Implements I/O for NanoVDB grids. Features optional BLOSC and ZIP file compression, support for multiple grids per file as well as multiple grid types.
More...
|
| template<typename BufferT> |
| void | writeGrid (const std::string &fileName, const GridHandle< BufferT > &handle, io::Codec codec=io::Codec::NONE, int verbose=0) |
| | Write a single grid to file (over-writing existing content of the file)
|
| template<typename BufferT = HostBuffer, template< typename... > class VecT = std::vector> |
| void | writeGrids (const std::string &fileName, const VecT< GridHandle< BufferT > > &handles, Codec codec=Codec::NONE, int verbose=0) |
| | Write multiple grids to file (over-writing existing content of the file)
|
| template<typename BufferT = HostBuffer> |
| GridHandle< BufferT > | readGrid (const std::string &fileName, int n=0, int verbose=0, const BufferT &buffer=BufferT()) |
| | Read and return one or all grids from a file into a single GridHandle.
|
| template<typename BufferT = HostBuffer> |
| GridHandle< BufferT > | readGrid (const std::string &fileName, const std::string &gridName, int verbose=0, const BufferT &buffer=BufferT()) |
| | Read and return the first grid with a specific name from a file.
|
| template<typename BufferT = HostBuffer, template< typename... > class VecT = std::vector> |
| VecT< GridHandle< BufferT > > | readGrids (const std::string &fileName, int verbose=0, const BufferT &buffer=BufferT()) |
| | Read all the grids in the file and return them as a vector of multiple GridHandles, each containing all grids encoded in the same segment of the file (i.e. they where written together). This method also works if the file contains a raw grid buffer in which case a single GridHandle is returned.
|
| uint64_t | stringHash (const char *cstr) |
| | Internal functions for compressed read/write of a NanoVDB GridHandle into a stream.
|
| uint64_t | stringHash (const std::string &str) |
| | Return a uint64_t hash key of a std::string.
|
| uint64_t | reverseEndianness (uint64_t val) |
| | Return a uint64_t with its bytes reversed so we can check for endianness.
|
| bool | hasGrid (const std::string &fileName, const std::string &gridName) |
| | Return true if the file contains a grid with the specified name.
|
| bool | hasGrid (std::istream &is, const std::string &gridName) |
| | Return true if the stream contains a grid with the specified name.
|
| std::vector< FileGridMetaData > | readGridMetaData (const std::string &fileName) |
| | Reads and returns a vector of meta data for all the grids found in the specified file.
|
| std::vector< FileGridMetaData > | readGridMetaData (std::istream &is) |
| | Reads and returns a vector of meta data for all the grids found in the specified stream.
|
| template<typename BufferT> |
| void | writeGrid (std::ostream &os, const GridHandle< BufferT > &handle, Codec codec) |
| template<typename BufferT = HostBuffer, template< typename... > class VecT = std::vector> |
| void | writeGrids (std::ostream &os, const VecT< GridHandle< BufferT > > &handles, Codec codec=Codec::NONE) |
| template<typename BufferT> |
| GridHandle< BufferT > | readGrid (std::istream &is, int n, const BufferT &pool) |
| template<typename BufferT> |
| GridHandle< BufferT > | readGrid (std::istream &is, const std::string &gridName, const BufferT &pool) |
| | Read a specific grid from an input stream given the name of the grid.
|
| template<typename BufferT = HostBuffer, template< typename... > class VecT = std::vector> |
| VecT< GridHandle< BufferT > > | readGrids (std::istream &is, const BufferT &pool=BufferT()) |
| template<typename T> |
| std::ostream & | operator<< (std::ostream &os, const nanovdb::math::BBox< nanovdb::math::Vec3< T > > &b) |
| std::ostream & | operator<< (std::ostream &os, const nanovdb::CoordBBox &b) |
| std::ostream & | operator<< (std::ostream &os, const nanovdb::Coord &ijk) |
| template<typename T> |
| std::ostream & | operator<< (std::ostream &os, const nanovdb::math::Vec3< T > &v) |
| template<typename T> |
| std::ostream & | operator<< (std::ostream &os, const nanovdb::math::Vec4< T > &v) |
Implements I/O for NanoVDB grids. Features optional BLOSC and ZIP file compression, support for multiple grids per file as well as multiple grid types.
- Author
- Ken Museth
- Date
- May 1, 2020
- Note
- This file does NOT depend on OpenVDB, but optionally on ZIP and BLOSC
NanoVDB files take on one of the two following formats: 1) multiple segments each with multiple grids (segments have easy to access metadata about its grids) 2) starting with verion 32.6.0 nanovdb files also support a raw buffer with one or more grids (just a dump of a raw grid buffer, so no new metadata in headers as when using segments mentioned above).
Example of case 1: | <---------------------------------— segment 1 with N grids --------------------------------------> | <— segment 2 ... FileHeader, FileMetaData0, gridName0...FileMetaDataN, gridNameN, compressed Grid0, ... compressed GridN FileHeader ... Example of case 2: | <– grid buffer ---> | Grid0, Grid1, ... GridN
Note that FileHeader and FileMetaData (both defined in NanoVDB.h) have fixed sizes of respectively 16B and 176B. However, GridNameX and GridX have variable sizes!
1.14.0