proxy_util.c revision 984519938db95cd681c159102a852e523da1d07e
/* ====================================================================
* The Apache Software License, Version 1.1
*
* Copyright (c) 2000 The Apache Software Foundation. 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
* distribution.
*
* 3. The end-user documentation included with the redistribution,
* if any, must include the following acknowledgment:
* "This product includes software developed by the
* Apache Software Foundation (http://www.apache.org/)."
* Alternately, this acknowledgment may appear in the software itself,
* if and wherever such third-party acknowledgments normally appear.
*
* 4. The names "Apache" and "Apache Software Foundation" must
* not be used to endorse or promote products derived from this
* software without prior written permission. For written
* permission, please contact apache@apache.org.
*
* 5. Products derived from this software may not be called "Apache",
* nor may "Apache" appear in their name, without prior written
* permission of the Apache Software Foundation.
*
* THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED 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 APACHE SOFTWARE FOUNDATION OR
* ITS 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.
* ====================================================================
*
* This software consists of voluntary contributions made by many
* individuals on behalf of the Apache Software Foundation. For more
* information on the Apache Software Foundation, please see
*
* Portions of this software are based upon public domain software
* originally written at the National Center for Supercomputing Applications,
* University of Illinois, Urbana-Champaign.
*/
/* Utility routines for Apache proxy */
#include "mod_proxy.h"
#include "http_main.h"
#include "http_log.h"
#include "util_uri.h"
#include "util_date.h" /* get ap_checkmask() decl. */
#include "apr_md5.h"
static struct per_thread_data *get_per_thread_data(void);
/* already called in the knowledge that the characters are hex digits */
int ap_proxy_hex2c(const char *x)
{
int i, ch;
#ifndef CHARSET_EBCDIC
ch = x[0];
if (apr_isdigit(ch))
i = ch - '0';
else if (apr_isupper(ch))
else
i <<= 4;
ch = x[1];
if (apr_isdigit(ch))
i += ch - '0';
else if (apr_isupper(ch))
else
return i;
#else /*CHARSET_EBCDIC*/
#endif /*CHARSET_EBCDIC*/
}
void ap_proxy_c2hex(int ch, char *x)
{
#ifndef CHARSET_EBCDIC
int i;
x[0] = '%';
if (i >= 10)
x[1] = ('A' - 10) + i;
else
x[1] = '0' + i;
i = ch & 0x0F;
if (i >= 10)
x[2] = ('A' - 10) + i;
else
x[2] = '0' + i;
#else /*CHARSET_EBCDIC*/
static const char ntoa[] = { "0123456789ABCDEF" };
ch &= 0xFF;
x[0] = '%';
x[3] = '\0';
#endif /*CHARSET_EBCDIC*/
}
/*
* canonicalise a URL-encoded string
*/
/*
* Convert a URL-encoded string to canonical form.
* It decodes characters which need not be encoded,
* and encodes those which must be encoded, and does not touch
* those which must not be touched.
*/
int isenc)
{
int i, j, ch;
char *y;
char *allowed; /* characters which should not be encoded */
/* N.B. in addition to :@&=, this allows ';' in an http path
* and '?' in an ftp path -- this may be revised
*
* Also, it makes a '+' character in a search string reserved, as
* it may be form-encoded. (Although RFC 1738 doesn't allow this -
* it only permits ; / ? : @ = & as reserved chars.)
*/
if (t == enc_path)
allowed = "$-_.+!*'(),;:@&=";
else if (t == enc_search)
allowed = "$-_.!*'(),;:@&=";
else if (t == enc_user)
allowed = "$-_.+!*'(),;@&=";
else if (t == enc_fpath)
allowed = "$-_.+!*'(),?:@&=";
else /* if (t == enc_parm) */
allowed = "$-_.+!*'(),?/:@&=";
if (t == enc_path)
reserved = "/";
else if (t == enc_search)
reserved = "+";
else
reserved = "";
for (i = 0, j = 0; i < len; i++, j++) {
/* always handle '/' first */
ch = x[i];
y[j] = ch;
continue;
}
/* decode it if not already done */
return NULL;
i += 2;
ap_proxy_c2hex(ch, &y[j]);
j += 2;
continue;
}
}
/* recode it, if necessary */
ap_proxy_c2hex(ch, &y[j]);
j += 2;
}
else
y[j] = ch;
}
y[j] = '\0';
return y;
}
/*
* Parses network-location.
