Spaces:
Sleeping
Sleeping
Add application file
Browse files
app.py
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
from PyPDF2 import PdfReader
|
| 3 |
+
|
| 4 |
+
def read_pdf(*inps):
|
| 5 |
+
pdf_file, start_index, end_index, text_rule = inps
|
| 6 |
+
reader = PdfReader(pdf_file)
|
| 7 |
+
pages = reader.pages
|
| 8 |
+
text = ""
|
| 9 |
+
for page in pages[int(start_index):int(end_index)+1]:
|
| 10 |
+
sub = page.extract_text()
|
| 11 |
+
if text_rule:
|
| 12 |
+
for rule in text_rule.split(";"):
|
| 13 |
+
sub = eval(rule)
|
| 14 |
+
text += sub
|
| 15 |
+
return text
|
| 16 |
+
|
| 17 |
+
with gr.Blocks() as demo:
|
| 18 |
+
with gr.Row():
|
| 19 |
+
with gr.Column():
|
| 20 |
+
pdf_file = gr.File(label="PDF", interactive=True)
|
| 21 |
+
with gr.Row():
|
| 22 |
+
start_index = gr.Number(label="start_page",value=0)
|
| 23 |
+
end_index = gr.Number(label="end_page",value=0)
|
| 24 |
+
text_rule = gr.Textbox(label="rule", value='sub.replace(" \n", "\n");sub.replace(" ", " ")')
|
| 25 |
+
submit = gr.Button(value="submit")
|
| 26 |
+
text_output = gr.Textbox(interactive=True)
|
| 27 |
+
|
| 28 |
+
inputs = [pdf_file, start_index, end_index, text_rule]
|
| 29 |
+
submit.click(fn=read_pdf, inputs=inputs, outputs=text_output)
|
| 30 |
+
|
| 31 |
+
demo.launch()
|