Files
test2/source/blender/modifiers/intern/MOD_util.cc

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

331 lines
9.3 KiB
C++
Raw Normal View History

/* SPDX-License-Identifier: GPL-2.0-or-later
* Copyright 2005 Blender Foundation. All rights reserved. */
/** \file
* \ingroup modifiers
*/
#include <cstring>
#include "BLI_utildefines.h"
#include "BLI_bitmap.h"
#include "BLI_math_matrix.h"
#include "BLI_math_vector.h"
#include "DNA_image_types.h"
#include "DNA_mesh_types.h"
#include "DNA_meshdata_types.h"
#include "DNA_modifier_types.h"
#include "DNA_object_types.h"
#include "DNA_scene_types.h"
#include "BKE_action.h" /* BKE_pose_channel_find_name */
#include "BKE_deform.h"
#include "BKE_editmesh.h"
#include "BKE_image.h"
#include "BKE_lattice.h"
#include "BKE_lib_id.h"
#include "BKE_mesh.h"
#include "BKE_mesh_wrapper.h"
#include "BKE_object.h"
#include "BKE_modifier.h"
#include "DEG_depsgraph.h"
#include "DEG_depsgraph_query.h"
#include "MOD_modifiertypes.h"
#include "MOD_util.h"
#include "MEM_guardedalloc.h"
#include "bmesh.h"
void MOD_init_texture(MappingInfoModifierData *dmd, const ModifierEvalContext *ctx)
{
Tex *tex = dmd->texture;
if (tex == nullptr) {
return;
}
if (tex->ima && BKE_image_is_animated(tex->ima)) {
BKE_image_user_frame_calc(tex->ima, &tex->iuser, DEG_get_ctime(ctx->depsgraph));
}
}
/* TODO: to be renamed to get_texture_coords once we are done with moving modifiers to Mesh. */
void MOD_get_texture_coords(MappingInfoModifierData *dmd,
const ModifierEvalContext *UNUSED(ctx),
Object *ob,
Mesh *mesh,
float (*cos)[3],
float (*r_texco)[3])
{
const int verts_num = mesh->totvert;
int i;
int texmapping = dmd->texmapping;
float mapref_imat[4][4];
if (texmapping == MOD_DISP_MAP_OBJECT) {
if (dmd->map_object != nullptr) {
Object *map_object = dmd->map_object;
if (dmd->map_bone[0] != '\0') {
bPoseChannel *pchan = BKE_pose_channel_find_name(map_object->pose, dmd->map_bone);
if (pchan) {
float mat_bone_world[4][4];
mul_m4_m4m4(mat_bone_world, map_object->obmat, pchan->pose_mat);
invert_m4_m4(mapref_imat, mat_bone_world);
}
else {
invert_m4_m4(mapref_imat, map_object->obmat);
}
}
else {
invert_m4_m4(mapref_imat, map_object->obmat);
}
}
else { /* if there is no map object, default to local */
texmapping = MOD_DISP_MAP_LOCAL;
}
}
/* UVs need special handling, since they come from faces */
if (texmapping == MOD_DISP_MAP_UV) {
if (CustomData_has_layer(&mesh->ldata, CD_MLOOPUV)) {
const MPoly *mpoly = BKE_mesh_polys(mesh);
Mesh: Remove redundant custom data pointers For copy-on-write, we want to share attribute arrays between meshes where possible. Mutable pointers like `Mesh.mvert` make that difficult by making ownership vague. They also make code more complex by adding redundancy. The simplest solution is just removing them and retrieving layers from `CustomData` as needed. Similar changes have already been applied to curves and point clouds (e9f82d3dc7ee, 410a6efb747f). Removing use of the pointers generally makes code more obvious and more reusable. Mesh data is now accessed with a C++ API (`Mesh::edges()` or `Mesh::edges_for_write()`), and a C API (`BKE_mesh_edges(mesh)`). The CoW changes this commit makes possible are described in T95845 and T95842, and started in D14139 and D14140. The change also simplifies the ongoing mesh struct-of-array refactors from T95965. **RNA/Python Access Performance** Theoretically, accessing mesh elements with the RNA API may become slower, since the layer needs to be found on every random access. However, overhead is already high enough that this doesn't make a noticible differenc, and performance is actually improved in some cases. Random access can be up to 10% faster, but other situations might be a bit slower. Generally using `foreach_get/set` are the best way to improve performance. See the differential revision for more discussion about Python performance. Cycles has been updated to use raw pointers and the internal Blender mesh types, mostly because there is no sense in having this overhead when it's already compiled with Blender. In my tests this roughly halves the Cycles mesh creation time (0.19s to 0.10s for a 1 million face grid). Differential Revision: https://developer.blender.org/D15488
