WF 4 Rehosted Designer - get foreach InArgument Value - workflow-foundation-4

After reading this article:
http://blogs.msdn.com/b/tilovell/archive/2009/12/29/the-trouble-with-system-activities-foreach-and-parallelforeach.aspx
I have defined the ForEachFactory as follows:
public class ForEachFactory<T> : IActivityTemplateFactory
{
public Activity Create(DependencyObject target)
{
return new ForEach<T>
{
DisplayName = "ForEachFromFactory",
Body = new ActivityAction<T>
{
Argument = new DelegateInArgument<T>("item")
}
};
}
}
All works well but is it possible to check how that DelegeateInArgument in my case named "item" changes its value ?
So if i have defined an array in the variables section and initialized with
{1, 2, 3} i need a way to check how the "item" takes value 1, 2 and then 3.
To be more accurate, i've added this pic, with a breakpoint on the WriteLine activity inside the foreach. When the execution will stop there, is there a way to find out what the value of item is ?
EDIT 1:
Possible solution in my case:
After struggling a bit more i found one interesting thing:
Adding one of my custom activities in the Body of the ForEach, i am able to get the value of the item like this :
So, my activity derives from : CodeActivity
Inside the protected override String[] Execute(CodeActivityContext context) i am doing this job.To be honest, this solves the thing somehow, but it is doable only in my custom activities. If i would put a WriteLine there for example, i would not be able to retrieve that value.

you can access the DelegeateInArgument of a ForEach activity by inspecting the ModelItem trees parent and checking for DelegeateInArgument's. If you need a specific code example to achieve this I may need a some time to code the example. As it has been a long time since I did this, see my question i asked over on msdn
So basically where your break point is, you can access the variable values as these are defined with n the scope of your activity as 'variables'. However the 'item' variable is actually only accessible from the parent loop activity. So you have to get the model item of the current executing activity and then traverse up the tree to find the parent containing the desired DelegateInArgument.
Can you flesh out exactly what you want to achieve? Is it that when your debugging the workflow in the re-hosted designer you want to display the variable values to the user as they change in the UI?
Edit - added tracking example
So as your wanting to display the variable values during execution of the workflow we need to use tracking to achieve this. In the example your using the author has already implemented some basic tracking. So to achieve the extended variable tracking you want you will need to alter the tracking profile.
Firstly amend the WorkflowDesignerHost.xaml.cs file alter the RunWorkflow method to define the SimulatorTrackingParticipant as below.
SimulatorTrackingParticipant simTracker = new SimulatorTrackingParticipant()
{
TrackingProfile = new TrackingProfile()
{
Name = "CustomTrackingProfile",
Queries =
{
new CustomTrackingQuery()
{
Name = all,
ActivityName = all
},
new WorkflowInstanceQuery()
{
**States = {all },**
},
new ActivityStateQuery()
{
// Subscribe for track records from all activities for all states
ActivityName = all,
States = { all },
**Arguments = {all},**
// Extract workflow variables and arguments as a part of the activity tracking record
// VariableName = "*" allows for extraction of all variables in the scope
// of the activity
Variables =
{
{ all }
}
}
}
}
};
This will now correctly capture all workflow instance states rather than just Started/Completed. You will also capture all Arguments on each activity that records tracking data rather than just the variables. This is important because the 'variable' were interested in is actually (as discussed earlier) a DelegateInArgument.
So once we have changed the tracking profile we also need to change the SimulatorTrackingParticipant.cs to extract the additional data we are now tracking.
