This must be really basic stuff since I did not find any discussion about it. However I have struggled with this for a while.
I have pretty basic many-to-many relation with extra field implemented like in this example (double one-to-many relation). This works nicely when creating new entities and saving them to database. I am now trying to create editing feature and came across some problems.
Lets say my main entity is called Recipe, which has many-to-many relation with Ingredient entity. Extra field(s) like "amount" are in RecipeIngredient entity. Recipe class has setRecipeIngredient method, which adds RecipeIngredient object to ingredients array.
Should I create some "clearRecipeIngredients" method to Recipe class, which would remove all RecipeIngredient objects? I would call this when editing Recipe, then creating new RecipeIngredient entities from my data and populating ingredients array like when creating new entity? I admit that my cascade settings are probably not set correctly this to work, but I try fixing it next.
Any related examples would be great.
Strictly speaking, as you mentioned, there is no many-to-many relationship here, but a one-to-many followed by a many-to-one.
About your question, I wouldn't perform a bulk "clear" each time I want to edit the recipe. Instead, I'd provide a fluent interface to mimic the steps that would be taken if you wanted to edit a paper-based recipe.
I provided a sample implementation below:
class Recipe
{
/**
* #OneToMany(targetEntity="RecipeIngredient", mappedBy="recipe")
*/
protected $recipeIngredients;
public function addIngredient(Ingredient $ingredient, $quantity)
{
// check if the ingredient already exists
// if it does, we'll just update the quantity
$recipeIngredient = $this->findRecipeIngredient($ingredient);
if ($recipeIngredient) {
$quantity += $recipeIngredient->getQuantity();
$recipeIngredient->updateQuantity($quantity);
}
else {
$recipeIngredient = new RecipeIngredient($this, $ingredient, $quantity);
$this->recipeIngredients[] = $recipeIngredient;
}
}
public function removeIngredient(Ingredient $ingredient)
{
$recipeIngredient = $this->findRecipeIngredient($ingredient);
if ($recipeIngredient) {
$this->recipeIngredients->removeElement($recipeIngredient);
}
}
public function updateIngredientQuantity(Ingredient $ingredient, $quantity)
{
$recipeIngredient = $this->findRecipeIngredient($ingredient);
if ($recipeIngredient) {
$recipeIngredient->updateQuantity($quantity);
}
}
protected function findRecipeIngredient(Ingredient $ingredient)
{
foreach ($this->recipeIngredients as $recipeIngredient) {
if ($recipeIngredient->getIngredient() === $ingredient) {
return $recipeIngredient;
}
}
return null;
}
}
Note: you'll need to setup cascade persist and orphan removal for this code to work properly.
Of course, if you take this approach, your UI shouldn't display a full form with all ingredients and quantities editable at once. Instead, all the ingredients should be listed, with a "delete" button on each line, as well as a "change quantity" button which would pop-up a (single-field) form to update the quantity, for example.
Related
My DAC and tables are defined as follows
Data structure
ParentTableDAC (FormView)
ChildDac1 (TAb1/Grid)
ChildDac2 (Tab2/Grid)
in ChildDAc2, ChildDAc1ID need to be shown as Selector, How this can be done?
We are facing issue if Data is not saved for Parent/ChildDAc1 then it is not available to ChildDAc2 Lookup
Update -
Business Scenario -
A work item has Multiple Task and multiple Steps to perform the work item.
now WORKITEMDAC is a Parent
TASKDAC and STEPDAC are the Child of the Parent WORKITEMDAC.
OK till yet everything is ok and Usual..
Now each step is suppose to be linked with the Task of the Parent WORKITEM.
SO in STEPS Grid a Selector is required to select the TASK.
Here I have proper Parent child relationship there is no issue with that,I have only one issue that is, I can select only the TASK which was already saved with the Parent WORKITEM, unsaved Tasks are not displayed in the selector.
SO My Question was, do we have any way to get the Task in the PXSelector query which is not yet saved?
Following Selector is used on STEPDAC on TaskID Column -
[PXSelector(typeof(Search<TASKDAC.TaskID, Where<TASKDAC.taskID, Equal<Current>>>), typeof(TASKDAC.taskCD), SubstituteKey = typeof(TASKDAC.taskCD))]
Note - TaskID column in TASKDAC is identity column and this DAC has WorkItemID defined as PArent.
