foot/sixel.c

1664 lines
51 KiB
C
Raw Normal View History

#include "sixel.h"
#include <string.h>
#include <limits.h>
#define LOG_MODULE "sixel"
#define LOG_ENABLE_DBG 0
#include "log.h"
#include "debug.h"
#include "grid.h"
#include "hsl.h"
#include "render.h"
#include "util.h"
#include "xmalloc.h"
#include "xsnprintf.h"
static size_t count;
void
sixel_fini(struct terminal *term)
{
free(term->sixel.image.data);
free(term->sixel.private_palette);
free(term->sixel.shared_palette);
}
void
sixel_init(struct terminal *term, int p1, int p2, int p3)
{
/*
* P1: pixel aspect ratio - unimplemented
* P2: background color mode
* - 0|2: empty pixels use current background color
* - 1: empty pixels remain at their current color (i.e. transparent)
* P3: horizontal grid size - ignored
*/
xassert(term->sixel.image.data == NULL);
xassert(term->sixel.palette_size <= SIXEL_MAX_COLORS);
/* Default aspect ratio is 2:1 */
const int pad = 1;
const int pan =
(p1 == 2) ? 5 :
(p1 == 3 || p1 == 4) ? 3 :
(p1 == 7 || p1 == 8 || p1 == 9) ? 1 : 2;
LOG_DBG("initializing sixel with "
"p1=%d (pan=%d, pad=%d, AR=%d:%d), "
"p2=%d (transparent=%d), "
"p3=%d (ignored)",
p1, pan, pad, pan, pad, p2, p2 == 1, p3);
term->sixel.state = SIXEL_DECSIXEL;
term->sixel.pos = (struct coord){0, 0};
term->sixel.row_byte_ofs = 0;
term->sixel.color_idx = 0;
term->sixel.pan = pan;
term->sixel.pad = pad;
term->sixel.param = 0;
term->sixel.param_idx = 0;
memset(term->sixel.params, 0, sizeof(term->sixel.params));
term->sixel.transparent_bg = p2 == 1;
term->sixel.image.data = NULL;
term->sixel.image.width = 0;
term->sixel.image.height = 6 * pan;
/* TODO: default palette */
if (term->sixel.use_private_palette) {
xassert(term->sixel.private_palette == NULL);
term->sixel.private_palette = xcalloc(
term->sixel.palette_size, sizeof(term->sixel.private_palette[0]));
term->sixel.palette = term->sixel.private_palette;
} else {
if (term->sixel.shared_palette == NULL) {
term->sixel.shared_palette = xcalloc(
term->sixel.palette_size, sizeof(term->sixel.shared_palette[0]));
} else {
/* Shared palette - do *not* reset palette for new sixels */
}
term->sixel.palette = term->sixel.shared_palette;
}
uint32_t bg = 0;
switch (term->vt.attrs.bg_src) {
case COLOR_RGB:
bg = 0xffu << 24 | term->vt.attrs.bg;
break;
case COLOR_BASE16:
case COLOR_BASE256:
bg = 0xffu << 24 | term->colors.table[term->vt.attrs.bg];
break;
case COLOR_DEFAULT:
if (term->colors.alpha == 0xffff)
bg = 0xffu << 24 | term->colors.bg;
else {
/* Alpha needs to be pre-multiplied */
uint32_t r = (term->colors.bg >> 16) & 0xff;
uint32_t g = (term->colors.bg >> 8) & 0xff;
uint32_t b = (term->colors.bg >> 0) & 0xff;
uint32_t alpha = term->colors.alpha;
r *= alpha; r /= 0xffff;
g *= alpha; g /= 0xffff;
b *= alpha; b /= 0xffff;
bg = (alpha >> 8) << 24 | (r & 0xff) << 16 | (g & 0xff) << 8 | (b & 0xff);
}
break;
}
term->sixel.default_bg = term->sixel.transparent_bg
? 0x00000000u
: bg;
count = 0;
}
2020-02-22 00:23:19 +01:00
void
sixel_destroy(struct sixel *sixel)
{
if (sixel->pix != NULL)
pixman_image_unref(sixel->pix);
2020-02-22 00:23:19 +01:00
free(sixel->data);
2020-02-22 00:23:19 +01:00
sixel->pix = NULL;
sixel->data = NULL;
}
void
sixel_destroy_all(struct terminal *term)
{
tll_foreach(term->normal.sixel_images, it)
sixel_destroy(&it->item);
tll_foreach(term->alt.sixel_images, it)
sixel_destroy(&it->item);
tll_free(term->normal.sixel_images);
tll_free(term->alt.sixel_images);
}
static void
sixel_erase(struct terminal *term, struct sixel *sixel)
{
for (int i = 0; i < sixel->rows; i++) {
int r = (sixel->pos.row + i) & (term->grid->num_rows - 1);
struct row *row = term->grid->rows[r];
if (row == NULL) {
/* A resize/reflow may cause row to now be unallocated */
continue;
}
row->dirty = true;
for (int c = sixel->pos.col; c < min(sixel->cols, term->cols); c++)
row->cells[c].attrs.clean = 0;
}
sixel_destroy(sixel);
}
/*
* Verify the sixels are sorted correctly.
*
* The sixels are sorted on their *end* row, in descending order. This
* invariant means the most recent sixels appear first in the list.
*/
static void
verify_list_order(const struct terminal *term)
{
#if defined(_DEBUG)
int prev_row = INT_MAX;
int prev_col = -1;
int prev_col_count = 0;
/* To aid debugging */
size_t UNUSED idx = 0;
tll_foreach(term->grid->sixel_images, it) {
int row = grid_row_abs_to_sb(
term->grid, term->rows, it->item.pos.row + it->item.rows - 1);
int col = it->item.pos.col;
int col_count = it->item.cols;
xassert(row <= prev_row);
if (row == prev_row) {
/* Allowed to be on the same row only if their columns
* don't overlap */
xassert(col + col_count <= prev_col ||
prev_col + prev_col_count <= col);
}
prev_row = row;
prev_col = col;
prev_col_count = col_count;
idx++;
}
#endif
}
/*
* Verifies there aren't any sixels that cross the scrollback
* wrap-around. This invariant means a sixel's absolute row numbers
* are strictly increasing.
*/
static void
verify_no_wraparound_crossover(const struct terminal *term)
{
#if defined(_DEBUG)
tll_foreach(term->grid->sixel_images, it) {
const struct sixel *six = &it->item;
xassert(six->pos.row >= 0);
xassert(six->pos.row < term->grid->num_rows);
int end = (six->pos.row + six->rows - 1) & (term->grid->num_rows - 1);
xassert(end >= six->pos.row);
}
#endif
}
/*
* Verify there aren't any sixels that cross the scrollback end. This
* invariant means a sixel's rebased row numbers are strictly
* increasing.
*/
static void
verify_scrollback_consistency(const struct terminal *term)
{
#if defined(_DEBUG)
tll_foreach(term->grid->sixel_images, it) {
const struct sixel *six = &it->item;
int last_row = -1;
for (int i = 0; i < six->rows; i++) {
int row_no = grid_row_abs_to_sb(
term->grid, term->rows, six->pos.row + i);
if (last_row != -1)
xassert(last_row < row_no);
last_row = row_no;
}
}
#endif
}
/*
* Verifies no sixel overlap with any other sixels.
*/
static void
verify_no_overlap(const struct terminal *term)
{
#if defined(_DEBUG)
tll_foreach(term->grid->sixel_images, it) {
const struct sixel *six1 = &it->item;
pixman_region32_t rect1;
pixman_region32_init_rect(
&rect1, six1->pos.col, six1->pos.row, six1->cols, six1->rows);
tll_foreach(term->grid->sixel_images, it2) {
const struct sixel *six2 = &it2->item;
if (six1 == six2)
continue;
pixman_region32_t rect2;
pixman_region32_init_rect(
&rect2, six2->pos.col,
six2->pos.row, six2->cols, six2->rows);
pixman_region32_t intersection;
pixman_region32_init(&intersection);
pixman_region32_intersect(&intersection, &rect1, &rect2);
xassert(!pixman_region32_not_empty(&intersection));
pixman_region32_fini(&intersection);
pixman_region32_fini(&rect2);
}
pixman_region32_fini(&rect1);
}
#endif
}
static void
verify_sixels(const struct terminal *term)
{
verify_no_wraparound_crossover(term);
verify_scrollback_consistency(term);
verify_no_overlap(term);
verify_list_order(term);
}
static void
sixel_insert(struct terminal *term, struct sixel sixel)
{
int end_row = grid_row_abs_to_sb(
term->grid, term->rows, sixel.pos.row + sixel.rows - 1);
tll_foreach(term->grid->sixel_images, it) {
int rebased = grid_row_abs_to_sb(
term->grid, term->rows, it->item.pos.row + it->item.rows - 1);
if (rebased < end_row) {
tll_insert_before(term->grid->sixel_images, it, sixel);
goto out;
}
}
tll_push_back(term->grid->sixel_images, sixel);
out:
#if defined(LOG_ENABLE_DBG) && LOG_ENABLE_DBG
LOG_DBG("sixel list after insertion:");
tll_foreach(term->grid->sixel_images, it) {
LOG_DBG(" rows=%d+%d", it->item.pos.row, it->item.rows);
}
#endif
verify_sixels(term);
}
void
sixel_scroll_up(struct terminal *term, int rows)
{
if (likely(tll_length(term->grid->sixel_images) == 0))
return;
tll_rforeach(term->grid->sixel_images, it) {
struct sixel *six = &it->item;
int six_start = grid_row_abs_to_sb(term->grid, term->rows, six->pos.row);
if (six_start < rows) {
sixel_erase(term, six);
tll_remove(term->grid->sixel_images, it);
} else {
/*
* Unfortunately, we cannot break here.
