Example config

.gitignore

@@ -1 +1,3 @@
 *.pyc
+venv
+config.yaml

config.yaml.example

@@ -0,0 +1,5 @@
+mysql:
+        hostname: localhost
+        username: root
+        password: ""
+        database: tahoes3

main.py

@@ -0,0 +1,39 @@
+import lxml
+
+import tornado.httpserver
+import tornado.ioloop
+import tornado.options
+import tornado.web
+
+from tornado.options import define, options
+define("port", default=8888, help="run on the given port", type=int)
+
+class ServiceHandler(tornado.web.RequestHandler):
+	def get(self):
+		E = ElementMaker(namespace="http://doc.s3.amazonaws.com/2006-03-01", nsmap=NSMAP)
+		Et = ElementMaker(namespace="http://doc.cryto.net/xml/tahoe-s3", nsmap=NSMAP)
+		
+		buckets = []
+		
+		
+		
+		doc = E.ListAllMyBucketsResult(
+			E.Owner(
+				E.Id("bcaf1ffd86f461ca5fb16fd081034f"),
+				E.DisplayName("webfile")
+			),
+			E.Buckets(
+				*buckets
+			)
+		)
+
+if __name__ == "__main__":
+	routes = [
+		(r"/", ServiceHandler)
+	]
+	
+	tornado.options.parse_command_line()
+	app = tornado.web.Application(handlers=routes)
+	http_server = tornado.httpserver.HTTPServer(app)
+	http_server.listen(options.port)
+	tornado.ioloop.IOLoop.instance().start()

test-hmac.py

@@ -0,0 +1,67 @@
+import hmac, hashlib, base64, urllib
+
+pubkey = "AKIAIOSFODNN7EXAMPLE"
+privkey = "wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY"
+
+def sign(key, message):
+	return hmac.new(key, message.encode('utf-8'), hashlib.sha1).digest()
+
+class S3SignatureGenerator(object):
+        def set_keys(self, access_key, secret_key):
+                self.access_key = access_key
+                self.secret_key = secret_key
+        
+        def generate(self, verb, bucket, path, headers=[], content_md5="", content_type=""):
+                date = ""
+                
+                elements = [
+                        verb,
+                        content_md5,
+                        content_type,
+                        date
+                ]
+                
+                if len(headers) > 0:
+                        elements.append(self.canonicalize_headers(headers))
+                        
+                elements.append(self.canonicalize_resource(bucket, path))
+                
+                sts = "\n".join()
+                
+        def canonicalize_resource(self, bucket=None, path="", subresources={}):
+                # TODO: Multi-object DELETE parameter?
+                if bucket is None:
+                        result = "/%s/" % path
+                else:
+                        result = "/%s%s" % (bucket, path)
+                
+                if len(subresources) > 0:
+                        subresources = sorted(subresources.items())
+                        subresource_strings = []
+                        
+                        for subresource, value in subresources:
+                                if subresource in ("acl", "lifecycle", "location", "logging", "notification", "partNumber", "policy", "requestPayment", "torrent", "uploadId", "uploads", "versionId", "versioning", "versions", "website"):
+                                        if value is None:
+                                                subresource_strings.append(subresource)
+                                        else:
+                                                # Not sure if this is a correct implementation of the encoding...
+                                                if subresource not in ("response-content-type", "response-content-language", "response-expires", "response-cache-control", "response-content-disposition", "response-content-encoding"):
+                                                        value = urllib.quote(str(value))
+                                                        
+                                                subresource_strings.append("%s=%s" % (subresource, value))
+                        
+                        if len(subresource_strings) > 0: # Don't add this if there were no qualifying subresource specifiers
+                                result += "?%s" % "&".join(subresource_strings)
+                        
+                return result
+                
+        def canonicalize_headers(self, ):
+                pass
+                
+        def sign(self, sts):
+                return base64.b64encode(hmac.new(self.secret_key, sts.encode("utf-8"), hashlib.sha1).digest())
+
+s = S3SignatureGenerator()
+print s.canonicalize_resource("testbucket", "/test/path", subresources={"torrent": None, "abc-id": 4})
+
+#print "AWS %s:%s" % (pubkey, base64.b64encode(sign(privkey, sts)))

test-xml.py

@@ -0,0 +1,33 @@
+from lxml.builder import ElementMaker
+from lxml import etree
+
+NSMAP = {None: "http://doc.s3.amazonaws.com/2006-03-01", "tahoe": "http://doc.cryto.net/xml/tahoe-s3"}
+
+E = ElementMaker(namespace="http://doc.s3.amazonaws.com/2006-03-01", nsmap=NSMAP)
+Et = ElementMaker(namespace="http://doc.cryto.net/xml/tahoe-s3", nsmap=NSMAP)
+
+buckets = (
+	E.Bucket(
+		E.Name("quotes"),
+		E.CreationDate("2006-02-03T16:45:09.000Z"),
+		Et.ReadCapability("ABC23423346"),
+		Et.WriteCapability("H98NG3040S"),
+		Et.VerifyCapability("3148J15990JE"),
+	),
+	E.Bucket(
+		E.Name("samples"),
+		E.CreationDate("2006-02-03T16:41:58.000Z")
+	)
+)
+
+doc = E.ListAllMyBucketsResult(
+	E.Owner(
+		E.Id("bcaf1ffd86f461ca5fb16fd081034f"),
+		E.DisplayName("webfile")
+	),
+	E.Buckets(
+		*buckets
+	)
+)
+
+print etree.tostring(doc, pretty_print=True)