This question already has answers here:
Merge specifications of different types in Criteria Query Specifications
(2 answers)
Closed 1 year ago.
Suppose I have two entities, one named Organization, another User. A User belongs to an Organization.
I have a this specification
Specification<Organization> isValid = ...
and I would like to look up a user by some criteria, including the company being valid. Is there an obvious way to turn a Specification<Organization> into a Specification<User>?
Something like
<X, Y> Specification<X> on(SingularAttribute<X, Y> attr, JoinType joinType, Specification<Y> spec) {
return (root, query, builder) -> {
final Join<X, Y> join = root.join(attr, joinType));
return spec.toPredicate(join, query, builder);
}
}
except of course this doesn't work, as toPredicate takes a Root<T>, not a From<?, T>.
In my case I had to entities Movie, Actor, Producers. Movie had multiple actors and one producers. To perform a search I used this:
private Specification<Movie> searchSpecification(String keyword) {
return ((root, criteriaQuery, criteriaBuilder) -> {
criteriaQuery.distinct(true);
if ((keyword ==null)) {
return null;
}
return criteriaBuilder.or(
criteriaBuilder.like(root.get("name"), "%" + keyword + "%"),
criteriaBuilder.like(root.get("plot"), "%" + keyword + "%"),
criteriaBuilder.like(root.join("actors").get("name"), "%" + keyword + "%"),
criteriaBuilder.like(root.join("producer").get("name"), "%" + keyword + "%")
);
});
}
You refer this to generate you own specification and join tables.
Related
I have been trying to pass the columns I want to select in but out of the box it appears it is not possible. I have tried things like
#Query("SELECT :columns FROM USERS u WHERE u.LOCALE = :locale AND u.id IN (:ids)")
Flux<Users> retrieveExportData(#Param("columns") String columns,
#Param("locale") String locale,
#Param("ids") String[] ids);
and with
private final R2dbcEntityTemplate template;
I tried to create my own query and but that was not working because it has to be of type Criteria and that was just creating a complexity that just was not worth it.
It would be nice if I could add the columns like
criteriaList.add(
Criteria.where("LOCALE").is(locale)
);
Criteria criteria = Criteria.from(criteriaList);
and execute it like
Flux<Users> users = this.template.select(User.class)
.matching(Query.query(criteria))
.all();
or just calling the repository like in my first example.
Has anyone been able to do this successfully?
----- update 1 -----
I tried doing like so:
import org.springframework.r2dbc.core.DatabaseClient;
DatabaseClient databaseClient = DatabaseClient.create(connectionFactory);
String sql = "SELECT " + columns + " FROM USERS u WHERE u.LOCALE=" + locale + " AND u.id IN (" + ids + ")";
return databaseClient.sql(sql)
.fetch()
.all().cast(User.class);
but Since Spring 4.3.6.RELEASE, LinkedCaseInsensitiveMap doesn't extend LinkedHashMap and HashMap, but only implements Map interface.
This results in a
Cannot cast org.springframework.util.LinkedCaseInsensitiveMap to User at java.base/java.lang.Class.cast
error.
I then tried the jooq approach suggested in the answers but it just produces syntax errors. Example
private final DSLContext ctx = DSL.using(connectionFactory);
private final Users users = ctx.newRecord(Users.USERS); <-- USERS not found
#Query("SELECT :columns FROM USERS u WHERE u.LOCALE = :locale AND u.id IN (:ids)")
public Flux<Users> retrieveExportData(
List<Field<?>> columns,
String locale,
String[] ids
) {
return Flux.from(ctx
.select(columns)
.from("USERS")
.where(users.LOCALE.eq(locale)) <--- LOCALE not found
.and(users.ID.in(ids)) <--- ID not found
).map(r -> r.into(Users.class)); <---- into not found
}
the library look promising. I will try to get it working.
You cannot replace a bind parameter (:columns) by syntactic elements like this, other than actual bind values. For this type of dynamic SQL, you'll have to resort to some sort of query building mechanism.
