It is my first time to use yii and unlike my old programming style, i notice that it use relationship automatically in its model.
public function relations()
{
return array(
'author'=>array(self::BELONGS_TO, 'User', 'author_id'),
'categories'=>array(self::MANY_MANY, 'Category',
'tbl_post_category(post_id, category_id)'),
);
}
I'm not used in doing this MySQL relationship. my old programming habit is connecting/manipulating the data to the php program itself.. To clarify my question, is this yii model relationship important? if i dont use this method, will i encounter problems?
Yii relations are very useful and if you work with it you will see that it will make you do less coding and make your code more readable.
while it is so much used in Yii applications, if you don't use relations, you won't get into any trouble, it is supposed to help you code and develop faster.
like if you looked at Yii blog, you have relation between Post model and Comments model, and you could go like this:
$post = Post::model()->findByPk( $id ); // find one post
$allCommentsRelated = $post->comments; // just one line for all search query and instanciating models
BTW in relations, there are two type of loading:
lazy loading (this is default mechanism)
eager loading
you have to know your scenario, and choose one that suites that scenario best
Related
I'm new to MVC 3 and Entity Framework so I'd like to know what is the best approach.
Basically, I have 3 entities: Course, Module and Chapter. Each is a parent of the next with a one to many relationship (A Course has many Modules and a Module has many Chapters). I have a column SortOrder for Modules and Chapters to have them ordered sequentially.
My idea was is to use partial views for the child entities when updating the parent.
I have 3 views in mind:
Create/Update Course: all basic details for a course
Course Modules (basically a different view for Update Course) which has an option to add multiple partial views, each creating a Module
Course Timeline (still a different view for update course) which lists all Modules (on separate divs) and has the option to add multiple partial views, each creating a Chapter
Does my plan sound right and plausible? I plan to use hidden fields to store IDs. I also want the saves to occur asynchronously.
Any piece of advise or information would be highly appreciated. Thanks!
I think this is what your after but not sure. For handling persistence of child/grandchild entities, you can do this in several ways. You can either perform crud operations on each entity separately. So that will involve for example saving the modules by themselves with a reference to the course, probably courseId.
Or you can look at saving just the aggregate root, which in this case looks like its your Course entity. This will involve Loading the course, populating the modules on the course, and for each module populate the chapters. Then when you `db.Courses.Add(newCourse); db.SaveChanges(); all the entities will be persisted. You have to make sure your foreign key and model references are setup correctly.
For example, to save child entities:
public ActionResult DoSomething(int courseId, Module newModule)
{
var course = someService.LoadCourse(courseId);
course.Modules.Add(newModule);
using (var db = new MyDbContext())
{
db.Courses.Add(course);
db.SaveChanges();
}
return RedirectToAction("Success");
}
Or you can save individually:
public ActionResult DoSomething(Module newModule)
{
using (var db = new MyDbContext())
{
//You will need to make sure newModule.CourseId is set correctly
db.Modules.Add(newModule);
db.SaveChanges();
}
return RedirectToAction("Success");
}
Depending on your views, you will be able to judge which way is best to go. Regarding asynchronous saving, you will be able to call these endpoints with jquery posting the models as json. On a side note, one thing to look at would be to create a custom Anti Forgery Token validator for json requests, example.
I've googled a lot and read as much as I can on the subject but I cannot find a direct answer to my question anywhere including here. My ZF2 application consists of roughly 7 different modules. 5 of the modules will need access to the same database configuration. The application itself has a database with roughly 124 different tables. So the idea here is to find the proper solution to write the least amount of code considering the setup.
What I'm trying to do is create a specific class to interface with the DB. Where the business logic is kept in the Module and note the controllers to keep everything more abstract and easier to maintain. By that I mean controller X should be able to create a new instance of for instance (Application\Model\DBInterface) and use the models functions to do inserts deletes updates joins selects and so forth. The reason I would like to do it this way is so that all modules installed can use the same interface without having to write endless DI statements everywhere. So what I will need is an example of how I can get the configuration for the DB (currently inside module.config.php + local.php(username / pw)) to be passed to the Application\Model\DBInterface dbConfig variable, and perhaps even an instance of the dbAdapter initialized from config if possible.
Alternatively I could potentially grab the configuration from the Application\Model\DBInterface if such a way exists.
If neither of the above is possible then I can always go back to the old way of doing things by reading an ini file for the db details and instantiating my db adapter that way.
Please keep in mind that I won't be injecting anything in the controllers as the controllers just use $db = new \Application\Model\DBInterface() so injecting into the controllers doesn't make much sense at all.
