Customizing queries with Spring Data Neo4j

Hantsy
4 min readNov 12, 2020

--

In the former post, we’ve created a simple project and used the declaration approach to create a Repository to access the Neo4j server. Similar to other Spring Data modules, Spring Data Neo4j provides a ReactiveNeo4jOperations for reactive applications to interact with Neo4j servers by a programmatic approach, and additionally, Spring Data Neo4j provides a low-level API abstraction to execute the Cypher Query Language through the ReactiveNeo4jClient bean.

This is the 4th post of the new Spring Data Neo4j series:

All source codes are hosted on my Github.

Conventional derived query methods

Like other Spring Data modules, Spring Data Neo4j also supports derived query methods.

For example, to find the posts by keyword that matches the title field, add the following method in the PostRepository interface.

interface PostRepository extends ReactiveNeo4jRepository<Post, Long> {    Flux<Post> findByTitleLike(String title);
}

Add a new test method in PostRespositoryTest to verify it.

@Test
void testFindByTitle() {
posts.findByTitleLike("one")
.as(StepVerifier::create)
.consumeNextWith(p -> assertEquals("Post one", p.getTitle()))
.verifyComplete();
}

Run the test, it should work as expected.

Customizing Query by @Query annotation

Like the @Query in Spring Data JPA, Spring Data Neo4j provides its own @Query annotation to attach the result of a custom Cypher Query to the method.

import org.springframework.data.neo4j.repository.query.Query;
//...

interface PostRepository extends ReactiveNeo4jRepository<Post, Long> {
@Query("MATCH(post:Post) WHERE post.title =~ $title RETURN post")
Flux<Post> findByTitleContains(String title);

//
...
}

Here we use a regex pattern in the where clause.

More details about the syntax of Cypher Query Language, please check the official Neo4j documentation.

Add a test method to verify it.

@Test
void testFindByQuery() {
posts.findByTitleContains("(?i).*" + "one" + ".*")
.as(StepVerifier::create)
.consumeNextWith(p -> assertEquals("Post one", p.getTitle()))
.verifyComplete();
}

Here the findByTitleContains method has to accept a Regex pattern.

ReactiveNeo4jClient

Once Spring Data Neo4j is configured in a reactive application, a ReactvieNeo4jClient bean is available in the Spring application context.

Like the R2dbc’s DatabaseClient , with ReactiveNeo4jClient, you can execute custom Cypher Queries and handle returning result freely.

For example, to find all posts, it can be done by the following method.

public Flux<Post> findAll() {
var query = """
MATCH (p:Post)
RETURN p.id as id, p.title as title, p.content as content, p.createdAt as createdAt, p.updatedAt as updatedAt
""";
return client
.query(query)
.fetchAs(Post.class).mappedBy((ts, r) ->
Post.builder()
.id(r.get("id").asLong())
.title(r.get("title").asString())
.content(r.get("content").asString())
.createdAt(r.get("createdAt").asLocalDateTime(null))
.updatedAt(r.get("updatedAt").asLocalDateTime(null))
.build()
)
.all();
}

In the above codes.

  • The query use a multi-lined text block(available in the latest Java 15) to define a Cypher query.
  • The client.query to execute the defined query.
  • The fetchAs to handle the returning result, similar to RowMapper in Jdbc/R2dbc to extract the result and wrap it into a POJO class.
  • The all will return a Flux , if you want to return a single result, use one instead.

The following is an example of basic CRUD operations.

