modify_codes.py

import csv
import random
import string

def generate_code():
    # Generate 8-character code: 5 letters and 3 numbers in specific positions
    letters = ''.join(random.choices(string.ascii_uppercase, k=5))
    numbers = ''.join(random.choices(string.digits, k=3))
    code = letters[0:2] + numbers[0] + letters[2] + numbers[1] + letters[3] + numbers[2] + letters[4]
    return code

def generate_study_name(number):
    # Generate study name with leading zeros
    return f"study-{str(number).zfill(7)}"

# Read and write the CSV file
with open('my_prolific_submission_history.csv', 'r', encoding='utf-8') as input_file:
    # Read all rows
    rows = list(csv.reader(input_file))

    # Get header row
    header = rows[0]

    # Find the indices of the Study and Completion Code columns
    study_index = header.index('Study')
    completion_code_index = header.index('Completion Code')

    # Process each data row
    study_counter = 1
    for row in rows[1:]:  # Skip header row
        # Replace study name if it exists
        if row[study_index].strip():
            row[study_index] = generate_study_name(study_counter)
            study_counter += 1

        # Replace completion code if it exists
        if row[completion_code_index].strip():
            row[completion_code_index] = generate_code()

# Write the modified data back to a new CSV file
with open('my_prolific_submission_history_modified.csv', 'w', newline='', encoding='utf-8') as output_file:
    writer = csv.writer(output_file)
    writer.writerows(rows)