Im trying to create a persistent sqllite db (creation tables once you install the app, deletion of db when you uninstall app)
I have a issue that I cant save my sub objects for example
public class ObjectInstanceResponseModel : GenericResponseModel
{
public ObservableCollection<ObjectInstanceData> ObjectInstances { get; set; }
}
public class ObjectInstanceData : GenericResponseModel
{
[PrimaryKey, AutoIncrement]
public int Id { get; set; }
[JsonProperty("idObjectinstance")]
public int IdObjectInstance { get; set; }
[JsonProperty("objectclass_idObjectclass")]
public int ObjectClassIdObjectClass { get; set; }
[JsonProperty("objectclassname")]
public string ObjectClassName { get; set; }
[JsonProperty("visibilitylevel")]
public int VisibilityLevel { get; set; }
[JsonProperty("showname")]
public bool ShowName { get; set; }
[JsonProperty("showicon")]
public bool ShowIcon { get; set; }
[JsonProperty("creationtime")]
public DateTimeOffset CreationTime { get; set; }
[JsonProperty("users_idUsers")]
public int UsersIdUsers { get; set; }
[JsonProperty("isfavorite")]
public bool? IsFavorite { get; set; }
[OneToMany("ObjectInstanceDataStrings")]
[JsonProperty("strings")]
public List<String> Strings { get; set; }
}
public class String
{
[PrimaryKey, AutoIncrement]
[JsonProperty("idObjectparameterstring")]
public int? IdObjectParameterString { get; set; }
[ForeignKey(typeof(ObjectInstanceData))]
public int ObjectInstanceDataStrings { get; set; }
[JsonProperty("stringvalue")]
public string StringValue { get; set; }
[JsonProperty("objectparameterset_idObjectparameterset")]
public int? ObjectParameterSetIdObjectParameterSet { get; set; }
[JsonProperty("showinballon")]
public bool? ShowInBallon { get; set; }
[JsonProperty("idClassparameter")]
public int IdClassParameter { get; set; }
[JsonProperty("classparametername")]
public string ClassParameterName { get; set; }
}
So, my class String is always empty, although there are some rows in the table that I created Strings..
Am I need a lazy loading for this?
I implemented sqllite through depedency service in my app.cs like this:
public partial class App : Application
{
public static SQLiteConnection DatabaseConnection { get; set; }
public App()
{
InitializeComponent();
DatabaseConnection = DependencyService.Get<IConnection>().GetConnection();
CreateTables();
}
private void CreateTables()
{
DatabaseConnection.CreateTable<ObjectInstanceData>();
DatabaseConnection.CreateTable<Models.Objects.String>();
}
protected override void OnStart()
{
}
protected override void OnSleep()
{
}
protected override void OnResume()
{
}
}
So, basically the logic should be when no internet, work with sql lite (keep local changes), and when internet come back upload that changes that kept in db, and erase data from tables.
You noticed that Im using response model for api.
So, Im calling from my FavoriteObjectViewModel this:
var response = await ApiServiceProvider.GetObjectInstances(null, true);
and in the ApiServiceProvider:
public static async Task<ObjectInstanceResponseModel> GetObjectInstances(string queryString = null, bool? onlyFavorites = null)
{
HttpResponseMessage response;
response = await apiClient.GetAsync(objectInstancesEndpoint);
var resultContent = await response.Content.ReadAsStringAsync();
var result = JsonConvert.DeserializeObject<ObservableCollection<ObjectInstanceData>>(resultContent);
objectInstanceResponse.ObjectInstances = result;
//after each api call I need to update db
//delete previous data, and add fresh data from api
App.DatabaseConnection.DeleteAll<ObjectInstanceData>();
foreach (var item in result)
{
App.DatabaseConnection.Insert(item);
if (item.Strings != null && item.Strings.Count > 0)
App.DatabaseConnection.InsertAll(item.Strings);
}
//I only get the data for ObjectInstanceData, Strings model is empty!
var objectsResponseDb = App.DatabaseConnection.GetAllWithChildren<ObjectInstanceData>();
objectInstanceResponse.Succeeded = true;
return objectInstanceResponse;
}
So, my questions are:
If I create tables each time in App.cs is that mean, that I not store data, when user quit application, and re-enter again?
