35.4.2 The INDEX file

The optional INDEX file provides a categorical view of the functions in the package. This file has a very simple format

• Lines beginning with ‘#’ are comments.
• The first non-comment line should look like this

  toolbox >> Toolbox name

• Lines beginning with an alphabetical character indicates a new category of functions.
• Lines starting with a white space character indicate that the function names on the line belong to the last mentioned category.

The format can be summarized with the following example

# A comment
toolbox >> Toolbox name
Category Name 1
 function1 function2 function3
 function4
Category Name 2
 function2 function5

If you wish to refer to a function that users might expect to find in your package but is not there, providing a work around or pointing out that the function is available elsewhere, you can use:

fn = workaround description

This workaround description will not appear when listing functions in the package with pkg describe but they will be published in the html documentation online. Workaround descriptions can use any html markup, but keep in mind that it will be enclosed in a bold-italic environment. For the special case of:

fn = use <code>alternate expression</code>

the bold-italic is automatically suppressed. You will need to use <code> even in references:

fn = use <a href="someothersite.html"><code>fn</code></a>

Sometimes functions are only partially compatible, in which case you can list the non-compatible cases separately. To refer to another function in the package, use <f>fn</f>. For example,

eig (a, b) = use <f>qz</f>

Since sites may have many missing functions, you can define a macro rather than typing the same link over and again.

$id = expansion

defines the macro id. You can use $id anywhere in the description and it will be expanded. For example,

$TSA = see <a href="link_to_spctools">SPC Tools</a>
arcov = $TSA <code>armcv</code>

id is any string of letters, numbers and _.