C: Linux Socket Programming, TCP, a simple HTTP client - 安東尼隨手記 - Yahoo!奇摩部落格

檢舉

C: Linux Socket Programming, TCP, a simple HTTP client

分類： Linux程式設計

2009/04/16 10:46

分享

Facebook
Plurk
YAHOO!

• 分享在我的 Facebook
• 分享在我的 Plurk
• 分享在我的即時通

Copy from http://coding.debuntu.org/c-linux-socket-programming-tcp-simple-http-client

1. #include <stdio.h>
2. #include <sys/socket.h>
3. #include <arpa/inet.h>
4. #include <stdlib.h>
5. #include <netdb.h>
6. #include <string.h>
7. int create_tcp_socket ( );
8. char *get_ip ( char *host );
9. char *build_get_query ( char *host, char *page );
10. void usage ( );
11. #define HOST "coding.debuntu.org"
12. #define PAGE "/"
13. #define PORT 80
14. #define USERAGENT "HTMLGET 1.0"
15. int main ( int argc, char **argv )
16. {
17. struct sockaddr_in *remote;
18. int sock;
19. int tmpres;
20. char *ip;
21. char *get;
22. char buf [BUFSIZ +1 ];
23. char *host;
24. char *page;
25. if (argc == 1 ) {
26. usage ( );
27. exit ( 2 );
28. }  
29. host = argv [ 1 ];
30. if (argc > 2 ) {
31. page = argv [ 2 ];
32. } else {
33. page = PAGE;
34. }
35. sock = create_tcp_socket ( );
36. ip = get_ip (host );
37. fprintf (stderr, "IP is %s\n", ip );
38. remote = ( struct sockaddr_in * )malloc ( sizeof ( struct sockaddr_in * ) );
39. remote ->sin_family = AF_INET;
40. tmpres = inet_pton (AF_INET, ip, ( void * ) ( & (remote ->sin_addr. s_addr ) ) );
41. if ( tmpres < 0 )  
42. {
43. perror ( "Can't set remote->sin_addr.s_addr" );
44. exit ( 1 );
45. } else if (tmpres == 0 )
46. {
47. fprintf (stderr, "%s is not a valid IP address\n", ip );
48. exit ( 1 );
49. }
50. remote ->sin_port = htons (PORT );
51. if (connect (sock, ( struct sockaddr * )remote, sizeof ( struct sockaddr ) ) < 0 ) {
52. perror ( "Could not connect" );
53. exit ( 1 );
54. }
55. get = build_get_query (host, page );
56. fprintf (stderr, "Query is:\n<<START>>\n%s<<END>>\n", get );
57. //Send the query to the server
58. int sent = 0;
59. while (sent < strlen (get ) )
60. {
61. tmpres = send (sock, get +sent, strlen (get ) -sent, 0 );
62. if (tmpres == -1 ) {
63. perror ( "Can't send query" );
64. exit ( 1 );
65. }
66. sent += tmpres;
67. }
68. //now it is time to receive the page
69. memset (buf, 0, sizeof (buf ) );
70. int htmlstart = 0;
71. char * htmlcontent;
72. while ( (tmpres = recv (sock, buf, BUFSIZ, 0 ) ) > 0 ) {
73. if (htmlstart == 0 )
74. {
75. /* Under certain conditions this will not work.
76.       * If the \r\n\r\n part is splitted into two messages
77.       * it will fail to detect the beginning of HTML content
78.       */
79. htmlcontent = strstr (buf, "\r\n\r\n" );
80. if (htmlcontent != NULL ) {
81. htmlstart = 1;
82. htmlcontent += 4;
83. }
84. } else {
85. htmlcontent = buf;
86. }
87. if (htmlstart ) {
88. fprintf (stdout, htmlcontent );
89. }
90. memset (buf, 0, tmpres );
91. }
92. if (tmpres < 0 )
93. {
94. perror ( "Error receiving data" );
95. }
96. free (get );
97. free (remote );
98. free (ip );
99. close (sock );
100. return 0;
101. }
102. void usage ( )
103. {
104. fprintf (stderr, "USAGE: htmlget host [page]\n\
105. \thost: the website hostname. ex: coding.debuntu.org\n\
106. \tpage: the page to retrieve. ex: index.html, default: /\n" );
107. }
108. int create_tcp_socket ( )
109. {
110. int sock;
111. if ( (sock = socket (AF_INET, SOCK_STREAM, IPPROTO_TCP ) ) < 0 ) {
112. perror ( "Can't create TCP socket" );
113. exit ( 1 );
114. }
115. return sock;
116. }
117. char *get_ip ( char *host )
118. {
119. struct hostent *hent;
120. int iplen = 15; //XXX.XXX.XXX.XXX
121. char *ip = ( char * )malloc (iplen +1 );
122. memset (ip, 0, iplen +1 );
123. if ( (hent = gethostbyname (host ) ) == NULL )
124. {
125. herror ( "Can't get IP" );
126. exit ( 1 );
127. }
128. if (inet_ntop (AF_INET, ( void * )hent ->h_addr_list [ 0 ], ip, iplen ) == NULL )
129. {
130. perror ( "Can't resolve host" );
131. exit ( 1 );
132. }
133. return ip;
134. }
135. char *build_get_query ( char *host, char *page )
136. {
137. char *query;
138. char *getpage = page;
139. char *tpl = "GET /%s HTTP/1.0\r\nHost: %s\r\nUser-Agent: %s\r\n\r\n";
140. if (getpage [ 0 ] == '/' ) {
141. getpage = getpage + 1;
142. fprintf (stderr, "Removing leading \"/\", converting %s to %s\n", page, getpage );
143. }
144. // -5 is to consider the %s %s %s in tpl and the ending \0
145. query = ( char * )malloc (strlen (host ) +strlen (getpage ) +strlen (USERAGENT ) +strlen (tpl ) -5 );
146. sprintf (query, tpl, getpage, host, USERAGENT );
147. return query;
148. }

To compile it, run:

$ gcc -o htmlget htmlget.c
$ ./htmlget
USAGE: htmlget host [page]host: the website hostname. ex: coding.debuntu.orgpage: the page to retrieve. ex: index.html, default: /

Informative messages and errors are printed to stderr. The content of the page is printed to stdout. Thus, to save the HTML content of a page to a file, you will need to run:

$ ./htmlget coding.debuntu.org category > /tmp/page.html