If you change the OnTrackingRecordReceived method to include the following sections these will capture variable data and also Argument data during execution.
protected void OnTrackingRecordReceived(TrackingRecord record, TimeSpan timeout)
{
System.Diagnostics.Debug.WriteLine(
String.Format("Tracking Record Received: {0} with timeout: {1} seconds.", record, timeout.TotalSeconds)
);
if (TrackingRecordReceived != null)
{
ActivityStateRecord activityStateRecord = record as ActivityStateRecord;
if (activityStateRecord != null)
{
IDictionary<string, object> variables = activityStateRecord.Variables;
StringBuilder vars = new StringBuilder();
if (variables.Count > 0)
{
vars.AppendLine("\n\tVariables:");
foreach (KeyValuePair<string, object> variable in variables)
{
vars.AppendLine(String.Format(
"\t\tName: {0} Value: {1}", variable.Key, variable.Value));
}
}
}
if (activityStateRecord != null)
{
IDictionary<string, object> arguments = activityStateRecord.Arguments;
StringBuilder args = new StringBuilder();
if (arguments.Count > 0)
{
args.AppendLine("\n\tArgument:");
foreach (KeyValuePair<string, object> argument in arguments)
{
args.AppendLine(String.Format(
"\t\tName: {0} Value: {1}", argument.Key, argument.Value));
}
}
//bubble up the args to the UI for the user to see!
}
if((activityStateRecord != null) && (!activityStateRecord.Activity.TypeName.Contains("System.Activities.Expressions")))
{
if (ActivityIdToWorkflowElementMap.ContainsKey(activityStateRecord.Activity.Id))
{
TrackingRecordReceived(this, new TrackingEventArgs(
record,
timeout,
ActivityIdToWorkflowElementMap[activityStateRecord.Activity.Id]
)
);
}
}
else
{
TrackingRecordReceived(this, new TrackingEventArgs(record, timeout,null));
}
}
}
Hope this helps!

Related

Why does Spring LDAP's LdapTemplate not return title, department & company attributes?

I'm using spring-ldap-core-2.3.1.RELEASE.jar over JDK 1.8 & Tomcat 8.0 to access AD information through LdapTemplate. The attributes such as title,department & company are not being returned by the ldapTemplate.search(..,.,..) method.
I'm using the following lines of code to search :-
LdapQuery ldapQuery = LdapQueryBuilder.query()
.where("objectclass").is("user")
.and("objectcategory").is("person")
.and("cn").like(strWildcardText+"*");
ldapTemplate.search(ldapQuery, new ADUserAttributesMapper());
Following is the ADUserAttributesMapper class :-
public class ADUserAttributesMapper implements AttributesMapper<ADUserBean> {
#Override
public ADUserBean mapFromAttributes(Attributes attributes) throws NamingException {
if(attributes==null) {
return null;
}
adUserBean.setName((attributes.get("name")!=null) ? attributes.get("name").get().toString() : null);
adUserBean.setCommonName((attributes.get("cn")!=null) ? attributes.get("cn").get().toString() : null);
adUserBean.setDisplayName((attributes.get("displayname")!=null) ? attributes.get("displayname").get().toString() : null);
adUserBean.setGivenName((attributes.get("givenname")!=null) ? attributes.get("givenname").get().toString() : null); // for FIRST NAME
adUserBean.setMiddleName((attributes.get("initials")!=null) ? attributes.get("initials").get().toString() : null); // for MIDDLE NAME / INITIALS
adUserBean.setLastName((attributes.get("sn")!=null) ? attributes.get("sn").get().toString() : null); // for LAST NAME
adUserBean.setDepartment((attributes.get("department")!=null) ? attributes.get("department").get().toString() : null);
adUserBean.setUserPrincipalName((attributes.get("userprincipalname")!=null) ? attributes.get("userprincipalname").get().toString() : null); // Logon Name
adUserBean.setsAMAccountName((attributes.get("samaccountname")!=null) ? attributes.get("samaccountname").get().toString() : null); // Logon Name (pre-Windows 2000)
adUserBean.setDistinguishedName((attributes.get("distinguishedname")!=null) ? attributes.get("distinguishedname").get().toString() : null);
adUserBean.setMailID((attributes.get("mail")!=null) ? attributes.get("mail").get().toString() : null);
adUserBean.setTitle((attributes.get("title")!=null) ? attributes.get("title").get().toString() : null); // Job Title
adUserBean.setTelephoneNumber((attributes.get("telephonenumber")!=null) ? attributes.get("telephonenumber").get().toString() : null);
adUserBean.setObjectCategory((attributes.get("objectcategory")!=null) ? attributes.get("objectcategory").get().toString() : null);
return adUserBean;
}
}
The title,department & company attributes belong to the Organization tab of the AD user properties as shown in the below image :-
Also, from the General tab the initials(initials) attribute is not being picked up/listed by Spring-LDAP's ldapTemplate. The LdapQueryBuilder.query() object has access to attributes(...) method that takes a string array of attribute names that are to be fetched. But even after mentioning them there explicitly, values for attributes such as initials, title, department & company are not returned.
