This repository has been archived on 2023-10-09. You can view files and clone it, but cannot push or open issues or pull requests.
Files
blender-archive/source/blender/blenkernel/intern/mesh_convert.c

1527 lines
46 KiB
C
Raw Normal View History

/*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
* of the License, or (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software Foundation,
* Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
*/
/** \file
* \ingroup bke
*/
#include "CLG_log.h"
2018-05-08 19:09:35 +02:00
#include "MEM_guardedalloc.h"
#include "DNA_scene_types.h"
#include "DNA_key_types.h"
#include "DNA_material_types.h"
#include "DNA_meta_types.h"
#include "DNA_object_types.h"
#include "DNA_mesh_types.h"
#include "DNA_curve_types.h"
#include "BLI_utildefines.h"
#include "BLI_math.h"
#include "BLI_listbase.h"
#include "BLI_edgehash.h"
#include "BLI_string.h"
#include "BKE_main.h"
#include "BKE_DerivedMesh.h"
#include "BKE_editmesh.h"
#include "BKE_key.h"
#include "BKE_library_query.h"
#include "BKE_mesh.h"
#include "BKE_mesh_runtime.h"
#include "BKE_modifier.h"
#include "BKE_displist.h"
#include "BKE_library.h"
#include "BKE_material.h"
#include "BKE_mball.h"
2018-05-08 19:09:35 +02:00
/* these 2 are only used by conversion functions */
#include "BKE_curve.h"
/* -- */
#include "BKE_object.h"
#include "DEG_depsgraph.h"
2018-05-08 19:09:35 +02:00
#include "DEG_depsgraph_query.h"
/* Define for cases when you want extra validation of mesh
* after certain modifications.
*/
// #undef VALIDATE_MESH
#ifdef VALIDATE_MESH
# define ASSERT_IS_VALID_MESH(mesh) \
(BLI_assert((mesh == NULL) || (BKE_mesh_is_valid(mesh) == true)))
#else
# define ASSERT_IS_VALID_MESH(mesh)
#endif
static CLG_LogRef LOG = {"bke.mesh_convert"};
void BKE_mesh_from_metaball(ListBase *lb, Mesh *me)
{
DispList *dl;
MVert *mvert;
MLoop *mloop, *allloop;
MPoly *mpoly;
const float *nors, *verts;
int a, *index;
dl = lb->first;
if (dl == NULL) {
return;
}
if (dl->type == DL_INDEX4) {
mvert = CustomData_add_layer(&me->vdata, CD_MVERT, CD_CALLOC, NULL, dl->nr);
allloop = mloop = CustomData_add_layer(&me->ldata, CD_MLOOP, CD_CALLOC, NULL, dl->parts * 4);
mpoly = CustomData_add_layer(&me->pdata, CD_MPOLY, CD_CALLOC, NULL, dl->parts);
me->mvert = mvert;
me->mloop = mloop;
me->mpoly = mpoly;
me->totvert = dl->nr;
me->totpoly = dl->parts;
a = dl->nr;
nors = dl->nors;
verts = dl->verts;
while (a--) {
copy_v3_v3(mvert->co, verts);
normal_float_to_short_v3(mvert->no, nors);
mvert++;
nors += 3;
verts += 3;
}
a = dl->parts;
index = dl->index;
while (a--) {
int count = index[2] != index[3] ? 4 : 3;
mloop[0].v = index[0];
mloop[1].v = index[1];
mloop[2].v = index[2];
if (count == 4) {
mloop[3].v = index[3];
}
mpoly->totloop = count;
mpoly->loopstart = (int)(mloop - allloop);
mpoly->flag = ME_SMOOTH;
mpoly++;
mloop += count;
me->totloop += count;
index += 4;
}
BKE_mesh_update_customdata_pointers(me, true);
BKE_mesh_calc_normals(me);
BKE_mesh_calc_edges(me, true, false);
}
}
/**
* Specialized function to use when we _know_ existing edges don't overlap with poly edges.
*/
2018-07-13 08:36:10 +02:00
static void make_edges_mdata_extend(
MEdge **r_alledge, int *r_totedge, const MPoly *mpoly, MLoop *mloop, const int totpoly)
{
int totedge = *r_totedge;
int totedge_new;
EdgeHash *eh;
unsigned int eh_reserve;
const MPoly *mp;
int i;
eh_reserve = max_ii(totedge, BLI_EDGEHASH_SIZE_GUESS_FROM_POLYS(totpoly));
eh = BLI_edgehash_new_ex(__func__, eh_reserve);
for (i = 0, mp = mpoly; i < totpoly; i++, mp++) {
BKE_mesh_poly_edgehash_insert(eh, mp, mloop + mp->loopstart);
}
totedge_new = BLI_edgehash_len(eh);
#ifdef DEBUG
/* ensure that there's no overlap! */
if (totedge_new) {
MEdge *medge = *r_alledge;
for (i = 0; i < totedge; i++, medge++) {
BLI_assert(BLI_edgehash_haskey(eh, medge->v1, medge->v2) == false);
}
}
#endif
if (totedge_new) {
EdgeHashIterator *ehi;
MEdge *medge;
unsigned int e_index = totedge;
*r_alledge = medge = (*r_alledge ?
