diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..dfc2545
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,107 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+.hypothesis/
+.pytest_cache/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# pyenv
+.python-version
+
+# celery beat schedule file
+celerybeat-schedule
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+
+# 
+temp
\ No newline at end of file
diff --git a/README.md b/README.md
index 7055cb3..26b1ddc 100644
--- a/README.md
+++ b/README.md
@@ -1,2 +1,55 @@
-# ncmdump
-将.ncm格式音频文件转换为flac格式，提供windows客户端和WEB两种使用方式。（Convert .ncm format audio files to flac format, providing two usage options: a Windows client and a web interface.）
+# ncm -> flac converter
+
+将.ncm格式音频文件转换为flac格式，提供windows客户端和WEB两种使用方式。
+
+## 1 环境
+
+基本环境：
+```bash
+pip install mutagen
+pip install pycryptodome
+```
+
+GUI额外环境：
+```bash
+pip install PyQt6
+pip install pyinstaller
+```
+
+WEB额外环境：
+```bash
+pip install streamlit
+```
+
+全安装：
+```bash
+pip install -r requirements.txt
+```
+
+## 2 使用
+
+### 2.1 GUI
+
+运行：
+```bash
+python gui.py
+```
+
+编译：
+```bash
+pyinstaller --onefile --add-data="file:file" -wF -i file/favicon-32x32.png -n "NCM转换器" .\gui.py
+```
+
+效果:
+![s1](./file/s1.gif)
+
+
+### 2.2 WEB
+
+运行：
+```bash
+streamlit run web.py --server.port 1111
+```
+
+效果:
+![s2](./file/s2.gif)
diff --git a/cplusplus/libncmdump.dll b/cplusplus/libncmdump.dll
new file mode 100644
index 0000000..0ad1b05
Binary files /dev/null and b/cplusplus/libncmdump.dll differ
diff --git a/cplusplus/ncmdumpdll.py b/cplusplus/ncmdumpdll.py
new file mode 100644
index 0000000..c8c8775
--- /dev/null
+++ b/cplusplus/ncmdumpdll.py
@@ -0,0 +1,51 @@
+import ctypes
+import os
+
+class NcmdumpDll:
+    def __init__(self, file_name):
+        dll_path = os.path.abspath(os.path.join(os.path.dirname(__file__), "libncmdump.dll"))
+        self.dll = ctypes.CDLL(dll_path)
+        
+        # Define function prototypes
+        self.dll.CreateNeteaseCrypt.argtypes = [ctypes.c_char_p]
+        self.dll.CreateNeteaseCrypt.restype = ctypes.c_void_p
+        
+        self.dll.Dump.argtypes = [ctypes.c_void_p]
+        self.dll.Dump.restype = ctypes.c_int
+        
+        self.dll.FixMetadata.argtypes = [ctypes.c_void_p]
+        self.dll.FixMetadata.restype = None
+        
+        self.dll.DestroyNeteaseCrypt.argtypes = [ctypes.c_void_p]
+        self.dll.DestroyNeteaseCrypt.restype = None
+        
+        # Convert file name to bytes
+        file_bytes = file_name.encode('utf-8')
+        
+        # Allocate memory and copy file name bytes
+        input_ptr = ctypes.create_string_buffer(file_bytes)
+        
+        # Create NeteaseCrypt instance
+        self.netease_crypt = self.dll.CreateNeteaseCrypt(input_ptr)
+    
+    def dump(self):
+        return self.dll.Dump(self.netease_crypt)
+    
+    def fix_metadata(self):
+        self.dll.FixMetadata(self.netease_crypt)
+    
+    def destroy(self):
+        self.dll.DestroyNeteaseCrypt(self.netease_crypt)
+
+    def process_file(self):
+        self.dump()
+        self.fix_metadata()
+        self.destroy()
+
+if __name__ == "__main__":
+    # 文件名
+    file_path = "YOASOBI.ncm"
+    # 创建 NeteaseCrypt 类的实例
+    netease_crypt = NcmdumpDll(file_path)
+    # 启动转换过程
+    netease_crypt.process_file()
diff --git a/file/bk.png b/file/bk.png
new file mode 100644
index 0000000..8e9a6d9
Binary files /dev/null and b/file/bk.png differ
