From 45583443c3f8ed8e93f382724ce1d154b55436e2 Mon Sep 17 00:00:00 2001
From: David Wendt <45795991+davidwendt@users.noreply.github.com>
Date: Wed, 29 Nov 2023 13:39:39 -0500
Subject: [PATCH] Remove deprecated nvtext::load_merge_pairs_file (#14460)

Remove deprecated `nvtext::load_merge_pairs_file` since it is not needed -- in favor `nvtext::load_merge_pairs` which accepts a strings column. Callers can use cuIO functions to load the file into a column.

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

Approvers:
  - Yunsong Wang (https://github.com/PointKernel)
  - Karthikeyan (https://github.com/karthikeyann)

URL: https://github.com/rapidsai/cudf/pull/14460
---
 cpp/include/nvtext/byte_pair_encoding.hpp | 37 --------------
 cpp/src/text/bpe/load_merge_pairs.cu      | 62 -----------------------
 2 files changed, 99 deletions(-)

diff --git a/cpp/include/nvtext/byte_pair_encoding.hpp b/cpp/include/nvtext/byte_pair_encoding.hpp
index 632a3cc279f..f9790a1a701 100644
--- a/cpp/include/nvtext/byte_pair_encoding.hpp
+++ b/cpp/include/nvtext/byte_pair_encoding.hpp
@@ -64,43 +64,6 @@ struct bpe_merge_pairs {
   bpe_merge_pairs();
 };
 
-/**
- * @brief Create a nvtext::bpe_merge_pairs from an input file.
- *
- * @deprecated Since 23.12
- *
- * The file should contain a pair of strings per line separated by
- * a single space.
- *
- * Example:
- * @code{.txt}
- * e n
- * i t
- * i s
- * e s
- * en t
- * c e
- * es t
- * en ce
- * T h
- * Th is
- * t est
- * s ent
- * ...
- * @endcode
- *
- * The pairs are expected to be ordered in the file by their rank
- * relative to each other. A pair earlier in the file has priority over
- * any pairs below it.
- *
- * @param filename_merges Local file path of pairs encoded in UTF-8.
- * @param mr Memory resource to allocate any returned objects.
- * @return A nvtext::bpe_merge_pairs object
- */
-[[deprecated]] std::unique_ptr<bpe_merge_pairs> load_merge_pairs_file(
-  std::string const& filename_merges,
-  rmm::mr::device_memory_resource* mr = rmm::mr::get_current_device_resource());
-
 /**
  * @brief Create a nvtext::bpe_merge_pairs from a strings column
  *
diff --git a/cpp/src/text/bpe/load_merge_pairs.cu b/cpp/src/text/bpe/load_merge_pairs.cu
index 80073df5804..6d223a7ddb7 100644
--- a/cpp/src/text/bpe/load_merge_pairs.cu
+++ b/cpp/src/text/bpe/load_merge_pairs.cu
@@ -37,53 +37,6 @@ namespace nvtext {
 namespace detail {
 namespace {
 
-/**
- * @brief Loads a text file of merge-pairs into a strings column.
- *
- * The line position in the file indicates the pair's rank.
- *
- * @code{.pseudo}
- * Format of the file:
- * #version ..
- * a1 a2
- * b1 b2
- * c1 c2
- * ...
- * @endcode
- *
- * @param filename_merges Path to text file containing merge-pairs
- * @return object containing table elements for the BPE function
- */
-std::unique_ptr<cudf::column> load_file_to_column(std::string const& filename_merges,
-                                                  rmm::cuda_stream_view stream,
-                                                  rmm::mr::device_memory_resource* mr)
-{
-  std::ifstream merges_file(filename_merges);
-  CUDF_EXPECTS(merges_file.good(), "Could not open " + filename_merges);
-
-  std::vector<char> chars{};
-  std::vector<cudf::size_type> offsets(1, 0);
-
-  std::string line;
-  std::getline(merges_file, line);
-  std::string version = "#version";
-  if (line.substr(0, version.size()).compare(version) == 0) { std::getline(merges_file, line); }
-
-  // This is a text file delimited only by CR/LF.
-  // TODO: Look into using the CSV reader to load the strings column instead.
-  while (!line.empty()) {
-    chars.insert(chars.end(), std::cbegin(line), std::cend(line));
-    offsets.push_back(offsets.back() + line.length());
-    std::getline(merges_file, line);
-  }
-
-  CUDF_EXPECTS(!chars.empty(), "No data found in " + filename_merges);
-
-  auto d_chars   = cudf::detail::make_device_uvector_async(chars, stream, mr);
-  auto d_offsets = cudf::detail::make_device_uvector_async(offsets, stream, mr);
-  return cudf::make_strings_column(d_chars, d_offsets, {}, 0);
-}
-
 std::unique_ptr<detail::merge_pairs_map_type> initialize_merge_pairs_map(
   cudf::column_device_view const& input, rmm::cuda_stream_view stream)
 {
@@ -146,14 +99,6 @@ std::unique_ptr<bpe_merge_pairs::bpe_merge_pairs_impl> create_bpe_merge_pairs_im
 
 }  // namespace
 
-std::unique_ptr<bpe_merge_pairs> load_merge_pairs_file(std::string const& filename_merges,
-                                                       rmm::cuda_stream_view stream,
-                                                       rmm::mr::device_memory_resource* mr)
-{
-  auto input_column = load_file_to_column(filename_merges, stream, mr);
-  return std::make_unique<bpe_merge_pairs>(std::move(input_column), stream, mr);
-}
-
 std::unique_ptr<bpe_merge_pairs> load_merge_pairs(cudf::strings_column_view const& merge_pairs,
                                                   rmm::cuda_stream_view stream,
                                                   rmm::mr::device_memory_resource* mr)
@@ -165,13 +110,6 @@ std::unique_ptr<bpe_merge_pairs> load_merge_pairs(cudf::strings_column_view cons
 
 }  // namespace detail
 
-std::unique_ptr<bpe_merge_pairs> load_merge_pairs_file(std::string const& filename_merges,
-                                                       rmm::mr::device_memory_resource* mr)
-{
-  CUDF_FUNC_RANGE();
-  return detail::load_merge_pairs_file(filename_merges, cudf::get_default_stream(), mr);
-}
-
 std::unique_ptr<bpe_merge_pairs> load_merge_pairs(cudf::strings_column_view const& merge_pairs,
                                                   rmm::cuda_stream_view stream,
                                                   rmm::mr::device_memory_resource* mr)