Fixes

FastFilter · Sep 16, 2023 · 5dfd4c5 · 5dfd4c5 · catap · Oct 1, 2023
1 parent ee55724
commit 5dfd4c5
Show file tree

Hide file tree

Showing 4 changed files with 8 additions and 19 deletions.
diff --git a/fastfilter/src/main/java/org/fastfilter/xor/Xor16.java b/fastfilter/src/main/java/org/fastfilter/xor/Xor16.java
@@ -44,6 +44,7 @@ public Xor16(long[] keys) {
         byte[] reverseH = new byte[size];
         int reverseOrderPos;
         long seed;
+        mainloop:
         do {
             seed = Hash.randomSeed();
             byte[] t2count = new byte[arrayLength];
@@ -54,11 +55,7 @@ public Xor16(long[] keys) {
                     t2[h] ^= k;
                     if (t2count[h] > 120) {
                         // probably something wrong with the hash function
-                        // let us not crash the system:
-                        for(int i = 0; i < fingerprints.length; i++) {
-                            fingerprints[i] = (short)0xFFFF;
-                        }
-                        return;
+                        continue mainloop;
                     }
                     t2count[h]++;
                 }

diff --git a/fastfilter/src/main/java/org/fastfilter/xor/Xor8.java b/fastfilter/src/main/java/org/fastfilter/xor/Xor8.java
@@ -49,6 +49,7 @@ public Xor8(long[] keys) {
         byte[] reverseH = new byte[size];
         int reverseOrderPos;
         long seed;
+        mainloop:
         do {
             seed = Hash.randomSeed();
             byte[] t2count = new byte[m];
@@ -59,11 +60,7 @@ public Xor8(long[] keys) {
                     t2[h] ^= k;
                     if (t2count[h] > 120) {
                         // probably something wrong with the hash function
-                        // let us not crash the system:
-                        for(int i = 0; i < fingerprints.length; i++) {
-                            fingerprints[i] = (byte)0xFF;
-                        }
-                        return;
+                        continue mainloop;
                     }
                     t2count[h]++;
                 }

diff --git a/fastfilter/src/main/java/org/fastfilter/xor/XorBinaryFuse8.java b/fastfilter/src/main/java/org/fastfilter/xor/XorBinaryFuse8.java
@@ -111,6 +111,7 @@ private void addAll(long[] keys) {
             blockBits++;
         }
         int block = 1 << blockBits;
+        mainloop:
         while (true) {
             reverseOrder[size] = 1;
             int[] startPos = new int[block];
@@ -145,12 +146,7 @@ private void addAll(long[] keys) {
             startPos = null;
             if (countMask < 0) {
                 // we have a possible counter overflow
-                // this branch is never taken except if there is a problem in the hash code
-                // in which case construction fails
-                for(int i = 0; i < fingerprints.length; i++) {
-                    fingerprints[i] = (byte)0xFF;
-                }
-                return;
+                continue mainloop;
             }
 
             reverseOrderPos = 0;

diff --git a/fastfilter/src/main/java/org/fastfilter/xorplus/XorPlus8.java b/fastfilter/src/main/java/org/fastfilter/xorplus/XorPlus8.java
@@ -119,6 +119,7 @@ public XorPlus8(long[] keys) {
         // then we try again, with a new random seed.
         long seed = 0;
         int attempts = 0;
+        mainloop:
         do {
             attempts++;
             if (attempts >= 100) {
@@ -145,9 +146,7 @@ public XorPlus8(long[] keys) {
                     int h = getHash(k, seed, hi);
                     t2[h] ^= k;
                     if (t2count[h] > 120) {
-                        // probably something wrong with the hash function; or, the keys[] array contains many copies
-                        // of the same value
-                        throw new IllegalArgumentException("More than 120 keys hashed to the same location; indicates duplicate keys, or a bad hash function");
+                        continue mainloop;
                     }
                     t2count[h]++;
                 }