Why is the model Strings empty? When I invoke var strings = App.DatabaseConnection.GetAllWithChildren<Models.Objects.String>(); I can see that there is data?
What is the best approach for doing offline "logging"? maybe there is a some better nuget for sqllite?
I don't know much about persistent databases but one thing I can tell you for sure:you're approach is wrong.
You should create a separate class for the database's logic like creating tables and instantiating the connection along with other methods for manipulating data.
In the App class you can create a static resource of the database class which you can call it and make use of the methods in it.
I'm not very good with explaining but here it is a very basic example of how it should look:https://learn.microsoft.com/en-us/xamarin/get-started/tutorials/local-database/?tutorial-step=2&tabs=vswin.
Related
I have these classes
public class Project
{
public Guid Id { get; set; }
public string DeployUrl { get; set; }
public List<Technology> Techs { get; set; } = new();
}
public class Technology
{
public Guid Id { get; set; }
public string Name { get; set; }
public string TechImg { get; set; }
[JsonIgnore]
public List<Project> Projects { get; set; } = new();
[JsonIgnore]
public DateTime CreatedAt { get; set; } = DateTime.UtcNow;
}
Which creates a many-to-many relationship. I want to add an endpoint to create a Project. The main issue here is the approach with which i'll do it. Since I don't send the full Technology object to the frontend, it would be impossible to send back a full Project as the request body for the endpoint since its Technology property would be missing fields, which posses an incompatibility with the actual entity in my server. Of course that the Technology instance already exists. So, what is the best approach here? Currently I'm sending a Project DTO that doesn't contain a List<Technology> but a List<Guid> of technologies, and in my repository I fetch from the DB all those technologies and add them manually. Is this correct? Thanks in advance.
public class ProjectPostDto
{
public string DeployUrl { get; set; }
public List<Guid> TechsIds { get; set; } = new();
}
public async Task<ActionResult> PostProject(ProjectPostDto projectDto)
{
try
{
Project newProject = FromPostDto(projectDto);
newProject.Techs = await Service.GetTechs(projectDto.TechsIds);
await Service.Create(newProject);
ProjectPostDto newProjectDto = ToPostDto(newProject);
return CreatedAtAction(nameof(GetProject), new { id = newProject.Id }, newProjectDto);
} catch (Exception)
{
return StatusCode(500, "Server error");
}
}
I want to push and get Recipe-Data from my Firebase Database in Xamarin.Forms with the Firesharp Plugin.
My Model Class is the Recipe Class:
public class Recipe
{
public string title { get; set; }
public string workTime { get; set; }
public string degreeOfDifficulty { get; set; }
public string caloriesPerPerson { get; set; }
public string preparation { get; set; }
public string cookingBakingTime { get; set; }
public string restTime { get; set; }
public int portions { get; set; }
public string pictureSource { get; set; }
public List<Ingredient> ingredients { get; set; }
public Recipe()
{
ingredients = new List<Ingredient>();
}
}
So Push Recipe-Objects to the Firebase DB works:
public async Task PushRecipe(Recipe recipe)
{
IFirebaseClient client = new FirebaseClient(config);
client.Push("Recipes", recipe);
}
Firebase Example
But when i want to get Data from the Database i get an Error..
public async Task<List<Recipe>> GetAllRecipes()
{
IFirebaseClient client = new FirebaseClient(config);
FirebaseResponse response = await client.GetAsync("Recipes");
try
{
List<Recipe> recipelist = response.ResultAs<List<Recipe>>();
return recipelist;
} catch (Exception e){
}
return null;
}
After this:
List<Recipe> recipelist = response.ResultAs<List<Recipe>>();
this Exception comes..
