nuttx-apps/netutils/thttpd/cgi-src/ssi.c

964 lines
22 KiB
C

/****************************************************************************
* netutils/thttpd/cgi-src/ssi.c
* Server-side-includes CGI program
*
* Copyright (C) 2009, 2015 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Derived from the file of the same name in the original THTTPD package:
*
* Copyright © 1995 by Jef Poskanzer <jef@mail.acme.com>.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
****************************************************************************/
/****************************************************************************
* Included Files
****************************************************************************/
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/time.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <errno.h>
#include <nuttx/lib/regex.h>
#include "config.h"
/****************************************************************************
* Pre-Processor Definitions
****************************************************************************/
#define ST_GROUND 0
#define ST_LESSTHAN 1
#define ST_BANG 2
#define ST_MINUS1 3
#define ST_MINUS2 4
#define SF_BYTES 0
#define SF_ABBREV 1
#define DI_CONFIG 0
#define DI_INCLUDE 1
#define DI_ECHO 2
#define DI_FSIZE 3
#define DI_FLASTMOD 4
#define BUFFER_SIZE 512
#define TIMEFMT_SIZE 80
#define MAX_TAGS 32
/****************************************************************************
* Private Function Prototypes
****************************************************************************/
static void read_file(FILE *instream, char *vfilename, char *filename);
/****************************************************************************
* Private Data
****************************************************************************/
static char *g_url;
static char g_timeformat[TIMEFMT_SIZE];
static char g_iobuffer1[BUFFER_SIZE];
static char g_iobuffer2[BUFFER_SIZE];
static char *g_tags[MAX_TAGS];
static int g_sizefmt;
static struct stat g_sb;
/****************************************************************************
* Private Functions
****************************************************************************/
static void internal_error(char *reason)
{
char *title = "500 Internal Error";
(void)printf("\
<HTML><HEAD><TITLE>%s</TITLE></HEAD>\n\
<BODY><H2>%s</H2>\n\
Something unusual went wrong during a server-side-includes request:\n\
<BLOCKQUOTE>\n\
%s\n\
</BLOCKQUOTE>\n\
</BODY></HTML>\n", title, title, reason);
}
static void not_found(char *filename)
{
char *title = "404 Not Found";
(void)printf("\
<HTML><HEAD><TITLE>%s</TITLE></HEAD>\n\
<BODY><H2>%s</H2>\n\
The requested server-side-includes filename, %s,\n\
does not seem to exist.\n\
</BODY></HTML>\n", title, title, filename);
}
static void not_found2(char *directive, char *tag, char *filename)
{
char *title = "Not Found";
(void)printf("\
<HR><H2>%s</H2>\n\
The filename requested in a %s %s directive, %s,\n\
does not seem to exist.\n\
<HR>\n", title, directive, tag, filename);
}
static void not_permitted(char *directive, char *tag, char *val)
{
char *title = "Not Permitted";
(void)printf("\
<HR><H2>%s</H2>\n\
The filename requested in the %s %s=%s directive\n\
may not be fetched.\n\
<HR>\n", title, directive, tag, val);
}
static void unknown_directive(char *filename, char *directive)
{
char *title = "Unknown Directive";
(void)printf("\
<HR><H2>%s</H2>\n\
The requested server-side-includes filename, %s,\n\
tried to use an unknown directive, %s.\n\
<HR>\n", title, filename, directive);
}
static void unknown_tag(char *filename, char *directive, char *tag)
{
char *title = "Unknown Tag";
(void)printf("\
<HR><H2>%s</H2>\n\
The requested server-side-includes filename, %s,\n\
tried to use the directive %s with an unknown tag, %s.\n\
<HR>\n", title, filename, directive, tag);
}
static void unknown_value(char *filename, char *directive, char *tag, char *val)
{
char *title = "Unknown Value";
(void)printf("\
<HR><H2>%s</H2>\n\
The requested server-side-includes filename, %s,\n\
tried to use the directive %s %s with an unknown value, %s.\n\
<HR>\n", title, filename, directive, tag, val);
}
static int get_filename(char *vfilename, char *filename,
char *directive, char *tag, char *val, char *fn,
int fnsize)
{
char *cp;
int vl;
int fl;
/* Used for the various commands that accept a file name. These commands
* accept two tags: virtual Gives a virtual path to a document on the
* server. file Gives a pathname relative to the current directory. ../
* cannot be used in this pathname, nor can absolute paths be used.
