You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@hive.apache.org by jc...@apache.org on 2020/08/10 14:16:31 UTC
[hive] branch master updated: HIVE-23940: Add TPCH tables (scale
factor 0.001) as qt datasets (Stamatis Zampetakis,
reviewed by Jesus Camacho Rodriguez)
This is an automated email from the ASF dual-hosted git repository.
jcamacho pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/hive.git
The following commit(s) were added to refs/heads/master by this push:
new ddee818 HIVE-23940: Add TPCH tables (scale factor 0.001) as qt datasets (Stamatis Zampetakis, reviewed by Jesus Camacho Rodriguez)
ddee818 is described below
commit ddee8180d51c7c6ba1366ea9be9bd003112821b8
Author: Stamatis Zampetakis <za...@gmail.com>
AuthorDate: Wed Jun 3 17:52:10 2020 +0200
HIVE-23940: Add TPCH tables (scale factor 0.001) as qt datasets (Stamatis Zampetakis, reviewed by Jesus Camacho Rodriguez)
---
data/files/datasets/lineitem/load.hive.sql | 2 +-
data/files/datasets/part/load.hive.sql | 2 +-
.../datasets/tpch_0_001.customer/load.hive.sql | 19 ++++
.../load.hive.sql | 10 ++-
.../files/datasets/tpch_0_001.nation/load.hive.sql | 15 ++++
.../files/datasets/tpch_0_001.orders/load.hive.sql | 20 +++++
data/files/datasets/tpch_0_001.part/load.hive.sql | 19 ++++
.../datasets/tpch_0_001.partsupp/load.hive.sql | 16 ++++
.../files/datasets/tpch_0_001.region/load.hive.sql | 14 +++
.../datasets/tpch_0_001.supplier/load.hive.sql | 18 ++++
data/files/lineitem.txt | 100 ---------------------
data/files/part_tiny.txt | 26 ------
data/files/tpch/sf0_001/customer.tbl.bz2 | Bin 0 -> 8627 bytes
data/files/tpch/sf0_001/lineitem.tbl.bz2 | Bin 0 -> 118989 bytes
data/files/tpch/sf0_001/nation.tbl.bz2 | Bin 0 -> 944 bytes
data/files/tpch/sf0_001/orders.tbl.bz2 | Bin 0 -> 30437 bytes
data/files/tpch/sf0_001/part.tbl.bz2 | Bin 0 -> 5141 bytes
data/files/tpch/sf0_001/partsupp.tbl.bz2 | Bin 0 -> 19353 bytes
data/files/tpch/sf0_001/region.tbl.bz2 | Bin 0 -> 282 bytes
data/files/tpch/sf0_001/supplier.tbl.bz2 | Bin 0 -> 791 bytes
data/files/tpch/tiny/lineitem.tbl.bz2 | Bin 0 -> 3524 bytes
data/files/tpch/tiny/part.tbl.bz2 | Bin 0 -> 1175 bytes
.../queries/clientpositive/ptf_general_queries.q | 2 +-
.../clientpositive/ptf_general_queries.q.out | 4 +-
.../special_character_in_tabnames_1.q | 4 +-
.../special_character_in_tabnames_quotes_1.q | 4 +-
.../queries/clientpositive/subquery_in_having.q | 2 +-
.../queries/clientpositive/tpch_sf0_001_datasets.q | 19 ++++
.../queries/clientpositive/tpch_tiny_datasets.q | 5 ++
.../queries/clientpositive/vector_case_when_1.q | 2 +-
.../test/queries/clientpositive/vectorized_ptf.q | 2 +-
.../llap/special_character_in_tabnames_1.q.out | 8 +-
.../special_character_in_tabnames_quotes_1.q.out | 8 +-
.../clientpositive/llap/subquery_in_having.q.out | 4 +-
.../llap/tpch_sf0_001_datasets.q.out | 78 ++++++++++++++++
.../clientpositive/llap/tpch_tiny_datasets.q.out | 18 ++++
.../clientpositive/llap/vector_case_when_1.q.out | 4 +-
.../clientpositive/llap/vectorized_ptf.q.out | 4 +-
.../clientpositive/spark/vectorized_ptf.q.out | 4 +-
39 files changed, 275 insertions(+), 158 deletions(-)
diff --git a/data/files/datasets/lineitem/load.hive.sql b/data/files/datasets/lineitem/load.hive.sql
index ab2f5fb..5174f62 100644
--- a/data/files/datasets/lineitem/load.hive.sql
+++ b/data/files/datasets/lineitem/load.hive.sql
@@ -17,7 +17,7 @@ CREATE TABLE lineitem (L_ORDERKEY INT,
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '|';
-LOAD DATA LOCAL INPATH '${hiveconf:test.data.dir}/lineitem.txt' OVERWRITE INTO TABLE lineitem;
+LOAD DATA LOCAL INPATH '${hiveconf:test.data.dir}/tpch/tiny/lineitem.tbl.bz2' OVERWRITE INTO TABLE lineitem;
analyze table lineitem compute statistics;
analyze table lineitem compute statistics for columns;
diff --git a/data/files/datasets/part/load.hive.sql b/data/files/datasets/part/load.hive.sql
index c24b505..a45c684 100644
--- a/data/files/datasets/part/load.hive.sql
+++ b/data/files/datasets/part/load.hive.sql
@@ -10,7 +10,7 @@ CREATE TABLE part(
p_comment STRING
);
-LOAD DATA LOCAL INPATH '${hiveconf:test.data.dir}/part_tiny.txt' overwrite into table part;
+LOAD DATA LOCAL INPATH '${hiveconf:test.data.dir}/tpch/tiny/part.tbl.bz2' overwrite into table part;
analyze table part compute statistics;
analyze table part compute statistics for columns;
diff --git a/data/files/datasets/tpch_0_001.customer/load.hive.sql b/data/files/datasets/tpch_0_001.customer/load.hive.sql
new file mode 100644
index 0000000..6e39a0d
--- /dev/null
+++ b/data/files/datasets/tpch_0_001.customer/load.hive.sql
@@ -0,0 +1,19 @@
+CREATE DATABASE IF NOT EXISTS tpch_0_001;
+
+CREATE TABLE tpch_0_001.CUSTOMER
+(
+ C_CUSTKEY INT NOT NULL,
+ C_NAME VARCHAR(25),
+ C_ADDRESS VARCHAR(40),
+ C_NATIONKEY INT NOT NULL, -- references N_NATIONKEY
+ C_PHONE CHAR(15),
+ C_ACCTBAL DECIMAL,
+ C_MKTSEGMENT CHAR(10),
+ C_COMMENT VARCHAR(117)
+)
+ ROW FORMAT DELIMITED FIELDS TERMINATED BY '|';
+
+LOAD DATA LOCAL INPATH '${hiveconf:test.data.dir}/tpch/sf0_001/customer.tbl.bz2' overwrite into table tpch_0_001.customer;
+
+analyze table tpch_0_001.customer compute statistics;
+analyze table tpch_0_001.customer compute statistics for columns;
diff --git a/data/files/datasets/lineitem/load.hive.sql b/data/files/datasets/tpch_0_001.lineitem/load.hive.sql
similarity index 72%
copy from data/files/datasets/lineitem/load.hive.sql
copy to data/files/datasets/tpch_0_001.lineitem/load.hive.sql
index ab2f5fb..ccdd20a 100644
--- a/data/files/datasets/lineitem/load.hive.sql
+++ b/data/files/datasets/tpch_0_001.lineitem/load.hive.sql
@@ -1,4 +1,6 @@
-CREATE TABLE lineitem (L_ORDERKEY INT,
+CREATE DATABASE IF NOT EXISTS tpch_0_001;
+
+CREATE TABLE tpch_0_001.lineitem (L_ORDERKEY INT,
L_PARTKEY INT,
L_SUPPKEY INT,
L_LINENUMBER INT,
@@ -17,7 +19,7 @@ CREATE TABLE lineitem (L_ORDERKEY INT,
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '|';
