Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,14 +1,19 @@
|
|
| 1 |
import streamlit as st
|
| 2 |
import PyPDF2
|
|
|
|
|
|
|
| 3 |
|
| 4 |
-
def extract_text_from_pdf(
|
| 5 |
text = ""
|
| 6 |
-
with
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
|
| 11 |
-
|
|
|
|
|
|
|
|
|
|
| 12 |
return text
|
| 13 |
|
| 14 |
def main():
|
|
|
|
| 1 |
import streamlit as st
|
| 2 |
import PyPDF2
|
| 3 |
+
import tempfile
|
| 4 |
+
import os
|
| 5 |
|
| 6 |
+
def extract_text_from_pdf(uploaded_file):
|
| 7 |
text = ""
|
| 8 |
+
with tempfile.NamedTemporaryFile(delete=False) as tmp_file:
|
| 9 |
+
tmp_file.write(uploaded_file.read())
|
| 10 |
+
tmp_file.seek(0)
|
| 11 |
+
with open(tmp_file.name, "rb") as f:
|
| 12 |
+
reader = PyPDF2.PdfFileReader(f)
|
| 13 |
+
num_pages = reader.numPages
|
| 14 |
+
for page_num in range(num_pages):
|
| 15 |
+
page = reader.getPage(page_num)
|
| 16 |
+
text += page.extractText()
|
| 17 |
return text
|
| 18 |
|
| 19 |
def main():
|