*/
vl = strlen(vfilename);
fl = strlen(filename);
if (strcmp(tag, "virtual") == 0)
{
if (strstr(val, "../") != (char *)0)
{
not_permitted(directive, tag, val);
return -1;
}
/* Figure out root using difference between vfilename and filename. */
if (vl > fl || strcmp(vfilename, &filename[fl - vl]) != 0)
{
return -1;
}
if (fl - vl + strlen(val) >= fnsize)
{
return -1;
}
(void)strncpy(fn, filename, fl - vl);
(void)strcpy(&fn[fl - vl], val);
}
else if (strcmp(tag, "file") == 0)
{
if (val[0] == '/' || strstr(val, "../") != (char *)0)
{
not_permitted(directive, tag, val);
return -1;
}
if (fl + 1 + strlen(val) >= fnsize)
{
return -1;
}
(void)strcpy(fn, filename);
cp = strrchr(fn, '/');
if (cp == (char *)0)
{
cp = &fn[strlen(fn)];
*cp = '/';
}
(void)strcpy(++cp, val);
}
else
{
unknown_tag(filename, directive, tag);
return -1;
}
return 0;
}
static int check_filename(char *filename)
{
static int inited = 0;
static char *cgi_pattern;
#ifdef CONFIG_AUTH_FILE
struct stat sb;
char *dirname;
char *authname;
char *cp;
int fnl;
int r;
#endif
if (!inited)
{
/* Get the cgi pattern. */
cgi_pattern = getenv("CGI_PATTERN");
#ifdef CGI_PATTERN
if (cgi_pattern == (char *)0)
{
cgi_pattern = CGI_PATTERN;
}
#endif /* CGI_PATTERN */
inited = 1;
}
/* ../ is not permitted. */
if (strstr(filename, "../") !=NULL)
{
return 0;
}
/* Ensure that we are not reading a basic auth password file. */
#ifdef CONFIG_AUTH_FILE
fnl = strlen(filename);
if (strcmp(filename, CONFIG_AUTH_FILE) == 0 ||
(fnl >= sizeof(CONFIG_AUTH_FILE) &&
strcmp(&filename[fnl - sizeof(CONFIG_AUTH_FILE) + 1], CONFIG_AUTH_FILE) == 0 &&
filename[fnl - sizeof(CONFIG_AUTH_FILE)] == '/'))
{
return 0;
}
/* Check for an auth file in the same directory. We can't do an actual **
* auth password check here because CGI programs are not given the **
* authorization header, for security reasons. So instead we just **
* prohibit access to all auth-protected files.
*/
dirname = strdup(filename);
if (dirname == (char *)0)
{
/* out of memory */
return 0;
}
cp = strrchr(dirname, '/');
if (cp == (char *)0)
{
(void)strcpy(dirname, ".");
}
else
{
*cp = '\0';
}
authname = malloc(strlen(dirname) + 1 + sizeof(CONFIG_AUTH_FILE));
if (!authname)
{
/* out of memory */
free(dirname);
return 0;
}
(void)sprintf(authname, "%s/%s", dirname, CONFIG_AUTH_FILE);
r = stat(authname, &sb);
free(dirname);
free(authname);
if (r == 0)
{
return 0;
}
#endif /* CONFIG_AUTH_FILE */
/* Ensure that we are not reading a CGI file. */
if (cgi_pattern != (char *)0 && match(cgi_pattern, filename))
{
return 0;
}
return 1;
}
static void show_time(time_t t, int gmt)
{
struct tm *tmP;
if (gmt)
{
tmP = gmtime(&t);
}
else
{
tmP = localtime(&t);
}
if (strftime(g_iobuffer2, BUFFER_SIZE, g_timeformat, tmP) > 0)
{
(void)puts(g_iobuffer2);
}
}
static void show_size(off_t size)
{
switch (g_sizefmt)
{
case SF_BYTES:
(void)printf("%ld", (long)size); /* spec says should have commas */
break;
case SF_ABBREV:
if (size < 1024)
{
(void)printf("%ld", (long)size);
}
else if (size < 1024 * 1024)
{
(void)printf("%ldK", (long)size / 1024L);
}
else if (size < 1024 * 1024 * 1024)
{
(void)printf("%ldM", (long)size / (1024L * 1024L));
}
else
{
(void)printf("%ldG", (long)size / (1024L * 1024L * 1024L));
}
break;
}
}
static void do_config(FILE *instream, char *vfilename, char *filename,
char *directive, char *tag, char *val)
{
/* The config directive controls various aspects of the file parsing. **
* There are two valid tags: g_timeformat Gives the server a new format to
* use when providing dates. This is a string compatible with the
* strftime library call. g_sizefmt Determines the formatting to be used
* when displaying the size of a file. Valid choices are bytes, for a
* formatted byte count (formatted as 1,234,567), or abbrev for an
* abbreviated version displaying the number of kilobytes or megabytes the
* file occupies.
