UT_COWWriteHandle< Data > Class Template Reference
#include <UT_COW.h>
List of all members.
Detailed Description
template<typename Data>
class UT_COWWriteHandle< Data >
A write handle can be thought of as a pointer to the actual data. It allows you to modify the underlying data. To do this, it will unique the plain-handle's data so no outstanding readlocks will be affected by the writes.
Building read handles while a write handle is active is valid so long as no dereference occurs during the lifetime of the read handle.
- Examples:
-
standalone/gengeovolume.C, and standalone/geo2voxel.C.
Definition at line 233 of file UT_COW.h.
Constructor & Destructor Documentation
Member Function Documentation
Test if this is a NULL pointer.
Definition at line 300 of file UT_COW.h.
Converting from a handle into a write handle can only be done once. This prevents other threads or methods from invalidating our uniqueness property. Note we up case from a *pointer* to a UT_COWHandle, not from a UT_COWHandle. This avoids people accidentally using an temporary as the source and underlines that write handles are like pointers while read handles are like values.
Definition at line 278 of file UT_COW.h.
Passing write handles by value is fine as the users understand they all share the same underlying data, so is responsible for thread safety, etc.
Definition at line 261 of file UT_COW.h.
Turns this into a null handle to drop references.
Definition at line 293 of file UT_COW.h.
The documentation for this class was generated from the following file:
