se.py

import json
from tqdm import tqdm
import time
from gpt_inf_se import se, se_inte, se_reas, se_sl

def main():
    total_inference_time = 0.0
    with open("data/java_xl.json", "r") as file:
        data = json.load(file)

    with open("last/sim_java_xl_v.json", "r") as file:
        data_sim = json.load(file)


    code_explanation_expert = {
        "role": "system", 
        "content": "You are an expert in code clone detection"
    }
    
    for i in tqdm(range(len(data)), desc="data processing"):
        inference_start_time = time.time()
        success = False
        while not success:
            try:
                codeA = data[i]['codeA']
                codeB = data[i]['codeB']
                data[i]['sim'] = data_sim[i]['output']
                data[i]['inte'] = se_inte(codeA= codeA, codeB=codeB, context=code_explanation_expert)
                data[i]['reas'] = se_reas(codeA= codeA, codeB=codeB, context=code_explanation_expert)
                data[i]['sl'] = se_sl(codeA= codeA, codeB=codeB, context=code_explanation_expert)
                data[i]['result'] = se(sim_= data[i]['sim'], reas_= data[i]['reas'], sl_= data[i]['sl'], inte_= data[i]['inte'], context=code_explanation_expert)

                with open(f"tmp/se_java_xl.json", 'w') as tp_file:
                    json.dump(data, tp_file, indent=4)

                success = True 
                inference_time = time.time() - inference_start_time
                total_inference_time += inference_time

            except Exception as e:
                time.sleep(30)
                print(f"Error processing index {i}: {e}")
                with open(f'error/e{i}.json', 'w') as err:
                    json.dump(data, err, indent=4)
                
    
    with open('results/se_java_xl.json', 'w') as fichier_sortie:
        json.dump(data, fichier_sortie, indent=4)
    print(f"Average Inference Time: {total_inference_time} seconds")


if __name__ == "__main__":
    main()