*/
if (strcmp(tag, "g_timeformat") == 0)
{
(void)strncpy(g_timeformat, val, TIMEFMT_SIZE - 1);
g_timeformat[TIMEFMT_SIZE - 1] = '\0';
}
else if (strcmp(tag, "g_sizefmt") == 0)
{
if (strcmp(val, "bytes") == 0)
{
g_sizefmt = SF_BYTES;
}
else if (strcmp(val, "abbrev") == 0)
{
g_sizefmt = SF_ABBREV;
}
else
{
unknown_value(filename, directive, tag, val);
}
}
else
{
unknown_tag(filename, directive, tag);
}
}
static void do_include(FILE *instream, char *vfilename, char *filename,
char *directive, char *tag, char *val)
{
FILE *instream2;
int ret;
/* Inserts the text of another document into the parsed document. */
ret = get_filename(vfilename, filename, directive, tag, val, g_iobuffer1, BUFFER_SIZE);
if (ret < 0)
{
return;
}
if (!check_filename(g_iobuffer1))
{
not_permitted(directive, tag, g_iobuffer1);
return;
}
instream2 = fopen(g_iobuffer1, "r");
if (instream2 == (FILE *) 0)
{
not_found2(directive, tag, g_iobuffer1);
return;
}
if (strcmp(tag, "virtual") == 0)
{
if (strlen(val) <BUFFER_SIZE)
{
(void)strcpy(g_iobuffer2, val);
}
else
{
(void)strcpy(g_iobuffer2, g_iobuffer1); /* same size, has to fit */
}
}
else
{
if (strlen(vfilename) + 1 + strlen(val) < BUFFER_SIZE)
{
char *cp;
(void)strcpy(g_iobuffer2, vfilename);
cp = strrchr(g_iobuffer2, '/');
if (cp == (char *)0)
{
cp = &g_iobuffer2[strlen(g_iobuffer2)];
*cp = '/';
}
(void)strcpy(++cp, val);
}
else
{
(void)strcpy(g_iobuffer2, g_iobuffer1); /* same size, has to fit */
}
}
read_file(instream2, g_iobuffer2, g_iobuffer1);
(void)fclose(instream2);
}
static void do_echo(FILE *instream, char *vfilename, char *filename,
char *directive, char *tag, char *val)
{
char *cp;
/* Prints the value of one of the include variables. Any dates are
* printed subject to the currently configured g_timeformat. The only valid
* tag is var, whose value is the name of the variable you wish to echo.
