I have some trouble with inheritence and ElasticsearchRepository.
My code (short version) :
Entities :
public abstract class father{
#Id
protected String identifier;
// constructors
#Field(type= FieldType.String, index = FieldIndex.not_analyzed)
public String uri;
// getter setter
}
/*
* Stored in elastic search
*/
#Document(indexName = "entity1", type = "entity1")
public abstract class sonA extends father{
// constructors
}
/*
* Stored in elastic search
*/
#Document(indexName = "entity2", type = "entity2")
public abstract class sonB extends father{
// constructors
}
Repository :
public interface DataBnfRepository extends
ElasticsearchRepository<SonA, String> {
public SonA findByUri(String uri);
#Query("{ \"filtered\":{ \"filter\":{\"term\":{\"uri\":\"?0\"}}}}")
public SonA findWithUri(String uri);
}
My trouble : I'm able to put data into elastic search but not to retreive them.
If I do a repository.findall() it works.
If I do a repository.findWithUri(uri) it does not work (null result)
Parse exception for findByUri
I tried to search for exemples :
https://github.com/spring-projects/spring-data-elasticsearch/blob/master/src/test/java/org/springframework/data/elasticsearch/entities/SampleInheritedEntity.java
But there's no repository.
What I have found: If i replace the uri by"foo", it works. So the problem is with https://www.elastic.co/guide/en/elasticsearch/reference/current/query-dsl-query-string-query.html#_reserved_characters
So i made a test with
uri = foo:bar => parse exception.
uri = "foo:bar" => parse exception.
uri = foo\:bar => parse exception.
uri = "foo\:bar" => parse exception
Before it was inherited I had no problems with the code.
I just put the URI in an object, repository.save(object) and repository.findWithUri(uri), it give me the object.
Any help/suggestion/comments is greatly appreciated. Please. Thank you so much
Try the following method.
findByFatherUri(uri);
This is how it works for Child Entities. I think it's the same for Extended Classes as well. From Spring Data's point of view, main class is SonA for DataBnfRepository and it's id is the one linked to the Repository when you use findOne method. You have to traverse through the child class to the field you are trying use to find the object. Since, I am not sure 100%, let me know if this works once you tried.
Related
Let me explain my problem with SpringData mongo, I have the following interface declared, I declared a custom query, with a projection to ignore the index, this example is only for illustration, in real life I will ignore a bunch of fields.
public interface MyDomainRepo extends MongoRepository<MyDomain, String> {
#Query(fields="{ index: 0 }")
MyDomain findByCode(String code);
}
In my MongoDB instance, the MyDomain has the following info, MyDomain(code="mycode", info=null, index=19), so when I use the findByCode from MyDomainRepo I got the following info MyDomain(code="mycode", info=null, index=null), so far so good, because this is expected behaviour, but the problem happens when..., I decided to save the findByCode return.
For instance, in the following example, I got the findByCode return and set the info property to myinfo and I got the object bellow.
MyDomain(code="mycode", info="myinfo", index=null)
So I used the save from MyDomainRepo, the index was ignored as expected by the projection, but, when I save it back, with or without an update, the SpringData Mongo, overridden the index property to null, and consequently, my record on the MongoDB instance is overridden too, the following example it's my MongoDB JSON.
{
"_id": "5f061f9011b7cb497d4d2708",
"info": "myinfo",
"_class": "io.springmongo.models.MyDomain"
}
There's a way to tell to SpringData Mongo, to simply ignores the null fields on saving?
Save is a replace operation and you won't be able to signal it to patch some fields. It will replace the document with whatever you send
Your option is to use the extension provided by Spring Data Repository to define custom repository methods
public interface MyDomainRepositoryCustom {
void updateNonNull(MyDomain myDomain);
}
public class MyDomainRepositoryImpl implements MyDomainRepositoryCustom {
private final MongoTemplate mongoTemplate;
#Autowired
public BookRepositoryImpl(MongoTemplate mongoTemplate) {
this.mongoTemplate = mongoTemplate;
}
#Override
public void updateNonNull(MyDomain myDomain) {
//Populate the fileds you want to patch
Update update = Update.update("key1", "value1")
.update("key2", "value2");
// you can you Update.fromDocument(Document object, String... exclude) to
// create you document as well but then you need to make use of `MongoConverter`
//to convert your domain to document.