* urlp on input the URL; on output the path, after the leading /
* password holder for password
* host holder for host
* port port number; only set if one is supplied.
*
* Returns an error string.
*/
char *
{
int i;
return "Malformed URL";
url = "";
else
*strp = '\0';
/* find password */
*strp = '\0';
return "Bad %-escape in URL (password)";
}
return "Bad %-escape in URL (username)";
}
}
}
*(strp++) = '\0';
for (i = 0; strp[i] != '\0'; i++)
if (!apr_isdigit(strp[i]))
break;
/* if (i == 0) the no port was given; keep default */
if (strp[i] != '\0') {
return "Bad port number in URL";
} else if (i > 0) {
if (*port > 65535)
return "Port number in URL > 65535";
}
}
if (*host == '\0')
return "Missing host in URL";
/* check hostname syntax */
for (i = 0; host[i] != '\0'; i++)
break;
/* must be an IP address */
#else
#endif
{
return "Bad IP address in URL";
}
/* if (strchr(host,'.') == NULL && domain != NULL)
host = pstrcat(p, host, domain, NULL);
*/
return NULL;
}
static const char * const lwday[7] =
{"Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday"};
/*
* If the date is a valid RFC 850 date or asctime() date, then it
* is converted to the RFC 1123 format, otherwise it is not modified.
* This routine is not very fast at doing conversions, as it uses
* sscanf and sprintf. However, if the date is already correctly
* formatted, then it exits very quickly.
*/
const char *
{
char *x = apr_pstrdup(p, x1);
q = strchr(x, ',');
/* check for RFC 850 date */
*q = '\0';
break;
*q = ',';
if (wk == 7)
return x; /* not a valid date */
if (q[4] != '-' || q[8] != '-' || q[11] != ' ' || q[14] != ':' ||
return x;
return x;
if (year < 70)
year += 2000;
else
year += 1900;
}
else {
/* check for acstime() date */
if (x[3] != ' ' || x[7] != ' ' || x[10] != ' ' || x[13] != ':' ||
x[16] != ':' || x[19] != ' ' || x[24] != '\0')
return x;
return x;
break;
if (wk == 7)
return x;
}
/* check date */
break;
if (mon == 12)
return x;
q = apr_palloc(p, 30);
return q;
}
/* NOTE: This routine is taken from http_protocol::getline()
* because the old code found in the proxy module was too
* difficult to understand and maintain.
*/
/* Get a line of protocol input, including any continuation lines
* caused by MIME folding (or broken clients) if fold != 0, and place it
* in the buffer s, of size n bytes, without the ending newline.
*
* Returns -1 on error, or the length of s.
*
* Note: Because bgets uses 1 char for newline and 1 char for NUL,
* the most we can get is (n - 2) actual characters if it
* was ended by a newline, or (n - 1) characters if the line
* length exceeded (n - 1). So, if the result == (n - 1),
* then the actual input line exceeded the buffer length,
* and it would be a good idea for the caller to puke 400 or 414.
*/
{
int retval;
int total = 0;
pos = s;
do {
if (retval <= 0)
/* retval is the number of characters read, not including NUL */
n -= retval; /* Keep track of how much of s is full */
*pos = '\0';
--total;
++n;
}
else
return total; /* if not, input line exceeded buffer size */
/* Continue appending if line folding is desired and
* the last line was not empty and we have room in the buffer and
* the next line begins with a continuation character.
*/
return total;
}
/*
* Reads headers from a buffer and returns an array of headers.
* Returns NULL on file error
* This routine tries to deal with too long lines and continuation lines.
* @@@: XXX: FIXME: currently the headers are passed thru un-merged.
* Is that okay, or should they be collapsed where possible?