-LOAD DATA LOCAL INPATH '${hiveconf:test.data.dir}/lineitem.txt' OVERWRITE INTO TABLE lineitem;
+LOAD DATA LOCAL INPATH '${hiveconf:test.data.dir}/tpch/sf0_001/lineitem.tbl.bz2' OVERWRITE INTO TABLE tpch_0_001.lineitem;
-analyze table lineitem compute statistics;
-analyze table lineitem compute statistics for columns;
+analyze table tpch_0_001.lineitem compute statistics;
+analyze table tpch_0_001.lineitem compute statistics for columns;
diff --git a/data/files/datasets/tpch_0_001.nation/load.hive.sql b/data/files/datasets/tpch_0_001.nation/load.hive.sql
new file mode 100644
index 0000000..8aba96e
--- /dev/null
+++ b/data/files/datasets/tpch_0_001.nation/load.hive.sql
@@ -0,0 +1,15 @@
+CREATE DATABASE IF NOT EXISTS tpch_0_001;
+
+CREATE TABLE tpch_0_001.NATION
+(
+ N_NATIONKEY INT NOT NULL,
+ N_NAME CHAR(25),
+ N_REGIONKEY INT NOT NULL, -- references R_REGIONKEY
+ N_COMMENT VARCHAR(152)
+)
+ ROW FORMAT DELIMITED FIELDS TERMINATED BY '|';
+
+LOAD DATA LOCAL INPATH '${hiveconf:test.data.dir}/tpch/sf0_001/nation.tbl.bz2' overwrite into table tpch_0_001.nation;
+
+analyze table tpch_0_001.nation compute statistics;
+analyze table tpch_0_001.nation compute statistics for columns;
diff --git a/data/files/datasets/tpch_0_001.orders/load.hive.sql b/data/files/datasets/tpch_0_001.orders/load.hive.sql
new file mode 100644
index 0000000..4008261
--- /dev/null
+++ b/data/files/datasets/tpch_0_001.orders/load.hive.sql
@@ -0,0 +1,20 @@
+CREATE DATABASE IF NOT EXISTS tpch_0_001;
+
+CREATE TABLE tpch_0_001.ORDERS
+(
+ O_ORDERKEY INT NOT NULL,
+ O_CUSTKEY INT NOT NULL, -- references C_CUSTKEY
+ O_ORDERSTATUS CHAR(1),
+ O_TOTALPRICE DECIMAL,
+ O_ORDERDATE DATE,
+ O_ORDERPRIORITY CHAR(15),
+ O_CLERK CHAR(15),
+ O_SHIPPRIORITY INT,
+ O_COMMENT VARCHAR(79)
+)
+ ROW FORMAT DELIMITED FIELDS TERMINATED BY '|';
+
+LOAD DATA LOCAL INPATH '${hiveconf:test.data.dir}/tpch/sf0_001/orders.tbl.bz2' overwrite into table tpch_0_001.orders;
+
+analyze table tpch_0_001.orders compute statistics;
+analyze table tpch_0_001.orders compute statistics for columns;
diff --git a/data/files/datasets/tpch_0_001.part/load.hive.sql b/data/files/datasets/tpch_0_001.part/load.hive.sql
new file mode 100644
index 0000000..3807dc0
--- /dev/null
+++ b/data/files/datasets/tpch_0_001.part/load.hive.sql
@@ -0,0 +1,19 @@
+CREATE DATABASE IF NOT EXISTS tpch_0_001;
+
+CREATE TABLE tpch_0_001.part(
+ p_partkey INT NOT NULL,
+ p_name STRING,
+ p_mfgr STRING,
+ p_brand STRING,
+ p_type STRING,
+ p_size INT,
+ p_container STRING,
+ p_retailprice DOUBLE,
+ p_comment STRING
+)
+ ROW FORMAT DELIMITED FIELDS TERMINATED BY '|';
+
+LOAD DATA LOCAL INPATH '${hiveconf:test.data.dir}/tpch/sf0_001/part.tbl.bz2' overwrite into table tpch_0_001.part;
+
+analyze table tpch_0_001.part compute statistics;
+analyze table tpch_0_001.part compute statistics for columns;
diff --git a/data/files/datasets/tpch_0_001.partsupp/load.hive.sql b/data/files/datasets/tpch_0_001.partsupp/load.hive.sql
new file mode 100644
index 0000000..87b4626
--- /dev/null
+++ b/data/files/datasets/tpch_0_001.partsupp/load.hive.sql
@@ -0,0 +1,16 @@
+CREATE DATABASE IF NOT EXISTS tpch_0_001;
+
+CREATE TABLE tpch_0_001.partsupp
+(
+ PS_PARTKEY INT NOT NULL, -- references P_PARTKEY
+ PS_SUPPKEY INT NOT NULL, -- references S_SUPPKEY
+ PS_AVAILQTY INT,
+ PS_SUPPLYCOST DECIMAL,
+ PS_COMMENT VARCHAR(199)
+)
+ ROW FORMAT DELIMITED FIELDS TERMINATED BY '|';
+
+LOAD DATA LOCAL INPATH '${hiveconf:test.data.dir}/tpch/sf0_001/partsupp.tbl.bz2' overwrite into table tpch_0_001.partsupp;
+
+analyze table tpch_0_001.partsupp compute statistics;
+analyze table tpch_0_001.partsupp compute statistics for columns;
diff --git a/data/files/datasets/tpch_0_001.region/load.hive.sql b/data/files/datasets/tpch_0_001.region/load.hive.sql
new file mode 100644
index 0000000..8b2e42c
--- /dev/null
+++ b/data/files/datasets/tpch_0_001.region/load.hive.sql
@@ -0,0 +1,14 @@
+CREATE DATABASE IF NOT EXISTS tpch_0_001;
+
+CREATE TABLE tpch_0_001.REGION
+(
+ R_REGIONKEY INT NOT NULL,
+ R_NAME CHAR(25),
+ R_COMMENT VARCHAR(152)
+)
+ ROW FORMAT DELIMITED FIELDS TERMINATED BY '|';
+
+LOAD DATA LOCAL INPATH '${hiveconf:test.data.dir}/tpch/sf0_001/region.tbl.bz2' overwrite into table tpch_0_001.region;
+
+analyze table tpch_0_001.region compute statistics;
+analyze table tpch_0_001.region compute statistics for columns;
diff --git a/data/files/datasets/tpch_0_001.supplier/load.hive.sql b/data/files/datasets/tpch_0_001.supplier/load.hive.sql
new file mode 100644
index 0000000..07c0f53
--- /dev/null
+++ b/data/files/datasets/tpch_0_001.supplier/load.hive.sql
@@ -0,0 +1,18 @@
+CREATE DATABASE IF NOT EXISTS tpch_0_001;
+
+CREATE TABLE tpch_0_001.SUPPLIER
+(
+ S_SUPPKEY INT NOT NULL,
+ S_NAME CHAR(25),
+ S_ADDRESS VARCHAR(40),
+ S_NATIONKEY INT NOT NULL, -- references N_NATIONKEY
+ S_PHONE CHAR(15),
+ S_ACCTBAL DECIMAL,
+ S_COMMENT VARCHAR(101)
+)
+ ROW FORMAT DELIMITED FIELDS TERMINATED BY '|';
+
+LOAD DATA LOCAL INPATH '${hiveconf:test.data.dir}/tpch/sf0_001/supplier.tbl.bz2' overwrite into table tpch_0_001.supplier;
+
+analyze table tpch_0_001.supplier compute statistics;
+analyze table tpch_0_001.supplier compute statistics for columns;
diff --git a/data/files/lineitem.txt b/data/files/lineitem.txt
deleted file mode 100644
index 1e11d95..0000000
--- a/data/files/lineitem.txt
+++ /dev/null
@@ -1,100 +0,0 @@
-1|155190|7706|1|17|21168.23|0.04|0.02|N|O|1996-03-13|1996-02-12|1996-03-22|DELIVER IN PERSON|TRUCK|egular courts above the|
-1|67310|7311|2|36|45983.16|0.09|0.06|N|O|1996-04-12|1996-02-28|1996-04-20|TAKE BACK RETURN|MAIL|ly final dependencies: slyly bold |
-1|63700|3701|3|8|13309.60|0.10|0.02|N|O|1996-01-29|1996-03-05|1996-01-31|TAKE BACK RETURN|REG AIR|riously. regular, express dep|
-1|2132|4633|4|28|28955.64|0.09|0.06|N|O|1996-04-21|1996-03-30|1996-05-16|NONE|AIR|lites. fluffily even de|
