Class UniqueRecomputeReport

Namespace: redb.Core.Models
Assembly: redb.Core.dll

What a recomputation of _values._unique for one key structure found. Produced when a
[RedbUnique] attribute appears on a populated structure, when the structure's encoder
version is behind UniqueKeyEncoder.Version, when rows are found with a value but no key
(a SQL-side writer touched them), and on an explicit RecomputeUniqueAsync.

Duplicates are reported, not thrown: the first row of each duplicate group (lowest value id)
keeps the key, the others stay with _unique IS NULL — outside the index, visible here,
and to a query WHERE _id_structure = … AND _unique IS NULL AND <column> IS NOT NULL.
Uniqueness is enforced for every row that has a key from this point on; the losers are the
application's to resolve.

public sealed class UniqueRecomputeReport

Inheritance

Object→UniqueRecomputeReport

Properties

Duplicates

Groups of rows sharing one key; each group's first id kept the key.

public IReadOnlyList<UniqueDuplicateGroup> Duplicates { get; init; }

HasDuplicates

public bool HasDuplicates { get; }

HashedRows

Rows that received a key (rows with a NULL value get none and are not counted).

public int HashedRows { get; init; }

LosingRows

Rows left without a key because another row already holds it.

public int LosingRows { get; }

PropertyName

public string PropertyName { get; init; }

SchemeId

public long SchemeId { get; init; }

SchemeName

public string SchemeName { get; init; }

StructureId

public long StructureId { get; init; }

TotalRows

Root scalar rows of the structure that were examined.

public int TotalRows { get; init; }