1434 lines
48 KiB
C++
1434 lines
48 KiB
C++
/* SPDX-FileCopyrightText: 2001-2002 NaN Holding BV. All rights reserved.
|
|
*
|
|
* SPDX-License-Identifier: GPL-2.0-or-later */
|
|
|
|
/** \file
|
|
* \ingroup DNA
|
|
* \brief ID and Library types, which are fundamental for SDNA.
|
|
*/
|
|
|
|
#pragma once
|
|
|
|
#include "DNA_ID_enums.h"
|
|
#include "DNA_defs.h"
|
|
#include "DNA_listBase.h"
|
|
|
|
#include "BLI_assert.h"
|
|
#include "BLI_compiler_typecheck.h"
|
|
|
|
#include <cstring>
|
|
#include <type_traits>
|
|
|
|
namespace blender {
|
|
|
|
namespace bke::id {
|
|
struct ID_Runtime;
|
|
}
|
|
namespace bke {
|
|
struct PreviewImageRuntime;
|
|
}
|
|
namespace bke::idprop {
|
|
struct IDPropertyGroupChildrenSet;
|
|
}
|
|
namespace bke::library {
|
|
struct LibraryRuntime;
|
|
}
|
|
|
|
struct FileData;
|
|
struct GHash;
|
|
struct ID;
|
|
struct Library;
|
|
struct PackedFile;
|
|
struct UniqueName_Map;
|
|
|
|
struct IDPropertyUIData {
|
|
/** Tool-tip / property description pointer. Owned by the #IDProperty. */
|
|
char *description = nullptr;
|
|
/** RNA `subtype`, used for every type except string properties (#PropertySubType). */
|
|
int rna_subtype = 0;
|
|
|
|
char _pad[4] = {};
|
|
};
|
|
|
|
/* DNA version of #EnumPropertyItem. */
|
|
struct IDPropertyUIDataEnumItem {
|
|
/* Unique identifier, used for string lookup. */
|
|
char *identifier = nullptr;
|
|
/* UI name of the item. */
|
|
char *name = nullptr;
|
|
/* Optional description. */
|
|
char *description = nullptr;
|
|
/* Unique integer value, should never change. */
|
|
int value = 0;
|
|
/* Optional icon. */
|
|
int icon = 0;
|
|
};
|
|
|
|
/* IDP_UI_DATA_TYPE_INT */
|
|
struct IDPropertyUIDataInt {
|
|
IDPropertyUIData base;
|
|
int *default_array = nullptr; /* Only for array properties. */
|
|
int default_array_len = 0;
|
|
|
|
int min = 0;
|
|
int max = 0;
|
|
int soft_min = 0;
|
|
int soft_max = 0;
|
|
int step = 0;
|
|
int default_value = 0;
|
|
|
|
int enum_items_num = 0;
|
|
IDPropertyUIDataEnumItem *enum_items = nullptr;
|
|
};
|
|
|
|
/** For #IDP_UI_DATA_TYPE_BOOLEAN Use `int8_t` because DNA does not support `bool`. */
|
|
struct IDPropertyUIDataBool {
|
|
IDPropertyUIData base;
|
|
int8_t *default_array = nullptr; /* Only for array properties. */
|
|
int default_array_len = 0;
|
|
char _pad[3] = {};
|
|
|
|
int8_t default_value = 0;
|
|
};
|
|
|
|
/** For #IDP_UI_DATA_TYPE_FLOAT */
|
|
struct IDPropertyUIDataFloat {
|
|
IDPropertyUIData base;
|
|
double *default_array = nullptr; /* Only for array properties. */
|
|
int default_array_len = 0;
|
|
char _pad[4] = {};
|
|
|
|
float step = 0;
|
|
int precision = 0;
|
|
|
|
double min = 0;
|
|
double max = 0;
|
|
double soft_min = 0;
|
|
double soft_max = 0;
|
|
double default_value = 0;
|
|
};
|
|
|
|
/** For #IDP_UI_DATA_TYPE_STRING */
|
|
struct IDPropertyUIDataString {
|
|
IDPropertyUIData base;
|
|
char *default_value = nullptr;
|
|
};
|
|
|
|
/** For #IDP_UI_DATA_TYPE_ID. */
|
|
struct IDPropertyUIDataID {
|
|
IDPropertyUIData base;
|
|
/**
|
|
* #ID_Type. With python-defined properties, this type is not enforced. A value of `0` means any
|
|
* ID type.
|
|
*
|
|
* However, when defined/edited from the UI (Custom Properties panel), it must/will be defined,
|
|
* as generic 'Any ID type' selection is a TODO UI-wise.
|
|
*/
|
|
short id_type = 0;
|
|
char _pad[6] = {};
|
|
};
|
|
|
|
struct IDPropertyData {
|
|
void *pointer = nullptr;
|
|
ListBaseT<struct IDProperty> group = {nullptr, nullptr};
|
|
/**
|
|
* Allows constant time lookup by name of the children in this group. This may be null if the
|
|
* group is empty. The order may not be exactly the same as in #group.
|
|
*/
|
|
bke::idprop::IDPropertyGroupChildrenSet *children_map = nullptr;
|
|
/** NOTE: a `double` is written into two 32bit integers. */
|
|
int val = 0, val2 = 0;
|
|
};
|
|
|
|
struct IDProperty {
|
|
struct IDProperty *next = nullptr, *prev = nullptr;
|
|
eIDPropertyType type = IDP_STRING;
|
|
/**
|
|
* #eIDPropertySubType when `type` is #IDP_STRING.
|
|
* #eIDPropertyType for all other types.
|
|
*/
|
|
char subtype = 0;
|
|
eIDPropertyFlag flag = {};
|
|
char name[/*MAX_IDPROP_NAME*/ 64] = "";
|
|
|
|
char _pad0[4] = {};
|
|
|
|
/** NOTE: alignment for 64 bits. */
|
|
IDPropertyData data;
|
|
|
|
/**
|
|
* Array length, and importantly string length + 1.
|
|
* the idea is to be able to reuse array reallocation functions on strings.
|
|
*/
|
|
int len = 0;
|
|
/**
|
|
* Strings and arrays are both buffered, though the buffer isn't saved.
|
|
* `totallen` is total length of allocated array/string, including a buffer.
|
|
* \note the buffering is mild; see #IDP_ResizeIDPArray for details.
|
|
*/
|
|
int totallen = 0;
|
|
|
|
IDPropertyUIData *ui_data = nullptr;
|
|
};
|
|
|
|
#define MAX_IDPROP_NAME 64
|
|
#define DEFAULT_ALLOC_FOR_NULL_STRINGS 64
|
|
|
|
/* add any future new id property types here. */
|
|
|
|
/* Static ID override structs. */
|
|
|
|
/* IDOverrideLibraryPropertyOperation->operation. */
|
|
enum eID_OverrideLib_Op : short {
|
|
/* Basic operations. */
|
|
/** Special value, forbids any overriding. */
|
|
LIBOVERRIDE_OP_NOOP = 0,
|
|
|
|
/** Fully replace local value by reference one. */
|
|
LIBOVERRIDE_OP_REPLACE = 1,
|
|
|
|
/* Numeric-only operations. */
|
|
/** Add local value to reference one. */
|
|
LIBOVERRIDE_OP_ADD = 101,
|
|
/** Subtract local value from reference one (needed due to unsigned values etc.). */
|
|
LIBOVERRIDE_OP_SUBTRACT = 102,
|
|
/** Multiply reference value by local one (more useful than diff for scales and the like). */
|
|
LIBOVERRIDE_OP_MULTIPLY = 103,
|
|
|
|
/* Collection-only operations. */
|
|
/** Insert after given reference's subitem. */
|
|
LIBOVERRIDE_OP_INSERT_AFTER = 201,
|
|
/** Insert before given reference's subitem. */
|
|
LIBOVERRIDE_OP_INSERT_BEFORE = 202,
|
|
/* We can add more if needed (move, delete, ...). */
|
|
|
|
/**
|
|
* Custom operation, generic liboverride code does not handle these, and expect all custom
|
|
* handling callbacks to be defined on properties when they have this type of operations.
|
|
*/
|
|
/* Note: Currently there is no dedicated 'extra data' for these custom operations.
