Make sure binwalk always opens files with utf-8 encoding

In Python 3.4, if the interpreter is launched with e.g. LANG=C environment,
binwalk fails with UnicodeDecodeError, since it tries to decode included
magic files (such as src/binwalk/magic/linux) with ascii codec. This patch
makes sure that utf-8 codec is always used.

src/binwalk/core/display.py

 # Code to handle displaying and logging of results.
 # Anything in binwalk that prints results to screen should use this class.
 
+import codecs
 import sys
 import csv as pycsv
 import datetime
@@ -29,7 +30,7 @@ class Display(object):
         self._configure_formatting()
 
         if log:
-            self.fp = open(log, "a")
+            self.fp = codecs.open(log, "a", encoding='utf-8')
             if csv:
                 self.csv = pycsv.writer(self.fp)
 

src/binwalk/core/magic.py

@@ -4,6 +4,7 @@
 
 __all__ = ['Magic']
 
+import codecs
 import re
 import struct
 import datetime
@@ -789,7 +790,7 @@ class Magic(object):
 
         Returns None.
         '''
-        fp = open(fname, "r")
+        fp = codecs.open(fname, "r", encoding='utf-8')
         lines = fp.readlines()
         self.parse(lines)
         fp.close()

src/binwalk/modules/extractor.py

@@ -2,6 +2,7 @@
 # This is automatically invoked by core.module code if extraction has been
 # enabled by the user; other modules need not reference this module directly.
 
+import codecs
 import os
 import re
 import sys
@@ -317,7 +318,7 @@ class Extractor(Module):
         '''
         try:
             # Process each line from the extract file, ignoring comments
-            with open(fname, 'r') as f:
+            with codecs.open(fname, 'r', encoding='utf-8') as f:
                 for rule in f.readlines():
                     self.add_rule(rule.split(self.COMMENT_DELIM, 1)[0])
         except KeyboardInterrupt as e: