remove ancient and broken 2.3 support code. (#681)

fixes #680

We replace a loop over every character in a string with a single
call to encode with a custom error-handler.

We move the call to the top-level, and only do it once we encode the
entire output.

2 files changed, 26 insertions, 26 deletions

diff --git a/rdflib/plugins/serializers/nquads.py b/rdflib/plugins/serializers/nquads.py
index 26cd85ef..66c5e6f1 100644
--- a/rdflib/plugins/serializers/nquads.py
+++ b/rdflib/plugins/serializers/nquads.py
@@ -4,7 +4,7 @@ from rdflib.term import Literal
 from rdflib.serializer import Serializer
 from rdflib.py3compat import b
 
-from rdflib.plugins.serializers.nt import _xmlcharref_encode, _quoteLiteral
+from rdflib.plugins.serializers.nt import _quoteLiteral
 
 __all__ = ['NQuadsSerializer']
 
@@ -36,11 +36,10 @@ def _nq_row(triple, context):
     if isinstance(triple[2], Literal):
         return u"%s %s %s %s .\n" % (triple[0].n3(),
                                      triple[1].n3(),
-                                     _xmlcharref_encode(
-                                         _quoteLiteral(triple[2])),
+                                     _quoteLiteral(triple[2]),
                                      context.n3())
     else:
         return u"%s %s %s %s .\n" % (triple[0].n3(),
                                      triple[1].n3(),
-                                     _xmlcharref_encode(triple[2].n3()),
+                                     triple[2].n3(),
                                      context.n3())
diff --git a/rdflib/plugins/serializers/nt.py b/rdflib/plugins/serializers/nt.py
index 6b3c6485..fb6a37cf 100644
--- a/rdflib/plugins/serializers/nt.py
+++ b/rdflib/plugins/serializers/nt.py
@@ -6,7 +6,9 @@ format.
 from rdflib.term import Literal
 from rdflib.serializer import Serializer
 from rdflib.py3compat import b
+
 import warnings
+import codecs
 
 __all__ = ['NTSerializer']
 
@@ -16,6 +18,10 @@ class NTSerializer(Serializer):
     Serializes RDF graphs to NTriples format.
     """
 
+    def __init__(self, store):
+        Serializer.__init__(self, store)
+        self.encoding = 'ascii' # n-triples are ascii encoded
+
     def serialize(self, stream, base=None, encoding=None, **args):
         if base is not None:
             warnings.warn("NTSerializer does not support base.")
@@ -23,7 +29,7 @@ class NTSerializer(Serializer):
             warnings.warn("NTSerializer does not use custom encoding.")
         encoding = self.encoding
         for triple in self.store:
-            stream.write(_nt_row(triple).encode(encoding, "replace"))
+            stream.write(_nt_row(triple).encode(self.encoding, "_rdflib_nt_escape"))
         stream.write(b("\n"))
 
 
@@ -32,11 +38,11 @@ def _nt_row(triple):
         return u"%s %s %s .\n" % (
             triple[0].n3(),
             triple[1].n3(),
-            _xmlcharref_encode(_quoteLiteral(triple[2])))
+            _quoteLiteral(triple[2]))
     else:
         return u"%s %s %s .\n" % (triple[0].n3(),
                                   triple[1].n3(),
-                                  _xmlcharref_encode(triple[2].n3()))
+                                  triple[2].n3())
 
 
 def _quoteLiteral(l):
@@ -62,23 +68,18 @@ def _quote_encode(l):
         .replace('"', '\\"')\
         .replace('\r', '\\r')
 
+def _nt_unicode_error_resolver(err):
+
+    """
+    Do unicode char replaces as defined in https://www.w3.org/TR/2004/REC-rdf-testcases-20040210/#ntrip_strings
+    """
+
+    def _replace_single(c):
+        c = ord(c)
+        fmt = u'\\u%04X' if c <= 0xFFFF else u'\\U%08X'
+        return fmt % c
+
+    string = err.object[err.start:err.end]
+    return ( "".join( _replace_single(c) for c in string ), err.end )
 
-# from <http://code.activestate.com/recipes/303668/>
-def _xmlcharref_encode(unicode_data, encoding="ascii"):
-    """Emulate Python 2.3's 'xmlcharrefreplace' encoding error handler."""
-    res = ""
-
-    # Step through the unicode_data string one character at a time in
-    # order to catch unencodable characters:
-    for char in unicode_data:
-        try:
-            char.encode(encoding, 'strict')
-        except UnicodeError:
-            if ord(char) <= 0xFFFF:
-                res += '\\u%04X' % ord(char)
-            else:
-                res += '\\U%08X' % ord(char)
-        else:
-            res += char
-
-    return res
+codecs.register_error('_rdflib_nt_escape', _nt_unicode_error_resolver)