IBM 000-8697 User Manual

Utilities

Before you use dbload, you first create an

ASCII

command file that names the

input data files and the tables that receive the data. The command file maps
fields from one or more input files into columns of one or more tables within
your database.

The command file contains only

FILE

 and

INSERT

 statements. Each

FILE

statement names an input data file. The

FILE

statement also defines the data

fields from the input file that will be inserted into the table. Each

INSERT

statement names a table that will receive the data. The

INSERT

statement also

defines how dbload will place the data described in the

FILE

 statement into

the columns of the table.

Within the command file, the

FILE

statement can appear in the following two

forms:

■

Delimiter form

■

Character-position form

Use the delimiter form of the

FILE

 statement when every field in the input

data row uses the same delimiter and every row ends with a

NEWLINE

character. This format is typical of data rows with variable-length fields. You
can also use the delimiter form of the

FILE

statement with fixed-length fields

as long as the data rows meet the delimiter and

NEWLINE

requirements. The

delimiter form of

FILE

and

INSERT

is easier to use than the character-position

form.

Use the character-position form of the

FILE

 statement when you cannot rely

on delimiters and you need to identify the input data fields by character
position within the input row. For example, you would use this form to
indicate that the first input data field begins at character position 1 and
continues until character position 20.

Another reason to use this second form is if you must translate a character
string into the null value. For example, if your input data file uses a sequence
of blanks to indicate a null value, you must use the second form if you want
to instruct dbload to substitute null at every occurrence of the blank-
character string.