|
|
* If needs arise, the padding in `IDOverrideLibraryPropertyOperation` could be used for some
|
|
* custom data... */
|
|
LIBOVERRIDE_OP_CUSTOM = 255,
|
|
};
|
|
|
|
/* IDOverrideLibraryPropertyOperation->flag. */
|
|
enum eID_OverrideLib_OpFlag : short {
|
|
/** User cannot remove that override operation. */
|
|
LIBOVERRIDE_OP_FLAG_MANDATORY = 1 << 0,
|
|
/** User cannot change that override operation. */
|
|
LIBOVERRIDE_OP_FLAG_LOCKED = 1 << 1,
|
|
|
|
/**
|
|
* For overrides of ID pointers: this override still matches (follows) the hierarchy of the
|
|
* reference linked data.
|
|
*/
|
|
LIBOVERRIDE_OP_FLAG_IDPOINTER_MATCH_REFERENCE = 1 << 8,
|
|
/**
|
|
* For overrides of ID pointers within RNA collections: this override is using the ID
|
|
* pointer in addition to the item name (to fully disambiguate the reference, since IDs from
|
|
* different libraries can have a same name).
|
|
*/
|
|
LIBOVERRIDE_OP_FLAG_IDPOINTER_ITEM_USE_ID = 1 << 9,
|
|
};
|
|
ENUM_OPERATORS(eID_OverrideLib_OpFlag)
|
|
|
|
/* IDOverrideLibraryProperty->tag and IDOverrideLibraryPropertyOperation->tag. */
|
|
enum eID_OverrideLib_PropTag : short {
|
|
/** This override property (operation) is unused and should be removed by cleanup process. */
|
|
LIBOVERRIDE_PROP_OP_TAG_UNUSED = 1 << 0,
|
|
|
|
/** This override property is forbidden and should be restored to its linked reference value. */
|
|
LIBOVERRIDE_PROP_TAG_NEEDS_RETORE = 1 << 1,
|
|
};
|
|
ENUM_OPERATORS(eID_OverrideLib_PropTag)
|
|
|
|
struct IDOverrideLibraryPropertyOperation {
|
|
struct IDOverrideLibraryPropertyOperation *next = nullptr, *prev = nullptr;
|
|
|
|
/* Type of override. */
|
|
eID_OverrideLib_Op operation = {};
|
|
eID_OverrideLib_OpFlag flag = {};
|
|
|
|
/** Runtime, tags are common to both #IDOverrideProperty and #IDOverridePropertyOperation. */
|
|
eID_OverrideLib_PropTag tag = {};
|
|
char _pad0[2] = {};
|
|
|
|
/* Sub-item references, if needed (for arrays or collections only).
|
|
* We need both reference and local values to allow e.g. insertion into RNA collections
|
|
* (constraints, modifiers...).
|
|
* In RNA collection case, if names are defined, they are used in priority.
|
|
* Names are pointers (instead of char[64]) to save some space, NULL or empty string when unset.
|
|
* Indices are -1 when unset.
|
|
*
|
|
* NOTE: For insertion operations in RNA collections, reference may not actually exist in the
|
|
* linked reference data. It is used to identify the anchor of the insertion operation (i.e. the
|
|
* item after or before which the new local item should be inserted), in the local override. */
|
|
char *subitem_reference_name = nullptr;
|
|
char *subitem_local_name = nullptr;
|
|
int subitem_reference_index = 0;
|
|
int subitem_local_index = 0;
|
|
/** Additional pointer to an ID. Only used and relevant when the related RNA collection stores ID
|
|
* pointers, to help disambiguate cases where several IDs from different libraries have the exact
|
|
* same name. */
|
|
struct ID *subitem_reference_id = nullptr;
|
|
struct ID *subitem_local_id = nullptr;
|
|
|
|
/**
|
|
* A UI-only label to represent that operation.
|
|
*
|
|
* Typically used for collection items, when the `subitem_reference_name`/`subitem_local_name`
|
|
* are not available or not usable from a UI PoV.
|
|
* See e.g. its usage by `rna_NodesModifierBake_override_diff` for geonodes packed bakes items.
|
|
*/
|
|
char *label = nullptr;
|
|
/**
|
|
* A UI-only longer tooltip to represent that operation.
|
|
*
|
|
* Same as `label` above, but for usage in tooltips and other longer text representations.
|
|
*/
|
|
char *tooltip = nullptr;
|
|
|
|
#ifdef __cplusplus
|
|
bool operator==(const IDOverrideLibraryPropertyOperation &b) const;
|
|
#endif
|
|
};
|
|
|
|
/* IDOverrideLibrary->flag */
|
|
enum eID_OverrideLib_Flag : uint32_t {
|
|
/**
|
|
* The override data-block should not be considered as part of an override hierarchy (generally
|
|
* because it was created as an single override, outside of any hierarchy consideration).
|
|
*/
|
|
LIBOVERRIDE_FLAG_NO_HIERARCHY = 1 << 0,
|
|
/**
|
|
* The override ID is required for the system to work (because of ID dependencies), but is not
|
|
* seen as editable by the user.
|
|
*/
|
|
LIBOVERRIDE_FLAG_SYSTEM_DEFINED = 1 << 1,
|
|
};
|
|
ENUM_OPERATORS(eID_OverrideLib_Flag)
|
|
|
|
/** A single overridden property, contain all operations on this one. */
|
|
struct IDOverrideLibraryProperty {
|
|
struct IDOverrideLibraryProperty *next = nullptr, *prev = nullptr;
|
|
|
|
/**
|
|
* Path from ID to overridden property.
|
|
* *Does not* include indices/names for final arrays/collections items.
|
|
*/
|
|
char *rna_path = nullptr;
|
|
|
|
/**
|
|
* List of overriding operations applied to this property.
|
|
* Recreated as part of the diffing, so do not store any of these elsewhere.
|
|
*/
|
|
ListBaseT<IDOverrideLibraryPropertyOperation> operations = {nullptr, nullptr};
|
|
|
|
/**
|
|
* Runtime, tags are common to both IDOverrideLibraryProperty and
|
|
* IDOverrideLibraryPropertyOperation. */
|
|
eID_OverrideLib_PropTag tag = {};
|
|
char _pad[2] = {};
|
|
|
|
/** The property type matching the rna_path. */
|
|
unsigned int rna_prop_type = 0;
|
|
};
|
|
|
|
struct IDOverrideLibraryRuntime;
|
|
|
|
/* Main container for all overriding data info of a data-block. */
|
|
struct IDOverrideLibrary {
|
|
/** Reference linked ID which this one overrides. */
|
|
struct ID *reference = nullptr;
|
|
ListBaseT<IDOverrideLibraryProperty> properties = {nullptr, nullptr};
|
|
|
|
/**
|
|
* Override hierarchy root ID. Usually the actual root of the hierarchy, but not always
|
|
* in degenerated cases.
|
|
*
|
|
* All liboverrides of a same hierarchy (e.g. a character collection) share the same root.
|
|
*/
|
|
struct ID *hierarchy_root = nullptr;
|
|
|
|
IDOverrideLibraryRuntime *runtime = nullptr;
|
|
|
|
eID_OverrideLib_Flag flag = {};
|
|
char _pad_1[4] = {};
|
|
};
|
|
|
|
/* watch it: Strip has identical beginning. */
|
|
/**
|
|
* ID is the first thing included in all serializable types. It
|
|
* provides a common handle to place all data in double-linked lists.
|
|
*/
|
|
|
|
/* 2 characters for ID code and 256 for actual name */
|
|
#define MAX_ID_NAME 258
|
|
|
|
/** #ID_Runtime_Remap.status */
|
|
enum eID_RemapStatus : char {
|
|
/** new_id is directly linked in current .blend. */
|
|
ID_REMAP_IS_LINKED_DIRECT = 1 << 0,
|
|
/** There was some skipped 'user_one' usages of old_id. */
|
|
ID_REMAP_IS_USER_ONE_SKIPPED = 1 << 1,
|
|
};
|
|
ENUM_OPERATORS(eID_RemapStatus)
|
|
|
|
struct IDHash {
|
|
char data[16] = "";
|
|
|
|
#ifdef __cplusplus
|
|
uint64_t hash() const
|
|
{
|
|
return *reinterpret_cast<const uint64_t *>(this->data);
|
|
}
|
|
|
|
static constexpr IDHash get_null()
|
|
{
|
|
return {};
|
|
}
|
|
bool is_null() const
|
|
{
|
|
return *this == IDHash::get_null();
|
|
}
|
|
|
|
friend bool operator==(const IDHash &a, const IDHash &b)
|
|
{
|
|
return memcmp(a.data, b.data, sizeof(a.data)) == 0;
|
|
}
|
|
|
|
friend bool operator!=(const IDHash &a, const IDHash &b)
|
|
{
|
|
return !(a == b);
|
|
}
|
|
|
|
#endif
|
|
};
|
|
|
|
struct ID {
|
|
/* There's a nasty circular dependency here.... 'void *' to the rescue! I
|
|
* really wonder why this is needed. */
|
|
void *next = nullptr, *prev = nullptr;
|
|
struct ID *newid = nullptr;
|
|
|
|
struct Library *lib = nullptr;
|
|
|
|
/** If the ID is an asset, this pointer is set. Owning pointer. */
|
|
struct AssetMetaData *asset_data = nullptr;
|
|
|
|
/**
|
|
* Main identifier for this data-block. Must be unique within the ID name-space (defined by its
|
|
* type, and owning #Library).