"{Newtonsoft.Json.JsonSerializationException: Cannot deserialize the current JSON object (e.g. {"name":"value"}) into type 'System.Collections.Generic.List`1[dignaBon_App.Models.Recipe]' because the type requires a JSON array (e.g. [1,2,3]) to deserialize correc…}"
I don´t understand why i can´t get Data from the Database back..
Can someone help me?
You need instatiate the "config" with your firebase secret address key. Go to your Firebase console for this.
Using Web API and OData, I have a service which exposes Data Transfer Objects instead of the Entity Framework entities.
I use AutoMapper to transform the EF Entities into their DTO counter parts using ProjectTo():
public class SalesOrdersController : ODataController
{
private DbContext _DbContext;
public SalesOrdersController(DbContext context)
{
_DbContext = context;
}
[EnableQuery]
public IQueryable<SalesOrderDto> Get(ODataQueryOptions<SalesOrderDto> queryOptions)
{
return _DbContext.SalesOrders.ProjectTo<SalesOrderDto>(AutoMapperConfig.Config);
}
[EnableQuery]
public IQueryable<SalesOrderDto> Get([FromODataUri] string key, ODataQueryOptions<SalesOrderDto> queryOptions)
{
return _DbContext.SalesOrders.Where(so => so.SalesOrderNumber == key)
.ProjectTo<SalesOrderDto>(AutoMapperConfig.Config);
}
}
AutoMapper (V4.2.1) is configured as follows, note the ExplicitExpansion() which prevents serialisation auto expanding navigation properties when they are not requested:
cfg.CreateMap<SalesOrderHeader, SalesOrderDto>()
.ForMember(dest => dest.SalesOrderLines, opt => opt.ExplicitExpansion());
cfg.CreateMap<SalesOrderLine, SalesOrderLineDto>()
.ForMember(dest => dest.MasterStockRecord, opt => opt.ExplicitExpansion())
.ForMember(dest => dest.SalesOrderHeader, opt => opt.ExplicitExpansion());
ExplicitExpansion() then creates a new problem where the following request throws an error:
/odatademo/SalesOrders('123456')?$expand=SalesOrderLines
The query specified in the URI is not valid. The specified type member 'SalesOrderLines' is not supported in LINQ to Entities
The navigation property SalesOrderLines is unknown to EF so this error is pretty much what I expected to happen. The question is, how do I handle this type of request?
The ProjectTo() method does have an overload that allows me to pass in an array of properties that require expansion, I found & modified the extension method ToNavigationPropertyArray to try and parse the request into a string array:
[EnableQuery]
public IQueryable<SalesOrderDto> Get([FromODataUri] string key, ODataQueryOptions<SalesOrderDto> queryOptions)
{
return _DbContext.SalesOrders.Where(so => so.SalesOrderNumber == key)
.ProjectTo<SalesOrderDto>(AutoMapperConfig.Config, null, queryOptions.ToNavigationPropertyArray());
}
public static string[] ToNavigationPropertyArray(this ODataQueryOptions source)
{
if (source == null) { return new string[]{}; }
var expandProperties = string.IsNullOrWhiteSpace(source.SelectExpand?.RawExpand) ? new List<string>().ToArray() : source.SelectExpand.RawExpand.Split(',');
for (var expandIndex = 0; expandIndex < expandProperties.Length; expandIndex++)
{
// Need to transform the odata syntax for expanding properties to something EF will understand:
// OData may pass something in this form: "SalesOrderLines($expand=MasterStockRecord)";
// But EF wants it like this: "SalesOrderLines.MasterStockRecord";
expandProperties[expandIndex] = expandProperties[expandIndex].Replace(" ", "");
expandProperties[expandIndex] = expandProperties[expandIndex].Replace("($expand=", ".");
expandProperties[expandIndex] = expandProperties[expandIndex].Replace(")", "");
}
var selectProperties = source.SelectExpand == null || string.IsNullOrWhiteSpace(source.SelectExpand.RawSelect) ? new List<string>().ToArray() : source.SelectExpand.RawSelect.Split(',');
//Now do the same for Select (incomplete)
var propertiesToExpand = expandProperties.Union(selectProperties).ToArray();
return propertiesToExpand;
}
This works for expand, so now I can handle a request like the following:
/odatademo/SalesOrders('123456')?$expand=SalesOrderLines
or a more complicated request like:
/odatademo/SalesOrders('123456')?$expand=SalesOrderLines($expand=MasterStockRecord)
However, more complicated request that try to combine $select with $expand will fail:
/odatademo/SalesOrders('123456')?$expand=SalesOrderLines($select=OrderQuantity)
Sequence contains no elements
So, the question is: am I approaching this the right way?
