48 lines
1.5 KiB
Python
48 lines
1.5 KiB
Python
|
import re
|
||
|
|
import os
|
||
|
|
from random import randrange
|
||
|
|
|
||
|
|
def new_upload(output_file):
|
||
|
|
"""
|
||
|
|
Extracts 1 to 3 sentences from input_file and writes them to output_file.
|
||
|
|
"""
|
||
|
|
num_sentences = randrange(1, 4)
|
||
|
|
input_file = '../HP1.txt'
|
||
|
|
index_file = '../read_index.txt'
|
||
|
|
|
||
|
|
def read_sentences(file_path, start_index, num_sentences):
|
||
|
|
sentences = []
|
||
|
|
with open(file_path, 'r') as file:
|
||
|
|
text = file.read()
|
||
|
|
sentences = re.split(r'(?<!\w\.\w.)(?<![A-Z][a-z]\.)(?<=\.|\?)\s', text)
|
||
|
|
return sentences[start_index:start_index + num_sentences], start_index + num_sentences
|
||
|
|
|
||
|
|
def write_sentences(sentences, output_path):
|
||
|
|
with open(output_path, 'w') as file:
|
||
|
|
for sentence in sentences:
|
||
|
|
file.write(sentence + ' ')
|
||
|
|
|
||
|
|
def read_index(index_file):
|
||
|
|
if not os.path.exists(index_file):
|
||
|
|
return 0
|
||
|
|
with open(index_file, 'r') as file:
|
||
|
|
return int(file.read().strip())
|
||
|
|
|
||
|
|
def write_index(index_file, index):
|
||
|
|
with open(index_file, 'w') as file:
|
||
|
|
file.write(str(index))
|
||
|
|
|
||
|
|
# Read the current index
|
||
|
|
start_index = read_index(index_file)
|
||
|
|
|
||
|
|
# Read 1 to 3 sentences from the input file
|
||
|
|
sentences, new_index = read_sentences(input_file, start_index, num_sentences)
|
||
|
|
|
||
|
|
# Write the sentences to the output file
|
||
|
|
write_sentences(sentences, output_file)
|
||
|
|
|
||
|
|
# Write the new index back to the index file
|
||
|
|
write_index(index_file, new_index)
|
||
|
|
|
||
|
|
print(f"Successfully wrote {len(sentences)} sentence(s) to {output_file}.")
|