@Component
@RequiredArgsConstructor
public class PostRepository {
private final ReactiveNeo4jClient client; public Mono<Long> count() {
var query = """
MATCH (p:Post) RETURN count(p)
""";
return client.query(query)
.fetchAs(Long.class)
.mappedBy((ts, r) -> r.get(0).asLong())
.one();
}
public Flux<Post> findAll() {
var query = """
MATCH (p:Post)
RETURN p.id as id, p.title as title, p.content as content, p.createdAt as createdAt, p.updatedAt as updatedAt
""";
return client
.query(query)
.fetchAs(Post.class).mappedBy((ts, r) ->
Post.builder()
.id(r.get("id").asLong())
.title(r.get("title").asString())
.content(r.get("content").asString())
.createdAt(r.get("createdAt").asLocalDateTime(null))
.updatedAt(r.get("updatedAt").asLocalDateTime(null))
.build()
)
.all();
}
public Flux<Post> findByTitleContains(String title) {
var query = """
MATCH (p:Post)
WHERE p.title =~ $title
RETURN p.id as id, p.title as title, p.content as content, p.createdAt as createdAt, p.updatedAt as updatedAt
""";
return client
.query(query)
.bind("(?!).*" + title + ".*").to("title")
.fetchAs(Post.class).mappedBy((ts, r) ->
Post.builder()
.id(r.get("id").asLong())
.title(r.get("title").asString())
.content(r.get("content").asString())
.createdAt(r.get("createdAt").asLocalDateTime(null))
.updatedAt(r.get("updatedAt").asLocalDateTime(null))
.build()
)
.all();
}
public Mono<Post> findById(Long id) {
var query = """
MATCH (p:Post)
WHERE p.id = $id
RETURN p.id as id, p.title as title, p.content as content, p.createdAt as createdAt, p.updatedAt as updatedAt
""";
return client
.query(query)
.bind(id).to("id")
.fetchAs(Post.class).mappedBy((ts, r) ->
Post.builder()
.id(r.get("id").asLong())
.title(r.get("title").asString())
.content(r.get("content").asString())
.createdAt(r.get("createdAt").asLocalDateTime(null))
.updatedAt(r.get("updatedAt").asLocalDateTime(null))
.build()
)
.one();
}
public Mono<Post> save(Post post) {
var query = """
MERGE (p:Post {id: $id})
ON CREATE SET p.createdAt=localdatetime(), p.title=$title, p.content=$content
ON MATCH SET p.updatedAt=localdatetime(), p.title=$title, p.content=$content
RETURN p.id as id, p.title as title, p.content as content, p.createdAt as createdAt, p.updatedAt as updatedAt
""";
return client.query(query)
.bind(post).with(data ->
Map.of(
"id", (data.getId() != null ? data.getId() : UUID.randomUUID().toString()),
"title", data.getTitle(),
"content", data.getContent()
)
)
.fetchAs(Post.class).mappedBy((ts, r) ->
Post.builder()
.id(r.get("id").asLong())
.title(r.get("title").asString())
.content(r.get("content").asString())
.createdAt(r.get("createdAt").asLocalDateTime(null))
.updatedAt(r.get("updatedAt").asLocalDateTime(null))
.build()
)
.one();
}
public Mono<Integer> deleteAll() {
var query = """
MATCH (m:Post) DETACH DELETE m
""";
return client.query(query)
.run()
.map(it -> it.counters().nodesDeleted());
} public Mono<Integer> deleteById(Long id) {
var query = """
MATCH (p:Post) WHERE p.id = $id
DETACH DELETE p
""";
return client
.query(query)
.bind(id).to("id")
.run()
.map(it -> it.counters().nodesDeleted());
}
}

Note: Please navigate to the data-neo4j and boot-neo4j-cypher repositories to check the above example codes.

ReactiveNeo4jOperations

Like other Spring Data modules, Spring Data Neo4j provides a ReactiveNeo4jOperations(and the implementation ReactiveNeo4jTemplate), it allows your to perform operations on Neo4j databases but by programmatic approaches.

Here is an example of PostRepository which is reimplemented by ReactiveNeo4jOperations.

@Component
@RequiredArgsConstructor
public class PostRepository {
private final ReactiveNeo4jOperations template;
public Mono<Long> count() {
return this.template.count(Post.class);
}
public Flux<Post> findAll() {
return this.template.findAll(Post.class);
}
public Mono<Post> findById(Long id) {
return this.template.findById(id, Post.class);
}
public Flux<Post> findByTitleContains(String title) {
var postNode = node("Post").named("p");
return this.template.findAll(
match(postNode)
.where(postNode.property("title").contains(literalOf(title)))
.returning(postNode)
.build(),
Post.class
);
}
public Mono<Post> save(Post post) {
return this.template.save(post);
}
public Flux<Post> saveAll(List<Post> data) {
return this.template.saveAll(data);
}
public Mono<Void> deleteById(Long id) {
return this.template.deleteById(id, Post.class);
}
public Mono<Void> deleteAll() {
return this.template.deleteAll(Post.class);
}
}

It it similar to the ReactiveNeo4jClient, but more simple. Have a look at the findAll, the literal queries are replaced by Java Query Criteria APIs.

Note: Please navigate to the data-neo4j and boot-neo4j repositories to check the above example codes.

Grab the source code from my github.

--

--

Hantsy

Self-employed technical consultant, solution architect and full-stack developer, open source contributor, freelancer and remote worker