|
|
*
|
|
* The first two bytes are always the #ID_Type code of the data-block's type.
|
|
*
|
|
* One critical usage is to reference external linked data. */
|
|
char name[/*MAX_ID_NAME*/ 258] = "";
|
|
|
|
/**
|
|
* ID_FLAG_... flags report on status of the data-block this ID belongs to
|
|
* (persistent, saved to and read from .blend).
|
|
*/
|
|
short flag = 0;
|
|
/**
|
|
* ID_TAG_... tags (runtime only, cleared at read time).
|
|
*/
|
|
int tag = 0;
|
|
int us = 0;
|
|
int icon_id = 0;
|
|
unsigned int recalc = 0;
|
|
/**
|
|
* Used by undo code. recalc_after_undo_push contains the changes between the
|
|
* last undo push and the current state. This is accumulated as IDs are tagged
|
|
* for update in the depsgraph, and only cleared on undo push.
|
|
*
|
|
* recalc_up_to_undo_push is saved to undo memory, and is the value of
|
|
* recalc_after_undo_push at the time of the undo push. This means it can be
|
|
* used to find the changes between undo states.
|
|
*/
|
|
unsigned int recalc_up_to_undo_push = 0;
|
|
unsigned int recalc_after_undo_push = 0;
|
|
|
|
/**
|
|
* A session-wide unique identifier for a given ID, that remain the same across potential
|
|
* re-allocations (e.g. due to undo/redo steps).
|
|
*/
|
|
unsigned int session_uid = 0;
|
|
|
|
/**
|
|
* This is only available on packed linked data-blocks. It is a hash of the contents the
|
|
* data-block including all its dependencies. It is computed when first packing the data-block
|
|
* and is not changed afterwards. It can be used to detect that packed data-blocks in two
|
|
* separate .blend files are the same.
|
|
*
|
|
* Two data-blocks with the same deep hash are assumed to be interchangeable, but not necessarily
|
|
* exactly the same. For example, it's possible to change node positions on packed data-blocks
|
|
* without changing the deep hash.
|
|
*/
|
|
IDHash deep_hash;
|
|
|
|
/**
|
|
* User-defined custom properties storage. Typically Accessed through the 'dict' syntax from
|
|
* Python.
|
|
*/
|
|
IDProperty *properties = nullptr;
|
|
|
|
/**
|
|
* System-defined custom properties storage. Used to store data dynamically defined either by
|
|
* Blender itself (e.g. the GeoNode modifier), or some python script, extension etc.
|
|
*
|
|
* Typically accessed through RNA paths (`C.object.my_dynamic_float_property = 33.3`), when
|
|
* wrapped/defined by RNA.
|
|
*/
|
|
IDProperty *system_properties = nullptr;
|
|
|
|
void *_pad1 = nullptr;
|
|
|
|
/** Reference linked ID which this one overrides. */
|
|
IDOverrideLibrary *override_library = nullptr;
|
|
|
|
/**
|
|
* Only set for data-blocks which are coming from copy-on-evaluation, points to
|
|
* the original version of it.
|
|
* Also used temporarily during memfile undo to keep a reference to old ID when found.
|
|
*/
|
|
struct ID *orig_id = nullptr;
|
|
|
|
/**
|
|
* Holds the #PyObject reference to the ID (initialized on demand).
|
|
*
|
|
* This isn't essential, it could be removed however it gives some advantages:
|
|
*
|
|
* - Every time the #ID is accessed a #BPy_StructRNA doesn't have to be created & destroyed
|
|
* (consider all the polling and drawing functions that access ID's).
|
|
*
|
|
* - When this #ID is deleted, the #BPy_StructRNA can be invalidated
|
|
* so accessing it from Python raises an exception instead of crashing.
|
|
*
|
|
* This is of limited benefit though, as it doesn't apply to non #ID data
|
|
* that references this ID (the bones of an armature or the modifiers of an object for
|
|
* example).
|
|
*/
|
|
void *py_instance = nullptr;
|
|
|
|
/**
|
|
* Weak reference to an ID in a given library file, used to allow re-using already appended data
|
|
* in some cases, instead of appending it again.
|
|
*
|
|
* May be NULL.
|
|
*/
|
|
struct LibraryWeakReference *library_weak_reference = nullptr;
|
|
|
|
/**
|
|
* Allocated runtime data, never written on disk or in undo steps.
|
|
*
|
|
* _Always_ valid for code handling IDs managed by the `BKE_lib_id` API.
|
|
*
|
|
* Internal low-level implementation of ID creation/copying/deletion, and code handling IDs
|
|
* themselves in non-standard ways (mainly the CoW IDs in depsgraph, and some temporary IDs in
|
|
* readfile) may have to manage this pointer themselves (see also #BKE_libblock_runtime_ensure
|
|
* and #BKE_libblock_free_runtime_data).
|
|
*/
|
|
bke::id::ID_Runtime *runtime = nullptr;
|
|
};
|
|
|
|
/**
|
|
* For each library file used, a Library struct is added to Main.
|
|
*/
|
|
struct Library {
|
|
#ifdef __cplusplus
|
|
/** See #ID_Type comment for why this is here. */
|
|
static constexpr ID_Type id_type = ID_LI;
|
|
#endif
|
|
|
|
ID id;
|
|
/** Path name used for reading, can be relative and edited in the outliner. */
|
|
char filepath[/*FILE_MAX*/ 1024] = "";
|
|
|
|
/** Flags defining specific characteristics of a library. See #LibraryFlag. */
|
|
uint16_t flag = 0;
|
|
/** #LibraryRuntime::tag need to be preserved in undo (memfile) case (#eLibrary_Tag). */
|
|
uint16_t undo_runtime_tag = 0;
|
|
char _pad[4] = {};
|
|
|
|
/**
|
|
* For archive library only (#LIBRARY_FLAG_IS_ARCHIVE): The main library owning it.
|
|
*
|
|
* `archive_parent_library` and `packedfile` should never be both non-null in a same Library ID.
|
|
*/
|
|
struct Library *archive_parent_library = nullptr;
|
|
|
|
/**
|
|
* Packed blend-file of the library, nullptr if not packed.
|
|
*
|
|
* \note Individual IDs may be packed even if the entire library is not packed.
|
|
*
|
|
* `archive_parent_library` and `packedfile` should never be both non-null in a same Library ID.
|
|
*/
|
|
struct PackedFile *packedfile = nullptr;
|
|
|
|
/**
|
|
* Runtime only data, never written in blendfile.
|
|
*
|
|
* Typically allocated when creating a new Library or reading it from a blendfile.
|
|
*/
|
|
bke::library::LibraryRuntime *runtime = nullptr;
|
|
|
|
void *_pad2 = nullptr;
|
|
};
|
|
|
|
/**
|
|
* A weak library/ID reference for local data that has been appended, to allow re-using that local
|
|
* data instead of creating a new copy of it in future appends.
|
|
*
|
|
* NOTE: This is by design a week reference, in other words code should be totally fine and perform
|
|
* a regular append if it cannot find a valid matching local ID.
|
|
*
|
|
* NOTE: There should always be only one single ID in current Main matching a given linked
|
|
* reference.