*/
{
int len;
char field[MAX_STRING_LEN];
/*
* Read header lines until we get the empty separator line, a read error,
* the connection closes (EOF), or we timeout.
*/
/* Buggy MS IIS servers sometimes return invalid headers
* (an extra "HTTP/1.0 200, OK" line sprinkled in between
* the usual MIME headers). Try to deal with it in a sensible
* way, but log the fact.
* XXX: The mask check is buggy if we ever see an HTTP/1.10 */
/* Nope, it wasn't even an extra HTTP header. Give up. */
return NULL;
}
"proxy: Ignoring duplicate HTTP header "
continue;
}
*value = '\0';
++value;
/* XXX: RFC2068 defines only SP and HT as whitespace, this test is
* wrong... and so are many others probably.
*/
while (apr_isspace(*value))
++value; /* Skip to start of value */
/* should strip trailing whitespace as well */
*end = '\0';
/* the header was too long; at the least we should skip extra data */
>= MAX_STRING_LEN - 1) {
/* soak up the extra data */
}
if (len == 0) /* time to exit the larger loop as well */
break;
}
}
return resp_hdrs;
}
{
int ok;
long total_bytes_rcvd, in_buffer;
register int n, o;
int written = 0, wrote_to_cache;
total_bytes_rcvd = 0;
if (c) ap_cache_el_data(c, &cachefp);
#if 0
#ifdef CHARSET_EBCDIC
#endif
#endif
/* Since we are reading from one buffer and writing to another,
* it is unsafe to do a soft_timeout here, at least until the proxy
* has its own timeout handler which can set both buffers to EOUT.
*/
/* works fine under win32, so leave it */
alternate_timeouts = 0;
#else
/* CHECKME! Since hard_timeout won't work in unix on sends with partial
* cache completion, we have to alternate between hard_timeout
* for reads, and soft_timeout for send. This is because we need
* to get a return from ap_bwrite to be able to continue caching.
* BUT, if we *can't* continue anyway, just use hard_timeout.
* (Also, if no cache file is written, use hard timeouts)
*/
alternate_timeouts = 0;
}
#endif
/* Loop and ap_bread() while we can successfully read and write,
* or (after the client aborted) while we can successfully
* read and finish the configured cache_completion.
*/
/* Read block from server */
{
if (c != NULL) {
"proxy: error reading from %s", c->name);
ap_proxy_cache_error(&c);
}
break;
}
else if(cntr == 0) break;
/* Write to cache first. */
/*@@@ XXX FIXME: Assuming that writing the cache file won't time out?!!? */
"proxy: error writing to cache");
ap_proxy_cache_error(&c);
} else {
written += n;
}
o = 0;
total_bytes_rcvd += cntr;
/* Write the block to the client, detect aborted transfers */
if (completion) {
/* when a send failure occurs, we need to decide
* whether to continue loading and caching the
* document, or to abort the whole thing
*/
(completion->cache_completion > 0) &&
if (!ok)
ap_proxy_cache_error(&c);
}
break;
}
o += cntr;
} /* while client alive and more data to send */
} /* loop and ap_bread while "ok" */
ap_rflush(r);
return total_bytes_rcvd;
}
/*
* Sends response line and headers. Uses the client fd and the
* headers_out array from the passed request_rec to talk to the client
* and to properly set the headers it sends for things such as logging.
*
* A timeout should be set before calling this routine.
*/
{
int i;
for (i = 0; i < ap_table_elts(t)->nelts; ++i) {
}
}
len = 2;
}
/*
* list is a comma-separated list of case-insensitive tokens, with
* optional whitespace around the tokens.
* The return returns 1 if the token val is found in the list, or 0
* otherwise.
