1.7

Setting boundaries using JavaScript
As soon as you select the On Script option as the trigger for establishing record boundaries
(see "Record boundaries" on page94), you are instructing the DataMapper to read the source
file sequentially and to trigger an event each and every time it hits a delimiter. (What a delimiter
is, depends on the source data and the settings for that data; see "Input data settings
(Delimiters)" on page93).
In other words, the script will be executed - by default - as many times as there are delimiters in
the input data.
If you know, for instance, that a PDF file only contains documents that are 3 pages long, your
script could keep count of the number of times it's been called since the last boundary was set
(that is, the count of delimiters that have been encountered). Each time the count is a multiple of
3, it could set a new record boundary. This is basically what happens when setting the trigger to
On Page and specifying 3 as the Number of Pages.
Note
Remember that a boundary script is being called on each new delimiter encountered by the
DataMapper parsing algorithm. If for instance a database query returns a million records, the script
will be executing a million times! Craft your script in such a way that it doesn't waste time
examining all possible conditions. Instead, it should terminate as soon as any condition it is
evaluating is false.
Accessing data
Data available inside each event
Every time a delimiter is encountered, an event is triggered and the script is executed. The
event gives the script access to the data between the current location - the start of a row, line or
page - and the next delimiter. So at the beginning of the process for a PDF or text file, you have
access to the first page only, and for a CSV or for tabular data, that would be the first row or
record.
This means that you can:
l Examine the data found in between delimiters for specific conditions.
l Examine specific regions of that data, or the available data as a whole.
Page 233