CodoMetis.ValueRanges.EFCore.PostgreSQL.NodaTime 8.0.0

CodoMetis.ValueRanges.EFCore.PostgreSQL.NodaTime

NodaTime support for the CodoMetis.ValueRanges EF Core plugin: maps the range and value-set types of CodoMetis.ValueRanges.NodaTime to PostgreSQL range, multirange and array columns, bridging through NpgsqlRange<T> and native arrays via Npgsql.EntityFrameworkCore.PostgreSQL.NodaTime.

Ranges and multiranges

Value sets (v6)

Usage

dotnet add package CodoMetis.ValueRanges.EFCore.PostgreSQL.NodaTime

options.UseNpgsql(connectionString, npgsql => npgsql.UseValueRangesNodaTime());

UseValueRangesNodaTime() implies both UseNodaTime() (the Npgsql NodaTime plugin, which maps the element types LocalDate, LocalDateTime and Instant) and UseValueRanges() (the base plugin) — neither needs to be called separately, and the BCL-based range and set types keep working in the same model. It registers the range types and the five value set types; nothing further is configured per property.

Plain LocalDate[]/List<Instant> properties keep their native Npgsql array mapping and coexist in the same model, and scaffolding still produces plain arrays — opting into a set type is a model decision.

The full range algebra translates from LINQ to SQL exactly as documented for the base package — operators (@>, &&, <<, -|-, …), bound accessors (lower, upper, lower_inc, upper_inc), Merge (range_merge), the RangeAgg/RangeIntersectAgg aggregates, Intersect/Union/Except, multirange operations including == equality, and the CreateFinite/CreateUnboundedStart/CreateUnboundedEnd factories as guarded range constructor calls:

var day = new LocalDate(2024, 6, 15);

reservations.Where(r => r.Period.Contains(day));            // r."Period" @> DATE '2024-06-15'
reservations.OrderBy(r => r.Period.LowerBound());           // ORDER BY lower(r."Period")
reservations.Where(r => r.Window.Overlaps(other));          // r."Window" && @other
reservations.GroupBy(r => r.CustomerId)
            .Select(g => g.Select(r => r.Period).RangeAgg()); // range_agg(r."Period")

The value-set algebra translates to the array operators — @>, &&, <@, cardinality, array_remove, array_cat:

reservations.Where(r => r.Holidays.Contains(day));           // r."Holidays" @> ARRAY['2024-06-15']::date[]
reservations.Where(r => r.Holidays.Overlaps(closures));      // r."Holidays" && @closures
reservations.Where(r => r.Holidays.IsSubsetOf(allowed));     // r."Holidays" <@ @allowed
reservations.Where(r => r.BillingMonths.Count > 2);          // cardinality(r."BillingMonths") > 2
reservations.Where(r => r.Holidays.Remove(day).IsEmpty);     // cardinality(array_remove(…)) = 0

Contains always translates as @> rather than = ANY, so a plain GIN index serves it. Intersect, Except and Add are client-side only — PostgreSQL's array type has no intersection, difference or sorted insert — and fail query translation by design.

Notes

• No normalization rules. The base package documents DateTimeKind reinterpretation for tsrange and UTC offset normalization for tstzrange. Neither applies here: LocalDateTime is wall-clock time by construction and Instant is an instant by construction — the value written is the value stored.
• Discrete canonicalization is compensated exactly as for DateRange: LocalDateRange.UpperBound() translates to upper(x) - 1, so server results always equal the in-memory results (verified against live PostgreSQL).
• Calendar normalization reaches bare operands. LocalDateSet/LocalDateTimeSet hold ISO elements, so a probe in another calendar is normalized before it is bound — Holidays.Contains(copticDate) queries the ISO date it denotes, not that calendar's year/month/day read as if they were ISO. YearMonthSet and YearMonthRange reject a non-ISO operand outright, since a non-ISO year-month has no lossless ISO equivalent. The same rule applies in memory (see the core NodaTime README).
• YearMonthSet (v6) persists first-of-month dates in a date[], exactly as YearMonthRange does for its bounds; reads validate alignment and throw on a partial-month date rather than silently shifting it to its month.
• Union is the one set operation whose SQL result is not canonical — it translates to array_cat, which concatenates. Harmless inside the operators above (all duplicate-insensitive) and on materialization (reads re-canonicalize), but Count over a union is refused rather than counting duplicates. Remove has no such caveat: array_remove leaves the array sorted and deduplicated.
• YearMonthRange (v5) is stored as a month-aligned daterange: [2025-01, 2025-03] becomes [2025-01-01, 2025-04-01), so no custom database type is needed and every operator works server-side. Bound elements convert through first-of-month dates (Contains(yearMonth) → @> DATE 'yyyy-MM-01'); UpperBound()'s upper(x) - 1 lands on the last day of the end month, which reads back as that month. Reads validate month alignment — a partial-month daterange throws instead of silently shifting. The factories cannot be constructed in SQL from column values (months are coarser than the date subtype); constant and parameter ranges work as usual.
• Precision: NodaTime carries nanoseconds, PostgreSQL stores microseconds — sub-microsecond precision is reduced at the database boundary.
• Instant.MinValue/MaxValue map to PostgreSQL -infinity/infinity by default (Npgsql rule): a finite bound that happens to be infinite, distinct from an unbounded side (upper_inf stays false).
• External data sources: when the application builds its own NpgsqlDataSource instead of letting EF create one, call UseNodaTime() on that data source builder as well — the same requirement Npgsql.EntityFrameworkCore.PostgreSQL.NodaTime documents.
• Store-name lookups: with the Npgsql NodaTime plugin active, resolving a mapping by store type name alone (e.g. scaffolding-style FindMapping("daterange")) is answered by Npgsql's own plugin (DateInterval). The range types always resolve by CLR type, which is unambiguous.
• Reverse engineering (dotnet ef dbcontext scaffold) does not produce these types — apply them manually after scaffolding, as with the base package.

License

MIT — see LICENSE.