*/
if (strcmp(tag, "var") != 0)
{
unknown_tag(filename, directive, tag);
}
else
{
if (strcmp(val, "DOCUMENT_NAME") == 0)
{
/* The current filename. */
(void)puts(filename);
}
else if (strcmp(val, "DOCUMENT_URI") == 0)
{
/* The virtual path to this file (such as /~robm/foo.shtml). */
(void)puts(vfilename);
}
else if (strcmp(val, "QUERY_STRING_UNESCAPED") == 0)
{
/* The unescaped version of any search query the client sent. */
cp = getenv("QUERY_STRING");
if (cp != (char *)0)
{
(void)puts(cp);
}
}
else if (strcmp(val, "DATE_LOCAL") == 0)
{
struct timeval tm;
/* The current date, local time zone. */
gettimeofday(&tm, NULL);
show_time(tm.tv_sec, 0);
}
else if (strcmp(val, "DATE_GMT") == 0)
{
struct timeval tm;
/* Same as DATE_LOCAL but in Greenwich mean time. */
gettimeofday(&tm, NULL);
show_time(tm.tv_sec, 1);
}
#if 0 /* fstat is not yet supported */
else if (strcmp(val, "LAST_MODIFIED") == 0)
{
/* The last modification date of the current document. */
if (fstat(fileno(instream), &g_sb) >= 0)
{
show_time(g_sb.st_mtime, 0);
}
}
#endif
else
{
/* Try an environment variable. */
cp = getenv(val);
if (cp == (char *)0)
{
unknown_value(filename, directive, tag, val);
}
else
{
(void)puts(cp);
}
}
}
}
static void do_fsize(FILE *instream, char *vfilename, char *filename,
char *directive, char *tag, char *val)
{
int ret;
/* Prints the size of the specified file. */
ret = get_filename(vfilename, filename, directive, tag, val, g_iobuffer1, BUFFER_SIZE);
if (ret < 0)
{
return;
}
if (stat(g_iobuffer1, &g_sb) < 0)
{
not_found2(directive, tag, g_iobuffer1);
return;
}
show_size(g_sb.st_size);
}
static void do_flastmod(FILE *instream, char *vfilename, char *filename,
char *directive, char *tag, char *val)
{
int ret;
/* Prints the last modification date of the specified file. */
ret = get_filename(vfilename, filename, directive, tag, val, g_iobuffer1, BUFFER_SIZE);
if (ret < 0)
{
return;
}
if (stat(g_iobuffer1, &g_sb) < 0)
{
not_found2(directive, tag, g_iobuffer1);
return;
}
show_time(g_sb.st_mtime, 0);
}
static void parse(FILE *instream, char *vfilename, char *filename, char *str)
{
char *directive;
char *cp;
int ntags;
int dirn;
int i;
char *val;
directive = str;
directive += strspn(directive, " \t\n\r");
ntags = 0;
cp = directive;
for (;;)
{
cp = strpbrk(cp, " \t\n\r\"");
if (cp == (char *)0)
{
break;
}
if (*cp == '"')
{
cp = strpbrk(cp + 1, "\"");
cp++;
if (*cp == '\0')
{
break;
}
}
*cp++ = '\0';
cp += strspn(cp, " \t\n\r");
if (*cp == '\0')
{
break;
}
if (ntags < MAX_TAGS)
{
g_tags[ntags++] = cp;
}
}
if (strcmp(directive, "config") == 0)
{
dirn = DI_CONFIG;
}
else if (strcmp(directive, "include") == 0)
{
dirn = DI_INCLUDE;
}
else if (strcmp(directive, "echo") == 0)
{
dirn = DI_ECHO;
}
else if (strcmp(directive, "fsize") == 0)
{
dirn = DI_FSIZE;
}
else if (strcmp(directive, "flastmod") == 0)
{
dirn = DI_FLASTMOD;
}
else
{
unknown_directive(filename, directive);
return;
}
for (i = 0; i < ntags; ++i)
{
if (i > 0)
{
putchar(' ');
}
val = strchr(g_tags[i], '=');
if (val == (char *)0)
{
val = "";
}
else
{
*val++ = '\0';
}
if (*val == '"' && val[strlen(val) - 1] == '"')
{
val[strlen(val) - 1] = '\0';
++val;
}
switch (dirn)
{
case DI_CONFIG:
do_config(instream, vfilename, filename, directive, g_tags[i], val);
break;
case DI_INCLUDE:
do_include(instream, vfilename, filename, directive, g_tags[i], val);
break;
case DI_ECHO:
do_echo(instream, vfilename, filename, directive, g_tags[i], val);
