provided.

  • base_uri (str, optional) – Base URI used to construct subject and object URIs.

Returns:

List of RDFLib Graphs, one per row.

Return type:

List[rdflib.Graph]

FAIRLinked.RDFTableConversion.csv_to_jsonld_template_filler.extract_from_folder(csv_folder, metadata_template, row_key_cols, orcid, output_base_folder, prop_column_pair_dict=None, ontology_graph=None, base_uri='https://cwrusdle.bitbucket.io/mds/')[source]

Processes all CSV files in a folder and converts each into RDF/JSON-LD files using a metadata template and optional object/datatype property mappings.

Parameters:

  • csv_folder (str) – Path to the folder containing CSV files.

  • metadata_template (dict) – JSON-LD metadata template with “@context” and “@graph” describing the RDF structure.

  • row_key_cols (list[str]) – List of CSV column names used to construct a unique key for each row.

  • orcid (str) – ORCID iD of the user (dashes will be removed automatically).

  • output_base_folder (str) – Directory where output subfolders (one per CSV) will be created for JSON-LD files.

  • prop_column_pair_dict (dict or None, optional) – Mapping from property key (e.g., predicate label) to list of (subject_column, object_column) tuples. These define additional object or datatype properties to inject based on CSV columns. If None, no extra connections are added.

  • ontology_graph (str or None, optional) – RDFLib graph object of ontology from which property URIs and types are resolved. Required only if prop_column_pair_dict is given.

  • base_uri (str, optional) – Base URI used to construct RDF subject and object URIs. Defaults to the CWRU MDS base.

Returns:

Writes JSON-LD files to disk. No return value.

Return type:

None

FAIRLinked.RDFTableConversion.csv_to_jsonld_template_filler.generate_prop_metadata_dict(ontology_graph)[source]

Generates a dictionary where the keys are human-readable labels of object/datatype properties, and the values are 2-tuples that contain the URI of that property in the first entry and the type (object/datatype) in second entry.

Parameters:

ontology_graph (RDFLib graph object of the ontology) – Path to the RDF/OWL ontology file.

Returns:

Dictionary of the form: {

”has material”: (”http://example.org/ontology#hasMaterial”, “Object Property”), “has value”: (”http://example.org/ontology#hasValue”, “Datatype Property”), …

}

Return type:

dict

FAIRLinked.RDFTableConversion.jsonld_batch_converter module

FAIRLinked.RDFTableConversion.jsonld_batch_converter.extract_row_from_jsonld(data: dict, filename: str)[source]

Extracts a single row of data from a JSON-LD object.

Parameters:

  • data (dict) – JSON-LD data parsed as dictionary.

  • filename (str) – Name of the file the data came from.

Returns:

(row_values, fair_types, units) as dictionaries.

Return type:

tuple

FAIRLinked.RDFTableConversion.jsonld_batch_converter.jsonld_directory_to_csv(input_dir: str, output_basename: str = 'merged_output', output_dir: str = 'outputs')[source]

Converts a directory of JSON-LD files into a tabular format (CSV, Parquet, Arrow). Each row represents a JSON-LD file with:

  • Column headers from skos:altLabel

  • Values from qudt:value

  • Extra header rows for FAIR type (@type) and units (qudt:hasUnit)

Parameters:

  • input_dir (str) – Directory containing JSON-LD files.

  • output_basename (str) – Base name for output files.

  • output_dir (str) – Output directory to save CSV, Parquet, and Arrow files.

Module contents