|
|
*/
|
|
struct LibraryWeakReference {
|
|
/** Expected to match a `Library.filepath`. */
|
|
char library_filepath[/*FILE_MAX*/ 1024] = "";
|
|
|
|
/** May be different from the current local ID name. */
|
|
char library_id_name[/*MAX_ID_NAME*/ 258] = "";
|
|
|
|
char _pad[2] = {};
|
|
};
|
|
|
|
/* PreviewImage.flag */
|
|
enum ePreviewImage_Flag {
|
|
PRV_CHANGED = (1 << 0),
|
|
/** If user-edited, do not auto-update this anymore! */
|
|
PRV_USER_EDITED = (1 << 1),
|
|
/* Rendering was invoked. Cleared on file read. */
|
|
PRV_RENDERING = (1 << 2),
|
|
};
|
|
|
|
/* PreviewImage.tag */
|
|
enum ePreviewImage_Tag : short {
|
|
/** Deferred preview is being loaded. */
|
|
PRV_TAG_DEFERRED_RENDERING = (1 << 1),
|
|
/** Deferred preview should be deleted asap. */
|
|
PRV_TAG_DEFERRED_DELETE = (1 << 2),
|
|
/** This deferred preview could not be loaded (e.g. not found on disk). */
|
|
PRV_TAG_DEFERRED_INVALID = (1 << 3),
|
|
/* Rendering was interrupted and needs restart. */
|
|
PRV_TAG_RESTART_RENDERING = (1 << 4),
|
|
};
|
|
ENUM_OPERATORS(ePreviewImage_Tag)
|
|
|
|
/**
|
|
* This type allows shallow copies. Use #BKE_previewimg_free() to release contained resources.
|
|
* Don't call this for shallow copies (or the original instance will have dangling pointers).
|
|
*/
|
|
struct PreviewImage {
|
|
DNA_DEFINE_CXX_METHODS(PreviewImage)
|
|
/* All values of 2 are really NUM_ICON_SIZES */
|
|
unsigned int w[2] = {};
|
|
unsigned int h[2] = {};
|
|
short flag[2] = {};
|
|
short changed_timestamp[2] = {};
|
|
unsigned int *rect[2] = {};
|
|
|
|
bke::PreviewImageRuntime *runtime = nullptr;
|
|
};
|
|
|
|
/**
|
|
* Amount of 'fake user' usages of this ID.
|
|
* Always 0 or 1.
|
|
*/
|
|
#define ID_FAKE_USERS(id) ((id_cast<const ID *>(id)->flag & ID_FLAG_FAKEUSER) ? 1 : 0)
|
|
/**
|
|
* Amount of defined 'extra' shallow, runtime-only usages of this ID (typically from UI).
|
|
* Always 0 or 1.
|
|
*
|
|
* \warning May not actually be part of the total #ID.us count, see #ID_EXTRA_REAL_USERS.
|
|
*/
|
|
#define ID_EXTRA_USERS(id) (id_cast<const ID *>(id)->tag & ID_TAG_EXTRAUSER ? 1 : 0)
|
|
/**
|
|
* Amount of real 'extra' shallow, runtime-only usages of this ID (typically from UI).
|
|
* Always 0 or 1.
|
|
*
|
|
* \note Actual number of usages added to #ID.us by these extra usages.
|
|
* May be 0 even if there are some 'extra' usages of this ID,
|
|
* when there are also other 'normal' reference-counting usages of it.
|
|
*/
|
|
#define ID_EXTRA_REAL_USERS(id) (id_cast<const ID *>(id)->tag & ID_TAG_EXTRAUSER_SET ? 1 : 0)
|
|
/**
|
|
* Amount of real usages of this ID (i.e. excluding the 'fake user' one, but including a potential
|
|
* 'extra' shallow/runtime usage).
|
|
*/
|
|
#define ID_REAL_USERS(id) (id_cast<const ID *>(id)->us - ID_FAKE_USERS(id))
|
|
/**
|
|
* Amount of 'normal' reference-counting usages of this ID
|
|
* (i.e. excluding the 'fake user' one, and a potential 'extra' shallow/runtime usage).
|
|
*/
|
|
#define ID_REFCOUNTING_USERS(id) (ID_REAL_USERS(id) - ID_EXTRA_REAL_USERS(id))
|
|
|
|
#define ID_CHECK_UNDO(id) (!ELEM(GS((id)->name), ID_SCR, ID_WM, ID_WS, ID_BR))
|
|
|
|
#define ID_BLEND_PATH(_bmain, _id) \
|
|
((_id)->lib ? BKE_main_blendfile_path_from_library(*(_id)->lib) : \
|
|
BKE_main_blendfile_path((_bmain)))
|
|
#define ID_BLEND_PATH_FROM_GLOBAL(_id) \
|
|
((_id)->lib ? BKE_main_blendfile_path_from_library(*(_id)->lib) : \
|
|
BKE_main_blendfile_path_from_global())
|
|
|
|
#define ID_MISSING(_id) ((id_cast<const ID *>(_id)->tag & ID_TAG_MISSING) != 0)
|
|
|
|
#define ID_IS_LINKED(_id) (id_cast<const ID *>(_id)->lib != NULL)
|
|
/**
|
|
* Indicates that this ID is linked but also packed into the current .blend file. Note that this
|
|
* just means that this specific ID and its dependencies are packed, not the entire library. So
|
|
* this is separate from #Library::packedfile.
|
|
*/
|
|
#define ID_IS_PACKED(_id) (ID_IS_LINKED(_id) && ((_id)->flag & ID_FLAG_LINKED_AND_PACKED))
|
|
|
|
#define ID_TYPE_SUPPORTS_ASSET_EDITABLE(id_type) \
|
|
ELEM(id_type, ID_BR, ID_TE, ID_NT, ID_IM, ID_PC, ID_MA)
|
|
|
|
#define ID_IS_EDITABLE(_id) \
|
|
((id_cast<const ID *>(_id)->lib == NULL) || \
|
|
((id_cast<const ID *>(_id)->lib->runtime->tag & LIBRARY_ASSET_EDITABLE) && \
|
|
ID_TYPE_SUPPORTS_ASSET_EDITABLE(GS(id_cast<const ID *>(_id)->name))))
|
|
|
|
/* Note that these are fairly high-level checks, should be used at user interaction level, not in
|
|
* BKE_library_override typically (especially due to the check on ID_TAG_EXTERN). */
|
|
#define ID_IS_OVERRIDABLE_LIBRARY_HIERARCHY(_id) \
|
|
(ID_IS_LINKED(_id) && !ID_MISSING(_id) && \
|
|
(BKE_idtype_get_info_from_id(id_cast<const ID *>(_id))->flags & IDTYPE_FLAGS_NO_LIBLINKING) == \
|
|
0 && \
|
|
!ELEM(GS((id_cast<const ID *>(_id))->name), ID_SCE))
|
|
#define ID_IS_OVERRIDABLE_LIBRARY(_id) \
|
|
(ID_IS_OVERRIDABLE_LIBRARY_HIERARCHY((_id)) && \
|
|
(id_cast<const ID *>(_id)->tag & ID_TAG_EXTERN) != 0)
|
|
|
|
/* NOTE: The three checks below do not take into account whether given ID is linked or not (when
|
|
* chaining overrides over several libraries). User must ensure the ID is not linked itself
|
|
* currently. */
|
|
/* TODO: add `_EDITABLE` versions of those macros (that would check if ID is linked or not)? */
|
|
#define ID_IS_OVERRIDE_LIBRARY_REAL(_id) \
|
|
(id_cast<const ID *>(_id)->override_library != NULL && \
|
|
id_cast<const ID *>(_id)->override_library->reference != NULL)
|
|
|
|
#define ID_IS_OVERRIDE_LIBRARY_VIRTUAL(_id) \
|
|
((id_cast<const ID *>(_id)->flag & ID_FLAG_EMBEDDED_DATA_LIB_OVERRIDE) != 0)
|
|
|
|
#define ID_IS_OVERRIDE_LIBRARY(_id) \
|
|
(ID_IS_OVERRIDE_LIBRARY_REAL(_id) || ID_IS_OVERRIDE_LIBRARY_VIRTUAL(_id))
|
|
|
|
#define ID_IS_OVERRIDE_LIBRARY_HIERARCHY_ROOT(_id) \
|
|
(!ID_IS_OVERRIDE_LIBRARY_REAL(_id) || \
|
|
(id_cast<ID *>(_id))->override_library->hierarchy_root == (id_cast<ID *>(_id)))
|
|
|
|
#define ID_IS_ASSET(_id) (id_cast<const ID *>(_id)->asset_data != NULL)
|
|
|
|
/* Check whether datablock type is covered by copy-on-evaluation. */
|
|
#define ID_TYPE_USE_COPY_ON_EVAL(_id_type) \
|
|
(!ELEM(_id_type, ID_LI, ID_SCR, ID_VF, ID_BR, ID_WM, ID_PAL, ID_PC, ID_WS, ID_IM))
|
|
|
|
/* Check whether data-block type requires copy-on-evaluation from #ID_RECALC_PARAMETERS.
