From 3a62b4e3ef14e62f1e5ff55069e133f0b1858d95 Mon Sep 17 00:00:00 2001 From: lishu Date: Tue, 19 Nov 2019 17:40:49 +0800 Subject: add http_host_parser --- src/HTTP_Parser.c | 288 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 288 insertions(+) create mode 100644 src/HTTP_Parser.c (limited to 'src/HTTP_Parser.c') diff --git a/src/HTTP_Parser.c b/src/HTTP_Parser.c new file mode 100644 index 0000000..cbb8437 --- /dev/null +++ b/src/HTTP_Parser.c @@ -0,0 +1,288 @@ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "HTTP_Common.h" + +typedef struct host_parser_t +{ + char* host_value; + char is_http_flag; + char get_host_flag; + char host_field_flag; + char host_value_flag; + uint32 host_valuelen; +}host_parser; + +int http_parser_callback_on_headers_field(struct http_parser * parser, const char * at, size_t length) +{ + if(((host_parser*)(parser->data))->get_host_flag==0 || ((host_parser*)(parser->data))->host_field_flag==1) return 0; + if(0==strncasecmp(at, "host",strlen("host"))) + { + ((host_parser*)(parser->data))->host_field_flag = 1; + } + return 0; +} + +int http_parser_callback_on_headers_value(struct http_parser * parser, const char * at, size_t length) +{ + if(((host_parser*)(parser->data))->get_host_flag==0 || ((host_parser*)(parser->data))->host_field_flag==0 || ((host_parser*)(parser->data))->host_value_flag==1) return 0; + if(0data))->host_value = (char*)malloc(length); + memcpy(((host_parser*)(parser->data))->host_value, at, length); + ((host_parser*)(parser->data))->host_valuelen = length; + ((host_parser*)(parser->data))->host_value_flag = 1; + } + return 0; +} + +int http_parser_callback_on_url(struct http_parser * parser, const char * at, size_t length) +{ + ((host_parser*)(parser->data))->is_http_flag = 1; + return 0; +} + +int http_parser_callback_on_status(struct http_parser * parser, const char * at, size_t length) +{ + ((host_parser*)(parser->data))->is_http_flag = 1; + return 0; +} + +static http_parser_settings http_setting = +{ + .on_message_begin = NULL, + .on_url = http_parser_callback_on_url, + .on_status = http_parser_callback_on_status, + .on_header_field = http_parser_callback_on_headers_field, + .on_header_value = http_parser_callback_on_headers_value, + .on_headers_complete = NULL, + .on_body = NULL, + .on_message_complete = NULL, + .on_chunk_header = NULL, + .on_chunk_complete = NULL +}; + +int http_host_parser(const char* buf, uint32 buflen, int http_dir, char** host) +{ + //buf前面的空格进行删除处理 + uint32 offset = 0; + http_deleteEmptyRow(&offset, (char*)(buf), buflen); + const char* pbuf = buf+offset; + uint32 pbuflen = buflen-offset; + + printf("buf=====================================================\n"); + printf("%s\n", buf); + + // 为结构体申请内存 + http_parser *parser = (http_parser*)calloc(1, sizeof(http_parser)); + host_parser* host_field = (host_parser*)calloc(1, sizeof(host_parser)); + int rec = -1; + + // 初始化解析器 + if(http_dir==DIR_C2S) + { + http_parser_init(parser, HTTP_REQUEST); + } + else + { + http_parser_init(parser, HTTP_RESPONSE); + } + parser->data = (void*)host_field; + + //获取host内容 + if(host!=NULL) + { + ((host_parser*)(parser->data))->get_host_flag = 1; + } + + //执行解析过程 + size_t parsed = http_parser_execute(parser, &http_setting, pbuf, (size_t)pbuflen); + printf("buflen: %d; parsed: %d\n", pbuflen, parsed); + if(((host_parser*)(parser->data))->is_http_flag == 1) + { + rec = 0; + if(((host_parser*)(parser->data))->get_host_flag == 1) + { + if(0host_valuelen) + { + rec = host_field->host_valuelen; + *host = memcasemem(pbuf, pbuflen, (const char*)host_field->host_value, host_field->host_valuelen); + } + } + } + else + { + rec = -1; + } + + if(NULL!=host_field) + { + if(NULL!