From 9eab31193abb8fbc35ecc26b2b1a93bda7e31882 Mon Sep 17 00:00:00 2001
From: Chris Cahoon <chris.cahoon@gmail.com>
Date: Thu, 9 Jul 2009 19:53:18 +0000
Subject: [PATCH] [soc2009/http-wsgi-improvements] Refactor setting 406 status
 codes in HttpResponse. Accessing HttpResponse.content now evaluates the
 status_code for side effects.

git-svn-id: http://code.djangoproject.com/svn/django/branches/soc2009/http-wsgi-improvements@11211 bcc190cf-cafb-0310-a4f2-bffc1f526a37
---
 django/http/__init__.py | 22 +++++++++++++---------
 django/http/charsets.py | 23 +++++++++++++++--------
 2 files changed, 28 insertions(+), 17 deletions(-)

diff --git a/django/http/__init__.py b/django/http/__init__.py
index 12ef104c39..a67dbdc949 100644
--- a/django/http/__init__.py
+++ b/django/http/__init__.py
@@ -13,7 +13,7 @@ except ImportError:
 from django.utils.datastructures import MultiValueDict, ImmutableList
 from django.utils.encoding import smart_str, iri_to_uri, force_unicode
 from django.http.multipartparser import MultiPartParser
-from django.http.charsets import get_response_encoding, get_codec
+from django.http.charsets import get_response_encoding, get_codec, UnsupportedCharset
 from django.conf import settings
 from django.core.files import uploadhandler
 from utils import *
@@ -367,8 +367,15 @@ class HttpResponse(object):
         self.set_cookie(key, max_age=0, path=path, domain=domain,
                         expires='Thu, 01-Jan-1970 00:00:00 GMT')
 
+    def _configure_body_encoding(self):
+        if not self._codec:
+            self._codec = get_codec(self._charset)
+        if not self._codec:
+            self._codec = UnsupportedCharset
+
     def _get_status_code(self):
-        if not self._valid_codec():
+        self._configure_body_encoding()
+        if self._codec is UnsupportedCharset:
             self._status_code = 406
             self._container = ['']
         return self._status_code
@@ -378,14 +385,9 @@ class HttpResponse(object):
 
     status_code = property(_get_status_code, _set_status_code)
 
-    def _valid_codec(self):
-        if not self._codec:
-            self._codec = get_codec(self._charset)
-        if not self._codec:
-            return False
-        return True
-
     def _get_content(self):
+        # Evaluate status_code for side effects
+        self._get_status_code()
         if self.has_header('Content-Encoding'):
             return ''.join(self._container)
         return smart_str(''.join(self._container), self._codec.name)
@@ -397,6 +399,8 @@ class HttpResponse(object):
     content = property(_get_content, _set_content)
 
     def __iter__(self):
+        # Evaluate status_code for side effects
+        self._get_status_code()
         self._iterator = iter(self._container)
         return self
 
diff --git a/django/http/charsets.py b/django/http/charsets.py
index 5ac333ed58..b4a4922e52 100644
--- a/django/http/charsets.py
+++ b/django/http/charsets.py
@@ -226,6 +226,13 @@ CHARSET_CODECS = {
     'windows-936': 'gbk'
 }
 
+class UnsupportedCharset(object):
+    """
+    Singleton class to indicate that our codec cannot be set due to an
+    unsupported charset in an Accept-Charset header.
+    """
+    pass
+
 def get_codec(charset):
     """
     Given the name or alias of a character set, find its Python codec if there is one.
@@ -285,6 +292,7 @@ def get_response_encoding(content_type, accept_charset_header):
     if not used_content_type:
         if not accept_charset_header: # No information to find a charset with.
             return None, None
+
         # Get list of matches for Accepted-Charsets.
         # [{ charset : q }, { charset : q }]
         match_iterator = ACCEPT_CHARSET_RE.finditer(accept_charset_header)
@@ -292,7 +300,7 @@ def get_response_encoding(content_type, accept_charset_header):
 
         # Remove charsets we cannot encode and whose q values are 0
         charsets = _process_accept_charset(accept_charset)
-        
+
         # Establish the prioritized charsets (ones we know about beforehand)
         default_charset = settings.DEFAULT_CHARSET
         fallback_charset = "ISO-8859-1"
@@ -309,11 +317,11 @@ def get_response_encoding(content_type, accept_charset_header):
         # or defaulting)
         else:
             charset = max_q_charset
-            
+
     codec = get_codec(charset)
+
     # We may reach here with no codec or no charset. We will change the status 
     # code in the HttpResponse.
-    #print charset, codec
     return charset, codec
 
 # NOTE -- make sure we are not duping the processing of q values
@@ -324,10 +332,10 @@ def _process_accept_charset(accept_charset):
     names, and excludes charsets without Python codecs and whose q values are 0.
     '''
     accepted_charsets = {}
-    
+
     default_value = 1
     wildcard = False
-    
+
     for potential in accept_charset:
         charset = potential["charset"].strip()            
         # The default quality value is 1
@@ -341,11 +349,10 @@ def _process_accept_charset(accept_charset):
         elif charset == "*" and q >= 0 and q <= 1:
             default_value = q
             wildcard = True
-            
+
     if settings.DEFAULT_CHARSET not in accepted_charsets:
         accepted_charsets[settings.DEFAULT_CHARSET] = default_value 
     if "ISO-8859-1" not in accepted_charsets and wildcard: 
         accepted_charsets["ISO-8859-1"] = default_value
-      
-      
+
     return accepted_charsets