Skip to content

fix: 키워드 페이징 조회를 단건 조회로 수정 #2216

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Soundbar91 merged 4 commits into from
Apr 8, 2026
Merged

fix: 키워드 페이징 조회를 단건 조회로 수정 #2216

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
6b47e27
fix: 페이징 조회를 단일 조회로 수정
Soundbar91 Apr 8, 2026
9514017
chore: 코드 포멧팅
Soundbar91 Apr 8, 2026
97bf224
fix: 미사용 리파지토리 메소드 삭제
Soundbar91 Apr 8, 2026
70b2b05
test: 테스트 코드 수정
Soundbar91 Apr 8, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions .../java/in/koreatech/koin/domain/community/keyword/repository/ArticleKeywordRepository.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,8 @@ public interface ArticleKeywordRepository extends Repository<ArticleKeyword, Int

List<ArticleKeyword> findAll(Pageable pageable);

List<ArticleKeyword> findAll();

@Query("""
SELECT new in.koreatech.koin.domain.community.article.dto.ArticleKeywordResult(k.id, k.keyword, COUNT(u))
FROM ArticleKeywordUserMap u
Expand Down
77 changes: 34 additions & 43 deletions src/main/java/in/koreatech/koin/domain/community/util/KeywordExtractor.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,15 +6,13 @@
import java.util.Map;
import java.util.stream.Collectors;

import org.springframework.data.domain.PageRequest;
import org.springframework.data.domain.Pageable;
import org.springframework.stereotype.Service;
import org.springframework.transaction.annotation.Transactional;

import in.koreatech.koin.common.event.ArticleKeywordEvent;
import in.koreatech.koin.domain.community.article.model.Article;
import in.koreatech.koin.domain.community.keyword.model.ArticleKeyword;
import in.koreatech.koin.domain.community.keyword.model.ArticleKeywordUserMap;
import in.koreatech.koin.common.event.ArticleKeywordEvent;
import in.koreatech.koin.domain.community.keyword.repository.ArticleKeywordRepository;
import in.koreatech.koin.domain.community.keyword.repository.ArticleKeywordUserMapRepository;
import lombok.RequiredArgsConstructor;
Expand All @@ -24,56 +22,49 @@
@Transactional(readOnly = true)
public class KeywordExtractor {

private static final int KEYWORD_BATCH_SIZE = 100;

private final ArticleKeywordRepository articleKeywordRepository;
private final ArticleKeywordUserMapRepository articleKeywordUserMapRepository;

public List<ArticleKeywordEvent> matchKeyword(List<Article> articles, Integer authorId) {
Map<Integer, Map<Integer, String>> matchedKeywordByUserIdByArticleId = new LinkedHashMap<>();
int offset = 0;
List<ArticleKeyword> keywords = articleKeywordRepository.findAll();

while (true) {
Pageable pageable = PageRequest.of(offset / KEYWORD_BATCH_SIZE, KEYWORD_BATCH_SIZE);
List<ArticleKeyword> keywords = articleKeywordRepository.findAll(pageable);
if (keywords.isEmpty()) {
return List.of();
}

if (keywords.isEmpty()) {
break;
}
List<Integer> keywordIds = keywords.stream()
.map(ArticleKeyword::getId)
.toList();
Map<Integer, List<ArticleKeywordUserMap>> userMapsByKeywordId = articleKeywordUserMapRepository
.findAllByArticleKeywordIdIn(keywordIds)
.stream()
.filter(keywordUserMap -> !keywordUserMap.getIsDeleted())
.collect(Collectors.groupingBy(
keywordUserMap -> keywordUserMap.getArticleKeyword().getId(),
LinkedHashMap::new,
Collectors.toList()
));
List<Integer> keywordIds = keywords.stream()
.map(ArticleKeyword::getId)
.toList();
Map<Integer, List<ArticleKeywordUserMap>> userMapsByKeywordId = articleKeywordUserMapRepository
.findAllByArticleKeywordIdIn(keywordIds)
.stream()
.filter(keywordUserMap -> !keywordUserMap.getIsDeleted())
.collect(Collectors.groupingBy(
keywordUserMap -> keywordUserMap.getArticleKeyword().getId(),
LinkedHashMap::new,
Collectors.toList()
));
Comment on lines +29 to +46
Copy link
Copy Markdown

@coderabbitai coderabbitai bot Apr 8, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

⚠️ Potential issue | 🟠 Major

Keep the paged scan on this request path.

This now loads the full ArticleKeyword table and then issues an unbounded findAllByArticleKeywordIdIn(keywordIds) over every keyword id before checking a few titles. src/main/java/in/koreatech/koin/domain/community/article/service/LostItemArticleService.java:173-186 and 254-261 call matchKeyword() during lost-item creation, so latency and memory now grow with total keyword volume instead of the current articles input. Please keep the batched/paged scan here, or move the full scan behind a cached/background path if this is only needed for testing.

🤖 Prompt for AI Agents 
Verify each finding against the current code and only fix it if needed.

