dbqcore

type Check struct {
	ID          string       `yaml:"id"`
	Description string       `yaml:"description,omitempty"` // optional
	OnFail      OnFailAction `yaml:"on_fail,omitempty"`     // optional (error, warn)
	Query       string       `yaml:"query,omitempty"`       // optional raw query
}

type ChecksConfig struct {
	Version     string       `yaml:"version"`
	Validations []Validation `yaml:"validations"`
}

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

type ColumnInfo struct {
	Name     string
	Type     string
	Comment  string
	Position uint
}

type ColumnMetrics struct {
	ColumnName          string   `json:"col_name"`
	ColumnComment       string   `json:"col_comment"`
	ColumnPosition      uint     `json:"col_position"`
	DataType            string   `json:"data_type"`
	NullCount           uint64   `json:"null_count"`
	BlankCount          *int64   `json:"blank_count,omitempty"`         // string only
	MinValue            *float64 `json:"min_value,omitempty"`           // numeric only
	MaxValue            *float64 `json:"max_value,omitempty"`           // numeric only
	AvgValue            *float64 `json:"avg_value,omitempty"`           // numeric only
	StddevValue         *float64 `json:"stddev_value,omitempty"`        // numeric only (Population StdDev)
	MostFrequentValue   *string  `json:"most_frequent_value,omitempty"` // pointer to handle NULL as most frequent
	ProfilingDurationMs int64    `json:"profiling_duration_ms"`
}

type ConfigDetails struct {
	Host     string `yaml:"host"`
	Port     int    `yaml:"port"`
	Username string `yaml:"username"`
	Password string `yaml:"password"`
	Database string `yaml:"database,omitempty"`
}

type DataSource struct {
	ID            string        `yaml:"id"`
	Type          string        `yaml:"type"`
	Configuration ConfigDetails `yaml:"configuration"`
	Datasets      []string      `yaml:"datasets"`
}

type DbqConfig struct {
	Version     string       `yaml:"version"`
	DataSources []DataSource `yaml:"datasources"`
}

type DbqConnector interface {
	Ping() (string, error)
	ImportDatasets(filter string) ([]string, error)
	ProfileDataset(dataset string, sample bool, maxConcurrent int) (*TableMetrics, error)
	RunCheck(check *Check, dataset string, defaultWhere string) (bool, string, error)
}

type OnFailAction string

type TableMetrics struct {
	ProfiledAt          int64                     `json:"profiled_at"`
	TableName           string                    `json:"table_name"`
	DatabaseName        string                    `json:"database_name"`
	TotalRows           uint64                    `json:"total_rows"`
	ColumnsMetrics      map[string]*ColumnMetrics `json:"columns_metrics"`
	RowsSample          []map[string]interface{}  `json:"rows_sample"`
	ProfilingDurationMs int64                     `json:"profiling_duration_ms"`
	DbqErrors           []error                   `json:"_dbq_errors"`
}

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

type Validation struct {
	Dataset string  `yaml:"dataset"`
	Where   string  `yaml:"where,omitempty"` // Optional where clause
	Checks  []Check `yaml:"checks"`
}