indexRangeError function

The RangeError the platform raises for an out-of-range index.

A plain RangeError, deliberately not an IndexError. IndexError is a RangeError subtype and looks like the better fit, but the SDK's List.[] does not use it: the VM raises a plain RangeError, and on IndexError does not catch an out-of-range list access. Raising IndexError here would make d4rt strictly more catchable than the platform, so a script written against d4rt with on IndexError would break once compiled.

length may be 0, in which case end is -1 — below start. RangeError.range accepts that and reports "Valid value range is empty", matching what the platform says for an empty container.

SCD93 removed every interpreter call site, and that is not a deprecation. The six list-index guards that called this were pre-empting a native operator: targetValue was already a List and indexValue already an int, so targetValue[indexValue] reaches the SDK's own [], which raises a better error than this composed — RangeError (length) on a read where this said (index), and UnsupportedError on an unmodifiable list where the bounds test pre-empted it with a RangeError. A helper is the right thing for a BRIDGE that must raise an index error for a container the SDK cannot be asked about; it was the wrong thing standing in front of a native list. The reasoning above about RangeError versus IndexError is what makes it safe to keep as API: delegation produces a plain RangeError too, so the helper and the platform still agree.