Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
import pdfplumber
|
| 3 |
+
|
| 4 |
+
def read_pdf(file, format_type):
|
| 5 |
+
if file is None:
|
| 6 |
+
return "Please upload a PDF file."
|
| 7 |
+
text = ""
|
| 8 |
+
with pdfplumber.open(file.name) as pdf:
|
| 9 |
+
for page in pdf.pages:
|
| 10 |
+
text += page.extract_text() or ""
|
| 11 |
+
if format_type == "HTML":
|
| 12 |
+
return f"<html><body><p>{text}</p></body></html>"
|
| 13 |
+
else:
|
| 14 |
+
return f"<xml><content>{text}</content></xml>"
|
| 15 |
+
|
| 16 |
+
app = gr.Interface(
|
| 17 |
+
fn=read_pdf,
|
| 18 |
+
inputs=[gr.File(label="Upload PDF"), gr.Radio(["HTML", "XML"], label="Output Format")],
|
| 19 |
+
outputs="text",
|
| 20 |
+
title="PDF to HTML/XML Converter"
|
| 21 |
+
)
|
| 22 |
+
|
| 23 |
+
app.launch()
|