Is there a better way to do this / optimized / am I doing it completely wrong? Anyone able to share some details please. I've spent way too much time on this already and definitely need help.
Okay, so #Ocramius did just let me know what my misconception with the initializers was and helped me out a bit in understanding it. So here is a probably working solution to your Problem. My understanding about your problem is:
"How to get a DbAdapter set for all your Models implementing a DbInterface". This is how you'd do it:
Step 1: Create invokables for all classes implementing the DbInterface. Create a factory for the default Zend\Db\Adapter\Adapter and then create an initializer for your DbInterface
Module.php getServiceConfig()
return array(
'invokables' => array(
'application-model-one' => 'Application\Model\One',
'application-model-two' => 'Application\Model\Two'
),
'factories' => array(
'Zend\Db\Adapter\Adapter' => 'Zend\Db\Adapter\AdapterServiceFactory'
),
'initializers' => array(
'DbInterfaceInitializer' => function($instance, $sm) {
if ($instance instanceof \Application\Model\DBInterface) {
$instance->setDbAdapter($sm->get('Zend\Db\Adapter\Adapter'));
}
},
)
)
The Zend\Db\Adapter\Adapter is using the top-level-configuration-array-key 'db' to automatically inject the dbParams
Step 2: Create your classes implementing your Interface
Application\Model(One|Two|N).php
namespace Application\Model;
class One implements DbInterface, \Zend\Db\Adapter\AdapterAwareInterface
{
/**
* #var \Zend\Db\Adapter\Adapter $dbAdapter
*/
protected $dbAdapter;
public function setDbAdapter(\Zend\Db\Adapter\Adapter $dbAdapter) {
$this->dbAdapter = $dbAdapter;
}
public function getDbAdapter() {
return $this->dbAdapter;
}
// More of your business logic or data here
}
Step 3: Access those classes with the ServiceLocator from your Controllers
SomeController.php someAction()
$dbOne = $this->getServiceLocator()->get('application-model-one');
$dbTwo = $this->getServiceLocator()->get('application-model-two');
// Adapter will automatically be injected
When accessing the invokable from the ServiceManager the initializer will be called. The initializer then will automatically call the Zend\Db\Adapter\Adapter, which in turn get's the parameters from the configuration key 'db'
You may get more information from once the tutorial Application as well as the blog of
samsonasik: ServiceManager Cheat-Sheet
I am doing a Rails 3 app that replaces a paper form for a company. The paper form spans two pages and contains a LOT of fields, checkboxes, drop downs, etc.
I am wondering how to model that in the DB - one approach is to just create a field in the DB for every field on the form (normalized of course). That will make it somewhat difficult to ad or remove fileds since a migration will be needed. An other approach is to do some kind of key/value store (no - MongoDB/CouchDB is not an option - MySQL is required). Doing key/value will be very flexible but will be a pain to query. And it will directly work against ActiveRecord?
Anyone have a great solution for this?
Regards,
Jacob
I would recommend that you model the most common attributes as separate database fields. Once you have setup as many fields as possible then fall back to using a key-value setup for your pseudo-random attributes. I'd recommend a simple approach of storing a Hash through the ActiveRecord method serialize. For example:
class TPS < ActiveRecord::Base
serialize :custom, Hash
end
#tps = TPS.create(:name => "Kevin", :ssn => "123-456-789", :custom => { :abc => 'ABC', :def => )'DEF' })
#tps.name # Kevin
#tps.ssn # 123-456-789
#tps.custom[:abc] # ABC
#tps.custom[:def] # DEF
If your form is fairly static, go ahead and make a model for it, that's a reasonable approach even if it seems rather rudimentary. It's not your fault the form is so complicated, you're just coming up with a solution that takes that into account. Migrations to make adjustments to this are really simple to implement and easy to understand.
Splitting it up into a key/value version would be better but would take a lot more engineering. If you expect that this form will be subject to frequent and radical revisions it may make more sense to build for the future in this regard. You can see an example of the sort of form-builder you might want to construct at something like WuFoo but of course building form builders is not to be taken lightly.
I asked a similar question a while back: Using the Data Mapper Pattern, Should the Entities (Domain Objects) know about the Mapper? However, it was generic and I'm really interested in how to accomplish a few things with Doctrine2 specifically.
Here's a simple example model: Each Thing can have a Vote from a User, a User may cast more than one Vote but only the last Vote counts. Because other data (Msssage, etc) is related to the Vote, when the second Vote is placed the original Vote can't just be updated, it needs to be replaced.