Perhaps look at jOOQ, which has R2DBC support? Your implementation would then look like this:
#Query("SELECT :columns FROM USERS u WHERE u.LOCALE = :locale AND u.id IN (:ids)")
public Flux<Users> retrieveExportData(
List<Field<?>> columns,
String locale,
String[] ids
) {
return Flux.from(ctx
.select(columns)
.from(USERS)
.where(USERS.LOCALE.eq(locale))
.and(USERS.ID.in(ids))
).map(r -> r.into(Users.class));
}
Disclaimer: I work for the company behind jOOQ.
you can write dynamic SQL like so.
import org.springframework.r2dbc.core.DatabaseClient;
DatabaseClient databaseClient = DatabaseClient.create(connectionFactory);
String sql = "SELECT " + columns + " FROM USERS u WHERE u.LOCALE=" + locale + " AND u.id IN (" + ids + ")";
return databaseClient.sql(sql)
.fetch()
.all().cast(User.class);
and you will need this dependency
implementation "org.springframework.boot:spring-boot-starter-data-r2dbc:2.6.2"
I have a LINQ query
var age = new int[]{1,2,3};
dbContext.TA.WHERE(x=> age.Contains( x.age)).ToList()
In an online article #11 (https://medium.com/swlh/entity-framework-common-performance-mistakes-cdb8861cf0e7) mentioned it is not a good practice as it creates many execution plan at the SQL server.
In this case, how should LINQ be revised so that I can do the same thing but minimize the amount of execution plans generated?
(note that I have no intention to convert it into a stored procedure and pass & join with the UDT as again it requires too many effort to do so)
That article offers some good things to keep in mind when writing expressions for EF. As a general rule that example is something to keep in mind, not a hard "never do this" kind of rule. It is a warning over writing queries that allow for multi-select and to avoid this when possible as it will be on the more expensive side.
In your example with something like "Ages", having a hard-coded list of values does not cause a problem because every execution uses the same list. (until the app is re-compiled with a new list, or you have code that changes the list for some reason.) Examples where it can be perfectly valid to use this is with something like Statuses where you have a status Enum. If there are a small number of valid statuses that a record can have, then declaring a common array of valid statuses to use in an Contains clause is fine:
public void DeleteEnquiry(int enquiryId)
{
var allowedStatuses = new[] { Statuses.Pending, Statuses.InProgress, Statuses.UnderReview };
var enquiry = context.Enquiries
.Where(x => x.EnquiryId == enquiryId && allowedStatuses.Contains(x.Status))
.SingleOrDefault();
try
{
if(enquiry != null)
{
enquiry.IsActive = false;
context.SaveChanges();
}
else
{
// Enquiry not found or invalid status.
}
}
catch (Exception ex) { /* handle exception */ }
}
The statuses in the list aren't going to change so the execution plan is static for that context.
The problem is where you accept something like a parameter with criteria that include a list for a Contains clause.
it is highly unlikely that someone would want to load data where a user could select ages "2, 4, and 6", but rather they would want to select something like: ">=2", or "<=6, or "2>=6" So rather than creating a method that accepts a list of acceptable ages:
public IEnumerable<Children> GetByAges(int[] ages)
{
return _dbContext.Children.Where(x => ages.Contains( x.Age)).ToList();
}
You would probably be better served with ranging the parameters:
private IEnumerable<Children> GetByAgeRange(int? minAge = null, int? maxAge = null)
{
var query = _dbContext.Children.AsQueryable();
if (minAge.HasValue)
query = query.Where(x => x.Age >= minAge.Value);
if (maxAge.HasValue)
query = query.Where(x => x.Age <= maxAge.Value);
return query.ToList();
}
private IEnumerable<Children> GetByAge(int age)
{
return _dbContext.Children.Where(x => x.Age == age).ToList();
}
I am searching for a API function corresponding to the "find in all diagrams"-function (Strg + U) in Enterprise Architect.
The class element provides the attribute diagrams which should return a collection of diagrams but it returns in my case always an empty list. Is it the wrong way?
EDIT:
I would be happy about a function that returns a collection of diagrams which include the element.
THE SOLUTION:
public List<EA.Diagram> getAllDiagramsOfElement(EA.Element element){
String xmlQueryResult = repository.SQLQuery(
"select dobj1.Diagram_ID " +
"from t_diagramobjects dobj1 " +
"where dobj1.Object_ID = " + element.ElementID+";");
XmlDocument xml = new XmlDocument();
xml.LoadXml(xmlQueryResult);
XmlNodeList xnList = xml.SelectNodes("/EADATA/Dataset_0/Data/Row");
List<EA.Diagram> result = new List<EA.Diagram>();
foreach (XmlNode xn in xnList){
result.Add(repository.GetDiagramByID(Convert.ToInt32(xn["Diagram_ID"].InnerText)));
}
return result;
}
With kind regards
MK
You might have to use a query,
Try this
select * from t_diagramobjects dobj1, t_diagramobjects dobj2 where dobj1.object_id=dobj2.object_id and dobj1.diagram_id!=dobj2.diagram_id;
In case you would like to stay with the API, you have to walk the packages in the model tree recursively, adding diagrams to a collection (ok, Dictionary object in VBScript).
