Flow PHP

SourceStatistics

Read onlyYes

FinalYes

Properties

$complete : bool
$declared : Statistics
$extractor : string
$rows : int

Methods

__construct() : mixed
normalize() : array{extractor: string, declared: array{rows: array, size: array}, rows: int, complete: bool, rows_error: null|float}: rows_error: rowsError() when it is a finite number; null without an estimate, for an incomplete read, or when the estimate was positive against 0 rows measured (infinitely off).
rowsError() : float|null: How far the declared estimate was off, relative to the rows measured. Null without an estimate: an upper bound alone is a guarantee, not a guess. Null for an incomplete read too: the declaration describes the whole source, the rows measured only part of it. Against 0 measured rows any positive estimate is infinitely off.

$complete


        public
            bool
    $complete

$declared


        public
            Statistics
    $declared

$extractor


        public
            string
    $extractor

$rows


        public
            int
    $rows

__construct()


    public
                    __construct(string $extractor, Statistics $declared, int $rows, bool $complete) : mixed

Parameters

$extractor : string: the source's short class name
$declared : Statistics
$rows : int
$complete : bool: every read of the source ran to its end with no limit or partition filter pushed into it

normalize()

rows_error: rowsError() when it is a finite number; null without an estimate, for an incomplete read, or when the estimate was positive against 0 rows measured (infinitely off).


    public
                    normalize() : array{extractor: string, declared: array{rows: array, size: array}, rows: int, complete: bool, rows_error: null|float}

Return values

array{extractor: string, declared: array{rows: array, size: array}, rows: int, complete: bool, rows_error: null|float}

rowsError()

How far the declared estimate was off, relative to the rows measured. Null without an estimate: an upper bound alone is a guarantee, not a guess. Null for an incomplete read too: the declaration describes the whole source, the rows measured only part of it. Against 0 measured rows any positive estimate is infinitely off.


    public
                    rowsError() : float|null

Return values

float|null