Hiperspace.SQL 2.5.28

Prefix Reserved
dotnet add package Hiperspace.SQL --version 2.5.28
                    
NuGet\Install-Package Hiperspace.SQL -Version 2.5.28
                    
This command is intended to be used within the Package Manager Console in Visual Studio, as it uses the NuGet module's version of Install-Package.
<PackageReference Include="Hiperspace.SQL" Version="2.5.28" />
                    
For projects that support PackageReference, copy this XML node into the project file to reference the package.
<PackageVersion Include="Hiperspace.SQL" Version="2.5.28" />
                    
Directory.Packages.props
<PackageReference Include="Hiperspace.SQL" />
                    
Project file
For projects that support Central Package Management (CPM), copy this XML node into the solution Directory.Packages.props file to version the package.
paket add Hiperspace.SQL --version 2.5.28
                    
#r "nuget: Hiperspace.SQL, 2.5.28"
                    
#r directive can be used in F# Interactive and Polyglot Notebooks. Copy this into the interactive tool or source code of the script to reference the package.
#:package Hiperspace.SQL@2.5.28
                    
#:package directive can be used in C# file-based apps starting in .NET 10 preview 4. Copy this into a .cs file before any lines of code to reference the package.
#addin nuget:?package=Hiperspace.SQL&version=2.5.28
                    
Install as a Cake Addin
#tool nuget:?package=Hiperspace.SQL&version=2.5.28
                    
Install as a Cake Tool

Hiperspace.SQL

Hiperspace.SQL is a full SQL query engine for Hiperspace, supporting the full range of joins, aggregations, and subqueries.

Hiperspace.SQL provides the same query functionality as a .NET client can use with LINQ queries, but without the need to write code in C#/F#

Hiperspace fully supports point-in-time "time travel" queries that are not possible with Python Data-Frames or DuckDB

Features

  • Hiperspace.SQL is not limited to queries of columns within a table, but supports the full navigation of properties of Hiperspace elements
  • Where a column is a complex object, it is returned as a JSON object
  • Executing a batch of SQL statements return columnar data frames (dictionary of column-name and array of values)
  • Explain SQL returns the execution plan, detailing the SetSpaces accessed and keys used for search (Key, Index, Scan)
  • The Parquet method returns a Parquet file that can be used with any Apache Parquet library, or added to DuckDB OLAP store

Data Dictionary

SCHEMA_TABLE

Column Name Data Type Description
TABLE_NAME string The name of the table
TABLE_TYPE string The type of the table in SCHEMA_PROPERTY

SCHEMA_COLUMN

Column Name Data Type Description
TABLE_NAME string The name of the table
COLUMN_NAME string The name of the column
COLUMN_TYPE string The type of the table in SCHEMA_PROPERTY

SCHEMA_PROPERTY

Column Name Data Type Description
TYPE_NAME string The Type Name
PROPERTY_NAME string The name of each property
PROPERTY_TYPE string reference to SCHEMA_PROPERTY.TYPE_NAME

Examples

Simple query

SELECT p.Name, p.Gender FROM Persons as p WHERE p.Name = 'Lucy'

Query parameters

SELECT p.Name, p.Gender FROM Persons as p WHERE p.Name = :name

Query batches

SELECT p.Name, p.Gender FROM Persons as p WHERE p.Name = :name;
SELECT Name as name, Father as father from   Persons ;

Joins

SELECT  p.Name, f.Name as Father, f.Father as GrandFather
FROM    Persons as p 
join    Persons as f on p.Father.Name = f.Name
WHERE   p.Name = :name

Aggregates

select  p.Father.Name, count(p.Name) as Children
from    Persons as p
group by p.Father.Name as f
having count(*) > 1;

Like expressions

select  p.Father.Name, count(p.Name) as Children
from    Persons as p
where   Name like 'L%' and Name like '%y' or (Name like '%u%' and Name like '_uc_')
group by p.Father.Name as f
having count(*) > 1;

