drop dataverse tpch if exists;
create dataverse tpch;
  

use dataverse tpch;

create type LineItemType as closed {
  l_orderkey: int32, 
  l_partkey: int32, 
  l_suppkey: int32, 
  l_linenumber: int32, 
  l_quantity: int32, 
  l_extendedprice: double,
  l_discount: double, 
  l_tax: double,
  l_returnflag: string, 
  l_linestatus: string, 
  l_shipdate: string,
  l_commitdate: string, 
  l_receiptdate: string, 
  l_shipinstruct: string, 
  l_shipmode: string, 
  l_comment: string
}

create type OrderType as closed {
  o_orderkey: int32, 
  o_custkey: int32, 
  o_orderstatus: string, 
  o_totalprice: double, 
  o_orderdate: string, 
  o_orderpriority: string,
  o_clerk: string, 
  o_shippriority: int32, 
  o_comment: string
}

create type CustomerType as closed {
  c_custkey: int32, 
  c_name: string, 
  c_address: string, 
  c_nationkey: int32, 
  c_phone: string, 
  c_acctbal: double, 
  c_mktsegment: string,
  c_comment: string
}

create type SupplierType as closed {
  s_suppkey: int32, 
  s_name: string,
  s_address: string,
  s_nationkey: int32,
  s_phone: string,
  s_acctbal: double,
  s_comment: string
}

create type NationType as closed {
  n_nationkey: int32,
  n_name: string,
  n_regionkey: int32,
  n_comment: string
}

create type RegionType as closed {
	r_regionkey: int32, 
	r_name: string, 
	r_comment: string
} 

create type PartType as closed {
  p_partkey: int32, 
  p_name: string, 
  p_mfgr: string,
  p_brand: string,
  p_type: string,
  p_size: int32,
  p_container: string,
  p_retailprice: double,
  p_comment: string
}

create type PartSuppType as closed {
  ps_partkey: int32, 
  ps_suppkey: int32,
  ps_availqty: int32,
  ps_supplycost: double,
  ps_comment: string 
}

create dataset LineItem(LineItemType)
  partitioned by key l_orderkey, l_linenumber;
create dataset Orders(OrderType)
  partitioned by key o_orderkey;
create dataset Supplier(SupplierType)
  partitioned by key s_suppkey;
create dataset Region(RegionType) 
  partitioned by key r_regionkey;
create dataset Nation(NationType) 
  partitioned by key n_nationkey;
create dataset Part(PartType)
  partitioned by key p_partkey;
create dataset Partsupp(PartSuppType)
  partitioned by key ps_partkey, ps_suppkey;  
create dataset Customer(CustomerType) 
  partitioned by key c_custkey;

load dataset LineItem 
using "edu.uci.ics.asterix.external.dataset.adapter.NCFileSystemAdapter"
(("path"="nc1://data/tpch0.001/lineitem.tbl"),("format"="delimited-text"),("delimiter"="|")) pre-sorted;

load dataset Orders 
using "edu.uci.ics.asterix.external.dataset.adapter.NCFileSystemAdapter"
(("path"="nc1://data/tpch0.001/orders.tbl"),("format"="delimited-text"),("delimiter"="|")) pre-sorted;

load dataset Supplier 
using "edu.uci.ics.asterix.external.dataset.adapter.NCFileSystemAdapter"
(("path"="nc1://data/tpch0.001/supplier.tbl"),("format"="delimited-text"),("delimiter"="|")) pre-sorted;

load dataset Region 
using "edu.uci.ics.asterix.external.dataset.adapter.NCFileSystemAdapter"
(("path"="nc1://data/tpch0.001/region.tbl"),("format"="delimited-text"),("delimiter"="|")) pre-sorted;

load dataset Nation 
using "edu.uci.ics.asterix.external.dataset.adapter.NCFileSystemAdapter"
(("path"="nc1://data/tpch0.001/nation.tbl"),("format"="delimited-text"),("delimiter"="|")) pre-sorted;

load dataset Part 
using "edu.uci.ics.asterix.external.dataset.adapter.NCFileSystemAdapter"
(("path"="nc1://data/tpch0.001/part.tbl"),("format"="delimited-text"),("delimiter"="|")) pre-sorted;

load dataset Partsupp 
using "edu.uci.ics.asterix.external.dataset.adapter.NCFileSystemAdapter"
(("path"="nc1://data/tpch0.001/partsupp.tbl"),("format"="delimited-text"),("delimiter"="|")) pre-sorted;

load dataset Customer 
using "edu.uci.ics.asterix.external.dataset.adapter.NCFileSystemAdapter"
(("path"="nc1://data/tpch0.001/customer.tbl"),("format"="delimited-text"),("delimiter"="|")) pre-sorted;


write output to nc1:"rttest/tpch_q13_customer_distribution.adm";

for $gco in (
	for $co in
		(
			for $c in dataset('Customer')
			return 
			{
				"c_custkey": $c.c_custkey, 
				 "o_orderkey_count": count(
						for $o in dataset('Orders')
   							where  $c.c_custkey = $o.o_custkey and not(like($o.o_comment,'%special%requests%'))
						return $o.o_orderkey)
			}
		)
	group by $c_custkey := $co.c_custkey with $co
		return{
			"c_custkey": $c_custkey, "c_count": sum(for $i in $co return $i.o_orderkey_count)
		}
)

group by $c_count := $gco.c_count with $gco
let $custdist := count($gco)
order by $custdist desc, $c_count desc
return {"c_count": $c_count, "custdist": $custdist}