*
* The sixels are sorted on their *end* row. This means
* there may be a sixel with a top row that will be
* scrolled out *anywhere* in the list (think of a huuuuge
* sixel that covers the entire scrollback)
*/
//break;
}
}
term_update_ascii_printer(term);
verify_sixels(term);
}
void
sixel_scroll_down(struct terminal *term, int rows)
{
if (likely(tll_length(term->grid->sixel_images) == 0))
return;
xassert(term->grid->num_rows >= rows);
tll_foreach(term->grid->sixel_images, it) {
struct sixel *six = &it->item;
int six_end = grid_row_abs_to_sb(
term->grid, term->rows, six->pos.row + six->rows - 1);
if (six_end >= term->grid->num_rows - rows) {
sixel_erase(term, six);
tll_remove(term->grid->sixel_images, it);
} else
break;
}
term_update_ascii_printer(term);
verify_sixels(term);
}
sixel: add support for overlapping sixels Writing a sixel on top of an already existing sixel currently has the following limitations in foot: * The parts of the first sixel that is covered by the new sixel are removed, completely. Even if the new sixel has transparent areas. I.e. writing a transparent sixel on top of another sixel *replaces* the first sixel with the new sixel, instead of layering them on top of each other. * The second sixel erases the first sixel cell-wise. That is, a sixel whose size isn’t a multiple of the cell dimensions will leave unsightly holes in the first sixel. This patch takes care of both issues. The first one is actually the easiest one: all we need to do is calculate the intersection, and blend the two images. To keep things relatively simple, we use the pixman image from the *new* image, and use the ‘OVER_REVERSE’ operation to blend the new image over the old one. That is, the old image is still split into four tiles (top, left, right, bottom), just like before. But instead of throwing away the fifth middle tile, we blend it with the new image. As an optimization, this is only done if the new image has transparency (P1=1). The second problem is solved by detecting when we’re erasing an area from the second image that is larger than the new image. In this case, we enlarge the new image, and copy the old image into the new one. Finally, when we enlarge the new image, there may be areas in the new image that is *not* covered by the old image. These areas are made transparent. The end result is: * Each cell is covered by at *most* 1 sixel image. I.e. the total numbers of sixels are finite. This is important for the ‘mpv --vo=sixel’ use case - we don’t want to end up with thousands of sixels layered on top of each other. * Writing an opaque sixel on top of another sixel has _almost_ zero performance impact. Especially if the two sixels have the same size, so that we don’t have to resize the new image. Again, important for the ‘mpv --vo=sixel’ use case. Closes #562
2021-06-06 21:38:31 +02:00
static void
blend_new_image_over_old(const struct terminal *term,
const struct sixel *six, pixman_region32_t *six_rect,
int row, int col, pixman_image_t **pix, bool *opaque)
{
xassert(pix != NULL);
xassert(opaque != NULL);
const int six_ofs_x = six->pos.col * term->cell_width;
const int six_ofs_y = six->pos.row * term->cell_height;
const int img_ofs_x = col * term->cell_width;
const int img_ofs_y = row * term->cell_height;
const int img_width = pixman_image_get_width(*pix);
const int img_height = pixman_image_get_height(*pix);
pixman_region32_t pix_rect;
pixman_region32_init_rect(
&pix_rect, img_ofs_x, img_ofs_y, img_width, img_height);
/* Blend the intersection between the old and new images */
pixman_region32_t intersection;
pixman_region32_init(&intersection);
pixman_region32_intersect(&intersection, six_rect, &pix_rect);
int n_rects = -1;
pixman_box32_t *boxes = pixman_region32_rectangles(
&intersection, &n_rects);
if (n_rects == 0)
goto out;
xassert(n_rects == 1);
pixman_box32_t *box = &boxes[0];
if (!*opaque) {
/*
* New image is transparent - blend on top of the old
* sixel image.
*/
pixman_image_composite32(
PIXMAN_OP_OVER_REVERSE,
six->pix, NULL, *pix,
box->x1 - six_ofs_x, box->y1 - six_ofs_y,
0, 0,
box->x1 - img_ofs_x, box->y1 - img_ofs_y,
box->x2 - box->x1, box->y2 - box->y1);
}
/*
* Since the old image is split into sub-tiles on a
* per-row basis, we need to enlarge the new image and
* copy the old image if the old image extends beyond the
* new image.
*
* The "bounding" coordinates are either the edges of the
* old image, or the next cell boundary, whichever comes
* first.
*/
int bounding_x = six_ofs_x + six->width > img_ofs_x + img_width
? min(
six_ofs_x + six->width,
(box->x2 + term->cell_width - 1) / term->cell_width * term->cell_width)
: box->x2;
int bounding_y = six_ofs_y + six->height > img_ofs_y + img_height
? min(
six_ofs_y + six->height,
(box->y2 + term->cell_height - 1) / term->cell_height * term->cell_height)
: box->y2;
/* The required size of the new image */
const int required_width = bounding_x - img_ofs_x;
const int required_height = bounding_y - img_ofs_y;
const int new_width = max(img_width, required_width);
const int new_height = max(img_height, required_height);
if (new_width <= img_width && new_height <= img_height)
goto out;
//LOG_INFO("enlarging: %dx%d -> %dx%d", img_width, img_height, new_width, new_height);
if (!six->opaque) {
/* Transparency is viral */
*opaque = false;
}
/* Create a new pixmap */
int stride = new_width * sizeof(uint32_t);
uint32_t *new_data = xmalloc(stride * new_height);
pixman_image_t *pix2 = pixman_image_create_bits_no_clear(
PIXMAN_a8r8g8b8, new_width, new_height, new_data, stride);
#if defined(_DEBUG)
/* Fill new image with an easy-to-recognize color (green) */
for (size_t i = 0; i < new_width * new_height; i++)
new_data[i] = 0xff00ff00;
#endif
/* Copy the new image, from its old pixmap, to the new pixmap */
pixman_image_composite32(
PIXMAN_OP_SRC,
*pix, NULL, pix2, 0, 0, 0, 0, 0, 0, img_width, img_height);
/* Copy the bottom tile of the old sixel image into the new pixmap */
pixman_image_composite32(
PIXMAN_OP_SRC,
six->pix, NULL, pix2,
box->x1 - six_ofs_x, box->y2 - six_ofs_y,
0, 0,
box->x1 - img_ofs_x, box->y2 - img_ofs_y,
bounding_x - box->x1, bounding_y - box->y2);
/* Copy the right tile of the old sixel image into the new pixmap */
pixman_image_composite32(
PIXMAN_OP_SRC,
six->pix, NULL, pix2,
box->x2 - six_ofs_x, box->y1 - six_ofs_y,
0, 0,
box->x2 - img_ofs_x, box->y1 - img_ofs_y,
bounding_x - box->x2, bounding_y - box->y1);
/*
* Ensure the newly allocated area is initialized.
*
* Some of it, or all, will have been initialized above, by the
* bottom and right tiles from the old sixel image. However, there
* may be areas in the new image that isn't covered by the old
* image. These areas need to be made transparent.
