* [Refactor] Moved Reverse Proxy Code

Moved the reverse proxy code from reqs.c into it's own files
(reverse_proxy.c).  The code in reqs.c is way too complicated, so I
want to move unrelated code into their own files to simplify the main
concepts in reqs.c.

8 files changed, 214 insertions, 157 deletions

diff --git a/configure.ac b/configure.ac
index 83d727d..2acbeb0 100644
--- a/configure.ac
+++ b/configure.ac
@@ -1,4 +1,4 @@
-dnl $Id: configure.ac,v 2.69 2004-08-24 18:43:17 rjkaes Exp $
+dnl $Id: configure.ac,v 2.70 2005-08-16 04:03:18 rjkaes Exp $
 
 dnl Devlopers, please strive to achieve this order:
 dnl
@@ -130,6 +130,7 @@ TP_ARG_ENABLE(reverse,
               [Enable reverse proxying (default is NO)]
               no)
 if test x"$reverse_enabled" = x"yes"; then
+    ADDITIONAL_OBJECTS="$ADDITIONAL_OBJECTS reverse_proxy.o"
     AC_DEFINE(REVERSE_SUPPORT)
 fi
 
diff --git a/src/Makefile.am b/src/Makefile.am
index becee65..3904cd3 100644
--- a/src/Makefile.am
+++ b/src/Makefile.am
@@ -1,4 +1,4 @@
-# $Id: Makefile.am,v 1.22 2005-07-12 17:39:40 rjkaes Exp $
+# $Id: Makefile.am,v 1.23 2005-08-16 04:03:19 rjkaes Exp $
 #
 # Copyright (C) 2000  Robert James Kaes (rjkaes@users.sourceforge.net)
 #
@@ -40,7 +40,8 @@ tinyproxy_SOURCES = \
 	utils.c utils.h \
 	vector.c vector.h
 
-EXTRA_tinyproxy_SOURCES = filter.c filter.h grammar.h
+EXTRA_tinyproxy_SOURCES = filter.c filter.h \
+	reverse_proxy.c reverse_proxy.h
 tinyproxy_DEPENDENCIES = @ADDITIONAL_OBJECTS@
 tinyproxy_LDADD = @ADDITIONAL_OBJECTS@
 
diff --git a/src/conffile.c b/src/conffile.c
index 453942f..1730998 100644
--- a/src/conffile.c
+++ b/src/conffile.c
@@ -1,4 +1,4 @@
-/* $Id: conffile.c,v 1.5 2005-08-15 03:54:31 rjkaes Exp $
+/* $Id: conffile.c,v 1.6 2005-08-16 04:03:19 rjkaes Exp $
  *
  * Parses the configuration file and sets up the config_s structure for
  * use by the application.  This file replaces the old grammar.y and
@@ -30,6 +30,7 @@
 #include "htmlerror.h"
 #include "log.h"
 #include "reqs.h"
+#include "reverse_proxy.h"
 
 /*
  * The configuration directives are defined in the structure below.  Each
diff --git a/src/reqs.c b/src/reqs.c
index e5e4061..ff65343 100644
--- a/src/reqs.c
+++ b/src/reqs.c
@@ -1,4 +1,4 @@
-/* $Id: reqs.c,v 1.120 2005-08-15 03:54:31 rjkaes Exp $
+/* $Id: reqs.c,v 1.121 2005-08-16 04:03:19 rjkaes Exp $
  *
  * This is where all the work in tinyproxy is actually done. Incoming
  * connections have a new child created for them. The child then
@@ -6,7 +6,7 @@
  * and then relays the bytes between the two.
  *
  * Copyright (C) 1998	    Steven Young
- * Copyright (C) 1999-2004  Robert James Kaes (rjkaes@users.sourceforge.net)
+ * Copyright (C) 1999-2005  Robert James Kaes (rjkaes@users.sourceforge.net)
  * Copyright (C) 2000       Chris Lightfoot (chris@ex-parrot.com)
  * Copyright (C) 2002       Petr Lampa (lampa@fit.vutbr.cz)
  *
@@ -39,6 +39,7 @@
 #include "text.h"
 #include "utils.h"
 #include "vector.h"
+#include "reverse_proxy.h"
 
