busybox/networking/httpd_ssi.c

/*
 * Copyright (c) 2009 Denys Vlasenko <vda.linux@googlemail.com>
 *
 * Licensed under GPLv2, see file LICENSE in this tarball for details.
 */

/*
 * This program is a CGI application. It processes server-side includes:
 * <!--#include file="file.html" -->
 *
 * Usage: put these lines in httpd.conf:
 *
 * *.html:/bin/httpd_ssi
 * *.htm:/bin/httpd_ssi
 */

/* Build a-la
i486-linux-uclibc-gcc \
-static -static-libgcc \
-D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 \
-Wall -Wshadow -Wwrite-strings -Wundef -Wstrict-prototypes -Werror \
-Wold-style-definition -Wdeclaration-after-statement -Wno-pointer-sign \
-Wmissing-prototypes -Wmissing-declarations \
-Os -fno-builtin-strlen -finline-limit=0 -fomit-frame-pointer \
-ffunction-sections -fdata-sections -fno-guess-branch-probability \
-funsigned-char \
-falign-functions=1 -falign-jumps=1 -falign-labels=1 -falign-loops=1 \
-march=i386 -mpreferred-stack-boundary=2 \
-Wl,-Map -Wl,link.map -Wl,--warn-common -Wl,--sort-common -Wl,--gc-sections \
httpd_ssi.c -o httpd_ssi
*/

/* Size (i386, static uclibc, approximate):
 * text    data     bss     dec     hex filename
 * 9487     160   68552   78199   13177 httpd_ssi
 *
 * Note: it wouldn't be too hard to get rid of stdio and strdup,
 * (especially that fgets() mangles NULs...)
 */

#include <sys/types.h>
#include <sys/stat.h>
#include <errno.h>
#include <fcntl.h>
#include <stdint.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <stdio.h>
#include <dirent.h>
#include <time.h>

static char* skip_whitespace(char *s)
{
        while (*s == ' ' || *s == '\t') ++s;

        return s;
}

static char line[64 * 1024];

static void process_includes(const char *filename)
{
	int curdir_fd;
	char *end;
	FILE *fp = fopen(filename, "r");
	if (!fp)
		exit(1);

	/* Ensure that nested includes are relative:
	 * if we include a/1.htm and it includes b/2.htm,
	 * we need to include a/b/2.htm, not b/2.htm
	 */
	curdir_fd = -1;
	end = strrchr(filename, '/');
	if (end) {
		curdir_fd = open(".", O_RDONLY);
		/* *end = '\0' would mishandle "/file.htm" */
		end[1] = '\0';
		chdir(filename);
	}

#define INCLUDE "<!--#include"
	while (fgets(line, sizeof(line), fp)) {
		unsigned preceding_len;
		char *include_directive;

		include_directive = strstr(line, INCLUDE);
		if (!include_directive) {
			fputs(line, stdout);
			continue;
		}
		preceding_len = include_directive - line;
		if (memchr(line, '\"', preceding_len)
		 || memchr(line, '\'', preceding_len)
		) {
			/* INCLUDE string may be inside "str" or 'str',
			 * ignore it */
			fputs(line, stdout);
			continue;
		}
		/* Small bug: we accept #includefile="file" too */
		include_directive = skip_whitespace(include_directive + sizeof(INCLUDE)-1);
		if (strncmp(include_directive, "file=\"", 6) != 0) {
			/* "<!--#include virtual=..."? - not supported */
			fputs(line, stdout);
			continue;
		}
		include_directive += 6; /* now it points to file name */
		end = strchr(include_directive, '\"');
		if (!end) {
			fputs(line, stdout);
			continue;
		}
		/* We checked that this is a valid include directive */

		/* Print everything before directive */
		if (preceding_len) {
			line[preceding_len] = '\0';
			fputs(line, stdout);
		}
		/* Save everything after directive */
		*end++ = '\0';
		end = strchr(end, '>');
		if (end)
			end = strdup(end + 1);

		/* FIXME:
		 * (1) are relative paths with /../ etc ok?
		 * (2) what to do with absolute paths?
		 * are they relative to doc root or to real root?
		 */
		process_includes(include_directive);

		/* Print everything after directive */
	        if (end) {
			fputs(end, stdout);
			free(end);
		}
	}
	if (curdir_fd >= 0)
		fchdir(curdir_fd);
	fclose(fp);
}

int main(int argc, char *argv[])
{
	if (!argv[1])
		return 1;