-1|24027|1534|5|24|22824.48|0.10|0.04|N|O|1996-03-30|1996-03-14|1996-04-01|NONE|FOB| pending foxes. slyly re|
-1|15635|638|6|32|49620.16|0.07|0.02|N|O|1996-01-30|1996-02-07|1996-02-03|DELIVER IN PERSON|MAIL|arefully slyly ex|
-2|106170|1191|1|38|44694.46|0.00|0.05|N|O|1997-01-28|1997-01-14|1997-02-02|TAKE BACK RETURN|RAIL|ven requests. deposits breach a|
-3|4297|1798|1|45|54058.05|0.06|0.00|R|F|1994-02-02|1994-01-04|1994-02-23|NONE|AIR|ongside of the furiously brave acco|
-3|19036|6540|2|49|46796.47|0.10|0.00|R|F|1993-11-09|1993-12-20|1993-11-24|TAKE BACK RETURN|RAIL| unusual accounts. eve|
-3|128449|3474|3|27|39890.88|0.06|0.07|A|F|1994-01-16|1993-11-22|1994-01-23|DELIVER IN PERSON|SHIP|nal foxes wake. |
-3|29380|1883|4|2|2618.76|0.01|0.06|A|F|1993-12-04|1994-01-07|1994-01-01|NONE|TRUCK|y. fluffily pending d|
-3|183095|650|5|28|32986.52|0.04|0.00|R|F|1993-12-14|1994-01-10|1994-01-01|TAKE BACK RETURN|FOB|ages nag slyly pending|
-3|62143|9662|6|26|28733.64|0.10|0.02|A|F|1993-10-29|1993-12-18|1993-11-04|TAKE BACK RETURN|RAIL|ges sleep after the caref|
-4|88035|5560|1|30|30690.90|0.03|0.08|N|O|1996-01-10|1995-12-14|1996-01-18|DELIVER IN PERSON|REG AIR|- quickly regular packages sleep. idly|
-5|108570|8571|1|15|23678.55|0.02|0.04|R|F|1994-10-31|1994-08-31|1994-11-20|NONE|AIR|ts wake furiously |
-5|123927|3928|2|26|50723.92|0.07|0.08|R|F|1994-10-16|1994-09-25|1994-10-19|NONE|FOB|sts use slyly quickly special instruc|
-5|37531|35|3|50|73426.50|0.08|0.03|A|F|1994-08-08|1994-10-13|1994-08-26|DELIVER IN PERSON|AIR|eodolites. fluffily unusual|
-6|139636|2150|1|37|61998.31|0.08|0.03|A|F|1992-04-27|1992-05-15|1992-05-02|TAKE BACK RETURN|TRUCK|p furiously special foxes|
-7|182052|9607|1|12|13608.60|0.07|0.03|N|O|1996-05-07|1996-03-13|1996-06-03|TAKE BACK RETURN|FOB|ss pinto beans wake against th|
-7|145243|7758|2|9|11594.16|0.08|0.08|N|O|1996-02-01|1996-03-02|1996-02-19|TAKE BACK RETURN|SHIP|es. instructions|
-7|94780|9799|3|46|81639.88|0.10|0.07|N|O|1996-01-15|1996-03-27|1996-02-03|COLLECT COD|MAIL| unusual reques|
-7|163073|3074|4|28|31809.96|0.03|0.04|N|O|1996-03-21|1996-04-08|1996-04-20|NONE|FOB|. slyly special requests haggl|
-7|151894|9440|5|38|73943.82|0.08|0.01|N|O|1996-02-11|1996-02-24|1996-02-18|DELIVER IN PERSON|TRUCK|ns haggle carefully ironic deposits. bl|
-7|79251|1759|6|35|43058.75|0.06|0.03|N|O|1996-01-16|1996-02-23|1996-01-22|TAKE BACK RETURN|FOB|jole. excuses wake carefully alongside of |
-7|157238|2269|7|5|6476.15|0.04|0.02|N|O|1996-02-10|1996-03-26|1996-02-13|NONE|FOB|ithely regula|
-32|82704|7721|1|28|47227.60|0.05|0.08|N|O|1995-10-23|1995-08-27|1995-10-26|TAKE BACK RETURN|TRUCK|sleep quickly. req|
-32|197921|441|2|32|64605.44|0.02|0.00|N|O|1995-08-14|1995-10-07|1995-08-27|COLLECT COD|AIR|lithely regular deposits. fluffily |
-32|44161|6666|3|2|2210.32|0.09|0.02|N|O|1995-08-07|1995-10-07|1995-08-23|DELIVER IN PERSON|AIR| express accounts wake according to the|
-32|2743|7744|4|4|6582.96|0.09|0.03|N|O|1995-08-04|1995-10-01|1995-09-03|NONE|REG AIR|e slyly final pac|
-32|85811|8320|5|44|79059.64|0.05|0.06|N|O|1995-08-28|1995-08-20|1995-09-14|DELIVER IN PERSON|AIR|symptotes nag according to the ironic depo|
-32|11615|4117|6|6|9159.66|0.04|0.03|N|O|1995-07-21|1995-09-23|1995-07-25|COLLECT COD|RAIL| gifts cajole carefully.|
-33|61336|8855|1|31|40217.23|0.09|0.04|A|F|1993-10-29|1993-12-19|1993-11-08|COLLECT COD|TRUCK|ng to the furiously ironic package|
-33|60519|5532|2|32|47344.32|0.02|0.05|A|F|1993-12-09|1994-01-04|1993-12-28|COLLECT COD|MAIL|gular theodolites|
-33|137469|9983|3|5|7532.30|0.05|0.03|A|F|1993-12-09|1993-12-25|1993-12-23|TAKE BACK RETURN|AIR|. stealthily bold exc|
-33|33918|3919|4|41|75928.31|0.09|0.00|R|F|1993-11-09|1994-01-24|1993-11-11|TAKE BACK RETURN|MAIL|unusual packages doubt caref|
-34|88362|871|1|13|17554.68|0.00|0.07|N|O|1998-10-23|1998-09-14|1998-11-06|NONE|REG AIR|nic accounts. deposits are alon|
-34|89414|1923|2|22|30875.02|0.08|0.06|N|O|1998-10-09|1998-10-16|1998-10-12|NONE|FOB|thely slyly p|
-34|169544|4577|3|6|9681.24|0.02|0.06|N|O|1998-10-30|1998-09-20|1998-11-05|NONE|FOB|ar foxes sleep |
-35|450|2951|1|24|32410.80|0.02|0.00|N|O|1996-02-21|1996-01-03|1996-03-18|TAKE BACK RETURN|FOB|, regular tithe|
-35|161940|4457|2|34|68065.96|0.06|0.08|N|O|1996-01-22|1996-01-06|1996-01-27|DELIVER IN PERSON|RAIL|s are carefully against the f|
-35|120896|8433|3|7|13418.23|0.06|0.04|N|O|1996-01-19|1995-12-22|1996-01-29|NONE|MAIL| the carefully regular |
-35|85175|7684|4|25|29004.25|0.06|0.05|N|O|1995-11-26|1995-12-25|1995-12-21|DELIVER IN PERSON|SHIP| quickly unti|
-35|119917|4940|5|34|65854.94|0.08|0.06|N|O|1995-11-08|1996-01-15|1995-11-26|COLLECT COD|MAIL|. silent, unusual deposits boost|
-35|30762|3266|6|28|47397.28|0.03|0.02|N|O|1996-02-01|1995-12-24|1996-02-28|COLLECT COD|RAIL|ly alongside of |
-36|119767|9768|1|42|75043.92|0.09|0.00|N|O|1996-02-03|1996-01-21|1996-02-23|COLLECT COD|SHIP| careful courts. special |
-37|22630|5133|1|40|62105.20|0.09|0.03|A|F|1992-07-21|1992-08-01|1992-08-15|NONE|REG AIR|luffily regular requests. slyly final acco|
-37|126782|1807|2|39|70542.42|0.05|0.02|A|F|1992-07-02|1992-08-18|1992-07-28|TAKE BACK RETURN|RAIL|the final requests. ca|
-37|12903|5405|3|43|78083.70|0.05|0.08|A|F|1992-07-10|1992-07-06|1992-08-02|DELIVER IN PERSON|TRUCK|iously ste|
-38|175839|874|1|44|84252.52|0.04|0.02|N|O|1996-09-29|1996-11-17|1996-09-30|COLLECT COD|MAIL|s. blithely unusual theodolites am|
-39|2320|9821|1|44|53782.08|0.09|0.06|N|O|1996-11-14|1996-12-15|1996-12-12|COLLECT COD|RAIL|eodolites. careful|
-39|186582|4137|2|26|43383.08|0.08|0.04|N|O|1996-11-04|1996-10-20|1996-11-20|NONE|FOB|ckages across the slyly silent|
-39|67831|5350|3|46|82746.18|0.06|0.08|N|O|1996-09-26|1996-12-19|1996-10-26|DELIVER IN PERSON|AIR|he carefully e|
-39|20590|3093|4|32|48338.88|0.07|0.05|N|O|1996-10-02|1996-12-19|1996-10-14|COLLECT COD|MAIL|heodolites sleep silently pending foxes. ac|
