top - download
⟦b024b59ab⟧ Wang Wps File
Length: 3642 (0xe3a)
Types: Wang Wps File
Notes: CPS/USM/005
Names: »3646A «
Derivation
└─⟦1347aae84⟧ Bits:30006174 8" Wang WCS floppy, CR 0315A
└─ ⟦this⟧ »3646A «
WangText
…02…CPS/USM/005
…02…BEP/830425…02……02…#
CAMPS DATA BASE LOADING
…02……02…CAMPS
T̲A̲B̲L̲E̲ ̲O̲F̲ ̲C̲O̲N̲T̲E̲N̲T̲S̲
1 DATA BASE LOADING BASED ON COMPILED
INIT STATEMENTS ................................
4
1.1 STEP 1 .....................................
4
1.1.1 Interactive Inputs .....................
4
1.1.2 Output to Screen .......................
5
1.2 STEP 2 .....................................
5
1.2.1 Interactive Inputs .....................
5
1.2.2 Output to Screen .......................
6
1̲ ̲ ̲D̲A̲T̲A̲ ̲B̲A̲S̲E̲ ̲L̲O̲A̲D̲I̲N̲G̲ ̲B̲A̲S̲E̲D̲ ̲O̲N̲ ̲C̲O̲M̲P̲I̲L̲E̲D̲ ̲I̲N̲I̲T̲ ̲S̲T̲A̲T̲E̲M̲E̲N̲T̲S̲
The loading procedure is handled by an interactive
PASCAL program LOAD ̲MERGE. This program operates in
two steps, see fig. 1. During the first step records
of similar length are read from max. 3 input fields.
The records are validated and sorted in increasing
order based on a specified primary key. The sorted
records are written on a specified output file (SUB
FILE) in blocks of 3K bytes. The first word in each
block is an integer, the value of which equals the
number of records allocated in the block. Futher, an
integer (= 0) is written before the first field in
each record. Non-completed blocks are filled with NULLs.
If specified, all primary keys are written separately
on a file. This list reflects the order in which the
records were sorted. (Applicable if primary keys are
character-strings).
In step 2 sub files, as created during step 1, are
stacked in specified order on a single output file.
By calling LOAD ̲MERGE it is possible to jump directly
to step 2.
1.1 S̲T̲E̲P̲ ̲1̲
When loading a SUB file the procedure will be as follows:
The link modules (max. 3) containing the compiled INIT
statements are copied into files INPUT1, INPUT2, and
INPUT3. Each of those files are then linked to an empty
compiled main module by activating the job files INIT
̲FILE ̲i, i = 1,2,3. The thus created input files, see
fig. 1, then serve as input to the SUB file loading.
1.1.1 I̲n̲t̲e̲r̲a̲c̲t̲i̲v̲e̲ ̲I̲n̲p̲u̲t̲s̲
1) Select step 1
2) Number of input files (max. 3)
3) Select first file (1,2, or 3)
4) Primary key location (bytes)
5) Primary key output
6) Record length (words)
7) Output file ID (SUB1, SUB1,...,SUBN) N = 10
add 2: Input files must contain records of similar
length
add 3: In current edition the input files may be read
as:
NUMBER OF FILES FILES READ
1 1 or 2 or 3
2 1 and 2 or 2 and 3
3 1 and 2 and 3
1.1.2 O̲u̲t̲p̲u̲t̲ ̲t̲o̲ ̲S̲c̲r̲e̲e̲n̲
1) Number of blocks delivered to specified sub file
Note that this number is an input parameter when
running step 2.
1.2 S̲T̲E̲P̲ ̲2̲
In this step sub files are stacked arbitrarily. Note
that different sub files may contain record entries
with different entry length.
When this step has been executed the output data are
saved on TOTAL ̲FILE.
1.2.1 I̲n̲t̲e̲r̲a̲c̲t̲i̲v̲e̲ ̲I̲n̲p̲u̲t̲s̲
1) Select step 2
2) Number of SUB files
3) Load order (SUB file IDs) and corresponding number
of blocks to be read from each file.
1.2.2 O̲u̲t̲p̲u̲t̲ ̲t̲o̲ ̲S̲c̲r̲e̲e̲n̲
1) File ID for single output file
1.3 LOAD ̲MERGE is filed on DPSVOL under directory LOAD
̲MERGE.D. This directory also contains the required
input/output files as described above.
