Uniforming string variable with loop

[Alessandro Castellani]

Hi everyone! I have encountered the following problem while doing some data cleaning.
I have a list of 30k + observations something like:

application_ id	firm_name
23	APPLE
24	APPLE
24	GOOGLE
26	APPLE INC.
27	APPLE & CO.
32	APPLE INC. (USA)
72	GOOGLE CORP.
75	GOOGLE CORP.

I would like to write a code conforming the different firm names, ideally to the shortest one (e.g. "APPLE").

My code is:

levelsof firm_name, local(names)

foreach n in `names' {
gen presence = strpos(firm_name, `n') > 0
replace firm_name = `n' if presence==1
drop presence
}

But I am getting "invalid name" type of error.

Many thanks for the help!

[Rich Goldstein]

I'm not sure exactly what you want, but it appears that just extracting the first word of name would work; here is some possible code (untested since you did not follow the advice in the FAQ and use -dataex-)

gen str fim_namew=word(firm_name,1)