app.py

import os
import gradio as gr
from summarize import summarize_text
from pdf2text import convert_PDF_to_Text
import logging

logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

EXAMPLES_DIR = "examples"

def load_examples():
    name_to_path = {}
    if os.path.exists(EXAMPLES_DIR):
        for file in os.listdir(EXAMPLES_DIR):
            if file.endswith(".txt"):
                name = file.replace(".txt", "")
                name_to_path[name] = os.path.join(EXAMPLES_DIR, file)
    logger.info(f"Loaded {len(name_to_path)} examples")
    return name_to_path

def get_example_text(example_name, name_to_path):
    path = name_to_path.get(example_name)
    if path and os.path.exists(path):
        with open(path, "r", encoding="utf-8") as f:
            return f.read()
    return ""

name_to_path = load_examples()

def summarize_interface(input_text, summary_length, summary_type, use_ocr):
    try:
        if input_text.endswith(".pdf") and os.path.exists(input_text):
            result_text = convert_PDF_to_Text(input_text, use_ocr=use_ocr)
        elif os.path.isfile(input_text):
            with open(input_text, "r", encoding="utf-8") as f:
                result_text = f.read()
        else:
            result_text = input_text
        summary = summarize_text(result_text, summary_length, summary_type)
        return summary
    except Exception as e:
        logger.exception("Summarization failed:")
        return f"❌ Summarization failed: {str(e)}"

example_names = list(name_to_path.keys())
default_example = example_names[0] if example_names else None

with gr.Blocks() as demo:
    gr.Markdown("# DocSummarizer
使用 AI 自動摘要你的文件 📄")

    with gr.Row():
        input_textbox = gr.Textbox(label="Text to Summarize (or PDF path)", lines=15, placeholder="輸入或貼上文字，或提供 txt/pdf 檔案路徑")
        with gr.Column():
            summary_length = gr.Slider(50, 1000, value=250, label="Summary Length")
            summary_type = gr.Radio(choices=["map", "map-reduce"], value="map-reduce", label="Summarization Strategy")
            use_ocr = gr.Checkbox(label="Use OCR for PDF", value=False)
            submit_button = gr.Button("Summarize")

    output_textbox = gr.Textbox(label="Summarized Output", lines=15)
    submit_button.click(fn=summarize_interface, inputs=[input_textbox, summary_length, summary_type, use_ocr], outputs=output_textbox)

    if default_example:
        with gr.Row():
            gr.Examples(
                examples=[[name] for name in example_names],
                inputs=input_textbox,
                label="📚 範例檔案",
                fn=lambda name: get_example_text(name, name_to_path),
                cache_examples=False
            )

demo.launch()