Improve performance for replace-multi for long strings (#12858)

Adds more efficient algorithm for multi-string version of `cudf::strings::replace` for longer strings (greater than 256 bytes on average in each row).

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

Approvers:
  - Mike Wilson (https://github.com/hyperbolic2346)
  - Yunsong Wang (https://github.com/PointKernel)
  - Bradley Dice (https://github.com/bdice)

URL: #12858

cpp/CMakeLists.txt

@@ -547,6 +547,7 @@ add_library(
   src/strings/regex/regex_program.cpp
   src/strings/repeat_strings.cu
   src/strings/replace/backref_re.cu
+  src/strings/replace/multi.cu
   src/strings/replace/multi_re.cu
   src/strings/replace/replace.cu
   src/strings/replace/replace_re.cu

cpp/benchmarks/string/replace.cpp

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2023, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -69,7 +69,7 @@ static void generate_bench_args(benchmark::internal::Benchmark* b)
   int const row_mult   = 8;
   int const min_rowlen = 1 << 5;
   int const max_rowlen = 1 << 13;
-  int const len_mult   = 4;
+  int const len_mult   = 2;
   generate_string_bench_args(b, min_rows, max_rows, row_mult, min_rowlen, max_rowlen, len_mult);
 }