Struct MaskData 

pub struct MaskData { /* private fields */ }

A raster mask stored as PNG-encoded bytes.

MaskData provides zero-copy access to PNG metadata (width, height, bit_depth) by reading the IHDR chunk directly, and full encode/decode for pixel data at 1-bit, 8-bit, and 16-bit depths.

PNG layout reference

[0..8]   PNG signature
[8..12]  IHDR chunk length (always 13)
[12..16] IHDR chunk type ("IHDR")
[16..20] width  (big-endian u32)
[20..24] height (big-endian u32)
[24]     bit_depth
[25]     color_type
...

Implementations

impl MaskData

pub fn from_png(png: Vec<u8>) -> Self

Creates a MaskData from raw PNG bytes.

The caller is responsible for ensuring the bytes represent a valid PNG. For validated construction, use from_png_checked.

pub fn from_png_checked(png: Vec<u8>) -> Result<Self, Error>

Creates a MaskData from raw PNG bytes with validation.

Validates that the bytes represent a valid grayscale PNG:

• Length >= 33 bytes (signature + IHDR chunk including CRC)
• PNG magic bytes at offset 0
• Color type byte (offset 25) is 0 (grayscale)

Errors

Returns an error if the bytes are not a valid grayscale PNG.

pub fn is_valid(&self) -> bool

Returns true if the underlying bytes contain a valid PNG signature and are long enough to have an IHDR chunk.

pub fn as_bytes(&self) -> &[u8]

Returns a reference to the underlying PNG bytes.

pub fn into_bytes(self) -> Vec<u8>

Consumes the MaskData and returns the underlying PNG bytes.

pub fn width(&self) -> u32

Returns the image width by reading the PNG IHDR chunk (bytes 16..20).

Returns 0 if the PNG data is too short or invalid.

pub fn height(&self) -> u32

Returns the image height by reading the PNG IHDR chunk (bytes 20..24).

Returns 0 if the PNG data is too short or invalid.

pub fn bit_depth(&self) -> u8

Returns the bit depth by reading the PNG IHDR chunk (byte 24).

Returns 0 if the PNG data is too short or invalid.

pub fn encode( pixels: &[u8], width: u32, height: u32, bit_depth: u8, ) -> Result<Self, Error>

Encodes raw 8-bit grayscale pixels into a PNG.

For bit_depth == 1, pixel values must be 0 or 1 and will be packed into 1-bit-per-pixel PNG rows (MSB first, 8 pixels per byte, with zero-padding on the last byte if width is not a multiple of 8).

For bit_depth == 8, pixels are encoded directly as 8-bit grayscale.

Errors

Returns an error if bit_depth is not 1 or 8, or if pixels.len() does not equal width * height.

pub fn encode_16bit( pixels: &[u16], width: u32, height: u32, ) -> Result<Self, Error>

Encodes raw 16-bit grayscale pixels into a PNG.

Each u16 value is written as two bytes in big-endian order, matching the PNG 16-bit grayscale format.

Errors

Returns an error if pixels.len() does not equal width * height.

pub fn decode(&self) -> Result<Vec<u8>, Error>

Decodes the PNG image to raw pixel bytes.

For 1-bit PNGs, each pixel is unpacked to a single byte (0 or 1). For 8-bit PNGs, pixel bytes are returned directly. For 16-bit PNGs, each pixel yields two bytes in big-endian order.

Errors

Returns an error if the PNG data is malformed or cannot be decoded.

Trait Implementations

impl Clone for MaskData

fn clone(&self) -> MaskData

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for MaskData

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.