WebSVN – Kolibri OS – Blame – /programs/develop/libraries/cairo/src/cairo-botor-scan-converter.c

Rev	Author	Line No.	Line
1892	serge	1	/*
		2	* Copyright © 2004 Carl Worth
		3	* Copyright © 2006 Red Hat, Inc.
		4	* Copyright © 2007 David Turner
		5	* Copyright © 2008 M Joonas Pihlaja
		6	* Copyright © 2008 Chris Wilson
		7	* Copyright © 2009 Intel Corporation
		8	*
		9	* This library is free software; you can redistribute it and/or
		10	* modify it either under the terms of the GNU Lesser General Public
		11	* License version 2.1 as published by the Free Software Foundation
		12	* (the "LGPL") or, at your option, under the terms of the Mozilla
		13	* Public License Version 1.1 (the "MPL"). If you do not alter this
		14	* notice, a recipient may use your version of this file under either
		15	* the MPL or the LGPL.
		16	*
		17	* You should have received a copy of the LGPL along with this library
		18	* in the file COPYING-LGPL-2.1; if not, write to the Free Software
		19	* Foundation, Inc., 51 Franklin Street, Suite 500, Boston, MA 02110-1335, USA
		20	* You should have received a copy of the MPL along with this library
		21	* in the file COPYING-MPL-1.1
		22	*
		23	* The contents of this file are subject to the Mozilla Public License
		24	* Version 1.1 (the "License"); you may not use this file except in
		25	* compliance with the License. You may obtain a copy of the License at
		26	* http://www.mozilla.org/MPL/
		27	*
		28	* This software is distributed on an "AS IS" basis, WITHOUT WARRANTY
		29	* OF ANY KIND, either express or implied. See the LGPL or the MPL for
		30	* the specific language governing rights and limitations.
		31	*
		32	* The Original Code is the cairo graphics library.
		33	*
		34	* The Initial Developer of the Original Code is Carl Worth
		35	*
		36	* Contributor(s):
		37	* Carl D. Worth
		38	* M Joonas Pihlaja
		39	* Chris Wilson
		40	*/
		41
		42	/* Provide definitions for standalone compilation */
		43	#include "cairoint.h"
		44
		45	#include "cairo-error-private.h"
		46	#include "cairo-list-private.h"
		47	#include "cairo-freelist-private.h"
		48	#include "cairo-combsort-private.h"
		49
		50	#include
		51
		52	#define STEP_X CAIRO_FIXED_ONE
		53	#define STEP_Y CAIRO_FIXED_ONE
		54	#define UNROLL3(x) x x x
		55
		56	#define STEP_XY (2STEP_XSTEP_Y) /* Unit area in the step. */
		57	#define AREA_TO_ALPHA(c) (((c)*255 + STEP_XY/2) / STEP_XY)
		58
		59	typedef struct _cairo_bo_intersect_ordinate {
		60	int32_t ordinate;
		61	enum { EXACT, INEXACT } exactness;
		62	} cairo_bo_intersect_ordinate_t;
		63
		64	typedef struct _cairo_bo_intersect_point {
		65	cairo_bo_intersect_ordinate_t x;
		66	cairo_bo_intersect_ordinate_t y;
		67	} cairo_bo_intersect_point_t;
		68
		69	struct quorem {
		70	cairo_fixed_t quo;
		71	cairo_fixed_t rem;
		72	};
		73
		74	struct run {
		75	struct run *next;
		76	int sign;
		77	cairo_fixed_t y;
		78	};
		79
		80	typedef struct edge {
		81	cairo_list_t link;
		82
		83	cairo_edge_t edge;
		84
		85	/* Current x coordinate and advancement.
		86	* Initialised to the x coordinate of the top of the
		87	* edge. The quotient is in cairo_fixed_t units and the
		88	* remainder is mod dy in cairo_fixed_t units.
		89	*/
		90	cairo_fixed_t dy;
		91	struct quorem x;
		92	struct quorem dxdy;
		93	struct quorem dxdy_full;
		94
		95	cairo_bool_t vertical;
		96	unsigned int flags;
		97
		98	int current_sign;
		99	struct run *runs;
		100	} edge_t;
		101
		102	enum {
		103	START = 0x1,
		104	STOP = 0x2,
		105	};
		106
		107	/* the parent is always given by index/2 */
		108	#define PQ_PARENT_INDEX(i) ((i) >> 1)
		109	#define PQ_FIRST_ENTRY 1
		110
		111	/* left and right children are index * 2 and (index * 2) +1 respectively */
		112	#define PQ_LEFT_CHILD_INDEX(i) ((i) << 1)
		113
		114	typedef enum {
		115	EVENT_TYPE_STOP,
		116	EVENT_TYPE_INTERSECTION,
		117	EVENT_TYPE_START
		118	} event_type_t;
		119
		120	typedef struct _event {
		121	cairo_fixed_t y;
		122	event_type_t type;
		123	} event_t;
		124
		125	typedef struct _start_event {
		126	cairo_fixed_t y;
		127	event_type_t type;
		128	edge_t *edge;
		129	} start_event_t;
		130
		131	typedef struct _queue_event {
		132	cairo_fixed_t y;
		133	event_type_t type;
		134	edge_t *e1;
		135	edge_t *e2;
		136	} queue_event_t;
		137
		138	typedef struct _pqueue {
		139	int size, max_size;
		140
		141	event_t **elements;
		142	event_t *elements_embedded[1024];
		143	} pqueue_t;
		144
		145	struct cell {
		146	struct cell *prev;
		147	struct cell *next;
		148	int x;
		149	int uncovered_area;
		150	int covered_height;
		151	};
		152
		153	typedef struct _sweep_line {
		154	cairo_list_t active;
		155	cairo_list_t stopped;
		156	cairo_list_t *insert_cursor;
		157	cairo_bool_t is_vertical;
		158
		159	cairo_fixed_t current_row;
		160	cairo_fixed_t current_subrow;
		161
		162	struct coverage {
		163	struct cell head;
		164	struct cell tail;
		165
		166	struct cell *cursor;
		167	int count;
		168
		169	cairo_freepool_t pool;
		170	} coverage;
		171
		172	struct event_queue {
		173	pqueue_t pq;
		174	event_t **start_events;
		175
		176	cairo_freepool_t pool;
		177	} queue;
		178
		179	cairo_freepool_t runs;
		180
		181	jmp_buf unwind;
		182	} sweep_line_t;
		183
		184	cairo_always_inline static struct quorem
		185	floored_divrem (int a, int b)
		186	{
		187	struct quorem qr;
		188	qr.quo = a/b;
		189	qr.rem = a%b;
		190	if ((a^b)<0 && qr.rem) {
		191	qr.quo--;
		192	qr.rem += b;
		193	}
		194	return qr;
		195	}
		196
		197	static struct quorem
		198	floored_muldivrem(int x, int a, int b)
		199	{
		200	struct quorem qr;
		201	long long xa = (long long)x*a;
		202	qr.quo = xa/b;
		203	qr.rem = xa%b;
		204	if ((xa>=0) != (b>=0) && qr.rem) {
		205	qr.quo--;
		206	qr.rem += b;
		207	}
		208	return qr;
		209	}
		210
		211	static cairo_fixed_t
		212	line_compute_intersection_x_for_y (const cairo_line_t *line,
		213	cairo_fixed_t y)
		214	{
		215	cairo_fixed_t x, dy;
		216
		217	if (y == line->p1.y)
		218	return line->p1.x;
		219	if (y == line->p2.y)
		220	return line->p2.x;
		221
		222	x = line->p1.x;
		223	dy = line->p2.y - line->p1.y;
		224	if (dy != 0) {
		225	x += _cairo_fixed_mul_div_floor (y - line->p1.y,
		226	line->p2.x - line->p1.x,
		227	dy);
		228	}
		229
		230	return x;
		231	}
		232
		233	/*
		234	* We need to compare the x-coordinates of a pair of lines for a particular y,
		235	* without loss of precision.
		236	*
		237	* The x-coordinate along an edge for a given y is:
		238	* X = A_x + (Y - A_y) * A_dx / A_dy
		239	*
		240	* So the inequality we wish to test is:
		241	* A_x + (Y - A_y) * A_dx / A_dy ∘ B_x + (Y - B_y) * B_dx / B_dy,
		242	* where ∘ is our inequality operator.
		243	*
		244	* By construction, we know that A_dy and B_dy (and (Y - A_y), (Y - B_y)) are
		245	* all positive, so we can rearrange it thus without causing a sign change:
		246	* A_dy * B_dy * (A_x - B_x) ∘ (Y - B_y) * B_dx * A_dy
		247	* - (Y - A_y) * A_dx * B_dy
		248	*
		249	* Given the assumption that all the deltas fit within 32 bits, we can compute
		250	* this comparison directly using 128 bit arithmetic. For certain, but common,
		251	* input we can reduce this down to a single 32 bit compare by inspecting the
		252	* deltas.
		253	*
		254	* (And put the burden of the work on developing fast 128 bit ops, which are
		255	* required throughout the tessellator.)
		256	*
		257	* See the similar discussion for _slope_compare().
		258	*/
		259	static int
		260	edges_compare_x_for_y_general (const cairo_edge_t *a,
		261	const cairo_edge_t *b,
		262	int32_t y)
		263	{
		264	/* XXX: We're assuming here that dx and dy will still fit in 32
		265	* bits. That's not true in general as there could be overflow. We
		266	* should prevent that before the tessellation algorithm
		267	* begins.
