Update app.py
Browse files
app.py
CHANGED
|
@@ -1,3 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
import gradio as gr
|
| 2 |
from PyPDF2 import PdfReader
|
| 3 |
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
|
@@ -17,7 +36,7 @@ def get_pdf_text(pdf_files):
|
|
| 17 |
text = ""
|
| 18 |
for pdf in pdf_files:
|
| 19 |
pdf_reader = PdfReader(pdf)
|
| 20 |
-
for page in
|
| 21 |
text += page.extract_text() or ""
|
| 22 |
return text
|
| 23 |
|
|
|
|
| 1 |
+
import subprocess
|
| 2 |
+
import sys
|
| 3 |
+
|
| 4 |
+
# Function to install required packages
|
| 5 |
+
def install(package):
|
| 6 |
+
subprocess.check_call([sys.executable, "-m", "pip", "install", package])
|
| 7 |
+
|
| 8 |
+
# Install necessary packages
|
| 9 |
+
try:
|
| 10 |
+
import PyPDF2
|
| 11 |
+
except ImportError:
|
| 12 |
+
install('PyPDF2')
|
| 13 |
+
|
| 14 |
+
try:
|
| 15 |
+
import gradio
|
| 16 |
+
except ImportError:
|
| 17 |
+
install('gradio')
|
| 18 |
+
|
| 19 |
+
# Your existing imports and code here
|
| 20 |
import gradio as gr
|
| 21 |
from PyPDF2 import PdfReader
|
| 22 |
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
|
|
|
| 36 |
text = ""
|
| 37 |
for pdf in pdf_files:
|
| 38 |
pdf_reader = PdfReader(pdf)
|
| 39 |
+
for page in pdf.pages:
|
| 40 |
text += page.extract_text() or ""
|
| 41 |
return text
|
| 42 |
|