|
PSPZap uses five files (including transform) as follows:
Input => Source
Construct => Transform => Target
Construct => Output
The transform file defines the mapping between the source
construct file and the target construct file. It also identifies how data in
the input file is transferred into the output data file, and how it should be
manipulated when this is done. The content of the transform file is highly
dependent upon the content of the two construct files (source and target).
All transform files have the same layout, and can be edited
using any text editor. Each record in a transform file has four components:
transform identifier, source construct description, target construct
description and construct map. A typical record looks like this:
| Z |
Comp |
COMP |
{S2(Z)}{S1(Z)}{S3} |
| _ |
____ |
____ |
___________________ |
| | |
| |
| |
| |
| | |
| |
| |
Construct Map |
| | |
| |
Target Construct |
|
| | |
Source Construct |
|
|
| Transform Identifier |
|
|
|
Briefly, the above record defines a mapping between the
source construct Comp and a
target construct COMP. This is
the default mapping as identified by the transform identifier
Z. The construct map specifies
that the transform proceeds by referencing lower-level transform records for
the second, first and third components (in that order) of the source
construct.
There is a lot of flexibility in transform records,
particularly in the construct map component. The PSPZap User Guide provides
much more detail on how you define transform records and their components. Both
transform and construct files may include comments and blank lines, so can be
quite easy to read. |