-39|54519|9530|5|43|63360.93|0.01|0.01|N|O|1996-10-17|1996-11-14|1996-10-26|COLLECT COD|MAIL|yly regular i|
-39|94368|6878|6|40|54494.40|0.06|0.05|N|O|1996-12-08|1996-10-22|1997-01-01|COLLECT COD|AIR|quickly ironic fox|
-64|85951|5952|1|21|40675.95|0.05|0.02|R|F|1994-09-30|1994-09-18|1994-10-26|DELIVER IN PERSON|REG AIR|ch slyly final, thin platelets.|
-65|59694|4705|1|26|42995.94|0.03|0.03|A|F|1995-04-20|1995-04-25|1995-05-13|NONE|TRUCK|pending deposits nag even packages. ca|
-65|73815|8830|2|22|39353.82|0.00|0.05|N|O|1995-07-17|1995-06-04|1995-07-19|COLLECT COD|FOB| ideas. special, r|
-65|1388|3889|3|21|27076.98|0.09|0.07|N|O|1995-07-06|1995-05-14|1995-07-31|DELIVER IN PERSON|RAIL|bove the even packages. accounts nag carefu|
-66|115118|7630|1|31|35126.41|0.00|0.08|R|F|1994-02-19|1994-03-11|1994-02-20|TAKE BACK RETURN|RAIL|ut the unusual accounts sleep at the bo|
-66|173489|3490|2|41|64061.68|0.04|0.07|A|F|1994-02-21|1994-03-01|1994-03-18|COLLECT COD|AIR| regular de|
-67|21636|9143|1|4|6230.52|0.09|0.04|N|O|1997-04-17|1997-01-31|1997-04-20|NONE|SHIP| cajole thinly expres|
-67|20193|5198|2|12|13358.28|0.09|0.05|N|O|1997-01-27|1997-02-21|1997-02-22|NONE|REG AIR| even packages cajole|
-67|173600|6118|3|5|8368.00|0.03|0.07|N|O|1997-02-20|1997-02-12|1997-02-21|DELIVER IN PERSON|TRUCK|y unusual packages thrash pinto |
-67|87514|7515|4|44|66066.44|0.08|0.06|N|O|1997-03-18|1997-01-29|1997-04-13|DELIVER IN PERSON|RAIL|se quickly above the even, express reques|
-67|40613|8126|5|23|35733.03|0.05|0.07|N|O|1997-04-19|1997-02-14|1997-05-06|DELIVER IN PERSON|REG AIR|ly regular deposit|
-67|178306|824|6|29|40144.70|0.02|0.05|N|O|1997-01-25|1997-01-27|1997-01-27|DELIVER IN PERSON|FOB|ultipliers |
-68|7068|9569|1|3|2925.18|0.05|0.02|N|O|1998-07-04|1998-06-05|1998-07-21|NONE|RAIL|fully special instructions cajole. furious|
-68|175180|2732|2|46|57738.28|0.02|0.05|N|O|1998-06-26|1998-06-07|1998-07-05|NONE|MAIL| requests are unusual, regular pinto |
-68|34980|7484|3|46|88089.08|0.04|0.05|N|O|1998-08-13|1998-07-08|1998-08-29|NONE|RAIL|egular dependencies affix ironically along |
-68|94728|2256|4|20|34454.40|0.07|0.01|N|O|1998-06-27|1998-05-23|1998-07-02|NONE|REG AIR| excuses integrate fluffily |
-68|82758|5267|5|27|47000.25|0.03|0.06|N|O|1998-06-19|1998-06-25|1998-06-29|DELIVER IN PERSON|SHIP|ccounts. deposits use. furiously|
-68|102561|5072|6|30|46906.80|0.05|0.06|N|O|1998-08-11|1998-07-11|1998-08-14|NONE|RAIL|oxes are slyly blithely fin|
-68|139247|1761|7|41|52735.84|0.09|0.08|N|O|1998-06-24|1998-06-27|1998-07-06|NONE|SHIP|eposits nag special ideas. furiousl|
-69|115209|7721|1|48|58761.60|0.01|0.07|A|F|1994-08-17|1994-08-11|1994-09-08|NONE|TRUCK|regular epitaphs. carefully even ideas hag|
-69|104180|9201|2|32|37893.76|0.08|0.06|A|F|1994-08-24|1994-08-17|1994-08-31|NONE|REG AIR|s sleep carefully bold, |
-69|137267|4807|3|17|22172.42|0.09|0.00|A|F|1994-07-02|1994-07-07|1994-07-03|TAKE BACK RETURN|AIR|final, pending instr|
-69|37502|2509|4|3|4318.50|0.09|0.04|R|F|1994-06-06|1994-07-27|1994-06-15|NONE|MAIL| blithely final d|
-69|92070|7089|5|42|44606.94|0.07|0.04|R|F|1994-07-31|1994-07-26|1994-08-28|DELIVER IN PERSON|REG AIR|tect regular, speci|
-69|18504|1006|6|23|32717.50|0.05|0.00|A|F|1994-10-03|1994-08-06|1994-10-24|NONE|SHIP|nding accounts ca|
-70|64128|9141|1|8|8736.96|0.03|0.08|R|F|1994-01-12|1994-02-27|1994-01-14|TAKE BACK RETURN|FOB|ggle. carefully pending dependenc|
-70|196156|1195|2|13|16277.95|0.06|0.06|A|F|1994-03-03|1994-02-13|1994-03-26|COLLECT COD|AIR|lyly special packag|
-70|179809|7361|3|1|1888.80|0.03|0.05|R|F|1994-01-26|1994-03-05|1994-01-28|TAKE BACK RETURN|RAIL|quickly. fluffily unusual theodolites c|
-70|45734|743|4|11|18477.03|0.01|0.05|A|F|1994-03-17|1994-03-17|1994-03-27|NONE|MAIL|alongside of the deposits. fur|
-70|37131|2138|5|37|39520.81|0.09|0.04|R|F|1994-02-13|1994-03-16|1994-02-21|COLLECT COD|MAIL|n accounts are. q|
-70|55655|3171|6|19|30602.35|0.06|0.03|A|F|1994-01-26|1994-02-17|1994-02-06|TAKE BACK RETURN|SHIP| packages wake pending accounts.|
-71|61931|1932|1|25|47323.25|0.09|0.07|N|O|1998-04-10|1998-04-22|1998-04-11|COLLECT COD|FOB|ckly. slyly|
-71|65916|3435|2|3|5645.73|0.09|0.07|N|O|1998-05-23|1998-04-03|1998-06-02|COLLECT COD|SHIP|y. pinto beans haggle after the|
-71|34432|1942|3|45|61489.35|0.00|0.07|N|O|1998-02-23|1998-03-20|1998-03-24|DELIVER IN PERSON|SHIP| ironic packages believe blithely a|
-71|96645|9155|4|33|54174.12|0.00|0.01|N|O|1998-04-12|1998-03-20|1998-04-15|NONE|FOB| serve quickly fluffily bold deposi|
-71|103255|5766|5|39|49071.75|0.08|0.06|N|O|1998-01-29|1998-04-07|1998-02-18|DELIVER IN PERSON|RAIL|l accounts sleep across the pack|
-71|195635|674|6|34|58841.42|0.04|0.01|N|O|1998-03-05|1998-04-22|1998-03-30|DELIVER IN PERSON|TRUCK|s cajole. |
-96|123076|613|1|23|25278.61|0.10|0.06|A|F|1994-07-19|1994-06-29|1994-07-25|DELIVER IN PERSON|TRUCK|ep-- carefully reg|
-96|135390|5391|2|30|42761.70|0.01|0.06|R|F|1994-06-03|1994-05-29|1994-06-22|DELIVER IN PERSON|TRUCK|e quickly even ideas. furiou|
-97|119477|1989|1|13|19454.11|0.00|0.02|R|F|1993-04-01|1993-04-04|1993-04-08|NONE|TRUCK|ayers cajole against the furiously|
-97|49568|2073|2|37|56149.72|0.02|0.06|A|F|1993-04-13|1993-03-30|1993-04-14|DELIVER IN PERSON|SHIP|ic requests boost carefully quic|
-97|77699|5221|3|19|31857.11|0.06|0.08|R|F|1993-05-14|1993-03-05|1993-05-25|TAKE BACK RETURN|RAIL|gifts. furiously ironic packages cajole. |
-98|40216|217|1|28|32373.88|0.06|0.07|A|F|1994-12-24|1994-10-25|1995-01-16|COLLECT COD|REG AIR| pending, regular accounts s|
-98|109743|7274|2|1|1752.74|0.00|0.00|A|F|1994-12-01|1994-12-12|1994-12-15|DELIVER IN PERSON|TRUCK|. unusual instructions against|
-98|44706|4707|3|14|23109.80|0.05|0.02|A|F|1994-12-30|1994-11-22|1995-01-27|COLLECT COD|AIR| cajole furiously. blithely ironic ideas |