Then you find all Diagramobjects from Diagrams. DiagramObjects then relate to Elements (remember, Element may be represented in more Diagrams).
Another approach could be to use Repository.SQLQuery method, which should return XML-formatted resultset (I didn't test that yet). But you'd need MSXML present on the machine to parse it (and keep up with the versions).
Generally, if you want to scan whole model and you don't need parent-child relationships, SQL should be better fit. And vice versa.
I have this same function in my Enterprise Architect Add-in Framework, implemented in the class ElementWrapper:
//returns a list of diagrams that somehow use this element.
public override HashSet<T> getUsingDiagrams<T>()
{
string sqlGetDiagrams = #"select distinct d.Diagram_ID from t_DiagramObjects d
where d.Object_ID = " + this.wrappedElement.ElementID;
List<UML.Diagrams.Diagram> allDiagrams = this.model.getDiagramsByQuery(sqlGetDiagrams).Cast<UML.Diagrams.Diagram>().ToList(); ; ;
HashSet<T> returnedDiagrams = new HashSet<T>();
foreach (UML.Diagrams.Diagram diagram in allDiagrams)
{
if (diagram is T)
{
T typedDiagram = (T)diagram;
if (!returnedDiagrams.Contains(typedDiagram))
{
returnedDiagrams.Add(typedDiagram);
}
}
}
return returnedDiagrams;
}
The function getDiagramsByQuery in the Model class looks like this
//returns a list of diagrams according to the given query.
//the given query should return a list of diagram id's
public List<Diagram> getDiagramsByQuery(string sqlGetDiagrams)
{
// get the nodes with the name "Diagram_ID"
XmlDocument xmlDiagramIDs = this.SQLQuery(sqlGetDiagrams);
XmlNodeList diagramIDNodes =
xmlDiagramIDs.SelectNodes(formatXPath("//Diagram_ID"));
List<Diagram> diagrams = new List<Diagram>();
foreach (XmlNode diagramIDNode in diagramIDNodes)
{
int diagramID;
if (int.TryParse(diagramIDNode.InnerText, out diagramID))
{
Diagram diagram = this.getDiagramByID(diagramID);
diagrams.Add(diagram);
}
}
return diagrams;
}
I am using table-valued functions with Entity Framework 5. I just received this error:
A parameter named 'EffectiveDate' already exists in the parameter collection. Parameter names must be unique in the parameter collection. Parameter name: parameter
It is being caused by me joining the calls to table-valued functions taking the same parameter.
Is this a bug/limitation with EF? Is there a workaround? Right now I am auto-generating the code (.edmx file).
It would be really nice if Microsoft would make parameter names unique, at least on a per-context basis.
I've created an issue for this here.
In the meantime, I was able to get this to work by tweaking a few functions in the .Context.tt file, so that it adds a GUID to each parameter name at runtime:
private void WriteFunctionImport(TypeMapper typeMapper, CodeStringGenerator codeStringGenerator, EdmFunction edmFunction, string modelNamespace, bool includeMergeOption) {
if (typeMapper.IsComposable(edmFunction))
{
#>
[EdmFunction("<#=edmFunction.NamespaceName#>", "<#=edmFunction.Name#>")]
<#=codeStringGenerator.ComposableFunctionMethod(edmFunction, modelNamespace)#>
{ var guid = Guid.NewGuid().ToString("N"); <#+
codeStringGenerator.WriteFunctionParameters(edmFunction, " + guid", WriteFunctionParameter);
#>
<#=codeStringGenerator.ComposableCreateQuery(edmFunction, modelNamespace)#>
} <#+
}
else
{
#>
<#=codeStringGenerator.FunctionMethod(edmFunction, modelNamespace, includeMergeOption)#>
{ <#+
codeStringGenerator.WriteFunctionParameters(edmFunction, "", WriteFunctionParameter);
#>
<#=codeStringGenerator.ExecuteFunction(edmFunction, modelNamespace, includeMergeOption)#>
} <#+
if (typeMapper.GenerateMergeOptionFunction(edmFunction, includeMergeOption))
{
WriteFunctionImport(typeMapper, codeStringGenerator, edmFunction, modelNamespace, includeMergeOption: true);
}
} }
...
