 |
 |
30 of 97 |  |
 |
 |
30
mistakenly “see” that extra character as being a valid delimiter also, a condition
that will wreak havoc upon the entire balance of the data file.
For example, let’s say we have a simple data list as shown below:
WILLIAM SMYTH
123 MAIN ST
SOMEWHERE PA
PATTERSON, LYNNE
456 BROADWAY, APT 2B
ANYWHERE NY
Within the example scenario we have defined, all commas are recognized as
delimiting characters for data packets, regardless of where they are located or
why. Because of the extra commas, which have been inadvertently included,
any attempt to export this data utilizing a comma-delimited file scheme and then
import the data into a pre-existing table would be disastrous, as shown below:
Last
First
Street
City
State
WILLIAM
SMYTH
123 MAIN ST
SOMEWHERE
PA
PATTERSON
LYNNE
456 BROADWAY
APT 2B
ANYWHERE
NY
Not only did our data wind up in the wrong place in the table, it seems as if we
have actually created a third ”phantom” record, where in reality only two exist.
This is because the import routine “saw” the extra commas as valid delimiters
and assumed that we wanted to import blank fields containing no data where no
field actually existed at all.
You will find that the tilde “~” is a fairly reliable choice to employ as a delimiting
character, since most databases rarely contain data which include this symbol,
nor is it commonly employed as a “wild card” character by any of the more
popular data manipulation programs.
Univers is capable of generating either type of file, Fixed-Length or Delimited,
when the desired data has been filtered through the appropriately configured
SelFil.