*/
pixman_region32_t uninitialized;
pixman_region32_init_rects(
&uninitialized,
(const pixman_box32_t []){
/* Extended image area on the right side */
{img_ofs_x + img_width, img_ofs_y, img_ofs_x + new_width, img_ofs_y + new_height},
/* Bottom */
{img_ofs_x, img_ofs_y + img_height, img_ofs_x + new_width, img_ofs_y + new_height}},
2);
/* Subtract the old sixel image, since the area(s) covered by the
* old image has already been copied, and *must* not be
* overwritten */
pixman_region32_t diff;
pixman_region32_init(&diff);
pixman_region32_subtract(&diff, &uninitialized, six_rect);
if (pixman_region32_not_empty(&diff)) {
pixman_image_t *src =
pixman_image_create_solid_fill(&(pixman_color_t){0});
int count = -1;
pixman_box32_t *rects = pixman_region32_rectangles(&diff, &count);
for (int i = 0; i < count; i++) {
pixman_image_composite32(
PIXMAN_OP_SRC,
src, NULL, pix2,
0, 0, 0, 0,
rects[i].x1 - img_ofs_x, rects[i].y1 - img_ofs_y,
rects[i].x2 - rects[i].x1,
rects[i].y2 - rects[i].y1);
}
pixman_image_unref(src);
*opaque = false;
}
pixman_region32_fini(&diff);
pixman_region32_fini(&uninitialized);
/* Use the new pixmap in place of the old one */
free(pixman_image_get_data(*pix));
pixman_image_unref(*pix);
*pix = pix2;
out:
pixman_region32_fini(&intersection);
pixman_region32_fini(&pix_rect);
}
static void
2020-06-27 15:29:47 +02:00
sixel_overwrite(struct terminal *term, struct sixel *six,
sixel: add support for overlapping sixels Writing a sixel on top of an already existing sixel currently has the following limitations in foot: * The parts of the first sixel that is covered by the new sixel are removed, completely. Even if the new sixel has transparent areas. I.e. writing a transparent sixel on top of another sixel *replaces* the first sixel with the new sixel, instead of layering them on top of each other. * The second sixel erases the first sixel cell-wise. That is, a sixel whose size isn’t a multiple of the cell dimensions will leave unsightly holes in the first sixel. This patch takes care of both issues. The first one is actually the easiest one: all we need to do is calculate the intersection, and blend the two images. To keep things relatively simple, we use the pixman image from the *new* image, and use the ‘OVER_REVERSE’ operation to blend the new image over the old one. That is, the old image is still split into four tiles (top, left, right, bottom), just like before. But instead of throwing away the fifth middle tile, we blend it with the new image. As an optimization, this is only done if the new image has transparency (P1=1). The second problem is solved by detecting when we’re erasing an area from the second image that is larger than the new image. In this case, we enlarge the new image, and copy the old image into the new one. Finally, when we enlarge the new image, there may be areas in the new image that is *not* covered by the old image. These areas are made transparent. The end result is: * Each cell is covered by at *most* 1 sixel image. I.e. the total numbers of sixels are finite. This is important for the ‘mpv --vo=sixel’ use case - we don’t want to end up with thousands of sixels layered on top of each other. * Writing an opaque sixel on top of another sixel has _almost_ zero performance impact. Especially if the two sixels have the same size, so that we don’t have to resize the new image. Again, important for the ‘mpv --vo=sixel’ use case. Closes #562
2021-06-06 21:38:31 +02:00
int row, int col, int height, int width,
pixman_image_t **pix, bool *opaque)
{
pixman_region32_t six_rect;
pixman_region32_init_rect(
&six_rect,
six->pos.col * term->cell_width, six->pos.row * term->cell_height,
six->width, six->height);
pixman_region32_t overwrite_rect;
pixman_region32_init_rect(
&overwrite_rect,
col * term->cell_width, row * term->cell_height,
width * term->cell_width, height * term->cell_height);
#if defined(_DEBUG)
sixel: add support for overlapping sixels Writing a sixel on top of an already existing sixel currently has the following limitations in foot: * The parts of the first sixel that is covered by the new sixel are removed, completely. Even if the new sixel has transparent areas. I.e. writing a transparent sixel on top of another sixel *replaces* the first sixel with the new sixel, instead of layering them on top of each other. * The second sixel erases the first sixel cell-wise. That is, a sixel whose size isn’t a multiple of the cell dimensions will leave unsightly holes in the first sixel. This patch takes care of both issues. The first one is actually the easiest one: all we need to do is calculate the intersection, and blend the two images. To keep things relatively simple, we use the pixman image from the *new* image, and use the ‘OVER_REVERSE’ operation to blend the new image over the old one. That is, the old image is still split into four tiles (top, left, right, bottom), just like before. But instead of throwing away the fifth middle tile, we blend it with the new image. As an optimization, this is only done if the new image has transparency (P1=1). The second problem is solved by detecting when we’re erasing an area from the second image that is larger than the new image. In this case, we enlarge the new image, and copy the old image into the new one. Finally, when we enlarge the new image, there may be areas in the new image that is *not* covered by the old image. These areas are made transparent. The end result is: * Each cell is covered by at *most* 1 sixel image. I.e. the total numbers of sixels are finite. This is important for the ‘mpv --vo=sixel’ use case - we don’t want to end up with thousands of sixels layered on top of each other. * Writing an opaque sixel on top of another sixel has _almost_ zero performance impact. Especially if the two sixels have the same size, so that we don’t have to resize the new image. Again, important for the ‘mpv --vo=sixel’ use case. Closes #562
2021-06-06 21:38:31 +02:00
pixman_region32_t cell_intersection;
pixman_region32_init(&cell_intersection);
pixman_region32_intersect(&cell_intersection, &six_rect, &overwrite_rect);
xassert(pixman_region32_not_empty(&cell_intersection));
pixman_region32_fini(&cell_intersection);
#endif
sixel: add support for overlapping sixels Writing a sixel on top of an already existing sixel currently has the following limitations in foot: * The parts of the first sixel that is covered by the new sixel are removed, completely. Even if the new sixel has transparent areas. I.e. writing a transparent sixel on top of another sixel *replaces* the first sixel with the new sixel, instead of layering them on top of each other. * The second sixel erases the first sixel cell-wise. That is, a sixel whose size isn’t a multiple of the cell dimensions will leave unsightly holes in the first sixel. This patch takes care of both issues. The first one is actually the easiest one: all we need to do is calculate the intersection, and blend the two images. To keep things relatively simple, we use the pixman image from the *new* image, and use the ‘OVER_REVERSE’ operation to blend the new image over the old one. That is, the old image is still split into four tiles (top, left, right, bottom), just like before. But instead of throwing away the fifth middle tile, we blend it with the new image. As an optimization, this is only done if the new image has transparency (P1=1). The second problem is solved by detecting when we’re erasing an area from the second image that is larger than the new image. In this case, we enlarge the new image, and copy the old image into the new one. Finally, when we enlarge the new image, there may be areas in the new image that is *not* covered by the old image. These areas are made transparent. The end result is: * Each cell is covered by at *most* 1 sixel image. I.e. the total numbers of sixels are finite. This is important for the ‘mpv --vo=sixel’ use case - we don’t want to end up with thousands of sixels layered on top of each other. * Writing an opaque sixel on top of another sixel has _almost_ zero performance impact. Especially if the two sixels have the same size, so that we don’t have to resize the new image. Again, important for the ‘mpv --vo=sixel’ use case. Closes #562
2021-06-06 21:38:31 +02:00
if (pix != NULL)
blend_new_image_over_old(term, six, &six_rect, row, col, pix, opaque);
pixman_region32_t diff;
pixman_region32_init(&diff);
pixman_region32_subtract(&diff, &six_rect, &overwrite_rect);
pixman_region32_fini(&six_rect);
pixman_region32_fini(&overwrite_rect);
int n_rects = -1;
pixman_box32_t *boxes = pixman_region32_rectangles(&diff, &n_rects);
for (int i = 0; i < n_rects; i++) {
LOG_DBG("box #%d: x1=%d, y1=%d, x2=%d, y2=%d", i,
boxes[i].x1, boxes[i].y1, boxes[i].x2, boxes[i].y2);
xassert(boxes[i].x1 % term->cell_width == 0);
xassert(boxes[i].y1 % term->cell_height == 0);
/* New image's position, in cells */
const int new_col = boxes[i].x1 / term->cell_width;
const int new_row = boxes[i].y1 / term->cell_height;
xassert(new_row < term->grid->num_rows);
/* New image's width and height, in pixels */
const int new_width = boxes[i].x2 - boxes[i].x1;
const int new_height = boxes[i].y2 - boxes[i].y1;
uint32_t *new_data = xmalloc(new_width * new_height * sizeof(uint32_t));
const uint32_t *old_data = six->data;
/* Pixel offsets into old image backing memory */
const int x_ofs = boxes[i].x1 - six->pos.col * term->cell_width;
const int y_ofs = boxes[i].y1 - six->pos.row * term->cell_height;
/* Copy image data, one row at a time */
for (size_t j = 0; j < new_height; j++) {
memcpy(
&new_data[(0 + j) * new_width],
&old_data[(y_ofs + j) * six->width + x_ofs],
new_width * sizeof(uint32_t));
}
pixman_image_t *new_pix = pixman_image_create_bits_no_clear(
PIXMAN_a8r8g8b8,
new_width, new_height, new_data, new_width * sizeof(uint32_t));
struct sixel new_six = {
.data = new_data,
.pix = new_pix,
.width = new_width,
.height = new_height,
.pos = {.col = new_col, .row = new_row},
.cols = (new_width + term->cell_width - 1) / term->cell_width,
.rows = (new_height + term->cell_height - 1) / term->cell_height,
.opaque = six->opaque,
};
#if defined(_DEBUG)
/* Assert we don't cross the scrollback wrap-around */
const int new_end = new_six.pos.row + new_six.rows - 1;
