Your code contains several errors. Have you tried running step 2 outside the macro? It will not work (the &FILELIST reference only gets initialized once, when the data step is compiled).
Rather than trying to loop inside the macro, I would write a small macro for a single import step (and test that), and then have the first data step write the code to execute for all files.
The macro would be something like this:
%macro bio_imp(l);
proc import datafile="&BIO.\&l." out=WORK.Labo
dbms=XLSX replace;
getnames=yes;
run;
proc append base=BDD_labo data=Labo force;
run;
%mend;
(I assume that you were trying to collect all the data into BDD_LABO with the last data step, but what your code does is just to overwrite with the newest read table). I did not put the (physical) filename into the output dataset name, as it is probably not a valid name for a SAS table.
Test the macro with a couple of files, to see that it works as intended.
To execute the imports, write code to execute all the macro calls, something like this:
proc delete data=BDD_labo;run; /* if the final data already exists, delete it */
data filelist;
length filename $256;
rc = filename('dir', "&BIO.");
if rc = 0 then
did = dopen('dir');
if did > 0 then do while(dread(did) = 0);
filename = dread(did);
output;
end;
rc = dclose(did);
run;
filename tempsas temp;
data _NULL_;
set filelist;
file tempsas;
put '%bio_imp(' filename ');';
run;
options mprint;
%include tempsas /source2;
You can also do it with CALL EXECUTE, but I much prefer writing to a temporary file, which you can then take a look at, and perhaps execute one line at a time when testing at first.
Click image to register for webinar
