sqlite之我见--C/C++ API接口示例

在之前的两篇博文中,分别介绍了SQLITE的基础知识和操作,C/C++ 的一些常用API

sqlite之我见--简单介绍与基本操作

sqlite之我见--C/C++ API接口介绍 


本文中,我会给大家用几个小程序示例SQLITE C/C++ API的使用。

1.我们看下最简单的sqlite程序,通过sqlite3_open, sqlite3_exec, sqlite3_close来实现一个简单的数据库操作。

/* 
 * File:   sqlite_test.cpp
 * Author: Carl
 *
 * Created on September 20, 2012, 3:28 PM
 */

#include <cstdlib>
#include <cstdio>
#include <sqlite3.h>

static int _sql_callback(void *notused, int argc, char **argv, char **szColName)
{
    int i = 0;
    
    printf("notused:0x%x, argc:%d\n", notused, argc);
    for (i = 0; i < argc; i++)
    {
        printf("%s = %s\n", szColName[i], argv[i] == 0 ? "NULL" : argv[i]);
    }
    printf("\n");
    
    return 0;
}

/*
 * 
 */
int main(int argc, char** argv)
{
    const char *sSQL1 = "create table users(userid varchar(20) PRIMARY KEY, age int, birthday datetime);";
    const char *sSQL2 = "insert into users values('wang', 20, '1989-5-4');";
    const char *sSQL3 = "select * from users;";
    
    sqlite3 *db = 0;
    char *pErrMsg = 0;
    int ret = 0;
    
    //连接数据库
    ret = sqlite3_open("./test.db", &db);
    if (ret != SQLITE_OK)
    {
        fprintf(stderr, "无法打开数据库:%s\n", sqlite3_errmsg(db));
        sqlite3_close(db);
        return 1;
    }
    printf("数据库连接成功\n");
    
    //执行建表SQL
    ret = sqlite3_exec(db, sSQL1, _sql_callback, 0, &pErrMsg);
    if (ret != SQLITE_OK)
    {
        fprintf(stderr, "SQL create error: %s\n", pErrMsg);
        sqlite3_free(pErrMsg); //这个要的哦,要不然会内存泄露的哦!!!
        sqlite3_close(db);
        return 1;
    }
    printf("数据库建表成功!!\n");
    
    //执行插入数据
    ret = sqlite3_exec(db, sSQL2, _sql_callback, 0, &pErrMsg);
    if (ret != SQLITE_OK)
    {
        fprintf(stderr, "SQL insert error: %s\n", pErrMsg);
        sqlite3_free(pErrMsg); //这个要的哦,要不然会内存泄露的哦!!!
        sqlite3_close(db);
        return 1;
    }
    printf("数据库插入数据成功!\n");
    
    //执行查询操作
    ret = sqlite3_exec(db, sSQL3, _sql_callback, 0, &pErrMsg);
    if (ret != SQLITE_OK)
    {
        fprintf(stderr, "SQL error: %s\n", pErrMsg);
        sqlite3_free(pErrMsg);
        sqlite3_close(db);
        return 1;
    }
    printf("数据库查询成功!!\n");
    
    //关闭数据库
    sqlite3_close(db);
    db = 0;
    
    return 0;
}


运行结果如下结果:

[carl@Fedora sqlite]$ g++ sqlite_test.cpp -lsqlite3
[carl@Fedora sqlite]$ ./a.out 
数据库连接成功
数据库建表成功!!
数据库插入数据成功!
notused:0x0, argc:3
userid = wang
age = 20
birthday = 1989-5-4

数据库查询成功!!
[carl@Fedora sqlite]$

2. 我们再看一个在sqlite上是有事务来实现原子操作的的例子,

代码如下:

/* 
 * File:   sqlite_test.cpp
 * Author: Carl
 *
 * Created on 2012年9月22日, 上午7:50
 */

#include <cstdlib>
#include <cstdio>
#include <sqlite3.h>

static int _sql_callback(void *notused, int argc, char **argv, char **szColName)
{
    int i = 0;
    
    printf("notused:0x%x, argc:%d\n", notused, argc);
    for (i = 0; i < argc; i++)
    {
        printf("%s = %s\n", szColName[i], argv[i] == 0 ? "NULL" : argv[i]);
    }
    printf("\n");
    
    return 0;
}


/*
 * 
 */
int main(int argc, char** argv)
{
    const char *sSQL1 = "create table test_for_cpp (id int, name varchar(10), age int);";
    char sql[100] = {0};
    sqlite3 *db = NULL;
    char *pErrMsg = NULL;
    int ret = 0;
    bool is_success = true;
    const char *sSQL3 = "select * from test_for_cpp;";  
    
    ret = sqlite3_open("./test.db", &db);
    if (SQLITE_OK != ret)
    {
        fprintf(stderr, "无法打开数据库: %s\n", sqlite3_errmsg(db));
        sqlite3_close(db);
        return 1;
    }
    printf("数据库连接成功\n");  
    
