allow whitespace in request URI's.

diff --git a/src/cache_cf.cc b/src/cache_cf.cc
index 0d858697a9968e1e1fd045ce412d21b49262c316..1c8a6cfbb939ce68870b50b8b4f2ee3c302a70b5 100644 (file)
--- a/src/cache_cf.cc
+++ b/src/cache_cf.cc
@@ -1,6 +1,6 @@
 
 /*
- * $Id: cache_cf.cc,v 1.307 1998/10/13 23:33:31 wessels Exp $
+ * $Id: cache_cf.cc,v 1.308 1998/10/14 21:11:55 wessels Exp $
  *
  * DEBUG: section 3     Configuration File Parsing
  * AUTHOR: Harvest Derived
@@ -1407,6 +1407,42 @@ parse_stringlist(wordlist ** list)
 
 #define free_wordlist wordlistDestroy
 
+#define free_uri_whitespace free_int
+
+static void
+parse_uri_whitespace(int *var)
+{
+    char *token = strtok(NULL, w_space);
+    if (token == NULL)
+       self_destruct();
+    if (!strcasecmp(token, "deny"))
+       *var = URI_WHITESPACE_DENY;
+    else if (!strcasecmp(token, "allow"))
+       *var = URI_WHITESPACE_ALLOW;
+    else if (!strcasecmp(token, "encode"))
+       *var = URI_WHITESPACE_ENCODE;
+    else if (!strcasecmp(token, "chop"))
+       *var = URI_WHITESPACE_CHOP;
+    else
+       self_destruct();
+}
+
+
+static void
+dump_uri_whitespace(StoreEntry * entry, const char *name, int var)
+{
+    char *s;
+    if (var == URI_WHITESPACE_ALLOW)
+       s = "allow";
+    else if (var == URI_WHITESPACE_ENCODE)
+       s = "encode";
+    else if (var == URI_WHITESPACE_CHOP)
+       s = "chop";
+    else
+       s = "deny";
+    storeAppendPrintf(entry, "%s %s\n", name, s);
+}
+
 #include "cf_parser.c"
 
 peer_t

diff --git a/src/cf.data.pre b/src/cf.data.pre
index 5787214759e2495ae23429d988b2013ec8b64696..07b6b8af0fcbcabe9725dfdb44b7886cc691cf63 100644 (file)
--- a/src/cf.data.pre
+++ b/src/cf.data.pre
@@ -1,6 +1,6 @@
 
 #
-# $Id: cf.data.pre,v 1.121 1998/10/13 23:42:03 wessels Exp $
+# $Id: cf.data.pre,v 1.122 1998/10/14 21:11:57 wessels Exp $
 #
 #
 # SQUID Internet Object Cache  http://squid.nlanr.net/Squid/
@@ -2657,4 +2657,28 @@ DOC_START
        objects.
 DOC_END
 
+NAME: uri_whitespace
+TYPE: uri_whitespace
+LOC: Config.uri_whitespace
+DEFAULT: deny
+DOC_START
+       What to do with requests that have whitespace characters in the
+       URI.  Options:
+
+       deny:   The request is denied.  The user receives an "Invalid
+               Request" message.
+       allow:  The request is allowed and the URI is not changed.  The
+               whitespace characters remain in the URI.  Note the
+               whitespace is passed to redirector processes if they
+               are in use.
+       encode: The request is allowed and the whitespace characters are
+               encoded according to RFC1738.  This could be considered
+               a violation of the HTTP/1.1
+               RFC because proxies are not allowed to rewrite URI's.
+       chop:   The request is allowed and the URI is chopped at the
+               first whitespace.  This might also be considered a
+               violation.
+uri_whitespace deny
+DOC_END
+
 EOF

diff --git a/src/client_side.cc b/src/client_side.cc
index 373bc89b8f91206c5c210a320d0d878c0f77ed33..c6d149e9f923465d6687c9c4f0d65ac607ab834d 100644 (file)
--- a/src/client_side.cc
+++ b/src/client_side.cc
@@ -1,6 +1,6 @@
 
 /*
- * $Id: client_side.cc,v 1.411 1998/10/13 23:39:08 wessels Exp $
+ * $Id: client_side.cc,v 1.412 1998/10/14 21:11:58 wessels Exp $
  *
  * DEBUG: section 33    Client-side Routines
  * AUTHOR: Duane Wessels
@@ -1882,16 +1882,25 @@ parseHttpRequest(ConnStateData * conn, method_t * method_p, int *status,
     debug(33, 5) ("parseHttpRequest: Method is '%s'\n", mstr);
     *method_p = method;
 
