From e4a4d863ab23066ccbd36f22f41119a524fad45e Mon Sep 17 00:00:00 2001
From: Advit Arora <advitarora2@gmail.com>
Date: Mon, 24 Aug 2026 20:43:20 +0530
Subject: [PATCH 1/3] GH-48977: [C++] Add a schema construction benchmark

---
 cpp/src/arrow/type_benchmark.cc | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)

diff --git a/cpp/src/arrow/type_benchmark.cc b/cpp/src/arrow/type_benchmark.cc
index f502c4b7c860..6544e3e4f990 100644
--- a/cpp/src/arrow/type_benchmark.cc
+++ b/cpp/src/arrow/type_benchmark.cc
@@ -170,6 +170,20 @@ static void SchemaEqualsWithMetadata(
   state.SetItemsProcessed(state.iterations() * 2);
 }
 
+static void SchemaConstruct(benchmark::State& state,  // NOLINT non-const reference
+                            bool duplicate_names) {
+  const int num_fields = static_cast<int>(state.range(0));
+  FieldVector fields(num_fields);
+  for (int i = 0; i < num_fields; ++i) {
+    fields[i] = field(duplicate_names ? "f" : "f" + std::to_string(i), int32());
+  }
+
+  for (auto _ : state) {
+    benchmark::DoNotOptimize(::arrow::schema(fields));
+  }
+  state.SetItemsProcessed(state.iterations() * num_fields);
+}
+
 // ------------------------------------------------------------------------
 // Micro-benchmark various error reporting schemes
 
@@ -541,6 +555,14 @@ BENCHMARK(TypeEqualsComplex);
 BENCHMARK(TypeEqualsWithMetadata);
 BENCHMARK(SchemaEquals);
 BENCHMARK(SchemaEqualsWithMetadata);
+BENCHMARK_CAPTURE(SchemaConstruct, distinct_names, /*duplicate_names=*/false)
+    ->Arg(100)
+    ->Arg(1000)
+    ->Arg(10000);
+BENCHMARK_CAPTURE(SchemaConstruct, duplicate_names, /*duplicate_names=*/true)
+    ->Arg(100)
+    ->Arg(1000)
+    ->Arg(10000);
 
 BENCHMARK(ErrorSchemeNoError);
 BENCHMARK(ErrorSchemeBool);

From 123aab5ba887ef03cbb372e19359977a828656a5 Mon Sep 17 00:00:00 2001
From: Advit Arora <advitarora2@gmail.com>
Date: Mon, 24 Aug 2026 20:43:20 +0530
Subject: [PATCH 2/3] GH-48977: [C++] Fix quadratic field name index
 construction on libc++

---
 cpp/src/arrow/type.cc | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/cpp/src/arrow/type.cc b/cpp/src/arrow/type.cc
index 1dedf1978dab..1de3f7ae7f9d 100644
--- a/cpp/src/arrow/type.cc
+++ b/cpp/src/arrow/type.cc
@@ -1333,8 +1333,10 @@ namespace {
 std::unordered_multimap<std::string_view, int> CreateNameToIndexMap(
     const FieldVector& fields) {
   std::unordered_multimap<std::string_view, int> name_to_index;
+  name_to_index.reserve(fields.size());
   for (size_t i = 0; i < fields.size(); ++i) {
-    name_to_index.emplace(fields[i]->name(), static_cast<int>(i));
+    const std::string_view name = fields[i]->name();
+    name_to_index.emplace_hint(name_to_index.find(name), name, static_cast<int>(i));
   }
   return name_to_index;
 }
@@ -2294,7 +2296,7 @@ Schema::Schema(FieldVector fields, std::shared_ptr<const KeyValueMetadata> metad
       impl_(new Impl(std::move(fields), Endianness::Native, std::move(metadata))) {}
 
 Schema::Schema(const Schema& schema)
-    : detail::Fingerprintable(), impl_(new Impl(*schema.impl_)) {}
+    : Schema(schema.impl_->fields_, schema.impl_->endianness_, schema.impl_->metadata_) {}
 
 Schema::~Schema() = default;
 
@@ -2568,7 +2570,9 @@ class SchemaBuilder::Impl {
   }
 
   Status AppendField(const std::shared_ptr<Field>& field) {
-    name_to_index_.emplace(field->name(), static_cast<int>(fields_.size()));
+    const std::string_view name = field->name();
+    name_to_index_.emplace_hint(name_to_index_.find(name), name,
+                                static_cast<int>(fields_.size()));
     fields_.push_back(field);
     return Status::OK();
   }

From 4cacc16929928103bb881859fdbd24571e5a7998 Mon Sep 17 00:00:00 2001
From: Advit Arora <advitarora2@gmail.com>
Date: Tue, 25 Aug 2026 10:15:16 +0530
Subject: [PATCH 3/3] GH-48977: [C++] Comment the hinted-insert motivation

---
 cpp/src/arrow/type.cc | 1 +
 1 file changed, 1 insertion(+)

diff --git a/cpp/src/arrow/type.cc b/cpp/src/arrow/type.cc
index 1de3f7ae7f9d..e841f0918097 100644
--- a/cpp/src/arrow/type.cc
+++ b/cpp/src/arrow/type.cc
@@ -1336,6 +1336,7 @@ std::unordered_multimap<std::string_view, int> CreateNameToIndexMap(
   name_to_index.reserve(fields.size());
   for (size_t i = 0; i < fields.size(); ++i) {
     const std::string_view name = fields[i]->name();
+    // The find() hint avoids libc++'s quadratic scan of equal keys on plain emplace.
     name_to_index.emplace_hint(name_to_index.find(name), name, static_cast<int>(i));
   }
   return name_to_index;