    ret = sqlite3_exec(db, sSQL1, NULL, 0, &pErrMsg);
    if (SQLITE_OK != ret)
    {
        fprintf(stderr, "SQL create error: %s\n", pErrMsg);
        sqlite3_free(pErrMsg);
        sqlite3_close(db);
        return 1;
    }
    printf("数据库建表成功!!\n");  
    
    sqlite3_exec(db, "begin;", _sql_callback, 0, &pErrMsg); //开启事务
    if (SQLITE_OK != ret)
    {
        fprintf(stderr, "SQL begin error: %s\n", pErrMsg);
        sqlite3_free(pErrMsg);
        sqlite3_close(db);
        return 1;
    }
    printf("数据库开启事务成功!!\n");  
    
    for (int i = 0; i < 10; i++)
    {
        sprintf(sql, "insert into test_for_cpp(id, name, age) values(%d, \"%s\", %d);", i, "Carl", i);
        ret = sqlite3_exec(db, sql, _sql_callback, 0, &pErrMsg);
        if (SQLITE_OK != ret)
        {
            is_success = false;
            fprintf(stderr, "for %d time error: %s\n", i, pErrMsg);
            sqlite3_free(pErrMsg);
            break;
        }
    }
    
    if (is_success)
    {
        sqlite3_exec(db, "commit;", 0, 0, 0);
        printf("数据库插入数据成功!\n");  
    }
    else
    {
        sqlite3_exec(db, "rollback;", 0, 0, 0);
        printf("数据库插入数据失败!\n");  
    }
    
    ret = sqlite3_exec(db, sSQL3, _sql_callback, 0, &pErrMsg);
    if (SQLITE_OK != ret)
    {
        fprintf(stderr, "SQL ERROR: %s\n", pErrMsg);
        sqlite3_free(pErrMsg);
        sqlite3_close(db);
        return 1;
    }
    printf("数据库查询成功!!\n");  
    
    sqlite3_close(db);
    db = 0;
    
    return 0;
}

运行结果:

[carl@Fedora sqlite]$ g++ sqlite_test.cpp -lsqlite3
[carl@Fedora sqlite]$ ./a.out 
数据库连接成功
数据库建表成功!!
数据库开启事务成功!!
数据库插入数据成功!
notused:0x0, argc:3
id = 0
name = Carl
age = 0

notused:0x0, argc:3
id = 1
name = Carl
age = 1

notused:0x0, argc:3
id = 2
name = Carl
age = 2

notused:0x0, argc:3
id = 3
name = Carl
age = 3

notused:0x0, argc:3
id = 4
name = Carl
age = 4

notused:0x0, argc:3
id = 5
name = Carl
age = 5

notused:0x0, argc:3
id = 6
name = Carl
age = 6

notused:0x0, argc:3
id = 7
name = Carl
age = 7

notused:0x0, argc:3
id = 8
name = Carl
age = 8

notused:0x0, argc:3
id = 9
name = Carl
age = 9

数据库查询成功!!
[carl@Fedora sqlite]$ 


3. 我们接着看一下如何能够更好的使用语句参数来操作sqlite数据库,用sqlite3_prepare_v2, sqlite3_bind_*, sqlite3_step, sqlite3_column_*等接口来实现对数据库的操作。

代码如下:里面的注释,有兴趣的可以试着打开试一下,但要记得注释掉相关的重复功能的语句哦。

/* 
 * File:   sqlite_test2.cpp
 * Author: Carl
 *
 * Created on September 21, 2012, 3:12 PM
 */

#include <cstdlib>
#include <cstdio>
#include <cstring>
#include <sqlite3.h>

static int _sql_callback(void *notused, int argc, char **argv, char **szColName)
{
    int i = 0;
    
    printf("notused:0x%x, argc:%d\n", notused, argc);
    for (i = 0; i < argc; i++)
    {
        printf("%s = %s\n", szColName[i], argv[i] == 0 ? "NULL" : argv[i]);
    }
    printf("\n");
    
    return 0;
}

/*
 * 
 */
int main(int argc, char** argv)
{
    sqlite3 *conn = NULL;
    sqlite3_stmt *stmt = NULL;
    char *err_msg = NULL;
    int ret = 0;
    
    char col_types[][10] = {"", "Interger", "Float", "Text", "Blob", "NULL"};
    
    ret = sqlite3_open("./test.db", &conn);
    if (SQLITE_OK != ret)
    {
        fprintf(stderr, "sqlite open err, %d\n", ret);
        return 1;
    }
    printf("打开数据库成功!!!\n");
    