The LDAP Browser plugin within the Eclipse IDE lists the title,department & company properties under the Organization tab without a problem.
Even the com4j API returns the title, department & company attributes.
Is there any configuration that is limiting the attribute(s) listing or is it a limitation with Spring-LDAP API itself? Are these attributes not part of BasicAttributes? How to fetch these attributes through Spring-LDAP?
UPDATE (01-Aug-2017):
The plain Java JNDI approach/code does NOT return department,company,title attributes (even with these attributes being explicitly mentioned in attributes string array), but surprisingly it does return the initials attribute value.
UPDATE (02-Aug-2017):
Similar to #Pierre's suggestion (below) tried the following code using SearchControls object :-
String strFilter= "(&(objectclass=top)(cn=cgma*))";
String[] attrs = new String[] {"cn","givenName","sn","initials","title","department","company"};
long maxResults = 10; // for example
SearchControls searchControls = new SearchControls();
searchControls.setSearchScope(SearchControls.SUBTREE_SCOPE);
searchControls.setReturningAttributes(attrs);
searchControls.setCountLimit(maxResults);
List<String> aLstOfADUsers = ldapTemplate.search("",strFilter,searchControls,new AttributesMapper<String>()
{
public String mapFromAttributes(Attributes attrs) throws NamingException {
try
{
System.out.println(attrs.toString());
return attrs.get("cn").get().toString();
}
catch(Exception ex) {
ex.printStackTrace();
return null;
}
}
});
return aLstOfADUsers;
Even this does not return the initials, title, company & department attribute values.
The person attributes might be internal attributes which you wouldn't get back by default. You can specify explicitly which attributes you want returned BUT not in the search method you're using (the one where you pass in an LdapQuery object). If you take a look at the org.springframework.ldap.core.LdapTemplate class, it doesn't seem like you can pass in the SearchControls object to the method signature you're using. So, to be able to specify attributes to fetch, replace this:
LdapQuery ldapQuery = LdapQueryBuilder.query()
.where("objectclass").is("user")
.and("objectcategory").is("person")
.and("cn").like(strWildcardText+"*");
ldapTemplate.search(ldapQuery, new ADUserAttributesMapper());
With this:
LikeFilter filter = new LikeFilter("cn", strWildcardText+"*");
// list of attributes to retrieve
String[] attrs = new String[] {"title","department","company"};
long maxResults = 10; // for example
SearchControls searchControls = new SearchControls();
searchControls.setSearchScope(SearchControls.SUBTREE_SCOPE);
searchControls.setReturningAttributes(attrs);
searchControls.setCountLimit(numResults);
ldapTemplate.search(DistinguishedName.EMPTY_PATH, filter.encode(), searchControls, new ADUserAttributesMapper());
The above should work. You could also try something like this (I haven't tried that yet):
ldapTemplate.search( "dc=yourorg,dc=com",
"(&(cn=" +strWildcardText + "*)(&(objectClass=person)(objectcategory=person)))",
SearchControls.SUBTREE_SCOPE,
new String[]{ "title","department","company" },
new ADUserAttributesMapper() );
Finally, to get ALL attributes back, ask to retrieve ALL attributes in the code above (my above example only asked for 3 attributes, this would return ALL of them):
String[] attrs = new String[]{"*","+"};
This is based on your AttributesMapper. I don't know what ADUserAttributesMapper is, so you'd have to provide that implementation.
