Changeset 876

Timestamp:

12/16/05 23:47:26

Author:

fumanchu

Message:

Better error message on "no acceptable encoding".

Files:

• trunk/cherrypy/filters/encodingfilter.py (modified) (2 diffs)
• trunk/cherrypy/test/test_decodingencoding_filter.py (modified) (1 diff)

trunk/cherrypy/filters/encodingfilter.py

@@ -5 +5 @@
-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.collapse_body()
-            def encode_body(encoding):
-                try:
-                    body = []
-                    for chunk in response.body:
-                        body.append(chunk.encode(encoding))
-                    response.body = body
-                except UnicodeError:
-                    # Try the next encoding
-                    return False
-                else:
-                    return True
-        
-        failmsg = "The response could not be encoded with %s"
-        
-        enc = conf('encoding_filter.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('encoding_filter.default_encoding', 'utf-8')
-        
-        encs = cherrypy.request.headers.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 before_finalize(self):
@@ -91 +16 @@
-            if ct.value.lower().startswith("text/"):
-                # Set "charset=..." param on response Content-Type header
-self.
+
-                cherrypy.response.headers["Content-Type"] = str(ct)
-
+
+def encode_stream(encoding):
+    """Encode a streaming response body.
+    
+    Use a generator wrapper, and just pray it works as the stream is
+    being written out.
+    """
+    def encoder(body):
+        for line in body:
+            yield line.encode(encoding)
+    cherrypy.response.body = encoder(cherrypy.response.body)
+    return True
+
+def encode_string(encoding):
+    """Encode a buffered response body."""
+    try:
+        body = []
+        for chunk in cherrypy.response.body:
+            body.append(chunk.encode(encoding))
+        cherrypy.response.body = body
+    except UnicodeError:
+        return False
+    else:
+        return True
+
+def find_acceptable_charset():
+    conf = cherrypy.config.get
+    response = cherrypy.response
+    
+    attempted_charsets = []
+    
+    stream = conf("streamResponse", False)
+    if stream:
+        encode = encode_stream
+    else:
+        response.collapse_body()
+        encode = encode_string
+    
+    failmsg = "The response could not be encoded with %s"
+    
+    enc = conf('encoding_filter.encoding', None)
+    if enc is not None:
+        # If specified, force this encoding to be used, or fail.
+        if encode(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('encoding_filter.default_encoding', 'utf-8')
+    
+    encs = cherrypy.request.headerMap.elements('Accept-Charset')
+    if encs is None:
+        # Any character-set is acceptable.
+        charsets = []
+        if encode(default_enc):
+            return default_enc
+        else:
+            raise cherrypy.HTTPError(500, failmsg % default_enc)
+    else:
+        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:
+                attempted_charsets.append(iso)
+                if encode(iso):
+                    return iso
+        
+        for element in encs:
+            if element.qvalue > 0:
+                if element.value == "*":
+                    # Matches any charset. Try our default.
+                    if default_enc not in attempted_charsets:
+                        attempted_charsets.append(default_enc)
+                        if encode(default_enc):
+                            return default_enc
+                else:
+                    encoding = element.value
+                    if encoding not in attempted_charsets:
+                        attempted_charsets.append(encoding)
+                        if encode(encoding):
+                            return encoding
+    
+    # No suitable encoding found.
+    ac = cherrypy.request.headers.get('Accept-Charset')
+    if ac is None:
+        msg = "Your client did not send an Accept-Charset header."
+    else:
+        msg = "Your client sent this Accept-Charset header: %s." % ac
+    msg += " We tried these charsets: %s." % ", ".join(attempted_charsets)
+    raise cherrypy.HTTPError(406, msg)

trunk/cherrypy/test/test_decodingencoding_filter.py

@@ -60 +60 @@
-        self.getPage('/mao_zedong', [('Accept-Charset', 'iso-8859-1, *;q=0')])
-        self.assertStatus("406 Not Acceptable")
-
+
+
+
-
-