def skipdata(self, schm, decoder):
fn = self.__skipfn.get(schm.gettype())
if fn is not None:
return fn(schm, decoder)
else:
raise schema.AvroException("Unknown type: " +
schema.stringval(schm))
def check(self, string):
schm = schema.parse(string)
st = schema.stringval(schm)
self.assertEquals(string.replace(" ", ""), st.replace(" ", ""))
#test __eq__
self.assertEquals(schm, schema.parse(string))
#test hashcode doesn't generate infinite recursion
schm.__hash__()
randomdata = self.__random(schm)
for i in range(1, 10):
self.checkser(schm, randomdata)
self.checkdatafile(schm)
def readdata(self, actual, expected, decoder):
if actual.gettype() == schema.UNION:
actual = actual.getelementtypes()[int(decoder.readlong())]
if expected.gettype() == schema.UNION:
expected = self._resolve(actual, expected)
if actual.gettype() == schema.NULL:
return None
fn = self.__readfn.get(actual.gettype())
if fn is not None:
return fn(actual, expected, decoder)
else:
raise schema.AvroException("Unknown type: "+schema.stringval(actual));
def check(self, string):
schm = schema.parse(string)
st = schema.stringval(schm)
self.assertEquals(string.replace(" ",""), st.replace(" ",""))
#test __eq__
self.assertEquals(schm, schema.parse(string))
#test hashcode doesn't generate infinite recursion
schm.__hash__()
randomdata = self.__random(schm)
for i in range(1,10):
self.checkser(schm, randomdata)
self.checkdatafile(schm)
def readdata(self, actual, expected, decoder):
if actual.gettype() == schema.UNION:
actual = actual.getelementtypes()[int(decoder.readlong())]
if expected.gettype() == schema.UNION:
expected = self._resolve(actual, expected)
if actual.gettype() == schema.NULL:
return None
fn = self.__readfn.get(actual.gettype())
if fn is not None:
return fn(actual, expected, decoder)
else:
raise schema.AvroException("Unknown type: " +
schema.stringval(actual))
def __init__(self, schm, writer, dwriter):
self.__writer = writer
self.__encoder = Encoder(writer)
self.__dwriter = dwriter
self.__dwriter.setschema(schm)
self.__count = 0 #entries in file
self.__blockcount = 0 #entries in current block
self.__buffer = cStringIO.StringIO()
self.__bufwriter = Encoder(self.__buffer)
self.__meta = dict()
self.__sync = uuid.uuid4().bytes
self.__meta["sync"] = self.__sync
self.__meta["codec"] = "null"
self.__meta["schema"] = schema.stringval(schm)
self.__writer.write(struct.pack(len(_MAGIC).__str__() + 's', _MAGIC))
def _defaultfieldvalue(self, schm, defaultnode):
if schm.gettype() == schema.RECORD:
record = self.createrecord(schm)
for field in schm.getfields().values():
v = defaultnode.get(field.getname())
if v is None:
v = field.getdefaultvalue()
if v is not None:
record[field.getname()] = self._defaultfieldvalue(
field.getschema(), v)
return record
elif schm.gettype() == schema.ENUM:
return defaultnode
elif schm.gettype() == schema.ARRAY:
array = list()
for node in defaultnode:
array.append(
self._defaultfieldvalue(schm.getelementtype(), node))
return array
elif schm.gettype() == schema.MAP:
map = dict()
for k, v in defaultnode.items():
map[k] = self._defaultfieldvalue(schm.getvaluetype(), v)
return map
elif schm.gettype() == schema.UNION:
return self._defaultfieldvalue(schm.getelementtypes()[0],
defaultnode)
elif schm.gettype() == schema.FIXED:
return defaultnode
elif schm.gettype() == schema.STRING:
return defaultnode
elif schm.gettype() == schema.BYTES:
return defaultnode
elif schm.gettype() == schema.INT:
return int(defaultnode)
elif schm.gettype() == schema.LONG:
return long(defaultnode)
elif schm.gettype() == schema.FLOAT:
return float(defaultnode)
elif schm.gettype() == schema.DOUBLE:
return float(defaultnode)
elif schm.gettype() == schema.BOOLEAN:
return bool(defaultnode)
elif schm.gettype() == schema.NULL:
return None
else:
raise schema.AvroException("Unknown type: " +
schema.stringval(actual))
def __init__(self, schm, writer, dwriter):
self.__writer = writer
self.__encoder = Encoder(writer)
self.__dwriter = dwriter
self.__dwriter.setschema(schm)
self.__count = 0 #entries in file
self.__blockcount = 0 #entries in current block
self.__buffer = cStringIO.StringIO()
self.__bufwriter = Encoder(self.__buffer)
self.__meta = dict()
self.__sync = uuid.uuid4().bytes
self.__meta["sync"] = self.__sync
self.__meta["codec"] = "null"
self.__meta["schema"] = schema.stringval(schm)
self.__writer.write(struct.pack(len(_MAGIC).__str__()+'s',
_MAGIC))
def _defaultfieldvalue(self, schm, defaultnode):
if schm.gettype() == schema.RECORD:
record = self.createrecord(schm)
for field in schm.getfields().values():
v = defaultnode.get(field.getname())
if v is None:
v = field.getdefaultvalue()
if v is not None:
record[field.getname()] = self._defaultfieldvalue(
field.getschema(), v)
return record
elif schm.gettype() == schema.ENUM:
return defaultnode
elif schm.gettype() == schema.ARRAY:
array = list()
for node in defaultnode:
array.append(self._defaultfieldvalue(schm.getelementtype(), node))
return array
elif schm.gettype() == schema.MAP:
map = dict()
for k,v in defaultnode.items():
map[k] = self._defaultfieldvalue(schm.getvaluetype(), v)
return map
elif schm.gettype() == schema.UNION:
return self._defaultfieldvalue(schm.getelementtypes()[0], defaultnode)
elif schm.gettype() == schema.FIXED:
return defaultnode
elif schm.gettype() == schema.STRING:
return defaultnode
elif schm.gettype() == schema.BYTES:
return defaultnode
elif schm.gettype() == schema.INT:
return int(defaultnode)
elif schm.gettype() == schema.LONG:
return long(defaultnode)
elif schm.gettype() == schema.FLOAT:
return float(defaultnode)
elif schm.gettype() == schema.DOUBLE:
return float(defaultnode)
elif schm.gettype() == schema.BOOLEAN:
return bool(defaultnode)
elif schm.gettype() == schema.NULL:
return None
else:
raise schema.AvroException("Unknown type: "+schema.stringval(actual))
def __raisematchException(self, actual, expected):
raise schema.AvroException("Expected " + schema.stringval(expected) +
", found " + schema.stringval(actual))
def __raisematchException(self, actual, expected):
raise schema.AvroException("Expected "+schema.stringval(expected)+
", found "+schema.stringval(actual))
def skipdata(self, schm, decoder):
fn = self.__skipfn.get(schm.gettype())
if fn is not None:
return fn(schm, decoder)
else:
raise schema.AvroException("Unknown type: "+schema.stringval(schm));
def __init__(self, schm=None, datum=None, msg=None):
if msg is None:
msg = "Not a "+schema.stringval(schm)+": "+datum.__str__()
schema.AvroException.__init__(self, msg)
def __init__(self, schm=None, datum=None, msg=None):
if msg is None:
msg = "Not a " + schema.stringval(schm) + ": " + datum.__str__()
schema.AvroException.__init__(self, msg)