diff --git a/data/files/part_tiny.txt b/data/files/part_tiny.txt
deleted file mode 100644
index f6973fd..0000000
--- a/data/files/part_tiny.txt
+++ /dev/null
@@ -1,26 +0,0 @@
-121152almond antique burnished rose metallicManufacturer#1Brand#14PROMO PLATED TIN2JUMBO BOX1173.15e pinto beans h
-121152almond antique burnished rose metallicManufacturer#1Brand#14PROMO PLATED TIN2JUMBO BOX1173.15e pinto beans h
-85768almond antique chartreuse lavender yellowManufacturer#1Brand#12LARGE BRUSHED STEEL34SM BAG1753.76refull
-110592almond antique salmon chartreuse burlywoodManufacturer#1Brand#15PROMO BURNISHED NICKEL6JUMBO PKG1602.59 to the furiously
-86428almond aquamarine burnished black steelManufacturer#1Brand#12STANDARD ANODIZED STEEL28WRAP BAG1414.42arefully
-65667almond aquamarine pink moccasin thistleManufacturer#1Brand#12LARGE BURNISHED STEEL42JUMBO CASE1632.66e across the expr
-105685almond antique violet chocolate turquoiseManufacturer#2Brand#22MEDIUM ANODIZED COPPER14MED CAN1690.68ly pending requ
-191709almond antique violet turquoise frostedManufacturer#2Brand#22ECONOMY POLISHED STEEL40MED BOX1800.7 haggle
-146985almond aquamarine midnight light salmonManufacturer#2Brand#23MEDIUM BURNISHED COPPER2SM CASE2031.98s cajole caref
-132666almond aquamarine rose maroon antiqueManufacturer#2Brand#24SMALL POLISHED NICKEL25MED BOX1698.66even
-195606almond aquamarine sandy cyan gainsboroManufacturer#2Brand#25STANDARD PLATED TIN18SM PKG1701.6ic de
-90681almond antique chartreuse khaki whiteManufacturer#3Brand#31MEDIUM BURNISHED TIN17SM CASE1671.68are slyly after the sl
-17273almond antique forest lavender goldenrodManufacturer#3Brand#35PROMO ANODIZED TIN14JUMBO CASE1190.27along the
-112398almond antique metallic orange dimManufacturer#3Brand#32MEDIUM BURNISHED BRASS19JUMBO JAR1410.39ole car
-40982almond antique misty red oliveManufacturer#3Brand#32ECONOMY PLATED COPPER1LG PKG1922.98c foxes can s
-144293almond antique olive coral navajoManufacturer#3Brand#34STANDARD POLISHED STEEL45JUMBO CAN1337.29ag furiously about
-49671almond antique gainsboro frosted violetManufacturer#4Brand#41SMALL BRUSHED BRASS10SM BOX1620.67ccounts run quick
-48427almond antique violet mint lemonManufacturer#4Brand#42PROMO POLISHED STEEL39SM CASE1375.42hely ironic i
-45261almond aquamarine floral ivory bisqueManufacturer#4Brand#42SMALL PLATED STEEL27WRAP CASE1206.26careful
-17927almond aquamarine yellow dodger mintManufacturer#4Brand#41ECONOMY BRUSHED COPPER7SM PKG1844.92ites. eve
-33357almond azure aquamarine papaya violetManufacturer#4Brand#41STANDARD ANODIZED TIN12WRAP CASE1290.35reful
-192697almond antique blue firebrick mintManufacturer#5Brand#52MEDIUM BURNISHED TIN31LG DRUM1789.69ickly ir
-42669almond antique medium spring khakiManufacturer#5Brand#51STANDARD BURNISHED TIN6MED CAN1611.66sits haggl
-155733almond antique sky peru orangeManufacturer#5Brand#53SMALL PLATED BRASS2WRAP DRUM1788.73furiously. bra
-15103almond aquamarine dodger light gainsboroManufacturer#5Brand#53ECONOMY BURNISHED STEEL46LG PACK1018.1packages hinder carefu
-78486almond azure blanched chiffon midnightManufacturer#5Brand#52LARGE BRUSHED BRASS23MED BAG1464.48hely blith
diff --git a/data/files/tpch/sf0_001/customer.tbl.bz2 b/data/files/tpch/sf0_001/customer.tbl.bz2
new file mode 100644
index 0000000..43c04d2
Binary files /dev/null and b/data/files/tpch/sf0_001/customer.tbl.bz2 differ
diff --git a/data/files/tpch/sf0_001/lineitem.tbl.bz2 b/data/files/tpch/sf0_001/lineitem.tbl.bz2
new file mode 100644
index 0000000..4686356
Binary files /dev/null and b/data/files/tpch/sf0_001/lineitem.tbl.bz2 differ
diff --git a/data/files/tpch/sf0_001/nation.tbl.bz2 b/data/files/tpch/sf0_001/nation.tbl.bz2
new file mode 100644
index 0000000..f930c05
Binary files /dev/null and b/data/files/tpch/sf0_001/nation.tbl.bz2 differ
diff --git a/data/files/tpch/sf0_001/orders.tbl.bz2 b/data/files/tpch/sf0_001/orders.tbl.bz2
new file mode 100644
index 0000000..fdcd226
Binary files /dev/null and b/data/files/tpch/sf0_001/orders.tbl.bz2 differ
diff --git a/data/files/tpch/sf0_001/part.tbl.bz2 b/data/files/tpch/sf0_001/part.tbl.bz2
new file mode 100644
index 0000000..f69b6df
Binary files /dev/null and b/data/files/tpch/sf0_001/part.tbl.bz2 differ
diff --git a/data/files/tpch/sf0_001/partsupp.tbl.bz2 b/data/files/tpch/sf0_001/partsupp.tbl.bz2
new file mode 100644
index 0000000..53363d3
Binary files /dev/null and b/data/files/tpch/sf0_001/partsupp.tbl.bz2 differ
diff --git a/data/files/tpch/sf0_001/region.tbl.bz2 b/data/files/tpch/sf0_001/region.tbl.bz2
new file mode 100644
index 0000000..6985fa0
Binary files /dev/null and b/data/files/tpch/sf0_001/region.tbl.bz2 differ
diff --git a/data/files/tpch/sf0_001/supplier.tbl.bz2 b/data/files/tpch/sf0_001/supplier.tbl.bz2
new file mode 100644
index 0000000..dd5d06b
Binary files /dev/null and b/data/files/tpch/sf0_001/supplier.tbl.bz2 differ
diff --git a/data/files/tpch/tiny/lineitem.tbl.bz2 b/data/files/tpch/tiny/lineitem.tbl.bz2
new file mode 100644
index 0000000..95fca19
Binary files /dev/null and b/data/files/tpch/tiny/lineitem.tbl.bz2 differ
diff --git a/data/files/tpch/tiny/part.tbl.bz2 b/data/files/tpch/tiny/part.tbl.bz2
new file mode 100644
index 0000000..4fd5f26
Binary files /dev/null and b/data/files/tpch/tiny/part.tbl.bz2 differ
diff --git a/itests/hive-blobstore/src/test/queries/clientpositive/ptf_general_queries.q b/itests/hive-blobstore/src/test/queries/clientpositive/ptf_general_queries.q
index 5ad65b1..2661446 100644
--- a/itests/hive-blobstore/src/test/queries/clientpositive/ptf_general_queries.q
+++ b/itests/hive-blobstore/src/test/queries/clientpositive/ptf_general_queries.q
@@ -13,7 +13,7 @@ CREATE TABLE part_tiny(
)
LOCATION '${hiveconf:test.blobstore.path.unique}/ptf_general_queries/part_tiny';
-LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' INTO TABLE part_tiny;
+LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' INTO TABLE part_tiny;
