MuPDF X.Y.Z

fitz/pixmap.h

Index

typedef

struct

enum

function

Pixmap

typedef fz_overprint

typedef struct fz_overprint fz_overprint

Pixmaps represent a set of pixels for a 2 dimensional region of a plane. Each pixel has n components per pixel. The components are in the order process-components, spot-colors, alpha, where there can be 0 of any of those types. The data is in premultiplied alpha when rendering, but non-premultiplied for colorspace conversions and rescaling.

function fz_pixmap_bbox

fz_irect
fz_pixmap_bbox (
        fz_context *ctx,
        const fz_pixmap *pix
)

Return the bounding box for a pixmap.

function fz_pixmap_width

int
fz_pixmap_width (
        fz_context *ctx,
        const fz_pixmap *pix
)

Return the width of the pixmap in pixels.

function fz_pixmap_height

int
fz_pixmap_height (
        fz_context *ctx,
        const fz_pixmap *pix
)

Return the height of the pixmap in pixels.

function fz_pixmap_x

int
fz_pixmap_x (
        fz_context *ctx,
        const fz_pixmap *pix
)

Return the x value of the pixmap in pixels.

function fz_pixmap_y

int
fz_pixmap_y (
        fz_context *ctx,
        const fz_pixmap *pix
)

Return the y value of the pixmap in pixels.

function fz_pixmap_size

size_t
fz_pixmap_size (
        fz_context *ctx,
        fz_pixmap *pix
)

Return sizeof fz_pixmap plus size of data, in bytes.

function fz_new_pixmap

fz_pixmap *
fz_new_pixmap (
        fz_context *ctx,
        fz_colorspace *cs,
        int w,
        int h,
        fz_separations *seps,
        int alpha
)

Create a new pixmap, with its origin at (0,0)

cs:
The colorspace to use for the pixmap, or NULL for an alpha plane/mask.
w:
The width of the pixmap (in pixels)
h:
The height of the pixmap (in pixels)
seps:
Details of separations.
alpha:
0 for no alpha, 1 for alpha.

Returns a pointer to the new pixmap. Throws exception on failure to allocate.

function fz_new_pixmap_with_bbox

fz_pixmap *
fz_new_pixmap_with_bbox (
        fz_context *ctx,
        fz_colorspace *colorspace,
        fz_irect bbox,
        fz_separations *seps,
        int alpha
)

Create a pixmap of a given size, location and pixel format.

The bounding box specifies the size of the created pixmap and where it will be located. The colorspace determines the number of components per pixel. Alpha is always present. Pixmaps are reference counted, so drop references using fz_drop_pixmap.

colorspace:
Colorspace format used for the created pixmap. The pixmap will keep a reference to the colorspace.
bbox:
Bounding box specifying location/size of created pixmap.
seps:
Details of separations.
alpha:
0 for no alpha, 1 for alpha.

Returns a pointer to the new pixmap. Throws exception on failure to allocate.

function fz_new_pixmap_with_data

fz_pixmap *
fz_new_pixmap_with_data (
        fz_context *ctx,
        fz_colorspace *colorspace,
        int w,
        int h,
        fz_separations *seps,
        int alpha,
        int stride,
        unsigned char *samples
)

Create a new pixmap, with its origin at (0,0) using the supplied data block.

cs:
The colorspace to use for the pixmap, or NULL for an alpha plane/mask.
w:
The width of the pixmap (in pixels)
h:
The height of the pixmap (in pixels)
seps:
Details of separations.
alpha:
0 for no alpha, 1 for alpha.
stride:
The byte offset from the pixel data in a row to the pixel data in the next row.
samples:
The data block to keep the samples in.

Returns a pointer to the new pixmap. Throws exception on failure to allocate.

function fz_new_pixmap_with_bbox_and_data

fz_pixmap *
fz_new_pixmap_with_bbox_and_data (
        fz_context *ctx,
        fz_colorspace *colorspace,
        fz_irect rect,
        fz_separations *seps,
        int alpha,
        unsigned char *samples
)

Create a pixmap of a given size, location and pixel format, using the supplied data block.

The bounding box specifies the size of the created pixmap and where it will be located. The colorspace determines the number of components per pixel. Alpha is always present. Pixmaps are reference counted, so drop references using fz_drop_pixmap.

colorspace:
Colorspace format used for the created pixmap. The pixmap will keep a reference to the colorspace.
rect:
Bounding box specifying location/size of created pixmap.
seps:
Details of separations.
alpha:
Number of alpha planes (0 or 1).
samples:
The data block to keep the samples in.

