I have 2 domain classes, User and Dog (for example)
class User {
String id
Map<String, Dog> dogs
}
class Dog {
String name
}
My Controller get as an input a json
{"key" : "dogKey", "userId" : "someId", "dogName" : "dog"}
def addDog(){
String key = request.JSON.key
User user = User.get(request.JSON.userId)
String dogName = request.JSON.dog
...
if(! user.dogs){
user.dogs = new HashMap<>(1)
}
user.dogs.put(key, new Dog(name: dogName))
user.save(flush: true)
}
after running the user data # Mongo is:
user:
{ _id:....,
dogs: {
"dogKey": null
}...
}
can someone please explain me what i'm missing?
Thanks!
Roy
may be dog reference is not save in database
def addDog(){
String key = request.JSON.key
User user = User.get(request.JSON.userId)
String dogName = request.JSON.dog
...
if(! user.dogs){
user.dogs = new HashMap<>(1)
}
Dog dog = new Dog(name: dogName)
dog.save(flush:true)
user.dogs.put(key, dog)
user.save(flush: true)
}
Related
I am trying to create indexes on a mongodb collection using mongock version 4.3.8. The mongockLock and mongockChangeLog collections are eventually created, but the migration itself does not run. mongockChangeLog table is empty. Query db.BOOKS.getIndexes(); returns a single string: with an index for id (the value of name in it is equal to _id_), which is most likely generated automatically. There are no obvious errors in the logs either.
Everything else seems to be set up correctly, I don't understand why my migration is not running.
build.gradle.kts:
implementation("com.github.cloudyrock.mongock:mongodb-springdata-v3-driver:4.3.8")
implementation("com.github.cloudyrock.mongock:mongock-spring-v5:4.3.8")
implementation(platform("com.github.cloudyrock.mongock:mongock-bom:4.3.8"))
application.yml:
mongock:
change-logs-scan-package:
- com.dreamland.configuration.migration
config file:
#EnableMongock
#Configuration
#EnableMongoAuditing
class MongoConfig
Migration file:
#ChangeLog(order = "1668471203")
class IndexChangeLog {
companion object : KLogging()
#ChangeSet(order = "1668471204", id = "1668471204_create_indexes", author = "Irish")
fun createIndexes(mongockTemplate: MongockTemplate) {
val indexOps = mongockTemplate.indexOps(BookObjectDocument::class.java)
Index().on("createdAt", Sort.Direction.DESC).let { indexOps.ensureIndex(it) }.also { log(it) }
Index().on("createdBy", Sort.Direction.ASC).let { indexOps.ensureIndex(it) }.also { log(it) }
Index().on("type", Sort.Direction.DESC).let { indexOps.ensureIndex(it) }.also { log(it) }
}
private fun log(indexName: String) {
logger.info { "Index '$indexName' was created successfully" }
}
}
Document class:
#Document(collection = "BOOKS_V2")
data class BookObjectDocument(
val type: BookObjectType?,
val description: String?
) {
#Id
lateinit var id: String
#CreatedDate
lateinit var createdAt: Instant
#CreatedBy
var createdBy: String? = null
}
I am trying to query two separate objects and return them as a single result set. I've tried using a Union and an Interface, but not sure what I'm doing wrong there.
My model:
class BaseActivity(models.Model):
class Meta:
abstract = True
name = models.CharField(db_column="activity_type_name", unique=True, max_length=250)
created_at = CreationDateTimeField()
modified_at = ModificationDateTimeField()
created_by = models.UUIDField()
modified_by = models.UUIDField()
deleted_at = models.DateTimeField(blank=True, null=True)
deleted_by = models.UUIDField(blank=True, null=True)
class Activity(BaseActivity):
class Meta:
db_table = "activity_type"
ordering = ("sort_order",)
id = models.UUIDField(
db_column="activity_type_id",
primary_key=True,
default=uuid.uuid4,
editable=False,
)
sort_order = models.IntegerField()
def __str__(self):
return self.name
class CustomActivity(BaseActivity):
class Meta:
db_table = "organization_custom_activity_type"
id = models.UUIDField(
db_column="organization_custom_activity_type",
primary_key=True,
default=uuid.uuid4,
editable=False,
)
farm = models.ForeignKey("farm.Farm", db_column="organization_id", on_delete=models.DO_NOTHING, related_name="custom_activity_farm")
My schema:
class FarmActivities(graphene.ObjectType):
id = graphene.String()
name = graphene.String()
class ActivityType(DjangoObjectType):
class Meta:
model = Activity
fields = ("id", "name", "requires_crop", "sort_order")
class CustomActivityType(DjangoObjectType):
class Meta:
model = CustomActivity
fields = ("id", "name", "farm")
And the query:
class Query(graphene.ObjectType):
get_farm_activities = graphene.Field(FarmActivities, farm=graphene.String(required=True))
def resolve_get_farm_activities(self, info, farm):
farm_activities = Activity.objects.values("id", "name").filter(
farmtypeactivityrel__farm_type__farm=farm, deleted_at=None
)
custom_activities = CustomActivity.objects.values("id", "name").filter(farm=farm, deleted_at=None)
return list(chain(farm_activities, custom_activities))
With this, I do get a list back from the query, but it's not going thru the resolver when I call getFarmActivities.
