- SCHEMANAME
- Specifies the schema of the table. You can specify a NULL value to use
the default schema name. The SCHEMANAME parameter takes
an input argument that is a VARCHAR(128) data type.
- TABLENAME
- Specifies the name of the table into which the data is to be imported.
This table cannot be a system table or a declared temporary table. The string
must exactly match the case of the table name. Specifying a NULL value results
in an error. The TABLENAME parameter takes an input
argument that is a VARCHAR(128) data type.
- INSERTCOLUMNS
- Specifies the comma-separated column names of the table into which the
data will be imported. You can specify a NULL value to import into all columns
of the table. The INSERTCOLUMNS parameter takes an input
argument that is a VARCHAR(32672) data type.
- COLUMNINDEXES
- Specifies the comma-separated column indexes (numbered from one) of the
input data fields that will be imported. You can specify a NULL value to use all
input data fields in the file. The COLUMNINDEXES parameter
takes an input argument that is a VARCHAR(32672) data type.
- FILENAME
- Specifies the name of the file that contains the data to be imported.
If the path is omitted, the current working directory is used. The specified
location of the file should refer to the server side location if you are using
the Network Server. Specifying a NULL value results in an error. The
FILENAME parameter takes an input argument that is a
VARCHAR(32672) data type.
- COLUMNDELIMITER
- Specifies a column delimiter. The specified character is used in place
of a comma to signify the end of a column. You can specify a NULL value to
use the default value of a comma. The COLUMNDELIMITER
parameter takes an input argument that is a CHAR(1) data type.
- CHARACTERDELIMITER
- Specifies a character delimiter. The specified character is used in place
of double quotation marks to enclose a character string. You can specify a
NULL value to use the default value of a double quotation mark. The
CHARACTERDELIMITER parameter takes an input argument that
is a CHAR(1) data type.
- CODESET
- Specifies the code set of the data in the input file. The code set name
should be one of the Java supported character encoding sets. Data is converted
from the specified code set to the database code set (UTF-8). You can specify
a NULL value to interpret the data file in the same code set as the JVM in
which it is being executed. The CODESET parameter takes
an input argument that is a VARCHAR(128) data type.
- REPLACE
- A non-zero value for the replace parameter will import in REPLACE mode,
while a zero value will import in INSERT mode. REPLACE mode deletes all existing
data from the table by truncating the table and inserts the imported data.
The table definition and the index definitions are not changed. You can import
with REPLACE mode only if the table already exists. INSERT mode adds the
imported data to the table without changing the existing table data. Specifying
a NULL value results in an error. The REPLACE parameter
takes an input argument that is a SMALLINT data type.
If you create a schema, table, or column name as a non-delimited identifier,
you must pass the name to the import procedure using all uppercase characters.
If you created a schema, table, or column name as a delimited identifier, you
must pass the name to the import procedure using the same case that was used
when it was created.