Returns a pointer to the new pixmap. Throws exception on failure to allocate.

function fz_new_pixmap_from_pixmap

fz_pixmap *
fz_new_pixmap_from_pixmap (
        fz_context *ctx,
        fz_pixmap *pixmap,
        const fz_irect *rect
)

Create a new pixmap that represents a subarea of the specified pixmap. A reference is taken to this pixmap that will be dropped on destruction.

The supplied rectangle must be wholly contained within the original pixmap.

Returns a pointer to the new pixmap. Throws exception on failure to allocate.

function fz_clone_pixmap

fz_pixmap *
fz_clone_pixmap (
        fz_context *ctx,
        const fz_pixmap *old
)

Clone a pixmap, copying the pixels and associated data to new storage.

The reference count of ‘old’ is unchanged.

function fz_keep_pixmap

fz_pixmap *
fz_keep_pixmap (
        fz_context *ctx,
        fz_pixmap *pix
)

Increment the reference count for the pixmap. The same pointer is returned.

Never throws exceptions.

function fz_drop_pixmap

void
fz_drop_pixmap (
        fz_context *ctx,
        fz_pixmap *pix
)

Decrement the reference count for the pixmap. When the reference count hits 0, the pixmap is freed.

Never throws exceptions.

function fz_pixmap_colorspace

fz_colorspace *
fz_pixmap_colorspace (
        fz_context *ctx,
        const fz_pixmap *pix
)

Return the colorspace of a pixmap

Returns colorspace.

function fz_pixmap_components

int
fz_pixmap_components (
        fz_context *ctx,
        const fz_pixmap *pix
)

Return the number of components in a pixmap.

Returns the number of components (including spots and alpha).

function fz_pixmap_colorants

int
fz_pixmap_colorants (
        fz_context *ctx,
        const fz_pixmap *pix
)

Return the number of colorants in a pixmap.

Returns the number of colorants (components, less any spots and alpha).

function fz_pixmap_spots

int
fz_pixmap_spots (
        fz_context *ctx,
        const fz_pixmap *pix
)

Return the number of spots in a pixmap.

Returns the number of spots (components, less colorants and alpha). Does not throw exceptions.

function fz_pixmap_alpha

int
fz_pixmap_alpha (
        fz_context *ctx,
        const fz_pixmap *pix
)

Return the number of alpha planes in a pixmap.

Returns the number of alphas. Does not throw exceptions.

function fz_pixmap_samples

unsigned char *
fz_pixmap_samples (
        fz_context *ctx,
        const fz_pixmap *pix
)

Returns a pointer to the pixel data of a pixmap.

Returns the pointer.

function fz_pixmap_stride

int
fz_pixmap_stride (
        fz_context *ctx,
        const fz_pixmap *pix
)

Return the number of bytes in a row in the pixmap.

function fz_set_pixmap_resolution

void
fz_set_pixmap_resolution (
        fz_context *ctx,
        fz_pixmap *pix,
        int xres,
        int yres
)

Set the pixels per inch resolution of the pixmap.

function fz_clear_pixmap_with_value

void
fz_clear_pixmap_with_value (
        fz_context *ctx,
        fz_pixmap *pix,
        int value
)

Clears a pixmap with the given value.

pix:
The pixmap to clear.
value:
Values in the range 0 to 255 are valid. Each component sample for each pixel in the pixmap will be set to this value, while alpha will always be set to 255 (non-transparent).

This function is horrible, and should be removed from the API and replaced with a less magic one.

function fz_fill_pixmap_with_color

void
fz_fill_pixmap_with_color (
        fz_context *ctx,
        fz_pixmap *pix,
        fz_colorspace *colorspace,
        float *color,
        fz_color_params color_params
)

Fill pixmap with solid color.

function fz_clear_pixmap_rect_with_value

void
fz_clear_pixmap_rect_with_value (
        fz_context *ctx,
        fz_pixmap *pix,
        int value,
        fz_irect r
)

Clears a subrect of a pixmap with the given value.

pix:
The pixmap to clear.
value:
Values in the range 0 to 255 are valid. Each component sample for each pixel in the pixmap will be set to this value, while alpha will always be set to 255 (non-transparent).
r:
the rectangle.

function fz_clear_pixmap

void
fz_clear_pixmap (
        fz_context *ctx,
        fz_pixmap *pix
)

Sets all components (including alpha) of all pixels in a pixmap to 0.

pix:
The pixmap to clear.

function fz_invert_pixmap

void
fz_invert_pixmap (
        fz_context *ctx,
        fz_pixmap *pix
)

Invert all the pixels in a pixmap. All components (process and spots) of all pixels are inverted (except alpha, which is unchanged).

function fz_invert_pixmap_alpha

void
fz_invert_pixmap_alpha (
        fz_context *ctx,
        fz_pixmap *pix
)

Invert the alpha of all the pixels in a pixmap.

function fz_invert_pixmap_luminance

void
fz_invert_pixmap_luminance (
        fz_context *ctx,
        fz_pixmap *pix
)

Transform the pixels in a pixmap so that luminance of each pixel is inverted, and the chrominance remains unchanged (as much as accuracy allows).

