> I was looking for the 'Dual Nature of NIL' article and noticed that using
> the wiki search functionality to search for 'NIL' yeilds... well, nil. No
> Kind of amusing and maybe not terribly surprising.
The reason is simple: The '+MupIdx' index class in "wiki/er.l" uses
'foldedWords' to dissect the text:
(de foldedWords (Mup)
(filter '((W) (>= (length W) 4))
(splitWords (in (blob Mup 'txt) (till))) ) ) ) )
As we see, it indexes only words which have a length of 4 characters or more.
The reason is to decrease the total index size (which may in fact not be
critical) and to avoid noise like "a", "the" and "and". This function could be
made more intelligent.