It feels very smelly that I would have to write something to parse and transform the ODataQueryOptions into something EF can understand.
It seems this is a rather popular topic:
odata-expand-dtos-and-entity-framework
how-to-specify-the-shape-of-results-with-webapi2-odata-with-expand
web-api-queryable-how-to-apply-automapper
how-do-i-map-an-odata-query-against-a-dto-to-another-entity
While most of these suggest using ProjectTo, none seem to address serialisation auto expanding properties, or how to handle expansion if ExplictExpansion has been configured.
Classes and Config below:
Entity Framework (V6.1.3) entities:
public class SalesOrderHeader
{
public string SalesOrderNumber { get; set; }
public string Alpha { get; set; }
public string Customer { get; set; }
public string Status { get; set; }
public virtual ICollection<SalesOrderLine> SalesOrderLines { get; set; }
}
public class SalesOrderLine
{
public string SalesOrderNumber { get; set; }
public string OrderLineNumber { get; set; }
public string Product { get; set; }
public string Description { get; set; }
public decimal OrderQuantity { get; set; }
public virtual SalesOrderHeader SalesOrderHeader { get; set; }
public virtual MasterStockRecord MasterStockRecord { get; set; }
}
public class MasterStockRecord
{
public string ProductCode { get; set; }
public string Description { get; set; }
public decimal Quantity { get; set; }
}
OData (V6.13.0) Data Transfer Objects:
public class SalesOrderDto
{
[Key]
public string SalesOrderNumber { get; set; }
public string Customer { get; set; }
public string Status { get; set; }
public virtual ICollection<SalesOrderLineDto> SalesOrderLines { get; set; }
}
public class SalesOrderLineDto
{
[Key]
[ForeignKey("SalesOrderHeader")]
public string SalesOrderNumber { get; set; }
[Key]
public string OrderLineNumber { get; set; }
public string LineType { get; set; }
public string Product { get; set; }
public string Description { get; set; }
public decimal OrderQuantity { get; set; }
public virtual SalesOrderDto SalesOrderHeader { get; set; }
public virtual StockDto MasterStockRecord { get; set; }
}
public class StockDto
{
[Key]
public string StockCode { get; set; }
public string Description { get; set; }
public decimal Quantity { get; set; }
}
OData Config:
var builder = new ODataConventionModelBuilder();
builder.EntitySet<StockDto>("Stock");
builder.EntitySet<SalesOrderDto>("SalesOrders");
builder.EntitySet<SalesOrderLineDto>("SalesOrderLines");
I have created an Automapper explicit navigation expansion utility function that should work with N-deph expands. Posting it here since it might help someone.