MEM_reallocN(*r_alledge, sizeof(MEdge) * (totedge + totedge_new)) :
MEM_calloc_arrayN(totedge_new, sizeof(MEdge), __func__));
medge += totedge;
totedge += totedge_new;
/* --- */
for (ehi = BLI_edgehashIterator_new(eh); BLI_edgehashIterator_isDone(ehi) == false;
BLI_edgehashIterator_step(ehi), ++medge, e_index++) {
BLI_edgehashIterator_getKey(ehi, &medge->v1, &medge->v2);
BLI_edgehashIterator_setValue(ehi, POINTER_FROM_UINT(e_index));
medge->crease = medge->bweight = 0;
medge->flag = ME_EDGEDRAW | ME_EDGERENDER;
}
BLI_edgehashIterator_free(ehi);
*r_totedge = totedge;
for (i = 0, mp = mpoly; i < totpoly; i++, mp++) {
MLoop *l = &mloop[mp->loopstart];
MLoop *l_prev = (l + (mp->totloop - 1));
int j;
for (j = 0; j < mp->totloop; j++, l++) {
/* lookup hashed edge index */
l_prev->e = POINTER_AS_UINT(BLI_edgehash_lookup(eh, l_prev->v, l->v));
l_prev = l;
}
}
}
BLI_edgehash_free(eh, NULL);
}
/* Initialize mverts, medges and, faces for converting nurbs to mesh and derived mesh */
/* return non-zero on error */
int BKE_mesh_nurbs_to_mdata(Object *ob,
MVert **r_allvert,
int *r_totvert,
MEdge **r_alledge,
int *r_totedge,
MLoop **r_allloop,
MPoly **r_allpoly,
int *r_totloop,
int *r_totpoly)
{
ListBase disp = {NULL, NULL};
if (ob->runtime.curve_cache) {
disp = ob->runtime.curve_cache->disp;
}
return BKE_mesh_nurbs_displist_to_mdata(ob,
&disp,
r_allvert,
r_totvert,
r_alledge,
r_totedge,
r_allloop,
r_allpoly,
NULL,
r_totloop,
r_totpoly);
}
/* BMESH: this doesn't calculate all edges from polygons,
* only free standing edges are calculated */
/* Initialize mverts, medges and, faces for converting nurbs to mesh and derived mesh */
/* use specified dispbase */
int BKE_mesh_nurbs_displist_to_mdata(Object *ob,
const ListBase *dispbase,
MVert **r_allvert,
int *r_totvert,
MEdge **r_alledge,
int *r_totedge,
MLoop **r_allloop,
MPoly **r_allpoly,
MLoopUV **r_alluv,
int *r_totloop,
int *r_totpoly)
{
Curve *cu = ob->data;
DispList *dl;
MVert *mvert;
MPoly *mpoly;
MLoop *mloop;
MLoopUV *mloopuv = NULL;
MEdge *medge;
const float *data;
int a, b, ofs, vertcount, startvert, totvert = 0, totedge = 0, totloop = 0, totpoly = 0;
int p1, p2, p3, p4, *index;
const bool conv_polys = ((CU_DO_2DFILL(cu) ==
false) || /* 2d polys are filled with DL_INDEX3 displists */
(ob->type == OB_SURF)); /* surf polys are never filled */
/* count */
dl = dispbase->first;
while (dl) {
if (dl->type == DL_SEGM) {
totvert += dl->parts * dl->nr;
totedge += dl->parts * (dl->nr - 1);
}
else if (dl->type == DL_POLY) {
if (conv_polys) {
totvert += dl->parts * dl->nr;
totedge += dl->parts * dl->nr;
}
}
else if (dl->type == DL_SURF) {
int tot;
totvert += dl->parts * dl->nr;
tot = (dl->parts - 1 + ((dl->flag & DL_CYCL_V) == 2)) *
(dl->nr - 1 + (dl->flag & DL_CYCL_U));
totpoly += tot;
totloop += tot * 4;
}
else if (dl->type == DL_INDEX3) {
int tot;
totvert += dl->nr;
tot = dl->parts;
totpoly += tot;
totloop += tot * 3;
}
dl = dl->next;
}
if (totvert == 0) {
/* error("can't convert"); */
/* Make Sure you check ob->data is a curve */
return -1;
}
*r_allvert = mvert = MEM_calloc_arrayN(totvert, sizeof(MVert), "nurbs_init mvert");
*r_alledge = medge = MEM_calloc_arrayN(totedge, sizeof(MEdge), "nurbs_init medge");
*r_allloop = mloop = MEM_calloc_arrayN(
totpoly, 4 * sizeof(MLoop), "nurbs_init mloop"); // totloop
*r_allpoly = mpoly = MEM_calloc_arrayN(totpoly, sizeof(MPoly), "nurbs_init mloop");
if (r_alluv) {
*r_alluv = mloopuv = MEM_calloc_arrayN(totpoly, 4 * sizeof(MLoopUV), "nurbs_init mloopuv");
}
/* verts and faces */
vertcount = 0;
dl = dispbase->first;
while (dl) {
const bool is_smooth = (dl->rt & CU_SMOOTH) != 0;
if (dl->type == DL_SEGM) {
startvert = vertcount;
a = dl->parts * dl->nr;
data = dl->verts;
while (a--) {
copy_v3_v3(mvert->co, data);
data += 3;
vertcount++;
mvert++;
}
for (a = 0; a < dl->parts; a++) {
ofs = a * dl->nr;
for (b = 1; b < dl->nr; b++) {
medge->v1 = startvert + ofs + b - 1;
medge->v2 = startvert + ofs + b;
medge->flag = ME_LOOSEEDGE | ME_EDGERENDER | ME_EDGEDRAW;
medge++;
}
}
}
else if (dl->type == DL_POLY) {
if (conv_polys) {
startvert = vertcount;
a = dl->parts * dl->nr;
data = dl->verts;
while (a--) {
copy_v3_v3(mvert->co, data);
data += 3;
vertcount++;
mvert++;
}
for (a = 0; a < dl->parts; a++) {
ofs = a * dl->nr;
for (b = 0; b < dl->nr; b++) {
medge->v1 = startvert + ofs + b;
if (b == dl->nr - 1) {
medge->v2 = startvert + ofs;
}
else {
medge->v2 = startvert + ofs + b + 1;
}
medge->flag = ME_LOOSEEDGE | ME_EDGERENDER | ME_EDGEDRAW;
medge++;
}
}
}
}
else if (dl->type == DL_INDEX3) {
startvert = vertcount;
a = dl->nr;
data = dl->verts;
while (a--) {
copy_v3_v3(mvert->co, data);
data += 3;
vertcount++;
mvert++;
}
a = dl->parts;
index = dl->index;
while (a--) {
mloop[0].v = startvert + index[0];
mloop[1].v = startvert + index[2];
mloop[2].v = startvert + index[1];
mpoly->loopstart = (int)(mloop - (*r_allloop));
mpoly->totloop = 3;
mpoly->mat_nr = dl->col;
if (mloopuv) {
int i;
for (i = 0; i < 3; i++, mloopuv++) {
mloopuv->uv[0] = (mloop[i].v - startvert) / (float)(dl->nr - 1);
mloopuv->uv[1] = 0.0f;
}
}
if (is_smooth) {
mpoly->flag |= ME_SMOOTH;
}
mpoly++;
