StringTokenizer skipping delimiters

[Kristol]

I've created a bean to parse through a pipe delimited file to load into a
table. Whenever it encounters a null between pipes it skips to the next value.
How do I force the tokenizer to recognize that null field as an actual field?
Below is an example line from the file and the program.

In the first line, count should return 14, it returns 11.

Thanks for your help!

30103|DENNIS DOLAN |BELL ATLANTIC|4530 BISHOP LN SUITE 108|LOUISVILLE|KY|40218|US||(502)
421 - 3963|D1|CVG|||
30214|PAUL DAY |BELL ATLANTIC|845 LANE ALLEN RD SUITE 8|LEXINGTON|KY|40504|US|
|(606) 244 - 6343|D1|CVG| | |

StringTokenizer token = new StringTokenizer(pRecord,"|", false);
int holdCounter = 0;
System.out.println("###count is " +token.countTokens() );

while (token.hasMoreTokens() ){
String s = token.nextToken();
if (s!=null && s.length() !=0) { } else {s=" ";}
System.out.println("###WHAT IS A TOKEN: CHECK TOKEN " +s);
holdCounter = holdCounter + 1;
}
return holdCounter;

[Paul Clapham]

Use the other constructor that asks the tokenizer to return delimiters as
tokens.

PC2

"Kristol" <Kristoltaylor@hotmail.com> wrote in message
news:3c90a096$1@10.1.10.29...
>
> I've created a bean to parse through a pipe delimited file to load into a
> table. Whenever it encounters a null between pipes it skips to the next
value.
> How do I force the tokenizer to recognize that null field as an actual
field?
> Below is an example line from the file and the program.
>
> In the first line, count should return 14, it returns 11.
>
> Thanks for your help!
>
> 30103|DENNIS DOLAN |BELL ATLANTIC|4530 BISHOP LN SUITE
108|LOUISVILLE|KY|40218|US||(502)
> 421 - 3963|D1|CVG|||
> 30214|PAUL DAY |BELL ATLANTIC|845 LANE ALLEN RD SUITE
8|LEXINGTON|KY|40504|US|
> |(606) 244 - 6343|D1|CVG| | |
>
> StringTokenizer token = new StringTokenizer(pRecord,"|", false);
> int holdCounter = 0;
> System.out.println("###count is " +token.countTokens() );
>
> while (token.hasMoreTokens() ){
> String s = token.nextToken();
> if (s!=null && s.length() !=0) { } else {s=" ";}
> System.out.println("###WHAT IS A TOKEN: CHECK TOKEN " +s);
> holdCounter = holdCounter + 1;
> }
> return holdCounter;
>

[Kristol]

Thanks for the help. I should have posted this originally. I've already tried
that. It still skips the null fields. You'll get a count of the delimiters
that have values in them plus the count of the values. No nulls. So using
the first line of the file that I'd included, count is 25, the count should
be 28 in this instance. This count is just the first step. I have to make
sure the load file contains the correct number of columns before I can then
load it into the table. If tokenizers skips nulls then my bigger problem
will be loading data in the wrong columns.

[Paul Clapham]

Yes, that's right, count() is not going to be "right". You have to actually
read the tokens -- which will include delimiters. This code can check for
pairs of adjacent delimiters and behave as if there were null tokens between
them.

PC2

"Kristol" <kristoltaylor@hotmail.com> wrote in message
news:3c90ca98$1@10.1.10.29...
>
> Thanks for the help. I should have posted this originally. I've already
tried
> that. It still skips the null fields. You'll get a count of the delimiters
> that have values in them plus the count of the values. No nulls. So using
> the first line of the file that I'd included, count is 25, the count
should
> be 28 in this instance. This count is just the first step. I have to make
> sure the load file contains the correct number of columns before I can
then
> load it into the table. If tokenizers skips nulls then my bigger problem
> will be loading data in the wrong columns.