public List<string> ProcessExpands(IEnumerable<SelectItem> items, string parentNavPath="")
{
var expandedPropsList = new List<String>();
if (items == null) return expandedPropsList;
foreach (var selectItem in items)
{
if (selectItem is ExpandedNavigationSelectItem)
{
var expandItem = selectItem as ExpandedNavigationSelectItem;
var navProperty = expandItem.PathToNavigationProperty?.FirstSegment?.Identifier;
expandedPropsList.Add($"{parentNavPath}{navProperty}");
//go recursively to subproperties
var subExpandList = ProcessExpands(expandItem?.SelectAndExpand?.SelectedItems, $"{parentNavPath}{navProperty}.");
expandedPropsList = expandedPropsList.Concat(subExpandList).ToList();
}
}
return expandedPropsList;
}
You can call it with :
var navExp = ProcessExpands(options?.SelectExpand?.SelectExpandClause?.SelectedItems)
it will return a list with ["Parent" ,"Parent.Child"]
I never really managed to work this one out. The ToNavigationPropertyArray() extension method helps a little, but does not handle infinite depth navigation.
The real solution is to create Actions or Functions to allow clients to request data requiring a more complicated query.
The other alternative is to make multiple smaller/simple calls then aggregate the data on the client, but this isn't really ideal.
When you want to mark something for explicit expansion in AutoMapper, you need to also opt-back-in when calling ProjectTo<>().
// map
cfg.CreateMap<SalesOrderHeader, SalesOrderDto>()
.ForMember(dest => dest.SalesOrderLines, opt => opt.ExplicitExpansion());
// updated controller
[EnableQuery]
public IQueryable<SalesOrderDto> Get()
{
return _dbContext.SalesOrders
.ProjectTo<SalesOrderDto>(
AutoMapperConfig.Config,
so => so.SalesOrderLines,
// ... additional opt-ins
);
}
While the AutoMapper wiki does state this, the example is perhaps a little misleading by not including the paired ExplicitExpansion() call.
To control which members are expanded during projection, set ExplicitExpansion in the configuration and then pass in the members you want to explicitly expand:
For the life of me I can't save my poco's after an update, the insert works though.
Below is the code in question:
public class Campaign : IHasId<int>, IAudit
{
public Campaign()
{
IsRetread = true;
IsDuplicate = true;
}
[AutoIncrement]
public int Id { get; set; } //CampaignID
public long CreatedDate { get; set; }
public long ModifiedDate { get; set; }
public string ModifiedBy { get; set; }
[Required]
public string Name { get; set; } //CampaignName
[Required]
public int CampaignTypeId { get; set; } //CampaignType
[Required]
public int CampaignDeliveryTypeId { get; set; } //LeadDeliveryType
public string CampaignPhone { get; set; }
[Required]
public int LeadProviderId { get; set; } //LeadProviderID
public int PhoneTriggerId { get; set; }
[Required]
public int CampaignResponseId { get; set; } //LeadResponse
[Required]
public int CampaignCostTypeId { get; set; } //CostTypeID
public decimal CostAmount { get; set; } //CostAmount
public decimal FixedCost { get; set; } //FixedCost
public string NoteMisc { get; set; } //NoteTxt
public string NoteAgent { get; set; } //MessagetoAgents
public bool IsDefaultCampaign { get; set; } //DefaultCampaign
public bool IsExceptionCampaign { get; set; } //ExceptionCampaign
public bool IsFirmOffer { get; set; } //Firm Offer
[Reference]
public CampaignCreative CampaignCreative { get; set; }
[Reference]
public List<CampaignRule> CampaignRules { get; set; }
/* These really should be a collection of rules */
public bool IsDuplicate { get; set; } //IsDuplicate
public bool IsRetread { get; set; } //IsRetread
public bool IsFactorTrustLeads { get; set; } //IsFactorTrustLeads
public bool IsFactorTrustApp { get; set; } //IsFactorTrustApp
}
Then the save, which works:
public long SaveCampaign(Campaign campaign)
{
using (var db = _connectionFactory.OpenDbConnection())
{
var rowId = db.Insert(campaign, true);
return rowId;
}
}
and the update, which never saves the changed poco:
public long UpdateCampaign(Campaign campaign)
{
using (var db = _connectionFactory.OpenDbConnection())
{
db.Save(campaign, true);
return campaign.Id;
}
}
There aren't any errors, and the only sql I see in the immediate window are select statement, no updates (I've never seen ANY other statements beside SELECTs)
Is there another way to see why this update fails?