// create `queryToMatchId` to mtach the id
mongoTemplate.updateFirst(queryToMatchId, update, MyDomain.class);
}
}
public interface MyDomainRepository extends MongoRepository<..., ...>,
MyDomainRepositoryCustom {
}
I'm new to GraphQL and I'm currently implementing a GraphQL API into an established Java code, using GraphQL-SPQR and I'm running into a couple issues when it comes extracting data from hierarchical classes.
The issues that I am running into are as follows.
Firstly I don't if there is an easy way to get all the data associated with a returned node. If there is, this would be most useful for my more complex classes.
Secondly when a method returns an abstract class, I only seem able to request the variables on the abstract class. I'm sure this should be possible I am just hitting my head against a wall.
As a simple example
public abstract class Animal {
private String name;
private int age;
// Constructor
#GraphQLQuery(name = "name")
public String getName() {
return name;
}
// Age getter
}
public class Dog extends Animal {
private String favouriteFood;
// Constructor
#GraphQLQuery(name = "favouriteFood")
public String getFavouriteFood() {
return favouriteFood;
}
}
public class Database {
#GraphQLQuery(name = "getanimal")
public Animal getAnimal(#GraphQLArgument(name = "animalname") String animalname) {
return database.get(name);
}
}
So in my first question what I am currently querying is.
"{animalname(name: \"Daisy\") {name age}}"
This works fine as expected. If you imagine the class however had 10 variables I would like to merely be able to write the equivalent of the following without having to look them up.
"{node(name: \"Daisy\") {ALL}}"
Is this possible?
In terms of my second question.
The follow query, throws an error ('Field 'favouriteFood' in type 'Animal' is undefined')
"{animalname(name: \"Bones\") {name age favouriteFood}}"
likewise (reading Inline Fragments of https://graphql.org/learn/queries/)
"{animalname(name: \"Bones\") {name age ... on Dog{favouriteFood}}}"
throws an error Unknown type Dog
This is annoying as I have a number of sub classes which could be returned and may require handling in different fashions. I think I can understand why this is occuring as GraphQL has no knowledge as to what the true class is, only the super class I have returned. However I'm wondering if there is a way to fix this.
Ultimately while I can get past both these issues by simply serialising all the data to JSON and sending it back, it kind of gets rid of the point of GraphQL and I would rather find an alternate solution.
Thank you for any response.
Apologies if these are basic questions.
Answering my own question to help anyone else who has this issue.
The abstract class needs to have #GraphQLInterface included, as shown below
#GraphQLInterface(name = "Animal ", implementationAutoDiscovery = true)
public abstract class Animal {
private String name;
private int age;
// Constructor
#GraphQLQuery(name = "name")
public String getName() {
return name;
}
// Age getter
}
The following code was found after much solution and was created by the creator of SPQR. Effectively, when setting up your schema you need to declare an interface mapping strategy. The code below can be copied wholesale with only the "nodeQuery" variable being replaced with the service you are using to containing your "#GraphQLQuery" and "#GraphQLMutation" methods.
final GraphQLSchema schema = new GraphQLSchemaGenerator()
.withInterfaceMappingStrategy(new InterfaceMappingStrategy() {
#Override
public boolean supports(final AnnotatedType interfase) {
return interfase.isAnnotationPresent(GraphQLInterface.class);
}
#Override
public Collection<AnnotatedType> getInterfaces(final AnnotatedType type) {
Class clazz = ClassUtils.getRawType(type.getType());
final Set<AnnotatedType> interfaces = new HashSet<>();
do {
final AnnotatedType currentType = GenericTypeReflector.getExactSuperType(type, clazz);
if (supports(currentType)) {
interfaces.add(currentType);
}
Arrays.stream(clazz.getInterfaces())
.map(inter -> GenericTypeReflector.getExactSuperType(type, inter))
.filter(this::supports).forEach(interfaces::add);
} while ((clazz = clazz.getSuperclass()) != Object.class && clazz != null);
return interfaces;
}
}).withOperationsFromSingleton(nodeQuery)// register the service
.generate(); // done ;)
graphQL = new GraphQL.Builder(schema).build();
As this solution took some hunting, I'm going to start a blog soon with the other solutions I've stumbled on.
