Use offsetalator in nvtext::tokenize_with_vocabulary (#15878)

Updates the `token_counts_fn` kernel in the `nvtext::tokenize_with_vocabulary` to use the offsetalator instead of hardcoded `size_type` for accessing strings offsets.

Authors:
  - David Wendt (https://github.com/davidwendt)

Approvers:
  - Bradley Dice (https://github.com/bdice)
  - Karthikeyan (https://github.com/karthikeyann)

URL: #15878

cpp/src/text/vocabulary_tokenize.cu

@@ -240,10 +240,10 @@ CUDF_KERNEL void token_counts_fn(cudf::column_device_view const d_strings,
     return;
   }
 
-  auto const offsets =
-    d_strings.child(cudf::strings_column_view::offsets_column_index).data<cudf::size_type>();
-  auto const offset      = offsets[str_idx + d_strings.offset()] - offsets[d_strings.offset()];
-  auto const chars_begin = d_strings.data<char>() + offsets[d_strings.offset()];
+  auto const offsets     = d_strings.child(cudf::strings_column_view::offsets_column_index);
+  auto const offsets_itr = cudf::detail::input_offsetalator(offsets.head(), offsets.type());
+  auto const offset = offsets_itr[str_idx + d_strings.offset()] - offsets_itr[d_strings.offset()];
+  auto const chars_begin = d_strings.data<char>() + offsets_itr[d_strings.offset()];
 
   auto const begin        = d_str.data();
   auto const end          = begin + d_str.size_bytes();