How views are regenerated

The process for regenerating the views on the installed system consists of the following automated steps:

1. For each language directory under /usr/lib/scohelp, run a find(1) command to locate all view stanza files under that language.

2. Read and parse the stanza files and load the information into a view database in memory (and optionally save to disk).

3. For languages other than English, additional view data is collected from the English stanza files to fill in the views. This is because typically not all content is translated, and having references to the English content is better than no references at all in such views.

4. Sort by ``Parent View'' field.

5. All of the records with the same ``Parent'' field are assembled into a topic view HTML file that is to be displayed in the left-hand frame.

6. The title of a view HTML file is from the ``Title'' field for the record whose ID is the parent name.

7. The filename of a generated view HTML file is taken from the URL name for the record whose ID is the parent name. Normally, the topicID for such is TopicViews.

8. The nodes within a view are formatted into groups based on node class first, then weight within a node class, and then alphabetically within a weight class.

9. The appropriate frameset HTML files are generated to accommodate a view list and introduction page.

10. The view HTML file is placed into /usr/lib/scohelp/LANG according to the pathname of the URL given for it. Most should reside in /usr/lib/scohelp/LANG/TopicViews.

See also config_views(1M).