Update
Thanks for updating with more detail. Leaving the original response below for those that may need it regarding parent/child.
I've never seen the selector defined with reference to Current<> and no Current WHAT. To be honest, I'm surprised it compiled. "Where Field = Current View Field" directs the selector to limit results to only values associated to the Current value of whatever field you specified, and it is retrieved from the "ViewName.Current" record in the graph. Your where clause does not tell it what "current" to match. Since you direct the selector to Search, I suspect that this is causing the selector to be unable to find a matching record and short-circuiting the save of the result for that field.
It sounds like TASKDAC is a "master data" type record, defining all possible tasks to be performed. Therefore, TaskID is not a key field in STEPDAC but rather a key field in TASKDAC. If you want to select "any" task, you don't need the where clause at all.
[PXSelector(
typeof(TASKDAC.TaskID),
typeof(TASKDAC.taskCD),
SubstituteKey = typeof(TASKDAC.taskCD)
)]
Let's assume for a moment that you designate a WORKITEMDAC.WorkOrderType that is used to filter acceptable tasks by the same field. (Only repair tasks can be assigned to a repair WorkOrderType, assembly to an assembly WorkOrderType, etc.) In this case, you would use the syntax you stated, but you would designate Where<TASKDAC.WorkOrderType, Equal<Current<WORKITEMDAC.WorkOrderType>>. If the task is limited to something in the STEPDAC instead, just swap out WORKITEMDAC with the STEPDAC field in the example below.
[PXSelector(
typeof(Search<TASKDAC.TaskID,
Where<TASKDAC.WorkOrderType, Equal<Current<WORKITEMDAC.WorkOrderType>>>>),
typeof(TASKDAC.taskCD),
SubstituteKey = typeof(TASKDAC.taskCD)
)]
Please update your PXSelector and advise if that fixes the issue or what error you get next.
Original Response
With so little to your question, the nature of your question is very unclear. It sounds like you are attempting to do something that should perhaps take a different approach. The definition of a selector should not impact the saving of a field, but the structure of a parent child relationship is critical to define properly. As such, it seems very odd that you would have a Child ID that differs from the parent and that it would be used in a selector.
Typically, the child DAC shares an ID with the parent DAC. We often define a LineNbr field in the child which is given a value by PXLineNbrAttribute, and the Key fields would be ID + LineNbr to identify a specific child record. We also set SyncPosition = true in the ASPX document on the grid so that the user interface stays completely sync'd with whichever record the user has selected in the grid. Parent Child relationships can be daisy chained, such as with SOOrder -> SOLine -> SOLineSplit, but ultimately, the ID field of a child is not a selector... the ID of the parent is.
Think of a selector as a means of looking up master data. Master data may be related to other data, but it would not have a parent in most cases. There are exceptions, but let's keep it simple. A purchase order, for instance, would not have parent, and the key field would be a common selector field. However, since a child must always relate to its parent and cannot change its parent, the ID of the child field would not be a selector. Other tables in Acumatica would refer back to some field or combination of fields to relate to the data of a child DAC.
From the T210 training guides, I stripped away a lot to show just this point.
RSSVRepairPrice - Parent
using System;
using PX.Data;
namespace PhoneRepairShop
{
[PXCacheName("Repair Price")]
public class RSSVRepairPrice : IBqlTable
{
#region PriceID
[PXDBIdentity]
public virtual int? PriceID { get; set; }
public abstract class priceID : PX.Data.BQL.BqlInt.Field<priceID> { }
#endregion
#region ServiceID
[PXDBInt(IsKey = true)]
[PXDefault]
[PXUIField(DisplayName = "Service", Required = true)]
[PXSelector(
typeof(Search<RSSVRepairService.serviceID>),
typeof(RSSVRepairService.serviceCD),
typeof(RSSVRepairService.description),
DescriptionField = typeof(RSSVRepairService.description),
SelectorMode = PXSelectorMode.DisplayModeText)]
public virtual int? ServiceID { get; set; }
public abstract class serviceID : PX.Data.BQL.BqlInt.Field<serviceID> { }
#endregion
#region RepairItemLineCntr
[PXDBInt()]
[PXDefault(0)]
public virtual Int32? RepairItemLineCntr { get; set; }
public abstract class repairItemLineCntr : PX.Data.BQL.BqlInt.Field<repairItemLineCntr> { }
#endregion
}
}
RSSVRepairItem - Child
using System;
using PX.Data;
using PX.Objects.IN;
using PX.Data.BQL.Fluent;
namespace PhoneRepairShop
{
[PXCacheName("Repair Item")]
public class RSSVRepairItem : IBqlTable
{
#region ServiceID
[PXDBInt(IsKey = true)]
[PXDBDefault(typeof(RSSVRepairPrice.serviceID))]
[PXParent(
typeof(SelectFrom<RSSVRepairPrice>.
