Use OpenAI to resort genres from a tree

sort_genres.py

import json
import csv
import openai

openai.api_key = ''


def map_collected_genres(collected_genres, verified_genres):
    response = openai.ChatCompletion.create(
        model="gpt-3.5-turbo",
        messages=[{
            "role":
            "system",
            "content":
            f"You will map the given genre to it's closest matching genre from given array below and reply in json mapping {verified_genres}"
        }, {
            "role": "user",
            "content": f"map this genres {collected_genres}"
        }])
    verified_genre = response.choices[0].message.content

    return verified_genre


if __name__ == "__main__":
    collected_genres = set()
    book_data = []

    with open('files_seeder/genres.json', 'r') as f:
        genres = json.load(f)

    genre_in_trees = []
    for key, value in genres.items():
        genre_in_trees.append(key)
        genre_in_trees = genre_in_trees + value

    with open('genres_assigned.csv', 'r') as f:
        csv_reader = csv.reader(f)
        for row in [row for row in csv_reader][1:]:
            book_data.append({
                'barcode': row[0],
                'name': row[1],
                'collected_genres': [],
                'verified_genres': []
            })
            for genre in row[2:]:
                if len(genre) > 2:
                    collected_genres.add(genre)
                    book_data[-1]['collected_genres'].append(genre)
    print(len(collected_genres))
    need_verification = list(collected_genres - set(genre_in_trees))

    verification_chunks = []
    for i in range(0, len(need_verification), 50):
        verification_chunks.append(need_verification[i:i + 50])
    returned_mapping = []
    for index, chunk in enumerate(verification_chunks):
        print(f'{index}, {len(verification_chunks)}', end='\r')
        returned_mapping.append(map_collected_genres(chunk, genre_in_trees))