regexp.rb in ronin-support-0.4.0

- old
+ new

@@ -26,35 +26,54 @@
 
   # Regular expression for finding MAC addresses in text
   MAC = /[0-9a-fA-F]{2}(?::[0-9a-fA-F]{2}){5}/
 
   # A regular expression for matching IPv4 Addresses.
-  IPv4 = /#{OCTET}(?:\.#{OCTET}){3}/
+  IPv4 = /#{OCTET}(?:\.#{OCTET}){3}(?:\/\d{1,2})?/
 
   # A regular expression for matching IPv6 Addresses.
-  IPv6 = /:(:[0-9a-f]{1,4}){1,7}|([0-9a-f]{1,4}::?){1,7}[0-9a-f]{1,4}(:#{IPv4})?/
+  IPv6 = union(
+    /(?:[0-9a-f]{1,4}:){6}#{IPv4}/,
+    /(?:[0-9a-f]{1,4}:){5}[0-9a-f]{1,4}:#{IPv4}/,
+    /(?:[0-9a-f]{1,4}:){5}:[0-9a-f]{1,4}:#{IPv4}/,
+    /(?:[0-9a-f]{1,4}:){1,1}(?::[0-9a-f]{1,4}){1,4}:#{IPv4}/,
+    /(?:[0-9a-f]{1,4}:){1,2}(?::[0-9a-f]{1,4}){1,3}:#{IPv4}/,
+    /(?:[0-9a-f]{1,4}:){1,3}(?::[0-9a-f]{1,4}){1,2}:#{IPv4}/,
+    /(?:[0-9a-f]{1,4}:){1,4}(?::[0-9a-f]{1,4}){1,1}:#{IPv4}/,
+    /:(?::[0-9a-f]{1,4}){1,5}:#{IPv4}/,
+    /(?:(?:[0-9a-f]{1,4}:){1,5}|:):#{IPv4}/,
+    /(?:[0-9a-f]{1,4}:){1,1}(?::[0-9a-f]{1,4}){1,6}(?:\/\d{1,3})?/,
+    /(?:[0-9a-f]{1,4}:){1,2}(?::[0-9a-f]{1,4}){1,5}(?:\/\d{1,3})?/,
+    /(?:[0-9a-f]{1,4}:){1,3}(?::[0-9a-f]{1,4}){1,4}(?:\/\d{1,3})?/,
+    /(?:[0-9a-f]{1,4}:){1,4}(?::[0-9a-f]{1,4}){1,3}(?:\/\d{1,3})?/,
+    /(?:[0-9a-f]{1,4}:){1,5}(?::[0-9a-f]{1,4}){1,2}(?:\/\d{1,3})?/,
+    /(?:[0-9a-f]{1,4}:){1,6}(?::[0-9a-f]{1,4}){1,1}(?:\/\d{1,3})?/,
+    /[0-9a-f]{1,4}(?::[0-9a-f]{1,4}){7}(?:\/\d{1,3})?/,
+    /:(?::[0-9a-f]{1,4}){1,7}(?:\/\d{1,3})?/,
+    /(?:(?:[0-9a-f]{1,4}:){1,7}|:):(?:\/\d{1,3})?/
+  )
 
   # A regular expression for matching IP Addresses.
   IP = /#{IPv4}|#{IPv6}/
 
   # Regular expression used to find host-names in text
   HOST_NAME = /(?:[a-zA-Z0-9]+(?:[_-][a-zA-Z0-9]+)*\.)+(?:#{union(Resolv::TLDS)})/i
 
   # Regular expression to match a word in the username of an email address
-  USER_NAME = /[A-Za-z](?:[A-Za-z0-9]+[\._-])*[A-Za-z0-9]+/
+  USER_NAME = /[A-Za-z](?:[A-Za-z0-9]*[\._-])*[A-Za-z0-9]+/
 
   # Regular expression to find email addresses in text
-  EMAIL_ADDR = /#{USER_NAME}(?:\.#{USER_NAME})*\@#{HOST_NAME}/
+  EMAIL_ADDR = /#{USER_NAME}\@#{HOST_NAME}/
 
   # Regular expression to find deliminators in text
   DELIM = /[;&\n\r]/
 
   # Regular expression to find identifier in text
-  IDENTIFIER = /[_a-zA-Z][a-zA-Z0-9_-]*/
+  IDENTIFIER = /[_]*[a-zA-Z]+[a-zA-Z0-9_-]*/
 
   # Regular expression to find File extensions in text
-  FILE_EXT = /(?:\.[A-Za-z0-9_-]+)+/
+  FILE_EXT = /(?:\.[A-Za-z0-9]+)+/
 
   # Regular expression to find File names in text
   FILE_NAME = /(?:[^\/\\\. ]|\\[\/\\ ])+/
 
   # Regular expression to find Files in text
@@ -62,28 +81,28 @@
 
   # Regular expression to find Directory names in text
   DIRECTORY = /(?:\.\.|\.|#{FILE})/
 
   # Regular expression to find local UNIX Paths in text
-  LOCAL_UNIX_PATH = /(?:#{DIRECTORY}\/)+#{DIRECTORY}\/?/
+  RELATIVE_UNIX_PATH = /(?:#{DIRECTORY}\/)+#{DIRECTORY}\/?/
 
   # Regular expression to find absolute UNIX Paths in text
   ABSOLUTE_UNIX_PATH = /(?:\/#{FILE})+\/?/
 
   # Regular expression to find UNIX Paths in text
-  UNIX_PATH = /#{ABSOLUTE_UNIX_PATH}|#{LOCAL_UNIX_PATH}/
+  UNIX_PATH = /#{ABSOLUTE_UNIX_PATH}|#{RELATIVE_UNIX_PATH}/
 
   # Regular expression to find local Windows Paths in text
-  LOCAL_WINDOWS_PATH = /(?:#{DIRECTORY}\\)+#{DIRECTORY}\\?/
+  RELATIVE_WINDOWS_PATH = /(?:#{DIRECTORY}\\)+#{DIRECTORY}\\?/
 
   # Regular expression to find absolute Windows Paths in text
-  ABSOLUTE_WINDOWS_PATH = /[A-Za-z]:(?:\\#{DIRECTORY})+\\?/
+  ABSOLUTE_WINDOWS_PATH = /[A-Za-z]:(?:\\#{FILE})+\\?/
 
   # Regular expression to find Windows Paths in text
-  WINDOWS_PATH = /#{ABSOLUTE_WINDOWS_PATH}|#{LOCAL_WINDOWS_PATH}/
+  WINDOWS_PATH = /#{ABSOLUTE_WINDOWS_PATH}|#{RELATIVE_WINDOWS_PATH}/
 
   # Regular expression to find local Paths in text
-  LOCAL_PATH = /#{LOCAL_UNIX_PATH}|#{LOCAL_WINDOWS_PATH}/
+  RELATIVE_PATH = /#{RELATIVE_UNIX_PATH}|#{RELATIVE_WINDOWS_PATH}/
 
   # Regular expression to find absolute Paths in text
   ABSOLUTE_PATH = /#{ABSOLUTE_UNIX_PATH}|#{ABSOLUTE_WINDOWS_PATH}/
 
   # Regular expression to find Paths in text