		268	*/
		269	int32_t dx;
		270	int32_t adx, ady;
		271	int32_t bdx, bdy;
		272	enum {
		273	HAVE_NONE = 0x0,
		274	HAVE_DX = 0x1,
		275	HAVE_ADX = 0x2,
		276	HAVE_DX_ADX = HAVE_DX \| HAVE_ADX,
		277	HAVE_BDX = 0x4,
		278	HAVE_DX_BDX = HAVE_DX \| HAVE_BDX,
		279	HAVE_ADX_BDX = HAVE_ADX \| HAVE_BDX,
		280	HAVE_ALL = HAVE_DX \| HAVE_ADX \| HAVE_BDX
		281	} have_dx_adx_bdx = HAVE_ALL;
		282
		283	/* don't bother solving for abscissa if the edges' bounding boxes
		284	* can be used to order them. */
		285	{
		286	int32_t amin, amax;
		287	int32_t bmin, bmax;
		288	if (a->line.p1.x < a->line.p2.x) {
		289	amin = a->line.p1.x;
		290	amax = a->line.p2.x;
		291	} else {
		292	amin = a->line.p2.x;
		293	amax = a->line.p1.x;
		294	}
		295	if (b->line.p1.x < b->line.p2.x) {
		296	bmin = b->line.p1.x;
		297	bmax = b->line.p2.x;
		298	} else {
		299	bmin = b->line.p2.x;
		300	bmax = b->line.p1.x;
		301	}
		302	if (amax < bmin) return -1;
		303	if (amin > bmax) return +1;
		304	}
		305
		306	ady = a->line.p2.y - a->line.p1.y;
		307	adx = a->line.p2.x - a->line.p1.x;
		308	if (adx == 0)
		309	have_dx_adx_bdx &= ~HAVE_ADX;
		310
		311	bdy = b->line.p2.y - b->line.p1.y;
		312	bdx = b->line.p2.x - b->line.p1.x;
		313	if (bdx == 0)
		314	have_dx_adx_bdx &= ~HAVE_BDX;
		315
		316	dx = a->line.p1.x - b->line.p1.x;
		317	if (dx == 0)
		318	have_dx_adx_bdx &= ~HAVE_DX;
		319
		320	#define L _cairo_int64x32_128_mul (_cairo_int32x32_64_mul (ady, bdy), dx)
		321	#define A _cairo_int64x32_128_mul (_cairo_int32x32_64_mul (adx, bdy), y - a->line.p1.y)
		322	#define B _cairo_int64x32_128_mul (_cairo_int32x32_64_mul (bdx, ady), y - b->line.p1.y)
		323	switch (have_dx_adx_bdx) {
		324	default:
		325	case HAVE_NONE:
		326	return 0;
		327	case HAVE_DX:
		328	/* A_dy * B_dy * (A_x - B_x) ∘ 0 */
		329	return dx; /* ady * bdy is positive definite */
		330	case HAVE_ADX:
		331	/* 0 ∘ - (Y - A_y) * A_dx * B_dy */
		332	return adx; /* bdy * (y - a->top.y) is positive definite */
		333	case HAVE_BDX:
		334	/* 0 ∘ (Y - B_y) * B_dx * A_dy */
		335	return -bdx; /* ady * (y - b->top.y) is positive definite */
		336	case HAVE_ADX_BDX:
		337	/* 0 ∘ (Y - B_y) * B_dx * A_dy - (Y - A_y) * A_dx * B_dy */
		338	if ((adx ^ bdx) < 0) {
		339	return adx;
		340	} else if (a->line.p1.y == b->line.p1.y) { /* common origin */
		341	cairo_int64_t adx_bdy, bdx_ady;
		342
		343	/* ∴ A_dx * B_dy ∘ B_dx * A_dy */
		344
		345	adx_bdy = _cairo_int32x32_64_mul (adx, bdy);
		346	bdx_ady = _cairo_int32x32_64_mul (bdx, ady);
		347
		348	return _cairo_int64_cmp (adx_bdy, bdx_ady);
		349	} else
		350	return _cairo_int128_cmp (A, B);
		351	case HAVE_DX_ADX:
		352	/* A_dy * (A_x - B_x) ∘ - (Y - A_y) * A_dx */
		353	if ((-adx ^ dx) < 0) {
		354	return dx;
		355	} else {
		356	cairo_int64_t ady_dx, dy_adx;
		357
		358	ady_dx = _cairo_int32x32_64_mul (ady, dx);
		359	dy_adx = _cairo_int32x32_64_mul (a->line.p1.y - y, adx);
		360
		361	return _cairo_int64_cmp (ady_dx, dy_adx);
		362	}
		363	case HAVE_DX_BDX:
		364	/* B_dy * (A_x - B_x) ∘ (Y - B_y) * B_dx */
		365	if ((bdx ^ dx) < 0) {
		366	return dx;
		367	} else {
		368	cairo_int64_t bdy_dx, dy_bdx;
		369
		370	bdy_dx = _cairo_int32x32_64_mul (bdy, dx);
		371	dy_bdx = _cairo_int32x32_64_mul (y - b->line.p1.y, bdx);
		372
		373	return _cairo_int64_cmp (bdy_dx, dy_bdx);
		374	}
		375	case HAVE_ALL:
		376	/* XXX try comparing (a->line.p2.x - b->line.p2.x) et al */
		377	return _cairo_int128_cmp (L, _cairo_int128_sub (B, A));
		378	}
		379	#undef B
		380	#undef A
		381	#undef L
		382	}
		383
		384	/*
		385	* We need to compare the x-coordinate of a line for a particular y wrt to a
		386	* given x, without loss of precision.
		387	*
		388	* The x-coordinate along an edge for a given y is:
		389	* X = A_x + (Y - A_y) * A_dx / A_dy
		390	*
		391	* So the inequality we wish to test is:
		392	* A_x + (Y - A_y) * A_dx / A_dy ∘ X
		393	* where ∘ is our inequality operator.
		394	*
		395	* By construction, we know that A_dy (and (Y - A_y)) are
		396	* all positive, so we can rearrange it thus without causing a sign change:
		397	* (Y - A_y) * A_dx ∘ (X - A_x) * A_dy
		398	*
		399	* Given the assumption that all the deltas fit within 32 bits, we can compute
		400	* this comparison directly using 64 bit arithmetic.
		401	*
		402	* See the similar discussion for _slope_compare() and
		403	* edges_compare_x_for_y_general().
		404	*/
		405	static int
		406	edge_compare_for_y_against_x (const cairo_edge_t *a,
		407	int32_t y,
		408	int32_t x)
		409	{
		410	int32_t adx, ady;
		411	int32_t dx, dy;
		412	cairo_int64_t L, R;
		413
		414	if (a->line.p1.x <= a->line.p2.x) {
		415	if (x < a->line.p1.x)
		416	return 1;
		417	if (x > a->line.p2.x)
		418	return -1;
		419	} else {
		420	if (x < a->line.p2.x)
		421	return 1;
		422	if (x > a->line.p1.x)
		423	return -1;
		424	}
		425
		426	adx = a->line.p2.x - a->line.p1.x;
		427	dx = x - a->line.p1.x;
		428
		429	if (adx == 0)
		430	return -dx;
		431	if (dx == 0 \|\| (adx ^ dx) < 0)
		432	return adx;
		433
		434	dy = y - a->line.p1.y;
		435	ady = a->line.p2.y - a->line.p1.y;
		436
		437	L = _cairo_int32x32_64_mul (dy, adx);
		438	R = _cairo_int32x32_64_mul (dx, ady);
		439
		440	return _cairo_int64_cmp (L, R);
		441	}
		442
		443	static int
		444	edges_compare_x_for_y (const cairo_edge_t *a,
		445	const cairo_edge_t *b,
		446	int32_t y)
		447	{
		448	/* If the sweep-line is currently on an end-point of a line,
		449	* then we know its precise x value (and considering that we often need to
		450	* compare events at end-points, this happens frequently enough to warrant
		451	* special casing).
		452	*/
		453	enum {
		454	HAVE_NEITHER = 0x0,
		455	HAVE_AX = 0x1,
		456	HAVE_BX = 0x2,
		457	HAVE_BOTH = HAVE_AX \| HAVE_BX
		458	} have_ax_bx = HAVE_BOTH;
		459	int32_t ax, bx;
		460
		461	/* XXX given we have x and dx? */
		462
		463	if (y == a->line.p1.y)
		464	ax = a->line.p1.x;
		465	else if (y == a->line.p2.y)
		466	ax = a->line.p2.x;
		467	else
		468	have_ax_bx &= ~HAVE_AX;
		469
		470	if (y == b->line.p1.y)
		471	bx = b->line.p1.x;
		472	else if (y == b->line.p2.y)
		473	bx = b->line.p2.x;
		474	else
		475	have_ax_bx &= ~HAVE_BX;
		476
		477	switch (have_ax_bx) {
		478	default:
		479	case HAVE_NEITHER:
		480	return edges_compare_x_for_y_general (a, b, y);
		481	case HAVE_AX:
		482	return -edge_compare_for_y_against_x (b, y, ax);
		483	case HAVE_BX:
		484	return edge_compare_for_y_against_x (a, y, bx);
		485	case HAVE_BOTH:
		486	return ax - bx;
		487	}
		488	}
		489
		490	static inline int
		491	slope_compare (const edge_t *a,
		492	const edge_t *b)
		493	{
		494	cairo_int64_t L, R;
		495	int cmp;
		496
		497	cmp = a->dxdy.quo - b->dxdy.quo;
		498	if (cmp)
		499	return cmp;
		500
		501	if (a->dxdy.rem == 0)
		502	return -b->dxdy.rem;
		503	if (b->dxdy.rem == 0)
		504	return a->dxdy.rem;
		505
		506	L = _cairo_int32x32_64_mul (b->dy, a->dxdy.rem);
		507	R = _cairo_int32x32_64_mul (a->dy, b->dxdy.rem);
		508	return _cairo_int64_cmp (L, R);
		509	}
		510
		511	static inline int
		512	line_equal (const cairo_line_t a, const cairo_line_t b)
		513	{
		514	return a->p1.x == b->p1.x && a->p1.y == b->p1.y &&
		515	a->p2.x == b->p2.x && a->p2.y == b->p2.y;
		516	}
		517
		518	static inline int
		519	sweep_line_compare_edges (const edge_t *a,
		520	const edge_t *b,
		521	cairo_fixed_t y)
		522	{
		523	int cmp;
		524
		525	if (line_equal (&a->edge.line, &b->edge.line))
		526	return 0;
		527
		528	cmp = edges_compare_x_for_y (&a->edge, &b->edge, y);
		529	if (cmp)
		530	return cmp;
		531
		532	return slope_compare (a, b);
		533	}
		534
		535	static inline cairo_int64_t
		536	det32_64 (int32_t a, int32_t b,
		537	int32_t c, int32_t d)
		538	{
		539	/* det = a * d - b * c */
		540	return _cairo_int64_sub (_cairo_int32x32_64_mul (a, d),
		541	_cairo_int32x32_64_mul (b, c));
		542	}
		543
		544	static inline cairo_int128_t
		545	det64x32_128 (cairo_int64_t a, int32_t b,
		546	cairo_int64_t c, int32_t d)
		547	{
		548	/* det = a * d - b * c */
		549	return _cairo_int128_sub (_cairo_int64x32_128_mul (a, d),
		550	_cairo_int64x32_128_mul (c, b));
		551	}
		552
		553	/* Compute the intersection of two lines as defined by two edges. The
		554	* result is provided as a coordinate pair of 128-bit integers.
