A useful feature of SAS Stored Processes is the ability to load multiple files in a single request. This could be a bunch of CSVs for processing, or even a `.sas` program that you want to `%include` along with some associated data (deliberate SAS injection). This is particularly useful when you are running a build process, eg for a web application, and you need to send across some SAS programs that should be compiled into Stored Processes and re-downloaded as a distributable SPK. If you send your executable .sas program as the first file, to a ‘master’ STP that has just one line of code (`%inc &_webin_fileref1/source2;`), then this becomes the only entry point you need when running from a non-SAS environment!

Anyway, here is the sample code:

curl -v -L -b cookiefile -c cookiefile \
    -H "Content-Type: multipart/form-data" \
    -F "[email protected]/runme.sas" \
    -F "[email protected]/processme.csv" \
    -F "[email protected]/processmeaswell.csv" \
    --output "/tmp/out.txt"

Things to note:

curl options

  • -v = verbose
  • -L = follow the redirects
  • -b / -c = cookie in / out
  • -H = http header
  • -F = Files for upload
  • –output = output filename


  • $SERVER = your SAS server, eg: https://yourserver.com/SASStoredProcess/do
  • $STP = full metadata path to STP, eg:  /my/meta/path/to/my/service
  • $USER = sas username, including @saspw suffix if an internal account
  • $PASS = sas password

SAS Considerations

The following automatically assigned macro variables will be helpful when accessing the files you have uploaded:

  • _WEBIN_FILE_COUNT – contains a count of the number of files that were uploaded
  • _WEBIN_FILEREF – a unique fileref that can be used to access the file
  • _WEBIN_FILENAME – the location in which the file is saved in SAS (eg in WORK)
  • _WEBIN_NAME – the name given for the file on upload (eg somefile1 / somefile2 in the example)

The above variables are actually macro variable arrays – so _WEBIN_FILENAME1 will be the first,  _WEBIN_FILENAME2 will be the second, and so on.

Further resources

If you want more background on building web applications with SAS, check out the following resources!

The following web apps are available from Boemska as open source projects:

The following web app is a super useful tool for SAS data capture and data quality, as well as EUC / Power BI integration:

Finally, if you are looking to optimise your web applications for maximum performance then be sure to check out Boemska ESM.

Happy coding!