New function for ULR parsing

mailnews/mime/public/nsIMimeURLUtils.h

@@ -1,58 +0,0 @@
-/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*-
- *
- * The contents of this file are subject to the Netscape Public License
- * Version 1.0 (the "NPL"); you may not use this file except in
- * compliance with the NPL.  You may obtain a copy of the NPL at
- * http://www.mozilla.org/NPL/
- *
- * Software distributed under the NPL is distributed on an "AS IS" basis,
- * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
- * for the specific language governing rights and limitations under the
- * NPL.
- *
- * The Initial Developer of this code under the NPL is Netscape
- * Communications Corporation.  Portions created by Netscape are
- * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
- * Reserved.
- */
-
-/********************************************************************************************************
- 
-     Wrapper class for various parsing URL parsing routines...
- 
-*********************************************************************************************************/
-
-#ifndef _nsIMimeURLUtils_h__
-#define _nsIMimeURLUtils_h__
-
-#include "nsISupports.h" 
-
-// {9C4DA768-07EB-11d3-8EE5-00A024669799}
-#define NS_IMIME_URLUTILS_IID \
-    { 0x9c4da768, 0x7eb, 0x11d3,              \
-    { 0x8e, 0xe5, 0x0, 0xa0, 0x24, 0x66, 0x97, 0x99 } }
-
-// {9C4DA772-07EB-11d3-8EE5-00A024669799}
-#define NS_IMIME_URLUTILS_CID \
-    { 0x9c4da772, 0x7eb, 0x11d3, \
-    { 0x8e, 0xe5, 0x0, 0xa0, 0x24, 0x66, 0x97, 0x99 } }
-
-
-class nsIMimeURLUtils : public nsISupports {
-public: 
-
-  static const nsIID& GetIID(void) { static nsIID iid = NS_IMIME_URLUTILS_IID; return iid; }
-
-  NS_IMETHOD    URLType(const char *URL, PRInt32  *retType) = 0;
-
-  NS_IMETHOD    ReduceURL (char *url, char **retURL) = 0;
-
-  NS_IMETHOD    ScanForURLs(const char *input, int32 input_size,
-                        char *output, int output_size, PRBool urls_only) = 0;
-
-  NS_IMETHOD    MakeAbsoluteURL(char * absolute_url, char * relative_url, char **retURL) = 0;
-
-  NS_IMETHOD    ScanHTMLForURLs(const char* input, char **retBuf) = 0;
-}; 
-
-#endif /* _nsIMimeURLUtils_h__ */

