Delta-Vector
commited on
Commit
•
fe2a73e
1
Parent(s):
b209c86
Delete thing.py
Browse files
thing.py
DELETED
@@ -1,64 +0,0 @@
|
|
1 |
-
import os
|
2 |
-
import re
|
3 |
-
import json
|
4 |
-
import uuid
|
5 |
-
def process_conversation(file_path):
|
6 |
-
# Read the TXT file
|
7 |
-
with open(file_path, "r", encoding="utf-8") as file:
|
8 |
-
lines = file.readlines()
|
9 |
-
# Store the conversation
|
10 |
-
conversation = []
|
11 |
-
is_user = True # To alternate between user ('human') and gpt
|
12 |
-
# Regex pattern to match speaker before the colon
|
13 |
-
pattern = re.compile(r"^(.*?):")
|
14 |
-
for line in lines:
|
15 |
-
line = line.strip() # Remove leading/trailing whitespace
|
16 |
-
if not line:
|
17 |
-
continue # Skip empty lines
|
18 |
-
# Match the speaker before the colon
|
19 |
-
match = pattern.match(line)
|
20 |
-
if match:
|
21 |
-
message_dict = {
|
22 |
-
"from": "human" if is_user else "gpt", # Use 'human' instead of 'user'
|
23 |
-
"value": line
|
24 |
-
}
|
25 |
-
conversation.append(message_dict)
|
26 |
-
# Alternate turns between 'human' and 'gpt'
|
27 |
-
is_user = not is_user
|
28 |
-
# Ensure the conversation ends with gpt's turn
|
29 |
-
if conversation and conversation[-1]["from"] == "human":
|
30 |
-
# If it ends on 'human', remove last entry
|
31 |
-
conversation.pop()
|
32 |
-
# Create structure with id and conversations
|
33 |
-
conversation_data = {
|
34 |
-
"id": str(uuid.uuid4()), # Generate a unique ID for the conversation
|
35 |
-
"conversations": conversation
|
36 |
-
}
|
37 |
-
return conversation_data
|
38 |
-
def write_json(conversation_data, output_path):
|
39 |
-
# Prepare the output list that wraps the conversation
|
40 |
-
output_json = [conversation_data]
|
41 |
-
# Write to JSON file, compact, all in single line
|
42 |
-
with open(output_path, "w", encoding="utf-8") as outfile:
|
43 |
-
json.dump(output_json, outfile, ensure_ascii=False, separators=(',', ':'))
|
44 |
-
def process_folder(input_folder, output_folder):
|
45 |
-
# Make sure the output folder exists
|
46 |
-
if not os.path.exists(output_folder):
|
47 |
-
os.makedirs(output_folder) # Create output folder if it doesn't exist
|
48 |
-
# Loop through all files in the input folder
|
49 |
-
for filename in os.listdir(input_folder):
|
50 |
-
if filename.endswith(".txt"): # Process only TXT files
|
51 |
-
input_file_path = os.path.join(input_folder, filename)
|
52 |
-
# Process the conversation
|
53 |
-
conversation_data = process_conversation(input_file_path)
|
54 |
-
# Construct a unique output JSON file name
|
55 |
-
output_file_name = f"{os.path.splitext(filename)[0]}.json" # Use same name, but with .json extension
|
56 |
-
output_file_path = os.path.join(output_folder, output_file_name)
|
57 |
-
# Write the JSON file
|
58 |
-
write_json(conversation_data, output_file_path)
|
59 |
-
print(f"Converted: {filename} --> {output_file_name}")
|
60 |
-
# Call the folder processing function
|
61 |
-
input_folder = r"C:\Users\User\Documents\datasets\Special- Chat (RP logs, convert to sharegpt)\todo" # Path to folder containing .txt files
|
62 |
-
output_folder = r"C:\Users\User\Documents\datasets\Special- Chat (RP logs, convert to sharegpt)\room\done" # Path to folder where JSONs will be saved
|
63 |
-
process_folder(input_folder, output_folder)
|
64 |
-
print(f"Conversion complete! JSONs saved to {output_folder}")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|