Where<RSSVRepairPrice.serviceID.IsEqual<
RSSVRepairItem.serviceID.FromCurrent>>
))]
public virtual int? ServiceID { get; set; }
public abstract class serviceID : PX.Data.BQL.BqlInt.Field<serviceID> { }
#endregion
#region LineNbr
[PXDBInt(IsKey = true)]
[PXLineNbr(typeof(RSSVRepairPrice.repairItemLineCntr))]
[PXUIField(DisplayName = "Line Nbr.", Visible = false)]
public virtual int? LineNbr { get; set; }
public abstract class lineNbr : PX.Data.BQL.BqlInt.Field<lineNbr> { }
#endregion
}
}
In the parent, notice the use of PXDBIdentityAttribute which specifies that the record identity is set by the database in PriceID. The "ID" field used by the child is taken from the ServiceID field which is defined as a selector for another master table. The use of IsKey = true on PXDBIntAttribute tells Acumatica that it can find the unique record by the combination of all the fields marked IsKey = true. (Note that the actual training guide includes another field as part of the key, but I simplified the example here.) Also notice the use of a PDXBInt field that has a default of 0 to be used as a Line Counter. (RepairItemLineCtr)
Now, in the child, notice that the key field of the parent is defined again in the DAC, but without the selector. The PXParentAttribute defines the relationship back to the parent DAC, and the field value is defaulted in from that parent. This is what causes the value to be saved in the record automatically. It also associates the relationship so that deleting the parent will delete the child as well. Notice the use of PXLineNbrAttribute to autoincrement the field of the parent DAC and set the value in the LineNbr field. It is the combination of these 2 fields that points you to the exact record.
There are cases where Acumatica uses the NoteID field (a unique GUID value) to identify a record of a child DAC, but this is a bit more complex. You can see it in the relationship of INItemPlan (Demand) to Purchase Orders.
If you are having trouble with Parent-Child (Master-Detail) relationships, I'd highly recommend reviewing the T210 training course.
To refer the unsaved data use the ISDirty =true on PXSelector.
also to save the Identity key valueof the unsaved data at referenced place, use [PXDBDefault()]
I'm using existing Task module under the Project module. I would like to
assign particular task to multiple workers. Meaning that group of people
will together complete the task.
I already have workers as users in my vtigercrm. So if i make a user selection
multiple for assigning single task it could be easier to handle this
use-case.
This use-case already have in Calendar module but i'm unable to understand how they implement.
I have a thorough knowledge on how to create custom user list drop-down in any module and i created lot more for my custom modules. Now stuck with the same scenario having multiple at a time.
I tried like this in module.php, but it is not working.
$users = Vtiger_Module::getInstance('Users');
$module = Vtiger_Module::getInstance('MyModule');
$module->unsetRelatedList($users, 'Users', 'get_related_list');
$module->setRelatedList($users, 'Users', array('SELECT'), 'get_related_list');
Can anyone help me to get it done?
It is possible when you understood the Invite users field in Calender envent creation. Here the field must be multi-selection.
You need to deal with vtiger_salesmanactivityrel table for inserting the multiple users and for showing in detail view.
Here is the code to handle inviteusers.