Here's the javadoc for this interface. http://docs.spring.io/spring-ldap/docs/current/apidocs/org/springframework/ldap/core/AttributesMapper.html
Change ldap port from 3268 to 389

How to skip displaying a content item in Orchard CMS?

I have a content part that provides a begin timestamp and end timestamp option. These 2 fields are used to define a period of time in which the content item should be displayed.
I now have difficulties to implement a skip approach whereas content items should not be displayed / skipped when the period of time does not span the current time.
Digging in the source code and trying to find an entry point for my approach resulted in the following content handler
public class SkipContentHandler : Orchard.ContentManagement.Handlers.ContentHandler
{
protected override void BuildDisplayShape(Orchard.ContentManagement.Handlers.BuildDisplayContext aContext)
{
if (...) // my condition to process only content shapes which need to be skipped
{
aContext.Shape = null; // return null shape to skip it
}
}
}
This works but there are several side effects
I had to alter the source code of BuildDisplayContext as the Shape is normally read only
List shape may displayed a wrong pager when it contains content items with my content part because the Count() call in ContainerPartDriver.Display() is executed before BuildDisplay()
calling the URL of a content item that is skipped results in an exception because View(null) is abigious
So, what would be the correct approach here or is there any module in existence that does the job? I couldn't find one.
This is a quite complex task. There are several steps needed to achieve a proper skipping of display items:
Create the part correctly
There are a few pitfalls here as when coming to the task of adding a part view one might utilize Orchards date time editor in connection with the DateTime properties. But this brings a heck of a lot of additional issues to the table but these don't really relate to the question.
If someone is interested in how to use Orchards date time editor then i can post this code too, but for now it would only blow up the code unnecessarly.
So here we go, the part class...
public class ValidityPart : Orchard.ContentManagement.ContentPart<ValidityPartRecord>
{
// public
public System.DateTime? ValidFromUtc
{
get { return Retrieve(r => r.ValidFromUtc); }
set { Store(r => r.ValidFromUtc, value); }
}
...
public System.DateTime? ValidTillUtc
{
get { return Retrieve(r => r.ValidTillUtc); }
set { Store(r => r.ValidTillUtc, value); }
}
...
public bool IsContentItemValid()
{
var lUtcNow = System.DateTime.UtcNow;
return (ValidFromUtc == null || ValidFromUtc.Value <= lUtcNow) && (ValidTillUtc == null || ValidTillUtc.Value >= lUtcNow);
}
...
}
...and the record class...
