- 62 -
In general though, entity references are used for representing
bigger entities like files,
documents, or text sections. The parser follows the references and pieces together the
entire document. So, let’s say you were writing an essay consisting of an introduction, a
body, and a conclusion. If you were to prepare each section separately, declaring each
as
an entity in your DTD, your XML document would refer to it as follows:
&introduction;
&body;
&conclusion;
The parser would then track down and read the actual introduction, body, and conclusion
entities and stuff them into a single document in memory. This enables you to take an
object-oriented approach to your XML documents.
If you’re asking the very important question of how we defined the entities,
wait a few
more pages: We get into that in the
“Making DTDs”
section.
So, the generic syntax of using an entity reference is as follows:
&entity_name;
It is an ampersand, followed by the entity’s name (declared in the DTD), followed by a
semicolon. No white spaces are allowed.
Processing Instructions
Processing instructions (PI)
are a hand-me-down syntax from SGML.
The syntax is used
to clearly mark up information sections that are specific to your application/application
processor (for example, a parser) and nothing else. You could think of it as a coded
memo format for your
particular processor/application. You use the memo to write specific instructions to your
application, instructions that you want to make sure are not thought of as conventional
data.
PIs are particularly useful if your XML document is being processed by more than one
parser (say a rich text format [RTF] engine or a CDF display engine).
In that case, you
may want to include some specific instructions to one engine that you don’t want the
other to pay attention to. The way this is done is to express processing instructions in a
markup syntax that is different from elements, comments, or attributes.
Processing
instructions are of the following generic format:
It is a greater-than sign, followed by a question mark, followed by the name of the
processor/application the PI is intended for, followed by the instruction and its value. The
instruction need not have a value, though it could.
Let’s take a look at an example. XML uses a PI for
one of its most fundamental
operations: the XML declaration. In order to tell an XML parser that the document you are
passing though it is in XML, you use a PI expressed as follows:
Remember to keep it case sensitive; using XML in capitals causes the parser to go
ballistic.
- 63 -
If you were using an RTF engine to
insert a page break, here’s the processing instruction
you would include in your XML document.