xassert(new_end < term->grid->num_rows);
#endif
sixel_insert(term, new_six);
}
pixman_region32_fini(&diff);
}
/* Row numbers are absolute */
static void
2020-06-27 15:29:47 +02:00
_sixel_overwrite_by_rectangle(
sixel: add support for overlapping sixels Writing a sixel on top of an already existing sixel currently has the following limitations in foot: * The parts of the first sixel that is covered by the new sixel are removed, completely. Even if the new sixel has transparent areas. I.e. writing a transparent sixel on top of another sixel *replaces* the first sixel with the new sixel, instead of layering them on top of each other. * The second sixel erases the first sixel cell-wise. That is, a sixel whose size isn’t a multiple of the cell dimensions will leave unsightly holes in the first sixel. This patch takes care of both issues. The first one is actually the easiest one: all we need to do is calculate the intersection, and blend the two images. To keep things relatively simple, we use the pixman image from the *new* image, and use the ‘OVER_REVERSE’ operation to blend the new image over the old one. That is, the old image is still split into four tiles (top, left, right, bottom), just like before. But instead of throwing away the fifth middle tile, we blend it with the new image. As an optimization, this is only done if the new image has transparency (P1=1). The second problem is solved by detecting when we’re erasing an area from the second image that is larger than the new image. In this case, we enlarge the new image, and copy the old image into the new one. Finally, when we enlarge the new image, there may be areas in the new image that is *not* covered by the old image. These areas are made transparent. The end result is: * Each cell is covered by at *most* 1 sixel image. I.e. the total numbers of sixels are finite. This is important for the ‘mpv --vo=sixel’ use case - we don’t want to end up with thousands of sixels layered on top of each other. * Writing an opaque sixel on top of another sixel has _almost_ zero performance impact. Especially if the two sixels have the same size, so that we don’t have to resize the new image. Again, important for the ‘mpv --vo=sixel’ use case. Closes #562
2021-06-06 21:38:31 +02:00
struct terminal *term, int row, int col, int height, int width,
pixman_image_t **pix, bool *opaque)
{
verify_sixels(term);
#if defined(_DEBUG)
pixman_region32_t overwrite_rect;
pixman_region32_init_rect(&overwrite_rect, col, row, width, height);
#endif
const int start = row;
const int end = row + height - 1;
/* We should never generate scrollback wrapping sixels */
xassert(end < term->grid->num_rows);
const int scrollback_rel_start = grid_row_abs_to_sb(
term->grid, term->rows, start);
bool UNUSED would_have_breaked = false;
tll_foreach(term->grid->sixel_images, it) {
struct sixel *six = &it->item;
const int six_start = six->pos.row;
const int six_end = (six_start + six->rows - 1);
const int six_scrollback_rel_end =
grid_row_abs_to_sb(term->grid, term->rows, six_end);
/* We should never generate scrollback wrapping sixels */
xassert(six_end < term->grid->num_rows);
if (six_scrollback_rel_end < scrollback_rel_start) {
/* All remaining sixels are *before* our rectangle */
would_have_breaked = true;
break;
}
#if defined(_DEBUG)
pixman_region32_t six_rect;
pixman_region32_init_rect(&six_rect, six->pos.col, six->pos.row, six->cols, six->rows);
pixman_region32_t intersection;
pixman_region32_init(&intersection);
pixman_region32_intersect(&intersection, &six_rect, &overwrite_rect);
const bool collides = pixman_region32_not_empty(&intersection);
#else
const bool UNUSED collides = false;
#endif
if ((start <= six_start && end >= six_start) || /* Crosses sixel start boundary */
(start <= six_end && end >= six_end) || /* Crosses sixel end boundary */
(start >= six_start && end <= six_end)) /* Fully within sixel range */
{
const int col_start = six->pos.col;
const int col_end = six->pos.col + six->cols - 1;
if ((col <= col_start && col + width - 1 >= col_start) ||
(col <= col_end && col + width - 1 >= col_end) ||
(col >= col_start && col + width - 1 <= col_end))
{
xassert(!would_have_breaked);
struct sixel to_be_erased = *six;
tll_remove(term->grid->sixel_images, it);
sixel: add support for overlapping sixels Writing a sixel on top of an already existing sixel currently has the following limitations in foot: * The parts of the first sixel that is covered by the new sixel are removed, completely. Even if the new sixel has transparent areas. I.e. writing a transparent sixel on top of another sixel *replaces* the first sixel with the new sixel, instead of layering them on top of each other. * The second sixel erases the first sixel cell-wise. That is, a sixel whose size isn’t a multiple of the cell dimensions will leave unsightly holes in the first sixel. This patch takes care of both issues. The first one is actually the easiest one: all we need to do is calculate the intersection, and blend the two images. To keep things relatively simple, we use the pixman image from the *new* image, and use the ‘OVER_REVERSE’ operation to blend the new image over the old one. That is, the old image is still split into four tiles (top, left, right, bottom), just like before. But instead of throwing away the fifth middle tile, we blend it with the new image. As an optimization, this is only done if the new image has transparency (P1=1). The second problem is solved by detecting when we’re erasing an area from the second image that is larger than the new image. In this case, we enlarge the new image, and copy the old image into the new one. Finally, when we enlarge the new image, there may be areas in the new image that is *not* covered by the old image. These areas are made transparent. The end result is: * Each cell is covered by at *most* 1 sixel image. I.e. the total numbers of sixels are finite. This is important for the ‘mpv --vo=sixel’ use case - we don’t want to end up with thousands of sixels layered on top of each other. * Writing an opaque sixel on top of another sixel has _almost_ zero performance impact. Especially if the two sixels have the same size, so that we don’t have to resize the new image. Again, important for the ‘mpv --vo=sixel’ use case. Closes #562
2021-06-06 21:38:31 +02:00
sixel_overwrite(term, &to_be_erased, start, col, height, width,
pix, opaque);
sixel_erase(term, &to_be_erased);
} else
xassert(!collides);
} else
xassert(!collides);
#if defined(_DEBUG)
pixman_region32_fini(&intersection);
pixman_region32_fini(&six_rect);
#endif
}
#if defined(_DEBUG)
pixman_region32_fini(&overwrite_rect);
#endif
}
void
2020-06-27 15:29:47 +02:00
sixel_overwrite_by_rectangle(
struct terminal *term, int row, int col, int height, int width)
{
if (likely(tll_length(term->grid->sixel_images) == 0))
return;
const int start = (term->grid->offset + row) & (term->grid->num_rows - 1);
const int end = (start + height - 1) & (term->grid->num_rows - 1);
const bool wraps = end < start;
if (wraps) {
int rows_to_wrap_around = term->grid->num_rows - start;
xassert(height - rows_to_wrap_around > 0);
sixel: add support for overlapping sixels Writing a sixel on top of an already existing sixel currently has the following limitations in foot: * The parts of the first sixel that is covered by the new sixel are removed, completely. Even if the new sixel has transparent areas. I.e. writing a transparent sixel on top of another sixel *replaces* the first sixel with the new sixel, instead of layering them on top of each other. * The second sixel erases the first sixel cell-wise. That is, a sixel whose size isn’t a multiple of the cell dimensions will leave unsightly holes in the first sixel. This patch takes care of both issues. The first one is actually the easiest one: all we need to do is calculate the intersection, and blend the two images. To keep things relatively simple, we use the pixman image from the *new* image, and use the ‘OVER_REVERSE’ operation to blend the new image over the old one. That is, the old image is still split into four tiles (top, left, right, bottom), just like before. But instead of throwing away the fifth middle tile, we blend it with the new image. As an optimization, this is only done if the new image has transparency (P1=1). The second problem is solved by detecting when we’re erasing an area from the second image that is larger than the new image. In this case, we enlarge the new image, and copy the old image into the new one. Finally, when we enlarge the new image, there may be areas in the new image that is *not* covered by the old image. These areas are made transparent. The end result is: * Each cell is covered by at *most* 1 sixel image. I.e. the total numbers of sixels are finite. This is important for the ‘mpv --vo=sixel’ use case - we don’t want to end up with thousands of sixels layered on top of each other. * Writing an opaque sixel on top of another sixel has _almost_ zero performance impact. Especially if the two sixels have the same size, so that we don’t have to resize the new image. Again, important for the ‘mpv --vo=sixel’ use case. Closes #562
2021-06-06 21:38:31 +02:00
_sixel_overwrite_by_rectangle(term, start, col, rows_to_wrap_around, width, NULL, NULL);
_sixel_overwrite_by_rectangle(term, 0, col, height - rows_to_wrap_around, width, NULL, NULL);
} else
