Though scholars in the original project do not mention this anywhere, it seems that CAMENA contains a surprisingly large number of words -- about 50 million of them. Since most of the words are Latin, this is very important (for example, the collection Croatiae auctores Latini contains 5 million words).
We checked the number of words by turning CAMENA XML texts into a BaseX database and counting words (tokens) with two XQuery scripts. The first one counts all tokens in the database:
let $counts :=
for $w in ft:tokens("camena-neolatinlit")
let $n := number($w//@count)
return $n
return sum($counts)
This query returns 5.0753733E7, that is 50,753,733 tokens.
The second XQuery script counts only the tokens below the text element:
let $counts :=
for $w in collection("camena-neolatinlit")//*:text
let $n := tokenize($w)
return count($n)
return sum($counts)
This query returns 50,458,045 tokens.