Literally the list returns:
ExecutionResult(data={'getFarmActivities': {'id': None, 'name': None}}, errors=None)
How to resolve graphene.Union Type?
That provided the hint I needed to get this working. I had to build a Union that would parse the model and not the schema type.
class FarmActivities(graphene.Union):
class Meta:
types = (ActivityType, CustomActivityType)
#classmethod
def resolve_type(cls, instance, info):
if isinstance(instance, Activity):
return ActivityType
if isinstance(instance, CustomActivity):
return CustomActivityType
return FarmActivities.resolve_type(instance, info)
Which then allowed me to run the query like so:
def resolve_get_farm_activities(self, info, farm):
farm_activities = Activity.objects.filter(
farmtypeactivityrel__farm_type__farm=farm
)
custom_activities = CustomActivity.objects.filter(farm=farm)
return list(chain(farm_activities, custom_activities))
And the API query:
query farmParam($farm: String!)
{
getFarmActivities(farm: $farm)
{
... on CustomActivityType
{
id
name
}
... on ActivityType
{
id
name
}
}
}
Say: there is already a given schema definition object:
const schema = { prop1: { type: String, maxLength: 8 }, prop2... };
Is it possible that: without declare the interface for each schema object, make a respective class which can produce documents with prop1:string, prop2... extracted from the schema.
I expect something like this in my app:
// schema definition:
const PersonSchema = { name: { type: String, maxLength: 8 } };
// class factory
const PersonClass = SchemaClassFactory(PersonSchema);
// instance with props defined in schema.
let person1 = new PersonClass();
person1.name = 'Jack';
let person2 = new PersonClass();
person2.name = 3; // should be wrong hinted by tslint.
How can I achieve that?
You can create a class for the schema object using a mapped type and conditional types to extract the shape of the object from the schema.
A possible solution is below, I am not sure I covered all the ways you can defined the schema in mongoose, but this should get you stared:
const PersonSchema = {
name: { type: String, maxLength: 8 },
age: { type: Number },
title: String,
id: ObjectID
};
type PrimitiveConstructor<T> = {
new (...a: any[]): any;
(...a: any[]): T
}
type Constructor<T> = {
new (...a: any[]): T;
}
type ExtractType<T> = {
[P in keyof T] :
T[P] extends PrimitiveConstructor<infer U>? U :
T[P] extends { type: PrimitiveConstructor<infer U> } ? U:
T[P] extends Constructor<infer U> ? U :
T[P] extends { type: Constructor<infer U> } ? U:
never
}
function createClass<T>(schema: T): new (data?: Partial<ExtractType<T>>) => ExtractType<T> {
// The class will not have the fields explicitly defined since we don't know them but that is fine
return new class {
// Optional constructor for assigning data to the fields, you can remove this if not needed
constructor(data?: any){
if(data) {
Object.assign(this, data);
}
}
} as any;
}
var PersonClass = createClass(PersonSchema);
type PersonClass = InstanceType<typeof PersonClass>
let p = new PersonClass();
p.name ="";
p.name = 2; // error
p.id = new ObjectID(10);
let p2 = new PersonClass({
name: "",
});
I've problems updating a domain class. I'm using Grails 3.0.9 and MongoDB (for Gorm 5.0.0.RC1)
In my build.gradle:
compile "org.grails.plugins:mongodb:5.0.0.RC1"
compile "org.mongodb:mongodb-driver:3.0.2"
compile "org.grails:grails-datastore-gorm-mongodb:5.0.0.RC1"
runtime 'org.springframework.data:spring-data-mongodb:1.8.1.RELEASE'
compile("org.grails:gorm-mongodb-spring-boot:5.0.0.RC1")
The test:
#Integration
class CompanyControllerIntegrationSpec extends Specification{
def grailsApplication
Company company
RestBuilder rest
def setupData() {
company = Company.buildWithoutSave().save(flush: true, failOnError: true)
}
def setup(){
rest = new RestBuilder()
}
def "test update a company" (){
