import gradio as gr
import re
from docx import Document
from docx.shared import Cm, Pt
from docx.enum.style import WD_STYLE_TYPE
from docx.oxml import OxmlElement
from docx.oxml.ns import qn
import tempfile
import os

def set_outline_level(paragraph, level: int = 0):
    p = paragraph._p
    pPr = p.get_or_add_pPr()
    outline = OxmlElement('w:outlineLvl')
    outline.set(qn('w:val'), str(level))
    pPr.append(outline)

def normalize_paragraph(text):
    # 處理段落內換行符號與多餘空格
    text = re.sub(r'[\r\n]+', ' ', text)
    text = re.sub(r'\s{2,}', ' ', text)
    return text.strip()

def process_paragraphs_with_cleanup(doc, combined_pattern):
    """處理段落並清理多餘空行，實現 ^p^p -> ^p 效果"""
    content_list = []
    prev_empty = False
    
    for para in doc.paragraphs:
        text = para.text.strip()
        if not text:
            # 只有當前一個段落不是空的時候，才保留這個空段落
            if not prev_empty:
                content_list.append(('empty', ''))
            prev_empty = True
        elif re.search(combined_pattern, text):
            content_list.append(('heading', text))
            prev_empty = False
        else:
            content_list.append(('paragraph', text))
            prev_empty = False
    
    return content_list

def rebuild_document(doc, content_list):
    """重建文檔內容"""
    for kind, text in content_list:
        if kind == 'heading':
            heading = doc.add_paragraph(text, style='Heading 1')
            heading.paragraph_format.page_break_before = True
            heading.paragraph_format.space_before = Cm(0)
            heading.paragraph_format.space_after = Cm(0.3)
            heading.paragraph_format.line_spacing = 1.0
            heading.paragraph_format.left_indent = Cm(0)
            heading.paragraph_format.first_line_indent = Cm(0)
            set_outline_level(heading, 0)
            for run in heading.runs:
                run.font.name = '新細明體'
                run.font.size = Pt(16)
        elif kind == 'paragraph':
            clean_text = normalize_paragraph(text)
            para = doc.add_paragraph(clean_text)
            para.paragraph_format.space_before = Cm(0)
            para.paragraph_format.space_after = Cm(0)
            para.paragraph_format.line_spacing = 1.0
            para.paragraph_format.left_indent = Cm(0)
            para.paragraph_format.first_line_indent = Cm(0.7)
            for run in para.runs:
                run.font.name = '新細明體'
        elif kind == 'empty':
            # 只有在真正需要保留空段落時才創建
            doc.add_paragraph('')

def format_docx(file, chapter_keywords, remove_empty_paragraphs=True):
    if file is None:
        return None, "請上傳一個 Word 文件"
    if not chapter_keywords.strip():
        return None, "請輸入章節分段方式（例如：章,節,話）"

    try:
        doc = Document(file.name)

        # 確保有 Heading 1 樣式
        styles = doc.styles
        if 'Heading 1' not in styles:
            heading_style = styles.add_style('Heading 1', WD_STYLE_TYPE.PARAGRAPH)
            heading_style.base_style = styles['Normal']
            heading_style.font.bold = True
            heading_style.font.size = Pt(16)

        # 章節關鍵字模式
        keywords = [k.strip() for k in chapter_keywords.split(',')]
        patterns = [f'第\s*[0-9一二三四五六七八九十百千萬壹貳參肆伍陸柒捌玖拾佰仟萬IVXLCDMivxlcdm]+\s*{k}' for k in keywords]
        combined_pattern = '|'.join(patterns)

        # 處理段落並清理多餘空行
        content_list = process_paragraphs_with_cleanup(doc, combined_pattern)
        
        # 如果選擇移除空段落，過濾掉所有空段落
        if remove_empty_paragraphs:
            content_list = [item for item in content_list if item[0] != 'empty']

        # 清空原始內容
        for p in doc.paragraphs:
            p._element.getparent().remove(p._element)

        # 重建段落
        rebuild_document(doc, content_list)

        output_path = tempfile.mktemp(suffix='.docx')
        doc.save(output_path)
        
        empty_status = "已移除所有空段落" if remove_empty_paragraphs else "保留單個空段落"
        return output_path, f"✅ 處理完成！找到章節關鍵字：{', '.join(keywords)}，{empty_status}"
    except Exception as e:
        return None, f"❌ 處理失敗：{str(e)}"

def create_interface():
    with gr.Blocks(title="Word 文件格式化工具", theme=gr.themes.Soft()) as demo:
        gr.HTML("""
        <div style="text-align: center; margin-bottom: 20px;">
            <h1>📄 Word 文件格式化工具</h1>
            <p>自動格式化您的 Word 文件，設定章節樣式和分頁</p>
        </div>
        """)

        with gr.Row():
            with gr.Column(scale=1):
                file_input = gr.File(label="上傳 Word 文件 (.docx)", file_types=[".docx"], file_count="single")
                chapter_input = gr.Textbox(label="章節分段方式", placeholder="章,節,話", value="章,節,話")
                remove_empty_checkbox = gr.Checkbox(
                    label="移除空段落", 
                    value=True, 
                    info="勾選時會移除所有空段落，取消勾選時會保留單個空段落（^p^p -> ^p）"
                )
                process_btn = gr.Button("🔄 開始處理", variant="primary", size="lg")
            with gr.Column(scale=1):
                status_output = gr.Textbox(label="處理狀態", interactive=False, lines=3)
                download_output = gr.File(label="下載處理後的文件", interactive=False)

        process_btn.click(
            fn=format_docx, 
            inputs=[file_input, chapter_input, remove_empty_checkbox], 
            outputs=[download_output, status_output]
        )

    return demo

if __name__ == "__main__":
    demo = create_interface()
    demo.launch(server_name="0.0.0.0", server_port=7860, share=True)