The last part of a document is usually the index, containing keywords and references into the document, where these keywords first appear or are described. Many source formats and converters support index creation, either when converting to PDF or HTML.
When going for PDF from a LATEXsource and using the hyperref macro package, a hyperlinked index can be generated by giving the ``hyperindex'' option:
\RequirePackage[hyperindex]{hyperref}
When going from FrameMaker to HTML using frame2html, this converter needs to be fixed to recognize index links at all, and then the index needs to be generated with the help of an external script. See the conversion report for Jörg Raasch's book for further details.[6]
As noted above during discussion of the look of links in the previous section, there are a number of ways to display links in the index, ranging from (virtual) page numbers, to images and chapter number/names as anchor, see Figures 16, 17, 18, 19 and 20.
Besides the look of the links (or better, their anchors) in the index, there's another point to pay attention to when generating an index, especially in HTML. When the index gets too long, it becomes slow to load and awkward to use. In such a case, at least a character-jump-bar has to be inserted, which allows jumping to the beginning of every set of index marks that begin with that character. The look of this bar may differ, but if present, it needs to be on the top of the index as well as at the end, similar to the navigation bar. Figure 23 shows some examples.
|
[width=0.8]bild15-1.eps
[width=0.8]bild15-2.eps
[width=0.8]bild15-3.eps |
Of course, inserting a character-jump-bar just helps navigating inside the index, it doesn't make it shorter or speed up loading. If required, the index can be broken into a single file for every set of index words starting with the same letter, with a jump-table for every file as the starting point, similar to the table of contents with two levels, see 4.9