Upload app.py

app.py

@@ -3,6 +3,7 @@ import re
 from docx import Document
 from docx.shared import Cm, Pt
 from docx.enum.text import WD_PARAGRAPH_ALIGNMENT
+from docx.enum.style import WD_STYLE_TYPE
 from docx.oxml.ns import qn
 from docx.oxml import OxmlElement
 import tempfile
@@ -36,8 +37,32 @@ def format_docx(file, chapter_keywords):
         # 合併所有模式
         combined_pattern = '|'.join(patterns)
         
+        # 先處理多餘的換行：標記要保留的段落
+        paragraphs_to_keep = []
+        previous_empty = False
+        
+        for paragraph in doc.paragraphs:
+            current_empty = len(paragraph.text.strip()) == 0
+            
+            # 如果當前段落不是空的，或者是第一個空段落，則保留
+            if not current_empty or not previous_empty:
+                paragraphs_to_keep.append(paragraph)
+            
+            previous_empty = current_empty
+        
+        # 刪除不需要的段落（從後往前刪除避免索引問題）
+        all_paragraphs = list(doc.paragraphs)
+        for paragraph in reversed(all_paragraphs):
+            if paragraph not in paragraphs_to_keep:
+                p = paragraph._element
+                p.getparent().remove(p)
+        
         # 處理每個段落
         for paragraph in doc.paragraphs:
+            # 跳過空段落
+            if len(paragraph.text.strip()) == 0:
+                continue
+                
             # 1. 將樣式改為無間距
             paragraph.style.paragraph_format.space_before = Cm(0)
             paragraph.style.paragraph_format.space_after = Cm(0)
@@ -49,6 +74,9 @@ def format_docx(file, chapter_keywords):
             
             # 3. 檢查是否為章節標題
             if re.search(combined_pattern, paragraph.text):
+                # 保存原始文字
+                original_text = paragraph.text
+                
                 # 嘗試設定標題 1 樣式（處理不同語言版本的樣式名稱）
                 try:
                     # 英文版本
@@ -62,16 +90,21 @@ def format_docx(file, chapter_keywords):
                             # 其他可能的名稱
                             paragraph.style = doc.styles['Heading1']
                         except KeyError:
-                            # 如果都找不到，手動設定樣式
-                            paragraph.style = doc.styles['Normal']
-                            # 手動設定標題樣式
-                            for run in paragraph.runs:
-                                run.font.bold = True
-                                run.font.size = Pt(16)  # 16pt
+                            # 如果都找不到，創建自定義標題樣式
+                            from docx.enum.style import WD_STYLE_TYPE
+                            try:
+                                heading_style = doc.styles.add_style('CustomHeading1', WD_STYLE_TYPE.PARAGRAPH)
+                                heading_style.font.bold = True
+                                heading_style.font.size = Pt(16)
+                                paragraph.style = heading_style
+                            except:
+                                # 最後手段：直接設定格式
+                                paragraph.style = doc.styles['Normal']
                 
-                # 確保粗體設定
+                # 確保粗體設定和字體大小
                 for run in paragraph.runs:
                     run.font.bold = True
+                    run.font.size = Pt(16)
                 
                 # 4. 在章節前分頁
                 paragraph.paragraph_format.page_break_before = True
@@ -141,6 +174,7 @@ def create_interface():
             <h3>🛠️ 處理功能說明</h3>
             <ul>
                 <li><strong>樣式調整：</strong>將所有段落設定為無間距樣式</li>
+                <li><strong>刪除多餘換行：</strong>自動合併連續的空行，避免過多空白</li>
                 <li><strong>首行縮排：</strong>每個段落首行縮排 2 個字元位置</li>
                 <li><strong>章節識別：</strong>自動識別章節標題（如「第一章」、「第二節」等）</li>
                 <li><strong>標題格式：</strong>章節標題設定為粗體標題 1 樣式（無縮排）</li>