unicode: Added iri_to_uri() utility function and use it in the databrowse app.

Should make getting this stuff right a bit easier.


git-svn-id: http://code.djangoproject.com/svn/django/branches/unicode@5278 bcc190cf-cafb-0310-a4f2-bffc1f526a37

django/contrib/databrowse/datastructures.py

@@ -3,12 +3,11 @@ These classes are light wrappers around Django's database API that provide
 convenience functionality and permalink functions for the databrowse app.
 """
 
-import urllib
 from django.db import models
 from django.utils import dateformat
 from django.utils.text import capfirst
 from django.utils.translation import get_date_formats
-from django.utils.encoding import smart_unicode, smart_str
+from django.utils.encoding import smart_unicode, smart_str, iri_to_uri
 
 EMPTY_VALUE = '(None)'
 
@@ -77,7 +76,7 @@ class EasyChoice(object):
         return smart_str(u'<EasyChoice for %s.%s>' % (self.model.model._meta.object_name, self.field.name))
 
     def url(self):
-        return '%s%s/%s/%s/%s/' % (self.model.site.root_url, self.model.model._meta.app_label, self.model.model._meta.module_name, self.field.field.name, urllib.quote(smart_str(self.value)))
+        return '%s%s/%s/%s/%s/' % (self.model.site.root_url, self.model.model._meta.app_label, self.model.model._meta.module_name, self.field.field.name, iri_to_uri(self.value))
 
 class EasyInstance(object):
     def __init__(self, easy_model, instance):
@@ -99,7 +98,7 @@ class EasyInstance(object):
         return self.instance._get_pk_val()
 
     def url(self):
-        return '%s%s/%s/objects/%s/' % (self.model.site.root_url, self.model.model._meta.app_label, self.model.model._meta.module_name, urllib.quote(smart_str(self.pk())))
+        return '%s%s/%s/objects/%s/' % (self.model.site.root_url, self.model.model._meta.app_label, self.model.model._meta.module_name, iri_to_uri(self.pk()))
 
     def fields(self):
         """
@@ -180,18 +179,18 @@ class EasyInstanceField(object):
             if self.field.rel.to in self.model.model_list:
                 lst = []
                 for value in self.values():
-                    url = '%s%s/%s/objects/%s/' % (self.model.site.root_url, m.model._meta.app_label, m.model._meta.module_name, urllib.quote(smart_str(value._get_pk_val())))
+                    url = '%s%s/%s/objects/%s/' % (self.model.site.root_url, m.model._meta.app_label, m.model._meta.module_name, iri_to_uri(value._get_pk_val()))
                     lst.append((smart_unicode(value), url))
             else:
                 lst = [(value, None) for value in self.values()]
         elif self.field.choices:
             lst = []
             for value in self.values():
-                url = '%s%s/%s/fields/%s/%s/' % (self.model.site.root_url, self.model.model._meta.app_label, self.model.model._meta.module_name, self.field.name, urllib.quote(smart_str(self.raw_value)))
+                url = '%s%s/%s/fields/%s/%s/' % (self.model.site.root_url, self.model.model._meta.app_label, self.model.model._meta.module_name, self.field.name, iri_to_uri(self.raw_value))
                 lst.append((value, url))
         elif isinstance(self.field, models.URLField):
             val = self.values()[0]
-            lst = [(val, urllib.quote(smart_str(val)))]
+            lst = [(val, iri_to_uri(val))]
         else:
             lst = [(self.values()[0], None)]
         return lst

django/utils/encoding.py

@@ -1,7 +1,17 @@
 import types
+import urllib
 from django.conf import settings
 from django.utils.functional import Promise
 
+class StrAndUnicode(object):
+    """
+    A class whose __str__ returns its __unicode__ as a UTF-8 bytestring.
+
+    Useful as a mix-in.
+    """
+    def __str__(self):
+        return self.__unicode__().encode('utf-8')
+
 def smart_unicode(s, encoding='utf-8', errors='strict'):
     """
     Returns a unicode object representing 's'. Treats bytestrings using the
@@ -45,12 +55,16 @@ def smart_str(s, encoding='utf-8', strings_only=False, errors='strict'):
     else:
         return s
 
-class StrAndUnicode(object):
+def iri_to_uri(iri):
     """
-    A class whose __str__ returns its __unicode__ as a UTF-8 bytestring.
+    Convert an Internationalized Resource Identifier (IRI) portion to a URI
+    portion that is suitable for inclusion in a URL.
 
-    Useful as a mix-in.
+    This is the algorithm from section 3.1 of RFC 3987.  However, since we are
+    assuming input is either UTF-8 or unicode already, we can simplify things a
+    little from the full method.
+
+    Returns an ASCII string containing the encoded result.
     """
-    def __str__(self):
-        return self.__unicode__().encode('utf-8')
+    return urllib.quote(smart_str(iri), safe='/#%[]')
 

tests/regressiontests/text/tests.py

@@ -1,5 +1,6 @@
-"""
-# Tests for stuff in django.utils.text.
+# coding: utf-8
+r"""
+# Tests for stuff in django.utils.text and other text munging util functions.
 
 >>> from django.utils.text import *
 
@@ -8,10 +9,17 @@
 [u'This', u'is', u'"a person"', u'test.']
 >>> print list(smart_split(r'''This is "a person's" test.'''))[2]
 "a person's"
->>> print list(smart_split(r'''This is "a person\\"s" test.'''))[2]
+>>> print list(smart_split(r'''This is "a person\"s" test.'''))[2]
 "a person"s"
 >>> list(smart_split('''"a 'one'''))
 [u'"a', u"'one"]
 >>> print list(smart_split(r'''all friends' tests'''))[1]
 friends'
+
+### iri_to_uri ###########################################################
+>>> from django.utils.encoding import iri_to_uri
+>>> iri_to_uri(u'red%09ros\xe9#red')
+'red%09ros%C3%A9#red'
+>>> iri_to_uri(u'/blog/for/J\xfcrgen M\xfcnster/')
+'/blog/for/J%C3%BCrgen%20M%C3%BCnster/'
 """