Thback commited on
Commit
2759478
ยท
1 Parent(s): 59b5f1e

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +19 -7
app.py CHANGED
@@ -26,13 +26,25 @@ def get_pdf_text(pdf_docs):
26
  # ์•„๋ž˜ ํ…์ŠคํŠธ ์ถ”์ถœ ํ•จ์ˆ˜๋ฅผ ์ž‘์„ฑ
27
 
28
  def get_text_file(text_docs):
29
- temp_dir = tempfile.TemporaryDirectory() # ์ž„์‹œ ๋””๋ ‰ํ† ๋ฆฌ๋ฅผ ์ƒ์„ฑํ•ฉ๋‹ˆ๋‹ค.
30
- temp_filepath = os.path.join(temp_dir.name, text_docs.name) # ์ž„์‹œ ํŒŒ์ผ ๊ฒฝ๋กœ๋ฅผ ์ƒ์„ฑํ•ฉ๋‹ˆ๋‹ค.
31
- with open(temp_filepath, "wb") as f: # ์ž„์‹œ ํŒŒ์ผ์„ ๋ฐ”์ด๋„ˆ๋ฆฌ ์“ฐ๊ธฐ ๋ชจ๋“œ๋กœ ์—ฝ๋‹ˆ๋‹ค.
32
- f.write(text_docs.getvalue()) # PDF ๋ฌธ์„œ์˜ ๋‚ด์šฉ์„ ์ž„์‹œ ํŒŒ์ผ์— ์”๋‹ˆ๋‹ค.
33
- text_loader = TextLoader(temp_filepath) # PyPDFLoader๋ฅผ ์‚ฌ์šฉํ•ด PDF๋ฅผ ๋กœ๋“œํ•ฉ๋‹ˆ๋‹ค.
34
- text_doc = text_loader.load() # ํ…์ŠคํŠธ๋ฅผ ์ถ”์ถœํ•ฉ๋‹ˆ๋‹ค.
35
- return text_doc # ์ถ”์ถœํ•œ ํ…์ŠคํŠธ๋ฅผ ๋ฐ˜ํ™˜ํ•ฉ๋‹ˆ๋‹ค.
 
 
 
 
 
 
 
 
 
 
 
 
36
 
37
 
38
 
 
26
  # ์•„๋ž˜ ํ…์ŠคํŠธ ์ถ”์ถœ ํ•จ์ˆ˜๋ฅผ ์ž‘์„ฑ
27
 
28
  def get_text_file(text_docs):
29
+ # ์ž„์‹œ ๋””๋ ‰ํ† ๋ฆฌ๋ฅผ ์ƒ์„ฑํ•ฉ๋‹ˆ๋‹ค.
30
+ temp_dir = tempfile.TemporaryDirectory()
31
+
32
+ # ์ž„์‹œ ํŒŒ์ผ ๊ฒฝ๋กœ๋ฅผ ์ƒ์„ฑํ•ฉ๋‹ˆ๋‹ค.
33
+ temp_filepath = os.path.join(temp_dir.name, "temp_file.txt")
34
+
35
+ # ํ…์ŠคํŠธ ๋ฌธ์„œ์˜ ๋‚ด์šฉ์„ ์ž„์‹œ ํŒŒ์ผ์— ์”๋‹ˆ๋‹ค.
36
+ with open(temp_filepath, "w", encoding="utf-8") as f:
37
+ f.write(text_docs.getvalue())
38
+
39
+ # ์ž„์‹œ ํŒŒ์ผ์—์„œ ํ…์ŠคํŠธ๋ฅผ ์ฝ์–ด์˜ต๋‹ˆ๋‹ค.
40
+ with open(temp_filepath, "r", encoding="utf-8") as f:
41
+ text_content = f.read()
42
+
43
+ # ์ž„์‹œ ๋””๋ ‰ํ† ๋ฆฌ๋ฅผ ์ž๋™์œผ๋กœ ์ •๋ฆฌํ•ฉ๋‹ˆ๋‹ค.
44
+ temp_dir.cleanup()
45
+
46
+ return [text_content] # ์ถ”์ถœํ•œ ํ…์ŠคํŠธ๋ฅผ ๋ฆฌ์ŠคํŠธ์— ๋‹ด์•„ ๋ฐ˜ํ™˜ํ•ฉ๋‹ˆ๋‹ค.
47
+
48
 
49
 
50