def company2
given:
setupData()
def id = company.id
when:
RestResponse response = rest.put("http://localhost:${grailsApplication.config.server.port}/${company.companyKey}/company") {
json {
name = "newName"
description = "new Description"
}
}
company2 = Company.findById(id)
then:
response.status == 200
response.json.name == "newName"
company2.name == "newName"
company2.description == "new Description"
}
def cleanup() {
Company.collection.remove(new BasicDBObject())
}
}}
The controller:
class CompanyController extends ExceptionController{
static allowedMethods = ['update':'PUT','show':'GET',
'updateNew':'PUT','showNew':'GET']
CompanyService companyService
def update(String companyKey){
def object = request.JSON?request.JSON:params
Company companyOut = companyService.update(object, companyKey)
render text:companyOut as JSON, status:HttpStatus.OK
}
}
The service:
class CompanyService {
def securityService
def grailsApplication
public Company update(object, String companyKey) throws ForbiddenException, InvalidRequestException, NotFoundException{
Company company = findByKey(companyKey)
if (object.name!=null)
company.name = object.name
if (object.description!=null)
company.description = object.description
if (object.enterprise!=null)
company.enterprise = object.enterprise
if (object.genKey!=null)
company.companyKey = UUID.randomUUID().toString()
if (!company.save(flush:true)){
println company.errors
throw new InvalidRequestException("Some parameters are missing or are invalid: "+company.errors.fieldErrors.field)
}
return company
}
public Company findByKey(String companyKey) throws NotFoundException, ForbiddenException {
if (!companyKey){
throw new ForbiddenException("The company key has not been given")
}
Company company = Company.findByCompanyKey(companyKey)
if (!company){
throw new NotFoundException("No company exists for the given key")
}
return company
}
}
The results of the test are:
- response.status is 200
- response.json.name is "newName"
- company.name is old name ("company 1")
If I don't do the cleanup, the database still have the old value. I've followed the save method, also inside Mongo gorm classes, and I've seen that one problem is that the fields are not being marked as dirty, but don't know why.
With other Domain classes that are similar to this one, the update is done without problems and the properties are marked as dirty.
I have a requirement where I have to check in the database if the value of a boolean variable(crawled) is false. If so I have to set it to true. I am finding the record based on the value of a string variable(website). I referenced this link but it didn't help. Please tell me what I am doing wrong.
I tried this:
def p = Website.findByWebsite(website);
if(p['crawled'] == true) {
println "already crawled"
} else {
mongo.website.update( p, new BasicDBObject( '$set', new BasicDBObject( 'crawled', 'false' ) ) )
println "updated object"
}
It gives me an error No such property: mongo for class: cmsprofiler.ResourceController
My domain class is as follows:
class Website{
String website
User user
Boolean crawled
static belongsTo = [user: User]
static constraints = {
website( url:true, unique: ['user'])
}
static hasMany = [resource: Resource]
static mapping = {resource cascade:"all-delete-orphan" }
}
you should be using
Website.mongo.update(...)
or let the framework inject it:
class ResourceController {
def mongo
def list(){
mongo.website.update(...)
}
}
This worked for me. Posting it here in case anyone else has similar requirement.
#Grab(group='com.gmongo', module='gmongo', version='0.9.3')
import com.gmongo.GMongo
#Transactional(readOnly = true)
class ResourceController {
def mongo = new GMongo()
def db = mongo.getDB("resources")
def p = Website.findByWebsite(website)
if(p['crawled']==false){
db.website.update([crawled:false],[$set:[crawled:true]])
}