catalogscan

type ColumnRow struct {
	Schema    string
	Table     string
	TableType string
	Column    string
	DataType  string
	Nullable  bool
}

type Connection struct {
	Instance string
	Bound    string
}

type Document struct {
	// Key is what an Operation's Document names it by.
	Key string
	// Filename is the name the document is kept under, extension included.
	Filename string
	MIMEType string
	Content  []byte
}

type Exclusion struct {
	ConnectionKind string   `json:"connection_kind"`
	Connection     string   `json:"connection"`
	Entries        []string `json:"entries"`
}

type Exclusions [][]string

type Loop struct {
	// contains filtered or unexported fields
}

type Operation struct {
	// Path is the operation's name under its connection: the operation id for
	// an API, the kind and dotted field path for GraphQL (query.repository).
	Path    []string
	Columns []catalog.ColumnInput
	// Document is the Key of the schema document the operation was read from,
	// or empty.
	Document string
}

type OperationReader interface {
	Connections() []Connection
	Operations(ctx context.Context, bound string) ([]Operation, error)
	// Documents are the connection's schema documents, keyed as each
	// Operation's Document names them. A connection with none returns none.
	Documents(ctx context.Context, bound string) ([]Document, error)
}

type OperationSource struct {
	// contains filtered or unexported fields
}

type Outcome struct {
	ConnectionKind string    `json:"connection_kind"`
	Connection     string    `json:"connection"`
	StartedAt      time.Time `json:"started_at"`
	FinishedAt     time.Time `json:"finished_at"`
	Datasets       int       `json:"datasets"`
	Removed        int       `json:"removed"`
	// Error is why the scan of this connection stopped. A failed scan marks
	// nothing removed: what it did not read is not evidence of absence.
	Error string `json:"error,omitempty"`
	// Unread are the parts of the connection the scan could not read, such
	// as a Trino catalog whose connector is down. The rest is cataloged, and
	// nothing under an unread part is marked removed.
	Unread []Unread `json:"unread"`
}

type Recorder interface {
	RecordOutcome(ctx context.Context, o Outcome) error
}

type Scanner struct {
	// contains filtered or unexported fields
}

type Settings struct {
	// ScanInterval is how often every connection is scanned, as a Go
	// duration ("6h"). Empty means DefaultScanInterval.
	ScanInterval string      `json:"scan_interval"`
	Exclusions   []Exclusion `json:"exclusions"`
	UpdatedBy    string      `json:"updated_by,omitempty"`
	UpdatedAt    *time.Time  `json:"updated_at,omitempty"`
}

type SettingsReader interface {
	Get(ctx context.Context) (Settings, error)
}

type Source interface {
	// Kind is the connection kind, the first part of every reference the
	// source produces.
	Kind() string
	// Connections are the live connections of this kind.
	Connections() []Connection
	// Datasets lists what the connection can see, skipping any dataset the
	// exclusions name. It reports each dataset through yield so a large
	// connection is not held in memory at once.
	Datasets(ctx context.Context, c Connection, skip Exclusions, yield func(catalog.DatasetInput) error) error
	// Types are the dataset types the source produces, the ones a scan of it
	// may mark removed.
	Types() []catalog.DatasetType
}

type Store struct {
	// contains filtered or unexported fields
}

type TrinoReader interface {
	Connections() []Connection
	Catalogs(ctx context.Context, bound string) ([]string, error)
	Columns(ctx context.Context, bound, catalogName string, fn func(ColumnRow) error) error
}

type TrinoSource struct {
	// contains filtered or unexported fields
}

type Unread struct {
	Path  string `json:"path"`
	Error string `json:"error"`
}

type UnreadError struct {
	Parts []Unread
}

type Writer interface {
	PutDataset(ctx context.Context, in catalog.DatasetInput) (*catalog.Dataset, error)
	MarkRemoved(ctx context.Context, r catalog.Removal) (int, error)
}