mloop += 3;
index += 3;
}
}
else if (dl->type == DL_SURF) {
startvert = vertcount;
a = dl->parts * dl->nr;
data = dl->verts;
while (a--) {
copy_v3_v3(mvert->co, data);
data += 3;
vertcount++;
mvert++;
}
for (a = 0; a < dl->parts; a++) {
if ((dl->flag & DL_CYCL_V) == 0 && a == dl->parts - 1) {
break;
}
if (dl->flag & DL_CYCL_U) { /* p2 -> p1 -> */
p1 = startvert + dl->nr * a; /* p4 -> p3 -> */
p2 = p1 + dl->nr - 1; /* -----> next row */
p3 = p1 + dl->nr;
p4 = p2 + dl->nr;
b = 0;
}
else {
p2 = startvert + dl->nr * a;
p1 = p2 + 1;
p4 = p2 + dl->nr;
p3 = p1 + dl->nr;
b = 1;
}
if ((dl->flag & DL_CYCL_V) && a == dl->parts - 1) {
p3 -= dl->parts * dl->nr;
p4 -= dl->parts * dl->nr;
}
for (; b < dl->nr; b++) {
mloop[0].v = p1;
mloop[1].v = p3;
mloop[2].v = p4;
mloop[3].v = p2;
mpoly->loopstart = (int)(mloop - (*r_allloop));
mpoly->totloop = 4;
mpoly->mat_nr = dl->col;
if (mloopuv) {
int orco_sizeu = dl->nr - 1;
int orco_sizev = dl->parts - 1;
int i;
/* exception as handled in convertblender.c too */
if (dl->flag & DL_CYCL_U) {
orco_sizeu++;
if (dl->flag & DL_CYCL_V) {
orco_sizev++;
}
}
else if (dl->flag & DL_CYCL_V) {
orco_sizev++;
}
for (i = 0; i < 4; i++, mloopuv++) {
/* find uv based on vertex index into grid array */
int v = mloop[i].v - startvert;
mloopuv->uv[0] = (v / dl->nr) / (float)orco_sizev;
mloopuv->uv[1] = (v % dl->nr) / (float)orco_sizeu;
/* cyclic correction */
if ((i == 1 || i == 2) && mloopuv->uv[0] == 0.0f) {
mloopuv->uv[0] = 1.0f;
}
if ((i == 0 || i == 1) && mloopuv->uv[1] == 0.0f) {
mloopuv->uv[1] = 1.0f;
}
}
}
if (is_smooth) {
mpoly->flag |= ME_SMOOTH;
}
mpoly++;
mloop += 4;
p4 = p3;
p3++;
p2 = p1;
p1++;
}
}
}
dl = dl->next;
}
if (totpoly) {
make_edges_mdata_extend(r_alledge, &totedge, *r_allpoly, *r_allloop, totpoly);
}
*r_totpoly = totpoly;
*r_totloop = totloop;
*r_totedge = totedge;
*r_totvert = totvert;
return 0;
}
2018-05-08 19:09:35 +02:00
Mesh *BKE_mesh_new_nomain_from_curve_displist(Object *ob, ListBase *dispbase)
{
Curve *cu = ob->data;
Mesh *mesh;
MVert *allvert;
MEdge *alledge;
MLoop *allloop;
MPoly *allpoly;
MLoopUV *alluv = NULL;
int totvert, totedge, totloop, totpoly;
bool use_orco_uv = (cu->flag & CU_UV_ORCO) != 0;
if (BKE_mesh_nurbs_displist_to_mdata(ob,
dispbase,
&allvert,
&totvert,
&alledge,
&totedge,
&allloop,
&allpoly,
(use_orco_uv) ? &alluv : NULL,
&totloop,
&totpoly) != 0) {
/* Error initializing mdata. This often happens when curve is empty */
return BKE_mesh_new_nomain(0, 0, 0, 0, 0);
}
mesh = BKE_mesh_new_nomain(totvert, totedge, 0, totloop, totpoly);
mesh->runtime.cd_dirty_vert |= CD_MASK_NORMAL;
memcpy(mesh->mvert, allvert, totvert * sizeof(MVert));
memcpy(mesh->medge, alledge, totedge * sizeof(MEdge));
memcpy(mesh->mloop, allloop, totloop * sizeof(MLoop));
memcpy(mesh->mpoly, allpoly, totpoly * sizeof(MPoly));
if (alluv) {
const char *uvname = "Orco";
CustomData_add_layer_named(&mesh->ldata, CD_MLOOPUV, CD_ASSIGN, alluv, totloop, uvname);
}
MEM_freeN(allvert);
MEM_freeN(alledge);
MEM_freeN(allloop);
MEM_freeN(allpoly);
return mesh;
2018-05-08 19:09:35 +02:00
}
Mesh *BKE_mesh_new_nomain_from_curve(Object *ob)
{
ListBase disp = {NULL, NULL};
2018-05-08 19:09:35 +02:00
if (ob->runtime.curve_cache) {
disp = ob->runtime.curve_cache->disp;
}
2018-05-08 19:09:35 +02:00
return BKE_mesh_new_nomain_from_curve_displist(ob, &disp);
2018-05-08 19:09:35 +02:00
}
/* this may fail replacing ob->data, be sure to check ob->type */
void BKE_mesh_from_nurbs_displist(Main *bmain,
Object *ob,
ListBase *dispbase,
const bool use_orco_uv,
const char *obdata_name,
bool temporary)
{
Object *ob1;
Mesh *me_eval = ob->runtime.mesh_eval;
Mesh *me;
Curve *cu;
MVert *allvert = NULL;
MEdge *alledge = NULL;
MLoop *allloop = NULL;
MLoopUV *alluv = NULL;
MPoly *allpoly = NULL;
int totvert, totedge, totloop, totpoly;
cu = ob->data;
if (me_eval == NULL) {
if (BKE_mesh_nurbs_displist_to_mdata(ob,
dispbase,
&allvert,
&totvert,
&alledge,
&totedge,
&allloop,
&allpoly,
(use_orco_uv) ? &alluv : NULL,
&totloop,
&totpoly) != 0) {
/* Error initializing */
return;
}
/* make mesh */
if (bmain != NULL) {
me = BKE_mesh_add(bmain, obdata_name);
}
else {
me = BKE_id_new_nomain(ID_ME, obdata_name);
}
me->totvert = totvert;
me->totedge = totedge;
me->totloop = totloop;
me->totpoly = totpoly;
me->mvert = CustomData_add_layer(&me->vdata, CD_MVERT, CD_ASSIGN, allvert, me->totvert);
me->medge = CustomData_add_layer(&me->edata, CD_MEDGE, CD_ASSIGN, alledge, me->totedge);
me->mloop = CustomData_add_layer(&me->ldata, CD_MLOOP, CD_ASSIGN, allloop, me->totloop);
me->mpoly = CustomData_add_layer(&me->pdata, CD_MPOLY, CD_ASSIGN, allpoly, me->totpoly);
if (alluv) {
const char *uvname = "Orco";
me->mloopuv = CustomData_add_layer_named(
&me->ldata, CD_MLOOPUV, CD_ASSIGN, alluv, me->totloop, uvname);
}
BKE_mesh_calc_normals(me);
}
else {
if (bmain != NULL) {
me = BKE_mesh_add(bmain, obdata_name);
}
else {
me = BKE_id_new_nomain(ID_ME, obdata_name);
}
ob->runtime.mesh_eval = NULL;
BKE_mesh_nomain_to_mesh(me_eval, me, ob, &CD_MASK_MESH, true);
}
me->totcol = cu->totcol;
me->mat = cu->mat;
/* Copy evaluated texture space from curve to mesh.
