Changeset 820

Timestamp:

11/13/05 20:54:18

Author:

fumanchu

Message:

Improved EncodingFilter?. As before, if "encodingFilter.encoding" is provided, that encoding is forced. But if not provided, the filter now inspects the Accept-Charset request header. Use an "encodingFilter.defaultEncoding" entry (default = utf-8) to handle "Accept-Charset: *".

Files:

• trunk/cherrypy/filters/encodingfilter.py (modified) (2 diffs)
• trunk/cherrypy/test/test_decodingencoding_filter.py (modified) (2 diffs)

trunk/cherrypy/filters/encodingfilter.py

@@ -2 +2 @@
-from basefilter import BaseFilter
-
+
-class EncodingFilter(BaseFilter):
-    """Filter that automatically encodes the response."""
+    
+    def find_acceptable_charset(self):
+        conf = cherrypy.config.get
+        response = cherrypy.response
+        
+        stream = conf("streamResponse", False)
+        if stream:
+            # Use a generator wrapper, and just pray it works as the
+            # stream is being written out.
+            def encode_body(encoding):
+                def encoder(body):
+                    for line in body:
+                        yield line.encode(encoding)
+                response.body = encoder(response.body)
+                return True
+        else:
+            response.body = ''.join([chunk for chunk in response.body])
+            def encode_body(encoding):
+                try:
+                    response.body = response.body.encode(encoding)
+                except UnicodeError:
+                    # Try the next encoding
+                    return False
+                else:
+                    return True
+        
+        failmsg = "The response could not be encoded with %s"
+        
+        enc = conf('encodingFilter.encoding', None)
+        if enc is not None:
+            # If specified, force this encoding to be used, or fail.
+            if encode_body(enc):
+                return enc
+            else:
+                raise cherrypy.HTTPError(500, failmsg % enc)
+        
+        # Parse the Accept_Charset request header, and try to provide one
+        # of the requested charsets (in order of user preference).
+        default_enc = conf('encodingFilter.defaultEncoding', 'utf-8')
+        
+        encs = cherrypy.request.headerMap.elements('Accept-Charset')
+        if encs is None:
+            # Any character-set is acceptable.
+            if encode_body(default_enc):
+                return default_enc
+            else:
+                raise cherrypy.HTTPError(500, failmsg % default_enc)
+        else:
+            newbody = None
+            charsets = [enc.value.lower() for enc in encs]
+            if "*" not in charsets:
+                # If no "*" is present in an Accept-Charset field, then all
+                # character sets not explicitly mentioned get a quality
+                # value of 0, except for ISO-8859-1, which gets a quality
+                # value of 1 if not explicitly mentioned.
+                iso = 'iso-8859-1'
+                if iso not in charsets:
+                    if encode_body(iso):
+                        return iso
+            
+            for element in encs:
+                if element.qvalue > 0:
+                    if element.value == "*":
+                        # Matches any charset. Try our default.
+                        if encode_body(default_enc):
+                            return default_enc
+                    else:
+                        if encode_body(element.value):
+                            return element.value
+        
+        # No suitable encoding found.
+        raise cherrypy.HTTPError(406)
-    
-    def beforeFinalize(self):
@@ -10 +83 @@
-            return
-        
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
+
+
+
+
+
+
+
+
+

trunk/cherrypy/test/test_decodingencoding_filter.py

@@ -10 +10 @@
-        yield europoundUnicode
-    index.exposed = True
+    
+    def mao_zedong(self):
+        return u"\u6bdb\u6cfd\u4e1c: Sing, Little Birdie?"
+    mao_zedong.exposed = True
-
-cherrypy.root = Root()
@@ -22 +26 @@
-import helper
-
-europoundUtf8 = europoundUnicode.encode('utf-8')
-
-class DecodingEncodingFilterTest(helper.CPWebCase):
-    
-    def testDecodingEncodingFilter(self):
+        europoundUtf8 = europoundUnicode.encode('utf-8')
-        self.getPage('/?param=%s' % europoundUtf8)
-        self.assertBody(europoundUtf8)
+        
+        # Default encoding should be utf-8
+        self.getPage('/mao_zedong')
+        self.assertBody("\xe6\xaf\x9b\xe6\xb3\xbd\xe4\xb8\x9c: "
+                        "Sing, Little Birdie?")
+        
+        # Ask for utf-16.
+        sing16 = ('\xff\xfe\xdbk\xfdl\x1cN:\x00 \x00S\x00i\x00n\x00g\x00,\x00 '
+                  '\x00L\x00i\x00t\x00t\x00l\x00e\x00 '
+                  '\x00B\x00i\x00r\x00d\x00i\x00e\x00?\x00')
+        self.getPage('/mao_zedong', [('Accept-Charset', 'utf-16')])
+        self.assertBody(sing16)
+        
+        # Ask for multiple encodings. ISO-8859-1 should fail, and utf-16
+        # should be produced.
+        self.getPage('/mao_zedong', [('Accept-Charset',
+                                      'iso-8859-1;q=1, utf-16;q=0.5')])
+        self.assertBody(sing16)
+        
+        # The "*" value should default to our defaultEncoding, utf-8
+        self.getPage('/mao_zedong', [('Accept-Charset', '*;q=1, utf-7;q=.2')])
+        self.assertBody("\xe6\xaf\x9b\xe6\xb3\xbd\xe4\xb8\x9c: "
+                        "Sing, Little Birdie?")
+        
+        # Only allow iso-8859-1, which should fail and raise 406.
+        self.getPage('/mao_zedong', [('Accept-Charset', 'iso-8859-1, *;q=0')])
+        self.assertStatus("406 Not Acceptable")
+        self.assertErrorPage(406)
-
-