Thank you, Stephen
View Last SQL Executed
The easiest approach is to just print out the last SQL Statement that was executed, i.e:
db.GetLastSql().Print();
This works well for most OrmLite API's which only execute a single SQL statement. But it wont show all SQL executed with OrmLite's higher-level API's like db.Save() which can execute multiple statements.
Use Captured OrmLite SQL Filter
Another approach is to capture the generated SQL using a Capture Exec Filter which instead of executing the SQL Statements will
using (var captured = new CaptureSqlFilter())
using (var db = _connectionFactory.OpenDbConnection())
{
db.CreateTable<Campaign>();
var rowId = db.Insert(campaign, true);
var sql = string.Join(";\n", captured.SqlStatements.ToArray());
sql.Print();
}
But as this only captures and doesn't execute the SQL it wont show the full story of API's that rely on reading the database (which will just return mocked/empty results).
Profile the DB Connection
A more comprehensive approach which is used by many SQL Profilers is to use a connection profiler which is just a wrapper around a DbConnection that in addition to executing the SQL will also capture and profile it.
Here's an example of profiling OrmLite by using ServiceStack's built-in Mini Profiler:
Container.Register<IDbConnectionFactory>(c =>
new OrmLiteConnectionFactory(
connectionString, SqlServerDialect.Provider) {
ConnectionFilter = x => new ProfiledDbConnection(x, Profiler.Current)
});
Which gets OrmLite to execute its SQL using a ProfiledDbConnection instead which can be later viewed in the Mini Profiler.
I'm wondering what's the best way to handle default values for relationships when making models. (Specifically EF4)
For example, my Organization has a default Contact and I was wondering which one was the best approach. I got these two options (or any other anyone suggests if better)
Using Relationship:
public class Contact
{
public int Id { get; set; }
public string FirstName { get; set; }
}
public class Organization
{
public int Id { get; set; }
public string Name { get; set; }
public ICollection<Contact> Contacts { get; set; }
//Use a relationship for the default contact?
public Contact DefaultContact { get; set; }
}
Using Value:
public class Contact
{
public int Id { get; set; }
public string FirstName { get; set; }
//Use value?
public boolean IsDefault { get; set; }
}
public class Organization
{
public int Id { get; set; }
public string Name { get; set; }
public ICollection<Contact> Contacts { get; set; }
}
I'd go with Option 1. While 2 is definitely easier to implement, it doesn't enforce rules such as "There cannot be 2 default contacts". I end up with something like the following:
public class Organization {
// ...
public virtual ICollection<Contact> { get;set; }
[ForeignKey("DefaultContactId")]
public Contact DefaultContact { get;set; }
public int? DefaultContactId { get;set; }
}
There's a limitation of this approach - it doesn't work nested deletes (see this question for more details). Because of this, you need to disable CascadeOnDelete for the 1-to-many relationship:
protected override void OnModelCreating(DbModelBuilder modelBuilder)
{
modelBuilder.Entity<Contact>().HasRequired(co => co.Organization).WithMany().WillCascadeOnDelete(false);
}
(Code done without testing, but should work)
The other problem with this is that it's not possible to add the Default Contact at the same time as you're adding the organization, as EF can't figure out the correct order of statements. You need to call .SaveChanges between each. You can still use a TransactionScope to overcome this, but it's not clean:
using (var ts = new TransactionScope())
{
Organization org = new Organization
{
// ...
Contacts = new Collection<Contact>()
}
org.Contacts = new Contact() {};
orgRepo.SaveChanges();
// Now wire up the default contact
org.DefaultContact = org.Contacts.First();
orgRepo.SaveChanges();
}