|
|
* Keep in sync with #BKE_id_eval_properties_copy. */
|
|
#define ID_TYPE_SUPPORTS_PARAMS_WITHOUT_COW(id_type) ELEM(id_type, ID_ME)
|
|
|
|
/* This used to be ELEM(id_type, ID_IP), currently there is no deprecated ID
|
|
* type. ID_IP was removed in Blender 5.0. */
|
|
#define ID_TYPE_IS_DEPRECATED(id_type) false
|
|
|
|
#ifdef GS
|
|
# undef GS
|
|
#endif
|
|
#define GS(a) (CHECK_TYPE_ANY(a, char *, const char *), (ID_Type)(*((const short *)(a))))
|
|
|
|
#define ID_NEW_SET(_id, _idn) \
|
|
(((id_cast<ID *>)(_id))->newid = (id_cast<ID *>)(_idn), \
|
|
((id_cast<ID *>)(_id))->newid->tag |= ID_TAG_NEW, \
|
|
((id_cast<ID *>)(_id))->newid)
|
|
#define ID_NEW_REMAP(a) \
|
|
if ((a) && (a)->id.newid) { \
|
|
*(void **)&(a) = (a)->id.newid; \
|
|
} \
|
|
((void)0)
|
|
|
|
/** id->flag (persistent). */
|
|
enum eID_Flag : short {
|
|
/** Don't delete the data-block even if unused. */
|
|
ID_FLAG_FAKEUSER = 1 << 9,
|
|
/**
|
|
* The data-block is a sub-data of another one.
|
|
* Direct persistent references are not allowed.
|
|
*/
|
|
ID_FLAG_EMBEDDED_DATA = 1 << 10,
|
|
/**
|
|
* Data-block is from a library and linked indirectly, with ID_TAG_INDIRECT
|
|
* tag set. But the current .blend file also has a weak pointer to it that
|
|
* we want to restore if possible, and silently drop if it's missing.
|
|
*/
|
|
ID_FLAG_INDIRECT_WEAK_LINK = 1 << 11,
|
|
/**
|
|
* The data-block is a sub-data of another one, which is an override.
|
|
* Note that this also applies to shape-keys, even though they are not 100% embedded data.
|
|
*/
|
|
ID_FLAG_EMBEDDED_DATA_LIB_OVERRIDE = 1 << 12,
|
|
/**
|
|
* The override data-block appears to not be needed anymore after resync with linked data, but it
|
|
* was kept around (because e.g. detected as user-edited).
|
|
*/
|
|
ID_FLAG_LIB_OVERRIDE_RESYNC_LEFTOVER = 1 << 13,
|
|
/**
|
|
* This ID was explicitly copied as part of a clipboard copy operation.
|
|
*
|
|
* When reading the clipboard back, this can be used to differentiate between ID's that are
|
|
* intended to be part of the clipboard, and ID's that are only there because they are
|
|
* dependencies of the copied ones.
|
|
*
|
|
* This flag should never be set for 'normal' IDs in the global Main database.
|
|
*/
|
|
ID_FLAG_CLIPBOARD_MARK = 1 << 14,
|
|
/**
|
|
* Indicates that this linked ID is packed into the current .blend file. This should never be set
|
|
* on local ID (without one with a null `ID::lib` pointer).
|
|
*/
|
|
ID_FLAG_LINKED_AND_PACKED = short(1u << 15),
|
|
};
|
|
ENUM_OPERATORS(eID_Flag)
|
|
|
|
/**
|
|
* id->tag (runtime-only).
|
|
*
|
|
* Those tags belong to three different categories, which have different expected handling in
|
|
* code:
|
|
*
|
|
* - RESET_BEFORE_USE: piece of code that wants to use such flag has to ensure they are properly
|
|
* "reset" first.
|
|
* - RESET_AFTER_USE: piece of code that wants to use such flag has to ensure they are properly
|
|
* "reset" after usage (though "lifetime" of those flags is a bit fuzzy, e.g. _RECALC ones are
|
|
* reset on depsgraph evaluation...).
|
|
* - RESET_NEVER: these flags are 'status' ones, and never actually need any reset (except on
|
|
* initialization during `.blend` file reading).
|
|
*
|
|
* \note These tags are purely runtime, so changing there value is not an issue. When adding new
|
|
* tags, please put them in the relevant category and always keep their values strictly increasing.
|
|
*/
|
|
enum eID_Tag : int {
|
|
/**
|
|
* Long-life tags giving important info about general ID management.
|
|
*
|
|
* These tags are typically not changed often, if ever, during an ID's life.
|
|
*/
|
|
|
|
/**
|
|
* ID is from current .blend file.
|
|
*
|
|
* RESET_NEVER
|
|
*/
|
|
ID_TAG_LOCAL = 0,
|
|
/**
|
|
* ID is from a library, but is used (linked) directly by current .blend file.
|
|
*
|
|
* RESET_NEVER
|
|
*/
|
|
ID_TAG_EXTERN = 1 << 0,
|
|
/**
|
|
* ID is from a library, and is only used (linked) indirectly through other libraries.
|
|
*
|
|
* RESET_NEVER
|
|
*/
|
|
ID_TAG_INDIRECT = 1 << 1,
|
|
|
|
/**
|
|
* ID is considered as runtime, and should not be saved when writing .blend file, nor influence
|
|
* (in)direct status of linked data.
|
|
*
|
|
* Only meaningful for IDs belonging to regular Main database, all other cases are implicitly
|
|
* considered runtime-only.
|
|
*
|
|
* RESET_NEVER
|
|
*/
|
|
ID_TAG_RUNTIME = 1 << 2,
|
|
|
|
/**
|
|
* ID is a place-holder, an 'empty shell' (because the real one could not be linked from its
|
|
* library e.g.).
|
|
*
|
|
* RESET_NEVER
|
|
*/
|
|
ID_TAG_MISSING = 1 << 3,
|
|
|
|
/**
|
|
* ID has an extra virtual user (aka 'ensured real', as set by e.g. some editors, not to be
|
|
* confused with the `ID_FLAG_FAKEUSER` flag).
|
|
*
|
|
* RESET_NEVER
|
|
*
|
|
* \note This tag does not necessarily mean the actual user count of the ID is increased, this is
|
|
* defined by #ID_TAG_EXTRAUSER_SET.
|
|
*/
|
|
ID_TAG_EXTRAUSER = 1 << 4,
|
|
/**
|
|
* ID actually has increased user-count for the extra virtual user.
|
|
*
|
|
* RESET_NEVER
|
|
*/
|
|
ID_TAG_EXTRAUSER_SET = 1 << 5,
|
|
|
|
/**
|
|
* ID is up-to-date regarding its reference (only for library overrides).
|
|
*
|
|
* RESET_NEVER
|
|
*/
|
|
ID_TAG_LIBOVERRIDE_REFOK = 1 << 6,
|
|
/**
|
|
* ID needs an auto-diffing execution, if enabled (only for library overrides).
|
|
*
|
|
* RESET_NEVER
|
|
*/
|
|
ID_TAG_LIBOVERRIDE_AUTOREFRESH = 1 << 7,
|
|
/**
|
|
* ID is a library override that needs re-sync to its linked reference.
|
|
*
|
|
* \note Also used by readfile code when creating a missing ID placeholder if it is detected as
|
|
* being a linked liboverride ID.
|
|
*
|
|
* RESET_NEVER
|
|
*/
|
|
ID_TAG_LIBOVERRIDE_NEED_RESYNC = 1 << 8,
|
|
|
|
/**
|
|
* Short-life tags used during specific processes, like blend-file reading.
|
|
*/
|
|
|
|
/**
|
|
* ID is newly duplicated/copied (see #ID_NEW_SET macro above).
|
|
*
|
|
* RESET_AFTER_USE
|
|
*
|
|
* \note Also used internally in `readfile.cc` to mark data-blocks needing do_versions.
|
|
*/
|
|
ID_TAG_NEW = 1 << 12,
|
|
/**
|
|
* ID is already existing. Set before linking, to distinguish between existing data-blocks and
|
|
* newly linked ones.
|
|
*
|
|
* RESET_AFTER_USE
|
|
*/
|
|
ID_TAG_PRE_EXISTING = 1 << 13,
|
|
|
|
/**
|
|
* ID is being re-used from the old Main (instead of read from memfile), during memfile undo
|
|
* processing, because it was detected as unchanged.
|
|
*
|
|
* \note Also means that such ID does not need to be lib-linked during undo readfile process.