-- Test DISTRIBUTE BY without any aggregate function
SELECT p_mfgr, p_name, p_size
diff --git a/itests/hive-blobstore/src/test/results/clientpositive/ptf_general_queries.q.out b/itests/hive-blobstore/src/test/results/clientpositive/ptf_general_queries.q.out
index 5e10716..a73c3ce 100644
--- a/itests/hive-blobstore/src/test/results/clientpositive/ptf_general_queries.q.out
+++ b/itests/hive-blobstore/src/test/results/clientpositive/ptf_general_queries.q.out
@@ -34,11 +34,11 @@ POSTHOOK: type: CREATETABLE
POSTHOOK: Input: ### test.blobstore.path ###/ptf_general_queries/part_tiny
POSTHOOK: Output: database:default
POSTHOOK: Output: default@part_tiny
-PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' INTO TABLE part_tiny
+PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' INTO TABLE part_tiny
PREHOOK: type: LOAD
#### A masked pattern was here ####
PREHOOK: Output: default@part_tiny
-POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' INTO TABLE part_tiny
+POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' INTO TABLE part_tiny
POSTHOOK: type: LOAD
#### A masked pattern was here ####
POSTHOOK: Output: default@part_tiny
diff --git a/ql/src/test/queries/clientpositive/special_character_in_tabnames_1.q b/ql/src/test/queries/clientpositive/special_character_in_tabnames_1.q
index ed998b7..3abd152 100644
--- a/ql/src/test/queries/clientpositive/special_character_in_tabnames_1.q
+++ b/ql/src/test/queries/clientpositive/special_character_in_tabnames_1.q
@@ -31,7 +31,7 @@ CREATE TABLE `p/a/r/t`(
p_comment STRING
);
-LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table `p/a/r/t`;
+LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table `p/a/r/t`;
CREATE TABLE `line/item` (L_ORDERKEY INT,
L_PARTKEY INT,
@@ -52,7 +52,7 @@ CREATE TABLE `line/item` (L_ORDERKEY INT,
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '|';
-LOAD DATA LOCAL INPATH '../../data/files/lineitem.txt' OVERWRITE INTO TABLE `line/item`;
+LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/lineitem.tbl.bz2' OVERWRITE INTO TABLE `line/item`;
create table `src/_/cbo` as select * from default.src;
diff --git a/ql/src/test/queries/clientpositive/special_character_in_tabnames_quotes_1.q b/ql/src/test/queries/clientpositive/special_character_in_tabnames_quotes_1.q
index 85d714e..df5cec8 100644
--- a/ql/src/test/queries/clientpositive/special_character_in_tabnames_quotes_1.q
+++ b/ql/src/test/queries/clientpositive/special_character_in_tabnames_quotes_1.q
@@ -32,7 +32,7 @@ CREATE TABLE "p/a/r/t"(
p_comment STRING
);
-LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table "p/a/r/t";
+LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table "p/a/r/t";
CREATE TABLE "line/item" (L_ORDERKEY INT,
L_PARTKEY INT,
@@ -53,7 +53,7 @@ CREATE TABLE "line/item" (L_ORDERKEY INT,
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '|';
-LOAD DATA LOCAL INPATH '../../data/files/lineitem.txt' OVERWRITE INTO TABLE "line/item";
+LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/lineitem.tbl.bz2' OVERWRITE INTO TABLE "line/item";
create table "src/_/cbo" as select * from default.src;
diff --git a/ql/src/test/queries/clientpositive/subquery_in_having.q b/ql/src/test/queries/clientpositive/subquery_in_having.q
index 9e72184..2861c4e 100644
--- a/ql/src/test/queries/clientpositive/subquery_in_having.q
+++ b/ql/src/test/queries/clientpositive/subquery_in_having.q
@@ -17,7 +17,7 @@ CREATE TABLE part_subq(
p_comment STRING
);
-LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table part_subq;
+LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table part_subq;
-- non agg, non corr
explain
diff --git a/ql/src/test/queries/clientpositive/tpch_sf0_001_datasets.q b/ql/src/test/queries/clientpositive/tpch_sf0_001_datasets.q
new file mode 100644
index 0000000..5625c4a
--- /dev/null
+++ b/ql/src/test/queries/clientpositive/tpch_sf0_001_datasets.q
@@ -0,0 +1,19 @@
+--! qt:dataset:tpch_0_001.customer
+--! qt:dataset:tpch_0_001.lineitem
+--! qt:dataset:tpch_0_001.nation
+--! qt:dataset:tpch_0_001.orders
+--! qt:dataset:tpch_0_001.part
+--! qt:dataset:tpch_0_001.partsupp
+--! qt:dataset:tpch_0_001.region
+--! qt:dataset:tpch_0_001.supplier
+
+USE tpch_0_001;
+
+SELECT COUNT(1) FROM customer;
+SELECT COUNT(1) FROM lineitem;
+SELECT COUNT(1) FROM nation;
+SELECT COUNT(1) FROM orders;
+SELECT COUNT(1) FROM part;
+SELECT COUNT(1) FROM partsupp;
+SELECT COUNT(1) FROM region;
+SELECT COUNT(1) FROM supplier;
\ No newline at end of file
diff --git a/ql/src/test/queries/clientpositive/tpch_tiny_datasets.q b/ql/src/test/queries/clientpositive/tpch_tiny_datasets.q
new file mode 100644
index 0000000..23e87ee
--- /dev/null
+++ b/ql/src/test/queries/clientpositive/tpch_tiny_datasets.q
@@ -0,0 +1,5 @@
+--! qt:dataset:lineitem
+--! qt:dataset:part
+
+SELECT COUNT(1) FROM lineitem;
+SELECT COUNT(1) FROM part;
\ No newline at end of file
diff --git a/ql/src/test/queries/clientpositive/vector_case_when_1.q b/ql/src/test/queries/clientpositive/vector_case_when_1.q
index d027524..5004fae 100644
--- a/ql/src/test/queries/clientpositive/vector_case_when_1.q
+++ b/ql/src/test/queries/clientpositive/vector_case_when_1.q
@@ -25,7 +25,7 @@ CREATE TABLE lineitem_test_txt (L_ORDERKEY INT,
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '|';
-LOAD DATA LOCAL INPATH '../../data/files/lineitem.txt' OVERWRITE INTO TABLE lineitem_test_txt;
+LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/lineitem.tbl.bz2' OVERWRITE INTO TABLE lineitem_test_txt;
CREATE TABLE lineitem_test STORED AS ORC AS SELECT * FROM lineitem_test_txt;
INSERT INTO TABLE lineitem_test VALUES (NULL,NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL);
diff --git a/ql/src/test/queries/clientpositive/vectorized_ptf.q b/ql/src/test/queries/clientpositive/vectorized_ptf.q
index c5a7ddf..1a78bf0 100644
