/**
* Joins array elements to string.
*
* @param arr Array.
* @return String.
*/
@Nullable
public static String compactArray(Object[] arr) {
if (arr == null || arr.length == 0) return null;
String sep = ", ";
StringBuilder sb = new StringBuilder();
for (Object s : arr) sb.append(s).append(sep);
if (sb.length() > 0) sb.setLength(sb.length() - sep.length());
return U.compact(sb.toString());
}
/* -------------------------------------------------------------------------------------------------------
This method:
This method:
-- Takes in three paramters:
1. array - this is the byte array that actually holds the document contents
2. md5Hases - holds the entire hash values of the document
3. Divisor1/Divisor2/divisor3... - main and back up divisors
5. The remainder we are looking for
6/7. min/max boundaries
-- We will start running the karb rabin algorithm
-- We will find the boundaries using mod values and once they equal the mod value we have stored
-- we also have the divsor2/3 .. which are backup divisors. If we don't find a boundary by the divisor1 once we hit the maxBoundary
-- we will see if we have one with divisor2, if not, then we will see if we have one with divisor3 and so on
-- We will hash everything in that hash boundary and store it
-------------------------------------------------------------------------------------------------------- */
private static void runTddd(
byte[] array,
ArrayList<Long> md5Hashes,
long divisor1,
long divisor2,
long divisor3,
long remainder,
long minBoundary,
long maxBoundary) {
int documentStart = 0; // used to keep track of where the boundaries are
boolean match = false; // used to ck if we encountered a match
int backUpBreakPoint = -1; // used to store the backup breakpoint
int secondBackUpBreakPoint = -1; // used with the divisor3
StringBuilder builder = new StringBuilder();
int i = documentStart + (int) minBoundary - 1; // so we start at the minimum
// loop through all the values in the document
for (; i < md5Hashes.size() - 1; ++i) {
// if ((i - documentStart + 1) < minBoundary) // if the size of this boundary is less than
// the min, continue looping
// continue;
/*-----------------------------------------------------------------
- If the mod of this equals the modvalue we defined, then
- this is a boundary
------------------------------------------------------------------*/
if ((md5Hashes.get(i - 1) + md5Hashes.get(i) + md5Hashes.get(i + 1)) % divisor1
== remainder) // ck if this equals the mod value
{
// Hash all the values in the range (documentStart,current(i))
// Remember we only want to hash the original VALUES from the array that contains the
// original
// content of the file. Not the hash values in the md5Hash Array
for (int j = documentStart; j <= i; ++j) {
builder.append(array[j]); // store everything upto the current value
}
String original = builder.toString();
// if the string is a perfect match ( hash and original string)
if (HashClass.is_string_match(original, table)) // iinsert the hash in the table)
coverage += i - documentStart + 1; // this is the amount of bytes we saved
documentStart = i + 1; // set this as the beginning of the new boundary
backUpBreakPoint = -1; // reset this
secondBackUpBreakPoint = -1;
numOfPieces++; // increment the num of pieces
i = i + (int) minBoundary - 1; // skip all the way here
builder.setLength(0); // reset the stringBuilder for the next round
} else if ((md5Hashes.get(i - 1) + md5Hashes.get(i) + md5Hashes.get(i + 1)) % divisor2
== remainder) { // check if this is the backup point
backUpBreakPoint = i; // this is the backup breakpoint
} else if ((md5Hashes.get(i - 1) + md5Hashes.get(i) + md5Hashes.get(i + 1)) % divisor2
== remainder) {
secondBackUpBreakPoint = i; // set the second backup point
}
if ((i - documentStart + 1) >= maxBoundary) { // we have reached the maximum
// ck if we have a backUpbreakpoint
int point;
if (backUpBreakPoint != -1) // if we do, set this as the boundary
point = backUpBreakPoint;
else if (secondBackUpBreakPoint != -1) {
point = secondBackUpBreakPoint; // if we don't have a first break point, find the second
} else point = i; // else this current value of i is the breakpoint
// Hash all the values in the range (documentStart,current(i))
// Remember we only want to hash the original VALUES from the array that contains the
// original
// content of the file. Not the hash values in the md5Hash Array
for (int j = documentStart; j <= point; ++j) {
builder.append(array[j]); // store everything upto the current value
}
String original = builder.toString();
if (HashClass.is_string_match(original, table))
coverage += point - documentStart + 1; // this is the amount of bytes we saved
numOfPieces++; // increment the num of pieces
documentStart = point + 1; // set this as the beginning of the new boundary
backUpBreakPoint = -1; // reset this
secondBackUpBreakPoint = -1; // reset the secondBackUp point
i = point + (int) minBoundary - 1; // skip all the way here ;
builder.setLength(0); // reset the stringBuilder for the next round
}
} // end of the for loop
// -------------------------------------------------------------------------------------------
// we are missing the last boundary, so hash that last value
// We will also check against our values of the strings we already have, and if we encountered
// this
// already, then we will simply increment the counter, otherwise we will insert it in the
// hashtable
// and increase our miss counter
// ----------------------------------------------------------------------------------------------
for (int j = documentStart; j < array.length; ++j) {
builder.append(array[j]);
}
// only compute hash and insert into our hashtable only if the string buffer isn't empty
String original = builder.toString();
if (HashClass.is_string_match(original, table))
coverage += array.length - documentStart; // this is the amount of bytes we saved
numOfPieces++; // increment the num of pieces
} // end of the method
