# HG changeset patch
# User paulb
# Date 1201995193 0
# Node ID c9c8bbed41882ff74c00ddd77d97367b59811216
# Parent  8a2827cf1d9828cbda2c22d326e36467ec955f01
[project @ 2008-02-02 23:33:13 by paulb]
Changed default_charset, added a safe_default_charset attribute.
Made the path/value decoding/encoding methods more modular.

diff -r 8a2827cf1d98 -r c9c8bbed4188 WebStack/Generic.py
--- a/WebStack/Generic.py	Sat Feb 02 23:32:31 2008 +0000
+++ b/WebStack/Generic.py	Sat Feb 02 23:33:13 2008 +0000
@@ -51,8 +51,15 @@
     """
 
     # The default charset ties output together with body field interpretation.
+    # It is also used to interpret URLs and paths.
 
-    default_charset = "iso-8859-1"
+    default_charset = "utf-8"
+
+    # The safe default charset provides some interpretation of incoming data of
+    # an unknown encoding. Generally, one should avoid making "last resort"
+    # interpretations, however.
+
+    safe_default_charset = "iso-8859-1"
 
     # The default path info is provided here, although the manipulated virtual
     # path info is an instance attribute set through instances of subclasses of
@@ -211,12 +218,28 @@
         """
 
         unquoted_path = urllib.unquote(path)
+        return self.decode_value(unquoted_path, encoding)
+
+    def decode_value(self, value, encoding=None):
+
+        """
+        From the given 'value', use the optional 'encoding' (if specified) to decode the
+        information and convert it to Unicode. Upon failure for a specified 'encoding'
+        or where 'encoding' is not specified, use the default character encoding to
+        perform the conversion.
+
+        Returns the 'value' as a Unicode value.
+        """
+
         if encoding is not None:
             try:
-                return unicode(unquoted_path, encoding)
+                return unicode(value, encoding)
             except UnicodeError:
                 pass
-        return unicode(unquoted_path, self.default_charset)
+        try:
+            return unicode(value, self.default_charset)
+        except UnicodeError:
+            return unicode(value, self.safe_default_charset)
 
     def encode_path(self, path, encoding=None):
 
@@ -226,10 +249,22 @@
         encoded" string.
         """
 
+        return urllib.quote(self.encode_value(path, encoding))
+
+    def encode_value(self, value, encoding=None):
+
+        """
+        Encode the given 'value', using the optional 'encoding' (if specified) or the
+        default encoding where 'encoding' is not specified, producing a plain string.
+        """
+
         if encoding is not None:
-            return urllib.quote(path.encode(encoding))
+            return value.encode(encoding)
         else:
-            return urllib.quote(path.encode(self.default_charset))
+            try:
+                return value.encode(self.default_charset)
+            except UnicodeError:
+                return value.encode(self.safe_default_charset)
 
     # Server-related methods.