VarOpt heapify: add validation for totalRWeight in full preamble (align with C++)

[Fengzdadi]

Description

In Java VarOptItemsSketch heapify, full‑preamble deserialization currently only checks that rCount > 0, but does not validate the totalRWeight field. C++ (var_opt_sketch_impl.hpp) rejects NaN or non‑positive total_wt_r, and Go was updated to follow C++ behavior. This creates a cross‑language inconsistency and allows corrupted or malicious bytes to silently produce NaN/negative tau and invalid estimates.

Current Java code

File: src/main/java/org/apache/datasketches/sampling/VarOptItemsSketch.java
Method: heapify (around where numPreLongs == Family.VAROPT.getMaxPreLongs())

Current logic:

double totalRWeight = 0.0;
if (numPreLongs == Family.VAROPT.getMaxPreLongs()) {
  if (rCount > 0) {
    totalRWeight = extractTotalRWeight(srcSeg);
  } else {
    throw new SketchesArgumentException(
        "Possible Corruption: " + Family.VAROPT.getMaxPreLongs()
        + " preLongs but no items in R region");
  }
}

Behavior in other languages

C++ (sampling/include/var_opt_sketch_impl.hpp):

• Full preamble → if r == 0 or total_wt_r is NaN or <= 0, throw

  "Possible corruption: deserializing in full mode but r = 0 or invalid R weight.
   Found r = ..., R region weight = ..."
  

Go (sampling/varopt_items_sketch.go):

• Full preamble → reject NaN or <= 0 totalRWeight (aligned with C++)

Proposed change

Add validation after extractTotalRWeight in Java:

double totalRWeight = 0.0;
if (numPreLongs == Family.VAROPT.getMaxPreLongs()) {
  if (rCount > 0) {
    totalRWeight = extractTotalRWeight(srcSeg);
    if (Double.isNaN(totalRWeight) || totalRWeight <= 0.0) {
      throw new SketchesArgumentException(
          "Possible Corruption: deserializing in full mode but r = 0 or invalid R weight. "
          + "Found r = " + rCount + ", R region weight = " + totalRWeight);
    }
  } else {
    throw new SketchesArgumentException(
        "Possible Corruption: " + Family.VAROPT.getMaxPreLongs()
        + " preLongs but no items in R region");
  }
}

Why this matters

• Prevents NaN/negative tau from silently propagating
• Aligns Java behavior with C++
• Only rejects invalid/corrupted bytes