sixel: add support for overlapping sixels Writing a sixel on top of an already existing sixel currently has the following limitations in foot: * The parts of the first sixel that is covered by the new sixel are removed, completely. Even if the new sixel has transparent areas. I.e. writing a transparent sixel on top of another sixel *replaces* the first sixel with the new sixel, instead of layering them on top of each other. * The second sixel erases the first sixel cell-wise. That is, a sixel whose size isn’t a multiple of the cell dimensions will leave unsightly holes in the first sixel. This patch takes care of both issues. The first one is actually the easiest one: all we need to do is calculate the intersection, and blend the two images. To keep things relatively simple, we use the pixman image from the *new* image, and use the ‘OVER_REVERSE’ operation to blend the new image over the old one. That is, the old image is still split into four tiles (top, left, right, bottom), just like before. But instead of throwing away the fifth middle tile, we blend it with the new image. As an optimization, this is only done if the new image has transparency (P1=1). The second problem is solved by detecting when we’re erasing an area from the second image that is larger than the new image. In this case, we enlarge the new image, and copy the old image into the new one. Finally, when we enlarge the new image, there may be areas in the new image that is *not* covered by the old image. These areas are made transparent. The end result is: * Each cell is covered by at *most* 1 sixel image. I.e. the total numbers of sixels are finite. This is important for the ‘mpv --vo=sixel’ use case - we don’t want to end up with thousands of sixels layered on top of each other. * Writing an opaque sixel on top of another sixel has _almost_ zero performance impact. Especially if the two sixels have the same size, so that we don’t have to resize the new image. Again, important for the ‘mpv --vo=sixel’ use case. Closes #562
2021-06-06 21:38:31 +02:00
_sixel_overwrite_by_rectangle(term, start, col, height, width, NULL, NULL);
term_update_ascii_printer(term);
}
/* Row numbers are relative to grid offset */
void
sixel_overwrite_by_row(struct terminal *term, int _row, int col, int width)
{
xassert(col >= 0);
xassert(_row >= 0);
xassert(_row < term->rows);
xassert(col >= 0);
xassert(col < term->grid->num_cols);
if (likely(tll_length(term->grid->sixel_images) == 0))
return;
if (col + width > term->grid->num_cols)
width = term->grid->num_cols - col;
const int row = (term->grid->offset + _row) & (term->grid->num_rows - 1);
const int scrollback_rel_row = grid_row_abs_to_sb(term->grid, term->rows, row);
tll_foreach(term->grid->sixel_images, it) {
struct sixel *six = &it->item;
const int six_start = six->pos.row;
const int six_end = (six_start + six->rows - 1) & (term->grid->num_rows - 1);
/* We should never generate scrollback wrapping sixels */
xassert(six_end >= six_start);
const int six_scrollback_rel_end =
grid_row_abs_to_sb(term->grid, term->rows, six_end);
if (six_scrollback_rel_end < scrollback_rel_row) {
/* All remaining sixels are *before* "our" row */
break;
}
if (row >= six_start && row <= six_end) {
const int col_start = six->pos.col;
const int col_end = six->pos.col + six->cols - 1;
if ((col <= col_start && col + width - 1 >= col_start) ||
(col <= col_end && col + width - 1 >= col_end) ||
(col >= col_start && col + width - 1 <= col_end))
{
struct sixel to_be_erased = *six;
tll_remove(term->grid->sixel_images, it);
sixel: add support for overlapping sixels Writing a sixel on top of an already existing sixel currently has the following limitations in foot: * The parts of the first sixel that is covered by the new sixel are removed, completely. Even if the new sixel has transparent areas. I.e. writing a transparent sixel on top of another sixel *replaces* the first sixel with the new sixel, instead of layering them on top of each other. * The second sixel erases the first sixel cell-wise. That is, a sixel whose size isn’t a multiple of the cell dimensions will leave unsightly holes in the first sixel. This patch takes care of both issues. The first one is actually the easiest one: all we need to do is calculate the intersection, and blend the two images. To keep things relatively simple, we use the pixman image from the *new* image, and use the ‘OVER_REVERSE’ operation to blend the new image over the old one. That is, the old image is still split into four tiles (top, left, right, bottom), just like before. But instead of throwing away the fifth middle tile, we blend it with the new image. As an optimization, this is only done if the new image has transparency (P1=1). The second problem is solved by detecting when we’re erasing an area from the second image that is larger than the new image. In this case, we enlarge the new image, and copy the old image into the new one. Finally, when we enlarge the new image, there may be areas in the new image that is *not* covered by the old image. These areas are made transparent. The end result is: * Each cell is covered by at *most* 1 sixel image. I.e. the total numbers of sixels are finite. This is important for the ‘mpv --vo=sixel’ use case - we don’t want to end up with thousands of sixels layered on top of each other. * Writing an opaque sixel on top of another sixel has _almost_ zero performance impact. Especially if the two sixels have the same size, so that we don’t have to resize the new image. Again, important for the ‘mpv --vo=sixel’ use case. Closes #562
2021-06-06 21:38:31 +02:00
sixel_overwrite(term, &to_be_erased, row, col, 1, width, NULL, NULL);
sixel_erase(term, &to_be_erased);
}
}
}
term_update_ascii_printer(term);
}
void
sixel_overwrite_at_cursor(struct terminal *term, int width)
{
if (likely(tll_length(term->grid->sixel_images) == 0))
return;
2020-06-27 15:29:47 +02:00
sixel_overwrite_by_row(
term, term->grid->cursor.point.row, term->grid->cursor.point.col, width);
}
void
sixel_cell_size_changed(struct terminal *term)
{
struct grid *g = term->grid;
term->grid = &term->normal;
tll_foreach(term->normal.sixel_images, it) {
struct sixel *six = &it->item;
six->rows = (six->height + term->cell_height - 1) / term->cell_height;
six->cols = (six->width + term->cell_width - 1) / term->cell_width;
}
term->grid = &term->alt;
tll_foreach(term->alt.sixel_images, it) {
struct sixel *six = &it->item;
six->rows = (six->height + term->cell_height - 1) / term->cell_height;
six->cols = (six->width + term->cell_width - 1) / term->cell_width;
}
term->grid = g;
}
void
sixel_reflow_grid(struct terminal *term, struct grid *grid)
{
/* Meh - the sixel functions we call use term->grid... */
struct grid *active_grid = term->grid;
term->grid = grid;
/* Need the “real” list to be empty from the beginning */
tll(struct sixel) copy = tll_init();
tll_foreach(grid->sixel_images, it)
tll_push_back(copy, it->item);
tll_free(grid->sixel_images);
tll_rforeach(copy, it) {
struct sixel *six = &it->item;
int start = six->pos.row;
int end = (start + six->rows - 1) & (grid->num_rows - 1);
if (end < start) {
/* Crosses scrollback wrap-around */
/* TODO: split image */
sixel_destroy(six);
continue;
}
if (six->rows > grid->num_rows) {
/* Image too large */
/* TODO: keep bottom part? */
sixel_destroy(six);
continue;
}
/* Drop sixels that now cross the current scrollback end
* border. This is similar to a sixel that have been
* scrolled out */
/* TODO: should be possible to optimize this */
bool sixel_destroyed = false;
int last_row = -1;
for (int j = 0; j < six->rows; j++) {
int row_no = grid_row_abs_to_sb(
term->grid, term->rows, six->pos.row + j);
if (last_row != -1 && last_row >= row_no) {
sixel_destroy(six);
sixel_destroyed = true;
break;
}
last_row = row_no;
}
if (sixel_destroyed) {
LOG_WARN("destroyed sixel that now crossed history");
continue;
}
sixel: add support for overlapping sixels Writing a sixel on top of an already existing sixel currently has the following limitations in foot: * The parts of the first sixel that is covered by the new sixel are removed, completely. Even if the new sixel has transparent areas. I.e. writing a transparent sixel on top of another sixel *replaces* the first sixel with the new sixel, instead of layering them on top of each other. * The second sixel erases the first sixel cell-wise. That is, a sixel whose size isn’t a multiple of the cell dimensions will leave unsightly holes in the first sixel. This patch takes care of both issues. The first one is actually the easiest one: all we need to do is calculate the intersection, and blend the two images. To keep things relatively simple, we use the pixman image from the *new* image, and use the ‘OVER_REVERSE’ operation to blend the new image over the old one. That is, the old image is still split into four tiles (top, left, right, bottom), just like before. But instead of throwing away the fifth middle tile, we blend it with the new image. As an optimization, this is only done if the new image has transparency (P1=1). The second problem is solved by detecting when we’re erasing an area from the second image that is larger than the new image. In this case, we enlarge the new image, and copy the old image into the new one. Finally, when we enlarge the new image, there may be areas in the new image that is *not* covered by the old image. These areas are made transparent. The end result is: * Each cell is covered by at *most* 1 sixel image. I.e. the total numbers of sixels are finite. This is important for the ‘mpv --vo=sixel’ use case - we don’t want to end up with thousands of sixels layered on top of each other. * Writing an opaque sixel on top of another sixel has _almost_ zero performance impact. Especially if the two sixels have the same size, so that we don’t have to resize the new image. Again, important for the ‘mpv --vo=sixel’ use case. Closes #562
2021-06-06 21:38:31 +02:00
/* Sixels that didnt overlap may now do so, which isnt
* allowed of course */
_sixel_overwrite_by_rectangle(
term, six->pos.row, six->pos.col, six->rows, six->cols,
&it->item.pix, &it->item.opaque);
if (it->item.data != pixman_image_get_data(it->item.pix)) {
it->item.data = pixman_image_get_data(it->item.pix);
it->item.width = pixman_image_get_width(it->item.pix);
it->item.height = pixman_image_get_height(it->item.pix);
it->item.cols = (it->item.width + term->cell_width - 1) / term->cell_width;
it->item.rows = (it->item.height + term->cell_height - 1) / term->cell_height;
}
sixel_insert(term, it->item);
}
tll_free(copy);
term->grid = active_grid;
}
void
sixel_reflow(struct terminal *term)
{
for (size_t i = 0; i < 2; i++) {
struct grid *grid = i == 0 ? &term->normal : &term->alt;
sixel_reflow_grid(term, grid);
}
}
void
sixel_unhook(struct terminal *term)
{
int pixel_row_idx = 0;
int pixel_rows_left = term->sixel.image.height;
const int stride = term->sixel.image.width * sizeof(uint32_t);
/*
* When sixel scrolling is enabled (the default), sixels behave
* pretty much like normal output; the sixel starts at the current
* cursor position and the cursor is moved to a point after the
* sixel.