*
* Note that we disable auto texture space feature since that will cause
* texture space to evaluate differently for curve and mesh, since curve
* uses CV to calculate bounding box, and mesh uses what is coming from
* tessellated curve.
*/
me->texflag = cu->texflag & ~CU_AUTOSPACE;
copy_v3_v3(me->loc, cu->loc);
copy_v3_v3(me->size, cu->size);
copy_v3_v3(me->rot, cu->rot);
BKE_mesh_texspace_calc(me);
cu->mat = NULL;
cu->totcol = 0;
/* Do not decrement ob->data usercount here,
* it's done at end of func with BKE_id_free_us() call. */
ob->data = me;
ob->type = OB_MESH;
/* other users */
if (bmain != NULL) {
ob1 = bmain->objects.first;
while (ob1) {
if (ob1->data == cu) {
ob1->type = OB_MESH;
id_us_min((ID *)ob1->data);
ob1->data = ob->data;
id_us_plus((ID *)ob1->data);
}
ob1 = ob1->id.next;
}
}
if (temporary) {
/* For temporary objects in BKE_mesh_new_from_object don't remap
* the entire scene with associated depsgraph updates, which are
* problematic for renderers exporting data. */
BKE_id_free(NULL, cu);
}
else {
BKE_id_free_us(bmain, cu);
}
}
void BKE_mesh_from_nurbs(Main *bmain, Object *ob)
{
Curve *cu = (Curve *)ob->data;
bool use_orco_uv = (cu->flag & CU_UV_ORCO) != 0;
ListBase disp = {NULL, NULL};
if (ob->runtime.curve_cache) {
disp = ob->runtime.curve_cache->disp;
}
BKE_mesh_from_nurbs_displist(bmain, ob, &disp, use_orco_uv, cu->id.name, false);
}
typedef struct EdgeLink {
struct EdgeLink *next, *prev;
void *edge;
} EdgeLink;
typedef struct VertLink {
Link *next, *prev;
unsigned int index;
} VertLink;
static void prependPolyLineVert(ListBase *lb, unsigned int index)
{
VertLink *vl = MEM_callocN(sizeof(VertLink), "VertLink");
vl->index = index;
BLI_addhead(lb, vl);
}
static void appendPolyLineVert(ListBase *lb, unsigned int index)
{
VertLink *vl = MEM_callocN(sizeof(VertLink), "VertLink");
vl->index = index;
BLI_addtail(lb, vl);
}
void BKE_mesh_to_curve_nurblist(const Mesh *me, ListBase *nurblist, const int edge_users_test)
{
MVert *mvert = me->mvert;
MEdge *med, *medge = me->medge;
MPoly *mp, *mpoly = me->mpoly;
MLoop *mloop = me->mloop;
int medge_len = me->totedge;
int mpoly_len = me->totpoly;
int totedges = 0;
int i;
/* only to detect edge polylines */
int *edge_users;
ListBase edges = {NULL, NULL};
/* get boundary edges */
edge_users = MEM_calloc_arrayN(medge_len, sizeof(int), __func__);
for (i = 0, mp = mpoly; i < mpoly_len; i++, mp++) {
MLoop *ml = &mloop[mp->loopstart];
int j;
for (j = 0; j < mp->totloop; j++, ml++) {
edge_users[ml->e]++;
}
}
/* create edges from all faces (so as to find edges not in any faces) */
med = medge;
for (i = 0; i < medge_len; i++, med++) {
if (edge_users[i] == edge_users_test) {
EdgeLink *edl = MEM_callocN(sizeof(EdgeLink), "EdgeLink");
edl->edge = med;
BLI_addtail(&edges, edl);
totedges++;
}
}
MEM_freeN(edge_users);
if (edges.first) {
while (edges.first) {
/* each iteration find a polyline and add this as a nurbs poly spline */
ListBase polyline = {NULL, NULL}; /* store a list of VertLink's */
bool closed = false;
int totpoly = 0;
MEdge *med_current = ((EdgeLink *)edges.last)->edge;
unsigned int startVert = med_current->v1;
unsigned int endVert = med_current->v2;
bool ok = true;
appendPolyLineVert(&polyline, startVert);
totpoly++;
appendPolyLineVert(&polyline, endVert);
totpoly++;
BLI_freelinkN(&edges, edges.last);
totedges--;
while (ok) { /* while connected edges are found... */
EdgeLink *edl = edges.last;
ok = false;
while (edl) {
EdgeLink *edl_prev = edl->prev;
med = edl->edge;
if (med->v1 == endVert) {
endVert = med->v2;
appendPolyLineVert(&polyline, med->v2);
totpoly++;
BLI_freelinkN(&edges, edl);
totedges--;
ok = true;
}
else if (med->v2 == endVert) {
endVert = med->v1;
appendPolyLineVert(&polyline, endVert);
totpoly++;
BLI_freelinkN(&edges, edl);
totedges--;
ok = true;
}
else if (med->v1 == startVert) {
startVert = med->v2;
prependPolyLineVert(&polyline, startVert);
totpoly++;
BLI_freelinkN(&edges, edl);
totedges--;
ok = true;
}
else if (med->v2 == startVert) {
startVert = med->v1;
prependPolyLineVert(&polyline, startVert);
totpoly++;
BLI_freelinkN(&edges, edl);
totedges--;
ok = true;
}
edl = edl_prev;
}
}
/* Now we have a polyline, make into a curve */
if (startVert == endVert) {
BLI_freelinkN(&polyline, polyline.last);
totpoly--;
closed = true;
}
/* --- nurbs --- */
{
Nurb *nu;
BPoint *bp;
VertLink *vl;
/* create new 'nurb' within the curve */
nu = (Nurb *)MEM_callocN(sizeof(Nurb), "MeshNurb");
nu->pntsu = totpoly;
nu->pntsv = 1;
nu->orderu = 4;
nu->flagu = CU_NURB_ENDPOINT | (closed ? CU_NURB_CYCLIC : 0); /* endpoint */
nu->resolu = 12;
nu->bp = (BPoint *)MEM_calloc_arrayN(totpoly, sizeof(BPoint), "bpoints");
/* add points */
vl = polyline.first;
for (i = 0, bp = nu->bp; i < totpoly; i++, bp++, vl = (VertLink *)vl->next) {
copy_v3_v3(bp->vec, mvert[vl->index].co);
bp->f1 = SELECT;
bp->radius = bp->weight = 1.0;
}
BLI_freelistN(&polyline);
/* add nurb to curve */
BLI_addtail(nurblist, nu);
}
/* --- done with nurbs --- */
}
}
}
void BKE_mesh_to_curve(Main *bmain, Depsgraph *depsgraph, Scene *UNUSED(scene), Object *ob)
{
/* make new mesh data from the original copy */
Scene *scene_eval = DEG_get_evaluated_scene(depsgraph);
Object *ob_eval = DEG_get_evaluated_object(depsgraph, ob);
Mesh *me_eval = mesh_get_eval_final(depsgraph, scene_eval, ob_eval, &CD_MASK_MESH);
ListBase nurblist = {NULL, NULL};
BKE_mesh_to_curve_nurblist(me_eval, &nurblist, 0);
BKE_mesh_to_curve_nurblist(me_eval, &nurblist, 1);
if (nurblist.first) {
Curve *cu = BKE_curve_add(bmain, ob->id.name + 2, OB_CURVE);
cu->flag |= CU_3D;
cu->nurb = nurblist;
id_us_min(&((Mesh *)ob->data)->id);
ob->data = cu;
ob->type = OB_CURVE;
BKE_object_free_derived_caches(ob);
}
}
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* Create a temporary object to be used for nurbs-to-mesh conversion.