In `@src/main/java/in/koreatech/koin/domain/community/util/KeywordExtractor.java`
around lines 29 - 46, The current implementation in KeywordExtractor loads all
ArticleKeyword rows and then calls
articleKeywordUserMapRepository.findAllByArticleKeywordIdIn over every id,
causing unbounded memory/latency when matchKeyword() is called during lost-item
creation; change this to a paged/batched scan: iterate ArticleKeyword pages (use
ArticleKeywordRepository pageable/streaming API) and for each page collect its
ids and call articleKeywordUserMapRepository.findAllByArticleKeywordIdIn for
just that batch (or use a repository method that accepts a page/stream), build
the userMapsByKeywordId incrementally (merging per-page results) and keep
filtering out deleted ArticleKeywordUserMap entries as before so memory and
latency scale with the input batch rather than the whole table.


for (Article article : articles) {
String title = article.getTitle();
for (ArticleKeyword keyword : keywords) {
if (!title.contains(keyword.getKeyword())) {
continue;
}
Map<Integer, String> matchedKeywordByUserId = matchedKeywordByUserIdByArticleId
.computeIfAbsent(article.getId(), ignored -> new LinkedHashMap<>());
Map<Integer, Map<Integer, String>> matchedKeywordByUserIdByArticleId = new LinkedHashMap<>();
for (Article article : articles) {
String title = article.getTitle();
for (ArticleKeyword keyword : keywords) {
if (!title.contains(keyword.getKeyword())) {
continue;
}
Map<Integer, String> matchedKeywordByUserId = matchedKeywordByUserIdByArticleId
.computeIfAbsent(article.getId(), ignored -> new LinkedHashMap<>());

for (ArticleKeywordUserMap keywordUserMap :
userMapsByKeywordId.getOrDefault(keyword.getId(), List.of())) {
Integer userId = keywordUserMap.getUser().getId();
matchedKeywordByUserId.merge(
userId,
keyword.getKeyword(),
this::pickHigherPriorityKeyword
);
}
for (ArticleKeywordUserMap keywordUserMap :
userMapsByKeywordId.getOrDefault(keyword.getId(), List.of())) {
Integer userId = keywordUserMap.getUser().getId();
matchedKeywordByUserId.merge(
userId,
keyword.getKeyword(),
this::pickHigherPriorityKeyword
);
}
}
offset += KEYWORD_BATCH_SIZE;
}

List<ArticleKeywordEvent> keywordEvents = new ArrayList<>();
Expand Down
19 changes: 7 additions & 12 deletions src/test/java/in/koreatech/koin/unit/domain/community/util/KeywordExtractorTest.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,6 @@
import org.mockito.InjectMocks;
import org.mockito.Mock;
import org.mockito.junit.jupiter.MockitoExtension;
import org.springframework.data.domain.Pageable;
import org.springframework.test.util.ReflectionTestUtils;

import in.koreatech.koin.common.event.ArticleKeywordEvent;
Expand Down Expand Up @@ -51,9 +50,8 @@ void matchKeyword_withMultipleMatchedKeywordsInSingleArticle_createsSingleEvent(
ArticleKeyword keywordA = createKeyword(1, "근로", subscriber);
ArticleKeyword keywordB = createKeyword(2, "근로장학", subscriber);

when(articleKeywordRepository.findAll(any(Pageable.class)))
.thenReturn(List.of(keywordA, keywordB))
.thenReturn(List.of());
when(articleKeywordRepository.findAll())
.thenReturn(List.of(keywordA, keywordB));
when(articleKeywordUserMapRepository.findAllByArticleKeywordIdIn(any()))
.thenReturn(List.of(
keywordA.getArticleKeywordUserMaps().get(0),
Expand All @@ -79,9 +77,8 @@ void matchKeyword_whenNoKeywordsMatch_returnsEmptyResult() {
User subscriber = UserFixture.id_설정_코인_유저(1);
ArticleKeyword keyword = createKeyword(1, "장학금", subscriber);

when(articleKeywordRepository.findAll(any(Pageable.class)))
.thenReturn(List.of(keyword))
.thenReturn(List.of());
when(articleKeywordRepository.findAll())
.thenReturn(List.of(keyword));
when(articleKeywordUserMapRepository.findAllByArticleKeywordIdIn(any()))
.thenReturn(List.of(keyword.getArticleKeywordUserMaps().get(0)));

Expand All @@ -106,9 +103,8 @@ void matchKeyword_withMultipleArticles_createsEventPerArticle() {
ArticleKeyword firstKeyword = createKeyword(1, "근로", firstSubscriber);
ArticleKeyword secondKeyword = createKeyword(2, "장학금", secondSubscriber);

when(articleKeywordRepository.findAll(any(Pageable.class)))
.thenReturn(List.of(firstKeyword, secondKeyword))
.thenReturn(List.of());
when(articleKeywordRepository.findAll())
.thenReturn(List.of(firstKeyword, secondKeyword));
when(articleKeywordUserMapRepository.findAllByArticleKeywordIdIn(any()))
.thenReturn(List.of(
firstKeyword.getArticleKeywordUserMaps().get(0),
Expand All @@ -128,8 +124,7 @@ void matchKeyword_withMultipleArticles_createsEventPerArticle() {
@DisplayName("등록된 키워드가 없으면 빈 결과를 반환한다.")
void matchKeyword_whenNoKeywordsExist_returnsEmptyResult() {
Article article = mock(Article.class);
when(article.getId()).thenReturn(1);
when(articleKeywordRepository.findAll(any(Pageable.class))).thenReturn(List.of());
when(articleKeywordRepository.findAll()).thenReturn(List.of());

List<ArticleKeywordEvent> result = keywordExtractor.matchKeyword(List.of(article), null);

Expand Down
Loading