--- a/ql/src/test/queries/clientpositive/vectorized_ptf.q
+++ b/ql/src/test/queries/clientpositive/vectorized_ptf.q
@@ -24,7 +24,7 @@ CREATE TABLE part_staging(
p_comment STRING
);
-LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table part_staging;
+LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table part_staging;
CREATE TABLE part_orc(
p_partkey INT,
diff --git a/ql/src/test/results/clientpositive/llap/special_character_in_tabnames_1.q.out b/ql/src/test/results/clientpositive/llap/special_character_in_tabnames_1.q.out
index 68f14bb..23b4752 100644
--- a/ql/src/test/results/clientpositive/llap/special_character_in_tabnames_1.q.out
+++ b/ql/src/test/results/clientpositive/llap/special_character_in_tabnames_1.q.out
@@ -88,11 +88,11 @@ POSTHOOK: query: CREATE TABLE `p/a/r/t`(
POSTHOOK: type: CREATETABLE
POSTHOOK: Output: database:db~!@#$%^&*(),<>
POSTHOOK: Output: db~!@#$%^&*(),<>@p/a/r/t
-PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table `p/a/r/t`
+PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table `p/a/r/t`
PREHOOK: type: LOAD
#### A masked pattern was here ####
PREHOOK: Output: db~!@#$%^&*(),<>@p/a/r/t
-POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table `p/a/r/t`
+POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table `p/a/r/t`
POSTHOOK: type: LOAD
#### A masked pattern was here ####
POSTHOOK: Output: db~!@#$%^&*(),<>@p/a/r/t
@@ -138,11 +138,11 @@ FIELDS TERMINATED BY '|'
POSTHOOK: type: CREATETABLE
POSTHOOK: Output: database:db~!@#$%^&*(),<>
POSTHOOK: Output: db~!@#$%^&*(),<>@line/item
-PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/lineitem.txt' OVERWRITE INTO TABLE `line/item`
+PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/lineitem.tbl.bz2' OVERWRITE INTO TABLE `line/item`
PREHOOK: type: LOAD
#### A masked pattern was here ####
PREHOOK: Output: db~!@#$%^&*(),<>@line/item
-POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/lineitem.txt' OVERWRITE INTO TABLE `line/item`
+POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/lineitem.tbl.bz2' OVERWRITE INTO TABLE `line/item`
POSTHOOK: type: LOAD
#### A masked pattern was here ####
POSTHOOK: Output: db~!@#$%^&*(),<>@line/item
diff --git a/ql/src/test/results/clientpositive/llap/special_character_in_tabnames_quotes_1.q.out b/ql/src/test/results/clientpositive/llap/special_character_in_tabnames_quotes_1.q.out
index 1be2e32..042294c 100644
--- a/ql/src/test/results/clientpositive/llap/special_character_in_tabnames_quotes_1.q.out
+++ b/ql/src/test/results/clientpositive/llap/special_character_in_tabnames_quotes_1.q.out
@@ -88,11 +88,11 @@ POSTHOOK: query: CREATE TABLE "p/a/r/t"(
POSTHOOK: type: CREATETABLE
POSTHOOK: Output: database:db~!@@#$%^&*(),<>
POSTHOOK: Output: db~!@@#$%^&*(),<>@p/a/r/t
-PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table "p/a/r/t"
+PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table "p/a/r/t"
PREHOOK: type: LOAD
#### A masked pattern was here ####
PREHOOK: Output: db~!@@#$%^&*(),<>@p/a/r/t
-POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table "p/a/r/t"
+POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table "p/a/r/t"
POSTHOOK: type: LOAD
#### A masked pattern was here ####
POSTHOOK: Output: db~!@@#$%^&*(),<>@p/a/r/t
@@ -138,11 +138,11 @@ FIELDS TERMINATED BY '|'
POSTHOOK: type: CREATETABLE
POSTHOOK: Output: database:db~!@@#$%^&*(),<>
POSTHOOK: Output: db~!@@#$%^&*(),<>@line/item
-PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/lineitem.txt' OVERWRITE INTO TABLE "line/item"
+PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/lineitem.tbl.bz2' OVERWRITE INTO TABLE "line/item"
PREHOOK: type: LOAD
#### A masked pattern was here ####
PREHOOK: Output: db~!@@#$%^&*(),<>@line/item
-POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/lineitem.txt' OVERWRITE INTO TABLE "line/item"
+POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/lineitem.tbl.bz2' OVERWRITE INTO TABLE "line/item"
POSTHOOK: type: LOAD
#### A masked pattern was here ####
POSTHOOK: Output: db~!@@#$%^&*(),<>@line/item
diff --git a/ql/src/test/results/clientpositive/llap/subquery_in_having.q.out b/ql/src/test/results/clientpositive/llap/subquery_in_having.q.out
index 4164d1b..b63c6c3 100644
--- a/ql/src/test/results/clientpositive/llap/subquery_in_having.q.out
+++ b/ql/src/test/results/clientpositive/llap/subquery_in_having.q.out
@@ -30,11 +30,11 @@ POSTHOOK: query: CREATE TABLE part_subq(
POSTHOOK: type: CREATETABLE
POSTHOOK: Output: database:default
POSTHOOK: Output: default@part_subq
-PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table part_subq
+PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table part_subq
PREHOOK: type: LOAD
#### A masked pattern was here ####
PREHOOK: Output: default@part_subq
-POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table part_subq
+POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table part_subq
POSTHOOK: type: LOAD
#### A masked pattern was here ####
POSTHOOK: Output: default@part_subq
diff --git a/ql/src/test/results/clientpositive/llap/tpch_sf0_001_datasets.q.out b/ql/src/test/results/clientpositive/llap/tpch_sf0_001_datasets.q.out
new file mode 100644
index 0000000..4efbd14
--- /dev/null
+++ b/ql/src/test/results/clientpositive/llap/tpch_sf0_001_datasets.q.out
@@ -0,0 +1,78 @@
+PREHOOK: query: USE tpch_0_001
+PREHOOK: type: SWITCHDATABASE
+PREHOOK: Input: database:tpch_0_001
+POSTHOOK: query: USE tpch_0_001
+POSTHOOK: type: SWITCHDATABASE
+POSTHOOK: Input: database:tpch_0_001
+PREHOOK: query: SELECT COUNT(1) FROM customer
+PREHOOK: type: QUERY
+PREHOOK: Input: tpch_0_001@customer
+#### A masked pattern was here ####
+POSTHOOK: query: SELECT COUNT(1) FROM customer
+POSTHOOK: type: QUERY
+POSTHOOK: Input: tpch_0_001@customer
+#### A masked pattern was here ####
+150
+PREHOOK: query: SELECT COUNT(1) FROM lineitem
+PREHOOK: type: QUERY
+PREHOOK: Input: tpch_0_001@lineitem