*
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
* This is more complex that it should be because BKE_mesh_from_nurbs_displist() will do more than
* simply conversion and will attempt to take over ownership of evaluated result and will also
* modify the input object. */
static Object *object_for_curve_to_mesh_create(Object *object)
{
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
Curve *curve = (Curve *)object->data;
/* Create object itself. */
Object *temp_object;
BKE_id_copy_ex(NULL, &object->id, (ID **)&temp_object, LIB_ID_COPY_LOCALIZE);
/* Remove all modifiers, since we don't want them to be applied. */
BKE_object_free_modifiers(temp_object, LIB_ID_CREATE_NO_USER_REFCOUNT);
/* Copy relevant evaluated fields of curve cache.
*
* Note that there are extra fields in there like bevel and path, but those are not needed during
* conversion, so they are not copied to save unnecessary allocations. */
if (object->runtime.curve_cache != NULL) {
temp_object->runtime.curve_cache = MEM_callocN(sizeof(CurveCache),
"CurveCache for curve types");
BKE_displist_copy(&temp_object->runtime.curve_cache->disp, &object->runtime.curve_cache->disp);
}
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* Constructive modifiers will use mesh to store result. */
if (object->runtime.mesh_eval != NULL) {
BKE_id_copy_ex(NULL,
&object->runtime.mesh_eval->id,
(ID **)&temp_object->runtime.mesh_eval,
LIB_ID_COPY_LOCALIZE);
}
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* Need to create copy of curve itself as well, it will be freed by underlying conversion
* functions.
*
* NOTE: Copies the data, but not the shapekeys. */
BKE_id_copy_ex(NULL, object->data, (ID **)&temp_object->data, LIB_ID_COPY_LOCALIZE);
Curve *temp_curve = (Curve *)temp_object->data;
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* Make sure texture space is calculated for a copy of curve, it will be used for the final
* result. */
BKE_curve_texspace_calc(temp_curve);
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* Temporarily set edit so we get updates from edit mode, but also because for text datablocks
* copying it while in edit mode gives invalid data structures. */
temp_curve->editfont = curve->editfont;
temp_curve->editnurb = curve->editnurb;
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
return temp_object;
}
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
static void curve_to_mesh_eval_ensure(Object *object)
{
if (object->runtime.curve_cache == NULL) {
object->runtime.curve_cache = MEM_callocN(sizeof(CurveCache), "CurveCache for Curve");
}
Curve *curve = (Curve *)object->data;
Curve remapped_curve = *curve;
Object remapped_object = *object;
remapped_object.data = &remapped_curve;
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* Clear all modifiers for the bevel object.
*
* This is because they can not be reliably evaluated for an original object (at least because
* the state of dependencies is not know).
*
* So we create temporary copy of the object which will use same data as the original bevel, but
* will have no modifiers. */
Object bevel_object = {NULL};
if (remapped_curve.bevobj != NULL) {
bevel_object = *remapped_curve.bevobj;
BLI_listbase_clear(&bevel_object.modifiers);
remapped_curve.bevobj = &bevel_object;
}
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* Same thing for taper. */
Object taper_object = {NULL};
if (remapped_curve.taperobj != NULL) {
taper_object = *remapped_curve.taperobj;
BLI_listbase_clear(&taper_object.modifiers);
remapped_curve.taperobj = &taper_object;
}
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* NOTE: We don't have dependency graph or scene here, so we pass NULL. This is all fine since
* they are only used for modifier stack, which we have explicitly disabled for all objects.
*
* TODO(sergey): This is a very fragile logic, but proper solution requires re-writing quite a
* bit of internal functions (BKE_mesh_from_nurbs_displist, BKE_mesh_nomain_to_mesh) and also
* Mesh From Curve operator.
* Brecht says hold off with that. */
BKE_displist_make_curveTypes_forRender(NULL,
NULL,
&remapped_object,
&remapped_object.runtime.curve_cache->disp,
&remapped_object.runtime.mesh_eval,
false,
NULL);
BKE_object_free_curve_cache(&bevel_object);
BKE_object_free_curve_cache(&taper_object);
}
static Mesh *mesh_new_from_curve_type_object(Object *object)
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
{
Curve *curve = object->data;
const bool uv_from_orco = (curve->flag & CU_UV_ORCO) != 0;
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
Object *temp_object = object_for_curve_to_mesh_create(object);
Curve *temp_curve = (Curve *)temp_object->data;
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* When input object is an original one, we don't have evaluated curve cache yet, so need to
* create it in the temporary object. */
if (!DEG_is_evaluated_object(object)) {
curve_to_mesh_eval_ensure(temp_object);
}
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* Reset pointers before conversion. */
temp_curve->editfont = NULL;
temp_curve->editnurb = NULL;
/* Convert to mesh. */
BKE_mesh_from_nurbs_displist(NULL,
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
temp_object,
&temp_object->runtime.curve_cache->disp,
uv_from_orco,
curve->id.name + 2,
true);
/* BKE_mesh_from_nurbs changes the type to a mesh, check it worked. If it didn't the curve did
* not have any segments or otherwise would have generated an empty mesh. */
if (temp_object->type != OB_MESH) {
BKE_id_free(NULL, temp_object);
return NULL;
}
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
Mesh *mesh_result = temp_object->data;
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
BKE_id_free(NULL, temp_object);
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* NOTE: Materials are copied in BKE_mesh_from_nurbs_displist(). */
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
return mesh_result;
}
static Mesh *mesh_new_from_mball_object(Object *object)
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
{
MetaBall *mball = (MetaBall *)object->data;
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* NOTE: We can only create mesh for a polygonized meta ball. This figures out all original meta
* balls and all evaluated child meta balls (since polygonization is only stored in the mother
* ball).