 /*
  * Maximum length of a HTTP line
@@ -497,75 +498,6 @@ upstream_get(char *host)
 }
 #endif
 
-#ifdef REVERSE_SUPPORT
-/*
- * Add entry to the reversepath list
- */
-void
-reversepath_add(const char *path, const char *url)
-{
-        struct reversepath *reverse;
-
-        if (url == NULL) {
-                log_message(LOG_WARNING,
-                            "Illegal reverse proxy rule: missing url");
-                return;
-        }
-
-        if (!strstr(url, "://")) {
-                log_message(LOG_WARNING,
-                            "Skipping reverse proxy rule: '%s' is not a valid url",
-                            url);
-                return;
-        }
-
-        if (path && *path != '/') {
-                log_message(LOG_WARNING,
-                            "Skipping reverse proxy rule: path '%s' doesn't start with a /",
-                            path);
-                return;
-        }
-
-        if (!(reverse = safemalloc(sizeof(struct reversepath)))) {
-                log_message(LOG_ERR,
-                            "Unable to allocate memory in reversepath_add()");
-                return;
-        }
-
-        if (!path)
-                reverse->path = safestrdup("/");
-        else
-                reverse->path = safestrdup(path);
-
-        reverse->url = safestrdup(url);
-
-        reverse->next = config.reversepath_list;
-        config.reversepath_list = reverse;
-
-        log_message(LOG_INFO,
-                    "Added reverse proxy rule: %s -> %s", reverse->path,
-                    reverse->url);
-}
-
-/*
- * Check if a request url is in the reversepath list
- */
-static struct reversepath *
-reversepath_get(char *url)
-{
-        struct reversepath *reverse = config.reversepath_list;
-
-        while (reverse) {
-                if (strstr(url, reverse->path) == url)
-                        return reverse;
-
-                reverse = reverse->next;
-        }
-
-        return NULL;
-}
-#endif
-
 /*
  * Create a connection for HTTP connections.
  */
@@ -616,16 +548,7 @@ process_request(struct conn_s *connptr, hashmap_t hashofheaders)
 {
         char *url;
         struct request_s *request;
-
-#ifdef REVERSE_SUPPORT
-        char *rewrite_url = NULL;
-        char *cookie = NULL;
-        char *cookieval;
-        struct reversepath *reverse;
-#endif
-
         int ret;
-
         size_t request_len;
 
         /* NULL out all the fields so frees don't cause segfaults. */
@@ -698,72 +621,24 @@ process_request(struct conn_s *connptr, hashmap_t hashofheaders)
                 return NULL;
         }
 #ifdef REVERSE_SUPPORT
-        /*
-         * Reverse proxy URL rewriting.
-         */
         if (config.reversepath_list != NULL) {
-                /* Reverse requests always start with a slash */
-                if (*url == '/') {
-                        /* First try locating the reverse mapping by request url */
-                        reverse = reversepath_get(url);
-                        if (reverse) {
-                                rewrite_url = safemalloc(strlen(url) +
-                                                         strlen(reverse->url) +
-                                                         1);
-                                strcpy(rewrite_url, reverse->url);
-                                strcat(rewrite_url,
-                                       url + strlen(reverse->path));
-                        } else if (config.reversemagic
-                                   && hashmap_entry_by_key(hashofheaders,
-                                                           "cookie",
-                                                           (void **)&cookie) >
-                                   0) {
-
-                                /* No match - try the magical tracking cookie next */
-                                if ((cookieval =
-                                     strstr(cookie, REVERSE_COOKIE "="))
-                                    && (reverse =
-                                        reversepath_get(cookieval +
-                                                        strlen(REVERSE_COOKIE) +
-                                                        1))) {
-
-                                        rewrite_url = safemalloc(strlen(url) +
-                                                                 strlen
-                                                                 (reverse->
-                                                                  url) + 1);
-                                        strcpy(rewrite_url, reverse->url);
-                                        strcat(rewrite_url, url + 1);
-
-                                        log_message(LOG_INFO,
-                                                    "Magical tracking cookie says: %s",
-                                                    reverse->path);
-                                }
-                        }
-                }
+                /*
+                 * Rewrite the URL based on the reverse path.  After calling
+                 * reverse_rewrite_url "url" can be freed since we either
+                 * have the newly rewritten URL, or something failed and
+                 * we'll be closing anyway.
+                 */
+                char *reverse_url;
 
