sums.py

import csv
import time
from openai import OpenAI
client = OpenAI()

# Define the input and output file names
input_folder = 'input.csv'
output_file_name = 'output.csv'
input_file_name = 'input.csv'


def get_completion(prompt, model="gpt-3.5-turbo"):
    messages = [{"role": "user", "content": prompt}]
            
    retry_limit = 3
    retry_interval = 60
    retry_count = 0
    #limit size of prompt
    while retry_count < retry_limit:
        try:
            response = client.chat.completions.create(
                model=model,
                messages=messages,
                max_tokens = 3000,
                temperature=0, # this is the degree of randomness of the model's output
            )
            retry_count = retry_limit
            
        except Exception as e:
            print(f"An error occurred: {str(e)}")
            retry_count += 1
            if retry_count < retry_limit:
                print(f"Retrying in {retry_interval} seconds...")
                time.sleep(retry_interval)
            else:
                print("Exceeded retry limit. Skipping...")
                response = "ERROR"
    #print(response.choices[0].message)
    return response.choices[0].message.content

def make_summary(stream):
    prompt = f"""
    Read the following transcript and tell me, yes or no, if \
    the transcript is likely an interview with a journalist \
    and a medical expert.<transcript>{stream[:3000]}<transcript> \
    """
    summary = ""
    summary = get_completion(prompt)
    return summary



def process_csv(encoding):
    with open(input_file_name, mode='r', newline='', encoding=encoding) as infile:
        reader = csv.DictReader(infile)
        rows_with_counts = []
        this_record = ""
        for row in reader:
            #print(row)
            try:
                summary = make_summary(row['SegmentTranscript'])
                # Add the summary to the row
                
                #Lets see a print out
                this_record = f"{row['Date']} | {row['Network']} | "
                print(f"{this_record}{summary}")
                
                row['summary'] = summary
                rows_with_counts.append(row)
            except UnicodeDecodeError as e:
                print(f"Error reading row {reader.line_num}: {e}. This row will be skipped.")
        return rows_with_counts, reader.fieldnames

# Try reading with utf-8, if it fails, try with latin1
try:
    rows_with_counts, fieldnames = process_csv('utf-8')
except UnicodeDecodeError:
    print("Failed to read the file with utf-8 encoding, trying with latin1...")
    rows_with_counts, fieldnames = process_csv('latin1')

# Define the field names for the output CSV
fieldnames = fieldnames + ['summary']

# Open the output CSV file and write the data with the additional 'count' column
with open(output_file_name, mode='w', newline='', encoding='utf-8') as outfile:
    writer = csv.DictWriter(outfile, fieldnames=fieldnames)
    writer.writeheader()
    writer.writerows(rows_with_counts)

print(f'Process completed. The output has been saved in "{output_file_name}"')