*
* We create empty mesh so scripters don't run into None objects. */
if (!DEG_is_evaluated_object(object) || object->runtime.curve_cache == NULL ||
BLI_listbase_is_empty(&object->runtime.curve_cache->disp)) {
return BKE_id_new_nomain(ID_ME, ((ID *)object->data)->name + 2);
}
Mesh *mesh_result = BKE_id_new_nomain(ID_ME, ((ID *)object->data)->name + 2);
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
BKE_mesh_from_metaball(&object->runtime.curve_cache->disp, mesh_result);
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* Copy materials. */
mesh_result->totcol = mball->totcol;
mesh_result->mat = MEM_dupallocN(mball->mat);
if (mball->mat != NULL) {
for (int i = mball->totcol; i-- > 0;) {
mesh_result->mat[i] = give_current_material(object, i + 1);
}
}
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
return mesh_result;
}
static Mesh *mesh_new_from_mesh_object(Object *object)
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
{
Mesh *mesh_input = object->data;
/* If we are in edit mode, use evaluated mesh from edit structure, matching to what
* viewport is using for visualization. */
if (mesh_input->edit_mesh != NULL && mesh_input->edit_mesh->mesh_eval_final) {
mesh_input = mesh_input->edit_mesh->mesh_eval_final;
}
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
Mesh *mesh_result = NULL;
BKE_id_copy_ex(NULL,
&mesh_input->id,
(ID **)&mesh_result,
LIB_ID_CREATE_NO_MAIN | LIB_ID_CREATE_NO_USER_REFCOUNT);
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
/* NOTE: Materials should already be copied. */
/* Copy original mesh name. This is because edit meshes might not have one properly set name. */
BLI_strncpy(mesh_result->id.name, ((ID *)object->data)->name, sizeof(mesh_result->id.name));
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
return mesh_result;
}
Mesh *BKE_mesh_new_from_object(Object *object)
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
{
Mesh *new_mesh = NULL;
switch (object->type) {
case OB_FONT:
case OB_CURVE:
case OB_SURF:
new_mesh = mesh_new_from_curve_type_object(object);
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
break;
case OB_MBALL:
new_mesh = mesh_new_from_mball_object(object);
break;
case OB_MESH:
new_mesh = mesh_new_from_mesh_object(object);
break;
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
default:
/* Object does not have geometry data. */
return NULL;
}
if (new_mesh == NULL) {
/* Happens in special cases like request of mesh for non-mother meta ball. */
return NULL;
}
/* The result must have 0 users, since it's just a mesh which is free-dangling data-block.
* All the conversion functions are supposed to ensure mesh is not counted. */
BLI_assert(new_mesh->id.us == 0);
Dependency graph API changes Main goal here is to make it obvious and predictable about what is going on. Summary of changes. - Access to dependency graph is now only possible to a fully evaluated graph. This is now done via context.evaluated_depsgraph_get(). The call will ensure both relations and datablocks are updated. This way we don't allow access to some known bad state of the graph, and also making explicit that getting update dependency graph is not cheap. - Access to evaluated ID is now possible via id.evaluated_get(). It was already possible to get evaluated ID via dependency graph, but that was a bit confusing why access to original is done via ID and to evaluated via depsgraph. If datablock is not covered by dependency graph it will be returned as-is. - Similarly, request for original from an ID which is not evaluated will return ID as-is. - Removed scene.update(). This is very expensive to update all the view layers. - Added depsgraph.update(). Now when temporary changes to objects are to be done, this is to happen on original object and then dependency graph is to be updated. - Changed object.to_mesh() to behave the following way: * When is used for original object modifiers are ignored. For meshes this acts similar to mesh-copy, not very useful but allows to keep code paths similar (i.e. for exporter which has Apply Modifiers option it's only matter choosing between original and evaluated object, the to_mesh() part can stay the same). For curves this gives a mesh which is constructed from displist without taking own modifiers and modifiers of bevel/taper objects into account. For metaballs this gives empty mesh. Polygonization of metaball is not possible from a single object. * When is used for evaluated object modifiers are always applied. In fact, no evaluation is happening, the mesh is either copied as-is, or constructed from current state of curve cache. Arguments to apply modifiers and calculate original coordinates (ORCO, aka undeformed coordinates) are removed. The ORCO is to be calculated as part of dependency graph evaluation. File used to regression-test (a packed Python script into .blend): {F7033464} Patch to make addons tests to pass: {F7033466} NOTE: I've included changes to FBX exporter, and those are addressing report T63689. NOTE: All the enabled-by-default addons are to be ported still, but first want to have agreement on this part of changes. NOTE: Also need to work on documentation for Python API, but, again, better be done after having agreement on this work. Reviewers: brecht, campbellbarton, mont29 Differential Revision: https://developer.blender.org/D4834
2019-05-09 11:26:49 +02:00
return new_mesh;
}
static int foreach_libblock_make_original_and_usercount_callback(void *user_data_v,
ID *id_self,
ID **id_p,
int cb_flag)
{
UNUSED_VARS(user_data_v, id_self, cb_flag);
if (*id_p == NULL) {
return IDWALK_RET_NOP;
}
*id_p = DEG_get_original_id(*id_p);
id_us_plus(*id_p);
return IDWALK_RET_NOP;
}
Mesh *BKE_mesh_new_from_object_to_bmain(Main *bmain, Object *object)
{
Mesh *mesh = BKE_mesh_new_from_object(object);
/* Make sure mesh only points original datablocks, also increase users of materials and other
* possibly referenced data-blocks.
*
* Going to original data-blocks is required to have bmain in a consistent state, where
* everything is only allowed to reference original data-blocks.
*
* user-count is required is because so far mesh was in a limbo, where library management does
* not perform any user management (i.e. copy of a mesh will not increase users of materials). */
BKE_library_foreach_ID_link(
NULL, &mesh->id, foreach_libblock_make_original_and_usercount_callback, NULL, IDWALK_NOP);
/* Append the mesh to bmain.