diff --git a/file/favicon-32x32.png b/file/favicon-32x32.png
new file mode 100644
index 0000000..d236703
Binary files /dev/null and b/file/favicon-32x32.png differ
diff --git a/file/s1.gif b/file/s1.gif
new file mode 100644
index 0000000..7e43ee9
Binary files /dev/null and b/file/s1.gif differ
diff --git a/file/s2.gif b/file/s2.gif
new file mode 100644
index 0000000..50b4f22
Binary files /dev/null and b/file/s2.gif differ
diff --git a/gui.py b/gui.py
new file mode 100644
index 0000000..221ac54
--- /dev/null
+++ b/gui.py
@@ -0,0 +1,69 @@
+import sys
+from PyQt6.QtWidgets import QApplication, QWidget, QVBoxLayout, QLabel
+from PyQt6.QtCore import Qt
+from PyQt6.QtGui import QDropEvent,QPixmap, QPainter, QFont, QColor,QFontMetrics,QIcon,QDragEnterEvent
+import os
+from ncmdump import dump
+
+class DragDropWidget(QWidget):
+    def __init__(self, parent=None):
+        super().__init__(parent)
+        self.setAcceptDrops(True)
+        self.init_ui()
+
+    def init_ui(self):
+        self.setWindowTitle("NCM转换器")
+        self.setGeometry(100, 100, 573, 573)  
+
+        self.setWindowIcon(QIcon(self.get_resource_path("file/favicon-32x32.png")))
+        # 加载图片
+        self.original_pixmap = QPixmap(self.get_resource_path("file/bk.png"))
+
+        self.label = QLabel(self)
+        self.label.setPixmap(self.original_pixmap)
+        self.label.setGeometry(0, 0, 573, 573)  
+        self.update_text('将ncm文件拖拽到此处')  # 初始文本内容
+    def update_text(self, text):
+        pixmap = self.original_pixmap.copy()  # 复制原始的 QPixmap 对象
+        painter = QPainter(pixmap)
+        painter.setPen(QColor('black'))  # 设置文本颜色
+        font = QFont('SimHei', 20)  # 设置字体和大小
+        painter.setFont(font)
+
+        font_metrics = QFontMetrics(font)
+        text_width = font_metrics.horizontalAdvance(text)
+        text_height = font_metrics.height()
+
+        x = (pixmap.width() - text_width) // 2
+        y = (pixmap.height() - text_height) // 2 + font_metrics.ascent()
+
+        painter.drawText(x, y, text)  # 在图片中居中绘制文本
+        painter.end()
+        self.label.setPixmap(pixmap)  # 更新 QLabel 中的图片
+
+    def get_resource_path(self,relative_path):
+        if hasattr(sys, '_MEIPASS'):
+            return os.path.join(sys._MEIPASS, relative_path)
+        return os.path.join(os.path.abspath("."), relative_path)
+
+
+    def dragEnterEvent(self, event: QDragEnterEvent):
+        if event.mimeData().hasUrls():
+            event.acceptProposedAction()
+
+    def dropEvent(self, event: QDropEvent):
+        for url in event.mimeData().urls():
+            file_path = url.toLocalFile()
+            if file_path.endswith(".ncm"):
+                try:
+                    dump(file_path)
+                    self.update_text(f"处理完成：{file_path}")
+                except Exception as e:
+                    self.update_text(f"处理文件时出错：{str(e)}")
+                    
+
+if __name__ == "__main__":
+    app = QApplication(sys.argv)
+    widget = DragDropWidget()
+    widget.show()
+    sys.exit(app.exec())
diff --git a/ncmdump/__init__.py b/ncmdump/__init__.py
new file mode 100644
index 0000000..71dcdf4
--- /dev/null
+++ b/ncmdump/__init__.py
@@ -0,0 +1 @@
+from .core import dump
\ No newline at end of file
diff --git a/ncmdump/app.py b/ncmdump/app.py
new file mode 100644
index 0000000..a8fea5e
--- /dev/null
+++ b/ncmdump/app.py
@@ -0,0 +1,127 @@
+# -*- coding: utf-8 -*-
+"""
+Created on Fri Sep 28 13:32:51 2018
+
+@author: Nzix
+"""
+
