I've read this spring-data-rest tutorial https://spring.io/guides/gs/accessing-data-rest/ and I can't see how this can be applied in a real world situation where we don't have just one object, but a graph of objects.
Let's say that we have an Order object which has one-to-many relationship to an Item which is categorized by a Category object. Let's say for the sake of it, that the Category is implemented in a tree-like structure (so it has a parent and some children; i.e. an Electronic category could have 2 children, Computer and TV, the former having another two children, motherboard and keyboards).
And let's say that all these relation are two ways (i.e. an Order can see it's items and an Item can access it's Order)
So when I request an Order object threw my REST service, I'm gone get the Order, all it's Items and each Item will have the whole graph of Category which will be linked to each Item and thus all the orders. So I'm basically returning the whole database.
I do understand that the bidirectional relations is not ideal but even if we suppress the many side of the relationship, when requesting an Order, we would still get
Order-Item-Category-Parent Category-Parent of Parent Category-etc...
So how do you stop a graph of objects being serialized?
Furthermore, you might not want to break the graph at a fix point.
For instance when I request an Order, I might want to see it's items and the category of each item, but definitely not the parent's category.
However, when I want to explicitly display a Category, I would then like to see it's parent. Get it?
Does somebody have some insight for me?
This is where DTOs are a good thing. :)
You build the object structure you need and return it via json. You have full control over this structure, you dont have bidirectional relations and you can just give that back as an object graph from your controller.
Another (good) side effect is that you decouple persistence and view completely. I.e both can evolve independently.
One (negative) side effect can be the increased maintenance. You have, at least, twice the classes to maintain and also the mapping in between. Frameworks like Dozer can help you with the mapping at least.
Another solution can be to implement some addons for Jackson that handle all these cases without the need of DTOs.
I myself created the Antpath filter to dynamically decide which path to filter out in Jackson:
https://github.com/Antibrumm/jackson-antpathfilter
But you will need some more things to consider if you like to work with domain entities directly.
cycle breaker (to avoid bidirectional relations for example)
lazy loading / open session in view pattern as the serialization can go everywhere
hibernate proxies which have method which are not serializable (session)
Related
Lets say we have a class "Car" than has different pieces of data ( maker, model, color, fabrication date, registration date, etc). The class has no method to get data, but it knows to as for it from another object (sent via constructor, let's cal it for short DS).- and the same for when needing to update changes.
A method getColor() would be implemented like this
if(! this->loaded('color')){
this->askDS('color') // this will do the necesarry work to generate a request to DS
}
return this->information('color');
Nothing too fancy so far. No comes the part i want to find out if it has a name, or if there are libraries / frameworks that do this already.
DS has a list of methods registered dinamically based on the class that needs data. For car we have:
input: car serial number, output: method to use to read the numbers to extract raw values
input: car raw color value, output: color code
input: car color code, manufacturer, year, mode, output:human-readable color (for example navy blue)
Now, DS or any method does not have an ordered list of using command to start from serial number and return the color blue, but if can construct a chain of methods that from one set of data, it can run them in order and get the desired data.
For our example above, DS runs 1,2,3 in that order and injects the data resulted from all methods into the class object that needed it.
Now if the car needs registration info, we have method (4) that gets that from the police database with an api request.
So, given:
- a type of model (class/object)
- a list of methods that take a fixed list of input(object properties) and give out a fixed list of output (object properties)
- a class DS that can glue the methods and run the needed ones for a model to get from property A (serial) to properby B (human readable colour) without the model or DS having a preconfigured way to get this data but finding it as needed.
does this have a name or is it already implemented somewhere ?