* We do it a bit longer way since there is no simple and clear way of adding existing datablock
* to the bmain. So we allocate new empty mesh in the bmain (which guarantess all the naming and
* orders and flags) and move the temporary mesh in place there. */
Mesh *mesh_in_bmain = BKE_mesh_add(bmain, mesh->id.name + 2);
BKE_mesh_nomain_to_mesh(mesh, mesh_in_bmain, NULL, &CD_MASK_MESH, true);
/* Make sure user count from BKE_mesh_add() is the one we expect here and bring it down to 0. */
BLI_assert(mesh_in_bmain->id.us == 1);
id_us_min(&mesh_in_bmain->id);
return mesh_in_bmain;
}
static void add_shapekey_layers(Mesh *mesh_dest, Mesh *mesh_src)
{
KeyBlock *kb;
Key *key = mesh_src->key;
int i;
if (!mesh_src->key) {
return;
}
/* ensure we can use mesh vertex count for derived mesh custom data */
if (mesh_src->totvert != mesh_dest->totvert) {
CLOG_ERROR(&LOG,
"vertex size mismatch (mesh/dm) '%s' (%d != %d)",
mesh_src->id.name + 2,
mesh_src->totvert,
mesh_dest->totvert);
return;
}
for (i = 0, kb = key->block.first; kb; kb = kb->next, i++) {
int ci;
float *array;
if (mesh_src->totvert != kb->totelem) {
CLOG_ERROR(&LOG,
"vertex size mismatch (Mesh '%s':%d != KeyBlock '%s':%d)",
mesh_src->id.name + 2,
mesh_src->totvert,
kb->name,
kb->totelem);
array = MEM_calloc_arrayN((size_t)mesh_src->totvert, 3 * sizeof(float), __func__);
}
else {
array = MEM_malloc_arrayN((size_t)mesh_src->totvert, 3 * sizeof(float), __func__);
memcpy(array, kb->data, (size_t)mesh_src->totvert * 3 * sizeof(float));
}
CustomData_add_layer_named(
&mesh_dest->vdata, CD_SHAPEKEY, CD_ASSIGN, array, mesh_dest->totvert, kb->name);
ci = CustomData_get_layer_index_n(&mesh_dest->vdata, CD_SHAPEKEY, i);
mesh_dest->vdata.layers[ci].uid = kb->uid;
}
}
Mesh *BKE_mesh_create_derived_for_modifier(struct Depsgraph *depsgraph,
Scene *scene,
Object *ob_eval,
ModifierData *md_eval,
int build_shapekey_layers)
{
Mesh *me = ob_eval->runtime.mesh_orig ? ob_eval->runtime.mesh_orig : ob_eval->data;
const ModifierTypeInfo *mti = modifierType_getInfo(md_eval->type);
Mesh *result;
KeyBlock *kb;
ModifierEvalContext mectx = {depsgraph, ob_eval, 0};
if (!(md_eval->mode & eModifierMode_Realtime)) {
return NULL;
}
if (mti->isDisabled && mti->isDisabled(scene, md_eval, 0)) {
return NULL;
}
if (build_shapekey_layers && me->key &&
(kb = BLI_findlink(&me->key->block, ob_eval->shapenr - 1))) {
BKE_keyblock_convert_to_mesh(kb, me);
}
if (mti->type == eModifierTypeType_OnlyDeform) {
int numVerts;
float(*deformedVerts)[3] = BKE_mesh_vertexCos_get(me, &numVerts);
BKE_id_copy_ex(NULL, &me->id, (ID **)&result, LIB_ID_COPY_LOCALIZE);
mti->deformVerts(md_eval, &mectx, result, deformedVerts, numVerts);
BKE_mesh_apply_vert_coords(result, deformedVerts);
if (build_shapekey_layers) {
add_shapekey_layers(result, me);
}
MEM_freeN(deformedVerts);
}
else {
Mesh *mesh_temp;
BKE_id_copy_ex(NULL, &me->id, (ID **)&mesh_temp, LIB_ID_COPY_LOCALIZE);
if (build_shapekey_layers) {
add_shapekey_layers(mesh_temp, me);
}
result = mti->applyModifier(md_eval, &mectx, mesh_temp);
ASSERT_IS_VALID_MESH(result);
if (mesh_temp != result) {
BKE_id_free(NULL, mesh_temp);
}
}
return result;
}
/* This is a Mesh-based copy of the same function in DerivedMesh.c */
static void shapekey_layers_to_keyblocks(Mesh *mesh_src, Mesh *mesh_dst, int actshape_uid)
{
KeyBlock *kb;
int i, j, tot;
if (!mesh_dst->key) {
return;
}
tot = CustomData_number_of_layers(&mesh_src->vdata, CD_SHAPEKEY);
for (i = 0; i < tot; i++) {
CustomDataLayer *layer =
&mesh_src->vdata.layers[CustomData_get_layer_index_n(&mesh_src->vdata, CD_SHAPEKEY, i)];
float(*cos)[3], (*kbcos)[3];
for (kb = mesh_dst->key->block.first; kb; kb = kb->next) {
if (kb->uid == layer->uid) {
break;
}
}
if (!kb) {
kb = BKE_keyblock_add(mesh_dst->key, layer->name);
kb->uid = layer->uid;
}
if (kb->data) {
MEM_freeN(kb->data);
}
cos = CustomData_get_layer_n(&mesh_src->vdata, CD_SHAPEKEY, i);
kb->totelem = mesh_src->totvert;
kb->data = kbcos = MEM_malloc_arrayN(kb->totelem, 3 * sizeof(float), __func__);
if (kb->uid == actshape_uid) {
MVert *mvert = mesh_src->mvert;
for (j = 0; j < mesh_src->totvert; j++, kbcos++, mvert++) {
copy_v3_v3(*kbcos, mvert->co);
}
}
else {
for (j = 0; j < kb->totelem; j++, cos++, kbcos++) {
copy_v3_v3(*kbcos, *cos);
}
}
}
for (kb = mesh_dst->key->block.first; kb; kb = kb->next) {
if (kb->totelem != mesh_src->totvert) {
if (kb->data) {
MEM_freeN(kb->data);
}
kb->totelem = mesh_src->totvert;
kb->data = MEM_calloc_arrayN(kb->totelem, 3 * sizeof(float), __func__);
CLOG_ERROR(&LOG, "lost a shapekey layer: '%s'! (bmesh internal error)", kb->name);
}
}
}
/* This is a Mesh-based copy of DM_to_mesh() */
void BKE_mesh_nomain_to_mesh(Mesh *mesh_src,
Mesh *mesh_dst,
Object *ob,
const CustomData_MeshMasks *mask,
bool take_ownership)
{
/* mesh_src might depend on mesh_dst, so we need to do everything with a local copy */
/* TODO(Sybren): the above claim came from DM_to_mesh();
* check whether it is still true with Mesh */
Mesh tmp = *mesh_dst;
int totvert, totedge /*, totface */ /* UNUSED */, totloop, totpoly;
int did_shapekeys = 0;
eCDAllocType alloctype = CD_DUPLICATE;
if (take_ownership /* && dm->type == DM_TYPE_CDDM && dm->needsFree */) {
