Files

200 lines
12 KiB
Markdown

# CLAUDE.md - DeepDrftData
Guidance for working in the DeepDrftData project (the SQL-side domain logic).
See the root `CLAUDE.md` for full architecture overview. This file covers what is specific to this project.
## One-line purpose
SQL-side domain logic for tracks. EF Core context, configurations, migrations, repository, manager, design-time factory. Consumed by `DeepDrftAPI` (the dual-database authority) and tests.
## Why this project exists
Separating domain logic from hosts so DeepDrftAPI can reuse `TrackManager` / `TrackRepository` / `DeepDrftContext` without embedding them directly in the host. Tests also consume this library.
**New SQL-side domain code goes here, not in the host projects.**
## Layout
```
DeepDrftData/
├── Data/
│ ├── DeepDrftContext.cs # EF DbContext
│ ├── DeepDrftContextFactory.cs # Design-time factory (reads environment/connections.json; Npgsql dummy fallback)
│ └── Configurations/
│ └── TrackConfiguration.cs # EF fluent configuration for TrackEntity
├── Migrations/ # EF-generated migrations (namespace DeepDrftData.Migrations)
├── Repositories/
│ └── TrackRepository.cs # Data access layer
├── TrackManager.cs # Service orchestrator (public interface: ITrackService)
└── DeepDrftData.csproj
```
## EF DbContext and configuration
`DeepDrftContext` targets **PostgreSQL** (Npgsql), connection string from `environment/connections.json` (loaded at runtime via `CredentialTools.ResolvePathOrThrow("connections", ...)` in `DeepDrftAPI/Program.cs`, key `ConnectionStrings:DefaultConnection`). The design-time factory (`DeepDrftContextFactory`) reads the same `environment/connections.json` when present and falls back to a Npgsql dummy connection string (`Host=localhost;Database=deepdrft-design-time;Username=dummy`) for CI or environments without the file, so `dotnet ef` commands work without a live database.
`TrackConfiguration` uses EF fluent API:
- Table name: `track` (singular)
- Columns: snake_case (`entry_key`, `track_name`, `artist`, `album`, `genre`, `release_date`, `image_path`)
- `EntryKey`: required, max 100 (the FileDatabase vault entry id)
- `TrackName`, `Artist`: required, max 200
- `Album`, `Genre`: optional, max 200 / 100
- `ReleaseDate`: optional `DateOnly`
- `ImagePath`: optional, max 500 (currently a free-form URL string; points to images vault in future)
- `DurationSeconds`: optional `double?` (nullable; populated at upload from vault audio; backfillable via `POST api/track/duration/backfill`; used for aggregate mix-runtime queries). Column: `duration_seconds`. Migration: `20260618155002_AddTrackDuration`.
## Service → Repository → DbContext shape
- **Service** (`TrackManager`, implements `ITrackService`): Public contract. Takes `TrackRepository`, catches exceptions at service boundary, returns `ResultContainer<T>` with DTO results.
- **Repository** (`TrackRepository`): Internal data access. Queries the DbContext. Throws on error (service catches).
- **DbContext** (`DeepDrftContext`): EF Core. Directly accessed by repository, never by service (pattern isolation).
Notable repository / service methods beyond the standard CRUD:
- `TrackRepository.GetHomeStatsAsync` / `ITrackService.GetHomeStats`: Returns `HomeStatsDto` — cut track count, per-`ReleaseType` cut release counts (zero-suppressed), mix release count, total mix runtime seconds (null durations counted as 0; tracks under a soft-deleted release excluded). Used by `StatsController`.
- `TrackRepository.UpdateDurationAsync` / `ITrackService.UpdateDuration`: Null-guarded duration write — skips rows where `DurationSeconds` is already set. Used by the one-time backfill (`POST api/track/duration/backfill`).
- `TrackRepository.SetDurationAsync` / `ITrackService.SetDuration`: Unconditional duration overwrite — no null guard, always stamps the new value. Used by the replace-audio path (`POST api/track/{id:long}/replace-audio`) where the existing non-null duration must be overwritten with the new audio's value. Returns a fail result when zero rows are affected (track removed between lookup and write).
- `ITrackService.FindOrCreateRelease` / `TrackManager.FindOrCreateRelease`: Finds the release row matching (title, artist) or creates one if none exists. Returns `ResultContainer<(ReleaseDto Release, bool WasCreated)>` — the `WasCreated` flag lets the upload CREATE path distinguish a freshly minted release from one returned because a concurrent upload won the insert race (the latter is treated as a duplicate and rejected with 409, not silently attached). `ITrackService.GetReleaseByTitleAndArtist` is the read-only counterpart used for the upload pre-flight check and the ATTACH-path validation.
## Phase 16 — anonymous telemetry domain (EventRepository / EventManager)
`EventRepository` and `EventManager` (with `IEventService` boundary) are the SQL-side domain for anonymous play/share telemetry (Phase 16 waves 16.1 + 16.3). Unlike `TrackRepository`, these entities have no soft-delete lifecycle and are not `BaseEntity`/`IEntity``EventRepository` is a plain context-backed repository against the same scoped `DeepDrftContext`.
- **`EventRepository`** (`Repositories/EventRepository.cs`): append-only writes to the `play_event` and `share_event` tables; incremental-on-write bump of the `play_counter` rollup (D6); server-side track→release resolution at write time (D4) — the client sends only the track `EntryKey`, the repository joins track→release and stamps the `release_id` on the row. Also owns the three distinct-listener aggregation queries added in wave 16.3: `CountDistinctListenersAsync()` (site-wide), `CountDistinctListenersForTrackAsync(trackEntryKey)`, `CountDistinctListenersForReleaseAsync(releaseId)` — each excludes null `anon_id` rows.
