tlsgy commited on
Commit
e8d3bfa
Β·
1 Parent(s): 0952935
Files changed (1) hide show
  1. app.py +26 -3
app.py CHANGED
@@ -29,14 +29,37 @@ def get_pdf_text(pdf_docs):
29
  # μ•„λž˜ ν…μŠ€νŠΈ μΆ”μΆœ ν•¨μˆ˜λ₯Ό μž‘μ„±
30
 
31
  def get_text_file(docs):
32
- pass
 
 
 
 
 
 
33
 
34
 
35
  def get_csv_file(docs):
36
- pass
 
 
 
 
 
 
37
 
38
  def get_json_file(docs):
39
- pass
 
 
 
 
 
 
 
 
 
 
 
40
 
41
 
42
  # λ¬Έμ„œλ“€μ„ μ²˜λ¦¬ν•˜μ—¬ ν…μŠ€νŠΈ 청크둜 λ‚˜λˆ„λŠ” ν•¨μˆ˜μž…λ‹ˆλ‹€.
 
29
  # μ•„λž˜ ν…μŠ€νŠΈ μΆ”μΆœ ν•¨μˆ˜λ₯Ό μž‘μ„±
30
 
31
  def get_text_file(docs):
32
+ temp_dir = tempfile.TemporaryDirectory() # μž„μ‹œ 디렉토리λ₯Ό μƒμ„±ν•©λ‹ˆλ‹€.
33
+ temp_filepath = os.path.join(temp_dir.name, docs.name) # μž„μ‹œ 파일 경둜λ₯Ό μƒμ„±ν•©λ‹ˆλ‹€.
34
+ with open(temp_filepath, "wb") as f: # μž„μ‹œ νŒŒμΌμ„ λ°”μ΄λ„ˆλ¦¬ μ“°κΈ° λͺ¨λ“œλ‘œ μ—½λ‹ˆλ‹€.
35
+ f.write(docs.getvalue())
36
+ loader = TextLoader(temp_filepath)
37
+ data = loader.load()
38
+ return data
39
 
40
 
41
  def get_csv_file(docs):
42
+ temp_dir = tempfile.TemporaryDirectory() # μž„μ‹œ 디렉토리λ₯Ό μƒμ„±ν•©λ‹ˆλ‹€.
43
+ temp_filepath = os.path.join(temp_dir.name, docs.name) # μž„μ‹œ 파일 경둜λ₯Ό μƒμ„±ν•©λ‹ˆλ‹€.
44
+ with open(temp_filepath, "wb") as f: # μž„μ‹œ νŒŒμΌμ„ λ°”μ΄λ„ˆλ¦¬ μ“°κΈ° λͺ¨λ“œλ‘œ μ—½λ‹ˆλ‹€.
45
+ f.write(docs.getvalue())
46
+ loader = CSVLoader(file_path=temp_filepath)
47
+ data = loader.load()
48
+ return data
49
 
50
  def get_json_file(docs):
51
+ temp_dir = tempfile.TemporaryDirectory() # μž„μ‹œ 디렉토리λ₯Ό μƒμ„±ν•©λ‹ˆλ‹€.
52
+ temp_filepath = os.path.join(temp_dir.name, docs.name) # μž„μ‹œ 파일 경둜λ₯Ό μƒμ„±ν•©λ‹ˆλ‹€.
53
+ with open(temp_filepath, "wb") as f: # μž„μ‹œ νŒŒμΌμ„ λ°”μ΄λ„ˆλ¦¬ μ“°κΈ° λͺ¨λ“œλ‘œ μ—½λ‹ˆλ‹€.
54
+ f.write(docs.getvalue())
55
+ loader = JSONLoader(
56
+ file_path=temp_filepath,
57
+ jq_schema='.messages[].content',
58
+ text_content=False
59
+ )
60
+ data = loader.load()
61
+ return data
62
+
63
 
64
 
65
  # λ¬Έμ„œλ“€μ„ μ²˜λ¦¬ν•˜μ—¬ ν…μŠ€νŠΈ 청크둜 λ‚˜λˆ„λŠ” ν•¨μˆ˜μž…λ‹ˆλ‹€.