stanfordnlp/CoreNLP · error · IllegalArgumentException

Array must be sorted!

Error message

Array must be sorted!

What it means

ArrayUtils.gapEncodeList computes gap encoding of an int array, which is only well-defined for monotonically non-decreasing sequences (each gap must be non-negative). It scans the array and throws IllegalArgumentException('Array must be sorted!') if any element is smaller than its predecessor.

Solutions

1. Sort the input array first with Arrays.sort(orig) before calling gapEncodeList.
2. If you must preserve the original order, sort a copy: int[] sorted = orig.clone(); Arrays.sort(sorted); gapEncodeList(sorted).
3. Pre-validate with a loop or IntStream check that no element is smaller than the previous one, and handle the unsorted case explicitly.

Example fix

// before
List<Byte> encoded = ArrayUtils.gapEncodeList(docIds); // docIds unsorted
// after
int[] sorted = docIds.clone();
Arrays.sort(sorted);
List<Byte> encoded = ArrayUtils.gapEncodeList(sorted);

for (int i = 1; i < orig.length; i++) {
  if (orig[i] < orig[i-1]) throw new IllegalArgumentException("input to gapEncodeList must be sorted");
}

try {
  List<Byte> out = ArrayUtils.gapEncodeList(orig);
} catch (IllegalArgumentException e) {
  int[] sorted = orig.clone(); Arrays.sort(sorted);
  List<Byte> out = ArrayUtils.gapEncodeList(sorted);
}

When it happens

Trigger: Calling ArrayUtils.gapEncodeList(int[] orig) where for some i>0, orig[i] < orig[i-1], i.e. the array contains a decreasing pair; also triggered transitively via encodedList on unsorted input.

Common situations: Passing positions/docIDs that were not sorted because they came from a HashMap iteration order; appending new items to an encoded list without re-sorting; assuming gapEncodeList sorts for you (it validates but does not sort).

Understand the failure class

Background: "Must be a positive integer", "Invalid value", "Unsupported": the invalid-argument-value error family, when a library rejects the value you pass — this error's family across 35 libraries.

Related errors

• Arrays do not have the same length.
• Can only operate on preterminals
• CoreMap must have either a Calendar or DocDate annotation
• dim should be an array of size 2.
• Neither element of pair comparable

AI-assisted analysis of stanfordnlp/CoreNLP@1b7edd19c4 (2026-09-10). Data as JSON: /api/errors/0cefd541496f461c. Report an issue: GitHub.

Appendix: source

Thrown at src/edu/stanford/nlp/util/ArrayUtils.java:40

  private static final Redwood.RedwoodChannels log = Redwood.channels(ArrayUtils.class);

  /**
   * Should not be instantiated
   */
  private ArrayUtils() {}

  public static byte[] gapEncode(int[] orig) {
    List<Byte> encodedList = gapEncodeList(orig);
    byte[] arr = new byte[encodedList.size()];
    int i = 0;
    for (byte b : encodedList) { arr[i++] = b; }
    return arr;
  }

  public static List<Byte> gapEncodeList(int[] orig) {
    for (int i = 1; i < orig.length; i++) {
      if (orig[i] < orig[i-1]) {
        throw new IllegalArgumentException("Array must be sorted!");
      }
    }

    List<Byte> bytes = new ArrayList<>();

    int index = 0;
    int prevNum = 0;
    byte currByte = 0 << 8;

    for (int f : orig) {
      String n = (f == prevNum ? "" : Integer.toString(f-prevNum, 2));
      for (int ii = 0; ii < n.length(); ii++) {
        if (index == 8) {
          bytes.add(currByte);
          currByte = 0 << 8;
          index = 0;
        }
        currByte <<= 1;

View on GitHub (pinned to 1b7edd19c4)