Geneva Document Specification

This is a draft standard, this notice will disappear once the specification is final.

A Geneva document is an ordered collection of elements. Geneva defines the following element types:

Pargraph
Listing
Table
Plaintext
Media
Section

Rich Text

A central component of all element types is rich text. Rich text is defined as a sequence of text tokens, each made up of a variable number of character strings and an attribute to signify its appearance. There are five different types of text tokens:

Token	Description
plain s	Render s in regular font.
bold s	Recommends to render s in bold font.
italic s	Recommends to render s in italic font.
fixed-width s	Recommends to render s in fixed-width font.
url s	Interpret s as a Uniform Resource Locator.
url s, u	Interpret u as a Uniform Resource Locator and s as its label.

Table 1. Text token types.

The occurrence of whitespace characters in text token strings is restricted by the following rules:

All whitespace character sequences are to be reduced to a single space character (ASCII 0x20 or equivalent).
For all token types except the plain type, discard prefixes and suffixes of whitespace character sequences.
For the first and last text tokens in a rich text sequence, discard prefixes and suffixes of whitespace character sequences respectively.

At least the following conceptual characters have to be recognized as whitespace:

Space
Tab
Newline (including Carriage Return)
Vertical Tab
Page break

Element Types

A paragraph consists of exactly one rich text sequence. It signifies a self-contained piece of text.

A listing consists of a finite sequence of rich text sequences. It signifies an ordered group of self-contained text pieces.

A table consists of a two-dimensional matrix of rich text sequences and a single rich text sequence being its description. It signifies a tabular relation of the matrix of rich text pieces.

A plaintext element consists of a verbatim character string and a single rich text sequence being its description. It signifies a sequence of characters which has to be preserved as is except for whitespace prefixes and suffixes (including newlines).

A media element consists of an Unique Resource Locator string and a single rich text sequence being its description. It signifies the embedment of an external resource.

A description as mentioned above, is a piece of text elaborating the contents of a given element.

A section consists of a Geneva document and a single rich text sequence being its heading. It signifies a continuous subsequence of the document, introduced by a headline (the heading).

Formal Definition

The table below defines a Geneva document formally using the modifed BNF syntax described in ANSI Common Lisp's Notational Conventions.¹

Symbol	Expression
document	document-element`*`
document-element	pargraph `\|` listing `\|` table `\|` plaintext `\|` media `\|` section
paragraph	text-token`+`
listing	rich-text`+`
table	rich-text table-row`+`
table-row	rich-text`+`
plaintext	rich-text string
media	rich-text string
section	rich-text document-element`*`
rich-text	text-token`*`
text-token	A text token, see “Rich Text”
string	A character string

Table 2. Formal definition of a Geneva document.

1. ANSI Common Lisp: Notational Conventions