*
* Furthermore, if the sixel reaches the bottom of the scrolling
* region, the terminal content is scrolled.
*
* When scrolling is disabled, sixels always start at (0,0), the
* cursor is not moved at all, and the terminal content never
* scrolls.
*/
const bool do_scroll = term->sixel.scrolling;
/* Number of rows we're allowed to use.
*
* When scrolling is enabled, we always allow the entire sixel to
* be emitted.
*
* When disabled, only the number of screen rows may be used. */
int rows_avail = do_scroll
? (term->sixel.image.height + term->cell_height - 1) / term->cell_height
: term->scroll_region.end;
/* Initial sixel coordinates */
int start_row = do_scroll ? term->grid->cursor.point.row : 0;
const int start_col = do_scroll ? term->grid->cursor.point.col : 0;
/* Total number of rows needed by image (+ optional newline at the end) */
const int rows_needed =
(term->sixel.image.height + term->cell_height - 1) / term->cell_height +
(term->sixel.cursor_right_of_graphics ? 0 : 1);
bool free_image_data = true;
/* We do not allow sixels to cross the scrollback wrap-around, as
* this makes intersection calculations much more complicated */
while (pixel_rows_left > 0 &&
rows_avail > 0 &&
rows_needed <= term->grid->num_rows)
{
const int cur_row = (term->grid->offset + start_row) & (term->grid->num_rows - 1);
const int rows_left_until_wrap_around = term->grid->num_rows - cur_row;
const int usable_rows = min(rows_avail, rows_left_until_wrap_around);
const int pixel_rows_avail = usable_rows * term->cell_height;
const int width = term->sixel.image.width;
const int height = min(pixel_rows_left, pixel_rows_avail);
uint32_t *img_data;
if (pixel_row_idx == 0 && height == pixel_rows_left) {
/* Entire image will be emitted as a single chunk - reuse
* the source buffer */
img_data = term->sixel.image.data;
free_image_data = false;
} else {
xassert(free_image_data);
img_data = xmalloc(height * stride);
memcpy(
img_data,
&((uint8_t *)term->sixel.image.data)[pixel_row_idx * stride],
height * stride);
}
struct sixel image = {
.data = img_data,
.width = width,
.height = height,
.rows = (height + term->cell_height - 1) / term->cell_height,
.cols = (width + term->cell_width - 1) / term->cell_width,
.pos = (struct coord){start_col, cur_row},
.opaque = !term->sixel.transparent_bg,
};
xassert(image.rows <= term->grid->num_rows);
xassert(image.pos.row + image.rows - 1 < term->grid->num_rows);
LOG_DBG("generating %s %dx%d pixman image at %d-%d",
image.opaque ? "opaque" : "transparent",
2020-10-02 20:54:45 +02:00
image.width, image.height,
image.pos.row, image.pos.row + image.rows);
image.pix = pixman_image_create_bits_no_clear(
sixel: add support for overlapping sixels Writing a sixel on top of an already existing sixel currently has the following limitations in foot: * The parts of the first sixel that is covered by the new sixel are removed, completely. Even if the new sixel has transparent areas. I.e. writing a transparent sixel on top of another sixel *replaces* the first sixel with the new sixel, instead of layering them on top of each other. * The second sixel erases the first sixel cell-wise. That is, a sixel whose size isn’t a multiple of the cell dimensions will leave unsightly holes in the first sixel. This patch takes care of both issues. The first one is actually the easiest one: all we need to do is calculate the intersection, and blend the two images. To keep things relatively simple, we use the pixman image from the *new* image, and use the ‘OVER_REVERSE’ operation to blend the new image over the old one. That is, the old image is still split into four tiles (top, left, right, bottom), just like before. But instead of throwing away the fifth middle tile, we blend it with the new image. As an optimization, this is only done if the new image has transparency (P1=1). The second problem is solved by detecting when we’re erasing an area from the second image that is larger than the new image. In this case, we enlarge the new image, and copy the old image into the new one. Finally, when we enlarge the new image, there may be areas in the new image that is *not* covered by the old image. These areas are made transparent. The end result is: * Each cell is covered by at *most* 1 sixel image. I.e. the total numbers of sixels are finite. This is important for the ‘mpv --vo=sixel’ use case - we don’t want to end up with thousands of sixels layered on top of each other. * Writing an opaque sixel on top of another sixel has _almost_ zero performance impact. Especially if the two sixels have the same size, so that we don’t have to resize the new image. Again, important for the ‘mpv --vo=sixel’ use case. Closes #562
2021-06-06 21:38:31 +02:00
PIXMAN_a8r8g8b8, image.width, image.height, img_data, stride);
pixel_row_idx += height;
pixel_rows_left -= height;
rows_avail -= image.rows;
if (do_scroll) {
/* Yes, truncate last row. This matches XTerms, and VT382s behavior */
const int linefeed_count = (image.height - 6 * term->sixel.pan + 1) / term->cell_height;
for (size_t i = 0; i < linefeed_count; i++)
term_linefeed(term);
/* Position text cursor if this is the last image chunk */
if (rows_avail == 0) {
term_cursor_to(
term,
term->grid->cursor.point.row,
(term->sixel.cursor_right_of_graphics
? min(image.pos.col + image.cols, term->cols - 1)
: image.pos.col));
}
}
/* Dirty touched cells, and scroll terminal content if necessary */
for (size_t i = 0; i < image.rows; i++) {
struct row *row = term->grid->rows[cur_row + i];
row->dirty = true;
for (int col = image.pos.col;
col < min(image.pos.col + image.cols, term->cols);
col++)
{
row->cells[col].attrs.clean = 0;
}
}
_sixel_overwrite_by_rectangle(
sixel: add support for overlapping sixels Writing a sixel on top of an already existing sixel currently has the following limitations in foot: * The parts of the first sixel that is covered by the new sixel are removed, completely. Even if the new sixel has transparent areas. I.e. writing a transparent sixel on top of another sixel *replaces* the first sixel with the new sixel, instead of layering them on top of each other. * The second sixel erases the first sixel cell-wise. That is, a sixel whose size isn’t a multiple of the cell dimensions will leave unsightly holes in the first sixel. This patch takes care of both issues. The first one is actually the easiest one: all we need to do is calculate the intersection, and blend the two images. To keep things relatively simple, we use the pixman image from the *new* image, and use the ‘OVER_REVERSE’ operation to blend the new image over the old one. That is, the old image is still split into four tiles (top, left, right, bottom), just like before. But instead of throwing away the fifth middle tile, we blend it with the new image. As an optimization, this is only done if the new image has transparency (P1=1). The second problem is solved by detecting when we’re erasing an area from the second image that is larger than the new image. In this case, we enlarge the new image, and copy the old image into the new one. Finally, when we enlarge the new image, there may be areas in the new image that is *not* covered by the old image. These areas are made transparent. The end result is: * Each cell is covered by at *most* 1 sixel image. I.e. the total numbers of sixels are finite. This is important for the ‘mpv --vo=sixel’ use case - we don’t want to end up with thousands of sixels layered on top of each other. * Writing an opaque sixel on top of another sixel has _almost_ zero performance impact. Especially if the two sixels have the same size, so that we don’t have to resize the new image. Again, important for the ‘mpv --vo=sixel’ use case. Closes #562
