Add files via upload

Author: anouarg88

c2po.py

@@ -0,0 +1,262 @@
+import tkinter as tk
+from tkinter import ttk, filedialog, messagebox
+import jieba
+from pypinyin import pinyin, Style
+import os
+import re
+
+class ChineseToPinyinConverter:
+    def __init__(self):
+        # 只在类Unix系统上启用并行分词
+        if os.name != 'nt':  # 如果不是Windows系统
+            try:
+                jieba.enable_parallel(4)
+            except:
+                pass  # 如果并行模式失败，继续使用单线程
+    
+    def convert_text(self, text):
+        """将文本中的汉字转换为拼音，按词语组合，保留原始标点符号和格式"""
+        try:
+            # 使用正则表达式分割文本，保留所有非汉字字符
+            pattern = re.compile(r'([\u4e00-\u9fff]+)|([^\u4e00-\u9fff]+)')
+            result = []
+            
+            for match in pattern.finditer(text):
+                chinese_part, non_chinese_part = match.groups()
+                
+                if chinese_part:
+                    # 使用jieba进行分词
+                    words = jieba.cut(chinese_part, cut_all=False)
+                    pinyin_words = []
+                    
+                    for word in words:
+                        # 只处理中文字符
+                        if any('\u4e00' <= char <= '\u9fff' for char in word):
+                            # 获取带音标的拼音
+                            pinyin_list = pinyin(word, style=Style.TONE, heteronym=False)
+                            # 将词语的拼音连在一起，不带空格
+                            pinyin_str = ''.join([item[0].lower() for item in pinyin_list])
+                            pinyin_words.append(pinyin_str)
+                        else:
+                            # 非中文字符直接保留
+                            pinyin_words.append(word)
+                    
+                    # 在词语之间添加空格
+                    result.append(' '.join(pinyin_words))
+                elif non_chinese_part:
+                    # 保留所有非汉字字符（包括标点、时间戳等）
+                    result.append(non_chinese_part)
+            
+            # 合并所有部分
+            converted_text = ''.join(result)
+            
+            # 修复：移除每行前后的多余空格
+            lines = converted_text.split('\n')
+            cleaned_lines = [line.strip() for line in lines]
+            return '\n'.join(cleaned_lines)
+            
+        except Exception as e:
+            raise Exception(f"转换过程中出错 (error): {str(e)}")
+
+class PinyinConverterApp:
+    def __init__(self, root):
+        self.root = root
+        self.root.title("C2PO: Chinese to Pinyin Offline 线下汉字转拼音工具")
+        self.root.geometry("580x650")
+        
+        # 初始化转换器
+        self.converter = ChineseToPinyinConverter()
+        
+        self.setup_ui()
+        
+    def setup_ui(self):
+        # 主框架
+        main_frame = ttk.Frame(self.root, padding="10")
+        main_frame.pack(fill=tk.BOTH, expand=True)
+        
+        # 输入标签
+        input_label = ttk.Label(main_frame, text="Input (Chinese): 输入文本（中文）:")
+        input_label.pack(anchor=tk.W, pady=(0, 5))
+        
+        # 输入文本框和滚动条
+        input_frame = ttk.Frame(main_frame)
+        input_frame.pack(fill=tk.BOTH, expand=True, pady=(0, 10))
+        
+        self.input_text = tk.Text(input_frame, height=10, wrap=tk.WORD)
+        input_scrollbar = ttk.Scrollbar(input_frame, orient=tk.VERTICAL, command=self.input_text.yview)
+        self.input_text.configure(yscrollcommand=input_scrollbar.set)
+        
+        self.input_text.pack(side=tk.LEFT, fill=tk.BOTH, expand=True)
+        input_scrollbar.pack(side=tk.RIGHT, fill=tk.Y)
+        
+        # 按钮框架
+        button_frame = ttk.Frame(main_frame)
+        button_frame.pack(fill=tk.X, pady=10)
+        
+        # 按钮
+        ttk.Button(button_frame, text="Import 导入文件", command=self.import_file).pack(side=tk.LEFT, padx=(0, 10))
+        ttk.Button(button_frame, text="Convert 转换为拼音", command=self.convert_text).pack(side=tk.LEFT, padx=(0, 10))
+        ttk.Button(button_frame, text="Copy 复制结果", command=self.copy_to_clipboard).pack(side=tk.LEFT, padx=(0, 10))
+        ttk.Button(button_frame, text="Save Output 保存结果", command=self.save_file).pack(side=tk.LEFT, padx=(0, 10))
+        ttk.Button(button_frame, text="Clear 清空", command=self.clear_text).pack(side=tk.LEFT)
+        
+        # 输出标签
+        output_label = ttk.Label(main_frame, text="Output 拼音输出:")
+        output_label.pack(anchor=tk.W, pady=(10, 5))
+        
+        # 输出文本框和滚动条
+        output_frame = ttk.Frame(main_frame)
+        output_frame.pack(fill=tk.BOTH, expand=True)
+        
+        self.output_text = tk.Text(output_frame, height=10, wrap=tk.WORD, state=tk.DISABLED)
+        output_scrollbar = ttk.Scrollbar(output_frame, orient=tk.VERTICAL, command=self.output_text.yview)
+        self.output_text.configure(yscrollcommand=output_scrollbar.set)
+        
+        self.output_text.pack(side=tk.LEFT, fill=tk.BOTH, expand=True)
+        output_scrollbar.pack(side=tk.RIGHT, fill=tk.Y)
+        
+        # 状态栏
+        self.status_var = tk.StringVar()
+        self.status_var.set("Ready 就绪")
+        status_bar = ttk.Label(main_frame, textvariable=self.status_var, relief=tk.SUNKEN)
+        status_bar.pack(fill=tk.X, pady=(10, 0))
+
