ICS 125: XML

Before XML, every program had its own file format that was totally unlike any other file format
Since everything was totally different, there was no good way to briefly describe the file formats
Since everything was totally different, there was no good way to build widely reusable utilities that work on them
The format of network messages sent between programs was also totally different for every program

XML stands for eXtensible Markable Language
XML is a standard way to represent any kind of data:
- The low-level syntax always the same
- The higher-level syntactic structures are described in another file (a DTD)
- XML looks like HTML, but you can define your own tags in a DTD
- XML is about representing data, not at all about how it looks on the screen
Since XML is standard and widely usable, people have made widely usable tools and libraries that process XML: many XML parsers are available for all programming languages
A DTD is like a grammar specification file

XML File: a data file, uses ASCII or other human readable characters
DTD declaration: each XML file can name the DTD that defines it's structure:
- <?xml version="1.0" encoding="ISO-8859-1"?>
- <!DOCTYPE note SYSTEM "note.dtd">
Element:
- Elements have a name
- Start-tag and end-tag with something in between: <elementname>something...</elementname>
- Elements can be nested: <elementname><nestedelement>something...</nestedelement></elementname>
- Every element has an end-tag. An element with no content can be written as <elementname></elementname> or as <elementname />
Attribute: a key-value pair within an opening tag. E.g.,
- <elementname attr="value" attr2="value" />
CDATA: text that is not elements. Can only occur between begin and end tags that allow CDATA

A DTD is like a grammar the defines the legal format of an XML file. The DTD tells a parser how to handle the XML file, and how to determine if it is valid.
Each XML file can start with a DOCTYPE line that refers to the DTD file, or the DTD can be directly defined inside the XML file, or an XML file may not specify a DTD.
<!ELEMENT element-name category>
- <!ELEMENT element-name EMPTY>
- <!ELEMENT element-name ANY>
- <!ELEMENT element-name (ele1,ele2+,ele3*,ele4?,(ele5|ele6))>
- <!ELEMENT element-name (#PCDATA|ele1)*)>
PCDATA: Parsed character data, it may include markup
CDATA: Non-parsed character data, any markup will be treated as literal characters
<!ATTLIST element-name attribute-name attribute-type default-value)>
- <!ATTLIST element-name attribute-name CDATA "default")>
- <!ATTLIST element-name attribute-name (enum1|enum2|...) enum1)>
- <!ATTLIST element-name attribute-name ID #REQUIRED>
- <!ATTLIST element-name attribute-name IDREF #REQUIRED>
- <!ATTLIST element-name attribute-name CDATA #IMPLIED>
- <!ATTLIST element-name attribute-name CDATA #FIXED "default">
A "well-formed" XML document is one that has matching tags, quoted attributes, and fits general XML syntax rules.
A "valid" XML document is one that follows the rules defined in it's DTD

There are two main types of XML parsers
- SAX, SAX2: Parse an XML document and trigger events for your application to process on-the-fly
- DOM: Parse an XML document and build a tree datastructure in memory that your application can process in any order
XML Parsers can have validation turned on or off