Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills.
Python agent-contextagent-memoryagent-pluginsagentic-ragcontext-databasedsh-pluginknowledge-basememorymemory-managementself-evolvingvector-database
// Copyright (c) 2026 Beijing Volcano Engine Technology Co., Ltd.
// SPDX-License-Identifier: AGPL-3.0
#pragma once
#include <vector>
#include <algorithm>
#include <numeric>
#include <iterator>
#include <type_traits>
namespace vectordb {
// Helper traits to check for random access iterator
template <typename It>
constexpr bool is_random_access_iterator_v =
std::is_same_v<typename std::iterator_traits<It>::iterator_category,
std::random_access_iterator_tag>;
// Core ZipSort: Branch optimized version supporting two sequences.
template <typename Comparator, typename T, typename U>
void ZipSortBranchOptimized(Comparator comp, T begin, T end, U begin1, U end1) {
static_assert(is_random_access_iterator_v<T>,
"First iterator must be random access");
static_assert(is_random_access_iterator_v<U>,
"All rest iterators must be random access");
const size_t n = end - begin;
if (n <= 1)
return;
std::vector<size_t> indices(n);
std::iota(indices.begin(), indices.end(), 0);
std::sort(indices.begin(), indices.end(),
[&](const size_t& a, const size_t& b) noexcept {
return comp(*(begin + a), *(begin + b));
});
auto rearrange = [&](auto it) noexcept {
using ValueType = typename std::iterator_traits<decltype(it)>::value_type;
std::vector<ValueType> temp(n);
for (size_t i = 0; i < n; ++i) {
temp[i] = *(it + indices[i]);
}
std::copy(temp.begin(), temp.end(), it);
};
rearrange(begin);
rearrange(begin1);
}
} // namespace vectordb