You can run hplsql tool from a command line to execute a SQL statement or HPL/SQL script.
Syntax:
hplsql -e 'query' | -f file 
      [-main procname]
      [-d | --define | -hiveconf | -hivevar var=value ...] 
      [-version | --version]
      [-trace | --trace]
      [-H | --help]
Parameters:
| Parameter | Description | HPL/SQL Version | |
| -e 'query' | SQL statements to execute | 0.1 | |
| -f file | Execute SQL statements from file | 0.1 | |
| -main procname | Entry point (procedure or function name) | 0.3.7 | |
| -d var=value | Variable definition | 0.1 | |
| --define var=value | 0.1 | ||
| -hiveconf var=value | 0.1 | ||
| -hivevar var=value | 0.1 | ||
| -version | --version | Print HPL/SQL version | 0.1 | 
| -trace | --trace | Print trace information | 0.1 | 
| -H | --help | Print help information and exit | 0.1 | 
Notes:
Example 1:
Executing HPL/SQL statements from a script:
hplsql -f script.sql
Example 2:
Executing HPL/SQL statements from command line:
hplsql -e "NVL(MAX_PARTITION_DATE(db.sales, local_dt, code='A'), CURRENT_DATE)"
Example 3:
Using variables:
hplsql -e "PRINT a || ', ' || b" -d a=Hello -d b=world
Result:
Hello, world