-    /* look for URL */
-    if ((url = strtok(NULL, "\r\n\t ")) == NULL) {
+    /* look for URL+HTTP/x.x */
+    if ((url = strtok(NULL, "\r\n")) == NULL) {
        debug(33, 1) ("parseHttpRequest: Missing URL\n");
        return parseHttpRequestAbort(conn, "error:missing-url");
     }
-    debug(33, 5) ("parseHttpRequest: Request is '%s'\n", url);
-
-    token = strtok(NULL, null_string);
-    for (t = token; t && *t && *t != '\n' && *t != '\r'; t++);
-    if (t == NULL || *t == '\0' || t == token || strncasecmp(token, "HTTP/", 5)) {
+    t = url + strlen(url);
+    assert(*t == '\0');
+    token = NULL;
+    while (t > url) {
+       t--;
+       if (isspace(*t) && !strncmp(t + 1, "HTTP/", 5)) {
+           token = t + 1;
+           break;
+       }
+    }
+    while (t > url && isspace(*t))
+       *(t--) = '\0';
+    debug(33, 5) ("parseHttpRequest: URI is '%s'\n", url);
+    if (token == NULL) {
        debug(33, 3) ("parseHttpRequest: Missing HTTP identifier\n");
 #if RELAXED_HTTP_PARSER
        http_ver = (float) 0.9; /* wild guess */
@@ -2030,7 +2039,10 @@ parseHttpRequest(ConnStateData * conn, method_t * method_p, int *status,
        strcpy(http->uri, url);
        http->flags.accel = 0;
     }
-    http->log_uri = xstrdup(http->uri);
+    if (!stringHasWhitespace(http->uri))
+       http->log_uri = xstrdup(http->uri);
+    else
+       http->log_uri = xstrdup(rfc1738_escape(http->uri));
     debug(33, 5) ("parseHttpRequest: Complete request received\n");
     if (free_request)
        safe_free(url);

diff --git a/src/defines.h b/src/defines.h
index 6ac6b86bbb54dcd275bc13c6bbd7f963596e4bb9..b2e3813455f2567f0444271eea98d3d9ecedbc8e 100644 (file)
--- a/src/defines.h
+++ b/src/defines.h
@@ -1,6 +1,6 @@
 
 /*
- * $Id: defines.h,v 1.66 1998/09/14 22:18:45 wessels Exp $
+ * $Id: defines.h,v 1.67 1998/10/14 21:12:00 wessels Exp $
  *
  *
  * SQUID Internet Object Cache  http://squid.nlanr.net/Squid/
@@ -258,3 +258,8 @@
 #define PEER_TCP_MAGIC_COUNT 10
 
 #define CLIENT_SOCK_SZ 4096
+
+#define URI_WHITESPACE_DENY 0
+#define URI_WHITESPACE_ALLOW 1
+#define URI_WHITESPACE_ENCODE 2
+#define URI_WHITESPACE_CHOP 3

diff --git a/src/protos.h b/src/protos.h
index d4aadad899978eac51a4b474a3418d56bcfbb09e..66a5d4e7d41c701b41b7e41676ddb2046e14a14d 100644 (file)
--- a/src/protos.h
+++ b/src/protos.h
@@ -1,6 +1,6 @@
 
 /*
- * $Id: protos.h,v 1.277 1998/10/12 21:41:01 wessels Exp $
+ * $Id: protos.h,v 1.278 1998/10/14 21:12:01 wessels Exp $
  *
  *
  * SQUID Internet Object Cache  http://squid.nlanr.net/Squid/
@@ -992,6 +992,8 @@ extern void pconnInit(void);
 extern int asnMatchIp(void *, struct in_addr);
 extern void asnInit(void);
 extern void asnFreeMemory(void);
+
+/* tools.c */
 extern void dlinkAdd(void *data, dlink_node *, dlink_list *);
 extern void dlinkAddTail(void *data, dlink_node *, dlink_list *);
 extern void dlinkDelete(dlink_node * m, dlink_list * list);
@@ -999,6 +1001,7 @@ extern void kb_incr(kb_t *, size_t);
 extern double gb_to_double(const gb_t *);
 extern const char *gb_to_str(const gb_t *);
 extern void gb_flush(gb_t *);  /* internal, do not use this */
+extern int stringHasWhitespace(const char *);
 