-                /* Forward proxy support off and no reverse path match found */
-                if (config.reverseonly && !rewrite_url) {
-                        log_message(LOG_ERR, "Bad request");
-                        indicate_http_error(connptr, 400, "Bad Request",
-                                            "detail",
-                                            "Request has an invalid URL", "url",
-                                            url, NULL);
+                reverse_url = reverse_rewrite_url(connptr, hashofheaders, url);
+                safefree(url);
 
-                        safefree(url);
+                if (!reverse_url) {
                         free_request_struct(request);
-
                         return NULL;
+                } else {
+                        url = reverse_url;
                 }
-
-                log_message(LOG_CONN, "Rewriting URL: %s -> %s",
-                            url, rewrite_url);
-
-                safefree(url);
-                url = rewrite_url;
-
-                /* Store reverse path so that the magical tracking cookie can be set */
-                if (config.reversemagic)
-                        connptr->reversepath = safestrdup(reverse->path);
         }
 #endif
 
diff --git a/src/reqs.h b/src/reqs.h
index e5288a0..35061be 100644
--- a/src/reqs.h
+++ b/src/reqs.h
@@ -1,4 +1,4 @@
-/* $Id: reqs.h,v 1.6 2005-07-12 17:39:44 rjkaes Exp $
+/* $Id: reqs.h,v 1.7 2005-08-16 04:03:19 rjkaes Exp $
  *
  * See 'reqs.c' for a detailed description.
  *
@@ -22,6 +22,5 @@
 extern void handle_connection(int fd);
 extern void add_connect_port_allowed(int port);
 extern void upstream_add(const char *host, int port, const char *domain);
-extern void reversepath_add(const char *path, const char *url);
 