mailnews/mime/src/nsMimeURLUtils.cpp

@@ -1295,3 +1295,226 @@ FAIL:
     return NS_OK;
 }
 
+//
+// Also skip '>' as part of host name 
+//
+nsresult
+nsMimeURLUtils::ParseURL(const char *url, int parts_requested, char **returnVal)
+{
+  char *rv=0,*colon, *slash, *ques_mark, *hash_mark;
+  char *atSign, *host, *passwordColon, *gtThan;
+  
+  assert(url); 
+  if(!url)
+  {
+    *returnVal = mime_SACat(&rv, "");
+    return NS_ERROR_FAILURE;
+  }
+  colon = PL_strchr(url, ':'); /* returns a const char */
+  
+  /* Get the protocol part, not including anything beyond the colon */
+  if (parts_requested & GET_PROTOCOL_PART) {
+    if(colon) {
+      char val = *(colon+1);
+      *(colon+1) = '\0';
+      mime_SACopy(&rv, url);
+      *(colon+1) = val;
+      
+      /* If the user wants more url info, tack on extra slashes. */
+      if( (parts_requested & GET_HOST_PART)
+        || (parts_requested & GET_USERNAME_PART)
+        || (parts_requested & GET_PASSWORD_PART)
+        )
+      {
+        if( *(colon+1) == '/' && *(colon+2) == '/')
+          mime_SACat(&rv, "//");
+        /* If there's a third slash consider it file:/// and tack on the last slash. */
+        if( *(colon+3) == '/' )
+          mime_SACat(&rv, "/");
+      }
+		  }
+  }
+  
+  
+  /* Get the username if one exists */
+  if (parts_requested & GET_USERNAME_PART) {
+    if (colon 
+      && (*(colon+1) == '/')
+      && (*(colon+2) == '/')
+      && (*(colon+3) != '\0')) {
+      
+      if ( (slash = PL_strchr(colon+3, '/')) != NULL)
+        *slash = '\0';
+      if ( (atSign = PL_strchr(colon+3, '@')) != NULL) {
+        *atSign = '\0';
+        if ( (passwordColon = PL_strchr(colon+3, ':')) != NULL)
+          *passwordColon = '\0';
+        mime_SACat(&rv, colon+3);
+        
+        /* Get the password if one exists */
+        if (parts_requested & GET_PASSWORD_PART) {
+          if (passwordColon) {
+            mime_SACat(&rv, ":");
+            mime_SACat(&rv, passwordColon+1);
+          }
+        }
+        if (parts_requested & GET_HOST_PART)
+          mime_SACat(&rv, "@");
+        if (passwordColon)
+          *passwordColon = ':';
+        *atSign = '@';
+      }
+      if (slash)
+        *slash = '/';
+    }
+  }
+  
+  /* Get the host part */
+  if (parts_requested & GET_HOST_PART) {
+    if(colon) {
+      if(*(colon+1) == '/' && *(colon+2) == '/')
+      {
+        slash = PL_strchr(colon+3, '/');
+        
+        if(slash)
+          *slash = '\0';
+        
+        if( (atSign = PL_strchr(colon+3, '@')) != NULL)
+          host = atSign+1;
+        else
+          host = colon+3;
+        
+        ques_mark = PL_strchr(host, '?');
+        
+        if(ques_mark)
+          *ques_mark = '\0';
+        
+        gtThan = PL_strchr(host, '>');
+        
+        if (gtThan)
+          *gtThan = '\0';
+        
+          /*
+          * Protect systems whose header files forgot to let the client know when
+          * gethostbyname would trash their stack.
+        */
+#ifndef MAXHOSTNAMELEN
+#ifdef XP_OS2
+#error Managed to get into NET_ParseURL without defining MAXHOSTNAMELEN !!!
+#endif
+#define MAXHOSTNAMELEN  256
+#endif
+        /* limit hostnames to within MAXHOSTNAMELEN characters to keep
+        * from crashing
+        */
+        if(PL_strlen(host) > MAXHOSTNAMELEN)
+        {
+          char * cp;
+          char old_char;
+          
+          cp = host+MAXHOSTNAMELEN;
+          old_char = *cp;
+          
+          *cp = '\0';
+          
+          mime_SACat(&rv, host);
+          
+          *cp = old_char;
+        }
+        else
+        {
+          mime_SACat(&rv, host);
+        }
+        
+        if(slash)
+          *slash = '/';
+        
+        if(ques_mark)
+          *ques_mark = '?';
+        
+        if (gtThan)
+          *gtThan = '>';
+      }
+    }
+  }
+  
+  /* Get the path part */
+  if (parts_requested & GET_PATH_PART) {
+    if(colon) {
+      if(*(colon+1) == '/' && *(colon+2) == '/')
+      {
+        /* skip host part */
+        slash = PL_strchr(colon+3, '/');
+      }
+      else 
+      {
+      /* path is right after the colon
+        */
+        slash = colon+1;
+      }
+      
+      if(slash)
+      {
+        ques_mark = PL_strchr(slash, '?');
+        hash_mark = PL_strchr(slash, '#');
+        
+        if(ques_mark)
+          *ques_mark = '\0';
+        
+        if(hash_mark)
+          *hash_mark = '\0';
+        
+        mime_SACat(&rv, slash);
+        
+        if(ques_mark)
+          *ques_mark = '?';
+        
+        if(hash_mark)
+          *hash_mark = '#';
+      }
+		  }
+  }
+		
+  if(parts_requested & GET_HASH_PART) {
+    hash_mark = PL_strchr(url, '#'); /* returns a const char * */
+    
+    if(hash_mark)
+		  {
+      ques_mark = PL_strchr(hash_mark, '?');
+      
+      if(ques_mark)
+        *ques_mark = '\0';
+      
+      mime_SACat(&rv, hash_mark);
+      
+      if(ques_mark)
+        *ques_mark = '?';
+		  }
+  }
+  
+  if(parts_requested & GET_SEARCH_PART) {
+    ques_mark = PL_strchr(url, '?'); /* returns a const char * */
+    
+    if(ques_mark)
+    {
+      hash_mark = PL_strchr(ques_mark, '#');
+      
+      if(hash_mark)
+        *hash_mark = '\0';
+      
+      mime_SACat(&rv, ques_mark);
+      
+      if(hash_mark)
+        *hash_mark = '#';
+    }
+  }
+  
+  /* copy in a null string if nothing was copied in
+	 */
+  if(!rv)
+	   mime_SACopy(&rv, "");
+
+  *returnVal = rv;
+
+  return NS_OK;
+}

mailnews/mime/src/nsMimeURLUtils.h

@@ -47,6 +47,8 @@ public:
     NS_IMETHOD    MakeAbsoluteURL(char * absolute_url, char * relative_url, char **retURL);
 
     NS_IMETHOD    ScanHTMLForURLs(const char* input, char **retBuf);
+
+    NS_IMETHOD    ParseURL(const char *url, int parts_requested, char **returnVal);
 };