+import argparse, os, sys, traceback, re
+sys.path.append(os.path.dirname(os.path.abspath(__file__)))
+
+if __package__ is None:
+    from core import dump
+else:
+    from .core import dump
+
+parser = argparse.ArgumentParser(
+    prog = 'ncmdump', add_help = False
+)
+parser.add_argument(
+    '-h', action = 'help',
+    help = 'show this help message and exit'
+)
+parser.add_argument(
+    'input', metavar = 'input', nargs = '*', default = ['.'],
+    help = 'ncm file or folder path'
+)
+parser.add_argument(
+    '-f', metavar = 'format', dest = 'format', default = '',
+    help = 'customize naming format'
+)
+parser.add_argument(
+    '-o', metavar = 'output', dest = 'output',
+    help = 'customize saving folder'
+)
+parser.add_argument(
+    '-d', dest = 'delete', action = 'store_true',
+    help = 'delete source after conversion'
+)
+group = parser.add_mutually_exclusive_group()
+group.add_argument(
+    '-c', dest = 'cover', action = 'store_true',
+    help = 'overwrite file with the same name'
+)
+group.add_argument(
+    '-r', dest = 'rename', action = 'store_true',
+    help = 'auto rename if file name conflicts'
+)
+args = parser.parse_args()
+
+def validate_name(name):
+    pattern = {u'\\': u'＼', u'/': u'／', u':': u'：', u'*': u'＊', u'?': u'？', u'"': u'＂', u'<': u'＜', u'>': u'＞', u'|': u'｜'}
+    for character in pattern:
+        name = name.replace(character, pattern[character])
+    return name
+
+def validate_collision(path):
+    index = 1
+    origin = path
+    while os.path.exists(path):
+        path = '({})'.format(index).join(os.path.splitext(origin))
+        index += 1
+    return path
+
+def name_format(path, meta):
+    information = {
+        'artist': ','.join([artist[0] for artist in meta.get('artist')]) if 'artist' in meta else None,
+        'title': meta.get('musicName'),
+        'album': meta.get('album')
+    }
+
+    def substitute(matched):
+        key = matched.group(1)
+        if key in information:
+            return information[key]
+        else:
+            return key
+
+    name = re.sub(r'%(.+?)%', substitute, args.format)
+    name = os.path.splitext(os.path.split(path)[1])[0] if not name else name
+    name = validate_name(name)
+    name += '.' + meta['format']
+    folder = args.output if args.output else os.path.dirname(path)
+    save = os.path.join(folder, name)
+    if args.rename: save = validate_collision(save)
+    return save
+
+def traverse(path):
+    path = os.path.abspath(path)
+    if not os.path.exists(path):
+        return []
+    elif os.path.isdir(path):
+        return sum([traverse(os.path.join(path, name)) for name in os.listdir(path)], [])
+    else:
+        return [path] if os.path.splitext(path)[-1] == '.ncm' else []
+
+def main():
+    if args.output:
+        args.output = os.path.abspath(args.output)
+        if not os.path.exists(args.output):
+            print('output does not exist')
+            exit()
+        if not os.path.isdir(args.output):
+            print('output is not a folder')
+            exit()
+
+    input_files = sum([traverse(path) for path in args.input], [])
+    files = sorted(set(input_files), key = input_files.index)
+
+    if sys.version[0] == '2':
+        files = [path.decode(sys.stdin.encoding) for path in files]
+
+    if not files:
+        print('empty input')
+        exit()
+
+    for path in files:
+        try:
+            save = dump(path, name_format, not args.cover)
