/** * RegExp helper function to start the recursive matching algorithm * and create the result Array object * * @return completion value * Returned value must be freed with ecma_free_completion_value */ ecma_completion_value_t ecma_regexp_exec_helper (ecma_value_t regexp_value, /**< RegExp object */ ecma_value_t input_string, /**< input string */ bool ignore_global) /**< ignore global flag */ { ecma_completion_value_t ret_value = ecma_make_empty_completion_value (); JERRY_ASSERT (ecma_is_value_object (regexp_value)); JERRY_ASSERT (ecma_is_value_string (input_string)); ecma_object_t *regexp_object_p = ecma_get_object_from_value (regexp_value); JERRY_ASSERT (ecma_object_get_class_name (regexp_object_p) == LIT_MAGIC_STRING_REGEXP_UL); ecma_property_t *bytecode_prop_p = ecma_get_internal_property (regexp_object_p, ECMA_INTERNAL_PROPERTY_REGEXP_BYTECODE); re_bytecode_t *bc_p = ECMA_GET_POINTER (re_bytecode_t, bytecode_prop_p->u.internal_property.value); ecma_string_t *input_string_p = ecma_get_string_from_value (input_string); lit_utf8_size_t input_string_size = ecma_string_get_size (input_string_p); MEM_DEFINE_LOCAL_ARRAY (input_utf8_buffer_p, input_string_size, lit_utf8_byte_t); ecma_string_to_utf8_string (input_string_p, input_utf8_buffer_p, (ssize_t) input_string_size); lit_utf8_iterator_t iterator = lit_utf8_iterator_create (input_utf8_buffer_p, input_string_size); re_matcher_ctx_t re_ctx; re_ctx.input_start_p = iterator.buf_p; re_ctx.input_end_p = iterator.buf_p + iterator.buf_size; /* 1. Read bytecode header and init regexp matcher context. */ re_ctx.flags = (uint8_t) re_get_value (&bc_p); if (ignore_global) { re_ctx.flags &= (uint8_t) ~RE_FLAG_GLOBAL; } JERRY_DDLOG ("Exec with flags [global: %d, ignoreCase: %d, multiline: %d]\n", re_ctx.flags & RE_FLAG_GLOBAL, re_ctx.flags & RE_FLAG_IGNORE_CASE, re_ctx.flags & RE_FLAG_MULTILINE); re_ctx.num_of_captures = re_get_value (&bc_p); JERRY_ASSERT (re_ctx.num_of_captures % 2 == 0); re_ctx.num_of_non_captures = re_get_value (&bc_p); /* We create an invalid iterator, that will be used to identify unused result values. */ lit_utf8_iterator_t unused_iter = lit_utf8_iterator_create (NULL, 0); unused_iter.buf_p = (lit_utf8_byte_t *) 1; MEM_DEFINE_LOCAL_ARRAY (saved_p, re_ctx.num_of_captures + re_ctx.num_of_non_captures, lit_utf8_iterator_t); for (uint32_t i = 0; i < re_ctx.num_of_captures + re_ctx.num_of_non_captures; i++) { saved_p[i] = unused_iter; } re_ctx.saved_p = saved_p; uint32_t num_of_iter_length = (re_ctx.num_of_captures / 2) + (re_ctx.num_of_non_captures - 1); MEM_DEFINE_LOCAL_ARRAY (num_of_iter_p, num_of_iter_length, uint32_t); for (uint32_t i = 0; i < num_of_iter_length; i++) { num_of_iter_p[i] = 0u; } bool is_match = false; re_ctx.num_of_iterations_p = num_of_iter_p; int32_t index = 0; ecma_length_t input_str_len = lit_utf8_string_length (iterator.buf_p, iterator.buf_size); if (iterator.buf_p && (re_ctx.flags & RE_FLAG_GLOBAL)) { ecma_string_t *magic_str_p = ecma_get_magic_string (LIT_MAGIC_STRING_LASTINDEX_UL); ecma_property_t *lastindex_prop_p = ecma_op_object_get_property (regexp_object_p, magic_str_p); ECMA_OP_TO_NUMBER_TRY_CATCH (lastindex_num, lastindex_prop_p->u.named_data_property.value, ret_value) index = ecma_number_to_int32 (lastindex_num); JERRY_ASSERT (iterator.buf_pos.offset == 0 && !iterator.buf_pos.is_non_bmp_middle); if (!lit_utf8_iterator_is_eos (&iterator) && index <= (int32_t) input_str_len && index > 0) { lit_utf8_iterator_advance (&iterator, (ecma_length_t) index); } ECMA_OP_TO_NUMBER_FINALIZE (lastindex_num); ecma_deref_ecma_string (magic_str_p); } /* 2. Try to match */ lit_utf8_iterator_t sub_iter = lit_utf8_iterator_create (NULL, 0); while (ecma_is_completion_value_empty (ret_value)) { if (index < 0 || index > (int32_t) input_str_len) { if (re_ctx.flags & RE_FLAG_GLOBAL) { ecma_string_t *magic_str_p = ecma_get_magic_string (LIT_MAGIC_STRING_LASTINDEX_UL); ecma_number_t *lastindex_num_p = ecma_alloc_number (); *lastindex_num_p = ECMA_NUMBER_ZERO; ecma_op_object_put (regexp_object_p, magic_str_p, ecma_make_number_value (lastindex_num_p), true); ecma_dealloc_number (lastindex_num_p); ecma_deref_ecma_string (magic_str_p); } is_match = false; break; } else { ECMA_TRY_CATCH (match_value, re_match_regexp (&re_ctx, bc_p, iterator, &sub_iter), ret_value); if (ecma_is_value_true (match_value)) { is_match = true; break; } if (!lit_utf8_iterator_is_eos (&iterator)) { lit_utf8_iterator_advance (&iterator, 1); } index++; ECMA_FINALIZE (match_value); } } if (iterator.buf_p && (re_ctx.flags & RE_FLAG_GLOBAL)) { ecma_string_t *magic_str_p = ecma_get_magic_string (LIT_MAGIC_STRING_LASTINDEX_UL); ecma_number_t *lastindex_num_p = ecma_alloc_number (); *lastindex_num_p = sub_iter.buf_pos.offset; ecma_op_object_put (regexp_object_p, magic_str_p, ecma_make_number_value (lastindex_num_p), true); ecma_dealloc_number (lastindex_num_p); ecma_deref_ecma_string (magic_str_p); } /* 3. Fill the result array or return with 'undefiend' */ if (ecma_is_completion_value_empty (ret_value)) { if (is_match) { ecma_completion_value_t result_array = ecma_op_create_array_object (0, 0, false); ecma_object_t *result_array_obj_p = ecma_get_object_from_completion_value (result_array); ecma_string_t *input_str_p = ecma_new_ecma_string_from_utf8 (iterator.buf_p, iterator.buf_size); re_set_result_array_properties (result_array_obj_p, input_str_p, re_ctx.num_of_captures / 2, index); ecma_deref_ecma_string (input_str_p); for (uint32_t i = 0; i < re_ctx.num_of_captures; i += 2) { ecma_string_t *index_str_p = ecma_new_ecma_string_from_uint32 (i / 2); /* Note: 'iter_p->buf_p == NULL' means the input is empty string */ if ((re_ctx.saved_p[i].buf_p != unused_iter.buf_p && re_ctx.saved_p[i + 1].buf_p != unused_iter.buf_p) && re_ctx.saved_p[i + 1].buf_pos.offset >= re_ctx.saved_p[i].buf_pos.offset) { ecma_length_t capture_str_len; capture_str_len = (ecma_length_t) re_ctx.saved_p[i + 1].buf_pos.offset - re_ctx.saved_p[i].buf_pos.offset; ecma_string_t *capture_str_p; if (capture_str_len > 0) { const lit_utf8_byte_t *utf8_str_p = re_ctx.saved_p[i].buf_p + re_ctx.saved_p[i].buf_pos.offset; capture_str_p = ecma_new_ecma_string_from_utf8 (utf8_str_p, capture_str_len); } else { capture_str_p = ecma_get_magic_string (LIT_MAGIC_STRING__EMPTY); } ecma_op_object_put (result_array_obj_p, index_str_p, ecma_make_string_value (capture_str_p), true); ecma_deref_ecma_string (capture_str_p); } else { ecma_op_object_put (result_array_obj_p, index_str_p, ecma_make_simple_value (ECMA_SIMPLE_VALUE_UNDEFINED), true); } ecma_deref_ecma_string (index_str_p); } ret_value = result_array; } else { ret_value = ecma_make_normal_completion_value (ecma_make_simple_value (ECMA_SIMPLE_VALUE_NULL)); } } MEM_FINALIZE_LOCAL_ARRAY (num_of_iter_p); MEM_FINALIZE_LOCAL_ARRAY (saved_p); MEM_FINALIZE_LOCAL_ARRAY (input_utf8_buffer_p); return ret_value; } /* ecma_regexp_exec_helper */
