generate table from Avro schema

gistfile1.txt

#!/usr/bin/python
import json
import argparse

def convertType(type):
        if type=="long":
                return "bigint"
        else:
                return type

def gen_columns(schema):
        ret = "(" + ",".join(['%s %s' % (field['name'],convertType(field['type'][0])) for (field) in schema['fields']]) + ")"
        return ret



parser = argparse.ArgumentParser(description='Generate a create external table DDL that includes column definitions from a given Avro schema')
parser.add_argument('table_name',help='Name of Hive table');
parser.add_argument('location',help='HDFS path that contains table data');
parser.add_argument('AvroSchema',help='HDFS path with Avro schema');
parser.add_argument('--partitions',help='If creating a partitioned table, specify the partitions');

args = parser.parse_args();

schema_file = file('/tmp/'+args.table_name+'.schema','r')

schema_obj = json.load(schema_file)



if args.partitions:
        partitions = "partitioned by " + args.partitions
else:
        partitions = ""


print """CREATE EXTERNAL TABLE %s
%s
%s
ROW FORMAT SERDE 'org.apache.hadoop.hive.serde2.avro.AvroSerDe'
STORED AS INPUTFORMAT 'org.apache.hadoop.hive.ql.io.avro.AvroContainerInputFormat'
OUTPUTFORMAT 'org.apache.hadoop.hive.ql.io.avro.AvroContainerOutputFormat'
location '%s'
TBLPROPERTIES (   'avro.schema.url'='%s')""" % (args.table_name,gen_columns(schema_obj),partitions,args.location,args.AvroSchema);