Null handling

select  p.Name, p.Father.Name
from    Persons as p
where   Name is not null

in query

SELECT p.Name, p.Gender 
FROM Persons as p 
WHERE p.Gender in (select p2.Gender from Persons as p2 where p2.Name = 'Lucy')

union

SELECT p.Name, p.Gender 
FROM Persons as p 
WHERE p.Name in ('Lucy', 'Mark')
union 
SELECT p.Name, p.Gender 
FROM Persons as p 
WHERE p.Name in ('Eve', 'Mary')

inline view

SELECT p.Name, p.Gender 
FROM Persons as p 
join (select p2.Gender from Persons as p2 where p2.Name = 'Lucy') as p3 on p.Gender = p3.Gender

dictionary query

select * from SCHEMA_TABLES;
select * from SCHEMA_COLUMNS;
select * from SCHEMA_PROPERTIES;

Hierarchy query

select  p.Name as Parent, c.Name as Child
from    Persons as p, p.MotherChild as c
where   p.Name = :name;

is equivalent to

select  p.Name as Parent, c.Name as Child
from    Persons as p JOIN Persons as c ON p.Name = c.Mother.Name
where   p.Name = :name;

since MotherChild is a set (of person) with each Person element, and any set can be joined with another set. In this case there is an implicit join between the set MotherChild and the Person p.

Graph

The Cousins example includes the property AllRelatives = allrelation(this) which uses the Cousins.Helper.AllRelations function to return the set of HiperEdge for all graph nodes that can be recursively found in Hiperspace for each person.

The following query finds all the relations for a person

select  p.Name as Person, 
        r.To.Name as Relation, 
        r.TypeName as Relationship,
        r.Length as Length,
        r.Width as Width
from    Persons as p,
        p.AllRelatives as r
where   p.Name = :name;

NB : There is no need to use a JOIN clause as p.AllRelatives includes the implicit join to Person.

the ability to navigate graph relationships is a unique feature of Hiperspace.SQL

API

The Hiperspace.SQL API can be called from any language that supports DOTNET interop, including Python (using pythonnet). Access via the Hiperspace.SQL.Engine object that is constructed with reference to any domain space.

Explain

Provides a detailed breakdown of the query execution plan

member engine.Explain (source, parameters : IDictionary<string,obj>) : string array = 

Execute

Executes the SQL queries and returns an array of Data Frames

member engine.Execute (source , parameters : IDictionary<string,obj>) : IDictionary<string, obj array> array  =

Parquet

Executes the SQL queries will an array of filenames (one for each statement) and returns the filenames after writing the results to the Apache Parquet files.

member this.Parquet (source, fileNames, parameters : IDictionary<string,obj>): string array = 
Product Compatible and additional computed target framework versions.
.NET net8.0 is compatible.  net8.0-android was computed.  net8.0-browser was computed.  net8.0-ios was computed.  net8.0-maccatalyst was computed.  net8.0-macos was computed.  net8.0-tvos was computed.  net8.0-windows was computed.  net9.0 is compatible.  net9.0-android was computed.  net9.0-browser was computed.  net9.0-ios was computed.  net9.0-maccatalyst was computed.  net9.0-macos was computed.  net9.0-tvos was computed.  net9.0-windows was computed.  net10.0 is compatible.  net10.0-android was computed.  net10.0-browser was computed.  net10.0-ios was computed.  net10.0-maccatalyst was computed.  net10.0-macos was computed.  net10.0-tvos was computed.  net10.0-windows was computed. 
Compatible target framework(s)
Included target framework(s) (in package)
Learn more about Target Frameworks and .NET Standard.

NuGet packages

This package is not used by any NuGet packages.

GitHub repositories

This package is not used by any popular GitHub repositories.

