Madcreator Rabinfingerprint command line vs java/scala program

Question

I'm using Madcreator's Rabinfingerprint java library in a scala program, https://github.com/themadcreator/rabinfingerprint.

Using the command line I generate an irreducible 53 degree polynomial:

java -jar rabinfingerprint.jar -polygen 53
31DF3F8C7356D3

Then fingerprint the file:

java -jar rabinfingerprint.jar -p 31DF3F8C7356D3 file.txt
1847CCD86D5DE5 file.txt

The problem I'm having is producing the same output, given the same input, using code.

// Can't create Polynomial from hex above, so convert hex to long and use that:
val polynomial = Polynomial.createFromLong(14037737891124947L)
// val p = java.lang.Long.toString(14037737891124947L, 16) 
// p = 31DF3F8C7356D3 same as command line input

val rabin = new RabinFingerprintLong(polynomial)
rabin.pushBytes(text.getBytes)

val fp = rabin.getFingerprintLong
println(java.lang.Long.toString(fp, 16))  // 1acc76a73eed1f
fp.toString // 7543159378603295

// Where is 1847CCD86D5DE5 ?

Answer 1

As I note in a comment above, if you're reading the contents of the file into a string, you need to be careful that you don't strip out line breaks or any other characters—if for example you're calling source.getLines.mkString you're going to lose the line breaks, which means you won't see the same results as the driver class, which uses Guava's ByteStreams.toByteArray.