=host_field->host_value) + { + free(host_field->host_value); + host_field->host_value = NULL; + } + free(host_field); + host_field = NULL; + } + return rec; +} + +/*http_host_parser_test 协议识别正常,host解析正常*/ +static const char * http_request = + "POST /gen_204 HTTP/1.1\r\n" + "Host: www.google.com\r\n" + "Connection: close\r\n" + "Content-Length: 0\r\n" + "Origin: https://www.google.com\r\n" + "User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/69.0.3497.100 Safari/537.36\r\n" + "Content-Type: text/plain;charset=UTF-8\r\n" + "Accept: */*\r\n" + "X-Client-Data: CJG2yQEIorbJAQjEtskBCKmdygEI2J3KAQjZncoBCKijygEY+aXKAQ==\r\n" + "Referer: https://www.google.com/\r\n" + "Accept-Encoding: gzip, deflate\r\n" + "Accept-Language: zh-CN,zh;q=0.9,en-US;q=0.8,en;q=0.7\r\n"; + +/*http_host_parser_test 协议识别正常,无host字段*/ +static const char * http_request_no_host = + "POST /gen_204 HTTP/1.1\r\n" + "Connection: close\r\n" + "Content-Length: 0\r\n" + "Origin: https://www.google.com\r\n" + "User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/69.0.3497.100 Safari/537.36\r\n" + "Content-Type: text/plain;charset=UTF-8\r\n" + "Accept: */*\r\n" + "X-Client-Data: CJG2yQEIorbJAQjEtskBCKmdygEI2J3KAQjZncoBCKijygEY+aXKAQ==\r\n" + "Referer: https://www.google.com/\r\n" + "Accept-Encoding: gzip, deflate\r\n" + "Accept-Language: zh-CN,zh;q=0.9,en-US;q=0.8,en;q=0.7\r\n"; + +/*http_host_parser_test 协议识别正常*/ +static const char * http_response = + "HTTP/1.1 200 OK\r\nAccept-Ranges: bytes\r\nVary: Accept-Encoding\r\nContent-Type: text/javascript; charset=UTF-8\r\nContent-Length: 4064\r\nAge: 8963\r\n"; + +/*http_host_parser_test 协议识别正常，host解析失败，不支持\r*/ +static const char * http_request_CR = + "POST /gen_204 HTTP/1.1\r" + "Host: www.google.com\r" + "Connection: close\r" + "Content-Length: 0\r" + "Origin: https://www.google.com\r" + "User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/69.0.3497.100 Safari/537.36\r\n" + "Content-Type: text/plain;charset=UTF-8\r"; + +/*http_host_parser_test 协议识别正常，解析失败*/ +static const char * http_reponse_CR = + "HTTP/1.1 200 OK\r" + "Accept-Ranges: bytes\r" + "Vary: Accept-Encoding\r\nContent-Type: text/javascript; charset=UTF-8\r\nContent-Length: 4064\r\nAge: 8963\r\n"; + +/*/*http_host_parser_test 协议识别情况正常，但是host为close，不支持此情况*/ +static const char * http_request_LF = + "POST /gen_204 HTTP/1.1\n" + "host: \n" + "Connection: close\n" + "Content-Length: 0\n" + "Origin: https://www.google.com\n"; + +/*http_host_parser_test 协议识别正常,host解析正常*/ +static const char * http_request_SPACE = + "POST /gen_204 HTTP/1.1\r\n" + "HOST: www. google.com \r\n"; + +/*http_host_parser_test 协议识别正常,host解析正常*/ +static const char * http_request_incpomplete = + "POST /gen_204 HTTP/1.1\n" + "Host: www.googl"; + +/*http_host_parser_test 协议识别正常,host解析正常*/ +static const char * http_request_with_body = + "POST /gen_204 HTTP/1.1\r\n" + "Host: www.google.com\r\n" + "Connection: close\r\n" + "Content-Length: 0\r\n" + "Origin: https://www.google.com\r\n" + "User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/69.0.3497.100 Safari/537.36\r\n" + "Content-Type: text/plain;charset=UTF-8\r\n" + "Accept: */*\r\n" + "X-Client-Data: CJG2yQEIorbJAQjEtskBCKmdygEI2J3KAQjZncoBCKijygEY+aXKAQ==\r\n" + "Referer: https://www.google.com/\r\n" + "Accept-Encoding: gzip, deflate\r\n" + "Accept-Language: zh-CN,zh;q=0.9,en-US;q=0.8,en;q=0.7\r\n" + "\r\n" + "abcdjhjkhdlaks"; + +/*http_host_parser_test 协议识别正常,host解析正常*/ +static const char * http_request_space_start = + "\r \nPOST /gen_204 HTTP/1.1\r\n" + "Host: www.google.com\r\n" + "Connection: close\r\n" + "Content-Length: 0\r\n" + "Origin: https://www.google.com\r\n" + "User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/69.0.3497.100 Safari/537.36\r\n" + "Content-Type: text/plain;charset=UTF-8\r\n" + "Accept: */*\r\n" + "X-Client-Data: CJG2yQEIorbJAQjEtskBCKmdygEI2J3KAQjZncoBCKijygEY+aXKAQ==\r\n" + "Referer: https://www.google.com/\r\n" + "Accept-Encoding: gzip, deflate\r\n" + "Accept-Language: zh-CN,zh;q=0.9,en-US;q=0.8,en;q=0.7\r\n" + "\r\n" + "abcdjhjkhdlaks"; + + +void http_host_parser_test(const char* test_buf, int test_buf_dir) +{ + char* host = NULL; + uint32 hostlen = 0; + int rec = -1; + + /*判断是否是http协议数据*/ + rec = http_host_parser((const char*)test_buf, (uint32)strlen(test_buf), test_buf_dir, NULL); + if(-1==rec) + { + printf("not http data\n"); + } + else if(0==rec) + { + printf("is http data\n"); + } + + /*判断是否是http协议数据，并获取host字段内容*/ + rec = http_host_parser((const char*)test_buf, (uint32)strlen(test_buf), test_buf_dir, &host); + if(-1==rec) + { + printf("not http data\n"); + } + else if(0<=rec) + { + printf("is http data\n"); + if(0