		555	*
		556	* Returns %CAIRO_BO_STATUS_INTERSECTION if there is an intersection or
		557	* %CAIRO_BO_STATUS_PARALLEL if the two lines are exactly parallel.
		558	*/
		559	static cairo_bool_t
		560	intersect_lines (const edge_t a, const edge_t b,
		561	cairo_bo_intersect_point_t *intersection)
		562	{
		563	cairo_int64_t a_det, b_det;
		564
		565	/* XXX: We're assuming here that dx and dy will still fit in 32
		566	* bits. That's not true in general as there could be overflow. We
		567	* should prevent that before the tessellation algorithm begins.
		568	* What we're doing to mitigate this is to perform clamping in
		569	* cairo_bo_tessellate_polygon().
		570	*/
		571	int32_t dx1 = a->edge.line.p1.x - a->edge.line.p2.x;
		572	int32_t dy1 = a->edge.line.p1.y - a->edge.line.p2.y;
		573
		574	int32_t dx2 = b->edge.line.p1.x - b->edge.line.p2.x;
		575	int32_t dy2 = b->edge.line.p1.y - b->edge.line.p2.y;
		576
		577	cairo_int64_t den_det;
		578	cairo_int64_t R;
		579	cairo_quorem64_t qr;
		580
		581	den_det = det32_64 (dx1, dy1, dx2, dy2);
		582
		583	/* Q: Can we determine that the lines do not intersect (within range)
		584	* much more cheaply than computing the intersection point i.e. by
		585	* avoiding the division?
		586	*
		587	* X = ax + t * adx = bx + s * bdx;
		588	* Y = ay + t * ady = by + s * bdy;
		589	* ∴ t * (adybdx - bdyadx) = bdx * (by - ay) + bdy * (ax - bx)
		590	* => t * L = R
		591	*
		592	* Therefore we can reject any intersection (under the criteria for
		593	* valid intersection events) if:
		594	* L^R < 0 => t < 0, or
		595	* L t > 1
		596	*
		597	* (where top/bottom must at least extend to the line endpoints).
		598	*
		599	* A similar substitution can be performed for s, yielding:
		600	* s * (adybdx - bdyadx) = ady * (ax - bx) - adx * (ay - by)
		601	*/
		602	R = det32_64 (dx2, dy2,
		603	b->edge.line.p1.x - a->edge.line.p1.x,
		604	b->edge.line.p1.y - a->edge.line.p1.y);
		605	if (_cairo_int64_negative (den_det)) {
		606	if (_cairo_int64_ge (den_det, R))
		607	return FALSE;
		608	} else {
		609	if (_cairo_int64_le (den_det, R))
		610	return FALSE;
		611	}
		612
		613	R = det32_64 (dy1, dx1,
		614	a->edge.line.p1.y - b->edge.line.p1.y,
		615	a->edge.line.p1.x - b->edge.line.p1.x);
		616	if (_cairo_int64_negative (den_det)) {
		617	if (_cairo_int64_ge (den_det, R))
		618	return FALSE;
		619	} else {
		620	if (_cairo_int64_le (den_det, R))
		621	return FALSE;
		622	}
		623
		624	/* We now know that the two lines should intersect within range. */
		625
		626	a_det = det32_64 (a->edge.line.p1.x, a->edge.line.p1.y,
		627	a->edge.line.p2.x, a->edge.line.p2.y);
		628	b_det = det32_64 (b->edge.line.p1.x, b->edge.line.p1.y,
		629	b->edge.line.p2.x, b->edge.line.p2.y);
		630
		631	/* x = det (a_det, dx1, b_det, dx2) / den_det */
		632	qr = _cairo_int_96by64_32x64_divrem (det64x32_128 (a_det, dx1,
		633	b_det, dx2),
		634	den_det);
		635	if (_cairo_int64_eq (qr.rem, den_det))
		636	return FALSE;
		637	#if 0
		638	intersection->x.exactness = _cairo_int64_is_zero (qr.rem) ? EXACT : INEXACT;
		639	#else
		640	intersection->x.exactness = EXACT;
		641	if (! _cairo_int64_is_zero (qr.rem)) {
		642	if (_cairo_int64_negative (den_det) ^ _cairo_int64_negative (qr.rem))
		643	qr.rem = _cairo_int64_negate (qr.rem);
		644	qr.rem = _cairo_int64_mul (qr.rem, _cairo_int32_to_int64 (2));
		645	if (_cairo_int64_ge (qr.rem, den_det)) {
		646	qr.quo = _cairo_int64_add (qr.quo,
		647	_cairo_int32_to_int64 (_cairo_int64_negative (qr.quo) ? -1 : 1));
		648	} else
		649	intersection->x.exactness = INEXACT;
		650	}
		651	#endif
		652	intersection->x.ordinate = _cairo_int64_to_int32 (qr.quo);
		653
		654	/* y = det (a_det, dy1, b_det, dy2) / den_det */
		655	qr = _cairo_int_96by64_32x64_divrem (det64x32_128 (a_det, dy1,
		656	b_det, dy2),
		657	den_det);
		658	if (_cairo_int64_eq (qr.rem, den_det))
		659	return FALSE;
		660	#if 0
		661	intersection->y.exactness = _cairo_int64_is_zero (qr.rem) ? EXACT : INEXACT;
		662	#else
		663	intersection->y.exactness = EXACT;
		664	if (! _cairo_int64_is_zero (qr.rem)) {
		665	/* compute ceiling away from zero */
		666	qr.quo = _cairo_int64_add (qr.quo,
		667	_cairo_int32_to_int64 (_cairo_int64_negative (qr.quo) ? -1 : 1));
		668	intersection->y.exactness = INEXACT;
		669	}
		670	#endif
		671	intersection->y.ordinate = _cairo_int64_to_int32 (qr.quo);
		672
		673	return TRUE;
		674	}
		675
		676	static int
		677	bo_intersect_ordinate_32_compare (int32_t a, int32_t b, int exactness)
		678	{
		679	int cmp;
		680
		681	/* First compare the quotient */
		682	cmp = a - b;
		683	if (cmp)
		684	return cmp;
		685
		686	/* With quotient identical, if remainder is 0 then compare equal */
		687	/* Otherwise, the non-zero remainder makes a > b */
		688	return -(INEXACT == exactness);
		689	}
		690
		691	/* Does the given edge contain the given point. The point must already
		692	* be known to be contained within the line determined by the edge,
		693	* (most likely the point results from an intersection of this edge
		694	* with another).
		695	*
		696	* If we had exact arithmetic, then this function would simply be a
		697	* matter of examining whether the y value of the point lies within
		698	* the range of y values of the edge. But since intersection points
		699	* are not exact due to being rounded to the nearest integer within
		700	* the available precision, we must also examine the x value of the
		701	* point.
		702	*
		703	* The definition of "contains" here is that the given intersection
		704	* point will be seen by the sweep line after the start event for the
		705	* given edge and before the stop event for the edge. See the comments
		706	* in the implementation for more details.
		707	*/
		708	static cairo_bool_t
		709	bo_edge_contains_intersect_point (const edge_t *edge,
		710	cairo_bo_intersect_point_t *point)
		711	{
		712	int cmp_top, cmp_bottom;
		713
		714	/* XXX: When running the actual algorithm, we don't actually need to
		715	* compare against edge->top at all here, since any intersection above
		716	* top is eliminated early via a slope comparison. We're leaving these
		717	* here for now only for the sake of the quadratic-time intersection
		718	* finder which needs them.
