From 9735d726ad71dfad91ce987875ea0dc989d1cd04 Mon Sep 17 00:00:00 2001
From: Aaditya Srinivasan <aadityasri03@gmail.com>
Date: Wed, 15 Apr 2026 01:54:05 +0530
Subject: [PATCH 1/2] Optimize base64_decode validation using lookup table

---
 cpp/src/arrow/vendored/base64.cpp | 25 +++++++++++++++++--------
 1 file changed, 17 insertions(+), 8 deletions(-)

diff --git a/cpp/src/arrow/vendored/base64.cpp b/cpp/src/arrow/vendored/base64.cpp
index db2f74ed98fc..d5f4638e3c1c 100644
--- a/cpp/src/arrow/vendored/base64.cpp
+++ b/cpp/src/arrow/vendored/base64.cpp
@@ -30,6 +30,8 @@
 */
 
 #include "arrow/util/base64.h"
+#include <array>
+#include <cstdint>
 #include <iostream>
 
 namespace arrow {
@@ -40,6 +42,17 @@ static const std::string base64_chars =
              "abcdefghijklmnopqrstuvwxyz"
              "0123456789+/";
 
+static const std::array<int8_t, 256> kBase64Lookup = [] {
+  std::array<int8_t, 256> table{};
+  table.fill(-1);
+
+  for (size_t i = 0; i < base64_chars.size(); ++i) {
+    table[static_cast<uint8_t>(base64_chars[i])] = static_cast<int8_t>(i);
+  }
+
+  return table;
+}();
+
 static std::string base64_encode(unsigned char const* bytes_to_encode, unsigned int in_len) {
   std::string ret;
   int i = 0;
@@ -119,22 +132,18 @@ Result<std::string> base64_decode(std::string_view encoded_string) {
         return Status::Invalid("Invalid base64 input: padding in wrong position");
       }
 
-      if (base64_chars.find(c) == std::string::npos) {
+      int8_t val = kBase64Lookup[static_cast<uint8_t>(c)];
+
+      if (val == -1) {
         return Status::Invalid("Invalid base64 input: character is not valid base64 character");
       }
 
-      char_array_4[i++] = c;
+      char_array_4[i++] = val;
     }
 
     in_++;
 
     if (i == 4) {
-      for (i = 0; i < 4; i++) {
-        if (char_array_4[i] != 0) {
-          char_array_4[i] = base64_chars.find(char_array_4[i]) & 0xff;
-        }
-      }
-
       char_array_3[0] = (char_array_4[0] << 2) + ((char_array_4[1] & 0x30) >> 4);
       char_array_3[1] = ((char_array_4[1] & 0xf) << 4) + ((char_array_4[2] & 0x3c) >> 2);
       char_array_3[2] = ((char_array_4[2] & 0x3) << 6) + char_array_4[3];

From a4e7a084c77f170d19b7434ff4b2df702e30ca30 Mon Sep 17 00:00:00 2001
From: Aaditya Srinivasan <aadityasri03@gmail.com>
Date: Sat, 18 Apr 2026 11:27:29 +0530
Subject: [PATCH 2/2] Use constexpr std::string_view for base64_chars

---
 cpp/src/arrow/vendored/base64.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/cpp/src/arrow/vendored/base64.cpp b/cpp/src/arrow/vendored/base64.cpp
index d5f4638e3c1c..f7b105e53303 100644
--- a/cpp/src/arrow/vendored/base64.cpp
+++ b/cpp/src/arrow/vendored/base64.cpp
@@ -37,7 +37,7 @@
 namespace arrow {
 namespace util {
 
-static const std::string base64_chars =
+constexpr std::string_view base64_chars =
              "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
              "abcdefghijklmnopqrstuvwxyz"
              "0123456789+/";