*/
{
int len, i;
const char *p;
if (p != NULL) {
i = p - list;
do
p++;
while (apr_isspace(*p));
}
else
i--;
return 1;
list = p;
}
return 0;
}
/*
* Converts 8 hex digits to a time integer
*/
int ap_proxy_hex2sec(const char *x)
{
int i, ch;
unsigned int j;
for (i = 0, j = 0; i < 8; i++) {
ch = x[i];
j <<= 4;
if (apr_isdigit(ch))
j |= ch - '0';
else if (apr_isupper(ch))
else
}
if (j == 0xffffffff)
return -1; /* so that it works with 8-byte ints */
else
return j;
}
/*
* Converts a time integer to 8 hex digits
*/
void ap_proxy_sec2hex(int t, char *y)
{
int i, ch;
unsigned int j = t;
for (i = 7; i >= 0; i--) {
ch = j & 0xF;
j >>= 4;
if (ch >= 10)
else
y[i] = ch + '0';
}
y[8] = '\0';
}
void ap_proxy_cache_error(ap_cache_el **c)
{
if (c && *c) {
ap_cache_el_finalize((*c));
*c = NULL;
}
}
{
apr_pstrcat(r->pool,
"The proxy server could not handle the request "
" ",
"Reason: <STRONG>",
"</STRONG>", NULL));
/* Allow "error-notes" string to be printed by ap_send_error_response() */
return statuscode;
}
/*
* This routine returns its own error message
*/
{
int i;
for (i = 0; host[i] != '\0'; i++)
break;
if (host[i] != '\0') {
return "Host not found";
}
else {
}
}
return NULL;
}
static const char *
{
int port = -1;
return r->hostname;
/* Set url to the first char after "scheme://" */
return NULL;
url = apr_pstrdup(r->pool, &url[1]); /* make it point to "//", which is what proxy_canon_netloc expects */
"%s", err);
return host; /* ought to return the port, too */
}
/* Return TRUE if addr represents an IP address (or an IP network address) */
{
long ip_addr[4];
int i, quads;
long bits;
/* if the address is given with an explicit netmask, use that */
/* Due to a deficiency in ap_inet_addr(), it is impossible to parse */
/* "partial" addresses (with less than 4 quads) correctly, i.e. */
/* 192.168.123 is parsed as 192.168.0.123, which is not what I want. */
/* I therefore have to parse the IP address manually: */
/* addr and mask were set by proxy_readmask() */
/*return 1; */
/* Parse IP addr manually, optionally allowing */
/* abbreviated net addresses like 192.168. */
/* Iterate over up to 4 (dotted) quads. */
char *tmp;
break;
if (!apr_isdigit(*addr))
return 0; /* no digit at start of quad */
return 0;
/* invalid octet */
return 0;
}
++addr; /* after the 4th quad, a dot would be illegal */
}
char *tmp;
++addr;
return 0;
return 0;
}
else {
/* Determine (i.e., "guess") netmask by counting the */
/* number of trailing .0's; reduce #quads appropriately */
/* (so that 192.168.0.0 is equivalent to 192.168.) */
--quads;
/* "IP Address should be given in dotted-quad form, optionally followed by a netmask (e.g., 192.168.111.0/24)"; */
if (quads < 1)
return 0;
/* every zero-byte counts as 8 zero-bits */
"Warning: NetMask not supplied with IP-Addr; guessing: %s/%ld\n",
}
"Warning: NetMask and IP-Addr disagree in %s/%ld\n",
" Set to %s/%ld\n",
}
if (*addr == '\0') {
return 1;
}
else
}
/* Return TRUE if addr represents an IP address (or an IP network address) */
{
int i;
int ip_addr[4];
char **ip_listptr;
const char *found;
const char *host = proxy_get_host_of_request(r);
return 0;
#if DEBUGGING
#endif
return 1;
}
#if DEBUGGING
else {
}
#endif
}
else {
#if DEBUGGING
#endif
return 0;
}
else
/* Try to deal with multiple IP addr's for a host */
#if DEBUGGING
#endif
return 1;
}
#if DEBUGGING
else {
}
#endif
}
}
return 0;
}
/* Return TRUE if addr represents a domain name */
{
int i;
/* Domain name must start with a '.' */
if (addr[0] != '.')
