Create app.py

app.py

@@ -0,0 +1,91 @@
+import os
+import tempfile
+from typing import List
+import gradio as gr
+from PyPDF2 import PdfReader, PdfWriter
+
+MAX_SIZE_BYTES = 500 * 1024 * 1024  # 500 MB
+
+def parse_page_ranges(ranges: str, num_pages: int) -> List[int]:
+    """
+    Convert a string like "1-3,5,8-10" into a sorted list of zero-based page indices.
+    """
+    pages = set()
+    for part in ranges.split(','):
+        part = part.strip()
+        if '-' in part:
+            start_str, end_str = part.split('-', 1)
+            start = max(1, int(start_str))
+            end = min(num_pages, int(end_str))
+            pages.update(range(start - 1, end))
+        else:
+            p = int(part)
+            if 1 <= p <= num_pages:
+                pages.add(p - 1)
+    return sorted(pages)
+
+
+def split_pdf(file, page_ranges: str):
+    # Validate file size
+    file_size = os.path.getsize(file.name)
+    if file_size > MAX_SIZE_BYTES:
+        return None, f"File size exceeds 500 MB limit ({file_size / (1024*1024):.2f} MB)."
+
+    reader = PdfReader(file.name)
+    num_pages = len(reader.pages)
+    try:
+        page_indices = parse_page_ranges(page_ranges, num_pages)
+    except Exception as e:
+        return None, f"Error parsing page ranges: {e}"
+
+    if not page_indices:
+        return None, "No valid pages selected."
+
+    writer = PdfWriter()
+    for idx in page_indices:
+        writer.add_page(reader.pages[idx])
+
+    out_path = tempfile.NamedTemporaryFile(delete=False, suffix=".pdf").name
+    with open(out_path, "wb") as f_out:
+        writer.write(f_out)
+
+    return out_path, None
+
+# Build Gradio interface
+with gr.Blocks(css="""
+    #header {text-align: center; margin-bottom: 20px;}
+    .input-row {display: flex; gap: 10px; margin-bottom: 20px;}
+    .button-row {text-align: center; margin-top: 20px;}
+""") as demo:
+    # Header
+    gr.HTML("<h1 id='header'>📄 PDF Splitter</h1>")
+    gr.Markdown("Upload a PDF (up to 500 MB) and extract specific pages using ranges like `1-3,5,7-9`.")
+
+    with gr.Row(elem_classes="input-row"):
+        pdf_input = gr.File(label="Select PDF file", file_types=['.pdf'])
+        page_input = gr.Textbox(label="Page ranges", placeholder="e.g. 1-3,5,7-9")
+
+    with gr.Row(elem_classes="button-row"):
+        split_button = gr.Button("Split PDF", variant="primary")
+
+    output_file = gr.File(label="Download Split PDF")
+    error_text = gr.Textbox(label="Error Message", interactive=False, visible=False)
+
+    def run_split(file, ranges):
+        if file is None or not ranges:
+            return None, "Please upload a PDF and specify page ranges.", True
+        out_path, error = split_pdf(file, ranges)
+        if error:
+            return None, error, True
+        return out_path, "", False
+
+    split_button.click(fn=run_split,
+                       inputs=[pdf_input, page_input],
+                       outputs=[output_file, error_text, error_text],
+                       api_name="split_pdf")
+
+    # Hide error box when no error
+    error_text.change(lambda msg: msg != "", inputs=error_text, outputs=error_text)
+
+if __name__ == "__main__":
+    demo.launch(server_name="0.0.0.0", server_port=int(os.environ.get("PORT", 7860)))