All components of all pixels are inverted (except alpha, which is unchanged). Only supports Grey and RGB bitmaps.

function fz_tint_pixmap

void
fz_tint_pixmap (
        fz_context *ctx,
        fz_pixmap *pix,
        int black,
        int white
)

Tint all the pixels in an RGB, BGR, or Gray pixmap.

black:
Map black to this hexadecimal RGB color.
white:
Map white to this hexadecimal RGB color.

function fz_invert_pixmap_rect

void
fz_invert_pixmap_rect (
        fz_context *ctx,
        fz_pixmap *image,
        fz_irect rect
)

Invert all the pixels in a given rectangle of a (premultiplied) pixmap. All components of all pixels in the rectangle are inverted (except alpha, which is unchanged).

function fz_invert_pixmap_raw

void
fz_invert_pixmap_raw (
        fz_context *ctx,
        fz_pixmap *pix
)

Invert all the pixels in a non-premultiplied pixmap in a very naive manner.

function fz_gamma_pixmap

void
fz_gamma_pixmap (
        fz_context *ctx,
        fz_pixmap *pix,
        float gamma
)

Apply gamma correction to a pixmap. All components of all pixels are modified (except alpha, which is unchanged).

gamma:
The gamma value to apply; 1.0 for no change.

function fz_convert_pixmap

fz_pixmap *
fz_convert_pixmap (
        fz_context *ctx,
        const fz_pixmap *pix,
        fz_colorspace *cs_des,
        fz_colorspace *prf,
        fz_default_colorspaces *default_cs,
        fz_color_params color_params,
        int keep_alpha
)

Convert an existing pixmap to a desired colorspace. Other properties of the pixmap, such as resolution and position are copied to the converted pixmap.

pix:
The pixmap to convert.
default_cs:
If NULL pix->colorspace is used. It is possible that the data may need to be interpreted as one of the color spaces in default_cs.
cs_des:
Desired colorspace, may be NULL to denote alpha-only.
prf:
Proofing color space through which we need to convert.
color_params:
Parameters that may be used in conversion (e.g. ri).
keep_alpha:
If 0 any alpha component is removed, otherwise alpha is kept if present in the pixmap.

function fz_is_pixmap_monochrome

int
fz_is_pixmap_monochrome (
        fz_context *ctx,
        fz_pixmap *pixmap
)

Check if the pixmap is a 1-channel image containing samples with only values 0 and 255

function fz_alpha_from_gray

fz_pixmap *
fz_alpha_from_gray (
        fz_context *ctx,
        fz_pixmap *gray
)

Implementation details: subject to change.

function fz_alpha_from_rgb

fz_pixmap *
fz_alpha_from_rgb (
        fz_context *ctx,
        fz_pixmap *color
)

function fz_decode_tile

void
fz_decode_tile (
        fz_context *ctx,
        fz_pixmap *pix,
        const float *decode
)

function fz_md5_pixmap

void
fz_md5_pixmap (
        fz_context *ctx,
        fz_pixmap *pixmap,
        unsigned char digest[16]
)

function fz_unpack_stream

fz_stream *
fz_unpack_stream (
        fz_context *ctx,
        fz_stream *src,
        int depth,
        int w,
        int h,
        int n,
        int indexed,
        int pad,
        int skip
)

struct fz_pixmap

struct fz_pixmap
{
        fz_storable storable;
        int x, y, w, h;
        unsigned char n;
        unsigned char s;
        unsigned char alpha;
        unsigned char flags;
        ptrdiff_t stride;
        fz_separations *seps;
        int xres, yres;
        fz_colorspace *colorspace;
        unsigned char *samples;
        fz_pixmap *underlying;
}

Pixmaps represent a set of pixels for a 2 dimensional region of a plane. Each pixel has n components per pixel. The components are in the order process-components, spot-colors, alpha, where there can be 0 of any of those types. The data is in premultiplied alpha when rendering, but non-premultiplied for colorspace conversions and rescaling.

