karthickg12 commited on
Commit
900dfe3
·
verified ·
1 Parent(s): 5dfef39

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +9 -9
app.py CHANGED
@@ -1,4 +1,5 @@
1
- import fitz
 
2
  import gradio as gr
3
  import torch
4
 
@@ -18,18 +19,17 @@ pipe = pipeline(
18
  print("Model loaded.")
19
 
20
 
21
- def extract_text(pdf_file):
22
 
 
23
  text = ""
24
 
25
- doc = fitz.open(pdf_file.name)
26
-
27
- for page in doc:
28
- text += page.get_text()
29
-
30
- doc.close()
31
 
32
- return text[:20000]
33
 
34
 
35
  def build_prompt(policy_text):
 
1
+
2
+ import pdfplumber
3
  import gradio as gr
4
  import torch
5
 
 
19
  print("Model loaded.")
20
 
21
 
 
22
 
23
+ def extract_text(pdf_file):
24
  text = ""
25
 
26
+ with pdfplumber.open(pdf_file.name) as pdf:
27
+ for page in pdf.pages:
28
+ page_text = page.extract_text()
29
+ if page_text:
30
+ text += page_text + "\n"
 
31
 
32
+ return text[:12000]
33
 
34
 
35
  def build_prompt(policy_text):