DocFlex Technology - Features - Data Processing
XML file data sources
Ability to use any XML files as a data sources, retrieve any character or numeric data
from them and generate by such data any kind of formatted/hypertext output
This features is available in DocFlex/XML.
Java application data sources
Using special drivers written with
DocFlex API, it is also
possible to connect to any other non-XML data sources provided by a Java application.
Such data sources (called Data Source Models
(or DSMs) and drivers to them --
can be processed in the same way as XML-files and allow to generate by them any kind
as rich examples of working with such application-provided data sources.
Single document from multiple data sources
By calling from one template another templates
it is possible to mix data from the different data sources and to generate by them a single documentation files.
The involved data sources can be of any possible nature, both XML and non-XML types.
Creating framed HTML documentation
allows to generate from one or more data sources a documentation that consists of many
separate HTML files interconnected by a common network of hyperlinks and displayed in separate frame windows.
This is achieved by creating special frameset templates.
Each frameset template consists of two major parts: the definition of a frameset structure
and the template body.
The frameset structure is defined in the form of a tree as shown on the following screenshot:
It is used to generate a frameset HTML file which contains no actual data but only a layout of frame windows.
The documents displayed in the windows are generated during interpretation of the template body,
which consists mostly of calls to the different document subtemplates:
Inserting graphic images (both static and provided by Java application)
DocFlex Technology supports inserting graphic images into the generated documentation.
The images can be either taken from the existing graphics files (via URLs, file pathnames or Java
resource names specified in the templates) – static images, or generated dynamically
by a Java application and, then, provided to the
by a custom written
(see DocFlex API for more details).
The supported graphics formats include JPEG, GIF, PNG, and WMF. Depending on the selected output
format and the generator's settings, the images can be either inserted as the references to the
graphics files stored externally near the generated documents (HTML, RTF) or embedded directly in the document
See Image Formatting
for more details.
For more information about image support, see also
DocFlex allows to generate hypertext links between any parts of the generated documentation
as well as to the external URLs.
The internal links are generated using the special hyperlink source- and target-anchors
defined within the properties of
Each anchor definition contains specifications for generating the anchor keys.
The following screenshot shows defining two hyperlink source anchors by a
component (the two are because this control generates a text hyperlink which loads two frame windows
at once; see description below):
And here is how the particular source anchor defined:
The following two screenshot show defining a hyperlink target anchor (there may be also several of them;
but in that case, this is only for convenience purpose):
The generation is subdivided into two phases:
During the first (estimation) phase, by the anchor definitions specified in the templates,
the possible anchor positions in the output documents are produced, together with their anchor keys.
These positions are stored in a single pool for all generated documents.
On the second (generation) phase, when the real output is generated, by each anchor having
a match in the anchor pool (i.e. another one of the opposite source/target type and with the same keys),
the real hyperlink anchor (source or target) is produced and inserted in the output.
Targeting frame windows
A hyperlink source definition may include a name of the targeted frame window (see screenshot
It is also possibility to define by the same
several hyperlink sources assigned with different document destinations and
targeted to different frame windows (see screenshot above).
This allows to load from a generated hyperlink several frame windows at once
with the different documents.
Hyperlinks within images (imagemaps)
For images provided by a Java application (via DSM Drivers),
it is possible to generate multiple hyperlinks within a single image.
To do this, the DSM Driver should also provide a hypertext imagemap for the image.
The following screenshot shows how various hyperlinks are defined for a complex diagram image
(specified in the
template included in DocFlex/XML | XSDDoc):
Supported data types
DocFlex recognizes the following base data types of XML-element/attribute values:
These data types are considered in many situations. For example, when defining
sorting of sequences of elements or attributes.
object' data type represents a reference to any Java object and is useful
when the data source is provided by a Java applications
and represents its internal data.
In addition, DocFlex recognizes attributes with element identifier data types:
The values of
ID attributes are used to distinguish particular elements.
IDREF attributes can be used to obtain the referenced elements
as well as for generation of hyperlinks.
Besides this, all values can be arrays of the above types.
FlexQuery-expressions are small scripts with Java-based syntax widely used in DocFlex templates
to specify all kinds of dynamically computed settings (such as dynamic
properties of template components,
filter conditions and so on).
The current FlexQuery implementation supports the following features:
Retrieving elements/attributes by Location Paths
- Location Paths
are special expressions used in DocFlex templates to specify searching of elements or attributes.
Location Paths applied in DocFlex are similar to those in XPath and looks like the following:
Step1 / Step2 / ... / StepN
- Each step may assume the form:
axis :: ETs [filter]
axis is the step's axis;
ETs is a list of matching element types (delimited with '|');
filter is a boolean FlexQuery-expression
StepN, in addition, may assume the form:
name is the name of an attribute.
However, there are several important differences between Location Paths used in DocFlex
and those used in XPath:
The axes borrowed from XPath are only the following:
attribute (only in the form:
A new axis
link was introduced, which has a form:
name is the name of an attribute whose type should be either
IDREF (an arrays of
IDREF). This axis selects all those elements of the XML document
(or DSM) whose identifiers are listed
in the attribute's value.
XPath predicates are replaced with a
filter, which can be any boolean
For more information see documentation:
Element Iterator (details) | Location Paths.
Iterations by elements, attributes, attribute values
The general data processing scheme employed in DocFlex consists in organizing nested iterations
by sets of elements collected from the data source. Such iterations are specified using
In addition to elements, it is possible to iterate by a set of attributes of a certain element or by multiple
values of a certain attribute. Organizing iterations by sets of attributes can be useful when the same processing
should be repeated for each attribute from the set. See
documentation for more details.
Collecting elements by Location Rules
This is the primary method used in DocFlex to specify collecting elements for iterations.
According to it, an
is specified with a set of
- Each Location Rule has the following form:
Matching Element Types → Location Path
- The rule is activated when the current context element has one of the
Matching Element Types.
Location Path is interpreted against the
and a set of new elements is produced.
Traversing element references
Unlike techniques available in XSL Transformations,
Location Rules allow to collect elements not only within a subtree attached to the
but also to involve in the search the elements accessible via reference attributes
(i.e. those with
This is achieved by using
link-axes (introduced in DocFlex)
together with specifying some Location Rules as recursive (i.e. repeatable an undefined number of times).
For more details, see documentation:
Element Iterator (details) | Recursive Location Rules.
The following screenshot shows specifying the collecting elements by Location Rules in an
Other methods of collecting elements
Besides Location Rules, DocFlex supports
two additional alternative methods for collecting iterated elements:
The set of elements is produced as a sequence of connected elements according to two
an expression calculating the first element (basing on the current
an expression producing all subsequent elements each next from the previous one
This allows to organize iterations by an arbitrary set of elements
provided by an external Java-class. The class should be written with the
For more information, see documentation:
Element Iterator (details) | Specifying Element Iteration Scope.
The information placed in reports can be presented in a particular sorting order.
This is achieved by specifying the sorting order for the iterated elements,
attributes or the attribute values:
Elements can be sorted:
Attributes can be sorted:
- by name
- by value (regarding its type)
by an arbitrary order defined in the
Multiple values of a single attribute can be sorted according to the value type.
For more information, see documentation:
Element Iterator (details) | Sorting
Template Components | Attribute Iterator.