public class ValidityPartRecord : Orchard.ContentManagement.Records.ContentPartRecord
{
// valid from value as UTC to use Orchard convention (see CommonPart table) and to be compatible with projections
// (date/time tokens work with UTC values, see https://github.com/OrchardCMS/Orchard/issues/6963 for a related issue)
public virtual System.DateTime? ValidFromUtc { get; set; }
// valid from value as UTC to use Orchard convention (see CommonPart table) and to be compatible with projections
// (date/time tokens work with UTC values, see https://github.com/OrchardCMS/Orchard/issues/6963 for a related issue)
public virtual System.DateTime? ValidTillUtc { get; set; }
}
Create a customized content query class
public class MyContentQuery : Orchard.ContentManagement.DefaultContentQuery
{
// public
public ContentQuery(Orchard.ContentManagement.IContentManager aContentManager,
Orchard.Data.ITransactionManager aTransactionManager,
Orchard.Caching.ICacheManager aCacheManager,
Orchard.Caching.ISignals aSignals,
Orchard.Data.IRepository<Orchard.ContentManagement.Records.ContentTypeRecord> aContentTypeRepository,
Orchard.IWorkContextAccessor aWorkContextAccessor)
: base(aContentManager, aTransactionManager, aCacheManager, aSignals, aContentTypeRepository)
{
mWorkContextAccessor = aWorkContextAccessor;
}
protected override void BeforeExecuteQuery(NHibernate.ICriteria aContentItemVersionCriteria)
{
base.BeforeExecuteQuery(aContentItemVersionCriteria);
// note:
// this method will be called each time a query for multiple items is going to be executed (e.g. content items of a container, layers, menus),
// this gives us the chance to add a validity criteria
var lWorkContext = mWorkContextAccessor.GetContext();
// exclude admin as content items should still be displayed / accessible when invalid as validity needs to be editable
if (lWorkContext == null || !Orchard.UI.Admin.AdminFilter.IsApplied(lWorkContext.HttpContext.Request.RequestContext))
{
var lUtcNow = System.DateTime.UtcNow;
// left outer join of ValidityPartRecord table as part is optional (not present on all content types)
var ValidityPartRecordCriteria = aContentItemVersionCriteria.CreateCriteria(
"ContentItemRecord.ValidityPartRecord", // string adopted from foreach loops in Orchard.ContentManagement.DefaultContentQuery.WithQueryHints()
NHibernate.SqlCommand.JoinType.LeftOuterJoin
);
// add validity criterion
ValidityPartRecordCriteria.Add(
NHibernate.Criterion.Restrictions.And(
NHibernate.Criterion.Restrictions.Or(
NHibernate.Criterion.Restrictions.IsNull("ValidFromUtc"),
NHibernate.Criterion.Restrictions.Le("ValidFromUtc", lUtcNow)
),
NHibernate.Criterion.Restrictions.Or(
NHibernate.Criterion.Restrictions.IsNull("ValidTillUtc"),
NHibernate.Criterion.Restrictions.Ge("ValidTillUtc", lUtcNow)
)
)
);
}
}
// private
Orchard.IWorkContextAccessor mWorkContextAccessor;
}
This essentially adds a left join of the validity part fields to the SQL query (content query) and extends the WHERE statement with the validity condition.
Please note that this step is only possible with the solution described the following issue: https://github.com/OrchardCMS/Orchard/issues/6978
Register the content query class
public class ContentModule : Autofac.Module
{
protected override void Load(Autofac.ContainerBuilder aBuilder)
{
aBuilder.RegisterType<MyContentQuery>().As<Orchard.ContentManagement.IContentQuery>().InstancePerDependency();
}
}
Create a customized content manager
public class ContentManager : Orchard.ContentManagement.DefaultContentManager
{
// public
public ContentManager(
Autofac.IComponentContext aContext,
Orchard.Data.IRepository<Orchard.ContentManagement.Records.ContentTypeRecord> aContentTypeRepository,
Orchard.Data.IRepository<Orchard.ContentManagement.Records.ContentItemRecord> aContentItemRepository,
Orchard.Data.IRepository<Orchard.ContentManagement.Records.ContentItemVersionRecord> aContentItemVersionRepository,
Orchard.ContentManagement.MetaData.IContentDefinitionManager aContentDefinitionManager,