2021-06-06 21:38:31 +02:00
term, image.pos.row, image.pos.col, image.rows, image.cols,
&image.pix, &image.opaque);
if (image.data != pixman_image_get_data(image.pix)) {
image.data = pixman_image_get_data(image.pix);
image.width = pixman_image_get_width(image.pix);
image.height = pixman_image_get_height(image.pix);
image.cols = (image.width + term->cell_width - 1) / term->cell_width;
image.rows = (image.height + term->cell_height - 1) / term->cell_height;
}
sixel_insert(term, image);
if (do_scroll)
start_row = term->grid->cursor.point.row;
else
start_row -= image.rows;
}
if (free_image_data)
free(term->sixel.image.data);
2020-02-22 10:46:35 +01:00
term->sixel.image.data = NULL;
term->sixel.image.width = 0;
term->sixel.image.height = 0;
term->sixel.pos = (struct coord){0, 0};
free(term->sixel.private_palette);
term->sixel.private_palette = NULL;
LOG_DBG("you now have %zu sixels in current grid",
tll_length(term->grid->sixel_images));
term_update_ascii_printer(term);
render_refresh(term);
}
static void
resize_horizontally(struct terminal *term, int new_width)
{
LOG_DBG("resizing image horizontally: %dx(%d) -> %dx(%d)",
term->sixel.image.width, term->sixel.image.height,
new_width, term->sixel.image.height);
if (unlikely(new_width > term->sixel.max_width)) {
LOG_WARN("maximum image dimensions exceeded, truncating");
new_width = term->sixel.max_width;
}
if (unlikely(term->sixel.image.width == new_width))
return;
uint32_t *old_data = term->sixel.image.data;
const int old_width = term->sixel.image.width;
const int height = term->sixel.image.height;
const int sixel_row_height = 6 * term->sixel.pan;
int alloc_height = (height + sixel_row_height - 1) / sixel_row_height * sixel_row_height;
xassert(new_width > 0);
xassert(alloc_height > 0);
/* Width (and thus stride) change - need to allocate a new buffer */
uint32_t *new_data = xmalloc(new_width * alloc_height * sizeof(uint32_t));
uint32_t bg = term->sixel.default_bg;
/* Copy old rows, and initialize new columns to background color */
for (int r = 0; r < height; r++) {
memcpy(&new_data[r * new_width],
&old_data[r * old_width],
old_width * sizeof(uint32_t));
for (int c = old_width; c < new_width; c++)
new_data[r * new_width + c] = bg;
}
free(old_data);
term->sixel.image.data = new_data;
term->sixel.image.width = new_width;
term->sixel.row_byte_ofs = term->sixel.pos.row * new_width;
}
static bool
resize_vertically(struct terminal *term, int new_height)
{
LOG_DBG("resizing image vertically: (%d)x%d -> (%d)x%d",
term->sixel.image.width, term->sixel.image.height,
term->sixel.image.width, new_height);
if (unlikely(new_height > term->sixel.max_height)) {
LOG_WARN("maximum image dimensions reached");
return false;
}
uint32_t *old_data = term->sixel.image.data;
const int width = term->sixel.image.width;
const int old_height = term->sixel.image.height;
int alloc_height = (new_height + 6 - 1) / 6 * 6;
xassert(new_height > 0);
if (unlikely(width == 0)) {
xassert(term->sixel.image.data == NULL);
term->sixel.image.height = new_height;
return true;
}
uint32_t *new_data = realloc(
old_data, width * alloc_height * sizeof(uint32_t));
if (new_data == NULL) {
LOG_ERRNO("failed to reallocate sixel image buffer");
return false;
}
uint32_t bg = term->sixel.default_bg;
/* Initialize new rows to background color */
for (int r = old_height; r < new_height; r++) {
for (int c = 0; c < width; c++)
new_data[r * width + c] = bg;
}
term->sixel.image.data = new_data;
term->sixel.image.height = new_height;
return true;
}
2020-02-22 10:46:35 +01:00
static bool
resize(struct terminal *term, int new_width, int new_height)
{
LOG_DBG("resizing image: %dx%d -> %dx%d",
term->sixel.image.width, term->sixel.image.height,
new_width, new_height);
if (unlikely(new_width > term->sixel.max_width)) {
LOG_WARN("maximum image width exceeded, truncating");
new_width = term->sixel.max_width;
}
if (unlikely(new_height > term->sixel.max_height)) {
LOG_WARN("maximum image height exceeded, truncating");
new_height = term->sixel.max_height;
}
uint32_t *old_data = term->sixel.image.data;
const int old_width = term->sixel.image.width;
const int old_height = term->sixel.image.height;
int alloc_new_width = new_width;
int alloc_new_height = (new_height + 6 - 1) / 6 * 6;
xassert(alloc_new_height >= new_height);
xassert(alloc_new_height - new_height < 6);
2020-02-22 10:46:35 +01:00
uint32_t *new_data = NULL;
uint32_t bg = term->sixel.default_bg;
2020-02-22 10:46:35 +01:00
if (new_width == old_width) {
/* Width (and thus stride) is the same, so we can simply
* re-alloc the existing buffer */
new_data = realloc(old_data, alloc_new_width * alloc_new_height * sizeof(uint32_t));
2020-02-22 10:46:35 +01:00
if (new_data == NULL) {
LOG_ERRNO("failed to reallocate sixel image buffer");
return false;
}
xassert(new_height > old_height);
2020-02-22 10:46:35 +01:00
} else {
/* Width (and thus stride) change - need to allocate a new buffer */
xassert(new_width > old_width);
new_data = xmalloc(alloc_new_width * alloc_new_height * sizeof(uint32_t));
2020-02-22 10:46:35 +01:00
/* Copy old rows, and initialize new columns to background color */
for (int r = 0; r < min(old_height, new_height); r++) {
2020-02-22 10:46:35 +01:00
memcpy(&new_data[r * new_width], &old_data[r * old_width], old_width * sizeof(uint32_t));
for (int c = old_width; c < new_width; c++)
new_data[r * new_width + c] = bg;
}
2020-02-22 10:46:35 +01:00
free(old_data);
}
/* Initialize new rows to background color */
for (int r = old_height; r < new_height; r++) {
for (int c = 0; c < new_width; c++)
new_data[r * new_width + c] = bg;
}
xassert(new_data != NULL);
2020-02-22 10:46:35 +01:00
term->sixel.image.data = new_data;
term->sixel.image.width = new_width;
term->sixel.image.height = new_height;
term->sixel.row_byte_ofs = term->sixel.pos.row * new_width;
2020-02-22 10:46:35 +01:00
return true;
}
static void
sixel_add(struct terminal *term, int col, int width, uint32_t color, uint8_t sixel)
{
xassert(term->sixel.pos.col < term->sixel.image.width);
xassert(term->sixel.pos.row < term->sixel.image.height);
size_t ofs = term->sixel.row_byte_ofs + col;
uint32_t *data = &term->sixel.image.data[ofs];
for (int i = 0; i < 6 * term->sixel.pan; i++, sixel >>= 1, data += width) {
if (sixel & 1) {
*data = color;
}
}
xassert(sixel == 0);
}
static void
sixel_add_many(struct terminal *term, uint8_t c, unsigned count)
{
int col = term->sixel.pos.col;
int width = term->sixel.image.width;
count *= term->sixel.pad;
if (unlikely(col + count - 1 >= width)) {
resize_horizontally(term, col + count);
width = term->sixel.image.width;
count = min(count, max(width - col, 0));
}
uint32_t color = term->sixel.color;
for (unsigned i = 0; i < count; i++, col++)
sixel_add(term, col, width, color, c);
term->sixel.pos.col = col;
}
static void
decsixel(struct terminal *term, uint8_t c)
{
switch (c) {
case '"':
term->sixel.state = SIXEL_DECGRA;
term->sixel.param = 0;
term->sixel.param_idx = 0;
break;
case '!':
term->sixel.state = SIXEL_DECGRI;
term->sixel.param = 0;
term->sixel.param_idx = 0;
break;
case '#':
term->sixel.state = SIXEL_DECGCI;
term->sixel.color_idx = 0;
term->sixel.param = 0;
term->sixel.param_idx = 0;
break;
case '$':
if (likely(term->sixel.pos.col <= term->sixel.max_width)) {
/*
* We set, and keep, col outside the image boundary when
* weve reached the maximum image height, to avoid also
* having to check the row vs image height in the common
* path in sixel_add().