		719	*/
		720
		721	cmp_top = bo_intersect_ordinate_32_compare (point->y.ordinate,
		722	edge->edge.top,
		723	point->y.exactness);
		724	if (cmp_top < 0)
		725	return FALSE;
		726
		727	cmp_bottom = bo_intersect_ordinate_32_compare (point->y.ordinate,
		728	edge->edge.bottom,
		729	point->y.exactness);
		730	if (cmp_bottom > 0)
		731	return FALSE;
		732
		733	if (cmp_top > 0 && cmp_bottom < 0)
		734	return TRUE;
		735
		736	/* At this stage, the point lies on the same y value as either
		737	* edge->top or edge->bottom, so we have to examine the x value in
		738	* order to properly determine containment. */
		739
		740	/* If the y value of the point is the same as the y value of the
		741	* top of the edge, then the x value of the point must be greater
		742	* to be considered as inside the edge. Similarly, if the y value
		743	* of the point is the same as the y value of the bottom of the
		744	* edge, then the x value of the point must be less to be
		745	* considered as inside. */
		746
		747	if (cmp_top == 0) {
		748	cairo_fixed_t top_x;
		749
		750	top_x = line_compute_intersection_x_for_y (&edge->edge.line,
		751	edge->edge.top);
		752	return bo_intersect_ordinate_32_compare (top_x, point->x.ordinate, point->x.exactness) < 0;
		753	} else { /* cmp_bottom == 0 */
		754	cairo_fixed_t bot_x;
		755
		756	bot_x = line_compute_intersection_x_for_y (&edge->edge.line,
		757	edge->edge.bottom);
		758	return bo_intersect_ordinate_32_compare (point->x.ordinate, bot_x, point->x.exactness) < 0;
		759	}
		760	}
		761
		762	static cairo_bool_t
		763	edge_intersect (const edge_t *a,
		764	const edge_t *b,
		765	cairo_point_t *intersection)
		766	{
		767	cairo_bo_intersect_point_t quorem;
		768
		769	if (! intersect_lines (a, b, &quorem))
		770	return FALSE;
		771
		772	if (a->edge.top != a->edge.line.p1.y \|\| a->edge.bottom != a->edge.line.p2.y) {
		773	if (! bo_edge_contains_intersect_point (a, &quorem))
		774	return FALSE;
		775	}
		776
		777	if (b->edge.top != b->edge.line.p1.y \|\| b->edge.bottom != b->edge.line.p2.y) {
		778	if (! bo_edge_contains_intersect_point (b, &quorem))
		779	return FALSE;
		780	}
		781
		782	/* Now that we've correctly compared the intersection point and
		783	* determined that it lies within the edge, then we know that we
		784	* no longer need any more bits of storage for the intersection
		785	* than we do for our edge coordinates. We also no longer need the
		786	* remainder from the division. */
		787	intersection->x = quorem.x.ordinate;
		788	intersection->y = quorem.y.ordinate;
		789
		790	return TRUE;
		791	}
		792
		793	static inline int
		794	event_compare (const event_t a, const event_t b)
		795	{
		796	return a->y - b->y;
		797	}
		798
		799	static void
		800	pqueue_init (pqueue_t *pq)
		801	{
		802	pq->max_size = ARRAY_LENGTH (pq->elements_embedded);
		803	pq->size = 0;
		804
		805	pq->elements = pq->elements_embedded;
		806	}
		807
		808	static void
		809	pqueue_fini (pqueue_t *pq)
		810	{
		811	if (pq->elements != pq->elements_embedded)
		812	free (pq->elements);
		813	}
		814
		815	static cairo_bool_t
		816	pqueue_grow (pqueue_t *pq)
		817	{
		818	event_t **new_elements;
		819	pq->max_size *= 2;
		820
		821	if (pq->elements == pq->elements_embedded) {
		822	new_elements = _cairo_malloc_ab (pq->max_size,
		823	sizeof (event_t *));
		824	if (unlikely (new_elements == NULL))
		825	return FALSE;
		826
		827	memcpy (new_elements, pq->elements_embedded,
		828	sizeof (pq->elements_embedded));
		829	} else {
		830	new_elements = _cairo_realloc_ab (pq->elements,
		831	pq->max_size,
		832	sizeof (event_t *));
		833	if (unlikely (new_elements == NULL))
		834	return FALSE;
		835	}
		836
		837	pq->elements = new_elements;
		838	return TRUE;
		839	}
		840
		841	static inline void
		842	pqueue_push (sweep_line_t sweep_line, event_t event)
		843	{
		844	event_t **elements;
		845	int i, parent;
		846
		847	if (unlikely (sweep_line->queue.pq.size + 1 == sweep_line->queue.pq.max_size)) {
		848	if (unlikely (! pqueue_grow (&sweep_line->queue.pq))) {
		849	longjmp (sweep_line->unwind,
		850	_cairo_error (CAIRO_STATUS_NO_MEMORY));
		851	}
		852	}
		853
		854	elements = sweep_line->queue.pq.elements;
		855	for (i = ++sweep_line->queue.pq.size;
		856	i != PQ_FIRST_ENTRY &&
		857	event_compare (event,
		858	elements[parent = PQ_PARENT_INDEX (i)]) < 0;
		859	i = parent)
		860	{
		861	elements[i] = elements[parent];
		862	}
		863
		864	elements[i] = event;
		865	}
		866
		867	static inline void
		868	pqueue_pop (pqueue_t *pq)
		869	{
		870	event_t **elements = pq->elements;
		871	event_t *tail;
		872	int child, i;
		873
		874	tail = elements[pq->size--];
		875	if (pq->size == 0) {
		876	elements[PQ_FIRST_ENTRY] = NULL;
		877	return;
		878	}
		879
		880	for (i = PQ_FIRST_ENTRY;
		881	(child = PQ_LEFT_CHILD_INDEX (i)) <= pq->size;
		882	i = child)
		883	{
		884	if (child != pq->size &&
		885	event_compare (elements[child+1],
		886	elements[child]) < 0)
		887	{
		888	child++;
		889	}
		890
		891	if (event_compare (elements[child], tail) >= 0)
		892	break;
		893
		894	elements[i] = elements[child];
		895	}
		896	elements[i] = tail;
		897	}
		898
		899	static inline void
		900	event_insert (sweep_line_t *sweep_line,
		901	event_type_t type,
		902	edge_t *e1,
		903	edge_t *e2,
		904	cairo_fixed_t y)
		905	{
		906	queue_event_t *event;
		907
		908	event = _cairo_freepool_alloc (&sweep_line->queue.pool);
		909	if (unlikely (event == NULL)) {
		910	longjmp (sweep_line->unwind,
		911	_cairo_error (CAIRO_STATUS_NO_MEMORY));
		912	}
		913
		914	event->y = y;
		915	event->type = type;
		916	event->e1 = e1;
		917	event->e2 = e2;
		918
		919	pqueue_push (sweep_line, (event_t *) event);
		920	}
		921
		922	static void
		923	event_delete (sweep_line_t *sweep_line,
		924	event_t *event)
		925	{
		926	_cairo_freepool_free (&sweep_line->queue.pool, event);
		927	}
		928
		929	static inline event_t *
		930	event_next (sweep_line_t *sweep_line)
		931	{
		932	event_t event, cmp;
		933
		934	event = sweep_line->queue.pq.elements[PQ_FIRST_ENTRY];
		935	cmp = *sweep_line->queue.start_events;
		936	if (event == NULL \|\|
		937	(cmp != NULL && event_compare (cmp, event) < 0))
		938	{
		939	event = cmp;
		940	sweep_line->queue.start_events++;
		941	}
		942	else
		943	{
		944	pqueue_pop (&sweep_line->queue.pq);
		945	}
		946
		947	return event;
		948	}
		949
		950	CAIRO_COMBSORT_DECLARE (start_event_sort, event_t *, event_compare)
		951
		952	static inline void
		953	event_insert_stop (sweep_line_t *sweep_line,
		954	edge_t *edge)
		955	{
		956	event_insert (sweep_line,
		957	EVENT_TYPE_STOP,
		958	edge, NULL,
		959	edge->edge.bottom);
		960	}
		961
		962	static inline void
		963	event_insert_if_intersect_below_current_y (sweep_line_t *sweep_line,
		964	edge_t *left,
		965	edge_t *right)
		966	{
		967	cairo_point_t intersection;
		968
		969	/* start points intersect */
		970	if (left->edge.line.p1.x == right->edge.line.p1.x &&
		971	left->edge.line.p1.y == right->edge.line.p1.y)
		972	{
		973	return;
		974	}
		975
		976	/* end points intersect, process DELETE events first */
		977	if (left->edge.line.p2.x == right->edge.line.p2.x &&
		978	left->edge.line.p2.y == right->edge.line.p2.y)
		979	{
		980	return;
		981	}
		982
		983	if (slope_compare (left, right) <= 0)
		984	return;
		985
		986	if (! edge_intersect (left, right, &intersection))
		987	return;
		988
		989	event_insert (sweep_line,
		990	EVENT_TYPE_INTERSECTION,
		991	left, right,
		992	intersection.y);
		993	}
		994
		995	static inline edge_t *
		996	link_to_edge (cairo_list_t *link)
		997	{
		998	return (edge_t *) link;
		999	}
		1000
		1001	static void
		1002	sweep_line_insert (sweep_line_t *sweep_line,
		1003	edge_t *edge)
		1004	{
		1005	cairo_list_t *pos;
		1006	cairo_fixed_t y = sweep_line->current_subrow;
		1007
		1008	pos = sweep_line->insert_cursor;
		1009	if (pos == &sweep_line->active)
		1010	pos = sweep_line->active.next;
		1011	if (pos != &sweep_line->active) {
		1012	int cmp;
		1013
		1014	cmp = sweep_line_compare_edges (link_to_edge (pos),
		1015	edge,
		1016	y);
		1017	if (cmp < 0) {
		1018	while (pos->next != &sweep_line->active &&
		1019	sweep_line_compare_edges (link_to_edge (pos->next),
		1020	edge,
		1021	y) < 0)
		1022	{
		1023	pos = pos->next;
		1024	}
		1025	} else if (cmp > 0) {
		1026	do {
		1027	pos = pos->prev;
