63 lines
2.1 KiB
Python
63 lines
2.1 KiB
Python
""" این روال جهت تبدیل فایل جیسون حاوی قوانین بازنویسی شده به یک متن ایجاد شده است """
|
|
from html import escape
|
|
from datetime import datetime
|
|
import time
|
|
import os
|
|
|
|
from funcs import read_from_json, save_to_file_by_address
|
|
|
|
|
|
counter = 0
|
|
total = 0
|
|
remained = 0
|
|
id = ''
|
|
|
|
if __name__ == "__main__":
|
|
|
|
base_address = os.getcwd()
|
|
#base_address = "/home/gpu/tnlp/jokar/llama" # terminal
|
|
|
|
sections_110_address = base_address + "/data/simplized_sentences_110_2.json"
|
|
|
|
datalist_110 = read_from_json(sections_110_address)
|
|
# import numpy as np
|
|
# np_sections = np.array(datalist_110)
|
|
start_time = time.time()
|
|
|
|
finall_text = ""
|
|
try:
|
|
|
|
for i , line in enumerate(datalist_110):
|
|
content_texts = ''
|
|
id = line["id"]
|
|
qanon_id = line["qanon_id"]
|
|
content_list = line["result"]
|
|
for item in content_list:
|
|
text = item.replace("*", "")
|
|
text = text.strip()
|
|
if text.endswith(" .."):
|
|
text = text.replace(" ..", ". ")
|
|
elif text.endswith(".."):
|
|
text = text.replace("..", ".\n")
|
|
content_texts += "".join(text)
|
|
finall_text += "".join(f"{content_texts}\nid:[{id}]\n\n")
|
|
print(i+1)
|
|
|
|
|
|
destination_address = base_address + "/data/simplized_sentences_110_02.txt"
|
|
destination_address2 = base_address + "/data/address_link.txt"
|
|
save_to_file_by_address(destination_address, finall_text)
|
|
|
|
|
|
|
|
except Exception as inst:
|
|
print(type(inst)) # the exception type
|
|
print(inst.args) # arguments stored in .args
|
|
print(inst) # __str__ allows args to be printed directly,
|
|
# but may be overridden in exception subclasses
|
|
print("Exception:=> %s -> %.2f " % (id , counter / total))
|
|
|
|
|
|
end_time = time.time()
|
|
print(f"elapsed time: {end_time-start_time}")
|
|
print(" *** finished! *** ") |