/* * Helper function for string indexOf and lastIndexOf functions * * This function implements string indexOf and lastIndexOf with required checks and conversions. * * See also: * ECMA-262 v5, 15.5.4.7 * ECMA-262 v5, 15.5.4.8 * * Used by: * - The String.prototype.indexOf routine. * - The String.prototype.lastIndexOf routine. * * @return uint32_t - (last)index of search string */ ecma_completion_value_t ecma_builtin_helper_string_prototype_object_index_of (ecma_value_t this_arg, /**< this argument */ ecma_value_t arg1, /**< routine's first argument */ ecma_value_t arg2, /**< routine's second argument */ bool firstIndex) /**< routine's third argument */ { ecma_completion_value_t ret_value = ecma_make_empty_completion_value (); /* 1 */ ECMA_TRY_CATCH (check_coercible_val, ecma_op_check_object_coercible (this_arg), ret_value); /* 2 */ ECMA_TRY_CATCH (to_str_val, ecma_op_to_string (this_arg), ret_value); /* 3 */ ECMA_TRY_CATCH (search_str_val, ecma_op_to_string (arg1), ret_value); /* 4 */ ECMA_OP_TO_NUMBER_TRY_CATCH (pos_num, arg2, ret_value); /* 6 */ ecma_string_t *original_str_p = ecma_get_string_from_value (to_str_val); const ecma_length_t original_len = ecma_string_get_length (original_str_p); const lit_utf8_size_t original_size = ecma_string_get_size (original_str_p); /* 4b, 5, 7 */ ecma_length_t start = ecma_builtin_helper_string_index_normalize (pos_num, original_len, firstIndex); /* 8 */ ecma_string_t *search_str_p = ecma_get_string_from_value (search_str_val); const ecma_length_t search_len = ecma_string_get_length (search_str_p); const lit_utf8_size_t search_size = ecma_string_get_size (search_str_p); ecma_number_t *ret_num_p = ecma_alloc_number (); *ret_num_p = ecma_int32_to_number (-1); /* 9 */ if (search_len <= original_len) { if (!search_len) { *ret_num_p = ecma_uint32_to_number (firstIndex ? 0 : original_len); } else { /* create utf8 string from original string and advance to position */ MEM_DEFINE_LOCAL_ARRAY (original_str_utf8_p, original_size, lit_utf8_byte_t); ecma_string_to_utf8_string (original_str_p, original_str_utf8_p, (ssize_t) (original_size)); lit_utf8_iterator_t original_it = lit_utf8_iterator_create (original_str_utf8_p, original_size); ecma_length_t index = start; lit_utf8_iterator_advance (&original_it, index); /* create utf8 string from search string */ MEM_DEFINE_LOCAL_ARRAY (search_str_utf8_p, search_size, lit_utf8_byte_t); ecma_string_to_utf8_string (search_str_p, search_str_utf8_p, (ssize_t) (search_size)); lit_utf8_iterator_t search_it = lit_utf8_iterator_create (search_str_utf8_p, search_size); /* iterate original string and try to match at each position */ bool searching = true; while (searching) { /* match as long as possible */ ecma_length_t match_len = 0; lit_utf8_iterator_t stored_original_it = original_it; while (match_len < search_len && index + match_len < original_len && lit_utf8_iterator_read_next (&original_it) == lit_utf8_iterator_read_next (&search_it)) { match_len++; } /* check for match */ if (match_len == search_len) { *ret_num_p = ecma_uint32_to_number (index); break; } else { /* inc/dec index and update iterators and search condition */ lit_utf8_iterator_seek_bos (&search_it); original_it = stored_original_it; if (firstIndex) { if ((searching = (index <= original_len - search_len))) { lit_utf8_iterator_incr (&original_it); index++; } } else { if ((searching = (index > 0))) { lit_utf8_iterator_decr (&original_it); index--; } } } } MEM_FINALIZE_LOCAL_ARRAY (search_str_utf8_p); MEM_FINALIZE_LOCAL_ARRAY (original_str_utf8_p); } } ecma_value_t new_value = ecma_make_number_value (ret_num_p); ret_value = ecma_make_normal_completion_value (new_value); ECMA_OP_TO_NUMBER_FINALIZE (pos_num); ECMA_FINALIZE (search_str_val); ECMA_FINALIZE (to_str_val); ECMA_FINALIZE (check_coercible_val); return ret_value; } /* ecma_builtin_helper_string_index_normalize */
