I've been checking ServiceStack's documentation, but I haven't found a way to do many to many relationships with ServiceStack.OrmLite, is it supported? Is there a workaround (without writing raw sql)?
I'd like to have something like this:
Article <- ArticleToTag -> Tag
Thanks!!
It's not implicitly handled automatically for you behind the scenes if that's what you mean? But as OrmLite is just a thin wrapper around ADO.NET interfaces anything is possible.
In OrmLite, by default every POCO maps 1:1 with a table. So if you wanted the table layout you would create it just as it looks in your database, e.g.
var article = new Article { ... };
var tag = new Tag { ... };
var articleTag = new ArticleTag { ArticleId = article.Id, TagId = tag.Id };
db.Insert(article, tag, articleTag);
Although you might want to take advantage of the built-in blobbing in OrmLite where any complex type just gets serialized and stored in a single text field. So you could do something like:
var article = { new Article { Tags = { "A","B","C" } };
Where Tags is just a List<string> and OrmLite will take care of transparently serializing it in the database field for you.
Related
This is a really trivial problem. I am just curious on how to deal with this in a "professional" manner.
I am trying to stick to variable naming convention. For NodeJs I am doing camelCasing. For database, I am using PostgreSQL and using underscore_casing.
Now the problem arises when I query data from PostgreSQL. I'll get a user object with following format,
{user_id: 1, account_type : "Admin"}
I can pass this object directly to server side-render and will have to use underscore casing to access account_type. Of course, I can manually create a new user JSON object with property userId and accountType but that is unnecessary work.
Is it possible to follow variable naming convention for both language and avoid having mixed variable names casing in some files? What is a good way to stay organized?
The are two good ways to approach this issue. The simplest one - do no conversion, use the exact database names. And the second one is to camel-case columns automatically.
Either way, you should always follow the underscore notation for all PostgreSQL declarations, as it will give you the option to activate camel-casing in your app at a later time, if it becomes necessary. Never use camel-case inside the database, or you will end up in a lot of pain later.
If you want the best of both worlds, follow the underscore notation for all PostgreSQL declarations, and convert to camel-case as you read data.
Below is an example of how to do it properly with pg-promise, copied from event receive example:
// Example below shows the fastest way to camelize column names:
const options = {
receive(e) {
camelizeColumns(e.data);
}
};
function camelizeColumns(data) {
const template = data[0];
for (var prop in template) {
const camel = pgp.utils.camelize(prop);
if (!(camel in template)) {
for (var i = 0; i < data.length; i++) {
const d = data[i];
d[camel] = d[prop];
delete d[prop];
}
}
}
}
Also see the following article: Pg-promise and case sensitivity in column names.
UPDATE
The code above has been updated for use of pg-promise v11 or later.
I've struggled with this too, and I've concluded that there's really no way to avoid this kind of ugliness unless you rewrite the objects that come from the database. Fortunately, that's not too difficult in Javascript:
const fromDBtoJS = (obj) => {
// declare a variable to hold the result
const result = {};
// iterate over the keys on the object
Object.keys(obj).forEach((key) => {
// adjust the key
const newKey = key.replace(/_[a-z]/g, (x) => x[1].toUpperCase());
// add the value from the old object with the new key
result[newKey] = obj[key];
});
// return the result
return result;
};
Here's a JSFiddle. The "replace" code above was found here
If you wanted to use classes for models in your application, you could incorporate this code into the constructor or database load method so it's all handled more-or-less automatically.
I'm running into some n+1 performance issues when iterating over a collection of ContentItems of a custom Type that I created solely through migrations.
ContentDefinitionManager.AlterPartDefinition("MyType", part => part
.WithField("MyField", field => field
...
)
);
ContentDefinitionManager.AlterTypeDefinition("MyType", type => type
.WithPart("MyType")
);
Every time I access a field of this part a new query is performed. I can use QueryHints to avoid this for the predefined parts
var myItems = _orchardServices.ContentManager.Query().ForType("MyType")
.WithQueryHints(new QueryHints().ExpandParts<LocalizationPart()
...
);
but can I do this for the ContentPart of my custom type too? This does not seem to work:
var myItems = _orchardServices.ContentManager.Query().ForType("MyType")
.WithQueryHints(new QueryHints().ExpandParts<ContentPart>()
...
);
How can I tell Orchard to just get everything in one go? I'd prefer to be able to do this without writing my own HQL or directly addressing the repositories.
Example:
var myItems = _orchardServices.ContentManager.Query().ForType("MyType");
#foreach(var item in myItems.Take(100)) {
foreach(var term in item.Content.MyItem.MyTaxonomyField.Terms) {
// Executes 100 queries
<div>#term.Name</div>
}
}
TaxonomyField doesn't store ids and using the TaxonomyService inside of the loop wouldn't improve performance. Right now, to work around this, I fetch all TermContentItems.Where(x => myItems.Select(i => i.Id).Contains(TermPartRecord.Id)) from the repository outside of the loop as well as a list of all the terms of the Taxonomy that the field is using. Then inside the loop:
var allTermsInThisField = termContentItems.Where(tci => tci.TermsPartRecord.Id == c.Id)
.Select(tci => terms.Where(t => t.Id == tci.TermRecord.Id).Single()).ToList()
I'm not a very experienced programmer but this was the only way I could see how to do this without digging into HQL and it seems overly complicated for my purposes. Can Orchard do this in less steps?
