Improve base-32 hash decoding performance with reverse map

The changes include:

* Defining nix32Chars as a constexpr char[].
* Adding a constexpr std::array<unsigned char, 256> (reverseNix32Map) to map characters to their base-32 digit values at compile time.
* Replacing the slow character search loop with a direct lookup using reverseNix32Map.
* Removing std::once_flag/isBase32 logic in references.cc in favor of reverseNix32Map

Signed-off-by: Alexander V. Nikolaev <avn@avnik.info>

Author: avnik

src/libutil/hash.cc

@@ -72,7 +72,22 @@ static std::string printHash16(const Hash & hash)
 }
 
 // omitted: E O U T
-const std::string nix32Chars = "0123456789abcdfghijklmnpqrsvwxyz";
+constexpr char nix32Chars[] = "0123456789abcdfghijklmnpqrsvwxyz";
+
+constexpr std::array<unsigned char, 256> makeReverseNix32Map()
+{
+    std::array<unsigned char, 256> map{};
+
+    for (size_t i = 0; i < map.size(); ++i)
+        map[i] = 0xFF; // invalid
+
+    for (unsigned char i = 0; i < 32; ++i)
+        map[static_cast<unsigned char>(nix32Chars[i])] = i;
+
+    return map;
+}
+
+constexpr const std::array<unsigned char, 256> reverseNix32Map = makeReverseNix32Map();
 
 static std::string printHash32(const Hash & hash)
 {
@@ -217,12 +232,11 @@ Hash::Hash(std::string_view rest, HashAlgorithm algo, bool isSRI)
 
         for (unsigned int n = 0; n < rest.size(); ++n) {
             char c = rest[rest.size() - n - 1];
-            unsigned char digit;
-            for (digit = 0; digit < nix32Chars.size(); ++digit) /* !!! slow */
-                if (nix32Chars[digit] == c)
-                    break;
-            if (digit >= 32)
-                throw BadHash("invalid base-32 hash '%s'", rest);
+            unsigned char digit = reverseNix32Map[static_cast<unsigned char>(c)];
+
+            if (digit == 0xFF)
+                throw BadHash("invalid base-32 hash: '%s'", rest);
+
             unsigned int b = n * 5;
             unsigned int i = b / 8;
             unsigned int j = b % 8;

src/libutil/include/nix/util/hash.hh

@@ -35,7 +35,7 @@ constexpr inline size_t regularHashSize(HashAlgorithm type)
 
 extern const StringSet hashAlgorithms;
 
-extern const std::string nix32Chars;
+extern const std::array<unsigned char, 256> reverseNix32Map;
 
 /**
  * @brief Enumeration representing the hash formats.

src/libutil/references.cc

@@ -13,20 +13,11 @@ static size_t refLength = 32; /* characters */
 
 static void search(std::string_view s, StringSet & hashes, StringSet & seen)
 {
-    static std::once_flag initialised;
-    static bool isBase32[256];
-    std::call_once(initialised, []() {
-        for (unsigned int i = 0; i < 256; ++i)
-            isBase32[i] = false;
-        for (unsigned int i = 0; i < nix32Chars.size(); ++i)
-            isBase32[(unsigned char) nix32Chars[i]] = true;
-    });
-
     for (size_t i = 0; i + refLength <= s.size();) {
         int j;
         bool match = true;
         for (j = refLength - 1; j >= 0; --j)
-            if (!isBase32[(unsigned char) s[i + j]]) {
+            if (reverseNix32Map[(unsigned char) s[i + j]] == 0xFF) {
                 i += j + 1;
                 match = false;
                 break;