An ontology triple (quad) store for RDF/OWL using Entity Framework 4

Ian — Wed, 12 May 2010 22:34:21 +0000

This weeks side-project was the creation of an ontology store using Entity Framework 4. An ontology store stores axioms consisting of Subject, Predicate, Object which are usually serialized as RDF, OWL, N3, … Whereas there’s lots of details about these serialization formats, the actual mechanics of how to store and manipulate them was somewhat harder to come by. Nevertheless, after much experimentation I came up with an Entity Model that can store Quads (Subject, Predicate, Object and Meta) or Quins (Subject, Predicate, Object, Meta, Graph). The addition of Meta allows one Axiom to reference another. The addition of Graph allows the store to be segmented making it easy to import some N3 or RDF into a graph, then flush that graph if it is no longer needed or if a newer version becomes available.

The store is currently hooked up to an Euler reasoner that can reason against it, lazily fetching just the necessary records from the SQL database that backs the Entity Model.

Here’s the EDMX showing how I modeled the Ontology Store:

SQL Server – error: 18456, severity: 14, state: 38 – Incorrect Login

Ian — Mon, 16 Nov 2009 23:57:00 +0000

Despite the error message this problem can be caused by something other than an Authorization failure. In fact, simply misspelling the Initial Catalog can cause this message to appear. I wish developers wouldn’t reuse error messages when the problem and solution is completely different.

Generate a SQL Compact Database from your Entity Model (EDMX)

Ian — Mon, 18 May 2009 16:55:00 +0000

I recently switched from SQL Lite to SQL Compact for some of the databases in a system I’m building. If you are using SQL Compact make sure you have the Hotfix from Microsoft that fixes the Where() clause. The release that shipped as SP1 is utterly useless – can’t even do a where clause on an nvarchar()! I don’t know why they didn’t recall it and replace it with the hotfix version!

Unfortunately SQL Server Management Studio doesn’t support generating scripts from a SQL Compact database, so if you design your database there you are out of luck when it comes to creating the scripts to auto-generate it when you deploy your code.

Steve Lasker’s Blog has a great article explaing how to embed these scripts as resources but it still doesn’t solve the issue of how to generate the script in the first place.

There are a few articles that explain how to use T4 to generate stored procedures from the EDMX but I didn’t see one to generate the actual database itself. So I built one and you can get it below.

To use this simply create a text file in your project called generate.tt, paste this text in. It will run and generate another file below itself containing the script to create each table it finds in each EDMX file in your project.

Limitations: Within the EDMX file there isn’t enough information to fully create a database. While the code below can generate foreign keys and primary keys from Associations and Keys defined in your model it has no idea what indexes you might want. So for now it just generates an index for every field that looks vaguely indexable and it leaves it up to you to delete the ones you don’t want.

// Generate SQL Server CE database tables from EDMX definition – suitable for some simple databases only

// Creates tables, adds primary keys, adds an index for every conceivable field, adds foreign key relationships for all associations

<#@ template language=”C#v3.5″ debug=”True” #>

<#@ output extension=”.sql” #>

<#@ assembly name=”System.Core” #>

<#@ assembly name=”System.Data” #>

<#@ assembly name=”System.Data.Entity” #>

<#@ assembly name=”System.Xml” #>

<#@ assembly name=”System.Xml.Linq” #>

<#@ import namespace=”System.Collections” #>

<#@ import namespace=”System.Collections.Generic” #>

<#@ import namespace=”System.Data.EntityClient” #>

<#@ import namespace=”System.Data.SqlClient” #>

<#@ import namespace=”System.Diagnostics” #>

<#@ import namespace=”System.IO” #>

<#@ import namespace=”System.Linq” #>

<#@ import namespace=”System.Text” #>

<#@ import namespace=”System.Text.RegularExpressions” #>

<#@ import namespace=”System.Xml.Linq” #>

// This will generate the code to build SQL COMPACT tables for every EDMX in the current directory

// Simply plug that code into SQL Server Management Studio or use it as a resource to build your database

// Get the current directory from the stack trace

string stackTraceFileName = new StackTrace(true).GetFrame(0).GetFileName();

string directoryName = Path.GetDirectoryName(stackTraceFileName);

——————————————————————————–

–

– This code was generated by a tool.

–

– Changes to this file may cause incorrect behavior and will be lost if

– the code is regenerated.

–

——————————————————————————–

string[] entityFrameworkFiles = Directory.GetFiles(directoryName, “*.edmx”);

foreach (string fileName in entityFrameworkFiles)

{

– Creating TSQL creation script <#= fileName #>.

