blob: c91eca5e3aedbd4d29d6e683b9e684a5feaf00b9 [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 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_q7_volume_shipping.adm";
declare function q7_volume_shipping_tmp(){
for $n1 in dataset('Nation')
for $n2 in dataset('Nation')
where $n2.n_name='GERMANY' or $n1.n_name='GERMANY'
return
{
"supp_nation": $n1.n_name,
"cust_nation": $n2.n_name,
"s_nationkey": $n1.n_nationkey,
"c_nationkey": $n2.n_nationkey
}
}
for $locs in (
for $loc in (
for $lo in
(
for $l in dataset('LineItem')
for $o in dataset('Orders')
where
$o.o_orderkey = $l.l_orderkey and $l.l_shipdate >= '1992-01-01'
and $l.l_shipdate <= '1996-12-31'
return
{
"l_shipdate": $l.l_shipdate,
"l_extendedprice": $l.l_extendedprice,
"l_discount": $l.l_discount,
"l_suppkey": $l.l_suppkey,
"o_custkey": $o.o_custkey
}
)
for $c in dataset('Customer')
where
$c.c_custkey = $lo.o_custkey
return
{
"l_shipdate": $lo.l_shipdate,
"l_extendedprice": $lo.l_extendedprice,
"l_discount": $lo.l_discount,
"l_suppkey": $lo.l_suppkey,
"c_nationkey": $c.c_nationkey
}
)
for $s in dataset('Supplier')
where
$s.s_suppkey = $loc.l_suppkey
return {
"l_shipdate": $loc.l_shipdate,
"l_extendedprice": $loc.l_extendedprice,
"l_discount": $loc.l_discount,
"c_nationkey": $loc.c_nationkey,
"s_nationkey": $s.s_nationkey
}
)
for $t in q7_volume_shipping_tmp()
where
$locs.c_nationkey = $t.c_nationkey and $locs.s_nationkey = $t.s_nationkey
let $l_year0 := year($locs.l_shipdate)
group by $supp_nation := $t.supp_nation, $cust_nation := $t.cust_nation, $l_year := $l_year0
with $locs
let $revenue := sum(for $i in $locs return $i.l_extendedprice * (1 - $i.l_discount))
order by $supp_nation, $cust_nation, $l_year
return {
"supp_nation": $supp_nation,
"cust_nation": $cust_nation,
"l_year": $l_year,
"revenue": $revenue
}