bool has_any_referenced_layers = CustomData_has_referenced(&mesh_src->vdata) ||
CustomData_has_referenced(&mesh_src->edata) ||
CustomData_has_referenced(&mesh_src->ldata) ||
CustomData_has_referenced(&mesh_src->fdata) ||
CustomData_has_referenced(&mesh_src->pdata);
if (!has_any_referenced_layers) {
alloctype = CD_ASSIGN;
}
}
CustomData_reset(&tmp.vdata);
CustomData_reset(&tmp.edata);
CustomData_reset(&tmp.fdata);
CustomData_reset(&tmp.ldata);
CustomData_reset(&tmp.pdata);
BKE_mesh_ensure_normals(mesh_src);
totvert = tmp.totvert = mesh_src->totvert;
totedge = tmp.totedge = mesh_src->totedge;
totloop = tmp.totloop = mesh_src->totloop;
totpoly = tmp.totpoly = mesh_src->totpoly;
tmp.totface = 0;
CustomData_copy(&mesh_src->vdata, &tmp.vdata, mask->vmask, alloctype, totvert);
CustomData_copy(&mesh_src->edata, &tmp.edata, mask->emask, alloctype, totedge);
CustomData_copy(&mesh_src->ldata, &tmp.ldata, mask->lmask, alloctype, totloop);
CustomData_copy(&mesh_src->pdata, &tmp.pdata, mask->pmask, alloctype, totpoly);
tmp.cd_flag = mesh_src->cd_flag;
tmp.runtime.deformed_only = mesh_src->runtime.deformed_only;
if (CustomData_has_layer(&mesh_src->vdata, CD_SHAPEKEY)) {
KeyBlock *kb;
int uid;
if (ob) {
kb = BLI_findlink(&mesh_dst->key->block, ob->shapenr - 1);
if (kb) {
uid = kb->uid;
}
else {
CLOG_ERROR(&LOG, "could not find active shapekey %d!", ob->shapenr - 1);
uid = INT_MAX;
}
}
else {
/* if no object, set to INT_MAX so we don't mess up any shapekey layers */
uid = INT_MAX;
}
shapekey_layers_to_keyblocks(mesh_src, mesh_dst, uid);
did_shapekeys = 1;
}
/* copy texture space */
if (ob) {
BKE_mesh_texspace_copy_from_object(&tmp, ob);
}
/* not all DerivedMeshes store their verts/edges/faces in CustomData, so
* we set them here in case they are missing */
/* TODO(Sybren): we could probably replace CD_ASSIGN with alloctype and
* always directly pass mesh_src->mxxx, instead of using a ternary operator. */
if (!CustomData_has_layer(&tmp.vdata, CD_MVERT)) {
CustomData_add_layer(&tmp.vdata,
CD_MVERT,
CD_ASSIGN,
(alloctype == CD_ASSIGN) ? mesh_src->mvert :
MEM_dupallocN(mesh_src->mvert),
totvert);
}
if (!CustomData_has_layer(&tmp.edata, CD_MEDGE)) {
CustomData_add_layer(&tmp.edata,
CD_MEDGE,
CD_ASSIGN,
(alloctype == CD_ASSIGN) ? mesh_src->medge :
MEM_dupallocN(mesh_src->medge),
totedge);
}
if (!CustomData_has_layer(&tmp.pdata, CD_MPOLY)) {
/* TODO(Sybren): assignment to tmp.mxxx is probably not necessary due to the
* BKE_mesh_update_customdata_pointers() call below. */
tmp.mloop = (alloctype == CD_ASSIGN) ? mesh_src->mloop : MEM_dupallocN(mesh_src->mloop);
tmp.mpoly = (alloctype == CD_ASSIGN) ? mesh_src->mpoly : MEM_dupallocN(mesh_src->mpoly);
CustomData_add_layer(&tmp.ldata, CD_MLOOP, CD_ASSIGN, tmp.mloop, tmp.totloop);
CustomData_add_layer(&tmp.pdata, CD_MPOLY, CD_ASSIGN, tmp.mpoly, tmp.totpoly);
}
/* object had got displacement layer, should copy this layer to save sculpted data */
/* NOTE: maybe some other layers should be copied? nazgul */
if (CustomData_has_layer(&mesh_dst->ldata, CD_MDISPS)) {
if (totloop == mesh_dst->totloop) {
MDisps *mdisps = CustomData_get_layer(&mesh_dst->ldata, CD_MDISPS);
CustomData_add_layer(&tmp.ldata, CD_MDISPS, alloctype, mdisps, totloop);
}
}
/* yes, must be before _and_ after tessellate */
BKE_mesh_update_customdata_pointers(&tmp, false);
/* since 2.65 caller must do! */
// BKE_mesh_tessface_calc(&tmp);
CustomData_free(&mesh_dst->vdata, mesh_dst->totvert);
CustomData_free(&mesh_dst->edata, mesh_dst->totedge);
CustomData_free(&mesh_dst->fdata, mesh_dst->totface);
CustomData_free(&mesh_dst->ldata, mesh_dst->totloop);
CustomData_free(&mesh_dst->pdata, mesh_dst->totpoly);
/* ok, this should now use new CD shapekey data,
* which should be fed through the modifier
* stack */
if (tmp.totvert != mesh_dst->totvert && !did_shapekeys && mesh_dst->key) {
CLOG_ERROR(&LOG, "YEEK! this should be recoded! Shape key loss!: ID '%s'", tmp.id.name);
if (tmp.key && !(tmp.id.tag & LIB_TAG_NO_MAIN)) {
id_us_min(&tmp.key->id);
}
tmp.key = NULL;
}
/* Clear selection history */
MEM_SAFE_FREE(tmp.mselect);
tmp.totselect = 0;
BLI_assert(ELEM(tmp.bb, NULL, mesh_dst->bb));
if (mesh_dst->bb) {
MEM_freeN(mesh_dst->bb);
tmp.bb = NULL;
}
/* skip the listbase */
MEMCPY_STRUCT_AFTER(mesh_dst, &tmp, id.prev);
if (take_ownership) {
if (alloctype == CD_ASSIGN) {
CustomData_free_typemask(&mesh_src->vdata, mesh_src->totvert, ~mask->vmask);
CustomData_free_typemask(&mesh_src->edata, mesh_src->totedge, ~mask->emask);
CustomData_free_typemask(&mesh_src->ldata, mesh_src->totloop, ~mask->lmask);
CustomData_free_typemask(&mesh_src->pdata, mesh_src->totpoly, ~mask->pmask);
}
BKE_id_free(NULL, mesh_src);
}
}
/* This is a Mesh-based copy of DM_to_meshkey() */
void BKE_mesh_nomain_to_meshkey(Mesh *mesh_src, Mesh *mesh_dst, KeyBlock *kb)
{
int a, totvert = mesh_src->totvert;
float *fp;
MVert *mvert;
if (totvert == 0 || mesh_dst->totvert == 0 || mesh_dst->totvert != totvert) {
return;
}
if (kb->data) {
MEM_freeN(kb->data);
}
kb->data = MEM_malloc_arrayN(mesh_dst->key->elemsize, mesh_dst->totvert, "kb->data");
kb->totelem = totvert;
fp = kb->data;
mvert = mesh_src->mvert;
for (a = 0; a < kb->totelem; a++, fp += 3, mvert++) {
copy_v3_v3(fp, mvert->co);
}
}