I have a struts 1 application.
A user without an account comes in and I redirect them to a page to set up their account.
The user came in with some parameters on the URL.
enter.do?sys=foo
The account creation page has a form with an iteration in it that uses the name sys.
public void setSys(Domain d, int i)
{...}
the result is is a stack dump as the framework tries to set foo into the bean.
java.lang.reflect.InvocationTargetException: Cannot set sys
at org.apache.commons.beanutils.BeanUtilsBean.setProperty(BeanUtilsBean.java:1025)
at org.apache.commons.beanutils.BeanUtilsBean.populate(BeanUtilsBean.java:811)
at org.apache.commons.beanutils.BeanUtils.populate(BeanUtils.java:298)
at org.apache.struts.util.RequestUtils.populate(RequestUtils.java:1252)
at org.apache.struts.action.RequestProcessor.processPopulate(RequestProcessor.java:821)
at org.apache.struts.action.RequestProcessor.process(RequestProcessor.java:254)
at org.apache.struts.action.ActionServlet.process(ActionServlet.java:1482)
at org.apache.struts.action.ActionServlet.doGet(ActionServlet.java:507)
at javax.servlet.http.HttpServlet.service(HttpServlet.java:731)
at javax.servlet.http.HttpServlet.service(HttpServlet.java:844)
at weblogic.servlet.internal.StubSecurityHelper$ServletServiceAction.run(StubSecurityHelper.java:242)
at weblogic.servlet.internal.StubSecurityHelper$ServletServiceAction.run(StubSecurityHelper.java:216)
at weblogic.servlet.internal.StubSecurityHelper.invokeServlet(StubSecurityHelper.java:132)
at weblogic.servlet.internal.ServletStubImpl.execute(ServletStubImpl.java:352)
at weblogic.servlet.internal.TailFilter.doFilter(TailFilter.java:25)
at weblogic.servlet.internal.FilterChainImpl.doFilter(FilterChainImpl.java:74)
at com.aaron.filter.AuthorizationFilter.doFilter(AuthorizationFilter.java:179)
How can I work around this name conflict? I could rename one of the fields, but I have still leave myself open for other name conflicts.
I ended up renaming the field to avoid the name conflict. If someone comes up with a different way please go ahead and post it.
Related
We call the microsoft exchange to set the extended property which in our case is an unique guid
microsoft.exchange.webservices.data.core.exception.service.remote.ServiceResponseException: An internal server error occurred. The operation failed., Invalid named property
Its been working great until now when some of our users are facing the above issue ....
val uId = getUniqueId();
val emailExtendedPropDef = new ExtendedPropertyDefinition(uId,"uniqueId", MapiPropertyType.String)
try {
email.setExtendedProperty(emailExtendedPropDef, uId.toString)
email.sendAndSaveCopy()
} catch {
case e: Exception =>
error(s"Exception in setting extended property for user $from", e)
throw e
}
trying to find the root cause of the issue, we are also thinking it might be related to throttling on Microsoft exchange for extended properties (Not sure how to prove if it's indeed throttling) any help to point us in the right direction will be of great help
Our use case is to able to retrieve the email when customer want's to reply back we want to retrieve that particular email to be included in users reply....currently we are using the uid to achieve that ....
we have been using the code as per the documentation here
https://learn.microsoft.com/en-us/previous-versions/office/developer/exchange-server-2010/dd633654(v%3Dexchg.80)
and also the documentation here
https://github.com/OfficeDev/ews-java-api/wiki/Getting-Started-Guide#extended-properties
Update : As per the comments we do understand that we have to treat extendedProperty as a column definition and update the same column ...but we couldn't figure out how to achieve this as...Any code samples to point us in the right direction will be of great help
Latest Update : We have deleted some of the extendedPropertyDefinition's but still facing the same invalid property could some one please point us in the right direction
Is it safe to say that getUniqueId returns a different guid on each call? If so, then that is the problem. Think of the Guid for an extended prop as a namespace. The exchange store limits the number of custom extended props to something like 32k per mailbox. So you are likely hitting that limit. But aside from that, the main reason for creating an extended property is so that you can refer to it later. But if you are basically discarding the namespace each time, you are leaving orphaned props on items. Without understanding your particular scenario, I can only say that the Guid should be thought of truly as a namespace. Choose one for your app/company/scenario and hard code it. They create all the named props you want within that namespace. For instance, "MyProp/String" in Guid namespace 1 is a different property than "MyProp/String" in Guid namespace 2.
I have a User model, for login and registration, its email field is used (everything vanilla from the devise gem).
I want (other) users to be able to e.g. add Users to a team, with the email-address as the identifier.
That is fine when the User is already existing (pseudo #team.users.add(User.find_by(email: other_users_email))) but I am unsure how to handle situations where the user does not yet exist (did not [yet] register).
When a (new) User sets up a new account, for the example above after successfull registration current_user.teams should show up correctly.
I do not want to force these potentially new users to use the system (e.g. using devise_invitable) and bother them with an email.