+            if save: print(os.path.split(save)[-1])
+            if args.delete: os.remove(path)
+        except KeyboardInterrupt:
+            exit()
+        except:
+            print(traceback.format_exc())
+
+if __name__ == '__main__':
+    main()
\ No newline at end of file
diff --git a/ncmdump/core.py b/ncmdump/core.py
new file mode 100644
index 0000000..4313405
--- /dev/null
+++ b/ncmdump/core.py
@@ -0,0 +1,133 @@
+# -*- coding: utf-8 -*-
+"""
+Created on Sun Jul 15 01:05:58 2018
+
+@author: Nzix
+"""
+
+import binascii, struct
+import base64, json
+import os
+
+from Crypto.Cipher import AES
+from Crypto.Util.Padding import unpad
+from Crypto.Util.strxor import strxor
+from mutagen import mp3, flac, id3
+
+def dump(input_path, output_path = None, skip = True):
+
+    output_path = (lambda path, meta: os.path.splitext(path)[0] + '.' + meta['format']) if not output_path else output_path
+    output_path_generator = (lambda path, meta: output_path) if not callable(output_path) else output_path
+
+    core_key = binascii.a2b_hex('687A4852416D736F356B496E62617857')
+    meta_key = binascii.a2b_hex('2331346C6A6B5F215C5D2630553C2728')
+
+    f = open(input_path, 'rb')
+
+    # magic header
+    header = f.read(8)
+    assert header == binascii.a2b_hex('4354454e4644414d')
+
+    f.seek(2, 1)
+
+    # key data
+    key_length = f.read(4)
+    key_length = struct.unpack('<I', bytes(key_length))[0]
+
+    key_data = bytearray(f.read(key_length))
+    key_data = bytes(bytearray([byte ^ 0x64 for byte in key_data]))
+
+    cryptor = AES.new(core_key, AES.MODE_ECB)
+    key_data = unpad(cryptor.decrypt(key_data), 16)[17:]
+    key_length = len(key_data)
+
+    # S-box (standard RC4 Key-scheduling algorithm)
+    key = bytearray(key_data)
+    S = bytearray(range(256))
+    j = 0
+
+    for i in range(256):
+        j = (j + S[i] + key[i % key_length]) & 0xFF
+        S[i], S[j] = S[j], S[i]
+
+    # meta data
+    meta_length = f.read(4)
+    meta_length = struct.unpack('<I', bytes(meta_length))[0]
+
+    if meta_length:
+        meta_data = bytearray(f.read(meta_length))
+        meta_data = bytes(bytearray([byte ^ 0x63 for byte in meta_data]))
+        identifier = meta_data.decode('utf-8')
+        meta_data = base64.b64decode(meta_data[22:])
+
+        cryptor = AES.new(meta_key, AES.MODE_ECB)
+        meta_data = unpad(cryptor.decrypt(meta_data), 16).decode('utf-8')
+        meta_data = json.loads(meta_data[6:])
+    else:
+        meta_data = {'format': 'flac' if os.fstat(f.fileno()).st_size > 1024 ** 2 * 16 else 'mp3'}
+
+    f.seek(5, 1)
+
+    # album cover
+    image_space = f.read(4)
+    image_space = struct.unpack('<I', bytes(image_space))[0]
+    image_size = f.read(4)
+    image_size = struct.unpack('<I', bytes(image_size))[0]
+    image_data = f.read(image_size) if image_size else None
+
+    f.seek(image_space - image_size, 1)
+
+    # media data
+    output_path = output_path_generator(input_path, meta_data)
+    if skip and os.path.exists(output_path): return
+
+    # stream cipher (modified RC4 Pseudo-random generation algorithm)
+    stream = [S[(S[i] + S[(i + S[i]) & 0xFF]) & 0xFF] for i in range(256)]
+    stream = bytes(stream[1:] + stream[:1]) * 64
+
+    m = open(output_path, 'wb')
+    while True:
+        data = f.read(16384)
+        if not data:
+            break
+        data = strxor(data, stream[:len(data)])
+        m.write(data)
+    m.close()
+    f.close()
+
+    # media tag
