Utils: Generalize is_string from POSIX/File

is_string from POSIX/File could be used in other situations, too.
So we move it to Utils, use it from Lint/MergeFiles and replace a
custom is_binary function.

1 files changed, 3 insertions, 7 deletions

diff --git a/src/lib/Bcfg2/Server/Lint/MergeFiles.py b/src/lib/Bcfg2/Server/Lint/MergeFiles.py
index bdb97cee2..3a6251594 100644
--- a/src/lib/Bcfg2/Server/Lint/MergeFiles.py
+++ b/src/lib/Bcfg2/Server/Lint/MergeFiles.py
@@ -6,6 +6,7 @@ import copy
 from difflib import SequenceMatcher
 import Bcfg2.Server.Lint
 from Bcfg2.Server.Plugins.Cfg import CfgGenerator
+from Bcfg2.Utils import is_string
 
 
 def threshold(val):
@@ -17,12 +18,6 @@ def threshold(val):
     return rv
 
 
-def is_binary(data):
-    """ Check if a given string contains only text or binary data. """
-    text_chars = bytearray([7, 8, 9, 10, 12, 13, 27] + range(0x20, 0x100))
-    return bool(data.translate(None, text_chars))
-
-
 class MergeFiles(Bcfg2.Server.Lint.ServerPlugin):
     """ find Probes or Cfg files with multiple similar files that
     might be merged into one """
@@ -56,7 +51,8 @@ class MergeFiles(Bcfg2.Server.Lint.ServerPlugin):
         for filename, entryset in self.core.plugins['Cfg'].entries.items():
             candidates = dict([(f, e) for f, e in entryset.entries.items()
                                if (isinstance(e, CfgGenerator) and
-                                   not is_binary(e.data) and
+                                   is_string(e.data,
+                                             Bcfg2.Options.setup.encoding) and
                                    f not in ignore and
                                    not f.endswith(".crypt"))])
             similar, identical = self.get_similar(candidates)