I followed the path of creating the User when a user with the given email does not yet exist, but then when the user actually tries to setup an account, it fails (email not unique).
Alternatively, I could remodel the TeamMember-part and let it optionally either store an email-adress or the reference to an existing User. Then what I would need is to check for "open" TeamMembers directly after User-Account-creation (so, TeamMembers with the given email). I could also do this on each requst, but that looks too expensive to me. There might be race conditions, but I could live with that (and check for the every-now-in-a-millenia-gap with a cron-job).
Any pointers? I am sure this is not that unusual.
I'd do this:
When a user A adds user B to a team by email, create the object for that user B, but set a flag, something like auto_created_and_inactive: true
When user B signs up on the site, you just have to handle this in your users#create: first, try to find an auto-created record and update it (set a password or whatever; also reset the flag). Or otherwise proceed with the usual route of creating a new record.
I have to admit that I did not yet tried #sergio-tulentsevs approach (implement RegistrationController#create). But to complete what I sketched in my question:
User model can define an after_confirmation method, which is called after ... confirmation! So, if I store every information about a potential user with a reference to his/her email-adress, once he/she registered I can query this information and e.g. complete Team-Memberships.
# app/models/user.rb
def after_confirmation
# (pseudo-code, did not try)
self.teams < TeamMembership.open.where(email: self.email)
end
I'm working with the Extension Library Managed Bean "UserBean". I'm trying a simple proof to test the ACL property of a panel with the userBean. Then I have three panels on the page with text in them. Here is one of the panels.
to display my current roles.
<xp:text escape="true" id="computedField3" value="#{javascript:return userBean.accessRoles.toString();}"></xp:text>
<xp:panel id="panelMarketing" style="border:1px solid red;">
<xp:this.acl>
<xp:acl>
<xp:this.loaded><![CDATA[${javascript:var roles = userBean.accessRoles.toString();
return roles.contains("[MARKETING]");}]]></xp:this.loaded>
</xp:acl>
</xp:this.acl>
<xp:label value="I'm in Marketing" id="label2"></xp:label>
</xp:panel>
I load the page while I have the 3 Roles assigned to me. Panel displays. Remove the role, all panels still display and the text field that returns the user roles still shows the old role that was assigned to me in the ACL. Seams like UserBean is not reading the ACL again.
Then I try CTRL F5, still nothing. I then open another browser, log into the app and still the same roles. Close the browsers, log back in and no change.
Is there a flush time or a way to force the userBean to update the users ACL roles?
The next day I log back in and the UserBean reflects my changes. I make another change and the UserBean does not reflect the change.
For the acl add an aclEntry and do not code the loaded property, like this ...
<xp:panel id="panelMarketing" style="border:1px solid red;">
<xp:this.acl>
<xp:acl>
<xp:this.entries>
<xp:aclEntry right="READER" type="ROLE">
<xp:this.name><![CDATA[[MARKETING]]]></xp:this.name>
</xp:aclEntry>
</xp:this.entries>
</xp:acl>
</xp:this.acl>
<xp:label value="I'm in Marketing" id="label8"></xp:label>
</xp:panel>
From reading the Extension Library book (see page 464) and the source code for the userBean implementation it appears that userBeans are cached in application scope with a default cache size set to 100 (because you can use userBean to access details about other users than just the current user).
The userBean works with the user being authenticated in the session. So the userBean is basically a shortcut for peopleBean.getPerson() in Java.
The peopleBean class has a refreshPerson(id) method and a clearcache(id) method. Try using one of these to force the peopleBean/userBean to refresh by doing either peopleBean.refreshPerson(#UserName()) or peopleBean.clearcache(#UserName()).
I know this is 2 years ago but I had the same problem and decided to investigate further.
It turns out there are several bugs in the underlying clearCache(String id) method which is supposed to clear a single cache entry (e.g. a person). Consequently had no effect when used on a application, session or request scope cache. (it did work for a global cache though so I suspect that was the only one they tested).
I have submitted a pull request to the Extension Library to fix the clearCache(string id) method so hopefully it gets accepted for the next release!
The clearCache() method which clears all entries does work so you could always try
peopleBean.clearCache();
and that should work (it does for me) however it clears every persons entry which shouldn't be a big problem, they will just load again I guess.
I have the following in my application:
Action Orm entity (From telerik open access)
Repository(Of Action)
AppService(Holds an instance of the repository)
when I need to save an instance, I send the instance to the AppService. the AppService then calls a validator to validate the instance to save. the validator is based on http://codeinsanity.com/archive/2008/12/02/a-framework-for-validation-and-business-rules.aspx
(full code on https://github.com/riteshrao/ncommon)
so basically my save function in the AppService looks like this
Public Sub AddAction(ByVal Item As Data.Model.Action)
Contract.Requires(Of ArgumentNullException)(Item IsNot Nothing, "Item is nothing.")
