From bc4e330ee7c50b6dca9a47555e428c2b3cc4712d Mon Sep 17 00:00:00 2001
From: RogerWork <roger.x.sun@qq.com>
Date: Mon, 16 Oct 2023 14:27:04 +0800
Subject: [PATCH] first comit

---
 .gitignore  |  3 +++
 db.py       | 32 ++++++++++++++++++++++++++++++++
 getFiles.py | 17 +++++++++++++++++
 getMd5.py   | 23 +++++++++++++++++++++++
 main.py     | 26 ++++++++++++++++++++++++++
 5 files changed, 101 insertions(+)
 create mode 100644 .gitignore
 create mode 100644 db.py
 create mode 100644 getFiles.py
 create mode 100644 getMd5.py
 create mode 100644 main.py

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..e1334f8
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,3 @@
+.idea/
+__pycache__/
+*.log
diff --git a/db.py b/db.py
new file mode 100644
index 0000000..e9547de
--- /dev/null
+++ b/db.py
@@ -0,0 +1,32 @@
+import pymysql
+from pymysql.cursors import DictCursor
+
+db_config = {
+    'host': 'home.rogersun.cn',
+    'user': 'root',
+    'password': 'Sxzgx1209',
+    'database': 'file_md5'
+}
+
+db_connection = pymysql.Connect(**db_config)
+db_cursor = db_connection.cursor(cursor=DictCursor)
+
+
+def insert_data(_data):
+    insert_sql = "INSERT INTO files VALUES (NUAL, %s, %s, %s, %s);"
+    db_cursor.execute(insert_sql, _data)
+    db_connection.commit()
+
+
+def get_md5_by_path(_path):
+    get_md5_sql = "SELECT md5 FROM files WHERE path = %s;"
+    db_cursor.execute(get_md5_sql, (_path,))
+    r = db_cursor.fetchone()
+    return r
+
+
+def is_exist(_md5):
+    is_exist_sql = "SELECT * FROM files WHERE md5 = %s;"
+    db_cursor.execute(is_exist_sql, (_md5,))
+    r = db_cursor.fetchone()
+    return r
diff --git a/getFiles.py b/getFiles.py
new file mode 100644
index 0000000..aeeec84
--- /dev/null
+++ b/getFiles.py
@@ -0,0 +1,17 @@
+import os
+
+
+def get_file_list(_path):
+    file_gen = os.walk(_path)
+    file_list = []
+    for item in file_gen:
+        if len(item[2]):
+            for file in item[2]:
+                file_list.append(os.path.join(item[0], file))
+    return file_list
+
+
+if __name__ == "__main__":
+    result = get_file_list(r"E:\Backup\Tools\Normal\浏览器")
+    for r in result:
+        print(r)
diff --git a/getMd5.py b/getMd5.py
new file mode 100644
index 0000000..f72ab8c
--- /dev/null
+++ b/getMd5.py
@@ -0,0 +1,23 @@
+import hashlib
+
+
+def get_file_md5(file_name):
+    """
+    计算文件的md5
+    :param file_name:
+    :return:
+    """
+    m = hashlib.md5()   # 创建md5对象
+    with open(file_name,'rb') as fobj:
+        while True:
+            data = fobj.read(4096)
+            if not data:
+                break
+            m.update(data)  # 更新md5对象
+
+    return m.hexdigest()    # 返回md5对象
+
+
+if __name__ == "__main__":
+    md5_code = get_file_md5("E:\Backup\Tools\SystemISO\kms.txt")
+    print(md5_code)
\ No newline at end of file
diff --git a/main.py b/main.py
new file mode 100644
index 0000000..9c2146f
--- /dev/null
+++ b/main.py
@@ -0,0 +1,26 @@
+from getFiles import get_file_list
+from getMd5 import get_file_md5
+from db import *
+
+
+def main_process():
+    disk = ""
+    cate = ""
+    path = ""
+
+    if disk == "" or cate == "" or path == "":
+        print("检查参数")
+        return
+
+    file_list = get_file_list(path)
+    for f in file_list:
+        if md5 := get_md5_by_path(f):
+            print(f"请注意！！！文件的md5({md5})已经存在: {f}")
+        else:
+            md5 = get_file_md5(f)
+            insert_data((cate, disk, f, md5))
+            print(f"文件的md5({md5})成功插入: {f}")
+
+
+if __name__ == "__main__":
+    main_process()