	/* Seen from busybox.net's Apache:
	 * HTTP/1.1 200 OK
	 * Date: Thu, 10 Sep 2009 18:23:28 GMT
	 * Server: Apache
	 * Accept-Ranges: bytes
	 * Connection: close
	 * Content-Type: text/html
	 */
	fputs(
		/* "Date: Thu, 10 Sep 2009 18:23:28 GMT\r\n" */
		/* "Server: Apache\r\n" */
		/* "Accept-Ranges: bytes\r\n" - do we really accept bytes?! */
		"Connection: close\r\n"
		"Content-Type: text/html\r\n"
		"\r\n",
		stdout
	);
	process_includes(argv[1]);
	return 0;
}
1	niro	984	/*
2			* Copyright (c) 2009 Denys Vlasenko <vda.linux@googlemail.com>
3			*
4			* Licensed under GPLv2, see file LICENSE in this tarball for details.
5			*/
6
7			/*
8			* This program is a CGI application. It processes server-side includes:
9			* <!--#include file="file.html" -->
10			*
11			* Usage: put these lines in httpd.conf:
12			*
13			* *.html:/bin/httpd_ssi
14			* *.htm:/bin/httpd_ssi
15			*/
16
17			/* Build a-la
18			i486-linux-uclibc-gcc \
19			-static -static-libgcc \
20			-D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 \
21			-Wall -Wshadow -Wwrite-strings -Wundef -Wstrict-prototypes -Werror \
22			-Wold-style-definition -Wdeclaration-after-statement -Wno-pointer-sign \
23			-Wmissing-prototypes -Wmissing-declarations \
24			-Os -fno-builtin-strlen -finline-limit=0 -fomit-frame-pointer \
25			-ffunction-sections -fdata-sections -fno-guess-branch-probability \
26			-funsigned-char \
27			-falign-functions=1 -falign-jumps=1 -falign-labels=1 -falign-loops=1 \
28			-march=i386 -mpreferred-stack-boundary=2 \
29			-Wl,-Map -Wl,link.map -Wl,--warn-common -Wl,--sort-common -Wl,--gc-sections \
30			httpd_ssi.c -o httpd_ssi
31			*/
32
33			/* Size (i386, static uclibc, approximate):
34			* text data bss dec hex filename
35			* 9487 160 68552 78199 13177 httpd_ssi
36			*
37			* Note: it wouldn't be too hard to get rid of stdio and strdup,
38			* (especially that fgets() mangles NULs...)
39			*/
40
41			#include <sys/types.h>
42			#include <sys/stat.h>
43			#include <errno.h>
44			#include <fcntl.h>
45			#include <stdint.h>
46			#include <stdlib.h>
47			#include <string.h>
48			#include <unistd.h>
49			#include <stdio.h>
50			#include <dirent.h>
51			#include <time.h>
52
53			static char* skip_whitespace(char *s)
54			{
55			while (s == ' ' \|\| s == '\t') ++s;
56
57			return s;
58			}
59
60			static char line[64 * 1024];
61
62			static void process_includes(const char *filename)
63			{
64			int curdir_fd;
65			char *end;
66			FILE *fp = fopen(filename, "r");
67			if (!fp)
68			exit(1);
69
70			/* Ensure that nested includes are relative:
71			* if we include a/1.htm and it includes b/2.htm,
72			* we need to include a/b/2.htm, not b/2.htm
73			*/
74			curdir_fd = -1;
75			end = strrchr(filename, '/');
76			if (end) {
77			curdir_fd = open(".", O_RDONLY);
78			/* end = '\0' would mishandle "/file.htm" /
79			end[1] = '\0';
80			chdir(filename);
81			}
82
83			#define INCLUDE "<!--#include"
84			while (fgets(line, sizeof(line), fp)) {
85			unsigned preceding_len;
86			char *include_directive;
87
88			include_directive = strstr(line, INCLUDE);
89			if (!include_directive) {
90			fputs(line, stdout);
91			continue;
92			}
93			preceding_len = include_directive - line;
94			if (memchr(line, '\"', preceding_len)
95			\|\| memchr(line, '\'', preceding_len)
96			) {
97			/* INCLUDE string may be inside "str" or 'str',
98			* ignore it */
99			fputs(line, stdout);
100			continue;
101			}
102			/* Small bug: we accept #includefile="file" too */
103			include_directive = skip_whitespace(include_directive + sizeof(INCLUDE)-1);
104			if (strncmp(include_directive, "file=\"", 6) != 0) {
105			/* "<!--#include virtual=..."? - not supported */
106			fputs(line, stdout);
107			continue;
108			}
109			include_directive += 6; /* now it points to file name */
110			end = strchr(include_directive, '\"');
111			if (!end) {
112			fputs(line, stdout);
113			continue;
114			}
115			/* We checked that this is a valid include directive */
116
117			/* Print everything before directive */
118			if (preceding_len) {
119			line[preceding_len] = '\0';
120			fputs(line, stdout);
121			}
122			/* Save everything after directive */
123			*end++ = '\0';
124			end = strchr(end, '>');
125			if (end)
126			end = strdup(end + 1);
127
128			/* FIXME:
129			* (1) are relative paths with /../ etc ok?
130			* (2) what to do with absolute paths?
131			* are they relative to doc root or to real root?
132			*/
133			process_includes(include_directive);
134
135			/* Print everything after directive */
136			if (end) {
137			fputs(end, stdout);
138			free(end);
139			}
140			}
141			if (curdir_fd >= 0)
142			fchdir(curdir_fd);
143			fclose(fp);
144			}
145
146			int main(int argc, char *argv[])
147			{
148			if (!argv[1])
149			return 1;
150
151			/* Seen from busybox.net's Apache:
152			* HTTP/1.1 200 OK
153			* Date: Thu, 10 Sep 2009 18:23:28 GMT
154			* Server: Apache
155			* Accept-Ranges: bytes
156			* Connection: close
157			* Content-Type: text/html
158			*/
159			fputs(
160			/* "Date: Thu, 10 Sep 2009 18:23:28 GMT\r\n" */
161			/* "Server: Apache\r\n" */
162			/* "Accept-Ranges: bytes\r\n" - do we really accept bytes?! */
163			"Connection: close\r\n"
164			"Content-Type: text/html\r\n"
165			"\r\n",
166			stdout
167			);
168			process_includes(argv[1]);
169			return 0;
170			}