+    def embed(item, content, type):
+        item.encoding = 0
+        item.type = type
+        item.mime = 'image/png' if content[0:4] == binascii.a2b_hex('89504E47') else 'image/jpeg'
+        item.data = content
+
+    if image_data:
+        if meta_data['format'] == 'flac':
+            audio = flac.FLAC(output_path)
+            image = flac.Picture()
+            embed(image, image_data, 3)
+            audio.clear_pictures()
+            audio.add_picture(image)
+        elif meta_data['format'] == 'mp3':
+            audio = mp3.MP3(output_path)
+            image = id3.APIC()
+            embed(image, image_data, 6)
+            audio.tags.add(image)
+        audio.save()
+
+    if meta_length:
+        if meta_data['format'] == 'flac':
+            audio = flac.FLAC(output_path)
+            audio['description'] = identifier
+        else:
+            audio = mp3.EasyMP3(output_path)
+            audio['title'] = 'placeholder'
+            audio.tags.RegisterTextKey('comment', 'COMM')
+            audio['comment'] = identifier
+        audio['title'] = meta_data['musicName']
+        audio['album'] = meta_data['album']
+        audio['artist'] = '/'.join([artist[0] for artist in meta_data['artist']])
+        audio.save()
+
+    return output_path
\ No newline at end of file
diff --git a/requirements.txt b/requirements.txt
new file mode 100644
index 0000000..1eaf19f
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1,9 @@
+mutagen
+pycryptodome
+
+# gui
+PyQt6
+pyinstaller
+
+# web
+streamlit
\ No newline at end of file
diff --git a/web.py b/web.py
new file mode 100644
index 0000000..b33cfef
--- /dev/null
+++ b/web.py
@@ -0,0 +1,56 @@
+import streamlit as st
+from ncmdump import dump
+import os
+
+TEMP = "./temp"
+
+
+def web_page():
+    st.title("NCM转换器")
+    
+    input_file = st.file_uploader("上传音频：", type=["ncm"])
+    if input_file is not None:
+        temp_input_audio = os.path.join(
+            TEMP,
+            os.path.splitext(os.path.basename(input_file.name))[0]+".ncm"
+        )
+        temp_output_audio_name = os.path.splitext(os.path.basename(input_file.name))[0]+".flac"
+        temp_output_audio = os.path.join(
+            TEMP,
+            os.path.splitext(os.path.basename(input_file.name))[0]+".flac"
+        )
+        if os.path.exists(temp_output_audio):
+            os.remove(temp_output_audio)
+
+        if not os.path.exists(temp_input_audio):      
+            with open(temp_input_audio, "wb") as f:
+                f.write(input_file.read())
+        else:
+            print("文件:{} 已存在，无需创建".format(temp_input_audio))
+
+    st.markdown("------------")   
+    if st.button("开始转换"):
+        if temp_input_audio is None:
+            st.warning("请先上传音频")
+            st.stop()     
+
+        with st.spinner('转换中'):
+            print(temp_input_audio)
+            temp_output_audio = dump(temp_input_audio)
+            print(temp_output_audio)
+            st.audio(temp_output_audio, format='audio/flac', start_time=0)    
+
+            with open(temp_output_audio, "rb") as audio_file:
+               audio_data = audio_file.read()
+
+            st.download_button(
+                label="点击下载flac",
+                data=audio_data,
+                file_name=temp_output_audio_name,
+                mime="audio/flac"
+            )
+
+if __name__ == "__main__":
+    if not os.path.exists(TEMP):
+        os.makedirs(TEMP)
+    web_page()
\ No newline at end of file