//    ret = sqlite3_prepare_v2(conn, "SELECT * FROM [test_for_cpp] WHERE [name]==:name", -1, &stmt, (const char **)&err_msg);
    ret = sqlite3_prepare_v2(conn, "SELECT * FROM [test_for_cpp] WHERE [name]==?2", -1, &stmt, (const char **)&err_msg);
    if (SQLITE_OK != ret)
    {
        fprintf(stderr, "sqlite prepare error: %s\n", err_msg);
        sqlite3_free(err_msg);
        sqlite3_close(conn);
        return 1;
    }
//    printf("数据库语句对象编译成功!!!%d\n", sqlite3_bind_parameter_index(stmt, ":name"));
    printf("数据库语句对象编译成功!!!\n");
    
    ret = sqlite3_bind_text(stmt, 2, "Carl", 4, SQLITE_STATIC);
    if (SQLITE_OK != ret)
    {
        fprintf(stderr, "sqlite bind error: %d\n", ret);
        sqlite3_close(conn);
        return 1;
    }
    printf("数据库语句对象bind成功!!!\n");
    
    while (ret = sqlite3_step(stmt), ret == SQLITE_ROW)
    {
        int col_count = sqlite3_column_count(stmt); //结果集中列的数量
        printf("列数:%d\t", col_count);
        const char *col_0_name = sqlite3_column_name(stmt, 0); //获取列名
        printf("列名:%s\t", col_0_name);
        int id = sqlite3_column_int(stmt, 0);
        printf("id值:%d\t", id);
        int id_type = sqlite3_column_type(stmt, 0); //获取列数据类型
        printf("id类型:%d\t", id_type);
        
        const char *col_2_name = sqlite3_column_name(stmt, 2);
        int age = sqlite3_column_int(stmt, 2);
        int age_type = sqlite3_column_type(stmt, 2);
        
        const char *col_1_name = sqlite3_column_name(stmt, 1);
        char name[80];
        strncpy(name, (const char *)sqlite3_column_text(stmt, 1), 80);
        int name_type = sqlite3_column_type(stmt, 1);
        
        //打印结果
        printf("col_count: %d, %s = %d(%s), %s = %s(%s), %s = %d(%s)\n", 
               col_count, col_0_name, id, col_types[id_type], col_1_name, name,
               col_types[name_type], col_2_name, age, col_types[age_type]);
    }
    
    fprintf(stderr, "sqlite step exit with %d\n", ret);
    sqlite3_finalize(stmt);
    sqlite3_close(conn);
    
    return 0;
}


运行结果如下:

[carl@Fedora sqlite]$ g++ sqlite_test2.cpp -lsqlite3
[carl@Fedora sqlite]$ ./a.out 
打开数据库成功!!!
数据库语句对象编译成功!!!
数据库语句对象bind成功!!!
列数:3	列名:id	id值:0	id类型:1	col_count: 3, id = 0(Interger), name = Carl(Text), age = 0(Interger)
列数:3	列名:id	id值:1	id类型:1	col_count: 3, id = 1(Interger), name = Carl(Text), age = 1(Interger)
列数:3	列名:id	id值:2	id类型:1	col_count: 3, id = 2(Interger), name = Carl(Text), age = 2(Interger)
列数:3	列名:id	id值:3	id类型:1	col_count: 3, id = 3(Interger), name = Carl(Text), age = 3(Interger)
列数:3	列名:id	id值:4	id类型:1	col_count: 3, id = 4(Interger), name = Carl(Text), age = 4(Interger)
列数:3	列名:id	id值:5	id类型:1	col_count: 3, id = 5(Interger), name = Carl(Text), age = 5(Interger)
列数:3	列名:id	id值:6	id类型:1	col_count: 3, id = 6(Interger), name = Carl(Text), age = 6(Interger)
列数:3	列名:id	id值:7	id类型:1	col_count: 3, id = 7(Interger), name = Carl(Text), age = 7(Interger)
列数:3	列名:id	id值:8	id类型:1	col_count: 3, id = 8(Interger), name = Carl(Text), age = 8(Interger)
列数:3	列名:id	id值:9	id类型:1	col_count: 3, id = 9(Interger), name = Carl(Text), age = 9(Interger)
sqlite step exit with 101 //101意思为SQLITE_DONE
[carl@Fedora sqlite]$ 


下面解释下为什么要学会这种能够细致控制sqlite的方法(即使用语句参数),主要有以下几个优点:

(1) 使用“语句参数”方式,具有更高的安全性,可以有效防止“SQL注入攻击”。 “SQL注入攻击”要想达到目的,就必须让attack value随着SQL命令字符串一起传送进SQL解析器。黑客如果在一条SQL命令字符串被送入到sqlite3_prepare函数之前,利用c字符串处理函数等途径将attack value注入其中,而在sqlite3_prepare函数之中进行解析(parse),就可以达到攻击目的。而使用“语句参数”方式,被传送到sqlite3_prepare函数的只是SQL命令字符串中的参数符号(如:“?”),而不是具体的值。在sqlite3_prepare函数执行之后,才会使用bind函数给参数符号绑定具体的值,这就可以避免attack value随着SQL命令字符串一起在sqlite3_prepare函数中被解析,从而有效躲避“SQL注入攻击”。
(2)使用“语句参数”方式,可以更快的完成值替换。

(3)使用“语句参数”方式,更节省内存。原因是使用如snprintf函数,需要一个SQL命令模板,一块足够大的输出缓存,而且字符串处理函数需要工作内存(working memory),除此之外对于整形,浮点型,特别是BLOBs,经常会占用更多的空间。


水平有限,如果有朋友发现错误,欢迎留言交流。
转载请保留本文链接,如果觉得我的文章能帮到您,请顶一下。,谢谢。

你可能感兴趣的:(sqlite之我见--C/C++ API接口示例)