Complex string match SPSS (v20) - arrays

Question

I got a problem I cant figure out using SPSS (v20).
There is a master list with 10.000 strings. Think of it as an array like so:
['Sao Paolo S.P.', 'IDE MUNICH', '1_New YORK', 'BabylonX', ...]
I have a dataset with a variable, that contains strings similar to the beforementioned array, but that are not the exact same ones.
Like so:
What I need to do is: check if OldString (from the dataset) is part of any of the strings in the master array.
Obviously 123 Babylon (from the dataset) will be related to BabylonX (from the array).
Obviously 1234 Sao Paolo S (from the dataset) will be related to Sao Paolo S.P. (from the array).
and so on...
If a match is detected, then the string from the (master)array should be filled in in a new variable NewString.
Is there any way to achieve that? Using VBA, Perl, PHP this is dead easy, but using SPSS I got no clue how to combine those steps.

Answer 1

The following syntax is a possible way to do loop the match comparison using char.index.
*First I'm turning your master list into a dataset -
this can be done differently depending how the data is
stored right now. In this example I just copy-pasted
from your post into the syntax.
data list free/masterstring (a20).
begin data
'Sao Paolo S.P.', 'IDE MUNICH', '1_New YORK', 'BabylonX'
end data.
*now I create a new syntax with a comparison command for each string in the list.
cd "c:\your path".
string cmd (a100).
compute cmd=concat("if char.index('",lower(rtrim(masterstring)),"',
lower(rtrim(mystring)))>0 matchedstr='",rtrim(masterstring),"'.").
write out="check strings.sps"/cmd.
* the syntax is ready, at this point you will go back to your original dataset.
For the example I'm creating a small example dataset.
data list free/mystring (a20).
begin data
"123 Babylon" "babylon" "Sao Paolo" "1234 Sao Paolo S"
end data.
*now we can run the syntax created earlier on the present dataset.
string matchedstr (a50).
insert file="check strings.sps".
exe.
What you should see in the result is that "babylon" was recognized as part of "BabylonX" (the command equalizes lower/upper cases) and therefore "BabylonX" appears in the matchedstring. The same for "Sao Paolo" and "Sao Paolo S.P.".
Note: if mystring matches more than one string in the list, the present syntax will only capture the last match.