Apache NiFi - ScanContent dictionary format not working - dictionary

Question

I'm trying to run a ScanContent processor on Apache Nifi, and whilst I can get the processor to run when scanning a text file, and using a .txt dictionary file with the search terms contained in it (and delimited by a newline character), I cannot get it to run when searching a file using the binary type of the processor for the dictionary file.
I am unsure whether I am simply using the wrong format for the binary dictionary file, or whether it needs to be encoded differently. I couldnt find any example dictionaries anywhere online that would be of any use (most things were related to the ScanAttributes instead).
The format of my dictionary file is:
(inside a .txt file)
32 00001001001000010000100001000000\n
The requirements according to the documentation are that the dictionary terms need to be a 4 byte integer, followed by the binary search term.
Does anyone have any experience of using this processor with a binary dictionary that might be able to help specify the format?

Answer 1

A binary dictionary file would typically be generated as the output of another program. There is an example in the ScanContent unit tests for how to accomplish this in Java:
#Test
public void testBinaryScan() throws IOException {
// Create dictionary file.
final String[] terms = new String[]{"hello", "good-bye"};
final ByteArrayOutputStream baos = new ByteArrayOutputStream();
try (final DataOutputStream dictionaryOut = new DataOutputStream(baos);) {
for (final String term : terms) {
final byte[] termBytes = term.getBytes("UTF-8");
dictionaryOut.writeInt(termBytes.length);
dictionaryOut.write(termBytes);
}
final byte[] termBytes = baos.toByteArray();
final Path dictionaryPath = Paths.get("target/dictionary");
Files.write(dictionaryPath, termBytes, StandardOpenOption.CREATE, StandardOpenOption.WRITE);
...