Optimize hashing for case-insensitive strings

6e07ea929418b1fae4257a2af37aa0ed5799f22a

1 files changed, 18 insertions, 4 deletions

diff --git a/library/cpp/case_insensitive_string/case_insensitive_string.cpp b/library/cpp/case_insensitive_string/case_insensitive_string.cpp
index 16c0f5ff7a..dce0ff4af8 100644
--- a/library/cpp/case_insensitive_string/case_insensitive_string.cpp
+++ b/library/cpp/case_insensitive_string/case_insensitive_string.cpp
@@ -2,15 +2,29 @@
 
 #include <library/cpp/digest/murmur/murmur.h>
 
-size_t THash<TCaseInsensitiveStringBuf>::operator()(TCaseInsensitiveStringBuf str) const noexcept {
-    TMurmurHash2A<size_t> hash;
-    for (size_t i = 0; i < str.size(); ++i) {
-        char lower = std::tolower(str[i]);
+#include <array>
+
+static size_t HashTail(TMurmurHash2A<size_t>& hash, const char* data, size_t size) {
+    for (size_t i = 0; i < size; ++i) {
+        char lower = std::tolower(data[i]);
         hash.Update(&lower, 1);
     }
     return hash.Value();
 }
 
+size_t THash<TCaseInsensitiveStringBuf>::operator()(TCaseInsensitiveStringBuf str) const noexcept {
+    TMurmurHash2A<size_t> hash;
+    std::array<char, sizeof(size_t)> buf;
+    size_t headSize = str.size() - str.size() % buf.size();
+    for (size_t i = 0; i < headSize; i += buf.size()) {
+        for (size_t j = 0; j < buf.size(); ++j) {
+            buf[j] = std::tolower(str[i + j]);
+        }
+        hash.Update(buf.data(), buf.size());
+    }
+    return HashTail(hash, str.data() + headSize, str.size() - headSize);
+}
+
 template <>
 void Out<TCaseInsensitiveString>(IOutputStream& o, const TCaseInsensitiveString& p) {
     o.Write(p.data(), p.size());