With regards to having a query that just returns all results. This is not possible in GraphQL. One workaround I might write is to have a endpoint that returns JSON of the entire object and the name of the object, then I can just use ObjectMapper to convert it back.
I hope this helps other people. I'm still looking into an answer for my first question and will update this post when I find one.
How do I set the MongoDB Document field name dynamically (without using #Field)?
#Document
public class Account {
private String username;
}
For example, field names should be capitalized. Result:
{"USERNAME": "hello"}
And I want this dynamic converter to work with any document, so a solution without using generics.
This a bit strange requirement. You can make use of Mongo Listener Life cycle events docs.
#Component
public class MongoListener extends AbstractMongoEventListener<Account> {
#Override
public void onBeforeSave(BeforeSaveEvent<Account> event) {
DBObject dbObject = event.getDBObject();
String username = (String) dbObject.get("username");// get the value
dbObject.put("USERNAME", username);
dbObject.removeField("username");
// You need to go through each and every field recursively in
// dbObject and then remove the field and then add the Field you
// want(with modification)
}
}
This is a bit cluncky, but I believe there is no clean way to do this.
NOTE: Go down in order to see the edited message.
I'm trying to imitate this query:
db.sentiments.aggregate([
{"$group" : {_id:{theme_id:"$theme",sentiment_id:"$sentiment"}, count:{$sum:1}}},
{"$sort":{"_id.theme_id":1}} ])
This is the code that I had generated in order to imitate it:
#RepositoryRestResource(collectionResourceRel = "sentiments", path = "sentiments")
public interface SentimentsRepository extends MongoRepository<Sentiments, String> {
Long countByTheme(#Param("theme") String theme);
#Query(value ="[\n" +
" {\"$group\" : {_id:{theme_id:\"$theme\",sentiment_id:\"$sentiment\"}, count:{$sum:1}}},\n" +
"\t{\"$sort\":{\"_id.theme_id\":1}}\n" +
"]",count = true)
List<Object> comptarSentiments();
}
Well this code is returning me this error:
"exception": "org.springframework.data.mongodb.UncategorizedMongoDbException",
"message": "Can't canonicalize query: BadValue unknown operator: $group; nested exception is com.mongodb.MongoException: Can't canonicalize query: BadValue unknown operator: $group",
Actually I'm a begginer in what refers to the use of Spring so I'm very lost, does any one know what should I do?
Thanks and sorry for my bad english, not my native language.
[EDIT]----------------------------------------
Just as the comment wrote by Shawn Clark It's not possible to do it this way, in order to achieve that you will need to create a customRepository.
What's the difference between Spring Data's MongoTemplate and MongoRepository?
I have been trying to do it this way but something doesn't seem to be correct, here is my new code:
#RepositoryRestResource(collectionResourceRel = "sentiments", path = "sentiments")
public interface SentimentsRepository extends CrudRepository<Sentiments, String>, CustomSentimentsRepository {
//Other methods...
}
public interface CustomSentimentsRepository {
List<CountResult> yourCustomMethod();
class CountResult{
String theme;
String sentiment;
int total;
}
}
public class SentimentsRepositoryImpl implements CustomSentimentsRepository {
private final MongoOperations operations;
#Autowired
public SentimentsRepositoryImpl(MongoOperations operations) {
Assert.notNull(operations, "MongoOperations must not be null!");
this.operations = operations;
}
#Override
public List<CountResult> yourCustomMethod(){
Aggregation agg = Aggregation.newAggregation(
Aggregation.group("theme","sentiment").count().as("total"),
Aggregation.project("theme","sentiment").and("total").previousOperation(),
Aggregation.sort(Sort.Direction.DESC, "theme")
);
//Convert the aggregation result into a List
AggregationResults<CountResult> groupResults
= operations.aggregate(agg,"sentiments", CountResult.class);
//List<CountResult> result = groupResults.getMappedResults();
return groupResults.getMappedResults();
}
}
I'm not even able to debbug this code and I'm always getting a 404.