		1028	} while (pos != &sweep_line->active &&
		1029	sweep_line_compare_edges (link_to_edge (pos),
		1030	edge,
		1031	y) > 0);
		1032	}
		1033	}
		1034	cairo_list_add (&edge->link, pos);
		1035	sweep_line->insert_cursor = &edge->link;
		1036	}
		1037
		1038	inline static void
		1039	coverage_rewind (struct coverage *cells)
		1040	{
		1041	cells->cursor = &cells->head;
		1042	}
		1043
		1044	static void
		1045	coverage_init (struct coverage *cells)
		1046	{
		1047	_cairo_freepool_init (&cells->pool,
		1048	sizeof (struct cell));
		1049	cells->head.prev = NULL;
		1050	cells->head.next = &cells->tail;
		1051	cells->head.x = INT_MIN;
		1052	cells->tail.prev = &cells->head;
		1053	cells->tail.next = NULL;
		1054	cells->tail.x = INT_MAX;
		1055	cells->count = 0;
		1056	coverage_rewind (cells);
		1057	}
		1058
		1059	static void
		1060	coverage_fini (struct coverage *cells)
		1061	{
		1062	_cairo_freepool_fini (&cells->pool);
		1063	}
		1064
		1065	inline static void
		1066	coverage_reset (struct coverage *cells)
		1067	{
		1068	cells->head.next = &cells->tail;
		1069	cells->tail.prev = &cells->head;
		1070	cells->count = 0;
		1071	_cairo_freepool_reset (&cells->pool);
		1072	coverage_rewind (cells);
		1073	}
		1074
		1075	inline static struct cell *
		1076	coverage_alloc (sweep_line_t *sweep_line,
		1077	struct cell *tail,
		1078	int x)
		1079	{
		1080	struct cell *cell;
		1081
		1082	cell = _cairo_freepool_alloc (&sweep_line->coverage.pool);
		1083	if (unlikely (NULL == cell)) {
		1084	longjmp (sweep_line->unwind,
		1085	_cairo_error (CAIRO_STATUS_NO_MEMORY));
		1086	}
		1087
		1088	tail->prev->next = cell;
		1089	cell->prev = tail->prev;
		1090	cell->next = tail;
		1091	tail->prev = cell;
		1092	cell->x = x;
		1093	cell->uncovered_area = 0;
		1094	cell->covered_height = 0;
		1095	sweep_line->coverage.count++;
		1096	return cell;
		1097	}
		1098
		1099	inline static struct cell *
		1100	coverage_find (sweep_line_t *sweep_line, int x)
		1101	{
		1102	struct cell *cell;
		1103
		1104	cell = sweep_line->coverage.cursor;
		1105	if (unlikely (cell->x > x)) {
		1106	do {
		1107	if (cell->prev->x < x)
		1108	break;
		1109	cell = cell->prev;
		1110	} while (TRUE);
		1111	} else {
		1112	if (cell->x == x)
		1113	return cell;
		1114
		1115	do {
		1116	UNROLL3({
		1117	cell = cell->next;
		1118	if (cell->x >= x)
		1119	break;
		1120	});
		1121	} while (TRUE);
		1122	}
		1123
		1124	if (cell->x != x)
		1125	cell = coverage_alloc (sweep_line, cell, x);
		1126
		1127	return sweep_line->coverage.cursor = cell;
		1128	}
		1129
		1130	static void
		1131	coverage_render_cells (sweep_line_t *sweep_line,
		1132	cairo_fixed_t left, cairo_fixed_t right,
		1133	cairo_fixed_t y1, cairo_fixed_t y2,
		1134	int sign)
		1135	{
		1136	int fx1, fx2;
		1137	int ix1, ix2;
		1138	int dx, dy;
		1139
		1140	/* Orient the edge left-to-right. */
		1141	dx = right - left;
		1142	if (dx >= 0) {
		1143	ix1 = _cairo_fixed_integer_part (left);
		1144	fx1 = _cairo_fixed_fractional_part (left);
		1145
		1146	ix2 = _cairo_fixed_integer_part (right);
		1147	fx2 = _cairo_fixed_fractional_part (right);
		1148
		1149	dy = y2 - y1;
		1150	} else {
		1151	ix1 = _cairo_fixed_integer_part (right);
		1152	fx1 = _cairo_fixed_fractional_part (right);
		1153
		1154	ix2 = _cairo_fixed_integer_part (left);
		1155	fx2 = _cairo_fixed_fractional_part (left);
		1156
		1157	dx = -dx;
		1158	sign = -sign;
		1159	dy = y1 - y2;
		1160	y1 = y2 - dy;
		1161	y2 = y1 + dy;
		1162	}
		1163
		1164	/* Add coverage for all pixels [ix1,ix2] on this row crossed
		1165	* by the edge. */
		1166	{
		1167	struct quorem y = floored_divrem ((STEP_X - fx1)*dy, dx);
		1168	struct cell *cell;
		1169
		1170	cell = sweep_line->coverage.cursor;
		1171	if (cell->x != ix1) {
		1172	if (unlikely (cell->x > ix1)) {
		1173	do {
		1174	if (cell->prev->x < ix1)
		1175	break;
		1176	cell = cell->prev;
		1177	} while (TRUE);
		1178	} else do {
		1179	UNROLL3({
		1180	if (cell->x >= ix1)
		1181	break;
		1182	cell = cell->next;
		1183	});
		1184	} while (TRUE);
		1185
		1186	if (cell->x != ix1)
		1187	cell = coverage_alloc (sweep_line, cell, ix1);
		1188	}
		1189
		1190	cell->uncovered_area += sign * y.quo * (STEP_X + fx1);
		1191	cell->covered_height += sign * y.quo;
		1192	y.quo += y1;
		1193
		1194	cell = cell->next;
		1195	if (cell->x != ++ix1)
		1196	cell = coverage_alloc (sweep_line, cell, ix1);
		1197	if (ix1 < ix2) {
		1198	struct quorem dydx_full = floored_divrem (STEP_X*dy, dx);
		1199
		1200	do {
		1201	cairo_fixed_t y_skip = dydx_full.quo;
		1202	y.rem += dydx_full.rem;
		1203	if (y.rem >= dx) {
		1204	++y_skip;
		1205	y.rem -= dx;
		1206	}
		1207
		1208	y.quo += y_skip;
		1209
		1210	y_skip *= sign;
		1211	cell->covered_height += y_skip;
		1212	cell->uncovered_area += y_skip*STEP_X;
		1213
		1214	cell = cell->next;
		1215	if (cell->x != ++ix1)
		1216	cell = coverage_alloc (sweep_line, cell, ix1);
		1217	} while (ix1 != ix2);
		1218	}
		1219	cell->uncovered_area += sign(y2 - y.quo)fx2;
		1220	cell->covered_height += sign*(y2 - y.quo);
		1221	sweep_line->coverage.cursor = cell;
		1222	}
		1223	}
		1224
		1225	inline static void
		1226	full_inc_edge (edge_t *edge)
		1227	{
		1228	edge->x.quo += edge->dxdy_full.quo;
		1229	edge->x.rem += edge->dxdy_full.rem;
		1230	if (edge->x.rem >= 0) {
		1231	++edge->x.quo;
		1232	edge->x.rem -= edge->dy;
		1233	}
		1234	}
		1235
		1236	static void
		1237	full_add_edge (sweep_line_t sweep_line, edge_t edge, int sign)
		1238	{
		1239	struct cell *cell;
		1240	cairo_fixed_t x1, x2;
		1241	int ix1, ix2;
		1242	int frac;
		1243
		1244	edge->current_sign = sign;
		1245
		1246	ix1 = _cairo_fixed_integer_part (edge->x.quo);
		1247
		1248	if (edge->vertical) {
		1249	frac = _cairo_fixed_fractional_part (edge->x.quo);
		1250	cell = coverage_find (sweep_line, ix1);
		1251	cell->covered_height += sign * STEP_Y;
		1252	cell->uncovered_area += sign * 2 * frac * STEP_Y;
		1253	return;
		1254	}
		1255
		1256	x1 = edge->x.quo;
		1257	full_inc_edge (edge);
		1258	x2 = edge->x.quo;
		1259
		1260	ix2 = _cairo_fixed_integer_part (edge->x.quo);
		1261
		1262	/* Edge is entirely within a column? */
		1263	if (likely (ix1 == ix2)) {
		1264	frac = _cairo_fixed_fractional_part (x1) +
		1265	_cairo_fixed_fractional_part (x2);
		1266	cell = coverage_find (sweep_line, ix1);
		1267	cell->covered_height += sign * STEP_Y;
		1268	cell->uncovered_area += sign * frac * STEP_Y;
		1269	return;
		1270	}
		1271
		1272	coverage_render_cells (sweep_line, x1, x2, 0, STEP_Y, sign);
		1273	}
		1274
		1275	static void
		1276	full_nonzero (sweep_line_t *sweep_line)
		1277	{
		1278	cairo_list_t *pos;
		1279
		1280	sweep_line->is_vertical = TRUE;
		1281	pos = sweep_line->active.next;
		1282	do {
		1283	edge_t left = link_to_edge (pos), right;
		1284	int winding = left->edge.dir;
		1285
		1286	sweep_line->is_vertical &= left->vertical;
		1287
		1288	pos = left->link.next;
		1289	do {
		1290	if (unlikely (pos == &sweep_line->active)) {
		1291	full_add_edge (sweep_line, left, +1);
		1292	return;
		1293	}
		1294
		1295	right = link_to_edge (pos);
		1296	pos = pos->next;
		1297	sweep_line->is_vertical &= right->vertical;
		1298
		1299	winding += right->edge.dir;
		1300	if (0 == winding) {
		1301	if (pos == &sweep_line->active \|\|
		1302	link_to_edge (pos)->x.quo != right->x.quo)
		1303	{
		1304	break;
		1305	}
		1306	}
		1307
		1308	if (! right->vertical)
		1309	full_inc_edge (right);
		1310	} while (TRUE);
		1311
		1312	full_add_edge (sweep_line, left, +1);
		1313	full_add_edge (sweep_line, right, -1);
		1314	} while (pos != &sweep_line->active);
		1315	}
		1316
		1317	static void
		1318	full_evenodd (sweep_line_t *sweep_line)
		1319	{
		1320	cairo_list_t *pos;
		1321
		1322	sweep_line->is_vertical = TRUE;
		1323	pos = sweep_line->active.next;
		1324	do {
		1325	edge_t left = link_to_edge (pos), right;
		1326	int winding = 0;
		1327
		1328	sweep_line->is_vertical &= left->vertical;
		1329
		1330	pos = left->link.next;
		1331	do {
		1332	if (pos == &sweep_line->active) {
		1333	full_add_edge (sweep_line, left, +1);
		1334	return;
		1335	}
		1336
		1337	right = link_to_edge (pos);
		1338	pos = pos->next;
		1339	sweep_line->is_vertical &= right->vertical;
		1340
		1341	if (++winding & 1) {
		1342	if (pos == &sweep_line->active \|\|
		1343	link_to_edge (pos)->x.quo != right->x.quo)
