[SOLR-7983] Utils.toUTF8 uses larger buffer than necessary for holding UTF8 data

• Type: Improvement
• Status: Closed
• Priority: Trivial
• Resolution: Fixed
• Fix Version/s: 5.4, 6.0
• Assignee: Shalin Shekhar Mangar
• Reporter: Shalin Shekhar Mangar
• Created: 27/Aug/15 14:11
• Updated: 09/May/16 18:51
• Resolved: 27/Aug/15 14:18

Description

Method in question:

public static byte[] toUTF8(CharArr out) {
    byte[] arr = new byte[out.size() << 2]; // is 4x the real worst-case upper-bound?
    int nBytes = ByteUtils.UTF16toUTF8(out, 0, out.size(), arr, 0);
    return Arrays.copyOf(arr, nBytes);
  }

The byte array should be at most out.size() * 3.