|
|
*
|
|
* RESET_AFTER_USE
|
|
*/
|
|
ID_TAG_UNDO_OLD_ID_REUSED_UNCHANGED = 1 << 17,
|
|
/**
|
|
* ID is being re-used from the old Main (instead of read from memfile), during memfile undo
|
|
* processing, because it is a 'NO_UNDO' type of ID.
|
|
*
|
|
* \note Also means that such ID does not need to be lib-linked during undo readfile process. It
|
|
* does need to be relinked in a different way however, doing a `session_uid`-based lookup into
|
|
* the newly read main database.
|
|
*
|
|
* RESET_AFTER_USE
|
|
*/
|
|
ID_TAG_UNDO_OLD_ID_REUSED_NOUNDO = 1 << 18,
|
|
/**
|
|
* ID has be re-read in-place, the ID address is the same as in the old main, but the content is
|
|
* different.
|
|
*
|
|
* RESET_AFTER_USE
|
|
*/
|
|
ID_TAG_UNDO_OLD_ID_REREAD_IN_PLACE = 1 << 19,
|
|
|
|
/* ------------------------------------------------------------------------------------------- */
|
|
/**
|
|
* Tags for special kind of IDs, not living in regular BMain data-base.
|
|
*
|
|
* IDs with such tags have typically a short life.
|
|
*/
|
|
|
|
/**
|
|
* ID is part of a temporary #Main which is expected to be freed in a short time-frame.
|
|
*
|
|
* RESET_NEVER
|
|
*
|
|
* Don't allow assigning this to non-temporary members (since it's likely to cause errors).
|
|
* When set #ID.session_uid isn't initialized, since the data isn't part of the session.
|
|
*/
|
|
ID_TAG_TEMP_MAIN = 1 << 20,
|
|
/** General ID management info, for freeing or copying behavior e.g. */
|
|
/**
|
|
* ID is not listed/stored in any #Main database.
|
|
*
|
|
* RESET_NEVER
|
|
*/
|
|
ID_TAG_NO_MAIN = 1 << 21,
|
|
/**
|
|
* ID is fully outside of any ID management area, and should be considered as a purely
|
|
* independent data.
|
|
*
|
|
* RESET_NEVER
|
|
*
|
|
* \note Only used by node-trees currently.
|
|
*/
|
|
ID_TAG_LOCALIZED = 1 << 22,
|
|
/**
|
|
* ID is an evaluated/localized version.
|
|
*
|
|
* RESET_NEVER
|
|
*
|
|
* \warning This should not be cleared on existing data.
|
|
* If support for this is needed, see #88026 as this flag controls memory ownership
|
|
* of physics *shared* pointers.
|
|
*/
|
|
ID_TAG_COPIED_ON_EVAL = 1 << 23,
|
|
/**
|
|
* ID is not the original evaluated ID created by the depsgraph, but has been re-allocated during
|
|
* the evaluation process of another ID.
|
|
*
|
|
* RESET_NEVER
|
|
*
|
|
* Typical example is object data, when evaluating the object's modifier stack the final obdata
|
|
* can be different than the evaluated initial obdata ID.
|
|
*/
|
|
ID_TAG_COPIED_ON_EVAL_FINAL_RESULT = 1 << 24,
|
|
|
|
/**
|
|
* ID management status tags related to non-standard BMain IDs.
|
|
*/
|
|
|
|
/**
|
|
* Datablock does not refcount usages of other IDs.
|
|
*
|
|
* RESET_NEVER
|
|
*/
|
|
ID_TAG_NO_USER_REFCOUNT = 1 << 25,
|
|
/**
|
|
* ID was not allocated by standard system (BKE_libblock_alloc), do not free its memory
|
|
* (usual type-specific freeing is called though).
|
|
*
|
|
* RESET_NEVER
|
|
*/
|
|
ID_TAG_NOT_ALLOCATED = 1 << 26,
|
|
|
|
/* ------------------------------------------------------------------------------------------- */
|
|
|
|
/**
|
|
* Free to use tag, often used in BKE code to mark IDs to be processed.
|
|
*
|
|
* RESET_BEFORE_USE
|
|
*
|
|
* \todo Make it a RESET_AFTER_USE too.
|
|
*/
|
|
ID_TAG_DOIT = int(1u << 31),
|
|
};
|
|
ENUM_OPERATORS(eID_Tag)
|
|
|
|
/**
|
|
* Most of ID tags are cleared on file write (i.e. also when storing undo steps), since they
|
|
* either have of very short lifetime (not expected to exist across undo steps), or are info that
|
|
* will be re-generated when reading undo steps.
|
|
*
|
|
* Linked/local tags also need to be preserved, since the addition of more complex 'linked data'
|
|
* modes & handlings (e.g. packed linked data are handled like local IDs when it comes to
|
|
* read/write of blend-files).
|
|
*
|
|
* However a few of these need to be explicitly preserved across undo steps.
|
|
*/
|
|
#define ID_TAG_KEEP_ON_UNDO \
|
|
(ID_TAG_EXTRAUSER | ID_TAG_MISSING | ID_TAG_RUNTIME | ID_TAG_LOCAL | ID_TAG_EXTERN | \
|
|
ID_TAG_INDIRECT)
|
|
|
|
/* Tag given ID for an update in all the dependency graphs. */
|
|
enum IDRecalcFlag {
|
|
/***************************************************************************
|
|
* Individual update tags, this is what ID gets tagged for update with. */
|
|
|
|
/* ** Object transformation changed. ** */
|
|
ID_RECALC_TRANSFORM = (1 << 0),
|
|
|
|
/* ** Geometry changed. **
|
|
*
|
|
* When object of armature type gets tagged with this flag, its pose is
|
|
* re-evaluated.
|
|
*
|
|
* When object of other type is tagged with this flag it makes the modifier
|
|
* stack to be re-evaluated.
|
|
*
|
|
* When object data type (mesh, curve, ...) gets tagged with this flag it
|
|
* makes all objects which shares this data-block to be updated.
|
|
*
|
|
* Note that the evaluation depends on the object-mode.
|
|
* So edit-mesh data for example only reevaluate with the updated edit-mesh.
|
|
* When geometry in the original ID has been modified #ID_RECALC_GEOMETRY_ALL_MODES
|
|
* must be used instead.
|
|
*
|
|
* When a collection gets tagged with this flag, all objects depending on the geometry and
|
|
* transforms on any of the objects in the collection are updated. */
|
|
ID_RECALC_GEOMETRY = (1 << 1),
|
|
|
|
/* ** Animation or time changed and animation is to be re-evaluated. ** */
|
|
ID_RECALC_ANIMATION = (1 << 2),
|
|
|
|
/* ** Particle system changed. ** */
|
|
/* Only do path-cache etc. */
|
|
ID_RECALC_PSYS_REDO = (1 << 3),
|
|
/* Reset everything including point-cache. */
|
|
ID_RECALC_PSYS_RESET = (1 << 4),
|
|
/* Only child settings changed. */
|
|
ID_RECALC_PSYS_CHILD = (1 << 5),
|
|
/* Physics type changed. */
|
|
ID_RECALC_PSYS_PHYS = (1 << 6),
|
|
|
|
/* ** Material and shading ** */
|
|
|
|
/* For materials and node trees this means that topology of the shader tree
|
|
* changed, and the shader is to be recompiled.
|
|
* For objects it means that the draw batch cache is to be redone. */
|
|
ID_RECALC_SHADING = (1 << 7),
|
|
/* TODO(sergey): Consider adding an explicit ID_RECALC_SHADING_PARAMETERS
|
|
* which can be used for cases when only socket value changed, to speed up
|
|
* redraw update in that case. */
|
|
|
|
/* Selection of the ID itself or its components (for example, vertices) did
|
|
* change, and all the drawing data is to be updated. */
|
|
ID_RECALC_SELECT = (1 << 9),
|
|
/* Flags on the base did change, and is to be copied onto all the copies of
|
|
* corresponding objects. */
|
|
ID_RECALC_BASE_FLAGS = (1 << 10),
|
|
ID_RECALC_POINT_CACHE = (1 << 11),
|
|
/* Only inform editors about the change. Is used to force update of editors
|
|
* when data-block which is not a part of dependency graph did change.
|
|
*
|
|
* For example, brush texture did change and the preview is to be
|
|
* re-rendered. */
|
|
ID_RECALC_EDITORS = (1 << 12),
|
|
|
|
/* ** Update evaluated copy component. **
|
|
*
|
|
* This is most generic tag which should only be used when nothing else matches.
|
|
* It is not to explicitly mixed in with other recalculation flags.
