mandlm/VectorSearch
Archived
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Compare commits

base: mandlm:532e52cc7f035fefb6fd435e9f9b7dcc85f08487
Branches Tags
mandlm:main
..
compare: mandlm:1441a869c2388eca1b7e073c97947e962d1d5c2a
Branches Tags
mandlm:main

2 Commits
532e52cc7f ... 1441a869c2

Author SHA1 Message Date
Michael Mandl
1441a869c2
refactor: clean up 2024-03-23 12:22:53 +01:00
Michael Mandl
bc58ca5af8
docs: Bucket 2024-03-23 12:01:46 +01:00
2 changed files with 15 additions and 5 deletions
Show Stats Expand all files Collapse all files

6
lib_vector_search/include/bucket_finder.h
View File

@ -5,6 +5,12 @@
#include <unordered_map>
/** A Bucket contains a partial WordRefList, split-up and (hash-)mapped by their
* first characters.
*
* It's meant to be read and written by a single thread without the need for
* synchronization.
*/
class Bucket {
private:
std::unordered_map<char, WordRefList> directory_;

14
lib_vector_search/src/bucket_finder.cpp
View File

@ -7,6 +7,8 @@
#include <thread>
#include <vector>
// class Bucket
void Bucket::insert(const WordList &word_list, size_t first_index,
size_t last_index) {
for (auto index = first_index; index < last_index; ++index) {
@ -31,6 +33,8 @@ WordRefList Bucket::find_prefix(std::string_view search_term) const {
return result;
}
// class BucketFinder
BucketFinder::BucketFinder(const WordList &word_list) { insert(word_list); }
void BucketFinder::insert(const WordList &word_list) {
@ -47,7 +51,7 @@ void BucketFinder::insert(const WordList &word_list) {
std::vector<std::thread> insert_threads;
for (auto bucket_index = 0; bucket_index < bucket_count; ++bucket_index) {
auto &bucket = buckets_[bucket_index];
auto &thread_bucket = buckets_[bucket_index];
const bool is_last_bucket = bucket_index == bucket_count - 1;
const size_t first_word_index = bucket_index * bucket_size;
@ -55,7 +59,7 @@ void BucketFinder::insert(const WordList &word_list) {
is_last_bucket ? word_list_size : first_word_index + bucket_size;
insert_threads.emplace_back([&, first_word_index, last_word_index] {
bucket.insert(word_list, first_word_index, last_word_index);
thread_bucket.insert(word_list, first_word_index, last_word_index);
});
}
@ -68,9 +72,9 @@ WordRefList BucketFinder::find_prefix(std::string_view search_term) const {
WordRefList search_results;
std::mutex search_results_mutex;
std::vector<std::thread> threads;
std::vector<std::thread> search_threads;
for (const auto &bucket : buckets_) {
threads.emplace_back([&] {
search_threads.emplace_back([&] {
auto thread_search_results = bucket.find_prefix(search_term);
if (!thread_search_results.empty()) {
const std::lock_guard result_lock(search_results_mutex);
@ -80,7 +84,7 @@ WordRefList BucketFinder::find_prefix(std::string_view search_term) const {
});
}
for (auto &thread : threads) {
for (auto &thread : search_threads) {
thread.join();
}