XFA Specification
Chapter 13, Dealing with Data in Different XML Formats
Extended Mapping Rules
366
<transform ref="book">
<rename>publication</rename>
<ifEmpty>remove</ifEmpty>
<whitespace>normalize<whitespace>
<presence>preserve</presence>
</transform>
the order of processing for a
book
element in the XML data document is:
1. the last-occurring
rename
transformation
publication
1. the higher-precedence
presence
transformation
dissolve
2. the higher-precedence
whitespace
transformation
normalize
3. the sole
ifEmpty
transformation
remove
The XFA Configuration DOM may contain a
transform
element with a
ref
value of
""
(the empty string),
which supplies a default transformation for all elements. The data loader applies this transformation to all
elements that do not match a
transform
element with a non-empty value for
ref
. For example, the
following fragment of an XFA configuration document:
<transform ref="">
<whitespace>trim</whitespace>
</transform>
<transform ref="address">
<whitespace>preserve</whitespace>
</transform>
causes leading and trailing white space to be trimmed from all
dataValue
node
value
properties except
for those originating from
address
elements, which keep their leading and trailing white space.
There can be multiple
transform
elements with a
ref
value of
""
. Conflicts between these are resolved
the same way as described above for
transform
elements having identical values for
ref
.
The attributes Element
This section defines an extended mapping rule that can be used to map XML attributes to
dataValue
nodes.
The XFA Configuration DOM may include an
attributes
element that overrides the default behavior.
See
“Config Specification” on page 676
attributes
element must
be one of the following keywords:
attribute keyword
delegate
ignore
preserve
Meaning
Allows an implementation-defined behavior, which may have the effect of the
ignore
or
preserve
keywords.
The data loader does not map attributes to
dataValue
nodes.
The data loader maps attributes to
dataValue
nodes, except those attributes
excluded by namespace as described in the following sections.
Invoking the extended mapping rule with
ignore
produces the same results as the default mapping
behavior for attributes as described in section
“Attributes” on page 133.