Reading Data from Files

Data-reading programs read data from a file, record-by-record, and assign that data to variables, relations, dimensions, and composites in your analytic workspace. When the records in the file contain dimension values, you can limit dimensions to these values with the FILEREAD command before assigning the data to a variable dimensioned by them.

Example 11-1 Using FILEREAD in a Data-Reading Program

Suppose you want to update unit sales data for the product dimension in an analytic workspace. The new sales information is stored in a file called units.dat, which has the layout shown in the following figure.

Text description of the illustration freadsmp.gif

The FILEREAD command that reads the sample units.dat file is shown below.

FILEREAD funit -
    COLUMN 1 WIDTH 8 district -
    COLUMN 9 WIDTH 8 product -
    COLUMN 17 WIDTH 6 units

This command is processed in these steps:

1. The field is read beginning in column 1, and district dimension is limited to the value read. When the value read is not a dimension value of district, an error occurs.
2. The second field is read, and the product dimension is limited.
3. The third field is read, and the value is assigned to the units variable in the cell corresponding to the district and product read in Steps 1 and 2.

The full program, with commands to open and close the file, is shown next.

DEFINE readit1 PROGRAM
LD Read a data file
VARIABLE funit INTEGER
TRAP ON error
funit = FILEOPEN('olapfiles/units.dat' READ)
FILEREAD funit -
   COLUMN 1 WIDTH 8 district -
   COLUMN 9 WIDTH 8 product -
   COLUMN 17 WIDTH 6 units
FILECLOSE funit
RETURN
error:
IF funit NE na
   THEN FILECLOSE funit
END

Reading Structured PRN Files

You can also use the data-reading commands to read structured PRN files, which are produced by many PC software products. In a PRN file, quoted text or a series of numbers demarcated by spaces or commas constitutes a field of the record. Instead of specifying the column in which a field starts, you can use the STRUCTURED keyword to specify that you are reading a structured file. You can also use one or more FIELD keywords to indicate the number of the field you want to read.

Example 11-2 Reading a Structured PRN File

Suppose you want to read sales data from the structured PRN file illustrated below.

     010195  "TENTS"   "BOSTON"    307  50808.96
     010195  "TENTS"   "ATLANTA"   279  46174.92
     010195  "TENTS"   "CHICAGO"   189  31279.78
     010195  "TENTS"   "DALLAS"    308  50974.46
     010195  "TENTS"   "DENVER"    215  35582.82
     010195  "TENTS"   "SEATTLE"   276  45678.41
     010195  "CANOES"  "BOSTON"    352  70489.44
     010195  "CANOES"  "ATLANTA"   281  56271.40
     010195  "CANOES"  "CHICAGO"   243  48661.74
     010195  "CANOES"  "DALLAS"    176  35244.72
     010195  "CANOES"  "DENVER"    222  44456.41
     010195  "CANOES"  "SEATTLE"   335  67085.12

The file has product values in the second field, district values in the third field, and sales data in the fifth field.

You can limit the month dimension to the desired month, and then use the following command to read the sales data from the first six records in the file.

FILEREAD unit STOPAFTER 6 STRUCTURED FIELD 2 product -
   district FIELD 5 sales