 #endif
diff --git a/src/reverse_proxy.c b/src/reverse_proxy.c
new file mode 100644
index 0000000..40f15f4
--- /dev/null
+++ b/src/reverse_proxy.c
@@ -0,0 +1,154 @@
+/* $Id: reverse_proxy.c,v 1.1 2005-08-16 04:03:19 rjkaes Exp $
+ *
+ * Allow tinyproxy to be used as a reverse proxy.
+ *
+ * Copyright (C) 1999-2005  Robert James Kaes (rjkaes@users.sourceforge.net)
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ */
+
+#include "tinyproxy.h"
+#include "reverse_proxy.h"
+
+#include "conns.h"
+#include "heap.h"
+#include "htmlerror.h"
+#include "log.h"
+
+/*
+ * Add entry to the reversepath list
+ */
+void
+reversepath_add(const char *path, const char *url)
+{
+        struct reversepath *reverse;
+
+        if (url == NULL) {
+                log_message(LOG_WARNING,
+                            "Illegal reverse proxy rule: missing url");
+                return;
+        }
+
+        if (!strstr(url, "://")) {
+                log_message(LOG_WARNING,
+                            "Skipping reverse proxy rule: '%s' is not a valid url",
+                            url);
+                return;
+        }
+
+        if (path && *path != '/') {
+                log_message(LOG_WARNING,
+                            "Skipping reverse proxy rule: path '%s' doesn't start with a /",
+                            path);
+                return;
+        }
+
+        if (!(reverse = safemalloc(sizeof(struct reversepath)))) {
+                log_message(LOG_ERR,
+                            "Unable to allocate memory in reversepath_add()");
+                return;
+        }
+
+        if (!path)
+                reverse->path = safestrdup("/");
+        else
+                reverse->path = safestrdup(path);
+
+        reverse->url = safestrdup(url);
+
+        reverse->next = config.reversepath_list;
+        config.reversepath_list = reverse;
+
+        log_message(LOG_INFO,
+                    "Added reverse proxy rule: %s -> %s", reverse->path,
+                    reverse->url);
+}
+
+/*
+ * Check if a request url is in the reversepath list
+ */
+struct reversepath *
+reversepath_get(char *url)
+{
+        struct reversepath *reverse = config.reversepath_list;
+
+        while (reverse) {
+                if (strstr(url, reverse->path) == url)
+                        return reverse;
+
+                reverse = reverse->next;
+        }
+
+        return NULL;
+}
+
+/*
+ * Rewrite the URL for reverse proxying.
+ */
+char *
+reverse_rewrite_url(struct conn_s *connptr, hashmap_t hashofheaders, char *url)
+{
+        char *rewrite_url = NULL;
+        char *cookie = NULL;
+        char *cookieval;
+        struct reversepath *reverse;
+
+        /* Reverse requests always start with a slash */
+        if (*url == '/') {
+                /* First try locating the reverse mapping by request url */
+                reverse = reversepath_get(url);
+                if (reverse) {
+                        rewrite_url = safemalloc(strlen(url) +
+                                                 strlen(reverse->url) + 1);
+                        strcpy(rewrite_url, reverse->url);
+                        strcat(rewrite_url, url + strlen(reverse->path));
+                } else if (config.reversemagic
+                           && hashmap_entry_by_key(hashofheaders,
+                                                   "cookie",
+                                                   (void **)&cookie) > 0) {
+
+                        /* No match - try the magical tracking cookie next */
+                        if ((cookieval = strstr(cookie, REVERSE_COOKIE "="))
+                            && (reverse =
+                                reversepath_get(cookieval +
+                                                strlen(REVERSE_COOKIE) + 1))) {
+
+                                rewrite_url = safemalloc(strlen(url) +
+                                                         strlen
+                                                         (reverse->url) + 1);
+                                strcpy(rewrite_url, reverse->url);
+                                strcat(rewrite_url, url + 1);
+
+                                log_message(LOG_INFO,
+                                            "Magical tracking cookie says: %s",
+                                            reverse->path);
+                        }
+                }
+        }
+
+        /* Forward proxy support off and no reverse path match found */
+        if (config.reverseonly && !rewrite_url) {
+                log_message(LOG_ERR, "Bad request");
+                indicate_http_error(connptr, 400, "Bad Request",
+                                    "detail",
+                                    "Request has an invalid URL", "url",
+                                    url, NULL);
+                return NULL;
+        }
+
+        log_message(LOG_CONN, "Rewriting URL: %s -> %s", url, rewrite_url);
+
+        /* Store reverse path so that the magical tracking cookie can be set */
+        if (config.reversemagic)
+                connptr->reversepath = safestrdup(reverse->path);
+
+        return rewrite_url;
+}
diff --git a/src/reverse_proxy.h b/src/reverse_proxy.h
new file mode 100644
index 0000000..f27a9de
--- /dev/null
+++ b/src/reverse_proxy.h
@@ -0,0 +1,36 @@
+/* $Id: reverse_proxy.h,v 1.1 2005-08-16 04:03:19 rjkaes Exp $
+ *
+ * See 'reverse_proxy.c' for a detailed description.
+ *
+ * Copyright (C) 2005  Robert James Kaes (rjkaes@users.sourceforge.net)
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ */
+
+#ifndef TINYPROXY_REVERSE_PROXY_H
+#define TINYPROXY_REVERSE_PROXY_H
+
+#include "conns.h"
+
+struct reversepath {
+        struct reversepath *next;
+        char *path;
+        char *url;
+};
+
+#define REVERSE_COOKIE "yummy_magical_cookie"
+
+extern void reversepath_add(const char *path, const char *url);
+extern struct reversepath *reversepath_get(char *url);
+extern char *reverse_rewrite_url(struct conn_s *connptr,
+                                 hashmap_t hashofheaders, char *url);
+
+#endif
diff --git a/src/tinyproxy.h b/src/tinyproxy.h
index 4cb091c..5e14857 100644
--- a/src/tinyproxy.h
+++ b/src/tinyproxy.h
@@ -1,4 +1,4 @@
-/* $Id: tinyproxy.h,v 1.47 2005-08-15 03:54:31 rjkaes Exp $
+/* $Id: tinyproxy.h,v 1.48 2005-08-16 04:03:19 rjkaes Exp $
  *
  * See 'tinyproxy.c' for a detailed description.
  *
@@ -38,16 +38,6 @@ struct upstream {
         in_addr_t ip, mask;
 };
 
-#ifdef REVERSE_SUPPORT
-struct reversepath {
-        struct reversepath *next;
-        char *path;
-        char *url;
-};
-
-#define REVERSE_COOKIE "yummy_magical_cookie"
-#endif
-
 /*
  * Hold all the configuration time information.
  */