Struct Channel 

pub struct Channel {
    pub hshift: u32,
    pub vshift: u32,
    pub component: i32,
    /* private fields */
}

A single channel in a modular image.

Channels store i32 values, which can represent:

• Raw pixel values (0-255 for 8-bit, 0-65535 for 16-bit)
• Prediction residuals (can be negative)
• Transformed values (after RCT, Squeeze, etc.)

Fields

hshift: u32

Horizontal subsampling shift (0 = no subsampling).

vshift: u32

Vertical subsampling shift (0 = no subsampling).

component: i32

Original color component index (-1 = unset). Used to look up per-component quantization tables in lossy modular encoding. Set by LfFrame: Y=0, X=1, B-Y=2. Propagated through Squeeze transforms.

Implementations

impl Channel

pub fn new(width: usize, height: usize) -> Result<Self>

Creates a new channel filled with zeros.

pub fn from_vec(data: Vec<i32>, width: usize, height: usize) -> Result<Self>

Creates a channel from existing data.

pub fn width(&self) -> usize

Returns the width of the channel.

pub fn height(&self) -> usize

Returns the height of the channel.

pub fn len(&self) -> usize

Returns the total number of pixels.

pub fn is_empty(&self) -> bool

Returns true if the channel has no pixels.

pub fn get(&self, x: usize, y: usize) -> i32

Returns a reference to the pixel at (x, y).

pub fn set(&mut self, x: usize, y: usize, value: i32)

Sets the pixel at (x, y).

pub fn row(&self, y: usize) -> &[i32]

Returns a reference to a row.

pub fn row_mut(&mut self, y: usize) -> &mut [i32]

Returns a mutable reference to a row.

pub fn data(&self) -> &[i32]

Returns a reference to the underlying data.

pub fn data_mut(&mut self) -> &mut [i32]

Returns a mutable reference to the underlying data.

pub fn get_clamped(&self, x: isize, y: isize) -> i32

Gets a pixel with boundary handling (returns 0 outside bounds).

pub fn extract_shifted_region( &self, rect_x0: usize, rect_y0: usize, rect_xsize: usize, rect_ysize: usize, ) -> Option<Channel>

Extracts a region from this channel, accounting for hshift/vshift.

The rect is specified in full-resolution image coordinates. It is downshifted by hshift/vshift and clamped to channel bounds. Returns None if the shifted region has zero area.

pub fn extract_grid_cell( &self, gx: usize, gy: usize, group_dim: usize, ) -> Option<Channel>

Extracts a grid cell region matching the decoder’s get_grid_rect logic.

Given a group position (gx, gy) and group_dim (the image-level group size, typically 256), computes the channel-space sub-region for this group.

This matches jxl-rs ModularChannel::get_grid_rect: grid_dim = (group_dim >> hshift, group_dim >> vshift) bx = gx * grid_dim.0, by = gy * grid_dim.1 size = (min(chan_w - bx, grid_dim.0), min(chan_h - by, grid_dim.1))

Returns None if the sub-region has zero area.

pub fn get_clamped_to_edge(&self, x: isize, y: isize) -> i32

Gets a pixel, clamping coordinates to valid range.

Trait Implementations

impl Clone for Channel

fn clone(&self) -> Channel

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Channel

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

Formats the value using the given formatter.