//Handling for invitees
$selected_users_string = $_REQUEST['inviteesid'];
$invitees_array = explode(';',$selected_users_string);
$this->insertIntoInviteeTable($module,$invitees_array);
function insertIntoInviteeTable($module,$invitees_array)
{
global $log,$adb;
$log->debug("Entering insertIntoInviteeTable(".$module.",".$invitees_array.") method ...");
if($this->mode == 'edit'){
$sql = "delete from vtiger_invitees where activityid=?";
$adb->pquery($sql, array($this->id));
}
foreach($invitees_array as $inviteeid)
{
if($inviteeid != '')
{
$query="insert into vtiger_invitees values(?,?)";
$adb->pquery($query, array($this->id, $inviteeid));
}
}
$log->debug("Exiting insertIntoInviteeTable method ...");
}
Hope it would give you a better idea.
When you try to close Swing application and you change same value in field, application ask you if you want to save changes.
My first question is why RAP application doesn't ask same question ?
Second and more important question is how to force validation of fields change and how to manipulate this.
for example :
I have table with rows and some fields below table. If I click on row some value is entered in fields. I can change this value and click button save.
I would like to force change validation on row click. So if changes ware applied and if I click on row, application should warn me that some changes are not saved.
But I should be able to manipulate what is change and what is not. For example if table on row click fill some data if fields this is not change, but if I entered value is same fields this is change.
I discovered method
checkSaveNeeded();
But it does nothing. (if I change values or not)
I see that every field has mathod
#Override
public final void checkSaveNeeded() {
if (isInitialized()) {
try {
propertySupport.setPropertyBool(PROP_SAVE_NEEDED, m_touched || execIsSaveNeeded());
}
catch (ProcessingException e) {
SERVICES.getService(IExceptionHandlerService.class).handleException(e);
}
}
}
so I should manipulate changes throw m_touched ?
How is this handled in Scout ?
ADD
I am looking for function that check for dirty fields and pops up message dialog, same as when closing form, and way to set fields dirty or not.
I look here and here, but all it describes is how values is stored for popup messages and dot how to fire this messages (validation).
My first question is why RAP application doesn't ask same question ?
I am not sure to know which message box you mean but it should be the case.
There are several questions about unsaved changes and form lifecycle in the Eclipse Scout Forum. I think you can find them with Google.
I have also taken the time to start to document it in the Eclipse Wiki:
Scout Field > Contribution to unsaved changes
Form lifecycle
I think you should implement execIsSaveNeeded() in the corresponding field. The default implementation in AbstractTableField uses the state of the rows, but you can imagine the logic you want.
#Order(10.0)
public class MyTableField extends AbstractTableField<MyTableField.Table> {
#Override
protected boolean execIsSaveNeeded() throws ProcessingException {
boolean result;
//some logic that computes if the table field contains modification (result = true) or not (result = false)
return result;
}
//...
I hope this helps.
I am looking for function that check for dirty fields and pops up message dialog, same as when closing form.
Are you speaking from this message box that appears when the user clicks on Cancel in the form?
There is no specific function that you can call for that. But you can check the beginning of the AbstractForm.doCancel() function. It is exactly what you are looking for.
I have rewritten it like this:
// ensure all fields have the right save-needed-state
checkSaveNeeded();
// find any fields that needs save
AbstractCollectingFieldVisitor<IFormField> collector = new AbstractCollectingFieldVisitor<IFormField>() {
#Override
public boolean visitField(IFormField field, int level, int fieldIndex) {
if (field instanceof IValueField && field.isSaveNeeded()) {
collect(field);
}
return true;
}
};
SomeTestForm.this.visitFields(collector);
MessageBox.showOkMessage("DEBUG", "You have " + collector.getCollectionCount() + " fields containing a change in your form", null);
I have changed the Visitor to collect all value fields with unchanged changes. But you can stick to the original visitField(..) implementation. You cannot use P_AbstractCollectingFieldVisitor because it is private, but you can have a similar field visitor somewhere else.
I am looking for a way to set fields dirty or not.
As I told you: execIsSaveNeeded() in each field for some custom logic. You can also call touch() / markSaved() on a field to indicate that it contains modifications or not. But unless you cannot do otherwise, I do not think that this is the correct approach for you.
I've got an issue with a nested property that uses a composite key.
