Data in an Executable Program

Store in memory with data structures
Containers: std::vector, std::list, std::array, c-array
Objects: struct, class
Database

External Data

Must store persistent data externally, e.g., files
Must exchange data with other programs
Example: One program produces data (a producer), another program converts the data to a web page (a consumer)
Need to store and exchange using a data format

Text Data Formats

CSV
JSON
YAML
XML

XML

General-purpose interchange data format
Represents structured and semi-structured data
Very applicable to text documents, e.g., poetry and source code
Wide variety of tools and approaches

XML Technologies

XPath - Addressing language
XSLT - Transformation language
DTD, XML Schema, Relax NG - Schema languages, specify structure and content
DOM, JDOM, XOM, SAX2, LINQ - Parsers for XML, APIs to build on

Views of XML

Viewpoint	Concept	Characteristics
data	XML is a data format	No mixed content Lots of attributes Often a memory dump of internal objects Often inflexible
document	XML is a document format	Mixed content Favors elements over attributes High-level view of mixed data Very flexible to additional elements and attributes

Simple List in XML

XML Declaration

First thing in an XML document
Optional, but always include
version - Although there is an XML 1.1, most XML is 1.0.
encoding - Text encoding
UTF-8 is the default and the most common. UTF-8 is an extension of ASCII.
standalone - Ignore any markup declarations in the DTD
Except for exceptional situations, these are the only values you will see

Elements

Part	Example
Element
Element Start Tag
Element End Tag
Empty Element

XML Element

XML is a markup language
There are no specific XML tag names defined in the XML standard
The specific XML element names (and attributes) are defined for each application, e.g., srcML elements
Defining these names is not trivial
Use standard sets of XML elements whenever possible, e.g., XHTML

Element Content

Content Type	Example	Value
Text		"cin" is the text content of the element name
Nested elements		Element "name" is nested inside of element expr
Element		The start tag, the end tag, and all content in between

Nested Elements in XML

Attributes

Elements can have attributes, which are name-value pairs listed in the start tag
E.g., name is "type"
E.g., value is "block"
Attributes must be unique to an element, i.e., multiple attributes with the same name are not allowed
The value of an attribute is a string and cannot have nested elements

Attributes in XML

Attributes vs. Subelements

	Example	Nested Text	Nested Elements	Purpose
Attributes		Yes	No	metadata
Subelements		Yes	Yes	data

well-formed

Requirement of XML
Exactly one root element, <students>...</students>
Elements form a tree structure and nest properly
All elements are an empty element or have a start tag and an end tag

well-formed Violations

More than one root element
Missing end tags
Not nested properly
XML processors are required to stop processing and report an error if a document is not well-formed

Escaped Characters

Unescaped	Escaped	Requirements
`<`	`<`	escape required
`&`	`&`	escape required
`>`	`>`	escape optional
`'`	`'`	escape depends on context
`"`	`"`	escape depends on context

Escaped Text

Unescaped:
Escaped:

Mixed Content

Element has both text and nested elements
Additional markup for a section item
Good for text with included markup

Namespaces

Namespace Declaration	`xmlns:s="https://mlcollard.net/Student"`
Prefix	`s`
URI	`https://mlcollard.net/Student`
Element in the namespace	`<s:grade>`
Default Namespace Declaration	`xmlns="https://mlcollard.net/Student"`

Namespaces with Default Prefix

Namespaces with Prefix

Namespace Matching

Element name matching is based on namespace URI, not on the prefix. Both of these are identical
Many XML processors (tools that work on XML) require a non-default prefix for certain functionality, e.g., XPath in libxml2

Multiple Namespaces

Whitespace

significant and insignificant
Newlines are normalized (i.e., converted) to Unix line endings even on Windows
- CR + LF → LF
Whitespace in tags is insignificant and is just a separator, i.e., the following are equivalent:
XML processors will often remove extra insignificant whitespace
White space in element content can be significant. Assume that it is.

Validating for well-formedness

Any XML producer must produce well-formed XML, or it cannot be processed
xmllint - Tool for checking if XML is well-formed
Ubuntu package: libxml2-utils
Preinstalled on macOS
To check if XML is well-formed: xmllint --noout data/demo.xml

Further Validation

Allowed elements, text content of elements, and nested elements are specific to each XML application (i.e., each XML format)
Must be defined in an XML grammar
DTD
XSchema
RelaxNG

Special XML Contents

XML comment
CDATA section
XML Processing Instructions

XML Comment

XML Comment Start	`<!--`
XML Comment End	`-->`
XML Comment	`<!-- <s:student> -->`

XML Terminology

Term	Expansion	Example	Relation to Namespaces
qName	qualified name	`s:student`	Unique
prefix	prefix	`s`	Shared
localName	local name	`student`	May exist in multiple
URI	Uniform Resource Identifier	`https://mlcollard.net/Student`	URLs are a subset of URIs more

XML Interface Concepts

Concept	Attributes	Notes
Start Document		Occurs before parsing
XML Declaration	version, encoding, standalone	Occurs once before the root
Element Start Tag	qName, prefix, localName
Element End Tag	qName, prefix, localName
Characters	characters	Includes entity references
Attribute	qName, prefix, localName, value
XML Namespace	prefix, uri
XML Comment	value
CDATA	characters	Somewhat rare, and not in our data
Processing Instruction	target, data	Rare, and not in our data
End Document		Occurs after parsing