I've implemented a very basic prototype and it works very nice and i think this implementation method has useful features:
if you have a set of methods that do sql queries and then your app switches to using an api, you only need to change the methods and don't have to touch any other part of the application
when looking for a chain of methods that resolve the 'need' the object has, you can find a method chain, run it, if it fails keep looking for another list of methods based on the currently available data - so if you have multiple sources for a piece of data, it can try multiple versions
starting from the above paragraph i could start with an app that only has sql queries for data retrieval - when i find out a part of the app overloads the sql server i could add a method to retrieve data from cache with a lower cost than the one from database (or multiple layered caches, each with different costs)
i could probably add business logi in the mix the same ways as cache, and based on the user location / options present different data
this requires less coding overall, and decouples the data source from the object, making each piece easier to mock/test
what is needed to make this fast is a caching solution for the discovered method chains, since matching hundreds of thousands of methods per model type would be time-consuming but I don't think this is very hard to do - just store all found chains in memory as you find them and some metadata to be able to resume a search from any point in time - when you update the methods, just clear the cache, take a performance hit for the first requests
Thank you for your time
What you describe sounds like a somewhat roundabout way of doing Dependency Injection. Quote:
"Passing the service to the client, rather than allowing a client to
build or find the service, is the fundamental requirement of the
pattern."
Depending on what language you're using, there should be several Dependency Injection frameworks/libraries available.
I have developed a system where various classes have attributes consisting of a custom formula. The formula can contain special tokens which refer to different types of object. For example an object of class FruitSalad may have the following attribute;
$contents = "[A12] + [B76]";
In somewhat abstract terms, this means "add apple 12 to banana 76". It can also get significantly more complex than that with as many as 15 or 20 references to other objects involved in one formula.
I have a trait which passes formulae such as this and each time it finds a reference to a model (i.e. "[A12]") it gets it from the database with A::find(12) and adds it to an array of component objects which can be used for other processes later on in the request.
So, in essence, it's a relationship. But instead of a pivot table to describe the relationship, there is a formula on the parent model which can include references to child models.
This is all working. Yay! But it's really inefficient because there are so many tiny queries to get single models as formulae are parsed. One request may quite easily result in hundreds of queries. Oops.
I see two potential options;
1. Get all my apples and bananas from the database at the start of the request and get them from an in-memory store instead of from the database when parsing a formula (is this the repository pattern??).
2. Create a custom relation type (something like hasManyFromFormula) which makes eager loading work so that the parsing becomes much simpler because the relevant apples and bananas would already be loaded into the parent model.
Is there a precedent for this? As for why I am doing it like this, it would a bit tough to explain in brief but suffice to say it is to support a highly configurable data retrieval system which supports as-yet unknown input data configurations.
Help!
Thanks,
Geoff
Am not completely sure if it is the best solution, but in the end I created a new directory class for basic components and then set it up in the app service provider as a singleton. The constructor for the directory class loaded all models of several relevant classes and made them available as collections throughout the app.
I have a user which is a Marionette Model.
A user has a list of clients which I want to represent as a Collection.
A user has a currentClient which is a reference to one of the models in the clients Collection.
My question is, does it make sense to have the Collection of clients as an attribute of user or should I create a requestor for the client list, passing the user?
If it makes sense I'd rather keep clients as an attribute as I want to register events to alter other Views based on the data inside the currentClient when currentClient is reassigned. I don't know if this is the optimal way to handle this case.
As a best practice, I would say trying to keep models "flat" will prevent headaches later on, especially when it comes to operations like toJSON (the nested collection will only become an array of Backbone models). Changes inside vanilla objects and arrays nested in models cannot be tied to Backbone's eventing system easily.
Your description lists many "has-a" and "has-many" relationships. What's "optimal" very much depends on you app's architecture. Thankfully, Backbone has a pretty good plugin community to help you shape how you want your app to behave:
Backbone.Relational is a popular plugin for handling these types of relationships. It abstracts the nested collection use case with a bit of configuration.
I have this idea of generating an array of user-links that will depend on user-roles.
The user can be a student or an admin.
What I have in mind is use a foreach loop to generate a list of links that is only available for certain users.
My problem is, I created a helper class called Navigation, but I am so certain that I MUST NOT hard-code the links in there, instead I want that helper class to just read an object sent from somewhere, and then will return the desired navigation array to a page.
Follow up questions, where do you think should i keep the links that will only be available for students, for admins. Should i just keep them in a text-file?
or if it is possible to create a controller that passes an array of links, for example
a method in nav_controller class -> studentLinks(){} that will send an array of links to the helper class, the the helper class will then send it to the view..