|
|
*/
|
|
ID_RECALC_SYNC_TO_EVAL = (1 << 13),
|
|
|
|
/* Sequences in the sequencer did change.
|
|
* Use this tag with a scene ID which owns the sequences. */
|
|
ID_RECALC_SEQUENCER_STRIPS = (1 << 14),
|
|
|
|
/* Runs on frame-change (used for seeking audio too). */
|
|
ID_RECALC_FRAME_CHANGE = (1 << 15),
|
|
|
|
ID_RECALC_AUDIO_FPS = (1 << 16),
|
|
ID_RECALC_AUDIO_VOLUME = (1 << 17),
|
|
ID_RECALC_AUDIO_MUTE = (1 << 18),
|
|
ID_RECALC_AUDIO_LISTENER = (1 << 19),
|
|
|
|
ID_RECALC_AUDIO = (1 << 20),
|
|
|
|
/* NOTE: This triggers copy-on-eval for types that require it.
|
|
* Exceptions to this can be added using #ID_TYPE_SUPPORTS_PARAMS_WITHOUT_COW,
|
|
* this has the advantage that large arrays stored in the idea data don't
|
|
* have to be copied on every update. */
|
|
ID_RECALC_PARAMETERS = (1 << 21),
|
|
|
|
/* Input has changed and data-block is to be reload from disk.
|
|
* Applies to movie clips to inform that copy-on-written version is to be refreshed for the new
|
|
* input file or for color space changes. */
|
|
ID_RECALC_SOURCE = (1 << 23),
|
|
|
|
/* Virtual recalc tag/marker required for undo in some cases, where actual data does not change
|
|
* and hence do not require an update, but conceptually we are dealing with something new.
|
|
*
|
|
* Current known case: linked IDs made local without requiring any copy. While their users do not
|
|
* require any update, they have actually been 'virtually' remapped from the linked ID to the
|
|
* local one.
|
|
*/
|
|
ID_RECALC_TAG_FOR_UNDO = (1 << 24),
|
|
|
|
/* The node tree has changed in a way that affects its output nodes. */
|
|
ID_RECALC_NTREE_OUTPUT = (1 << 25),
|
|
|
|
/* Hierarchy of collection and object within collection changed. */
|
|
ID_RECALC_HIERARCHY = (1 << 26),
|
|
|
|
/* Provisioned flags.
|
|
*
|
|
* Not for actual use. The idea of them is to have all bits of the `IDRecalcFlag` defined to a
|
|
* known value, silencing sanitizer warnings when checking bits of the ID_RECALC_ALL. */
|
|
ID_RECALC_PROVISION_27 = (1 << 27),
|
|
ID_RECALC_PROVISION_28 = (1 << 28),
|
|
ID_RECALC_PROVISION_29 = (1 << 29),
|
|
ID_RECALC_PROVISION_30 = (1 << 30),
|
|
ID_RECALC_PROVISION_31 = (1u << 31),
|
|
|
|
/***************************************************************************
|
|
* Pseudonyms, to have more semantic meaning in the actual code without
|
|
* using too much low-level and implementation specific tags. */
|
|
|
|
/* Update animation data-block itself, without doing full re-evaluation of
|
|
* all dependent objects. */
|
|
ID_RECALC_ANIMATION_NO_FLUSH = ID_RECALC_SYNC_TO_EVAL,
|
|
|
|
/* Ensure geometry of object and edit modes are both up-to-date in the evaluated data-block.
|
|
* Example usage is when mesh validation modifies the non-edit-mode data,
|
|
* which we want to be copied over to the evaluated data-block. */
|
|
ID_RECALC_GEOMETRY_ALL_MODES = ID_RECALC_GEOMETRY | ID_RECALC_SYNC_TO_EVAL,
|
|
|
|
/***************************************************************************
|
|
* Aggregate flags, use only for checks on runtime.
|
|
* Do NOT use those for tagging. */
|
|
|
|
/* Identifies that SOMETHING has been changed in this ID. */
|
|
ID_RECALC_ALL = (0xffffffff),
|
|
|
|
/* Identifies that something in particle system did change. */
|
|
ID_RECALC_PSYS_ALL = (ID_RECALC_PSYS_REDO | ID_RECALC_PSYS_RESET | ID_RECALC_PSYS_CHILD |
|
|
ID_RECALC_PSYS_PHYS),
|
|
|
|
};
|
|
|
|
/* To filter ID types (filter_id). 64 bit to fit all types. */
|
|
#define FILTER_ID_AC (1ULL << 0)
|
|
#define FILTER_ID_AR (1ULL << 1)
|
|
#define FILTER_ID_BR (1ULL << 2)
|
|
#define FILTER_ID_CA (1ULL << 3)
|
|
#define FILTER_ID_CU_LEGACY (1ULL << 4)
|
|
#define FILTER_ID_GD_LEGACY (1ULL << 5)
|
|
#define FILTER_ID_GR (1ULL << 6)
|
|
#define FILTER_ID_IM (1ULL << 7)
|
|
#define FILTER_ID_LA (1ULL << 8)
|
|
#define FILTER_ID_LS (1ULL << 9)
|
|
#define FILTER_ID_LT (1ULL << 10)
|
|
#define FILTER_ID_MA (1ULL << 11)
|
|
#define FILTER_ID_MB (1ULL << 12)
|
|
#define FILTER_ID_MC (1ULL << 13)
|
|
#define FILTER_ID_ME (1ULL << 14)
|
|
#define FILTER_ID_MSK (1ULL << 15)
|
|
#define FILTER_ID_NT (1ULL << 16)
|
|
#define FILTER_ID_OB (1ULL << 17)
|
|
#define FILTER_ID_PAL (1ULL << 18)
|
|
#define FILTER_ID_PC (1ULL << 19)
|
|
#define FILTER_ID_SCE (1ULL << 20)
|
|
#define FILTER_ID_SPK (1ULL << 21)
|
|
#define FILTER_ID_SO (1ULL << 22)
|
|
#define FILTER_ID_TE (1ULL << 23)
|
|
#define FILTER_ID_TXT (1ULL << 24)
|
|
#define FILTER_ID_VF (1ULL << 25)
|
|
#define FILTER_ID_WO (1ULL << 26)
|
|
#define FILTER_ID_PA (1ULL << 27)
|
|
#define FILTER_ID_CF (1ULL << 28)
|
|
#define FILTER_ID_WS (1ULL << 29)
|
|
#define FILTER_ID_LP (1ULL << 31)
|
|
#define FILTER_ID_CV (1ULL << 32)
|
|
#define FILTER_ID_PT (1ULL << 33)
|
|
#define FILTER_ID_VO (1ULL << 34)
|
|
#define FILTER_ID_SIM (1ULL << 35)
|
|
#define FILTER_ID_KE (1ULL << 36)
|
|
#define FILTER_ID_SCR (1ULL << 37)
|
|
#define FILTER_ID_WM (1ULL << 38)
|
|
#define FILTER_ID_LI (1ULL << 39)
|
|
#define FILTER_ID_GP (1ULL << 40)
|
|
#define FILTER_ID_IP (1ULL << 41)
|
|
|
|
#define FILTER_ID_ALL \
|
|
(FILTER_ID_AC | FILTER_ID_AR | FILTER_ID_BR | FILTER_ID_CA | FILTER_ID_CU_LEGACY | \
|
|
FILTER_ID_GD_LEGACY | FILTER_ID_GR | FILTER_ID_IM | FILTER_ID_LA | FILTER_ID_LS | \
|
|
FILTER_ID_LT | FILTER_ID_MA | FILTER_ID_MB | FILTER_ID_MC | FILTER_ID_ME | FILTER_ID_MSK | \
|
|
FILTER_ID_NT | FILTER_ID_OB | FILTER_ID_PA | FILTER_ID_PAL | FILTER_ID_PC | FILTER_ID_SCE | \
|
|
FILTER_ID_SPK | FILTER_ID_SO | FILTER_ID_TE | FILTER_ID_TXT | FILTER_ID_VF | FILTER_ID_WO | \
|
|
FILTER_ID_CF | FILTER_ID_WS | FILTER_ID_LP | FILTER_ID_CV | FILTER_ID_PT | FILTER_ID_VO | \
|
|
FILTER_ID_SIM | FILTER_ID_KE | FILTER_ID_SCR | FILTER_ID_WM | FILTER_ID_LI | FILTER_ID_GP | \
|
|
FILTER_ID_IP)
|
|
|
|
/**
|
|
* This enum defines the index assigned to each type of IDs in the array returned by
|
|
* #BKE_main_lists_get, and by extension, controls the default order in which each ID type is
|
|
* processed during standard 'foreach' looping over all IDs of a #Main data-base.
