Add %TypedArray%.prototype.subarray([ begin [, end ] ]) support.

[AnthonyCalandra]

This patch attempts to add support for the %TypedArray%.prototype.subarray function according to Section 22.2.3.26 of the ES2015 spec.

JerryScript-DCO-1.0-Signed-off-by: AnthonyCalandra anthony@anthony-calandra.com

[AnthonyCalandra]

On this line I get the following assertion failure:

ICE: Assertion 'ecma_typedarray_helper_is_typedarray (builtin_id)' failed at /Users/acalandra/jerryscript/jerry-core/ecma/builtin-objects/typedarray/ecma-builtin-typedarray-helpers.c(ecma_typedarray_helper_dispatch_construct):180.
Error: ERR_FAILED_INTERNAL_ASSERTION

Some help solving this issue would be greatly appreciated as I can't seem to figure out why this isn't a typedarray.

[rerobika]

@AnthonyCalandra The main reason is that src_builtin_id is incorrrent and that causes the assertion error.
uint8_t src_builtin_id = ecma_get_object_builtin_id (src_typedarray_p); returns with ECMA_BUILTIN_ID__COUNT which means the object is not builtin.

Currently I cannot find any helper function in the code base to get the builtin_id from a typedarray object, so my suggestion is to create a function like:

static uint8_t
ecma_typedarray_helper_get_builtin_id (ecma_object_t *obj_p) /**< typedarray object **/
{
  #define TYPEDARRAY_ID_CASE(magic_id, builtin_id) \
    case magic_id: \
    { \
      return builtin_id; \
    }

    switch (ecma_object_get_class_name (obj_p))
    {
      TYPEDARRAY_ID_CASE (LIT_MAGIC_STRING_INT8_ARRAY_UL, ECMA_BUILTIN_ID_INT8ARRAY)
      TYPEDARRAY_ID_CASE (LIT_MAGIC_STRING_UINT8_ARRAY_UL, ECMA_BUILTIN_ID_UINT8ARRAY)
      TYPEDARRAY_ID_CASE (LIT_MAGIC_STRING_UINT8_CLAMPED_ARRAY_UL, ECMA_BUILTIN_ID_UINT8CLAMPEDARRAY)
      TYPEDARRAY_ID_CASE (LIT_MAGIC_STRING_INT16_ARRAY_UL, ECMA_BUILTIN_ID_INT16ARRAY)
      TYPEDARRAY_ID_CASE (LIT_MAGIC_STRING_UINT16_ARRAY_UL, ECMA_BUILTIN_ID_UINT16ARRAY)
      TYPEDARRAY_ID_CASE (LIT_MAGIC_STRING_INT32_ARRAY_UL, ECMA_BUILTIN_ID_INT32ARRAY)
      TYPEDARRAY_ID_CASE (LIT_MAGIC_STRING_UINT32_ARRAY_UL, ECMA_BUILTIN_ID_UINT32ARRAY)
      TYPEDARRAY_ID_CASE (LIT_MAGIC_STRING_FLOAT32_ARRAY_UL, ECMA_BUILTIN_ID_FLOAT32ARRAY)
  #if CONFIG_ECMA_NUMBER_TYPE == CONFIG_ECMA_NUMBER_FLOAT64
      TYPEDARRAY_ID_CASE (LIT_MAGIC_STRING_FLOAT64_ARRAY_UL, ECMA_BUILTIN_ID_FLOAT64ARRAY)
  #endif /* CONFIG_ECMA_NUMBER_TYPE == CONFIG_ECMA_NUMBER_FLOAT64 */
      default:
      {
        JERRY_UNREACHABLE ();
      }
    }
  #undef TYPEDARRAY_ID_CASE
} /* ecma_typedarray_helper_get_builtin_id */

(Note: this function is the opposite of ecma_typedarray_helper_get_magic_string (uint8_t builtin_id))

I hope it solves your problem!

[AnthonyCalandra]

Thanks Robert, I found your comment very helpful.

[rerobika]

I think the int64_t cast can be eliminated like this:

  ecma_length_t subarray_length = 0;
  if (end_index_uint32 > begin_index_uint32)
  {
    // Cannot underflow
    subarray_length = end_index_uint32 - begin_index_uint32;
  }
  // Otherwise the subarray_length must be 0

[LaszloLango]

We should return here if ret_value contains an error. The ECMA_OP_TO_NUMBER_TRY_CATCH might fail, so it can fill the ret_value with an error.

[AnthonyCalandra]

Done

[LaszloLango]

LGTM

[zherczeg]

What happens if the original Int32Array is mapped only to a subarray of the arraybuffer? I would like a test case for that as well.

[AnthonyCalandra]

Done. If the new tests aren't what you requested please put your request into pseudocode.

[zherczeg]

LGTM