		1344	{
		1345	break;
		1346	}
		1347	}
		1348
		1349	if (! right->vertical)
		1350	full_inc_edge (right);
		1351	} while (TRUE);
		1352
		1353	full_add_edge (sweep_line, left, +1);
		1354	full_add_edge (sweep_line, right, -1);
		1355	} while (pos != &sweep_line->active);
		1356	}
		1357
		1358	static void
		1359	render_rows (cairo_botor_scan_converter_t *self,
		1360	sweep_line_t *sweep_line,
		1361	int y, int height,
		1362	cairo_span_renderer_t *renderer)
		1363	{
		1364	cairo_half_open_span_t spans_stack[CAIRO_STACK_ARRAY_LENGTH (cairo_half_open_span_t)];
		1365	cairo_half_open_span_t *spans = spans_stack;
		1366	struct cell *cell;
		1367	int prev_x, cover;
		1368	int num_spans;
		1369	cairo_status_t status;
		1370
		1371	if (unlikely (sweep_line->coverage.count == 0)) {
		1372	status = renderer->render_rows (renderer, y, height, NULL, 0);
		1373	if (unlikely (status))
		1374	longjmp (sweep_line->unwind, status);
		1375	return;
		1376	}
		1377
		1378	/* Allocate enough spans for the row. */
		1379
		1380	num_spans = 2*sweep_line->coverage.count+2;
		1381	if (unlikely (num_spans > ARRAY_LENGTH (spans_stack))) {
		1382	spans = _cairo_malloc_ab (num_spans, sizeof (cairo_half_open_span_t));
		1383	if (unlikely (spans == NULL)) {
		1384	longjmp (sweep_line->unwind,
		1385	_cairo_error (CAIRO_STATUS_NO_MEMORY));
		1386	}
		1387	}
		1388
		1389	/* Form the spans from the coverage and areas. */
		1390	num_spans = 0;
		1391	prev_x = self->xmin;
		1392	cover = 0;
		1393	cell = sweep_line->coverage.head.next;
		1394	do {
		1395	int x = cell->x;
		1396	int area;
		1397
		1398	if (x > prev_x) {
		1399	spans[num_spans].x = prev_x;
		1400	spans[num_spans].coverage = AREA_TO_ALPHA (cover);
		1401	++num_spans;
		1402	}
		1403
		1404	cover += cell->covered_heightSTEP_X2;
		1405	area = cover - cell->uncovered_area;
		1406
		1407	spans[num_spans].x = x;
		1408	spans[num_spans].coverage = AREA_TO_ALPHA (area);
		1409	++num_spans;
		1410
		1411	prev_x = x + 1;
		1412	} while ((cell = cell->next) != &sweep_line->coverage.tail);
		1413
		1414	if (prev_x <= self->xmax) {
		1415	spans[num_spans].x = prev_x;
		1416	spans[num_spans].coverage = AREA_TO_ALPHA (cover);
		1417	++num_spans;
		1418	}
		1419
		1420	if (cover && prev_x < self->xmax) {
		1421	spans[num_spans].x = self->xmax;
		1422	spans[num_spans].coverage = 0;
		1423	++num_spans;
		1424	}
		1425
		1426	status = renderer->render_rows (renderer, y, height, spans, num_spans);
		1427
		1428	if (unlikely (spans != spans_stack))
		1429	free (spans);
		1430
		1431	coverage_reset (&sweep_line->coverage);
		1432
		1433	if (unlikely (status))
		1434	longjmp (sweep_line->unwind, status);
		1435	}
		1436
		1437	static void
		1438	full_repeat (sweep_line_t *sweep)
		1439	{
		1440	edge_t *edge;
		1441
		1442	cairo_list_foreach_entry (edge, edge_t, &sweep->active, link) {
		1443	if (edge->current_sign)
		1444	full_add_edge (sweep, edge, edge->current_sign);
		1445	else if (! edge->vertical)
		1446	full_inc_edge (edge);
		1447	}
		1448	}
		1449
		1450	static void
		1451	full_reset (sweep_line_t *sweep)
		1452	{
		1453	edge_t *edge;
		1454
		1455	cairo_list_foreach_entry (edge, edge_t, &sweep->active, link)
		1456	edge->current_sign = 0;
		1457	}
		1458
		1459	static void
		1460	full_step (cairo_botor_scan_converter_t *self,
		1461	sweep_line_t *sweep_line,
		1462	cairo_fixed_t row,
		1463	cairo_span_renderer_t *renderer)
		1464	{
		1465	int top, bottom;
		1466
		1467	top = _cairo_fixed_integer_part (sweep_line->current_row);
		1468	bottom = _cairo_fixed_integer_part (row);
		1469	if (cairo_list_is_empty (&sweep_line->active)) {
		1470	cairo_status_t status;
		1471
		1472	status = renderer->render_rows (renderer, top, bottom - top, NULL, 0);
		1473	if (unlikely (status))
		1474	longjmp (sweep_line->unwind, status);
		1475
		1476	return;
		1477	}
		1478
		1479	if (self->fill_rule == CAIRO_FILL_RULE_WINDING)
		1480	full_nonzero (sweep_line);
		1481	else
		1482	full_evenodd (sweep_line);
		1483
		1484	if (sweep_line->is_vertical \|\| bottom == top + 1) {
		1485	render_rows (self, sweep_line, top, bottom - top, renderer);
		1486	full_reset (sweep_line);
		1487	return;
		1488	}
		1489
		1490	render_rows (self, sweep_line, top++, 1, renderer);
		1491	do {
		1492	full_repeat (sweep_line);
		1493	render_rows (self, sweep_line, top, 1, renderer);
		1494	} while (++top != bottom);
		1495
		1496	full_reset (sweep_line);
		1497	}
		1498
		1499	cairo_always_inline static void
		1500	sub_inc_edge (edge_t *edge,
		1501	cairo_fixed_t height)
		1502	{
		1503	if (height == 1) {
		1504	edge->x.quo += edge->dxdy.quo;
		1505	edge->x.rem += edge->dxdy.rem;
		1506	if (edge->x.rem >= 0) {
		1507	++edge->x.quo;
		1508	edge->x.rem -= edge->dy;
		1509	}
		1510	} else {
		1511	edge->x.quo += height * edge->dxdy.quo;
		1512	edge->x.rem += height * edge->dxdy.rem;
		1513	if (edge->x.rem >= 0) {
		1514	int carry = edge->x.rem / edge->dy + 1;
		1515	edge->x.quo += carry;
		1516	edge->x.rem -= carry * edge->dy;
		1517	}
		1518	}
		1519	}
		1520
		1521	static void
		1522	sub_add_run (sweep_line_t sweep_line, edge_t edge, int y, int sign)
		1523	{
		1524	struct run *run;
		1525
		1526	run = _cairo_freepool_alloc (&sweep_line->runs);
		1527	if (unlikely (run == NULL))
		1528	longjmp (sweep_line->unwind, _cairo_error (CAIRO_STATUS_NO_MEMORY));
		1529
		1530	run->y = y;
		1531	run->sign = sign;
		1532	run->next = edge->runs;
		1533	edge->runs = run;
		1534
		1535	edge->current_sign = sign;
		1536	}
		1537
		1538	inline static cairo_bool_t
		1539	edges_coincident (edge_t left, edge_t right, cairo_fixed_t y)
		1540	{
		1541	/* XXX is compare_x_for_y() worth executing during sub steps? */
		1542	return line_equal (&left->edge.line, &right->edge.line);
		1543	//edges_compare_x_for_y (&left->edge, &right->edge, y) >= 0;
		1544	}
		1545
		1546	static void
		1547	sub_nonzero (sweep_line_t *sweep_line)
		1548	{
		1549	cairo_fixed_t y = sweep_line->current_subrow;
		1550	cairo_fixed_t fy = _cairo_fixed_fractional_part (y);
		1551	cairo_list_t *pos;
		1552
		1553	pos = sweep_line->active.next;
		1554	do {
		1555	edge_t left = link_to_edge (pos), right;
		1556	int winding = left->edge.dir;
		1557
		1558	pos = left->link.next;
		1559	do {
		1560	if (unlikely (pos == &sweep_line->active)) {
		1561	if (left->current_sign != +1)
		1562	sub_add_run (sweep_line, left, fy, +1);
		1563	return;
		1564	}
		1565
		1566	right = link_to_edge (pos);
		1567	pos = pos->next;
		1568
		1569	winding += right->edge.dir;
		1570	if (0 == winding) {
		1571	if (pos == &sweep_line->active \|\|
		1572	! edges_coincident (right, link_to_edge (pos), y))
		1573	{
		1574	break;
		1575	}
		1576	}
		1577
		1578	if (right->current_sign)
		1579	sub_add_run (sweep_line, right, fy, 0);
		1580	} while (TRUE);
		1581
		1582	if (left->current_sign != +1)
		1583	sub_add_run (sweep_line, left, fy, +1);
		1584	if (right->current_sign != -1)
		1585	sub_add_run (sweep_line, right, fy, -1);
		1586	} while (pos != &sweep_line->active);
		1587	}
		1588
		1589	static void
		1590	sub_evenodd (sweep_line_t *sweep_line)
		1591	{
		1592	cairo_fixed_t y = sweep_line->current_subrow;
		1593	cairo_fixed_t fy = _cairo_fixed_fractional_part (y);
		1594	cairo_list_t *pos;
		1595
		1596	pos = sweep_line->active.next;
		1597	do {
		1598	edge_t left = link_to_edge (pos), right;
		1599	int winding = 0;
		1600
		1601	pos = left->link.next;
		1602	do {
		1603	if (unlikely (pos == &sweep_line->active)) {
		1604	if (left->current_sign != +1)
		1605	sub_add_run (sweep_line, left, fy, +1);
		1606	return;
		1607	}
		1608
		1609	right = link_to_edge (pos);
		1610	pos = pos->next;
		1611
		1612	if (++winding & 1) {
		1613	if (pos == &sweep_line->active \|\|
		1614	! edges_coincident (right, link_to_edge (pos), y))
		1615	{
		1616	break;
		1617	}
		1618	}
		1619
		1620	if (right->current_sign)
		1621	sub_add_run (sweep_line, right, fy, 0);
		1622	} while (TRUE);
		1623
		1624	if (left->current_sign != +1)
		1625	sub_add_run (sweep_line, left, fy, +1);
		1626	if (right->current_sign != -1)
		1627	sub_add_run (sweep_line, right, fy, -1);
		1628	} while (pos != &sweep_line->active);
		1629	}
		1630
		1631	cairo_always_inline static void
		1632	sub_step (cairo_botor_scan_converter_t *self,
		1633	sweep_line_t *sweep_line)
		1634	{
		1635	if (cairo_list_is_empty (&sweep_line->active))