/** * The String.prototype object's 'indexOf' routine * * See also: * ECMA-262 v5, 15.5.4.7 * * @return completion value * Returned value must be freed with ecma_free_completion_value. */ static ecma_completion_value_t ecma_builtin_string_prototype_object_index_of (ecma_value_t this_arg, /**< this argument */ ecma_value_t arg1, /**< routine's first argument */ ecma_value_t arg2) /**< routine's second argument */ { ecma_completion_value_t ret_value = ecma_make_empty_completion_value (); /* 1 */ ECMA_TRY_CATCH (check_coercible_val, ecma_op_check_object_coercible (this_arg), ret_value); /* 2 */ ECMA_TRY_CATCH (to_str_val, ecma_op_to_string (this_arg), ret_value); /* 3 */ ECMA_TRY_CATCH (search_str_val, ecma_op_to_string (arg1), ret_value); /* 4 */ ECMA_OP_TO_NUMBER_TRY_CATCH (pos_num, arg2, ret_value); /* 5 */ ecma_string_t *original_str_p = ecma_get_string_from_value (to_str_val); const ecma_length_t original_len = ecma_string_get_length (original_str_p); const lit_utf8_size_t original_size = ecma_string_get_size (original_str_p); /* 4b, 6 */ ecma_length_t start = ecma_builtin_helper_string_index_normalize (pos_num, original_len); /* 7 */ ecma_string_t *search_str_p = ecma_get_string_from_value (search_str_val); const ecma_length_t search_len = ecma_string_get_length (search_str_p); const lit_utf8_size_t search_size = ecma_string_get_size (search_str_p); ecma_number_t *ret_num_p = ecma_alloc_number (); *ret_num_p = ecma_int32_to_number (-1); /* 8 */ if (search_len <= original_len) { if (!search_len) { *ret_num_p = ecma_uint32_to_number (0); } else { /* create utf8 string from original string and advance to start position */ MEM_DEFINE_LOCAL_ARRAY (original_str_utf8_p, original_size, lit_utf8_byte_t); ecma_string_to_utf8_string (original_str_p, original_str_utf8_p, (ssize_t) (original_size)); lit_utf8_iterator_t original_it = lit_utf8_iterator_create (original_str_utf8_p, original_size); ecma_length_t index = start; lit_utf8_iterator_advance (&original_it, index); /* create utf8 string from search string */ MEM_DEFINE_LOCAL_ARRAY (search_str_utf8_p, search_size, lit_utf8_byte_t); ecma_string_to_utf8_string (search_str_p, search_str_utf8_p, (ssize_t) (search_size)); lit_utf8_iterator_t search_it = lit_utf8_iterator_create (search_str_utf8_p, search_size); /* iterate original string and try to match at each position */ bool found = false; while (!found && index <= original_len - search_len) { ecma_length_t match_len = 0; lit_utf8_iterator_pos_t stored_original_pos = lit_utf8_iterator_get_pos (&original_it); while (match_len < search_len && lit_utf8_iterator_read_next (&original_it) == lit_utf8_iterator_read_next (&search_it)) { match_len++; } /* Check for match */ if (match_len == search_len) { *ret_num_p = ecma_uint32_to_number (index); found = true; } else { /* reset iterators */ lit_utf8_iterator_seek_bos (&search_it); lit_utf8_iterator_seek (&original_it, stored_original_pos); lit_utf8_iterator_incr (&original_it); } index++; } MEM_FINALIZE_LOCAL_ARRAY (search_str_utf8_p); MEM_FINALIZE_LOCAL_ARRAY (original_str_utf8_p); } } ecma_value_t new_value = ecma_make_number_value (ret_num_p); ret_value = ecma_make_normal_completion_value (new_value); ECMA_OP_TO_NUMBER_FINALIZE (pos_num); ECMA_FINALIZE (search_str_val); ECMA_FINALIZE (to_str_val); ECMA_FINALIZE (check_coercible_val); return ret_value; } /* ecma_builtin_string_prototype_object_index_of */