[llvm] [ADT] Add TrieRawHashMap (PR #69528)

Justin Bogner via llvm-commits llvm-commits at lists.llvm.org
Sat Oct 26 17:40:55 PDT 2024


================
@@ -0,0 +1,122 @@
+//===- TrieHashIndexGenerator.h ---------------------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIB_SUPPORT_TRIEHASHINDEXGENERATOR_H
+#define LLVM_LIB_SUPPORT_TRIEHASHINDEXGENERATOR_H
+
+#include "llvm/ADT/ArrayRef.h"
+#include <optional>
+
+namespace llvm {
+
+/// The utility class that helps computing the index of the object inside trie
+/// from its hash. The generator can be configured with the number of bits
+/// used for each level of trie structure with \c NumRootsBits and \c
+/// NumSubtrieBits.
+/// For example, try computing indexes for a 16-bit hash 0x1234 with 8-bit root
+/// and 4-bit sub-trie:
+///
+///   IndexGenerator IndexGen{8, 4, Hash};
+///   size_t index1 = IndexGen.next(); // index 18 in root node.
+///   size_t index2 = IndexGen.next(); // index 3 in sub-trie level 1.
+///   size_t index3 = IndexGen.next(); // index 4 in sub-tire level 2.
+///
+/// This is used by different trie implementation to figure out where to
+/// insert/find the object in the data structure.
+struct IndexGenerator {
+  size_t NumRootBits;
+  size_t NumSubtrieBits;
+  ArrayRef<uint8_t> Bytes;
+  std::optional<size_t> StartBit = std::nullopt;
+
+  // Get the number of bits used to generate current index.
+  size_t getNumBits() const {
+    assert(StartBit);
+    size_t TotalNumBits = Bytes.size() * 8;
+    assert(*StartBit <= TotalNumBits);
+    return std::min(*StartBit ? NumSubtrieBits : NumRootBits,
+                    TotalNumBits - *StartBit);
+  }
+
+  // Get the index of the object in the next level of trie.
+  size_t next() {
+    if (!StartBit) {
+      // Compute index for root when StartBit is not set.
+      StartBit = 0;
+      return getIndex(Bytes, *StartBit, NumRootBits);
+    }
+    if (*StartBit < Bytes.size() * 8) {
+      // Compute index for sub-trie.
+      *StartBit += *StartBit ? NumSubtrieBits : NumRootBits;
+      assert((*StartBit - NumRootBits) % NumSubtrieBits == 0);
+      return getIndex(Bytes, *StartBit, NumSubtrieBits);
+    }
+    // All the bits are consumed.
+    return end();
+  }
+
+  // Provide a hint to speed up the index generation by providing the
+  // information of the hash in current level. For example, if the object is
+  // known to have \c Index on a level that already consumes first n \c Bits of
+  // the hash, it can start index generation from this level by calling \c hint
+  // function.
+  size_t hint(unsigned Index, unsigned Bit) {
+    assert(Bit < Bytes.size() * 8);
+    assert(Bit == 0 || (Bit - NumRootBits) % NumSubtrieBits == 0);
+    StartBit = Bit;
+    return Index;
+  }
+
+  // Utility funciton for looking up the index in the trie for an object that
----------------
bogner wrote:

```suggestion
  // Utility function for looking up the index in the trie for an object that
```

https://github.com/llvm/llvm-project/pull/69528


More information about the llvm-commits mailing list