Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
from pdfminer.high_level import extract_pages, extract_text
|
| 3 |
+
from pdfminer.layout import LTTextContainer
|
| 4 |
+
|
| 5 |
+
title = 'PDF Text Extraction Demo'
|
| 6 |
+
|
| 7 |
+
def extract_text_from_pdf(pdf_file_path):
|
| 8 |
+
extracted_text = ""
|
| 9 |
+
for page_layout in extract_pages(pdf_file_path):
|
| 10 |
+
for element in page_layout:
|
| 11 |
+
if isinstance(element, LTTextContainer):
|
| 12 |
+
extracted_text += element.get_text()
|
| 13 |
+
return extracted_text
|
| 14 |
+
|
| 15 |
+
def extract_text_from_pdf_file(pdf_file):
|
| 16 |
+
extracted_text = extract_text_from_pdf(pdf_file.name)
|
| 17 |
+
return extracted_text
|
| 18 |
+
|
| 19 |
+
iface = gr.Interface(fn=extract_text_from_pdf_file,
|
| 20 |
+
inputs="file",
|
| 21 |
+
outputs="text",
|
| 22 |
+
title=title,
|
| 23 |
+
description="Upload a PDF file to extract its text.",
|
| 24 |
+
theme="peach")
|
| 25 |
+
|
| 26 |
+
iface.launch()
|