DoExtractTables(fileName);

}

<#+

XNamespace edmxns = “http://schemas.microsoft.com/ado/2007/06/edmx”;

XNamespace edmns = “http://schemas.microsoft.com/ado/2006/04/edm”;

XNamespace ssdlns = “http://schemas.microsoft.com/ado/2006/04/edm/ssdl”;

private void DoExtractTables(string edmxFilePath)

{

XDocument edmxDoc = XDocument.Load(edmxFilePath);

XElement edmxElement = edmxDoc.Element(edmxns + “Edmx”); // element

XElement runtimeElement = edmxElement.Element(edmxns + “Runtime”); // element

XElement storageModelsElement = runtimeElement.Element(edmxns + “StorageModels”); // element

XElement ssdlSchemaElement = storageModelsElement.Element(ssdlns + “Schema”);

string entityContainerName = runtimeElement

.Element(edmxns + “ConceptualModels”) // element

.Element(edmns + “Schema”) // element

.Element(edmns + “EntityContainer”) // element

.Attribute(“Name”).Value;

string ssdlNamespace = ssdlSchemaElement.Attribute(“Namespace”).Value;

// Get a list of tables from the SSDL.

XElement entityContainerElement = ssdlSchemaElement.Element(ssdlns + “EntityContainer”);

IEnumerable entitySets = entityContainerElement.Elements(ssdlns + “EntitySet”);

IEnumerable entityTypes = ssdlSchemaElement.Elements(ssdlns + “EntityType”);

IEnumerable associations = ssdlSchemaElement.Elements(ssdlns + “Association”);

string defaultSchema = entityContainerElement.Attribute(“Name”).Value;

foreach (XElement table in entitySets)

{

string tableName = table.Attribute(“Name”).Value;

DROP Table <#= tableName #>

CREATE TABLE <#= tableName #> (

<#+

XElement entityType = entityTypes.First(et => et.Attribute(“Name”).Value == tableName);

IEnumerable properties = entityType.Elements(ssdlns + “Property”);

int i = 0;

int count = properties.Count();

// GET THE PRIMARY KEY INFORMATION

XElement key = entityType.Element(ssdlns + “Key”);

string pkstr = “”;

if (key != null)

{

pkstr = string.Join(“,”, key.Elements(ssdlns + “PropertyRef”).Select(pk => pk.Attribute(“Name”).Value).ToArray());

count++; // To add a comma on last field definition

}

// // Accumulate a list of indexes to add as we go along examining the properties

// List indexesToAdd = new List();

foreach (XElement property in properties)

{

bool last = (i == count-1);

string propertyName = property.Attribute(“Name”).Value;

string propertyType = property.Attribute(“Type”).Value;

string nullField = “”;

if (property.Attribute(“Nullable”) != null && property.Attribute(“Nullable”).Value == “false”)

nullField = ” NOT NULL”;

string maxLength = “”;

if (property.Attribute(“MaxLength”) != null)

maxLength = “(” + property.Attribute(“MaxLength”).Value + “)”;

string comma = last ? “” : “,”;

<#= propertyName #> <#= propertyType #><#= maxLength #><#= nullField #><#= comma #>

<#+

i++;

} // each property

// Was their a primary key? If so, add it

if (pkstr != “”)

{

CONSTRAINT <#=”PK_”+tableName #> PRIMARY KEY ( <#= pkstr #> )

<#+

}

)

<#+

// Now add all the indexes we might need … ———————————————————————–

foreach (XElement property in properties)

{

string comment = “”;

string propertyName = property.Attribute(“Name”).Value;

if (pkstr.Contains(propertyName))

{

comment = “already in primary key, no need to index again”;

}

int maxLength = -1;

if (property.Attribute(“MaxLength”) != null)

int.TryParse(property.Attribute(“MaxLength”).Value, out maxLength);

if (maxLength > 256)

{

comment = “too long to meaningfully index”;

}

string commentPrefix = comment != “” ? “– ” : “”;

// TODO: More rules here about what to index and whether it’s unique or not …

// CREATE [UNIQUE] INDEX …

if (comment != “”)

{

<#=commentPrefix #><#=comment #>

<#+

}

<#=commentPrefix #>CREATE INDEX IX_<#=tableName #>_<#=propertyName #> ON <#=tableName #> (<#=propertyName #>);

<#+

} // each property

} // Each table

// Now add all the associations between tables … ——————————————————————-

foreach (XElement association in associations)

{

string associationName = association.Attribute(“Name”).Value;

XElement referentialConstraint = association.Elements(ssdlns + “ReferentialConstraint”).First();

XElement principalRole = referentialConstraint.Element(ssdlns + “Principal”);

string thisTable = principalRole.Attribute(“Role”).Value;

string thisField = principalRole.Element(ssdlns + “PropertyRef”).Attribute(“Name”).Value;

XElement dependentRole = referentialConstraint.Element(ssdlns + “Dependent”);

string otherTable = dependentRole.Attribute(“Role”).Value;

string otherField = dependentRole.Element(ssdlns + “PropertyRef”).Attribute(“Name”).Value;

ALTER TABLE <#= otherTable #>

ADD CONSTRAINT <#= associationName #>

FOREIGN KEY (<#= otherField #>)

REFERENCES <#= thisTable #> (<#= thisField #>)

ON UPDATE CASCADE

ON DELETE CASCADE

<#+

} // Each association

} // Class

Ian Mercer» SQL

An ontology triple (quad) store for RDF/OWL using Entity Framework 4

SQL Server – error: 18456, severity: 14, state: 38 – Incorrect Login

Generate a SQL Compact Database from your Entity Model (EDMX)