2022-09-05 11:56:34 -05:00
const MPoly *mp;
const MLoop *mloop = BKE_mesh_loops(mesh);
BLI_bitmap *done = BLI_BITMAP_NEW(verts_num, __func__);
const int polys_num = mesh->totpoly;
char uvname[MAX_CUSTOMDATA_LAYER_NAME];
CustomData_validate_layer_name(&mesh->ldata, CD_MLOOPUV, dmd->uvlayer_name, uvname);
const MLoopUV *mloop_uv = static_cast<const MLoopUV *>(
CustomData_get_layer_named(&mesh->ldata, CD_MLOOPUV, uvname));
/* verts are given the UV from the first face that uses them */
for (i = 0, mp = mpoly; i < polys_num; i++, mp++) {
uint fidx = mp->totloop - 1;
do {
uint lidx = mp->loopstart + fidx;
uint vidx = mloop[lidx].v;
if (!BLI_BITMAP_TEST(done, vidx)) {
/* remap UVs from [0, 1] to [-1, 1] */
r_texco[vidx][0] = (mloop_uv[lidx].uv[0] * 2.0f) - 1.0f;
r_texco[vidx][1] = (mloop_uv[lidx].uv[1] * 2.0f) - 1.0f;
BLI_BITMAP_ENABLE(done, vidx);
}
} while (fidx--);
}
MEM_freeN(done);
return;
}
/* if there are no UVs, default to local */
texmapping = MOD_DISP_MAP_LOCAL;
}
const MVert *mv = BKE_mesh_verts(mesh);
for (i = 0; i < verts_num; i++, mv++, r_texco++) {
switch (texmapping) {
case MOD_DISP_MAP_LOCAL:
copy_v3_v3(*r_texco, cos != nullptr ? *cos : mv->co);
break;
case MOD_DISP_MAP_GLOBAL:
mul_v3_m4v3(*r_texco, ob->obmat, cos != nullptr ? *cos : mv->co);
break;
case MOD_DISP_MAP_OBJECT:
mul_v3_m4v3(*r_texco, ob->obmat, cos != nullptr ? *cos : mv->co);
mul_m4_v3(mapref_imat, *r_texco);
break;
}
if (cos != nullptr) {
cos++;
}
}
}
void MOD_previous_vcos_store(ModifierData *md, const float (*vert_coords)[3])
{
2012-05-06 13:38:33 +00:00
while ((md = md->next) && md->type == eModifierType_Armature) {
ArmatureModifierData *amd = (ArmatureModifierData *)md;
if (amd->multi && amd->vert_coords_prev == nullptr) {
amd->vert_coords_prev = static_cast<float(*)[3]>(MEM_dupallocN(vert_coords));
}
else {
break;
}
}
/* lattice/mesh modifier too */
}
Mesh *MOD_deform_mesh_eval_get(Object *ob,
2018-05-12 08:21:07 +02:00
struct BMEditMesh *em,
Mesh *mesh,
const float (*vertexCos)[3],
const int verts_num,
const bool use_orco)
{
if (mesh != nullptr) {
/* pass */
}
else if (ob->type == OB_MESH) {
if (em) {
mesh = BKE_mesh_wrapper_from_editmesh_with_coords(
em, nullptr, vertexCos, static_cast<const Mesh *>(ob->data));
}
else {
/* TODO(sybren): after modifier conversion of DM to Mesh is done, check whether
* we really need a copy here. Maybe the CoW ob->data can be directly used. */
Mesh *mesh_prior_modifiers = BKE_object_get_pre_modified_mesh(ob);
mesh = (Mesh *)BKE_id_copy_ex(nullptr,
&mesh_prior_modifiers->id,
nullptr,
(LIB_ID_COPY_LOCALIZE | LIB_ID_COPY_CD_REFERENCE));
mesh->runtime.deformed_only = 1;
}
if (em != nullptr) {
/* pass */
}
/* TODO(sybren): after modifier conversion of DM to Mesh is done, check whether
* we really need vertexCos here. */
else if (vertexCos) {
BKE_mesh_vert_coords_apply(mesh, vertexCos);
}
if (use_orco) {
BKE_mesh_orco_ensure(ob, mesh);
}
}
else if (ELEM(ob->type, OB_FONT, OB_CURVES_LEGACY, OB_SURF)) {
/* TODO(sybren): get evaluated mesh from depsgraph once
* that's properly generated for curves. */
2018-05-08 17:28:43 +02:00
mesh = BKE_mesh_new_nomain_from_curve(ob);
2020-09-10 09:45:25 +10:00
/* Currently, that may not be the case every time
* (texts e.g. tend to give issues,
* also when deforming curve points instead of generated curve geometry... ). */
if (mesh != nullptr && mesh->totvert != verts_num) {
BKE_id_free(nullptr, mesh);
mesh = nullptr;
}
}
if (mesh && mesh->runtime.wrapper_type == ME_WRAPPER_TYPE_MDATA) {