break;
case DI_FSIZE:
do_fsize(instream, vfilename, filename, directive, g_tags[i], val);
break;
case DI_FLASTMOD:
do_flastmod(instream, vfilename, filename, directive, g_tags[i], val);
break;
}
}
}
static void slurp(FILE *instream, char *vfilename, char *filename)
{
int state;
int ich;
int i;
/* Now slurp in the rest of the comment from the input file. */
i = 0;
state = ST_GROUND;
while ((ich = getc(instream)) != EOF)
{
switch (state)
{
case ST_GROUND:
if (ich == '-')
{
state = ST_MINUS1;
}
break;
case ST_MINUS1:
if (ich == '-')
{
state = ST_MINUS2;
}
else
{
state = ST_GROUND;
}
break;
case ST_MINUS2:
if (ich == '>')
{
g_iobuffer1[i - 2] = '\0';
parse(instream, vfilename, filename, g_iobuffer1);
return;
}
else if (ich != '-')
{
state = ST_GROUND;
}
break;
}
if (i < BUFFER_SIZE - 1)
{
g_iobuffer1[i++] = (char)ich;
}
}
}
static void read_file(FILE *instream, char *vfilename, char *filename)
{
int ich;
int state;
/* Copy it to output, while running a state-machine to look for SSI
* directives.
*/
state = ST_GROUND;
while ((ich = getc(instream)) != EOF)
{
switch (state)
{
case ST_GROUND:
if (ich == '<')
{
state = ST_LESSTHAN;
continue;
}
break;
case ST_LESSTHAN:
if (ich == '!')
{
state = ST_BANG;
continue;
}
else
{
state = ST_GROUND;
putchar('<');
}
break;
case ST_BANG:
if (ich == '-')
{
state = ST_MINUS1;
continue;
}
else
{
state = ST_GROUND;
(void)puts("<!");
}
break;
case ST_MINUS1:
if (ich == '-')
{
state = ST_MINUS2;
continue;
}
else
{
state = ST_GROUND;
(void)puts("<!-");
}
break;
case ST_MINUS2:
if (ich == '#')
{
slurp(instream, vfilename, filename);
state = ST_GROUND;
continue;
}
else
{
state = ST_GROUND;
(void)puts("<!--");
}
break;
}
putchar((char)ich);
}
}
/****************************************************************************
* Public Functions
****************************************************************************/
#ifdef CONFIG_THTTPD_BINFS
int ssi_main(int argc, char *argv[])
#else
int main(int argc, char *argv[])
#endif
{
FILE *instream;
char *script_name;
char *path_info;
char *path_translated;
int errcode = 0;
/* Default formats. */
(void)strcpy(g_timeformat, "%a %b %e %T %Z %Y");
g_sizefmt = SF_BYTES;
/* The MIME type has to be text/html. */
(void)puts("Content-type: text/html\n\n");
/* Get the name that we were run as. */
script_name = getenv("SCRIPT_NAME");
if (!script_name)
{
internal_error("Couldn't get SCRIPT_NAME environment variable.");
return 1;
}
/* Append the PATH_INFO, if any, to get the full URL. */
path_info = getenv("PATH_INFO");
if (!path_info)
{
path_info = "";
}
g_url = (char*)malloc(strlen(script_name) + strlen(path_info) + 1);
if (!g_url)
{
internal_error("Out of memory.");
return 2;
}
(void)sprintf(g_url, "%s%s", script_name, path_info);
/* Get the name of the file to parse. */
path_translated = getenv("PATH_TRANSLATED");
if (!path_translated)
{
internal_error("Couldn't get PATH_TRANSLATED environment variable.");
errcode = 3;
goto errout_with_g_url;
}
if (!check_filename(path_translated))
{
not_permitted("initial", "PATH_TRANSLATED", path_translated);
errcode = 4;
goto errout_with_g_url;
}
/* Open it. */
instream = fopen(path_translated, "r");
if (!instream)
{
not_found(path_translated);
errcode = 5;
goto errout_with_g_url;
}
/* Read and handle the file. */
read_file(instream, path_info, path_translated);
(void)fclose(instream);
errout_with_g_url:
free(g_url);
return errcode;
}