pvary commented on a change in pull request #2512:
URL: https://github.com/apache/hive/pull/2512#discussion_r677496547
##########
File path:
iceberg/iceberg-handler/src/test/java/org/apache/iceberg/mr/hive/TestHiveIcebergStorageHandlerWithEngine.java
##########
@@ -2271,6 +2274,116 @@ public void testStatWithPartitionedCTAS() {
checkColStat("target", "dept");
}
+ @Test
+ public void testAsOfTimestamp() throws IOException, InterruptedException {
+ Table table = prepareTableWithVersions(2);
+
+ List<Object[]> rows = shell.executeStatement(
+ "SELECT * FROM customers FOR SYSTEM_TIME AS OF '" +
timestampAfterSnapshot(table, 0) + "'");
+
+ Assert.assertEquals(3, rows.size());
+
+ rows = shell.executeStatement(
+ "SELECT * FROM customers FOR SYSTEM_TIME AS OF '" +
timestampAfterSnapshot(table, 1) + "'");
+
+ Assert.assertEquals(4, rows.size());
+
+ AssertHelpers.assertThrows("should throw exception",
IllegalArgumentException.class,
+ "Cannot find a snapshot older than 1970-01-01 00:00:00", () -> {
+ shell.executeStatement("SELECT * FROM customers FOR SYSTEM_TIME AS
OF '1970-01-01 00:00:00'");
+ });
+ }
+
+ @Test
+ public void testAsOfVersion() throws IOException, InterruptedException {
+ Table table = prepareTableWithVersions(2);
+
+ HistoryEntry first = table.history().get(0);
+ List<Object[]> rows =
+ shell.executeStatement("SELECT * FROM customers FOR SYSTEM_VERSION AS
OF " + first.snapshotId());
+
+ Assert.assertEquals(3, rows.size());
+
+ HistoryEntry second = table.history().get(1);
+ rows = shell.executeStatement("SELECT * FROM customers FOR SYSTEM_VERSION
AS OF " + second.snapshotId());
+
+ Assert.assertEquals(4, rows.size());
+
+ AssertHelpers.assertThrows("should throw exception",
IllegalArgumentException.class,
+ "Cannot find snapshot with ID 1234", () -> {
+ shell.executeStatement("SELECT * FROM customers FOR SYSTEM_VERSION
AS OF 1234");
+ });
+ }
+
+ @Test
+ public void testAsOfTimestampWithJoins() throws IOException,
InterruptedException {
+ Table table = prepareTableWithVersions(4);
+
+ List<Object[]> rows = shell.executeStatement("SELECT * FROM " +
+ "customers FOR SYSTEM_TIME AS OF '" + timestampAfterSnapshot(table, 0)
+ "' fv, " +
+ "customers FOR SYSTEM_TIME AS OF '" + timestampAfterSnapshot(table, 1)
+ "' sv " +
+ "WHERE fv.first_name=sv.first_name");
+
+ Assert.assertEquals(4, rows.size());
+
+ rows = shell.executeStatement("SELECT * FROM " +
+ "customers FOR SYSTEM_TIME AS OF '" + timestampAfterSnapshot(table,
1) + "' sv, " +
+ "customers FOR SYSTEM_TIME AS OF '" + timestampAfterSnapshot(table,
2) + "' tv " +
+ "WHERE sv.first_name=tv.first_name");
+
+ Assert.assertEquals(8, rows.size());
+
+ rows = shell.executeStatement("SELECT * FROM " +
+ "customers FOR SYSTEM_TIME AS OF '" + timestampAfterSnapshot(table, 2)
+ "' sv, " +
+ "customers lv " +
+ "WHERE sv.first_name=lv.first_name");
+
+ Assert.assertEquals(14, rows.size());
+ }
+
+ /**
+ * Creates the 'customers' table with the default records and creates extra
snapshots by inserting one more line
+ * into the table.
+ * @param versions The number of snapshots we want to create
+ * @return The table created
+ * @throws IOException When there is a problem during table creation
+ * @throws InterruptedException When there is a problem during adding new
data to the table
+ */
+ private Table prepareTableWithVersions(int versions) throws IOException,
InterruptedException {
+ Table table = testTables.createTable(shell, "customers",
HiveIcebergStorageHandlerTestUtils.CUSTOMER_SCHEMA,
+ fileFormat, HiveIcebergStorageHandlerTestUtils.CUSTOMER_RECORDS);
+
+ for (int i = 0; i < versions - 1; ++i) {
+ // Just wait a little so we definitely will not have the same timestamp
for the snapshots
+ Thread.sleep(100);
+ shell.executeStatement("INSERT INTO customers values(" +
+ (i + HiveIcebergStorageHandlerTestUtils.CUSTOMER_RECORDS.size()) +
",'Alice','Green_" + i + "')");
+ }
+
+ table.refresh();
+
+ return table;
+ }
+
+ /**
+ * Get the timestamp string which we can use in the queries. The timestamp
will be after the given snapshot
+ * and before the next one
+ * @param table The table which we want to query
+ * @param snapshotPosition The position of the last snapshot we want to see
in the query results
+ * @return The timestamp which we can use in the queries
+ */
+ private String timestampAfterSnapshot(Table table, int snapshotPosition) {
+ List<HistoryEntry> history = table.history();
+ long snapshotTime = history.get(snapshotPosition).timestampMillis();
+ long time = snapshotTime + 100;
+ if (history.size() > snapshotPosition + 1) {
+ time = snapshotTime + ((history.get(snapshotPosition +
1).timestampMillis() - snapshotTime) / 2);
+ }
+
+ SimpleDateFormat simpleDateFormat = new SimpleDateFormat("yyyy-MM-dd
HH:mm:ss.SSS000000");
Review comment:
> Also, maybe one more test where we insert into a table using time
travel? Not sure it has added value, but I can imagine CTAS being used w/ time
travel frequently
Added tests for CTAS and INSERT
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