		1636	return;
		1637
		1638	if (self->fill_rule == CAIRO_FILL_RULE_WINDING)
		1639	sub_nonzero (sweep_line);
		1640	else
		1641	sub_evenodd (sweep_line);
		1642	}
		1643
		1644	static void
		1645	coverage_render_runs (sweep_line_t sweep, edge_t edge,
		1646	cairo_fixed_t y1, cairo_fixed_t y2)
		1647	{
		1648	struct run tail;
		1649	struct run *run = &tail;
		1650
		1651	tail.next = NULL;
		1652	tail.y = y2;
		1653
		1654	/* Order the runs top->bottom */
		1655	while (edge->runs) {
		1656	struct run *r;
		1657
		1658	r = edge->runs;
		1659	edge->runs = r->next;
		1660	r->next = run;
		1661	run = r;
		1662	}
		1663
		1664	if (run->y > y1)
		1665	sub_inc_edge (edge, run->y - y1);
		1666
		1667	do {
		1668	cairo_fixed_t x1, x2;
		1669
		1670	y1 = run->y;
		1671	y2 = run->next->y;
		1672
		1673	x1 = edge->x.quo;
		1674	if (y2 - y1 == STEP_Y)
		1675	full_inc_edge (edge);
		1676	else
		1677	sub_inc_edge (edge, y2 - y1);
		1678	x2 = edge->x.quo;
		1679
		1680	if (run->sign) {
		1681	int ix1, ix2;
		1682
		1683	ix1 = _cairo_fixed_integer_part (x1);
		1684	ix2 = _cairo_fixed_integer_part (x2);
		1685
		1686	/* Edge is entirely within a column? */
		1687	if (likely (ix1 == ix2)) {
		1688	struct cell *cell;
		1689	int frac;
		1690
		1691	frac = _cairo_fixed_fractional_part (x1) +
		1692	_cairo_fixed_fractional_part (x2);
		1693	cell = coverage_find (sweep, ix1);
		1694	cell->covered_height += run->sign * (y2 - y1);
		1695	cell->uncovered_area += run->sign * (y2 - y1) * frac;
		1696	} else {
		1697	coverage_render_cells (sweep, x1, x2, y1, y2, run->sign);
		1698	}
		1699	}
		1700
		1701	run = run->next;
		1702	} while (run->next != NULL);
		1703	}
		1704
		1705	static void
		1706	coverage_render_vertical_runs (sweep_line_t sweep, edge_t edge, cairo_fixed_t y2)
		1707	{
		1708	struct cell *cell;
		1709	struct run *run;
		1710	int height = 0;
		1711
		1712	for (run = edge->runs; run != NULL; run = run->next) {
		1713	if (run->sign)
		1714	height += run->sign * (y2 - run->y);
		1715	y2 = run->y;
		1716	}
		1717
		1718	cell = coverage_find (sweep, _cairo_fixed_integer_part (edge->x.quo));
		1719	cell->covered_height += height;
		1720	cell->uncovered_area += 2 * _cairo_fixed_fractional_part (edge->x.quo) * height;
		1721	}
		1722
		1723	cairo_always_inline static void
		1724	sub_emit (cairo_botor_scan_converter_t *self,
		1725	sweep_line_t *sweep,
		1726	cairo_span_renderer_t *renderer)
		1727	{
		1728	edge_t *edge;
		1729
		1730	sub_step (self, sweep);
		1731
		1732	/* convert the runs into coverages */
		1733
		1734	cairo_list_foreach_entry (edge, edge_t, &sweep->active, link) {
		1735	if (edge->runs == NULL) {
		1736	if (! edge->vertical) {
		1737	if (edge->flags & START) {
		1738	sub_inc_edge (edge,
		1739	STEP_Y - _cairo_fixed_fractional_part (edge->edge.top));
		1740	edge->flags &= ~START;
		1741	} else
		1742	full_inc_edge (edge);
		1743	}
		1744	} else {
		1745	if (edge->vertical) {
		1746	coverage_render_vertical_runs (sweep, edge, STEP_Y);
		1747	} else {
		1748	int y1 = 0;
		1749	if (edge->flags & START) {
		1750	y1 = _cairo_fixed_fractional_part (edge->edge.top);
		1751	edge->flags &= ~START;
		1752	}
		1753	coverage_render_runs (sweep, edge, y1, STEP_Y);
		1754	}
		1755	}
		1756	edge->current_sign = 0;
		1757	edge->runs = NULL;
		1758	}
		1759
		1760	cairo_list_foreach_entry (edge, edge_t, &sweep->stopped, link) {
		1761	int y2 = _cairo_fixed_fractional_part (edge->edge.bottom);
		1762	if (edge->vertical) {
		1763	coverage_render_vertical_runs (sweep, edge, y2);
		1764	} else {
		1765	int y1 = 0;
		1766	if (edge->flags & START)
		1767	y1 = _cairo_fixed_fractional_part (edge->edge.top);
		1768	coverage_render_runs (sweep, edge, y1, y2);
		1769	}
		1770	}
		1771	cairo_list_init (&sweep->stopped);
		1772
		1773	_cairo_freepool_reset (&sweep->runs);
		1774
		1775	render_rows (self, sweep,
		1776	_cairo_fixed_integer_part (sweep->current_row), 1,
		1777	renderer);
		1778	}
		1779
		1780	static void
		1781	sweep_line_init (sweep_line_t *sweep_line,
		1782	event_t **start_events,
		1783	int num_events)
		1784	{
		1785	cairo_list_init (&sweep_line->active);
		1786	cairo_list_init (&sweep_line->stopped);
		1787	sweep_line->insert_cursor = &sweep_line->active;
		1788
		1789	sweep_line->current_row = INT32_MIN;
		1790	sweep_line->current_subrow = INT32_MIN;
		1791
		1792	coverage_init (&sweep_line->coverage);
		1793	_cairo_freepool_init (&sweep_line->runs, sizeof (struct run));
		1794
		1795	start_event_sort (start_events, num_events);
		1796	start_events[num_events] = NULL;
		1797
		1798	sweep_line->queue.start_events = start_events;
		1799
		1800	_cairo_freepool_init (&sweep_line->queue.pool,
		1801	sizeof (queue_event_t));
		1802	pqueue_init (&sweep_line->queue.pq);
		1803	sweep_line->queue.pq.elements[PQ_FIRST_ENTRY] = NULL;
		1804	}
		1805
		1806	static void
		1807	sweep_line_delete (sweep_line_t *sweep_line,
		1808	edge_t *edge)
		1809	{
		1810	if (sweep_line->insert_cursor == &edge->link)
		1811	sweep_line->insert_cursor = edge->link.prev;
		1812
		1813	cairo_list_del (&edge->link);
		1814	if (edge->runs)
		1815	cairo_list_add_tail (&edge->link, &sweep_line->stopped);
		1816	edge->flags \|= STOP;
		1817	}
		1818
		1819	static void
		1820	sweep_line_swap (sweep_line_t *sweep_line,
		1821	edge_t *left,
		1822	edge_t *right)
		1823	{
		1824	right->link.prev = left->link.prev;
		1825	left->link.next = right->link.next;
		1826	right->link.next = &left->link;
		1827	left->link.prev = &right->link;
		1828	left->link.next->prev = &left->link;
		1829	right->link.prev->next = &right->link;
		1830	}
		1831
		1832	static void
		1833	sweep_line_fini (sweep_line_t *sweep_line)
		1834	{
		1835	pqueue_fini (&sweep_line->queue.pq);
		1836	_cairo_freepool_fini (&sweep_line->queue.pool);
		1837	coverage_fini (&sweep_line->coverage);
		1838	_cairo_freepool_fini (&sweep_line->runs);
		1839	}
		1840
		1841	static cairo_status_t
		1842	botor_generate (cairo_botor_scan_converter_t *self,
		1843	event_t **start_events,
		1844	cairo_span_renderer_t *renderer)
		1845	{
		1846	cairo_status_t status;
		1847	sweep_line_t sweep_line;
		1848	cairo_fixed_t ybot;
		1849	event_t *event;
		1850	cairo_list_t left, right;
		1851	edge_t e1, e2;
		1852	int bottom;
		1853
		1854	sweep_line_init (&sweep_line, start_events, self->num_edges);
		1855	if ((status = setjmp (sweep_line.unwind)))
		1856	goto unwind;
		1857
		1858	ybot = self->extents.p2.y;
		1859	sweep_line.current_subrow = self->extents.p1.y;
		1860	sweep_line.current_row = _cairo_fixed_floor (self->extents.p1.y);
		1861	event = *sweep_line.queue.start_events++;
		1862	do {
		1863	/* Can we process a full step in one go? */
		1864	if (event->y >= sweep_line.current_row + STEP_Y) {
		1865	bottom = _cairo_fixed_floor (event->y);
		1866	full_step (self, &sweep_line, bottom, renderer);
		1867	sweep_line.current_row = bottom;
		1868	sweep_line.current_subrow = bottom;
		1869	}
		1870
		1871	do {
		1872	if (event->y > sweep_line.current_subrow) {
		1873	sub_step (self, &sweep_line);
		1874	sweep_line.current_subrow = event->y;
		1875	}
		1876
		1877	do {
		1878	/* Update the active list using Bentley-Ottmann */
		1879	switch (event->type) {
		1880	case EVENT_TYPE_START:
		1881	e1 = ((start_event_t *) event)->edge;
		1882
		1883	sweep_line_insert (&sweep_line, e1);
		1884	event_insert_stop (&sweep_line, e1);
		1885
		1886	left = e1->link.prev;
		1887	right = e1->link.next;
		1888
		1889	if (left != &sweep_line.active) {
		1890	event_insert_if_intersect_below_current_y (&sweep_line,
		1891	link_to_edge (left), e1);
		1892	}
		1893
		1894	if (right != &sweep_line.active) {
		1895	event_insert_if_intersect_below_current_y (&sweep_line,
		1896	e1, link_to_edge (right));
		1897	}
		1898
		1899	break;
		1900
		1901	case EVENT_TYPE_STOP:
		1902	e1 = ((queue_event_t *) event)->e1;
		1903	event_delete (&sweep_line, event);
		1904
		1905	left = e1->link.prev;
		1906	right = e1->link.next;
		1907
		1908	sweep_line_delete (&sweep_line, e1);
		1909
		1910	if (left != &sweep_line.active &&
		1911	right != &sweep_line.active)
		1912	{
		1913	event_insert_if_intersect_below_current_y (&sweep_line,
		1914	link_to_edge (left),
		1915	link_to_edge (right));
		1916	}
		1917
		1918	break;
		1919
		1920	case EVENT_TYPE_INTERSECTION:
		1921	e1 = ((queue_event_t *) event)->e1;
