pdf-splitter / app.py
bluenevus's picture
Update app.py
ba2629d verified
raw
history blame
5.94 kB
import base64
import io
import os
import zipfile
from threading import Thread
import dash
import dash_bootstrap_components as dbc
from dash import dcc, html, Input, Output, State, callback, MATCH, ALL
from dash.exceptions import PreventUpdate
from PyPDF2 import PdfReader, PdfWriter
app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP])
# Global variables
generated_file = None
progress = 0
# Layout
app.layout = dbc.Container([
html.H1("PDF Splitter", className="my-4"),
dcc.Upload(
id='upload-pdf',
children=html.Div(['Drag and Drop or ', html.A('Select PDF')]),
style={
'width': '100%',
'height': '60px',
'lineHeight': '60px',
'borderWidth': '1px',
'borderStyle': 'dashed',
'borderRadius': '5px',
'textAlign': 'center',
'margin': '10px'
},
multiple=False
),
html.Div(id='pdf-name'),
dbc.Card([
dbc.CardBody([
html.Div(id='ranges-container', children=[
dbc.Row([
dbc.Col(dbc.Input(id={'type': 'range-input', 'index': 0}, type='text', placeholder='Enter page range (e.g., 1-3)'), width=10),
dbc.Col(dbc.Button("Remove", id={'type': 'remove-range', 'index': 0}, color="danger", size="sm"), width=2),
], className="mb-2"),
]),
dbc.Button("Add Range", id='add-range', color="secondary", className="mt-2"),
])
], className="my-3"),
dbc.Button("Split PDF", id='split-button', color="primary", className="mt-3", disabled=True),
dbc.Progress(id='progress-bar', className="my-3"),
dbc.Button("Download ZIP", id='download-button', color="success", className="mt-3", disabled=True),
dcc.Download(id="download-zip"),
], fluid=True)
@callback(
Output('pdf-name', 'children'),
Output('split-button', 'disabled'),
Input('upload-pdf', 'contents'),
State('upload-pdf', 'filename')
)
def update_output(contents, filename):
if contents is not None:
return f"Selected file: {filename}", False
return "No file selected", True
@callback(
Output('ranges-container', 'children'),
Input('add-range', 'n_clicks'),
Input({'type': 'remove-range', 'index': ALL}, 'n_clicks'),
State('ranges-container', 'children'),
prevent_initial_call=True
)
def manage_ranges(add_clicks, remove_clicks, existing_ranges):
ctx = dash.callback_context
triggered_id = ctx.triggered[0]['prop_id'].split('.')[0]
if triggered_id == 'add-range':
new_index = len(existing_ranges)
new_range = dbc.Row([
dbc.Col(dbc.Input(id={'type': 'range-input', 'index': new_index}, type='text', placeholder='Enter page range (e.g., 1-3)'), width=10),
dbc.Col(dbc.Button("Remove", id={'type': 'remove-range', 'index': new_index}, color="danger", size="sm"), width=2),
], className="mb-2")
existing_ranges.append(new_range)
elif 'remove-range' in triggered_id:
remove_index = json.loads(triggered_id)['index']
existing_ranges = [range for range in existing_ranges if json.loads(range['props']['children'][1]['props']['children']['props']['id'])['index'] != remove_index]
return existing_ranges
def process_pdf(contents, filename, ranges):
global generated_file, progress
progress = 0
try:
content_type, content_string = contents.split(',')
decoded = base64.b64decode(content_string)
pdf = PdfReader(io.BytesIO(decoded))
writers = []
for range_str in ranges:
start, end = map(int, range_str.split('-'))
writer = PdfWriter()
for i in range(start - 1, min(end, len(pdf.pages))):
writer.add_page(pdf.pages[i])
writers.append(writer)
zip_buffer = io.BytesIO()
with zipfile.ZipFile(zip_buffer, 'w', zipfile.ZIP_DEFLATED) as zf:
for i, writer in enumerate(writers):
progress = (i + 1) / len(writers) * 100
pdf_buffer = io.BytesIO()
writer.write(pdf_buffer)
pdf_buffer.seek(0)
zf.writestr(f"split_{i+1}.pdf", pdf_buffer.getvalue())
zip_buffer.seek(0)
generated_file = zip_buffer.getvalue()
progress = 100
except Exception as e:
print(f"Error processing PDF: {str(e)}")
progress = -1
@callback(
Output('progress-bar', 'value'),
Output('download-button', 'disabled'),
Input('split-button', 'n_clicks'),
State('upload-pdf', 'contents'),
State('upload-pdf', 'filename'),
State({'type': 'range-input', 'index': ALL}, 'value'),
prevent_initial_call=True
)
def split_pdf(n_clicks, contents, filename, ranges):
if not contents or not ranges:
raise PreventUpdate
ranges = [r for r in ranges if r] # Filter out empty ranges
thread = Thread(target=process_pdf, args=(contents, filename, ranges))
thread.start()
return 0, True
@callback(
Output('progress-bar', 'value', allow_duplicate=True),
Output('download-button', 'disabled', allow_duplicate=True),
Input('progress-bar', 'value'),
prevent_initial_call=True
)
def update_progress(value):
global progress
if progress == 100:
return 100, False
elif progress == -1:
return 0, True
else:
return progress, True
@callback(
Output("download-zip", "data"),
Input("download-button", "n_clicks"),
prevent_initial_call=True
)
def download_zip(n_clicks):
global generated_file
if generated_file is not None:
return dcc.send_bytes(generated_file, "split_pdfs.zip")
raise PreventUpdate
if __name__ == '__main__':
print("Starting the Dash application...")
app.run(debug=True, host='0.0.0.0', port=7860)
print("Dash application has finished running.")