BLI_assert(mesh->totvert == verts_num);
}
return mesh;
}
2018-06-29 19:02:19 +02:00
void MOD_get_vgroup(
Mesh: Remove redundant custom data pointers For copy-on-write, we want to share attribute arrays between meshes where possible. Mutable pointers like `Mesh.mvert` make that difficult by making ownership vague. They also make code more complex by adding redundancy. The simplest solution is just removing them and retrieving layers from `CustomData` as needed. Similar changes have already been applied to curves and point clouds (e9f82d3dc7ee, 410a6efb747f). Removing use of the pointers generally makes code more obvious and more reusable. Mesh data is now accessed with a C++ API (`Mesh::edges()` or `Mesh::edges_for_write()`), and a C API (`BKE_mesh_edges(mesh)`). The CoW changes this commit makes possible are described in T95845 and T95842, and started in D14139 and D14140. The change also simplifies the ongoing mesh struct-of-array refactors from T95965. **RNA/Python Access Performance** Theoretically, accessing mesh elements with the RNA API may become slower, since the layer needs to be found on every random access. However, overhead is already high enough that this doesn't make a noticible differenc, and performance is actually improved in some cases. Random access can be up to 10% faster, but other situations might be a bit slower. Generally using `foreach_get/set` are the best way to improve performance. See the differential revision for more discussion about Python performance. Cycles has been updated to use raw pointers and the internal Blender mesh types, mostly because there is no sense in having this overhead when it's already compiled with Blender. In my tests this roughly halves the Cycles mesh creation time (0.19s to 0.10s for a 1 million face grid). Differential Revision: https://developer.blender.org/D15488
2022-09-05 11:56:34 -05:00
Object *ob, struct Mesh *mesh, const char *name, const MDeformVert **dvert, int *defgrp_index)
{
if (mesh) {
*defgrp_index = BKE_id_defgroup_name_index(&mesh->id, name);
if (*defgrp_index != -1) {
Mesh: Remove redundant custom data pointers For copy-on-write, we want to share attribute arrays between meshes where possible. Mutable pointers like `Mesh.mvert` make that difficult by making ownership vague. They also make code more complex by adding redundancy. The simplest solution is just removing them and retrieving layers from `CustomData` as needed. Similar changes have already been applied to curves and point clouds (e9f82d3dc7ee, 410a6efb747f). Removing use of the pointers generally makes code more obvious and more reusable. Mesh data is now accessed with a C++ API (`Mesh::edges()` or `Mesh::edges_for_write()`), and a C API (`BKE_mesh_edges(mesh)`). The CoW changes this commit makes possible are described in T95845 and T95842, and started in D14139 and D14140. The change also simplifies the ongoing mesh struct-of-array refactors from T95965. **RNA/Python Access Performance** Theoretically, accessing mesh elements with the RNA API may become slower, since the layer needs to be found on every random access. However, overhead is already high enough that this doesn't make a noticible differenc, and performance is actually improved in some cases. Random access can be up to 10% faster, but other situations might be a bit slower. Generally using `foreach_get/set` are the best way to improve performance. See the differential revision for more discussion about Python performance. Cycles has been updated to use raw pointers and the internal Blender mesh types, mostly because there is no sense in having this overhead when it's already compiled with Blender. In my tests this roughly halves the Cycles mesh creation time (0.19s to 0.10s for a 1 million face grid). Differential Revision: https://developer.blender.org/D15488
2022-09-05 11:56:34 -05:00
