Differences

This shows you the differences between two versions of the page.

@@ Line 1: / Line 1: @@
+====== PART_COUNT Function ======
+PART_COUNT function returns the number partitions in the specified table.
+**Syntax**:
+<code language="sql">
+PART_COUNT([db_name.]table_name, part_col=filter, ...);
+</code>
+**Parameters:**
+| [dbname.]table_name | Identifier, variable or expression |
+| part_col=filter | One or more partition filters |
+**Notes**:
+  * HPL/SQL uses the following Hive statement to get the partition information:
+<code language="sql">
+SHOW PARTITIONS db_name.tab_name [PARTITION (part_col=filter, ...)]
+</code>
+**Return Value:**
+  * The number of partitions
+  * NULL if the table does not exist or an error occurs
+**Return Type:**
+INT
+**Examples:**
+Table //db.orders// is partitioned by //dt// and //region// columns and has the following partitions:
+| dt=2015-09-01/region=1 |
+| dt=2015-09-01/region=2 |
+| dt=2015-09-02/region=1 |
+| dt=2015-09-03/region=3 |
+| dt=2015-09-03/region=2 |
+Get the total number of partitions:
+<code language="sql">
+part_count(db.orders);
+--
+</code>
+Get the number of partitions in //region 1//:
+<code language="sql">
+part_count(db.orders, region='1');
+--
+</code>
+**Compatibility**: HPL/SQL extension.
+**Version:** HPL/SQL 0.3.13
+See also:
+  * [[part-count-by|PART_COUNT_BY]] - Get the number of partitions (group by)
+  * [[part-loc|PART_LOC]] - Get the location of a partition

HPL/SQL - Procedural SQL on Hadoop, NoSQL and RDBMS