+    def import_file(self):
+        """Import 导入文件"""
+        file_path = filedialog.askopenfilename(
+            title="Choose file 选择文件",
+            filetypes=[("All files 所有文件", "*.*"), ("TXT file 文本文件", "*.txt"), ("SRT file SRT文件", "*.srt"), ("JSON file JSON文件", "*.json")]
+        )
+        
+        if file_path:
+            try:
+                # 尝试多种编码
+                encodings = ['utf-8', 'gbk', 'gb2312', 'utf-16']
+                content = None
+                
+                for encoding in encodings:
+                    try:
+                        with open(file_path, 'r', encoding=encoding) as file:
+                            content = file.read()
+                        break
+                    except UnicodeDecodeError:
+                        continue
+                
+                if content is None:
+                    messagebox.showerror("Error 错误", "Encoding error 无法读取文件，可能是编码问题")
+                    return
+                
+                self.input_text.delete(1.0, tk.END)
+                self.input_text.insert(1.0, content)
+                self.status_var.set(f"Imported 已导入文件: {os.path.basename(file_path)}")
+                
+            except Exception as e:
+                messagebox.showerror("Error 错误", f"Error while reading: 读取文件时出错: {str(e)}")
+                self.status_var.set("Import failed! 导入文件失败")
+    
+    def convert_text(self):
+        """转换文本为拼音"""
+        input_content = self.input_text.get(1.0, tk.END).strip()
+        
+        if not input_content:
+            messagebox.showwarning("Warning 警告", "Nothing to convert 请输入要转换的中文文本")
+            return
+        
+        try:
+            self.status_var.set("Converting 正在转换...")
+            self.root.update_idletasks()  # 强制更新UI
+            
+            # 转换文本
+            pinyin_content = self.converter.convert_text(input_content)
+            
+            # 更新输出文本框
+            self.output_text.config(state=tk.NORMAL)
+            self.output_text.delete(1.0, tk.END)
+            self.output_text.insert(1.0, pinyin_content)
+            self.output_text.config(state=tk.DISABLED)
+            
+            self.status_var.set("Conversion complete 转换完成")
+            
+        except Exception as e:
+            messagebox.showerror("Error 错误", str(e))
+            self.status_var.set("Conversion failed 转换失败")
+    
+    def copy_to_clipboard(self):
+        """Copy to clipboard 复制结果到剪贴板"""
+        output_content = self.output_text.get(1.0, tk.END).strip()
+        
+        if not output_content:
+            messagebox.showwarning("Warning 警告", "Nothing to copy 没有内容可复制")
+            return
+        
+        try:
+            # 清空剪贴板并添加新内容
+            self.root.clipboard_clear()
+            self.root.clipboard_append(output_content)
+            # 确保剪贴板内容在程序退出后仍然可用
+            self.root.update()
+            
+            self.status_var.set("Copied to clipboard 已复制到剪贴板")
+            messagebox.showinfo("Success 成功", "Content has been copied to clipboard! 内容已复制到剪贴板！")
+            
+        except Exception as e:
+            messagebox.showerror("Error 错误", f"Error while copying to clipboard: 复制到剪贴板时出错: {str(e)}")
+            self.status_var.set("Copy failed 复制失败")
+    
+    def save_file(self):
+        """Save output to file 保存结果到文件"""
+        output_content = self.output_text.get(1.0, tk.END).strip()
+        
+        if not output_content:
+            messagebox.showwarning("Error 警告", "Nothing to save \n 没有内容可保存")
+            return
+        
+        file_path = filedialog.asksaveasfilename(
+            title="Save file 保存文件",
+            defaultextension=".txt",
+            filetypes=[("TXT file 文本文件", "*.txt"), ("All files 所有文件", "*.*")]
+        )
+        
+        if file_path:
+            try:
+                with open(file_path, 'w', encoding='utf-8') as file:
+                    file.write(output_content)
+                
+                self.status_var.set(f"File saved as 文件已保存: {os.path.basename(file_path)}")
+                messagebox.showinfo("Success 成功", "File saved 文件保存成功！")
+                
+            except Exception as e:
+                messagebox.showerror("Error 错误", f"Error while saving 保存文件时出错: {str(e)}")
+                self.status_var.set("Saving failed 保存失败")
+    
+    def clear_text(self):
+        """Clear all text 清空所有文本"""
+        self.input_text.delete(1.0, tk.END)
+        self.output_text.config(state=tk.NORMAL)
+        self.output_text.delete(1.0, tk.END)
+        self.output_text.config(state=tk.DISABLED)
+        self.status_var.set("Cleared 已清空")
+
+def main():
+    # 检查依赖
+    try:
+        import jieba
+        from pypinyin import pinyin, Style
+    except ImportError as e:
+        print("Missing package, please install: 缺少必要的依赖库，请安装：")
+        print("pip install jieba pypinyin")
+        return
+    
+    # 创建主窗口
+    root = tk.Tk()
+    app = PinyinConverterApp(root)
+    
+    # 添加测试文本
+    test_text = """此处可填需要转拼音的汉字内容"""
+    
+    app.input_text.insert(1.0, test_text)
+    
+    root.mainloop()
+
+if __name__ == "__main__":
+    main()