Select first N words of a local macro in Stata

Question

Let's say I have the macro:

local words = "first second third fourth"

I want to select the first 3 words of it. I know that I can select, say the second, the word using:

local select: word 2  of `words' 
di "`select'"

However, I would like to have a line that selects the first N words (in this case N=3) of the said macro. For now, I am using a loop structure that seems quite overkill for such a simple procedure.

local words = "first second third fourth"
local select: word 1  of `words' 
forvalues i=2/3 {
    local rest : word `i' of `words' 
    local select `select'  `rest'
}

di "`select'"
first second third

Any better solution? Thank you

Answer 1

If you knew there were 4 words you could look for the last space and split before and after. I am not picking up that you know that in general.

Here is another general method.

local words = "first second third fourth"
tokenize "`words'"
local select "`1' `2' `3'"

di "`select'"
first second third

Here is another general method:

local words = "first second third fourth"
mata : tokens = tokens(st_local("words"))
mata : st_local("select", invtokens(tokens[1..3]))

di "`select'"
first second third

That is easier to extend to larger numbers of words (just replace 3).