Orchard.Caching.ICacheManager aCacheManager,
System.Func<Orchard.ContentManagement.IContentManagerSession> aContentManagerSession,
System.Lazy<Orchard.ContentManagement.IContentDisplay> aContentDisplay,
System.Lazy<Orchard.Data.ITransactionManager> aTransactionManager,
System.Lazy<System.Collections.Generic.IEnumerable<Orchard.ContentManagement.Handlers.IContentHandler>> aHandlers,
System.Lazy<System.Collections.Generic.IEnumerable<Orchard.ContentManagement.IIdentityResolverSelector>> aIdentityResolverSelectors,
System.Lazy<System.Collections.Generic.IEnumerable<Orchard.Data.Providers.ISqlStatementProvider>> aSqlStatementProviders,
Orchard.Environment.Configuration.ShellSettings aShellSettings,
Orchard.Caching.ISignals aSignals,
Orchard.IWorkContextAccessor aWorkContextAccessor)
: base(aContext, aContentTypeRepository, aContentItemRepository, aContentItemVersionRepository, aContentDefinitionManager, aCacheManager, aContentManagerSession,
aContentDisplay, aTransactionManager, aHandlers, aIdentityResolverSelectors, aSqlStatementProviders, aShellSettings, aSignals)
{
mWorkContextAccessor = aWorkContextAccessor;
}
public override ContentItem Get(int aId, Orchard.ContentManagement.VersionOptions aOptions, Orchard.ContentManagement.QueryHints aHints)
{
var lResult = base.Get(aId, aOptions, aHints);
if (lResult != null)
{
// note:
// the validity check is done here (after the query has been executed!) as changing base.GetManyImplementation() to
// apply the validity critera directly to the query (like in ContentQuery) will not work due to a second attempt to retrieve the
// content item from IRepository<> (see base.GetManyImplementation(), comment "check in memory") when the query
// returns no data (and the query should not return data when the validity critera is false)
//
// http://stackoverflow.com/q/37841249/3936440
var lWorkContext = mWorkContextAccessor.GetContext();
// exclude admin as content items should still be displayed / accessible when invalid as validity needs to be editable
if (lWorkContext == null || !Orchard.UI.Admin.AdminFilter.IsApplied(lWorkContext.HttpContext.Request.RequestContext))
{
var lValidityPart = lResult.As<ValidityPart>();
if (lValidityPart != null)
{
if (lValidityPart.IsContentItemValid())
{
// content item is valid
}
else
{
// content item is not valid, return null (adopted from base.Get())
lResult = null;
}
}
}
}
return lResult;
}
// private
Orchard.IWorkContextAccessor mWorkContextAccessor;
}
Steps 2-4 are needed when having content items whereas the content type has a Container and Containable part or even content items which are processed / displayed separately. Here you normally cannot customize the content query that is executed behind the scenes.
Steps 2-4 are not needed if you use the Projection module. But again, this brings a few other issues to the table as reported in this issue: https://github.com/OrchardCMS/Orchard/issues/6979

Create tree using multithreading and recursion

This question is specifically related creating a tree using multithreading and recursion.
I have got the code running that will create the tree using recursion but the time required to create that tree is more than I want to spend.
The reason for the slowness of that is because I am calling TaxonomyManager in Ektron CMS which takes a little bit to return and all the calls add up quickly. I was wondering if there is a way to create a tree using multithreading.
(I don't have the code at present with me but I will add that code as soon as I get access to that code).
If I go this route what the chances of me corrupting the tree as the tree is one root node and multithreading is going to add those nodes to that node at some point.
Thanks for any input anyone may have.
Edit: Added code. TaxonomyNodes is my class doesn't have a lot of properties. Has Id,Name,Description, Path (Stores the path in similar way as Ektron), HasChildren flag, ParentId, and public List Children.