x, y:
The minimum x and y coord of the region in pixels.
w, h:
The width and height of the region in pixels.
n:
The number of color components in the image. n = num composite colors + num spots + num alphas
s:
The number of spot channels in the image.
alpha:
0 for no alpha, 1 for alpha present.
flags:
flag bits. Bit 0: If set, draw the image with linear interpolation. Bit 1: If set, free the samples buffer when the pixmap is destroyed.
stride:
The byte offset from the data for any given pixel to the data for the same pixel on the row below.
seps:
NULL, or a pointer to a separations structure. If NULL, s should be 0.
xres, yres:
Image resolution in dpi. Default is 96 dpi.
colorspace:
Pointer to a colorspace object describing the colorspace the pixmap is in. If NULL, the image is a mask.
samples:
Pointer to the first byte of the pixmap sample data. This is typically a simple block of memory w * h * n bytes of memory in which the components are stored linearly, but with the use of appropriate stride values, scanlines can be stored in different orders, and have different amounts of padding. The first n bytes are components 0 to n-1 for the pixel at (x,y). Each successive n bytes gives another pixel in scanline order as we move across the line. The start of each scanline is offset the start of the previous one by stride bytes.

enum fz_pixmap_flags

enum fz_pixmap_flags
{
        FZ_PIXMAP_FLAG_INTERPOLATE = 1,
        FZ_PIXMAP_FLAG_FREE_SAMPLES = 2
}

function fz_warp_pixmap

fz_pixmap *
fz_warp_pixmap (
        fz_context *ctx,
        fz_pixmap *src,
        fz_quad points,
        int width,
        int height
)

Create a new pixmap from a warped section of another.

Colorspace, resolution etc are inherited from the original. points give the corner points within the original pixmap of a (convex) quadrilateral. These corner points will be ‘warped’ to be the corner points of the returned bitmap, which will have the given width/height.

function fz_autowarp_pixmap

fz_pixmap *
fz_autowarp_pixmap (
        fz_context *ctx,
        fz_pixmap *src,
        fz_quad points
)

As for fz_warp_pixmap, where width/height are automatically ‘guessed’.

function fz_detect_document

int
fz_detect_document (
        fz_context *ctx,
        fz_quad *points,
        fz_pixmap *src
)

Search for a “document” within a pixmap (greyscale or rgb, no alpha).

points should point to an array of 4 fz_points.

If the function return false, no document was found. If true, points has been updated to include the corner positions of the detected document within the src image.

function fz_clone_pixmap_area_with_different_seps

fz_pixmap *
fz_clone_pixmap_area_with_different_seps (
        fz_context *ctx,
        fz_pixmap *src,
        const fz_irect *bbox,
        fz_colorspace *dcs,
        fz_separations *seps,
        fz_color_params color_params,
        fz_default_colorspaces *default_cs
)

Convert between different separation results.

function fz_new_pixmap_from_alpha_channel

fz_pixmap *
fz_new_pixmap_from_alpha_channel (
        fz_context *ctx,
        fz_pixmap *src
)

Extract alpha channel as a separate pixmap. Returns NULL if there is no alpha channel in the source.

function fz_new_pixmap_from_color_and_mask

fz_pixmap *
fz_new_pixmap_from_color_and_mask (
        fz_context *ctx,
        fz_pixmap *color,
        fz_pixmap *mask
)

Combine a pixmap without an alpha channel with a soft mask.

function fz_scale_pixmap

fz_pixmap *
fz_scale_pixmap (
        fz_context *ctx,
        fz_pixmap *src,
        float x,
        float y,
        float w,
        float h,
        const fz_irect *clip
)

Scale the pixmap up or down in size to fit the rectangle. Will return NULL if the scaling factors are out of range. This applies fancy filtering and will anti-alias the edges for subpixel positioning if using non-integer coordinates. If the clip rectangle is set, the returned pixmap may be subset to fit the clip rectangle. Pass NULL to the clip if you want the whole pixmap scaled.

function fz_subsample_pixmap

void
fz_subsample_pixmap (
        fz_context *ctx,
        fz_pixmap *tile,
        int factor
)

Reduces size to: tile->w => (tile->w + 2^factor-1) / 2^factor tile->h => (tile->h + 2^factor-1) / 2^factor

function fz_copy_pixmap_rect

void
fz_copy_pixmap_rect (
        fz_context *ctx,
        fz_pixmap *dest,
        fz_pixmap *src,
        fz_irect r,
        const fz_default_colorspaces *default_cs
)

Copies r (clipped to both src and dest) in src to dest.