		1922	e2 = ((queue_event_t *) event)->e2;
		1923
		1924	event_delete (&sweep_line, event);
		1925	if (e1->flags & STOP)
		1926	break;
		1927	if (e2->flags & STOP)
		1928	break;
		1929
		1930	/* skip this intersection if its edges are not adjacent */
		1931	if (&e2->link != e1->link.next)
		1932	break;
		1933
		1934	left = e1->link.prev;
		1935	right = e2->link.next;
		1936
		1937	sweep_line_swap (&sweep_line, e1, e2);
		1938
		1939	/* after the swap e2 is left of e1 */
		1940	if (left != &sweep_line.active) {
		1941	event_insert_if_intersect_below_current_y (&sweep_line,
		1942	link_to_edge (left), e2);
		1943	}
		1944
		1945	if (right != &sweep_line.active) {
		1946	event_insert_if_intersect_below_current_y (&sweep_line,
		1947	e1, link_to_edge (right));
		1948	}
		1949
		1950	break;
		1951	}
		1952
		1953	event = event_next (&sweep_line);
		1954	if (event == NULL)
		1955	goto end;
		1956	} while (event->y == sweep_line.current_subrow);
		1957	} while (event->y < sweep_line.current_row + STEP_Y);
		1958
		1959	bottom = sweep_line.current_row + STEP_Y;
		1960	sub_emit (self, &sweep_line, renderer);
		1961	sweep_line.current_subrow = bottom;
		1962	sweep_line.current_row = sweep_line.current_subrow;
		1963	} while (TRUE);
		1964
		1965	end:
		1966	/* flush any partial spans */
		1967	if (sweep_line.current_subrow != sweep_line.current_row) {
		1968	sub_emit (self, &sweep_line, renderer);
		1969	sweep_line.current_row += STEP_Y;
		1970	sweep_line.current_subrow = sweep_line.current_row;
		1971	}
		1972	/* clear the rest */
		1973	if (sweep_line.current_subrow < ybot) {
		1974	bottom = _cairo_fixed_integer_part (sweep_line.current_row);
		1975	status = renderer->render_rows (renderer,
		1976	bottom, _cairo_fixed_integer_ceil (ybot) - bottom,
		1977	NULL, 0);
		1978	}
		1979
		1980	unwind:
		1981	sweep_line_fini (&sweep_line);
		1982
		1983	return status;
		1984	}
		1985
		1986	static cairo_status_t
		1987	_cairo_botor_scan_converter_generate (void *converter,
		1988	cairo_span_renderer_t *renderer)
		1989	{
		1990	cairo_botor_scan_converter_t *self = converter;
		1991	start_event_t stack_events[CAIRO_STACK_ARRAY_LENGTH (start_event_t)];
		1992	start_event_t *events;
		1993	event_t *stack_event_ptrs[ARRAY_LENGTH (stack_events) + 1];
		1994	event_t **event_ptrs;
		1995	struct _cairo_botor_scan_converter_chunk *chunk;
		1996	cairo_status_t status;
		1997	int num_events;
		1998	int i, j;
		1999
		2000	num_events = self->num_edges;
		2001	if (unlikely (0 == num_events)) {
		2002	return renderer->render_rows (renderer,
		2003	_cairo_fixed_integer_floor (self->extents.p1.y),
		2004	_cairo_fixed_integer_ceil (self->extents.p2.y) -
		2005	_cairo_fixed_integer_floor (self->extents.p1.y),
		2006	NULL, 0);
		2007	}
		2008
		2009	events = stack_events;
		2010	event_ptrs = stack_event_ptrs;
		2011	if (unlikely (num_events >= ARRAY_LENGTH (stack_events))) {
		2012	events = _cairo_malloc_ab_plus_c (num_events,
		2013	sizeof (start_event_t) + sizeof (event_t *),
		2014	sizeof (event_t *));
		2015	if (unlikely (events == NULL))
		2016	return _cairo_error (CAIRO_STATUS_NO_MEMORY);
		2017
		2018	event_ptrs = (event_t **) (events + num_events);
		2019	}
		2020
		2021	j = 0;
		2022	for (chunk = &self->chunks; chunk != NULL; chunk = chunk->next) {
		2023	edge_t *edge;
		2024
		2025	edge = chunk->base;
		2026	for (i = 0; i < chunk->count; i++) {
		2027	event_ptrs[j] = (event_t *) &events[j];
		2028
		2029	events[j].y = edge->edge.top;
		2030	events[j].type = EVENT_TYPE_START;
		2031	events[j].edge = edge;
		2032
		2033	edge++, j++;
		2034	}
		2035	}
		2036
		2037	status = botor_generate (self, event_ptrs, renderer);
		2038
		2039	if (events != stack_events)
		2040	free (events);
		2041
		2042	return status;
		2043	}
		2044
		2045	static edge_t *
		2046	botor_allocate_edge (cairo_botor_scan_converter_t *self)
		2047	{
		2048	struct _cairo_botor_scan_converter_chunk *chunk;
		2049
		2050	chunk = self->tail;
		2051	if (chunk->count == chunk->size) {
		2052	int size;
		2053
		2054	size = chunk->size * 2;
		2055	chunk->next = _cairo_malloc_ab_plus_c (size,
		2056	sizeof (edge_t),
		2057	sizeof (struct _cairo_botor_scan_converter_chunk));
		2058	if (unlikely (chunk->next == NULL))
		2059	return NULL;
		2060
		2061	chunk = chunk->next;
		2062	chunk->next = NULL;
		2063	chunk->count = 0;
		2064	chunk->size = size;
		2065	chunk->base = chunk + 1;
		2066	self->tail = chunk;
		2067	}
		2068
		2069	return (edge_t *) chunk->base + chunk->count++;
		2070	}
		2071
		2072	static cairo_status_t
		2073	botor_add_edge (cairo_botor_scan_converter_t *self,
		2074	const cairo_edge_t *edge)
		2075	{
		2076	edge_t *e;
		2077	cairo_fixed_t dx, dy;
		2078
		2079	e = botor_allocate_edge (self);
		2080	if (unlikely (e == NULL))
		2081	return _cairo_error (CAIRO_STATUS_NO_MEMORY);
		2082
		2083	cairo_list_init (&e->link);
		2084	e->edge = *edge;
		2085
		2086	dx = edge->line.p2.x - edge->line.p1.x;
		2087	dy = edge->line.p2.y - edge->line.p1.y;
		2088	e->dy = dy;
		2089
		2090	if (dx == 0) {
		2091	e->vertical = TRUE;
		2092	e->x.quo = edge->line.p1.x;
		2093	e->x.rem = 0;
		2094	e->dxdy.quo = 0;
		2095	e->dxdy.rem = 0;
		2096	e->dxdy_full.quo = 0;
		2097	e->dxdy_full.rem = 0;
		2098	} else {
		2099	e->vertical = FALSE;
		2100	e->dxdy = floored_divrem (dx, dy);
		2101	if (edge->top == edge->line.p1.y) {
		2102	e->x.quo = edge->line.p1.x;
		2103	e->x.rem = 0;
		2104	} else {
		2105	e->x = floored_muldivrem (edge->top - edge->line.p1.y,
		2106	dx, dy);
		2107	e->x.quo += edge->line.p1.x;
		2108	}
		2109
		2110	if (_cairo_fixed_integer_part (edge->bottom) - _cairo_fixed_integer_part (edge->top) > 1) {
		2111	e->dxdy_full = floored_muldivrem (STEP_Y, dx, dy);
		2112	} else {
		2113	e->dxdy_full.quo = 0;
		2114	e->dxdy_full.rem = 0;
		2115	}
		2116	}
		2117
		2118	e->x.rem = -e->dy;
		2119	e->current_sign = 0;
		2120	e->runs = NULL;
		2121	e->flags = START;
		2122
		2123	self->num_edges++;
		2124
		2125	return CAIRO_STATUS_SUCCESS;
		2126	}
		2127
		2128	static cairo_status_t
		2129	_cairo_botor_scan_converter_add_edge (void *converter,
		2130	const cairo_point_t *p1,
		2131	const cairo_point_t *p2,
		2132	int top, int bottom,
		2133	int dir)
		2134	{
		2135	cairo_botor_scan_converter_t *self = converter;
		2136	cairo_edge_t edge;
		2137
		2138	edge.line.p1 = *p1;
		2139	edge.line.p2 = *p2;
		2140	edge.top = top;
		2141	edge.bottom = bottom;
		2142	edge.dir = dir;
		2143
		2144	return botor_add_edge (self, &edge);
		2145	}
		2146
		2147	static cairo_status_t
		2148	_cairo_botor_scan_converter_add_polygon (void *converter,
		2149	const cairo_polygon_t *polygon)
		2150	{
		2151	cairo_botor_scan_converter_t *self = converter;
		2152	cairo_status_t status;
		2153	int i;
		2154
		2155	for (i = 0; i < polygon->num_edges; i++) {
		2156	status = botor_add_edge (self, &polygon->edges[i]);
		2157	if (unlikely (status))
		2158	return status;
		2159	}
		2160
		2161	return CAIRO_STATUS_SUCCESS;
		2162	}
		2163
		2164	static void
		2165	_cairo_botor_scan_converter_destroy (void *converter)
		2166	{
		2167	cairo_botor_scan_converter_t *self = converter;
		2168	struct _cairo_botor_scan_converter_chunk chunk, next;
		2169
		2170	for (chunk = self->chunks.next; chunk != NULL; chunk = next) {
		2171	next = chunk->next;
		2172	free (chunk);
		2173	}
		2174	}
		2175
		2176	void
		2177	_cairo_botor_scan_converter_init (cairo_botor_scan_converter_t *self,
		2178	const cairo_box_t *extents,
		2179	cairo_fill_rule_t fill_rule)
		2180	{
		2181	self->base.destroy = _cairo_botor_scan_converter_destroy;
		2182	self->base.add_edge = _cairo_botor_scan_converter_add_edge;
		2183	self->base.add_polygon = _cairo_botor_scan_converter_add_polygon;
		2184	self->base.generate = _cairo_botor_scan_converter_generate;
		2185
		2186	self->extents = *extents;
		2187	self->fill_rule = fill_rule;
		2188
		2189	self->xmin = _cairo_fixed_integer_floor (extents->p1.x);
		2190	self->xmax = _cairo_fixed_integer_ceil (extents->p2.x);
		2191
		2192	self->chunks.base = self->buf;
		2193	self->chunks.next = NULL;
		2194	self->chunks.count = 0;
		2195	self->chunks.size = sizeof (self->buf) / sizeof (edge_t);
		2196	self->tail = &self->chunks;
		2197
		2198	self->num_edges = 0;
		2199	}

Subversion Repositories Kolibri OS

(root)/programs/develop/libraries/cairo/src/cairo-botor-scan-converter.c @ 3959 – Rev 1892