public List<TaxonomyNodes> CreateTree()
{
try
{
TaxonomyManager tManager = new TaxonomyManager();
TaxonomyCriteria criteria = new TaxonomyCriteria();
criteria.AddFilter(TaxonomyProperty.ParentId, CriteriaFilterOperator.EqualTo, 0);
criteria.OrderByDirection = EkEnumeration.OrderByDirection.Ascending;
criteria.OrderByField = TaxonomyProperty.Id;
List<TaxonomyData> tDataList = tManager.GetList(criteria);
int index = 0;
if (tDataList != null)
{
foreach (TaxonomyData item in tDataList)
{
if (item.Name != "Companies" && item.Name != "Content Information Centers")
root.Insert(index++, new TaxonomyNodes() { ParentId = 0, TaxonomyId = item.Id, TaxonomyDescription = item.Description, TaxonomyName = item.Name, TaxonomyPath = item.Path, HasChildren = item.HasChildren, Children = new List<TaxonomyNodes>() });
}
}
index = 0;
foreach (TaxonomyNodes itemT in root)
{
itemT.Children = CreateNodes(itemT.TaxonomyId, itemT);
}
return root;
}
catch (Exception)
{
throw;
}
}
private List<TaxonomyNodes> CreateNodes(long taxonomyId, TaxonomyNodes itemToAddTo)
{
try
{
TaxonomyManager tManager = new TaxonomyManager();
TaxonomyCriteria criteria = new TaxonomyCriteria();
criteria.AddFilter(TaxonomyProperty.ParentId, CriteriaFilterOperator.EqualTo, taxonomyId);
criteria.OrderByDirection = EkEnumeration.OrderByDirection.Ascending;
criteria.OrderByField = TaxonomyProperty.Id;
List<TaxonomyData> tDataList = tManager.GetList(criteria);
List<TaxonomyNodes> node = new List<TaxonomyNodes>();
if (tDataList != null)
{
foreach (TaxonomyData item in tDataList)
{
node.Add(new TaxonomyNodes() { ParentId = taxonomyId, Children = null, TaxonomyId = item.Id, TaxonomyDescription = item.Description, TaxonomyName = item.Name, TaxonomyPath = item.Path, HasChildren = item.HasChildren });
itemToAddTo.Children = node;
if (item.HasChildren)
{
CreateNodes(item.Id, node[node.Count - 1]);
}
else
{
return node;
}
}
}
return node;
}
catch (Exception)
{
throw;
}
}
Rather than get into multithreading, which, though it may work, is not officially supported by Ektron's APIs and may present other challenges, I would recommend some form of caching or other storage for the data that does not require recursive DB calls.
Options:
1 - Ektron's Taxonomy APIs do include a GetTree method, which can pull the entire tree, up to a specified number of levels, and child items in a single API call rather than recursively. This may perform better and would be easily cached.
2 - Ektron provides API-level caching that can be readily enabled in the web.config by changing
<framework defaultContainer="Default" childContainer="BusinessObjects" />
To:
<framework defaultContainer="Cache" childContainer="BusinessObjects" />
3 - Use an eSync Strategy which will output the data you need (better to use your own streamlined objects than to store Ektron's with all the extra data) to something like an XML file. See this sample, http://developer.ektron.com/Templates/CodeLibraryDetail.aspx?id=1989&blogid=116, which I wrote to do this very thing. It hooks into the DB sync complete event and triggers a console application which writes the entire Taxonomy structure out to an XML file.

Access Variable<int> in Custom Activity in WF4

I need to count how many times i iterate in a flowchart flow, but i need to be able to read and preferably write to the variable in a custom activity.
My current attempt is declaring the var in design view with scope of the entire Flowchart, default value 0 and incrementing using an Assign activity. But i cannot figure out how i can access the variable in a custom activity without resetting it.
My attempt to access the var is something like whats described in the answer here: Declare Variable<T> variable in a CodeActivity in windows workflow 4.0
Only i don't use a default value for the var when declaring though. Still it seems the var is not in any way related to the var i have defined in design view. I have also tried defining it in code only but then i cannot access it in for example a regular Assign activity.
So what can i do to be able to use the var as a "global" variable?
Thanks.
The most intuitive and perhaps correct way of doing it is to pass the variable that you're declaring on Flowchart level to inside your custom activity. Then you can do whatever you want with it's value and return it.
