1.4

line in your input data. Setting a maximum data line length that greatly exceeds the length
of the longest line in your input data may increase execution time.
l
Page delimiter type: Defines the delimiter between each page of data. Multiple such
pages can be part of a Source Record, as defined by the Boundaries.
l
On lines: Triggers a new page in the Data Sample after a static number of lines
(called Lines per Page), or using a Form Feed character.
l
On text: Triggers a new page in the Data Sample when a specific string is found in
a certain location.
l
Word to find: Compares the text value with the value in the Source Record.
l
Match case: Activates case-sensitive text comparison.
l
Location: Choose Selected area or Entire width to use the value of the
current data selection as the text value.
l
Left/Right: Use the spin buttons to set the start and stop columns to the
current data selection (Selected area) in the Source Record.
l
Lines before/after: Defines the delimiter a certain number of lines before or
after the current line. This is useful if the text triggering the delimiter is not on
the first line of each page.
For a XML File
Delimiters for XML files are related to XML Elements, or locations in its hierarchy.
l
Use root element: Locks the XML Elements option to the top-level element. If there is
only one top-level element, there will only be one record before the Boundaries are set.
l
XML elements: Displays a list containing all the elements in the XML file. Selecting an
element causes a new page of data to be created every time an instance of this element
is encountered.
The Boundaries Section
When the Data Source is received by the DataMapper, it has no boundaries to tell the
DataMapper if it contains different records or where each of those records begins and ends.
That's because boundaries are not actual bits of data (like a character or a field would be);
boundaries are a logical structure that exists outside the Data Source (note that some formats
like PDF/VT actually include structured information, but those are the exception rather than the
norm). Boundaries are therefore a form of metadata. You could very well use the same exacted
data with a different boundary structure, in order to extract different information. Think, for
instance, of an Invoice Run stored in a PDF. You could very well be using a structure where
each invoice is a single record or you could group all invoices for one customer into a single
Page 117