|
|
*
|
|
* About Order:
|
|
* ------------
|
|
*
|
|
* This is (loosely) defined with a relationship order in mind, from lowest level (ID types using,
|
|
* referencing almost no other ID types) to highest level (ID types potentially using many other ID
|
|
* types).
|
|
*
|
|
* So e.g. it ensures that this dependency chain is respected:
|
|
* #Material <- #Mesh <- #Object <- #Collection <- #Scene
|
|
*
|
|
* Default order of processing of IDs in 'foreach' macros (#FOREACH_MAIN_ID_BEGIN and the like),
|
|
* built on top of #BKE_main_lists_get, is actually reversed compared to the order defined here,
|
|
* since processing usually needs to happen on users before it happens on used IDs (when freeing
|
|
* e.g.).
|
|
*
|
|
* DO NOT rely on this order as being full-proofed dependency order, there are many cases were it
|
|
* can be violated (most obvious cases being custom properties and drivers, which can reference any
|
|
* other ID types).
|
|
*
|
|
* However, this order can be considered as an optimization heuristic, especially when processing
|
|
* relationships in a non-recursive pattern: in typical cases, a vast majority of those
|
|
* relationships can be processed fine in the first pass, and only few additional passes are
|
|
* required to address all remaining relationship cases.
|
|
* See e.g. how #BKE_library_unused_linked_data_set_tag is doing this.
|
|
*/
|
|
enum eID_Index {
|
|
/* Special case: Library, should never ever depend on any other type. */
|
|
INDEX_ID_LI = 0,
|
|
|
|
/* Animation types, might be used by almost all other types. */
|
|
INDEX_ID_AC,
|
|
|
|
/* Grease Pencil, special case, should be with the other obdata, but it can also be used by many
|
|
* other ID types, including node trees e.g.
|
|
* So there is no proper place for those, for now keep close to the lower end of the processing
|
|
* hierarchy, but we may want to re-evaluate that at some point. */
|
|
INDEX_ID_GD_LEGACY,
|
|
|
|
/* Node trees, abstraction for procedural data, potentially used by many other ID types.
|
|
*
|
|
* NOTE: While node trees can also use many other ID types, they should not /own/ any of those,
|
|
* while they are being owned by many other ID types. This is why they are placed here. */
|
|
INDEX_ID_NT,
|
|
|
|
/* File-wrapper types, those usually 'embed' external files in Blender, with no dependencies to
|
|
* other ID types. */
|
|
INDEX_ID_VF,
|
|
INDEX_ID_TXT,
|
|
INDEX_ID_SO,
|
|
|
|
/* Image/movie types, can be used by shading ID types, but also directly by Objects, Scenes, etc.
|
|
*/
|
|
INDEX_ID_MSK,
|
|
INDEX_ID_IM,
|
|
INDEX_ID_MC,
|
|
|
|
/* Shading types. */
|
|
INDEX_ID_TE,
|
|
INDEX_ID_MA,
|
|
INDEX_ID_LS,
|
|
INDEX_ID_WO,
|
|
|
|
/* Simulation-related types. */
|
|
INDEX_ID_CF,
|
|
INDEX_ID_PA,
|
|
|
|
/* Shape Keys snow-flake, can be used by several obdata types. */
|
|
INDEX_ID_KE,
|
|
|
|
/* Object data types. */
|
|
INDEX_ID_AR,
|
|
INDEX_ID_ME,
|
|
INDEX_ID_CU_LEGACY,
|
|
INDEX_ID_MB,
|
|
INDEX_ID_CV,
|
|
INDEX_ID_PT,
|
|
INDEX_ID_VO,
|
|
INDEX_ID_LT,
|
|
INDEX_ID_LA,
|
|
INDEX_ID_CA,
|
|
INDEX_ID_SPK,
|
|
INDEX_ID_LP,
|
|
INDEX_ID_GP,
|
|
|
|
/* Collection and object types. */
|
|
INDEX_ID_OB,
|
|
INDEX_ID_GR,
|
|
|
|
/* Preset-like, not-really-data types, can use many other ID types but should never be used by
|
|
* any actual data type (besides Scene, due to tool settings). */
|
|
INDEX_ID_PAL,
|
|
INDEX_ID_PC,
|
|
INDEX_ID_BR,
|
|
|
|
/* Scene, after preset-like ID types because of tool settings. */
|
|
INDEX_ID_SCE,
|
|
|
|
/* UI-related types, should never be used by any other data type. */
|
|
INDEX_ID_SCR,
|
|
INDEX_ID_WS,
|
|
INDEX_ID_WM,
|
|
|
|
/* Special values, keep last. */
|
|
INDEX_ID_NULL,
|
|
};
|
|
|
|
#define INDEX_ID_MAX (INDEX_ID_NULL + 1)
|
|
|
|
#ifdef __cplusplus
|
|
namespace dna {
|
|
namespace detail {
|
|
template<typename, typename = void> struct has_ID_member : std::false_type {};
|
|
template<typename T> struct has_ID_member<T, std::void_t<decltype(&T::id)>> : std::true_type {};
|
|
template<typename T> constexpr bool has_ID_as_first_member()
|
|
{
|
|
if constexpr (std::is_standard_layout_v<T> && has_ID_member<T>::value) {
|
|
return offsetof(T, id) == 0 && std::is_same_v<decltype(T::id), ID>;
|
|
}
|
|
else {
|
|
return false;
|
|
}
|
|
}
|
|
} // namespace detail
|
|
|
|
/**
|
|
* Type trait to check if a type is a ID data-block. It just actually checks whether the type has
|
|
* #ID is first data member, which should be good enough in practice.
|
|
*/
|
|
template<typename T>
|
|
constexpr bool is_ID_v = detail::has_ID_as_first_member<T>() || std::is_same_v<T, ID>;
|
|
|
|
} // namespace dna
|
|
|
|
namespace dna::detail {
|
|
template<typename Dst, typename Src, typename SrcRuntime>
|
|
constexpr void id_cast_assert([[maybe_unused]] SrcRuntime *src)
|
|
{
|
|
static_assert(dna::is_ID_v<Src>);
|
|
static_assert(dna::is_ID_v<Dst>);
|
|
if constexpr (std::is_same_v<Src, ID> && !std::is_same_v<Dst, ID>) {
|
|
/* Runtime check for when converting from #ID to subtype like #Object. */
|
|
BLI_assert(src == nullptr || GS(src->name) == Dst::id_type);
|
|
}
|
|
else if constexpr (!std::is_same_v<Src, ID> && std::is_same_v<Dst, ID>) {
|
|
/* Converting from subtype like #Object to #ID is always allowed. */
|
|
}
|
|
else {
|
|
/* Converting between the same types is always allowed. */
|
|
static_assert(std::is_same_v<Src, Dst>);
|
|
}
|
|
}
|
|
} // namespace dna::detail
|
|
|
|
/**
|
|
* A drop-in replacement for `reinterpret_cast` that does additional checks:
|
|
* - Static check that the source and destination types are data-block types.
|
|
* - Run-time assert when down-casting from #ID to e.g. #Object.
|
|
*
|
|
* \note This can't be used with forward-declared types as the type information is necessary for
|
|
* the additional checks. For the same reason, it also can't be used to convert from void pointers.
|
|
*/
|
|
template<typename Dst, typename Src> inline Dst id_cast(Src &&id)
|
|
{
|
|
using DstDecay = std::decay_t<Dst>;
|
|
using SrcDecay = std::decay_t<Src>;
|
|
static_assert(std::is_pointer_v<SrcDecay> == std::is_pointer_v<DstDecay>);
|
|
if constexpr (std::is_pointer_v<SrcDecay>) {
|
|
dna::detail::id_cast_assert<std::decay_t<std::remove_pointer_t<DstDecay>>,
|
|
std::decay_t<std::remove_pointer_t<SrcDecay>>>(id);
|
|
}
|
|
else {
|
|
static_assert(std::is_lvalue_reference_v<Src> && std::is_lvalue_reference_v<Dst>);
|
|
dna::detail::id_cast_assert<DstDecay, SrcDecay>(&id);
|
|
}
|
|
/* This also makes sure that we don't cast away constness. */
|
|
return reinterpret_cast<Dst>(id);
|
|
}
|
|
|
|
#endif
|
|
|
|
} // namespace blender
|