When I'm editing a model that has multiple instances of nested properties (with a composite key) and want to update it to have fewer by leaving them blank, cfWheels does not remove the ones that are not used anymore, and maintains the old value. Is there a way of forcing the deletion of these without calling delete on the nested model?
I've been doing by deleting all nested properties, and then update() creates the records that are needed, but the big issue with that is that when I have code in between that fails, It just deletes the items, which as you know can be very very bad.
In your init call to nestedProperties(), try adding the allowDelete option:
nestedProperties(association="comments", allowDelete=true);
Then if a model within that collection has a property called _delete that is set to true, CFWheels will delete that record.
I'm not sure of your model because you don't include any details in your question, but you could probably run a beforeValidationOnUpdate callback that checks criteria on the nested models and sets _delete = true when the record needs to be deleted.
For example:
// Post.cfc
component extends="Model" {
function init() {
hasMany("comments");
nestedProperties(association="comments", allowDelete=true);
beforeValidationOnUpdate("removeBlankComments");
}
private function removeBlankComments() {
if (StructKeyExists(this, "comments") && IsArray(this.comments)) {
for (local.i = 1; local.i < ArrayLen(this.comments); local.i++) {
if (!Len(this.comments[local.i].message)) {
this.comments[local.i]._delete = true;
}
}
}
}
}
Not sure if this will give you any problems with the nested composite key. Sometimes nested properties are a little kludgy with "special" cases.
i think you forgot to mention allowDelete attribute in nestedProperties by defalut allowDelete is set as false in wheels and does not delete the composite key form table. you have to set it true.
for example in model you have to do some thing like this.
<cfset hasMany(name="campaignlanguages",shortcut="languages", dependent="deleteAll") />
<cfsetnestedProperties(associations="campaignlanguages",allowDelete="true")/>
you can find more details here
I have been wondering why my mappings and the controller actions are not working. For this I need to refer to my previous post, where I have described my entities and database schema, which can be found here. I need to start a new post since there were no further updates and I thought this the only way to get attention of the Doctrine + Zend Pros.
As described in my previous post, I have a Zend Form, the user can enter teamId and teamName, further he has the choice to select multiple players from the drop down list on the form and can allocate players to the team. So basically that is my goal to achieve with Doctrine and Zend. For that I wrote my entities described in the previous post and right now I want to add the code from my controller to persist the entities.
Controller:
public function addAction()
{
$form = new TeamForm($this->getEntityManager());
$form->get('submit')->setAttribute('value', 'Add');
$request = $this->getRequest();
if ($request->isPost())
{
$team = new Team();
$player = new Player();
$teamPlayers = new TeamPlayer();
$form->setInputFilter($typeset->getInputFilter());
$form->setData($request->getPost());
if ($form->isValid())
{
$team->populate($form->getData());
$teamPlayers->setPlayer($player);
$teamPlayers->setTeam($team);
$this->getEntityManager()->persist($teamPlayers);
$this->getEntityManager()->flush();
//Reroute to the index page once the data is successfully added
}
}
//return form array
return array(
'form' => $form
);
}
So that is basically what Im doing in my controller to save the entities into two tables (team Table and teamPlayer Table), as already the player Table is populated with data. So I want to add players to the team and assign that values in these two tables.
Right now I can see my form and when I enter the data and press submit nothing is happening, i can see the form with no action. When the data is successfully saved into the database then I reroute it to the index page which is not happening.
Any help would be appreciated, to point out the error Im making either in mapping section or in the controller side.
The official documentation especially from Doctrine 2 is too global and is particulary not that clear for my requirement.
let's try to solve it by updating this answer by steps :
step 1 :
your words implied to me that , you might have some issues in from validation , so lets check if you are passing this $form->isValid()
if ($form->isValid())
{
$team->populate($form->getData());
$teamPlayers->setPlayer($player);
$teamPlayers->setTeam($team);
$this->getEntityManager()->persist($teamPlayers);
$this->getEntityManager()->flush();
//Reroute to the index page once the data is successfully added
}else{
var_dump($form->getMessages());
}
I can also suggest you to use doctrine commandline : doctrine orm:validate-schema , this command will help you to check if your entity mapping is ok plus your database mapping is also okay , i see it as handy to to debug your doctrine2 entity
ps : I havn't read your entities in depth yet