import csv

from rdflib import Graph
from rdflib.namespace import SKOS


csv.field_size_limit(131072 * 100)


def load_eurovoc_notations(graph_filename):
    g = Graph()
    g.parse(graph_filename, format='xml')
    print("Graph parsed")

    result = {}
    for concept, notation in g.subject_objects(SKOS.notation):
        result[str(notation)] = str(concept)
    # result = {}
    return result


def convert_ids_to_uris(ids):
    uris = []
    for id in ids:
        if len(id) == 6:
            uris.append(
                f'<http://eurovoc.europa.eu/{id.lstrip("0")}>'
            )

        else:
            mapped_uri = notatation_uri_map[id]
            if mapped_uri is not None:
                uris.append(
                    "<" + mapped_uri + ">"
                )
            else:
                print(f"Could not map notation {id} to uri")
    return uris


def cleanup(text):
    return " ".join(text.replace('"', ' ').split())


def convert_eurovoc_csv(input_file, output_file):
    """
    Convert CSV file with eurovoc subject URIs to proper format.

    Args:
        input_file (str): Path to input CSV file
        output_file (str): Path to output CSV file
    """

    # Print out header to CSV
    with open(output_file, 'w', newline='', encoding='utf-8') as outfile:
        print("document_id,url,date,type,title,text,subject_uris", file=outfile)


    # Read the input CSV and process each row
    with open(input_file, 'r', newline='', encoding='utf-8') as infile:
        reader = csv.DictReader(infile)

        for row in reader:
            # Create a new row with processed data
            processed_row = row.copy()

            # Quote the title and text fields
            processed_row['title'] = f'"{cleanup(row["title"])}"'
            processed_row['text'] = f'"{cleanup(row["text"])}"'

            # Process subject_uris - convert IDs to full URIs
            if row['subject_uris']:
                ids = row['subject_uris'].split(';')
                uris = convert_ids_to_uris(ids)
                processed_row['subject_uris'] =  '"' + ' '.join(uris) + '"'
            else:
                processed_row['subject_uris'] = '""'

            # Write to output CSV
            with open(output_file, 'a', newline='', encoding='utf-8') as outfile:
                print(",".join(processed_row.values()), file=outfile)

            # processed_rows.append(processed_row)
            # if len(processed_rows) >= 10:
            #     break


# Usage example:
if __name__ == "__main__":
    notatation_uri_map = load_eurovoc_notations("eurovoc-skos-ap-eu.rdf")
    convert_eurovoc_csv("european_parliament_eurovoc_2025_en.csv", 'european_parliament_eurovoc_2025_en_converted.csv')