return 0;
/* rfc1035 says DNS names must consist of "[-a-zA-Z0-9]" and '.' */
continue;
#if 0
if (addr[i] == ':') {
"@@@@ handle optional port in proxy_is_domainname()");
/* @@@@ handle optional port */
}
#endif
if (addr[i] != '\0')
return 0;
/* Strip trailing dots */
addr[i] = '\0';
return 1;
}
/* Return TRUE if host "host" is in domain "domain" */
{
const char *host = proxy_get_host_of_request(r);
return 0;
/* @@@ do this within the setup? */
/* Ignore trailing dots in domain comparison: */
--d_len;
--h_len;
}
/* Return TRUE if addr represents a host name */
{
int i;
/* Host names must not start with a '.' */
if (addr[0] == '.')
return 0;
/* rfc1035 says DNS names must consist of "[-a-zA-Z0-9]" and '.' */
#if 0
if (addr[i] == ':') {
"@@@@ handle optional port in proxy_is_hostname()");
/* @@@@ handle optional port */
}
#endif
return 0;
/* Strip trailing dots */
addr[i] = '\0';
return 1;
}
/* Return TRUE if host "host" is equal to host2 "host2" */
{
const char *host2 = proxy_get_host_of_request(r);
int h2_len;
int h1_len;
return 0; /* oops! */
#if 0
unsigned long *ip_list;
/* Try to deal with multiple IP addr's for a host */
if (*ip_list == ? ? ? ? ? ? ? ? ? ? ? ? ?)
return 1;
#endif
/* Ignore trailing dots in host2 comparison: */
--h2_len;
--h1_len;
}
/* Return TRUE if addr is to be matched as a word */
{
return 1;
}
/* Return TRUE if string "str2" occurs literally in "str1" */
{
const char *host = proxy_get_host_of_request(r);
}
{
int i;
for (i = 0; host[i] != '\0'; i++)
break;
if (host[i] == '\0') {
}
do
{
} while (APR_STATUS_IS_EINTR(rv));
if (rv != APR_SUCCESS)
{
}
return rv;
}
/* This function is called by ap_table_do() for all header lines */
/* (from proxy_http.c and proxy_ftp.c) */
/* It is passed a table_do_args struct pointer and a MIME field and value pair */
{
struct request_rec *r = (struct request_rec *)p;
return 1;
if (!r->assbackwards)
return 1; /* tell ap_table_do() to continue calling us for more headers */
}
/* send a text line to one or two BUFF's; return line length */
{
return len;
}
#if defined WIN32
{
switch (reason) {
case DLL_PROCESS_ATTACH:
case DLL_THREAD_ATTACH: /* intentional no break */
break;
case DLL_THREAD_DETACH:
if (memptr) {
TlsSetValue (tls_index, 0);
}
break;
}
return TRUE;
}
#endif
static struct per_thread_data *get_per_thread_data(void)
{
#if 0
#if defined(WIN32)
#else
return &sptd;
#endif
#endif
return NULL;
}
/* This function is completely bogus. This should become a part of the
* cache filter when it is finished. RBB
*/
{
char buffer[500];
"Sending cache file for %s", c->name);
return HTTP_INTERNAL_SERVER_ERROR;
/* send the response */
}
/* send headers */
len = 2;
/* send data */
return HTTP_INTERNAL_SERVER_ERROR;
return OK;
}
{
else
if((r->status != HTTP_OK && r->status != HTTP_MOVED_PERMANENTLY && r->status != HTTP_NOT_MODIFIED) ||
(r->status == HTTP_NOT_MODIFIED) ||
r->header_only ||
{
"proxy: Response is not cacheable: %s", r->unparsed_uri);
return 0;
}
return 1;
}
/*
* what responses should we not cache?
* Unknown status responses and those known to be uncacheable
* 304 HTTP_NOT_MODIFIED response when we have no valid cache file, or
* 200 HTTP_OK response from HTTP/1.0 and up without a Last-Modified header, or
* HEAD requests, or
* requests with an Authorization header, or
*/
/* @@@ XXX FIXME: is the test "r->status != HTTP_MOVED_PERMANENTLY" correct?
* or shouldn't it be "ap_is_HTTP_REDIRECT(r->status)" ? -MnKr */
int ap_proxy_cache_update(ap_cache_el *c)
{
if(!h) return DECLINED;
"proxy: Cache finalized: %s", c->name);
return DECLINED;
}