Currently Thing has this function:
public function addVote($vote)
{
$vote->entity = $this;
}
And Vote takes care of setting up the relationship:
public function setThing(Model_Thing $thing)
{
$this->thing = $thing;
$thing->votes[] = $this;
}
It seems to me that ensuring a User only has the last Vote counted is something the Thing should ensure, and not some service layer.
So to keep that in the Model, the new Thing function:
public function addVote($vote)
{
foreach($this->votes as $v){
if($v->user === $vote->user){
//remove vote
}
}
$vote->entity = $this;
}
So how do I remove the Vote from within the Domain Model? Should I relax Vote::setThing() to accept a NULL? Should I involve some kind of service layer that Thing can use to remove the vote? Once the votes start accumulating, that foreach is going to be slow - should a service layer be used to allow Thing to search for a Vote without having to load the entire collection?
I'm definitely leaning toward using a light service layer; however, is there a better way to handle this type of thing with Doctrine2, or am I heading in the right direction?
I vote for the service layer. I've often struggled with trying to add as much logic on the Entity itself, and simply frustrated myself. Without access to the EntityManager, you're simply not able to perform query logic, and you'll find yourself using a lot of O(n) operations or lazy-loading entire relationship sets when you only need a few records (which is super lame when compared to all the advantages DQL offers).
If you need some assistance getting over the idea that the Anemic Domain Model is always an anti-pattern, see this presentation by Matthew Weier O'Phinney or this question.
And while I could be misinterpreting the terminology, I'm not completely convinced that Entities have to be the only objects allowed in your Domain Model. I would easily consider that the sum of Entity objects and their Services constitutes the Model. I think the anti-pattern arises when you end up writing a service layer that pays little to no attention to separation of concerns.
I've often flirted with the idea of having all my entity objects proxy some methods to the service layer:
public function addVote($vote)
{
$this->_service->addVoteToThing($vote, $thing);
}
However, since Doctrine does not have any kind callback event system on object hydration, I haven't found an elegant way to inject the service object.
My advice would be to put all the query logic into an EntityRepository and then make an interface out of it sort of like:
class BlogPostRepository extends EntityRepository implements IBlogPostRepository {}
that way you can use the interface in your unit-tests for the service objects and no dependency on the EntityManager is required.
This really is an architectural question. I feel like I'm going about this the wrong way and wanted some input on best practices.
Let's say I have a Transactions table and a TransactionTypes table. Views will submit the appropriate transaction data which is processed in my controller. The problem is that the logic in the controller may be a bit complex and the TransactionType is not provided by the view inputs, but computed in the controller. (Which may be part of my problem).
For example, let's say that the View submits a ViewModel that would map to a TransactionType of "Withdrawal". However, the controller detects that it needs to change this to an Overdraft" as funds aren't sufficient. What I don't want to do is this:
transaction.TypeId =
DataContext.TransactionTypes.Single(x => x.type == "Overdraft").id;
... as I'll be embedding string literals in my code. Right?
OK, so I could map the values to strong types that would allow me to do this:
class TranTypes
{
public const long Deposit = 1;
public const long Withdrawal = 2;
public const long Overdraft = 3;
}
...
transaction.TypeId =
DataContext.TransactionTypes.Single(x => x.id == TranTypes.Overdraft);
Now, if my lookups change in the DB, I have one place that I can update the mappings and my controllers still have insight into the model.
But this feels awkward too.
I feel like what I really want is for the Linq To SQL auto-code generation to be able to generate the association so I can just refer to strongly-typed names (Deposit, Withdrawal, and Draft) and be assured that it will always return the current values for these in the database. Changes made to the lookup table during runtime would result in problems, but it still seems so much cleaner.
What should I be digesting to understand how best to structure this?
Thanks in advance for enlarging my brain. :-)
Dont worry about whether you have an embedded string or a strongy typed value - either is perfectly acceptable - which ever makes sense fror your database design.
What you should do, however, is write a single routine in a repository or helper class that you can then call from whatever controller or action requires it - if anything changes there is only one place to make the change.
One simple approach I've always liked is the Enum approach.
public enum TransactionType {
Overdraft
}
transaction.TypeId =
DataContext.TransactionTypes.Single(x => x.type == TransactionType.Overdraft.ToString()).id;
It's pretty simple, but I like it.
A more sophisticated approach (not sure if this works with Linq to SQL, but more sophisticated ORMs support it (like EF, DO .NET, LLBLGen, etc.) is to use inheritance in your data model, with discriminators.
That is, have a subclass of TransactionType called OverdraftTransactionType with a discriminator (the key) that identifies different types of TransactionTypes from each other.
Random link:
http://weblogs.asp.net/zeeshanhirani/archive/2008/08/16/single-table-inheritance-in-entity-framework.aspx