blob: ad16d3cbbd4ac7d49c42816b4786f15077a6a402 [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_q16_parts_supplier_relationship.adm";
declare function tmp(){
for $psp in (
for $ps in dataset('Partsupp')
for $p in dataset('Part')
where $p.p_partkey = $ps.ps_partkey and $p.p_brand != 'Brand#45'
and not(like($p.p_type, 'MEDIUM POLISHED%'))
return
{
"p_brand": $p.p_brand,
"p_type": $p.p_type,
"p_size": $p.p_size,
"ps_suppkey": $ps.ps_suppkey
}
)
for $s in dataset('Supplier')
where $psp.ps_suppkey = $s.s_suppkey and not(like($s.s_comment, '%Customer%Complaints%'))
return
{
"p_brand": $psp.p_brand,
"p_type": $psp.p_type,
"p_size": $psp.p_size,
"ps_suppkey": $psp.ps_suppkey
}
}
for $t2 in
(
for $t in tmp()
where $t.p_size = 49 or $t.p_size = 14 or $t.p_size = 23 or
$t.p_size = 45 or $t.p_size = 19 or $t.p_size = 3 or
$t.p_size = 36 or $t.p_size = 9
group by $p_brand1:= $t.p_brand, $p_type1 := $t.p_type,
$p_size1:= $t.p_size, $ps_suppkey1:=$t.ps_suppkey with $t
return
{
"p_brand": $p_brand1,
"p_type": $p_type1,
"p_size": $p_size1,
"ps_suppkey": $ps_suppkey1
} )
group by $p_brand := $t2.p_brand, $p_type := $t2.p_type, $p_size := $t2.p_size with $t2
let $supplier_cnt := count(for $i in $t2 return $i.ps_suppkey)
order by $supplier_cnt desc, $p_brand, $p_type, $p_size
return {
"p_brand": $p_brand,
"p_type": $p_type,
"p_size": $p_size,
"supplier_cnt": $supplier_cnt
}