public void WriteFunctionParameters(EdmFunction edmFunction, string nameSuffix, Action<string, string, string, string> writeParameter)
{
var parameters = FunctionImportParameter.Create(edmFunction.Parameters, _code, _ef);
foreach (var parameter in parameters.Where(p => p.NeedsLocalVariable))
{
var isNotNull = parameter.IsNullableOfT ? parameter.FunctionParameterName + ".HasValue" : parameter.FunctionParameterName + " != null";
var notNullInit = "new ObjectParameter(\"" + parameter.EsqlParameterName + "\"" + nameSuffix + ", " + parameter.FunctionParameterName + ")";
var nullInit = "new ObjectParameter(\"" + parameter.EsqlParameterName + "\"" + nameSuffix + ", typeof(" + parameter.RawClrTypeName + "))";
writeParameter(parameter.LocalVariableName, isNotNull, notNullInit, nullInit);
}
}
...
public string ComposableCreateQuery(EdmFunction edmFunction, string modelNamespace)
{
var parameters = _typeMapper.GetParameters(edmFunction);
return string.Format(
CultureInfo.InvariantCulture,
"return ((IObjectContextAdapter)this).ObjectContext.CreateQuery<{0}>(\"[{1}].[{2}]({3})\"{4});",
_typeMapper.GetTypeName(_typeMapper.GetReturnType(edmFunction), modelNamespace),
edmFunction.NamespaceName,
edmFunction.Name,
string.Join(", ", parameters.Select(p => "#" + p.EsqlParameterName + "\" + guid + \"").ToArray()),
_code.StringBefore(", ", string.Join(", ", parameters.Select(p => p.ExecuteParameterName).ToArray())));
}
Not a bug. Maybe a limitation or an omission. Apparently this use case has never been taken into account. EF could use auto-created parameter names, but, yeah, it just doesn't.
You'll have to resort to calling one of the functions with .AsEnumerable(). For some reason, this must be the first function in the join (as I have experienced). If you call the second function with .AsEnumerable() it is still translated to SQL and the name collision still occurs.
I'm attempting to parse SQL using the TSql100Parser provided by microsoft. Right now I'm having a little trouble using it the way it seems to be intended to be used. Also, the lack of documentation doesn't help. (example: http://msdn.microsoft.com/en-us/library/microsoft.data.schema.scriptdom.sql.tsql100parser.aspx )
When I run a simple SELECT statement through the parser it returns a collection of TSqlStatements which contains a SELECT statement.
Trouble is, the TSqlSelect statement doesn't contain attributes such as a WHERE clause, even though the clause is implemented as a class. http://msdn.microsoft.com/en-us/library/microsoft.data.schema.scriptdom.sql.whereclause.aspx
The parser does recognise the WHERE clause as such, looking at the token stream.
So, my question is, am I using the parser correctly? Right now the token stream seems to be the most useful feature of the parser...
My Test project:
public static void Main(string[] args)
{
var parser = new TSql100Parser(false);
IList<ParseError> Errors;
IScriptFragment result = parser.Parse(
new StringReader("Select col from T1 where 1 = 1 group by 1;" +
"select col2 from T2;" +
"select col1 from tbl1 where id in (select id from tbl);"),
out Errors);
var Script = result as TSqlScript;
foreach (var ts in Script.Batches)
{
Console.WriteLine("new batch");
foreach (var st in ts.Statements)
{
IterateStatement(st);
}
}
}
static void IterateStatement(TSqlStatement statement)
{
Console.WriteLine("New Statement");
if (statement is SelectStatement)
{
PrintStatement(sstmnt);
}
}
Yes, you are using the parser correctly.
As Damien_The_Unbeliever points out, within the SelectStatement there is a QueryExpression property which will be a QuerySpecification object for your third select statement (with the WHERE clause).
This represents the 'real' SELECT bit of the query (whereas the outer SelectStatement object you are looking at has just got the 'WITH' clause (for CTEs), 'FOR' clause (for XML), 'ORDER BY' and other bits)
The QuerySpecification object is the object with the FromClauses, WhereClause, GroupByClause etc.
So you can get to your WHERE Clause by using:
((QuerySpecification)((SelectStatement)statement).QueryExpression).WhereClause
which has a SearchCondition property etc. etc.