 #if USE_HTCP
 extern void htcpInit(void);

diff --git a/src/structs.h b/src/structs.h
index 4692832b5b3f661f5fec6931a50978abf9e96124..503528200fe37105307a4f9bd8d45d713e3bbcaf 100644 (file)
--- a/src/structs.h
+++ b/src/structs.h
@@ -1,6 +1,6 @@
 
 /*
- * $Id: structs.h,v 1.239 1998/10/10 14:57:43 wessels Exp $
+ * $Id: structs.h,v 1.240 1998/10/14 21:12:02 wessels Exp $
  *
  *
  * SQUID Internet Object Cache  http://squid.nlanr.net/Squid/
@@ -438,6 +438,7 @@ struct _SquidConfig {
        int http_min_poll;
     } comm_incoming;
     int max_open_disk_fds;
+    int uri_whitespace;
 };
 
 struct _SquidConfig2 {

diff --git a/src/tools.cc b/src/tools.cc
index 133787632fe6930b5aa633b343003e5a60e318c6..3d2e3f87444009df0920a897f8e51a4075af461b 100644 (file)
--- a/src/tools.cc
+++ b/src/tools.cc
@@ -1,6 +1,6 @@
 
 /*
- * $Id: tools.cc,v 1.169 1998/10/12 20:26:41 wessels Exp $
+ * $Id: tools.cc,v 1.170 1998/10/14 21:12:03 wessels Exp $
  *
  * DEBUG: section 21    Misc Functions
  * AUTHOR: Harvest Derived
@@ -830,3 +830,9 @@ debugObj(int section, int level, const char *label, void *obj, ObjPackMethod pm)
     packerClean(&p);
     memBufClean(&mb);
 }
+
+int
+stringHasWhitespace(const char *s)
+{
+    return (strcspn(s, w_space) != strlen(s));
+}

diff --git a/src/url.cc b/src/url.cc
index a5fa027456387ac709d5679de2ef20c54d46d7e8..7d91a0ec93fd117f956b4172530f7a5f404c9121 100644 (file)
--- a/src/url.cc
+++ b/src/url.cc
@@ -1,6 +1,6 @@
 
 /*
- * $Id: url.cc,v 1.108 1998/08/26 19:53:41 wessels Exp $
+ * $Id: url.cc,v 1.109 1998/10/14 21:12:04 wessels Exp $
  *
  * DEBUG: section 23    URL Parsing
  * AUTHOR: Duane Wessels
@@ -216,7 +216,7 @@ urlParse(method_t method, char *url)
     } else if (!strncmp(url, "urn:", 4)) {
        return urnParse(method, url);
     } else {
-       if (sscanf(url, "%[^:]://%[^/]%s", proto, host, urlpath) < 2)
+       if (sscanf(url, "%[^:]://%[^/]%[^\r\n]", proto, host, urlpath) < 2)
            return NULL;
        protocol = urlParseProtocol(proto);
        port = urlDefaultPort(protocol);
@@ -256,6 +256,22 @@ urlParse(method_t method, char *url)
        return NULL;
     }
 #endif
+    if (stringHasWhitespace(urlpath)) {
+       debug(23, 1) ("urlParse: URI has whitespace: {%s}\n", url);
+       switch (Config.uri_whitespace) {
+       case URI_WHITESPACE_DENY:
+           return NULL;
+       case URI_WHITESPACE_ALLOW:
+           break;
+       case URI_WHITESPACE_ENCODE:
+           t = rfc1738_escape(urlpath);
+           xstrncpy(urlpath, t, MAX_URL);
+           break;
+       case URI_WHITESPACE_CHOP:
+           *(urlpath + strcspn(urlpath, w_space)) = '\0';
+           break;
+       }
+    }
     request = requestCreate(method, protocol, urlpath);
     xstrncpy(request->host, host, SQUIDHOSTNAMELEN);
     xstrncpy(request->login, login, MAX_LOGIN_SZ);
@@ -310,7 +326,7 @@ urlCanonicalClean(const request_t * request)
     char *t;
     if (request->protocol == PROTO_URN) {
        snprintf(buf, MAX_URL, "urn:%s", strBuf(request->urlpath));
-    } else
+    } else {
        switch (request->method) {
        case METHOD_CONNECT:
            snprintf(buf, MAX_URL, "%s:%d", request->host, request->port);
@@ -339,6 +355,9 @@ urlCanonicalClean(const request_t * request)
                *(++t) = '\0';
            break;
        }
+    }
+    if (stringHasWhitespace(buf))
+       xstrncpy(buf, rfc1738_escape(buf), MAX_URL);
     return buf;
 }