*/
term->sixel.pos.col = 0;
}
break;
case '-':
term->sixel.pos.row += 6 * term->sixel.pan;
term->sixel.pos.col = 0;
term->sixel.row_byte_ofs += term->sixel.image.width * 6 * term->sixel.pan;
if (term->sixel.pos.row >= term->sixel.image.height) {
if (!resize_vertically(term, term->sixel.pos.row + 6 * term->sixel.pan))
term->sixel.pos.col = term->sixel.max_width + 1 * term->sixel.pad;
}
break;
case '?': case '@': case 'A': case 'B': case 'C': case 'D': case 'E':
case 'F': case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R': case 'S':
case 'T': case 'U': case 'V': case 'W': case 'X': case 'Y': case 'Z':
case '[': case '\\': case ']': case '^': case '_': case '`': case 'a':
case 'b': case 'c': case 'd': case 'e': case 'f': case 'g': case 'h':
case 'i': case 'j': case 'k': case 'l': case 'm': case 'n': case 'o':
case 'p': case 'q': case 'r': case 's': case 't': case 'u': case 'v':
case 'w': case 'x': case 'y': case 'z': case '{': case '|': case '}':
case '~':
sixel_add_many(term, c - 63, 1);
break;
case ' ':
case '\n':
case '\r':
break;
default:
2020-06-09 17:34:04 +02:00
LOG_WARN("invalid sixel character: '%c' at idx=%zu", c, count);
break;
}
}
static void
decgra(struct terminal *term, uint8_t c)
{
switch (c) {
case '0': case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
term->sixel.param *= 10;
term->sixel.param += c - '0';
break;
2020-02-22 10:46:35 +01:00
case ';':
if (term->sixel.param_idx < ALEN(term->sixel.params))
term->sixel.params[term->sixel.param_idx++] = term->sixel.param;
term->sixel.param = 0;
break;
default: {
if (term->sixel.param_idx < ALEN(term->sixel.params))
term->sixel.params[term->sixel.param_idx++] = term->sixel.param;
int nparams = term->sixel.param_idx;
unsigned pan = nparams > 0 ? term->sixel.params[0] : 0;
unsigned pad = nparams > 1 ? term->sixel.params[1] : 0;
unsigned ph = nparams > 2 ? term->sixel.params[2] : 0;
unsigned pv = nparams > 3 ? term->sixel.params[3] : 0;
pan = pan > 0 ? pan : 1;
pad = pad > 0 ? pad : 1;
pv *= pan;
ph *= pad;
term->sixel.pan = pan;
term->sixel.pad = pad;
LOG_DBG("pan=%u, pad=%u (aspect ratio = %u), size=%ux%u",
pan, pad, pan / pad, ph, pv);
if (ph >= term->sixel.image.height && pv >= term->sixel.image.width &&
ph <= term->sixel.max_height && pv <= term->sixel.max_width)
{
resize(term, ph, pv);
}
term->sixel.state = SIXEL_DECSIXEL;
decsixel(term, c);
break;
}
}
}
static void
decgri(struct terminal *term, uint8_t c)
{
switch (c) {
case '0': case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
term->sixel.param *= 10;
term->sixel.param += c - '0';
break;
case '?': case '@': case 'A': case 'B': case 'C': case 'D': case 'E':
case 'F': case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R': case 'S':
case 'T': case 'U': case 'V': case 'W': case 'X': case 'Y': case 'Z':
case '[': case '\\': case ']': case '^': case '_': case '`': case 'a':
case 'b': case 'c': case 'd': case 'e': case 'f': case 'g': case 'h':
case 'i': case 'j': case 'k': case 'l': case 'm': case 'n': case 'o':
case 'p': case 'q': case 'r': case 's': case 't': case 'u': case 'v':
case 'w': case 'x': case 'y': case 'z': case '{': case '|': case '}':
case '~': {
unsigned count = term->sixel.param;
if (likely(count > 0))
sixel_add_many(term, c - 63, count);
else if (unlikely(count == 0))
sixel_add_many(term, c - 63, 1);
term->sixel.state = SIXEL_DECSIXEL;
break;
}
default:
term->sixel.state = SIXEL_DECSIXEL;
sixel_put(term, c);
break;
}
}
static void
decgci(struct terminal *term, uint8_t c)
{
switch (c) {
case '0': case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
term->sixel.param *= 10;
term->sixel.param += c - '0';
break;
case ';':
if (term->sixel.param_idx < ALEN(term->sixel.params))
term->sixel.params[term->sixel.param_idx++] = term->sixel.param;
term->sixel.param = 0;
break;
default: {
if (term->sixel.param_idx < ALEN(term->sixel.params))
term->sixel.params[term->sixel.param_idx++] = term->sixel.param;
int nparams = term->sixel.param_idx;
if (nparams > 0)
term->sixel.color_idx = min(term->sixel.params[0], term->sixel.palette_size - 1);
if (nparams > 4) {
unsigned format = term->sixel.params[1];
int c1 = term->sixel.params[2];
int c2 = term->sixel.params[3];
int c3 = term->sixel.params[4];
switch (format) {
case 1: { /* HLS */
int hue = min(c1, 360);
int lum = min(c2, 100);
int sat = min(c3, 100);
/*
* Sixels HLS use the following primary color hues:
* blue: 0°
* red: 120°
* green: 240°
*
* While standard HSL uses:
* red: 0°
* green: 120°
* blue: 240°
*/
hue = (hue + 240) % 360;
uint32_t rgb = hsl_to_rgb(hue, sat, lum);
2020-02-22 21:18:55 +01:00
LOG_DBG("setting palette #%d = HLS %hhu/%hhu/%hhu (0x%06x)",
term->sixel.color_idx, hue, lum, sat, rgb);
term->sixel.palette[term->sixel.color_idx] = 0xffu << 24 | rgb;
break;
}
case 2: { /* RGB */
uint8_t r = 255 * min(c1, 100) / 100;
uint8_t g = 255 * min(c2, 100) / 100;
uint8_t b = 255 * min(c3, 100) / 100;
LOG_DBG("setting palette #%d = RGB %hhu/%hhu/%hhu",
term->sixel.color_idx, r, g, b);
term->sixel.palette[term->sixel.color_idx] =
0xffu << 24 | r << 16 | g << 8 | b;
break;
}
}
} else
term->sixel.color = term->sixel.palette[term->sixel.color_idx];
term->sixel.state = SIXEL_DECSIXEL;
decsixel(term, c);
break;
}
}
}
void
sixel_put(struct terminal *term, uint8_t c)
{
switch (term->sixel.state) {
case SIXEL_DECSIXEL: decsixel(term, c); break;
case SIXEL_DECGRA: decgra(term, c); break;
case SIXEL_DECGRI: decgri(term, c); break;
case SIXEL_DECGCI: decgci(term, c); break;
}
count++;
}
void
sixel_colors_report_current(struct terminal *term)
{
char reply[24];
size_t n = xsnprintf(reply, sizeof(reply), "\033[?1;0;%uS", term->sixel.palette_size);
term_to_slave(term, reply, n);
LOG_DBG("query response for current color count: %u", term->sixel.palette_size);
}
void
sixel_colors_reset(struct terminal *term)
{
LOG_DBG("sixel palette size reset to %u", SIXEL_MAX_COLORS);
free(term->sixel.palette);
term->sixel.palette = NULL;
term->sixel.palette_size = SIXEL_MAX_COLORS;
sixel_colors_report_current(term);
}
void
sixel_colors_set(struct terminal *term, unsigned count)
{
unsigned new_palette_size = min(max(2, count), SIXEL_MAX_COLORS);
LOG_DBG("sixel palette size set to %u", new_palette_size);
free(term->sixel.private_palette);
free(term->sixel.shared_palette);
term->sixel.private_palette = NULL;
term->sixel.shared_palette = NULL;
term->sixel.palette_size = new_palette_size;
sixel_colors_report_current(term);
}
void
sixel_colors_report_max(struct terminal *term)
{
char reply[24];
size_t n = xsnprintf(reply, sizeof(reply), "\033[?1;0;%uS", SIXEL_MAX_COLORS);
term_to_slave(term, reply, n);
LOG_DBG("query response for max color count: %u", SIXEL_MAX_COLORS);
}
void
sixel_geometry_report_current(struct terminal *term)
{
char reply[64];
size_t n = xsnprintf(reply, sizeof(reply), "\033[?2;0;%u;%uS",
min(term->cols * term->cell_width, term->sixel.max_width),
min(term->rows * term->cell_height, term->sixel.max_height));
term_to_slave(term, reply, n);
LOG_DBG("query response for current sixel geometry: %ux%u",
term->sixel.max_width, term->sixel.max_height);
}
void
sixel_geometry_reset(struct terminal *term)
{
LOG_DBG("sixel geometry reset to %ux%u", SIXEL_MAX_WIDTH, SIXEL_MAX_HEIGHT);
term->sixel.max_width = SIXEL_MAX_WIDTH;
term->sixel.max_height = SIXEL_MAX_HEIGHT;
sixel_geometry_report_current(term);
}
void
sixel_geometry_set(struct terminal *term, unsigned width, unsigned height)
{
LOG_DBG("sixel geometry set to %ux%u", width, height);
term->sixel.max_width = width;
term->sixel.max_height = height;
sixel_geometry_report_current(term);
}
void
sixel_geometry_report_max(struct terminal *term)
{
unsigned max_width = term->sixel.max_width;
unsigned max_height = term->sixel.max_height;
char reply[64];
size_t n = xsnprintf(reply, sizeof(reply), "\033[?2;0;%u;%uS", max_width, max_height);
term_to_slave(term, reply, n);
LOG_DBG("query response for max sixel geometry: %ux%u",
max_width, max_height);
}