http_uri处理源码

心已赠人 2022-05-08 12:18 114阅读 0赞

**目录**

1 接口总览

2 构造对象

3 析构对象

4 解析url参数

5 测试

--------------------

# 1 接口总览 #

url处理里面一共3个接口分别是构造对象，url解析以及析构对象

http_uri *http_uri_new(void);   
    void http_uri_destroy(http_uri *a_uri);
    int http_uri_parse(char *a_uri,http_uri *a_request);

#  2 构造对象 #

http\_uri\_new构造一个url对象相应结构体在前文已经介绍，这里配置一个默认端口为80

/*
     * 返回值:分配好资源的http_uri内容
     */
    http_uri *
    http_uri_new(void)
    {
        http_uri *l_return = NULL;
    
        l_return = (http_uri *)malloc(sizeof(http_uri));
        l_return->full = NULL;
        l_return->proto = NULL;
        l_return->host = NULL;
        l_return->port = 80;
        l_return->resource = NULL;
        return l_return;
    }

# 3 析构对象 #

当分配的资源使用完毕需要释放资源，则使用http\_uri\_destroy接口

/*
     * 输入参数:http_uri_new分配的http_uri资源
     */
    void
    http_uri_destroy(http_uri *a_uri)
    {
        if (a_uri->full) 
        {
            free(a_uri->full);
            a_uri->full = NULL;
        }
    
        if (a_uri->proto) 
        {
            free(a_uri->proto);
            a_uri->proto = NULL;
        }
    
        if (a_uri->host) 
        {
            free(a_uri->host);
            a_uri->host = NULL;
        }
    
        if (a_uri->resource) 
        {
            free(a_uri->resource);
            a_uri->resource = NULL;
        }
        free(a_uri);
    }

# 4 解析url参数 #

这里首先用到了一个枚举数据类型：记录解析项

typedef enum uri_parse_state_tag
    {
        parse_state_read_host = 0,
        parse_state_read_port,
        parse_state_read_resource
    } uri_parse_state;

http\_uri\_parse源码并不复杂，不过代码中考虑的比较全面，用起来还可以，不过这个解析函数写的并不全面，仅限于“http://ip:pirt/action”这种类型的URL，携带参数的无法解析，并且这个代码解析url的功能，个人感觉写的有些冗余。

/*
     * 输入参数:a_string，url字符串
     * 输出参数:a_uri，http_uri对象资源
     * 返回值: 0-成功 -1-失败
     */
    int
    http_uri_parse(char *a_string,http_uri *a_uri)
    {
        uri_parse_state l_state = parse_state_read_host;
        char *l_start_string = NULL;
        char *l_end_string = NULL;
        char  l_temp_port[6];
    
        /* init the array */
        memset(l_temp_port, 0, 6);
        if (a_string == NULL)
            goto ec;
    
        if (a_uri) 
        {
            a_uri->full = strdup(a_string);
        }
    
        //获取http://1.2.3.4:80/api/login中的第一个":"出现的位置
        l_start_string = strchr(a_string, ':');
        /* 检查确定是否存在":" */
        if (!l_start_string)
            goto ec;
        if (a_uri)//将url中的http字段复制到a_uri->proto里面
        {
            a_uri->proto = (char *)malloc(l_start_string - a_string + 1);
            memcpy(a_uri->proto, a_string, (l_start_string - a_string));
            a_uri->proto[l_start_string - a_string] = '\0';
        }
        /* 检查l_start_string指针的前3个字符是否是"://" */
        if (strncmp(l_start_string, "://", 3) != 0)
            goto ec;
        /* 将l_start_string和l_end_string指针指向url中剩余的部分，也就是ip地址或者域名的首位置,
        这里面l_end_string指针更类似于一个游标，l_start_string为参考起始位置*/
        l_start_string = l_end_string = &l_start_string[3];
        while(*l_end_string)
        {
            if (l_state == parse_state_read_host)
            {
                if (*l_end_string == ':')
                {
                    l_state = parse_state_read_port;//状态修改为获取端口
                    /*这块代码为什么进行了两次((l_end_string - l_start_string) == 0)检测，不理解*/
                    if ((l_end_string - l_start_string) == 0)
                        goto ec;
                    /* allocate space */
                    if ((l_end_string - l_start_string) == 0)
                        goto ec;
                    /* only do this if a uri was passed in */
                    if (a_uri)
                    {
                        a_uri->host = (char *)malloc(l_end_string - l_start_string + 1);
                        memcpy(a_uri->host, l_start_string, (l_end_string - l_start_string));
                        a_uri->host[l_end_string - l_start_string] = '\0';
                    }
                    /* reset the counters */
                    l_end_string++;//跳过检查的字符":"
                    l_start_string = l_end_string;
                    continue;
                }
                else if (*l_end_string == '/')
                {
                    l_state = parse_state_read_resource;
                    if ((l_end_string - l_start_string) == 0)
                        goto ec;
                    if (a_uri)
                    {
                        a_uri->host = (char *)malloc(l_end_string - l_start_string + 1);
                        memcpy(a_uri->host, l_start_string, (l_end_string - l_start_string));
                        a_uri->host[l_end_string - l_start_string] = '\0';
                    }
                    l_start_string = l_end_string;
                    continue;
                }
            }
            else if (l_state == parse_state_read_port)
            {
                if (*l_end_string == '/')
                {
                    l_state = parse_state_read_resource;
                    /* l_temp_port空间为6如果大于5说明错误 */
                    if (l_end_string - l_start_string > 5)
                        goto ec;
                    /* 始末位置相同则数据不存在 */
                    if ((l_end_string - l_start_string) == 0)
                        goto ec;
                    /* 蒋书记放进缓冲区 */
                    memcpy(l_temp_port, l_start_string, l_end_string - l_start_string);
                    /* convert it. */
                    if (a_uri)
                        a_uri->port = atoi(l_temp_port);
                    l_start_string = l_end_string;
                    continue;
                }
                else if (isdigit(*l_end_string) == 0)//若参数为阿拉伯数字0~9，则返回非0值，否则返回0
                {
                    /* 域名或者IP后面的冒号后面肯定是端口，在没有检索到"/"之前肯定为整数 ，不为整数则说明不是端口号，url出错*/
                    goto ec;
                }
            }
            /* next.. */
            l_end_string++;
            continue;
        }
    