Based on the information I have found you can't do that complex of a #Query on a MongoRepository method. In this case you would want to create a class and implement your comptarSentiments() method using the mongoTemplate to query the data store with your aggregate function. Then create a controller class that exposes a REST endpoint and have it call the repository.
Once you get to doing complex queries in Mongo you lose the ease of #RepositoryRestResource and have to go back to wiring the REST endpoint to the repository yourself.
Spring Data REST : custom query for MongoDB repository
Implementing custom methods of Spring Data repository and exposing them through REST
I finally managed to solve the problem, seems like it was related with the controller and the type of the atribute "total" from the innerClass CountResult, it needs to be a String (this is very important, otherwise the Aggregation.project will fail). Here goes the final code:
public interface CustomSentimentsRepository {
List<CountResult> myCountGroupByThemeAndSentiment();
class CountResult{
public String theme;
public String sentiment;
public String total;
}
}
public class SentimentsRepositoryImpl implements CustomSentimentsRepository {
private final MongoTemplate mongoTemplate;
#Autowired
public SentimentsRepositoryImpl(MongoTemplate mongoTemplate) {
this.mongoTemplate = mongoTemplate;
}
#Override
public List<CountResult> myCountGroupByThemeAndSentiment(){
Aggregation agg = Aggregation.newAggregation(
Aggregation.group("theme","sentiment").count().as("total"),
Aggregation.project("theme","sentiment").andInclude("total"),
Aggregation.sort(Sort.Direction.ASC,"theme","sentiment")
);
AggregationResults<CountResult> groupResults
= mongoTemplate.aggregate(agg,"sentiments", CountResult.class);
return groupResults.getMappedResults();
}
}
#RepositoryRestResource(collectionResourceRel = "sentiments", path = "sentiments")
public interface SentimentsRepository extends CrudRepository<Sentiments, String>, CustomSentimentsRepository {
//Other methods
}
#RestController
#RequestMapping(value = "sentiments/search")
public class ChartsController {
#Autowired
private SentimentsRepository sentimentsRepository;
#RequestMapping(value = "myCountGroupByThemeAndSentiment", method = RequestMethod.GET)
public ResponseEntity<?> yourCustomMethod() {
List<?> count=sentimentsRepository.myCountGroupByThemeAndSentiment();
return new ResponseEntity(count, HttpStatus.OK);
}
}
You can use #Aggrgation available in spring data mongodb 2.2.X versions:
#Aggregation(pipeline = {"{ '$group': { '_id' : '$lastname', names : { $addToSet : '$?0' } } }", "{ '$sort' : { 'lastname' : -1 } }"}) List<PersonAggregate> groupByLastnameAnd(String property);
How can I parameterize a SpringData ElasticSearch index at runtime?
For example, the data model:
#Document(indexName = "myIndex")
public class Asset {
#Id
public String id;
// ...
}
and the repository:
public interface AssetRepository extends ElasticsearchCrudRepository<Asset, String> {
Asset getAssetById(String assetId);
}
I know I can replace myIndex with a parameter, but that parameter will be resolved during instantiation / boot. We have the same Asset structure for multiple clients / tenants, which have their own index. What I need is something like this:
public interface AssetRepository extends ElasticsearchCrudRepository<Asset, String> {
Asset getAssetByIdFromIndex(String assetId, String index);
}
or this
repoInstance.forIndex("myOtherIndex").getAssetById("123");
I know this does not work out of the box, but is there any way to programmatically 'hack' it?
Even though the bean is init at boot time, you can still achieve it by spring expression language:
#Bean
Name name() {
return new Name();
}
#Document(indexName="#{name.name()}")
public class Asset{}
You can change the bean's property to change the index you want to save/search:
assetRepo.save(new Asset(...));
name.setName("newName");
assetRepo.save(new Asset(...));
What should be noticed is not to share this bean in multiple thread, which may mess up your index.
Here is a working example.
org.springframework.data.elasticsearch.repository.ElasticSearchRepository has a method
FacetedPage<T> search(SearchQuery searchQuery);
where SearchQuery can take multiple indices to be used for searching.
I hope it answers