Dim validateResult As Rules.ValidationResult = _ActionValidator.Validate(Item)
If Not validateResult.IsValid Then
Throw New Validation.ValidationException(validateResult)
End If
Try
_ActionRepository.Add(Item)
_unitOfWork.SaveChanges()
Catch ex As Exception
_unitOfWork.ClearChanges()
Throw New DataServiceException(ex.Message, ex)
End Try
End Sub
for checking properties of the Action item the sample code works great. my question begins when i need to make sure that the same action is not added twice to the DB for the same customer (ie. id is difference, name is the same and customer is the same)
as I see it I have a few options:
option 1: check for a duplicate action using something like
function(validatedItem) item.Customer.Actions.Any(function(item) item.id<>validatedItem.id andalso item.name=validatedItem.name))
basically I go from the action being saved back to the customer and then back to all his actions and check if any action exists with a different id and same name
the down sides are:
a. for this to work, when accessing the customer property of the item, the entire customer object is read from DB which is redundant in this case
b. the Any function is being evaluated on the client as item.Customer.Actions returns IList(Of Action)
Option 2: let the validation class have access to the action repository. then i could simply do something like
'assume I already have validatedItem
repository.Any(function(item) item.id<>validatedItem.id and item.customerid=validatedItem.customerid and item.name=validatedItem.name)
this will result in an Exists query being sent to the DB but the downside(?) is that the validation framework should not access the repository directly (as far as I have seen in the very few examples i could find that do use validation and ORM)
Option 3: let the validation class have access to the AppService and use the AppService to check for existence of a duplicate.
problems:
a. I create a circular reference (AppService->Validation Class->AppService)
b. I need to create a lot of useless functions in the AppService for loading items based on criteria that is only relevant for the validation
Any ideas what is the best course here?
The simplest is not to check duplicates in the database from your domain.
When a collection of entities is part of you aggregate then it is a non-issue since you would not permit the duplicate to be added to the collection. Since the aggregate is stored as a whole you would never run into the problem.
For scenarios where you do not want a duplicate, say, e-mail address and no collection of the entities is represented by an aggregate (such as the Users in a system) you can just let the database enforce the uniqueness. Simply pick up the exception and report back. In many instances your validation would not be able to enforce the uniqueness simply because it doesn't have/implement the required locks that a database system would have.
So I'd simply leave that up to the database.
I am trying to implement the beforUpdate event in grails domain classes I need to audit log both the old and new values of the Domains attributes. I see that we can use the isDirty check or use Domain.dirtyPropertyNames which return list of properties which are dirty in the domain. and getPersistentValue gets the old value in the table so I can have both values..
For implementing this I will be using the beforUpdate event in the domain class and call a logging service from there, passing it the id of the User domain. now using this ID I can get user Instance in Service and then check if any fields are dirty using the above specified method? or do I need to log the Audit when I am actually doing the Update in the UserController's update def?
Which is the better approach?
I want to confirm if this the right approach..
Also what other things I need to take care for, Like:
1) if the attributes are domain object references and not simple types.
2) any other things I need to take care like not flushing out the hibernate session, thinking of implementing this in call to service from the domain class.
Regards,
Priyank
Edit: I tried this in the beforeUpdate event in User domain that I want to Audit log for Update activity..
def beforeUpdate = {
GraauditService service = AH.getApplication().getMainContext().getBean(''graauditService)
service.saveUserUpdateEntry(this.id); // id property of User domain...
}
In the method in Service I do:
def saveUserUpdateEntry(Long id){
User grauser = User.get(id);
println ("user="+ grauser)
println "Dirty Properties -: ${grauser.dirtyPropertyNames}"
println "Changed value for firstName = -: ${ grauser.firstName}"
println "Database value for firstName = -: ${ grauser.getPersistentValue('firstName')}"
}
I try to do the update from the UI for email, firstname, lastname and get the following on console:
user=com.gra.register.User : 1
Dirty Properties -: [eMail, firstName, lastName]
Changed value for firstName = -: sefser
Database value for firstName = -: administer
user=com.gra.register.User : 1
Dirty Properties -: []
Changed value for firstName = -: sefser
Database value for firstName = -: sefser
possible nonthreadsafe access to session
I am not able to know:
1) Why am I getting 2 sets ... is the event called twice once before commit and once after commit...??
2) how to remove or handle the Hibernate exception (tried to use withNew session in the function but no difference
Thanks in Advance..
Rather than using GORM event handlers for audit logging, use audit logging plugin. This will take away a lot of pain of yours.
Hope this helps.
or
If You want much finer control over what you are doing you should consider using a subclass of Hibernate's EmptyInterceptor. This will serve tow purposes for you
Will give you much finer control over what and how you are doing the audit logging
Will place all your logic for audit logging at one place, which will help you maintaining your code.
Click here to see the API for EmptyInterceptor.
Note: Hibernate does not ship any implementation in this class and also don't provide any subclass of this which might provide you the default behavior. So you will have to write a custom implementation.