An example of a custom increment activity (this is how Assign activity works too):
public class IncrementActivity : CodeActivity<int>
{
[RequiredArgument]
public InArgument<int> CountVariable { get; set; }
protected override int Execute(CodeActivityContext context)
{
// Do whatever logic you want here
return CountVariable.Get(context) + 1;
}
}
Here it is a sample of usage using a Sequence (the same when using Flowchart):
var countVar = new Variable<int>("count");
var activity = new Sequence
{
Variables =
{
// declare counter variable at global scope
countVar
},
Activities =
{
new WriteLine { Text = new VisualBasicValue<string>(#"""Count: "" & count") },
new IncrementActivity { CountVariable = countVar, Result = countVar },
new WriteLine { Text = new VisualBasicValue<string>(#"""Count: "" & count") },
new IncrementActivity { CountVariable = countVar, Result = countVar },
new WriteLine { Text = new VisualBasicValue<string>(#"""Count: "" & count") },
new IncrementActivity { CountVariable = countVar, Result = countVar }
}
};
Output:
Count: 0
Count: 1
Count: 2
Note that is simpler through visual designer as you don't have to directly use VisualBasicValue<string> to build print string. Other than that, is exactly the same!

How to get a diff of pending changes to a model in ASP.NET MVC 2

I am working on an ASP.Net MVC app and I want to show a confirmation page after the user edits some data. What I would like to show is a list of the pending changes that the user made to the model.
For example,
Are you sure you want to make the following changes:
FieldName:
Previous Value: XXX
New Value: YYY
I know I can read my stored value from the database and compare it with the POSTed object but I want this to work generally. What would be some good ways to approach this?
To clarify, I am looking for a general way to get a "diff" of the pending changes. I already know how to get the previous and pending changes. Kind of like how TryUpdateModel() can attempt to update any Model with posted values. I'd like a magical GetPendingModelChanges() method that can return a list of something like new PendingChange { Original = "XXX", NewValue = "YYY"} objects.
You might be doing this already but I wouldn't send my model to the view, create a viewmodel. In this case I would map the model data to the viewmodel twice, my viewmodel might contain OrderInput and OrderInputOrig. Then stick OrderInputOrig in hidden fields. On post back you can compare the values and then redirect, if something changed, to a display view with the original and the changes for confirmation.
Maybe something like this:
[HttpPost]
public ActionResult Edit(CustomerInput cutomerInput)
{
var changes = PublicInstancePropertiesEqual(cutomerInput.OriginalCustomer, cutomerInput.Customer);
if (changes != null)
{
cutomerInput.WhatChangeds = changes;
return View("ConfirmChanges", cutomerInput);
}
return View();
}
public ActionResult ConfirmChanges(CustomerInput customerInput)
{
return View(customerInput);
}
from: Comparing object properties in c#
public static Dictionary<string, WhatChanged> PublicInstancePropertiesEqual<T>(T self, T to, params string[] ignore) where T : class
{
Dictionary<string, WhatChanged> changes = null;
if (self != null && to != null)
{
var type = typeof(T);
var ignoreList = new List<string>(ignore);
foreach (System.Reflection.PropertyInfo pi in type.GetProperties(System.Reflection.BindingFlags.Public | System.Reflection.BindingFlags.Instance))
{
if (!ignoreList.Contains(pi.Name))
{
var selfValue = type.GetProperty(pi.Name).GetValue(self, null);
var toValue = type.GetProperty(pi.Name).GetValue(to, null);
if (selfValue != toValue && (selfValue == null || !selfValue.Equals(toValue)))
{
if (changes == null)
changes = new Dictionary<string, WhatChanged>();
changes.Add(pi.Name, new WhatChanged
{
OldValue = selfValue,
NewValue=toValue
});
}
}
}
return changes;
}
return null;
}
Coming in very late here, but I created a library to do this on MVC models and providing "readable" diffs for humans using MVC ModelMetadata:
https://github.com/paultyng/ObjectDiff
It gives me output when I save a Model similar to:
Status: 'Live', was 'Inactive'
Phone: '123-456-7898', was '555-555-5555'
Etc.
use the TempData Dictionary.
TempData["previousValue"];
TempData["newValue"];

Resources