TCC web generator (webgen): requirements

The purpose of the web generator is to provide centralized control over the appearance of pages in documents structured as HTML webs (such as the TCC help system), and to reduce the drudgery of writing and maintaining webs.

This program assumes that the web conforms to a strict outline structure, and that structure is described by a single Plan file that defines the title and path of each Web page.

Page creators will use a shorthand notation to write source pages, and this notation will be expanded to the full HTML according to style rules defined in one place. Among the affected page elements are:

• Standard artwork such as header or footer logos, and copyright notices.
• Titles and main page headings are filled in automatically from the Plan file.
• Standard navigational links in the footer (and possibly also a header button-bar), such as Next, Previous, and at least one See also link to go up a level, are generated according to the Plan file.
• Cross-reference links in the body of a page will, by default, follow Jon Price's rule that the text of a link should match the title of the referenced page. In the source notation, you need only refer to the pathname of a file, and the link text will be generated automatically by looking up that pathname in the Plan file.
• Variant link text may be specified if the author desires, but it will be possible to obtain a report of all such links, showing both the standard and variant text.
• The page author can specify additional See also links to be added to the default up-a-level link.
• Modification timestamps will be extracted from RCS keywords in the source files.