+#### A masked pattern was here ####
+POSTHOOK: query: SELECT COUNT(1) FROM lineitem
+POSTHOOK: type: QUERY
+POSTHOOK: Input: tpch_0_001@lineitem
+#### A masked pattern was here ####
+6005
+PREHOOK: query: SELECT COUNT(1) FROM nation
+PREHOOK: type: QUERY
+PREHOOK: Input: tpch_0_001@nation
+#### A masked pattern was here ####
+POSTHOOK: query: SELECT COUNT(1) FROM nation
+POSTHOOK: type: QUERY
+POSTHOOK: Input: tpch_0_001@nation
+#### A masked pattern was here ####
+25
+PREHOOK: query: SELECT COUNT(1) FROM orders
+PREHOOK: type: QUERY
+PREHOOK: Input: tpch_0_001@orders
+#### A masked pattern was here ####
+POSTHOOK: query: SELECT COUNT(1) FROM orders
+POSTHOOK: type: QUERY
+POSTHOOK: Input: tpch_0_001@orders
+#### A masked pattern was here ####
+1500
+PREHOOK: query: SELECT COUNT(1) FROM part
+PREHOOK: type: QUERY
+PREHOOK: Input: tpch_0_001@part
+#### A masked pattern was here ####
+POSTHOOK: query: SELECT COUNT(1) FROM part
+POSTHOOK: type: QUERY
+POSTHOOK: Input: tpch_0_001@part
+#### A masked pattern was here ####
+200
+PREHOOK: query: SELECT COUNT(1) FROM partsupp
+PREHOOK: type: QUERY
+PREHOOK: Input: tpch_0_001@partsupp
+#### A masked pattern was here ####
+POSTHOOK: query: SELECT COUNT(1) FROM partsupp
+POSTHOOK: type: QUERY
+POSTHOOK: Input: tpch_0_001@partsupp
+#### A masked pattern was here ####
+800
+PREHOOK: query: SELECT COUNT(1) FROM region
+PREHOOK: type: QUERY
+PREHOOK: Input: tpch_0_001@region
+#### A masked pattern was here ####
+POSTHOOK: query: SELECT COUNT(1) FROM region
+POSTHOOK: type: QUERY
+POSTHOOK: Input: tpch_0_001@region
+#### A masked pattern was here ####
+5
+PREHOOK: query: SELECT COUNT(1) FROM supplier
+PREHOOK: type: QUERY
+PREHOOK: Input: tpch_0_001@supplier
+#### A masked pattern was here ####
+POSTHOOK: query: SELECT COUNT(1) FROM supplier
+POSTHOOK: type: QUERY
+POSTHOOK: Input: tpch_0_001@supplier
+#### A masked pattern was here ####
+10
diff --git a/ql/src/test/results/clientpositive/llap/tpch_tiny_datasets.q.out b/ql/src/test/results/clientpositive/llap/tpch_tiny_datasets.q.out
new file mode 100644
index 0000000..548bb29
--- /dev/null
+++ b/ql/src/test/results/clientpositive/llap/tpch_tiny_datasets.q.out
@@ -0,0 +1,18 @@
+PREHOOK: query: SELECT COUNT(1) FROM lineitem
+PREHOOK: type: QUERY
+PREHOOK: Input: default@lineitem
+#### A masked pattern was here ####
+POSTHOOK: query: SELECT COUNT(1) FROM lineitem
+POSTHOOK: type: QUERY
+POSTHOOK: Input: default@lineitem
+#### A masked pattern was here ####
+100
+PREHOOK: query: SELECT COUNT(1) FROM part
+PREHOOK: type: QUERY
+PREHOOK: Input: default@part
+#### A masked pattern was here ####
+POSTHOOK: query: SELECT COUNT(1) FROM part
+POSTHOOK: type: QUERY
+POSTHOOK: Input: default@part
+#### A masked pattern was here ####
+26
diff --git a/ql/src/test/results/clientpositive/llap/vector_case_when_1.q.out b/ql/src/test/results/clientpositive/llap/vector_case_when_1.q.out
index 7adb53b..0bbc643 100644
--- a/ql/src/test/results/clientpositive/llap/vector_case_when_1.q.out
+++ b/ql/src/test/results/clientpositive/llap/vector_case_when_1.q.out
@@ -40,11 +40,11 @@ FIELDS TERMINATED BY '|'
POSTHOOK: type: CREATETABLE
POSTHOOK: Output: database:default
POSTHOOK: Output: default@lineitem_test_txt
-PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/lineitem.txt' OVERWRITE INTO TABLE lineitem_test_txt
+PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/lineitem.tbl.bz2' OVERWRITE INTO TABLE lineitem_test_txt
PREHOOK: type: LOAD
#### A masked pattern was here ####
PREHOOK: Output: default@lineitem_test_txt
-POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/lineitem.txt' OVERWRITE INTO TABLE lineitem_test_txt
+POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/lineitem.tbl.bz2' OVERWRITE INTO TABLE lineitem_test_txt
POSTHOOK: type: LOAD
#### A masked pattern was here ####
POSTHOOK: Output: default@lineitem_test_txt
diff --git a/ql/src/test/results/clientpositive/llap/vectorized_ptf.q.out b/ql/src/test/results/clientpositive/llap/vectorized_ptf.q.out
index b2f1c4e..f284192 100644
--- a/ql/src/test/results/clientpositive/llap/vectorized_ptf.q.out
+++ b/ql/src/test/results/clientpositive/llap/vectorized_ptf.q.out
@@ -34,11 +34,11 @@ POSTHOOK: query: CREATE TABLE part_staging(
POSTHOOK: type: CREATETABLE
POSTHOOK: Output: database:default
POSTHOOK: Output: default@part_staging
-PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table part_staging
+PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table part_staging
PREHOOK: type: LOAD
#### A masked pattern was here ####
PREHOOK: Output: default@part_staging
-POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table part_staging
+POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table part_staging
POSTHOOK: type: LOAD
#### A masked pattern was here ####
POSTHOOK: Output: default@part_staging
diff --git a/ql/src/test/results/clientpositive/spark/vectorized_ptf.q.out b/ql/src/test/results/clientpositive/spark/vectorized_ptf.q.out
index 6c7e3a5..e31529a 100644
--- a/ql/src/test/results/clientpositive/spark/vectorized_ptf.q.out
+++ b/ql/src/test/results/clientpositive/spark/vectorized_ptf.q.out
@@ -34,11 +34,11 @@ POSTHOOK: query: CREATE TABLE part_staging(
POSTHOOK: type: CREATETABLE
POSTHOOK: Output: database:default
POSTHOOK: Output: default@part_staging
-PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table part_staging
+PREHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table part_staging
PREHOOK: type: LOAD
#### A masked pattern was here ####
PREHOOK: Output: default@part_staging
-POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/part_tiny.txt' overwrite into table part_staging
+POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/tpch/tiny/part.tbl.bz2' overwrite into table part_staging
POSTHOOK: type: LOAD
#### A masked pattern was here ####
POSTHOOK: Output: default@part_staging