View Source Pdf.Reader.Filter.Flate (ExPDF v1.0.1)

FlateDecode filter — zlib inflate, with optional PNG and TIFF predictor un-filtering.

Predictor support

Predictor values in /DecodeParms:

1 (default) — no predictor.
2 — TIFF Predictor 2 (horizontal differencing), applied row-by-row.
10 — PNG None (row type 0 prefix consumed and discarded).
11 — PNG Sub.
12 — PNG Up.
13 — PNG Average.
14 — PNG Paeth.
15 — PNG Optimal (decoder treats row-type byte as the actual filter; this is the same as having per-row filter selection — just read the type byte per row).

Key	Default	Meaning
`"Predictor"`	`1`	Predictor type (1 = none)
`"Columns"`	`1`	Row width in samples
`"Colors"`	`1`	Number of color components per sample
`"BitsPerComponent"`	`8`	Bits per component

Inflate a zlib-compressed binary, then apply any configured predictor.

@spec decode(binary(), map()) :: {:ok, binary()} | {:error, term()}

Inflate a zlib-compressed binary, then apply any configured predictor.