Version Downloads Last Updated
2.5.28 51 12/31/2025
2.5.26 164 12/21/2025
2.5.21 411 12/10/2025
2.5.18 656 12/3/2025
2.5.8 165 11/15/2025
2.5.2 193 11/6/2025
2.5.0 179 10/20/2025
2.4.6 187 9/23/2025
2.4.4 266 8/7/2025
2.4.2 154 7/28/2025
2.4.0 184 7/10/2025
2.3.8 175 7/1/2025
2.3.4 177 6/5/2025
2.2.2 194 5/5/2025
2.2.1 245 4/14/2025
2.2.0 128 3/29/2025
2.1.9 257 3/5/2025
2.1.6 157 2/15/2025
2.0.0 129 1/14/2025
1.0.5 146 11/15/2024
1.0.2 146 11/1/2024
1.0.1 191 10/18/2024

https://www.cepheis.com/hiperspace/20251231
# Overview
Change to the `GetSequence` and `UseSequence` functions to take a parameter to allow the use of `KeyType` for Sequences

Extended  `Session2(DateTime? AsAt = null)`  to `Session2(DateTime? AsAt = null, DateTime? DeltaFrom = null)` to allow the DeltaFrom parameter to be used in addition to AsAt.
## Obsolete
Several interfaces function have been marked as `Obsolete` and will be removed in the next version.

Currently **Hiperspace** uses a key prefix convention to distinguish *immutable* and *versioned* elements, *VectorSpaces* (for nearest neighbor search) and *sequences*

| Prefix | Description |
|-|-|
| 00| The meta data dictionary for the *Hiperspace*, used to ensure that incompatible libraries cannot corrupt a store |
| 0...| Versioned element that includes a *timestamp* suffix at the end of the key |
| 00...| VectorSpace indexes|
| 001...|  Sequence numbers for mutable items|
| |Keys without a prefix are immutable elements without version|

Performance testing of **Hiperspace.DB**  has highlighted that the rebuilding of the `MetaMap` used to translate between protobuf *tag/length/value* to *tag/value/length* for index search can be expensive top build when a store has experienced multiple schema changes.  The **next** version will change the prefix to:

| Prefix | Description |
|-|-|
|1| The meta data dictionary for the *Hiperspace*, used to ensure that incompatible libraries cannot corrupt a store |
|2| cached `MetaMap` entries for key encode/decode|
|3| validated client library *fingerprint* to avoid the need for `MetaMap` build |
|4| immutable elements without version suffix|
|5| Versioned element that includes a *timestamp* suffix at the end of the key |
|6| VectorSpace indexes|
|7|  Sequence numbers for mutable items|

The `#id` for {`Node`, `Edge`, `VectorNode`, `HiperEdge`,`PathMessage`} will also be moved to a *reserved* set of low values with all *domain* keys moved to higher values

-----

## SetSpace
Added two functions to get a sequence counter from the driver:
* `GetSequenceAsync<T>(T element)` to get the current sequence number associated with the domain key
* `UseSequenceAsync<T>(T element)` to get a new  sequence number for the with the domain key

**NB** While the main use of Sequence numbers is to provide an alternate (*GPGPU friendly*) sequence number for a `Node`, it can be applied to any key value.
```
// get an order number that is unique within all orders
var orderNo = await space.Orders.UseSequenceAsync (new Order { });
var orderNo = await space.Orders.UseSequenceAsync (new Order.KeyType { });

// get a line number within the owning Order {1,2,3..}
var lineNo = await space.Lines.UseSequenceAsync (new Line { Order = order });
var lineNo = await space.Lines.UseSequenceAsync (new Line.KeyType { Order = order.self });
```

`Get` and `GetASync()` functions now handle not found errors by returning `null` for scenarios where a reference object cannot be found.  This is especially useful for views where not found is better handled by `Horizon<>` filters that

-----

## Bind Version
All versions of Bind Function have been marked as Obsolete, with a preference to use the optimistic locking instead (that include the original AsAt copied to AsWas for comparison).
Versions of Bind functions that include the `priorVersion` DateTime value use optimistic concurrency control to prevent updates to a version that has been changed by another session

## Nearest
Hiperspace provides nearest neighbor search using a VectorSpace `aspect` for geospatial and RAG search.  These functions have been extended to include a maximum distance search

-----

## Hilang
The Hilang ahead of time compiler has been updated to remove indexes that are a subset of the element key, and will not be used


-----

## Hiperspace.Rocks
The **RocksDB** driver has been upgraded to verion 10.4.2.63147