- **`EventManager` / `IEventService`** (`EventManager.cs`): `RecordPlay(trackEntryKey, bucket, anonId, ct)` and `RecordShare(targetType, targetKey, channel, anonId, ct)` return NetBlocks `Result`. Wave 16.3 added three distinct-count members returning `ResultContainer<int>`: `GetDistinctListenerCount()`, `GetDistinctListenerCountForTrack(trackEntryKey)`, `GetDistinctListenerCountForRelease(releaseId)`. Registered scoped in `DeepDrftAPI/Program.cs`. Migration: `20260619155610_AddPlayShareTelemetry` (authored; not yet applied — Daniel-gated). The `anon_id` columns and covering indexes on `play_event`/`share_event` are part of this migration — no additional migration was needed for 16.3.
Example:
```csharp
// TrackManager.GetPaged (public via ITrackService)
// Returns PagedResult<TrackDto> — the repository outputs entities, the service outputs DTOs
public async Task<ResultContainer<PagedResult<TrackDto>>> GetPaged(
int pageNumber = 1,
int pageSize = 20,
string? sortColumn = null,
bool sortDescending = false,
CancellationToken cancellationToken = default)
{
try
{
var parameters = new PagingParameters<TrackEntity>
{
Page = pageNumber,
PageSize = pageSize,
OrderBy = GetOrderExpression(sortColumn), // Maps string to LINQ expression
IsDescending = sortDescending
};
var result = await _repository.GetPagedAsync(parameters, cancellationToken);
// Convert to DTO before returning
var dtoResult = new PagedResult<TrackDto>(
result.Items.Select(TrackConverter.Convert).ToList(),
result.TotalCount,
result.PageNumber,
result.PageSize);
return ResultContainer<PagedResult<TrackDto>>.CreatePassResult(dtoResult);
}
catch (Exception e)
{
return ResultContainer<PagedResult<TrackDto>>.CreateFailResult(e.Message);
}
}
```
## Pagination convention
`PagingParameters<T>` holds:
- `Page` (1-based, default 1)
- `PageSize` (default 20, capped at 100)
- `OrderBy: Expression<Func<T, object>>?` (LINQ expression for sorting)
- `IsDescending`
`TrackManager.GetPaged` (the public service method) maps a string `sortColumn` (from the API query) to an expression via a switch:
```csharp
private static Expression<Func<TrackEntity, object>> GetOrderExpression(string? sortColumn)
=> sortColumn switch
{
"TrackName" => e => e.TrackName,
"Artist" => e => e.Artist,
"Album" => e => e.Album ?? "", // Nulls sort to end
"Genre" => e => e.Genre ?? "",
"ReleaseDate" => e => e.ReleaseDate ?? DateOnly.MaxValue,
_ => e => e.Id // Default to ID
};
```
Add new sort columns by extending this switch and the corresponding column names in the API.
## EF Migration commands
Run from the solution root:
```bash
# Add a migration
dotnet ef migrations add MigrationName --project DeepDrftData --startup-project DeepDrftAPI
# Apply to database
dotnet ef database update --project DeepDrftData --startup-project DeepDrftAPI
```
The design-time factory means you can also run `dotnet ef ... --project DeepDrftData` standalone for local development (it doesn't need the startup project).
## Migrations namespace
Migrations live in the `DeepDrftData.Migrations` namespace. Migration files are auto-generated and rarely edited by hand.
## Connection string
- **DeepDrftAPI**: `environment/connections.json``ConnectionStrings:DefaultConnection`
- Always PostgreSQL (Npgsql) — both production and local development.
The design-time factory reads `environment/connections.json` when present; falls back to a Npgsql dummy for CI.
## Service registration
In `DeepDrftAPI/Program.cs`:
```csharp
services.AddDbContext<DeepDrftContext>(options =>
options.UseNpgsql(configuration.GetConnectionString("DefaultConnection")));
services.AddScoped<TrackRepository>();
services.AddScoped<TrackManager>();
services.AddScoped<ITrackService>(sp => sp.GetRequiredService<TrackManager>());
```
This pattern allows callers to depend on `ITrackService` (the public interface) without knowing about `TrackManager` (the implementation).
## Important patterns
- **Required properties**: `EntryKey`, `TrackName`, `Artist` are `required` strings. This compile-time guarantee prevents half-built entities from reaching the database.
- **Optional properties**: `Album?`, `Genre?`, `ReleaseDate?`, `ImagePath?` are nullable. Queries must handle nulls (e.g., `Album ?? ""` in the sort expression to push nulls to end).
- **Result types**: Services return `ResultContainer<T>` or `Result` from NetBlocks. No exceptions propagate to callers — the service catches and wraps.
- **Async operations**: All database methods are async (`GetPagedAsync`, `GetByIdAsync`, etc.). Sync is not available.
## Development commands
```bash
# Build
dotnet build DeepDrftData
# Add migration (from solution root)
dotnet ef migrations add MigrationName --project DeepDrftData --startup-project DeepDrftAPI
# Apply migration
dotnet ef database update --project DeepDrftData --startup-project DeepDrftAPI
# Run API (dual-database host consuming this service)
dotnet run --project DeepDrftAPI
```
## What does NOT live here
- HTTP controllers or middleware (in host projects)
- Blazor components or rendering logic (in host projects)
- FileDatabase or binary content code (in `DeepDrftContent.Services`)
- Host-specific wiring or configuration (in host `Program.cs` / `appsettings.json`)
- Host-internal services like `UnifiedTrackService` (in host `Services/` folders)
When working with this project, focus on the data layer (repository, manager, EF configuration) and ensure all new SQL logic is testable and reusable by multiple consumers (Content API, Public API, CLI) without host-specific dependencies.