blob: 091dda6d099b4f4eb4acf3b54a2c29df7507a398 [file] [log] [blame]
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 nodegroup group1 if not exists on nc1, nc2;
create dataset LineItem(LineItemType)
partitioned by key l_orderkey, l_linenumber on group1;
create dataset Orders(OrderType)
partitioned by key o_orderkey on group1;
create dataset Supplier(SupplierType)
partitioned by key s_suppkey on group1;
create dataset Region(RegionType)
partitioned by key r_regionkey on group1;
create dataset Nation(NationType)
partitioned by key n_nationkey on group1;
create dataset Part(PartType)
partitioned by key p_partkey on group1;
create dataset Partsupp(PartSuppType)
partitioned by key ps_partkey, ps_suppkey on group1;
create dataset Customer(CustomerType)
partitioned by key c_custkey on group1;
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_q20_potential_part_promotion.adm";
for $t3 in (
for $t2 in (
for $l in dataset('LineItem')
group by $l_partkey:=$l.l_partkey, $l_suppkey:=$l.l_suppkey with $l
return {
"l_partkey": $l_partkey,
"l_suppkey": $l_suppkey,
"sum_quantity": 0.5 * sum(for $i in $l return $i.l_quantity)
})
for $pst1 in (
for $ps in dataset('Partsupp')
for $t1 in
(
for $p in dataset('Part')
distinct by $p.p_partkey
return {"p_partkey": $p.p_partkey}
)
where $ps.ps_partkey = $t1.p_partkey
return {"ps_suppkey": $ps.ps_suppkey, "ps_partkey": $ps.ps_partkey, "ps_availqty": $ps.ps_availqty}
)
where $pst1.ps_partkey = $t2.l_partkey and $pst1.ps_suppkey = $t2.l_suppkey
and $pst1.ps_availqty > $t2.sum_quantity
distinct by $pst1.ps_suppkey
return {"ps_suppkey": $pst1.ps_suppkey}
)
for $t4 in (
for $n in dataset('Nation')
for $s in dataset('Supplier')
where $s.s_nationkey = $n.n_nationkey
return {"s_name": $s.s_name, "s_address": $s.s_address, "s_suppkey": $s.s_suppkey}
)
where $t3.ps_suppkey = $t4.s_suppkey
order by $t4.s_name
return {"s_name": $t4.s_name, "s_address": $t4.s_address}