        /*
         1.如果l_state仍然为parse_state_read_host说明url是"http://x.x.x.x",这也是合法的，因为http_uri_new默认端口是80
         1.如果l_state为parse_state_read_port说明url是"http://x.x.x.x:port",这也是合法的，说明访问的resource是根目录
        */
        if (l_state == parse_state_read_host)
        {
            if ((l_end_string - l_start_string) == 0)
                goto ec;
            if (a_uri)
            {
                a_uri->host = (char *)malloc(l_end_string - l_start_string + 1);
                memcpy(a_uri->host, l_start_string, (l_end_string - l_start_string));
                a_uri->host[l_end_string - l_start_string] = '\0';
                a_uri->resource = strdup("/");
            }
        }
        else if (l_state == parse_state_read_port)
        {
            if (strlen(l_start_string) == 0)
                goto ec;
            if (a_uri)
            {
                a_uri->port = atoi(l_start_string);
                a_uri->resource = strdup("/");
            }
        }
        else if (l_state == parse_state_read_resource)
        {
            if (strlen(l_start_string) == 0)
            {
                if (a_uri)
                    a_uri->resource = strdup("/");
            }
            else
            {
                if (a_uri)
                    a_uri->resource = strdup(l_start_string);
            }
        }
        else
        {
            goto ec;
        }
        return 0;
    ec:
        return -1;
    }

代码处理流程如下：

![70][]

# 5 测试 #

写了一段测试代码，用起来还不错，代码如下：

void test_url(char *url)
    {
        http_uri *uri = http_uri_new();
        http_uri_parse(url,uri);
        if (uri->full) 
            printf("uri->full:\t%s\n",uri->full);
    
        if (uri->proto) 
            printf("uri->proto:\t%s\n",uri->proto);
    
        if (uri->host) 
            printf("uri->host:\t%s\n",uri->host);
    
        if (uri->resource) 
            printf("uri->resource:\t%s\n",uri->resource);
        printf("uri->port:\t%d\n",uri->port);
        http_uri_destroy(uri);
    }
    
    int main(int argc, char **argv)
    {
        printf("\n#######test:http://1.1.1.1:9999/api/log#######\n");
        test_url("http://1.1.1.1:9999/api/log");
        printf("\n#######test:http://1.1.1.1:9999/##############\n");
        test_url("http://1.1.1.1:9999/");
        printf("\n#######test:http://1.1.1.1:9999###############\n");
        test_url("http://1.1.1.1:9999");
        printf("\n#######test:http://1.1.1.1#####################\n");
        test_url("http://1.1.1.1");
        return 0;
    }

代码执行结果如下：

#######test:http://1.1.1.1:9999/api/log#######
    uri->full:      http://1.1.1.1:9999/api/log
    uri->proto:     http
    uri->host:      1.1.1.1
    uri->resource:  /api/log
    uri->port:      9999
    
    #######test:http://1.1.1.1:9999/##############
    uri->full:      http://1.1.1.1:9999/
    uri->proto:     http
    uri->host:      1.1.1.1
    uri->resource:  /
    uri->port:      9999
    
    #######test:http://1.1.1.1:9999###############
    uri->full:      http://1.1.1.1:9999
    uri->proto:     http
    uri->host:      1.1.1.1
    uri->resource:  /
    uri->port:      9999
    
    #######test:http://1.1.1.1#####################
    uri->full:      http://1.1.1.1
    uri->proto:     http
    uri->host:      1.1.1.1
    uri->resource:  /
    uri->port:      80

[70]: /images/20220508/8dcc1aed7b614deda7cfde1fe85fe940.png