Sorry if I'm quite crazy at explaining. Do you have any related resources?
From your description it seems that you are building some education-related system. It would make sense to create implementation in such way, that you can later expand the project. Seems reasonable to expect addition of "lectors" as a role later.
Then again .. I am not sure how extensive your knowledge about MVC design pattern is.
That said, in this situation I would consider two ways to solve this:
View requests current user's status from model layer and, based on the response, requests additional data. Then view uses either admin or user templates and creates the response.
You can either hardcode the specific navigation items in the templates, from which you build the response, or the lit of available navigation items can be a part of the additional information that you requested from model layer.
The downside for this method is, that every time you need, when you need to add another group, you will have to rewrite some (if not all) view classes.
Wrap the structures from model layer in a containment object (the basis of implementation available in this post), which would let you restrict, what data is returned.
When using this approach, the views aways request all the available information from model layer, but some of it will return null, in which case the template would not be applied. To implement this, the list of available navigation items would have to be provided by model layer.
P.S. As you might have noticed from this description, view is not a template and model is not a class.
It really depends on what you're already using and the scale of your project. If you're using a db - stick it there. If you're using xml/json/yaml/whatever - store it in a file with corresponding format. If you have neither - hardcode it. What I mean - avoid using multiple technologies to store data. Also, if the links won't be updated frequently and the users won't be able to customize them I'd hardcode them. There's no point in creating something very complex for the sake of dynamics if the app will be mostly static.
Note that this question doesn't quite fit in stackoverflow. programmers.stackexchange.com would probably be a better fit
I am creating a document based project using Core Data and have run into what may simply be a conceptual issue for me, as while I am not new to Cocoa, this is my first attempt to utilize Core Data. What I am trying to accomplish should be relatively simple: with each new document launched, I would like a new instance of one of my model objects created that serves as a "root" object.
What I have done is add an NSObjectController to my xib, set its mode to Entity Name (with the correct entity name provided), checked off "Prepares Content", and bound its managed object context to File's Owner with managedObjectContext as the model key path. To test this, I bound the title of my main window to the object controller, with controller key as selection and model key path as one of the keys in my entity.
I know I can create my root object programmatically, but am trying to adopt the mediator pattern as is recommended by Apple. I have seen the instructions in the department-employee tutorial under the "adopting the mediator pattern" section and the steps detailed are exactly what I believe I have done.
Any thoughts?
Edit:
Perhaps I did not state the problem correctly. The models are created in Core Data and the relationships are setup as I need them to be (with a "root", children and leaves, using to-one parent relationships, to-many children relationships and an isLeaf boolean attribute). My issue is ensuring that this root object is instantiated as a singleton every time a new document is launched. There should be exactly a 1:1 relationship between the root object and the current document, that root object must always exist and be available without any user interaction to create it, and child nodes that are created and attached to the root are the data objects that are used and manipulated by the application.
I have implemented the above functionality programatically, but in keeping with Core Data principles, would like to adopt the mediator pattern completely and not manage any creation of data objects within my application logic.
If you want a "root" managed object like you would find in linked-list or tree, then you have to set that up in data model itself.
By default, a Core Data data model has no particular hierarchy among objects. Objects may be related but no object is logically "above" or "below" another one. You can reach in object in any relationship by starting with any other object and walking the relationship/s back to the desired object.
A hierarchy of managed objects needs a tree like structure like this:
Tree{
nodeName:string
parent<-->>Tree.children
children<<-->Tree.parent
}
... so that the "root" object is the sole Tree instances that has parent==nil.
Having said all this, I would point out that the Apple docs you refer to say that it is best NOT to use this type of built in hierarchy for most cases. It's just a simplification used for purposes of demonstration (and I think it is a bad one.)
The data model is intended to model/simulate the real-world objects, conditions or events that the app deals with. As such, the logical relationships between the entities/objects in the model/graph should reflect the real-world relationships. In this case, unless the real-world things you are modeling exist in a hierarchy with a real-world "root" object, condition or event, then your model shouldn't have one either.