Quick glance around would indicate that it contains a QueryExpression, which could be a QuerySpecification, which does have the Where clause attached to it.
if someone lands here and wants to know how to get the whole elements of a select statement the following code explain that:
QuerySpecification spec = (QuerySpecification)(((SelectStatement)st).QueryExpression);
StringBuilder sb = new StringBuilder();
sb.AppendLine("Select Elements");
foreach (var elm in spec.SelectElements)
sb.Append(((Identifier)((Column)((SelectColumn)elm).Expression).Identifiers[0]).Value);
sb.AppendLine();
sb.AppendLine("From Elements");
foreach (var elm in spec.FromClauses)
sb.Append(((SchemaObjectTableSource)elm).SchemaObject.BaseIdentifier.Value);
sb.AppendLine();
sb.AppendLine("Where Elements");
BinaryExpression binaryexp = (BinaryExpression)spec.WhereClause.SearchCondition;
sb.Append("operator is " + binaryexp.BinaryExpressionType);
if (binaryexp.FirstExpression is Column)
sb.Append(" First exp is " + ((Identifier)((Column)binaryexp.FirstExpression).Identifiers[0]).Value);
if (binaryexp.SecondExpression is Literal)
sb.Append(" Second exp is " + ((Literal)binaryexp.SecondExpression).Value);
I had to split a SELECT statement into pieces. My goal was to COUNT how many record a query will return. My first solution was to build a sub query such as
SELECT COUNT(*) FROM (select id, name from T where cat='A' order by id) as QUERY
The problem was that in this case the order clause raises the error "The ORDER BY clause is not valid in views, inline functions, derived tables, sub-queries, and common table expressions, unless TOP or FOR XML is also specified"
So I built a parser that split a SELECT statment into fragments using the TSql100Parser class.
using Microsoft.Data.Schema.ScriptDom.Sql;
using Microsoft.Data.Schema.ScriptDom;
using System.IO;
...
public class SelectParser
{
public string Parse(string sqlSelect, out string fields, out string from, out string groupby, out string where, out string having, out string orderby)
{
TSql100Parser parser = new TSql100Parser(false);
TextReader rd = new StringReader(sqlSelect);
IList<ParseError> errors;
var fragments = parser.Parse(rd, out errors);
fields = string.Empty;
from = string.Empty;
groupby = string.Empty;
where = string.Empty;
orderby = string.Empty;
having = string.Empty;
if (errors.Count > 0)
{
var retMessage = string.Empty;
foreach (var error in errors)
{
retMessage += error.Identifier + " - " + error.Message + " - position: " + error.Offset + "; ";
}
return retMessage;
}
try
{
// Extract the query assuming it is a SelectStatement
var query = ((fragments as TSqlScript).Batches[0].Statements[0] as SelectStatement).QueryExpression;
// Constructs the From clause with the optional joins
from = (query as QuerySpecification).FromClauses[0].GetString();
// Extract the where clause
where = (query as QuerySpecification).WhereClause.GetString();
// Get the field list
var fieldList = new List<string>();
foreach (var f in (query as QuerySpecification).SelectElements)
fieldList.Add((f as SelectColumn).GetString());
fields = string.Join(", ", fieldList.ToArray());
// Get The group by clause
groupby = (query as QuerySpecification).GroupByClause.GetString();
// Get the having clause of the query
having = (query as QuerySpecification).HavingClause.GetString();
// Get the order by clause
orderby = ((fragments as TSqlScript).Batches[0].Statements[0] as SelectStatement).OrderByClause.GetString();
}
catch (Exception ex)
{
return ex.ToString();
}
return string.Empty;
}
}
public static class Extension
{
/// <summary>
/// Get a string representing the SQL source fragment
/// </summary>
/// <param name="statement">The SQL Statement to get the string from, can be any derived class</param>
/// <returns>The SQL that represents the object</returns>
public static string GetString(this TSqlFragment statement)
{
string s = string.Empty;
if (statement == null) return string.Empty;
for (int i = statement.FirstTokenIndex; i <= statement.LastTokenIndex; i++)
{
s += statement.ScriptTokenStream[i].Text;
}
return s;
}
}
And to use this class simply:
string fields, from, groupby, where, having, orderby;
SelectParser selectParser = new SelectParser();
var retMessage = selectParser.Parse("SELECT * FROM T where cat='A' Order by Id desc",
out fields, out from, out groupby, out where, out having, out orderby);