X-Git-Url: https://git.korewanetadesu.com/?p=python-collate.git;a=blobdiff_plain;f=collate%2F_abcollator.py;h=12575eaed844603f039009a62262fe66b8d29ace;hp=a6ec268d9bb6512bffd6b1d752088798fde6ea09;hb=92fc0878bc7b75741a3434d17310e390a9304e70;hpb=96cd5d3ad9dd1390c7739a6c3b9fa03ac3a2b4ff

diff --git a/collate/_abcollator.py b/collate/_abcollator.py
index a6ec268..12575ea 100644
--- a/collate/_abcollator.py
+++ b/collate/_abcollator.py
@@ -1,14 +1,33 @@
+import collate.strings
+
 class Collator(object):
+    encoding = "ascii"
+
     def cmp(self, string1, string2):
         """Return negative if a < b, zero if a == b, positive if a > b."""
         return cmp(self.key(string1), self.key(string2))
 
+    def key(self, string):
+        return string
+
     def words(self, string):
-        """Split the string into separate words.
+        """Split the string along word boundries."""
+        if isinstance(string, str):
+            string = string.decode(self.encoding, 'replace')
+        return string.split()
 
-        This split is done using Unicode's definition of whitespace.
+    def sortemekey(self, string):
+        """Return a key based on sortemes of a string.
+
+        If the string is a str instance, it is decoded to a unicode
+        instance according to the 'encoding' attribute of the
+        Collator.
         """
-        return string.split()
+        if isinstance(string, str):
+            string = string.decode(self.encoding, 'replace')
+
+        # Shove the sortkeyed original string on the end to resolve
+        # ties intelligently.
+        return (collate.strings.sortemes(string, self.key),
+                self.key(string))
 
-    def wordkeys(self, string):
-        return map(self.key, self.words)