*dvert = BKE_mesh_deform_verts(mesh);
}
else {
*dvert = nullptr;
}
}
else {
*defgrp_index = BKE_object_defgroup_name_index(ob, name);
if (*defgrp_index != -1 && ob->type == OB_LATTICE) {
*dvert = BKE_lattice_deform_verts_get(ob);
}
else {
*dvert = nullptr;
}
}
}
void MOD_depsgraph_update_object_bone_relation(struct DepsNodeHandle *node,
Object *object,
const char *bonename,
const char *description)
{
if (object == nullptr) {
return;
}
if (bonename[0] != '\0' && object->type == OB_ARMATURE) {
DEG_add_object_relation(node, object, DEG_OB_COMP_EVAL_POSE, description);
}
else {
DEG_add_object_relation(node, object, DEG_OB_COMP_TRANSFORM, description);
}
}
void modifier_type_init(ModifierTypeInfo *types[])
{
#define INIT_TYPE(typeName) (types[eModifierType_##typeName] = &modifierType_##typeName)
INIT_TYPE(None);
INIT_TYPE(Curve);
INIT_TYPE(Lattice);
INIT_TYPE(Subsurf);
INIT_TYPE(Build);
INIT_TYPE(Array);
INIT_TYPE(Mirror);
INIT_TYPE(EdgeSplit);
INIT_TYPE(Bevel);
INIT_TYPE(Displace);
INIT_TYPE(UVProject);
INIT_TYPE(Decimate);
INIT_TYPE(Smooth);
INIT_TYPE(Cast);
INIT_TYPE(Wave);
INIT_TYPE(Armature);
INIT_TYPE(Hook);
INIT_TYPE(Softbody);
INIT_TYPE(Cloth);
INIT_TYPE(Collision);
INIT_TYPE(Boolean);
INIT_TYPE(MeshDeform);
INIT_TYPE(Ocean);
INIT_TYPE(ParticleSystem);
INIT_TYPE(ParticleInstance);
INIT_TYPE(Explode);
INIT_TYPE(Shrinkwrap);
INIT_TYPE(Mask);
INIT_TYPE(SimpleDeform);
INIT_TYPE(Multires);
INIT_TYPE(Surface);
INIT_TYPE(Fluid);
INIT_TYPE(ShapeKey);
INIT_TYPE(Solidify);
INIT_TYPE(Screw);
INIT_TYPE(Warp);
INIT_TYPE(WeightVGEdit);
INIT_TYPE(WeightVGMix);
INIT_TYPE(WeightVGProximity);
INIT_TYPE(DynamicPaint);
INIT_TYPE(Remesh);
INIT_TYPE(Skin);
INIT_TYPE(LaplacianSmooth);
INIT_TYPE(Triangulate);
INIT_TYPE(UVWarp);
INIT_TYPE(MeshCache);
INIT_TYPE(LaplacianDeform);
INIT_TYPE(Wireframe);
INIT_TYPE(Weld);
INIT_TYPE(DataTransfer);
INIT_TYPE(NormalEdit);
INIT_TYPE(CorrectiveSmooth);
INIT_TYPE(MeshSequenceCache);
INIT_TYPE(SurfaceDeform);
2018-05-25 22:24:24 +05:30
INIT_TYPE(WeightedNormal);
INIT_TYPE(MeshToVolume);
INIT_TYPE(VolumeDisplace);
INIT_TYPE(VolumeToMesh);
Geometry Nodes: initial scattering and geometry nodes This is the initial merge from the geometry-nodes branch. Nodes: * Attribute Math * Boolean * Edge Split * Float Compare * Object Info * Point Distribute * Point Instance * Random Attribute * Random Float * Subdivision Surface * Transform * Triangulate It includes the initial evaluation of geometry node groups in the Geometry Nodes modifier. Notes on the Generic attribute access API The API adds an indirection for attribute access. That has the following benefits: * Most code does not have to care about how an attribute is stored internally. This is mainly necessary, because we have to deal with "legacy" attributes such as vertex weights and attributes that are embedded into other structs such as vertex positions. * When reading from an attribute, we generally don't care what domain the attribute is stored on. So we want to abstract away the interpolation that that adapts attributes from one domain to another domain (this is not actually implemented yet). Other possible improvements for later iterations include: * Actually implement interpolation between domains. * Don't use inheritance for the different attribute types. A single class for read access and one for write access might be enough, because we know all the ways in which attributes are stored internally. We don't want more different internal structures in the future. On the contrary, ideally we can consolidate the different storage formats in the future to reduce the need for this indirection. * Remove the need for heap allocations when creating attribute accessors. It includes commits from: * Dalai Felinto * Hans Goudey * Jacques Lucke * Léo Depoix
2020-12-02 13:25:25 +01:00
INIT_TYPE(Nodes);
#undef INIT_TYPE
}