I need to write a service that connects to CRM, and returns with a list of all of the entity available on the server (custom or otherwise).
How can I do this? To be clear, I am not looking to return all data for all entities. Just a list of every type, regardless of whether any actually exist.
You need to use RetrieveAllEntitiesRequest
RetrieveAllEntitiesRequest request = new RetrieveAllEntitiesRequest()
{
EntityFilters = EntityFilters.Entity,
RetrieveAsIfPublished = true
};
// service is the IOrganizationService
RetrieveAllEntitiesResponse response = (RetrieveAllEntitiesResponse)service.Execute(request);
foreach (EntityMetadata currentEntity in response.EntityMetadata)
{
string logicalName = currentEntity.LogicalName;
// your logic here
}
note that you will get also system or hidden entities, like wizardpage or recordcountsnapshot
You will probably find these sections of the MSDN useful:
Customize Entity Metadata (lookout for the samples linked on that page).
Retrieve and Detect Changes to Metadata.
I'm trying to retrieve a list of entities from CRM, but I'd like to get each one with the related entities. So far, I've the following code:
FilterExpression filterExpression = new FilterExpression();
ConditionExpression condition = new ConditionExpression(Constants.ModifiedOnAttribute, ConditionOperator.GreaterEqual, lastSync);
filterExpression.AddCondition(condition);
QueryExpression query = new QueryExpression()
{
EntityName = entityName,
ColumnSet = new ColumnSet(attributesMetadata.Select(att => att.Name).ToArray<string>()),
Criteria = filterExpression,
Distinct = false,
NoLock = true
};
RetrieveMultipleRequest multipleRequest = new RetrieveMultipleRequest();
multipleRequest.Query = queryExpression;
RetrieveMultipleResponse response = (RetrieveMultipleResponse)proxy.Execute(multipleRequest);
In the variable response, I can see the EntityCollection attribute, but inside, Related entities always come empty.
I'd like to know if it is possible to retrieve the set of a given entities, with the related entities, using RetrieveMultipleRequest, rather than go one by one using RetrieveRequest.
One approach to retreive related entities data - adding LinkEntities to your query. Example below will make you an idea how to make this:
LinkEntity linkEntity = new LinkEntity("email", "new_emails", "activityid", "new_relatedemail", JoinOperator.Inner);
linkEntity.Columns.AddColumn("versionnumber");
linkEntity.Columns.AddColumn("new_emailsid");
linkEntity.EntityAlias = "related";
query = new QueryExpression("email");
query.ColumnSet.AddColumn("activityid");
query.ColumnSet.AddColumn("versionnumber");
query.Criteria.AddCondition("modifiedon", ConditionOperator.NotNull);
query.LinkEntities.Add(linkEntity);
And then you can access attributes from related entities using EntityAlias you specified above:
foreach (Entity entity in entities.Entities)
{
if ((long)(entity["related.versionnumber"] as AliasedValue).Value > 0)
{
stop = false;
}
}
The RetrieveMultipleRequest is for returning multiple instances of a particular type of entity. I have spent a year using the CRM SDK from C# and I have found no way of populating those related entity collections in a single query. This basically leaves you with two options:
Use the AliasedValue as SergeyS recommends. Remember when querying 1:Many relationships, be aware that you could be returning multiple results for the same parent entity. This is what I use most of the time.
Perform a second query for each relationship you want access to. You'll probably get better performance if you can use an IN statement in your second query, based on the results of the first, rather than performing a separate query for each result of the first.
Below is some pseudo code to show the difference.
var contacts = GetContacts();
// One Request to get the cars for the contacts
var cars = GetCarsWhereContactIdIn(contacts.Select( c => c.new_ContactId));
foreach(var c in contacts){
c.new_Cars.AddRange(cars.where(car => car.new_contactId = c.ContactId));
}
// Verses
var contacts = GetContacts();
foreach(var c in contacts){
// One Request for each contact
c.new_Cars.AddRange(GetCarsForContact(c.ContactId));
}
I'm looking for a way to get all records where deleted is set to true on a particular table. How might I accomplish this?
Note: Using auto-generated class by SubSonic. Not T-SQL.
The auto-generated SubSonic classes don't support querying logical deletes. But you can do this (version 2.1/2.2 syntax):
public partial class TableClassCollection
{
public TableClassCollection LoadAll(bool suppressLogicalDeletes)
{
SubSonic.SqlQuery q = new SubSonic.Select(TableClass.Schema)
.From(TableClass.Schema);
if (suppressLogicalDeletes)
{
q.Where(TableClass.DeletedColumn).IsEqualTo(false);
}
return q.ExecuteAsCollection<TableClassCollection>();
}
}
More examples at subsonicproject.com
I've never heard of SubSonic before, but a quick Google search turned up: Select Queries in SubSonic.
So, using that page as a guide, it sounds like you'd be able to write your query as:
FooCollection deletedFoos = // use the generated collection class
DB.Select().From("FooTable") // table name goes here
.Where("deleted").IsEqualTo(true) // might need 1, depends on database?
.ExecuteAsCollection<FooCollection>(); // should match the type above
not a lot of detail in your question, but assuming there's a column named "deleted," it would look something like this:
select * from tableName where deleted = true