#124941 charset detector does not work with single line attachment in certain situation

We need to call MimeInlineText_open_dam after ibuffer has been all processed.
Because there are 2 return path, we need to call this function each in each
path.
r=ducarroz, sr=sspitzer, a=asa

mailnews/mime/src/mimetext.cpp

@@ -214,32 +214,36 @@ MimeInlineText_parse_eof (MimeObject *obj, PRBool abort_p)
 
   MimeInlineText *text = (MimeInlineText *) obj;
 
+  /* If there is still data in the ibuffer, that means that the last line of
+   this part didn't end in a newline; so push it out anyway (this means that
+   the parse_line method will be called with a string with no trailing
+   newline, which isn't the usual case.)  We do this here, rather than in 
+   MimeObject_parse_eof, because MimeObject likes to shove things through
+   parse_line, and we have to shove it through the magic rotating-and-converting
+   code.  So, we do that and digest the buffer before MimeObject has a chance
+   to do the wrong thing.  See bug #26276 for more painful details.
+  */
+  if (!abort_p && obj->ibuffer_fp > 0)
+  {
+    status = MimeInlineText_rotate_convert_and_parse_line (obj->ibuffer,
+                                                           obj->ibuffer_fp,
+                                                           obj);
+    obj->ibuffer_fp = 0;
+    if (status < 0)
+    {
+      //we haven't find charset yet? Do it before return
+      if (text->inputAutodetect)
+        status = MimeInlineText_open_dam(nsnull, 0, obj);
+
+      obj->closed_p = PR_TRUE;
+      return status;
+    }
+  }
+
   //we haven't find charset yet? now its the time
   if (text->inputAutodetect)
      status = MimeInlineText_open_dam(nsnull, 0, obj);
-   
-  /* If there is still data in the ibuffer, that means that the last line of
-	 this part didn't end in a newline; so push it out anyway (this means that
-	 the parse_line method will be called with a string with no trailing
-	 newline, which isn't the usual case.)  We do this here, rather than in 
-	 MimeObject_parse_eof, because MimeObject likes to shove things through
-	 parse_line, and we have to shove it through the magic rotating-and-converting
-	 code.  So, we do that and digest the buffer before MimeObject has a chance
-	 to do the wrong thing.  See bug #26276 for more painful details.
-   */
-  if (!abort_p &&
-	  obj->ibuffer_fp > 0)
-	{
-	  status = MimeInlineText_rotate_convert_and_parse_line (obj->ibuffer,
-	  															 obj->ibuffer_fp,
-	  															 obj);
-	  obj->ibuffer_fp = 0;
-	  if (status < 0)
-		{
-		  obj->closed_p = PR_TRUE;
-		  return status;
-		}
-	}
+ 
   return ((MimeObjectClass*)&MIME_SUPERCLASS)->parse_eof (obj, abort_p);
 }