Record Class FloatPtr

java.lang.Object
java.lang.Record
club.doki7.ffm.ptr.FloatPtr
All Implemented Interfaces:
IPointer, Iterable<Float>
@ValueBasedCandidate @UnsafeConstructor public record FloatPtr(@NotNull MemorySegment segment) extends Record implements IPointer, Iterable<Float>

Represents a pointer to 32-bit float(s) in native memory

The property segment() should always be not-null (segment != NULL && !segment.equals(MemorySegment.NULL)), and properly aligned to MemoryLayout.byteAlignment() bytes. To represent null pointer, you may use a Java null instead. See the documentation of IPointer.segment() for more details.

The constructor of this class is marked as UnsafeConstructor, because it does not perform any runtime check. The constructor can be useful for automatic code generators. For normal users, checked(MemorySegment) is a good safe alternative.

  • Constructor Details

    • FloatPtr

      public FloatPtr(@NotNull @NotNull MemorySegment segment)
      Creates an instance of a FloatPtr record class.
      Parameters:
      segment - the value for the segment record component

  • Method Details

    • size

      public long size()

    • read

      public float read()

    • write

      public void write(float value)

    • read

      public float read(long index)

    • write

      public void write(long index, float value)

    • write

      public void write(float @NotNull [] array)

    • writeV

      public void writeV(float value0, float @NotNull ... values)

    • reinterpret

      @Unsafe @NotNull public @NotNull FloatPtr reinterpret(long newSize)

      Assume the FloatPtr is capable of holding at least newSize floats, create a new view FloatPtr that uses the same backing storage as this FloatPtr, but with the new size. Since there is actually no way to really check whether the new size is valid, while buffer overflow is undefined behavior, this method is marked as Unsafe.

      This method could be useful when handling data returned from some C API, where the size of the data is not known in advance.

      If the size of the underlying segment is actually known in advance and correctly set, and you want to create a shrunk view, you may use slice(long) (with validation) instead.

    • offset

      @NotNull public @NotNull FloatPtr offset(long offset)

    • slice

      @NotNull public @NotNull FloatPtr slice(long start, long end)
      Note that this function uses the List.subList(int, int) semantics (left inclusive, right exclusive interval), not MemorySegment.asSlice(long, long) semantics (offset + newSize). Be careful with the difference.

    • slice

      @NotNull public @NotNull FloatPtr slice(long end)

    • iterator

      @NotNull public @NotNull Iterator<Float> iterator()
      Specified by:
      iterator in interface Iterable<Float>

    • checked

      @Contract("null -> null") @Nullable public static @Nullable FloatPtr checked(@Nullable @Nullable MemorySegment segment)

      Create a new FloatPtr using segment as backing storage, with argument validation.

      This function does not ensure segment's size to be a multiple of Float.BYTES, since that several trailing bytes could be automatically ignored by size() method, and usually these bytes does not interfere with FFI operations. If segment is not big enough to hold at least one float, that segment is simply considered "empty". See the documentation of IPointer.segment() for more details.

      Parameters:
      segment - the MemorySegment to use as the backing storage
      Returns:
      null if segment is null or MemorySegment.NULL, otherwise a new FloatPtr that uses segment as backing storage
      Throws:
      IllegalArgumentException - if segment is not native or not properly aligned

    • checked

      @NotNull public static @NotNull FloatPtr checked(@NotNull @NotNull FloatBuffer buffer)

      Create a new FloatPtr using the same backing storage as buffer, with argument validation.

      The main difference between this static method and the allocate(Arena, FloatBuffer) method is that this method does not copy the contents of the buffer into a newly allocated MemorySegment. Instead, the newly created FloatPtr will use the same backing storage as buffer. Thus, modification from one side will be visible on the other side.

      Be careful with java.nio buffer types' Buffer.position() property: only the "remaining" (from Buffer.position() to Buffer.limit()) part of buffer will be referred. If you have ever read from buffer, and you want all the contents of buffer to be referred, you may want to call Buffer.rewind().

      When handling data types consisting of multiple bytes, also be careful with endianness and FloatBuffer.order() property. FloatPtr always uses the native endianness. So if buffer uses a different endianness, you may want to convert it to the native endianness first.

      Parameters:
      buffer - the FloatBuffer to use as the backing storage
      Returns:
      a new FloatPtr that uses buffer as its backing storage
      Throws:
      IllegalArgumentException - if buffer is not direct, or its backing storage is not properly aligned

    • allocate

      @NotNull public static @NotNull FloatPtr allocate(@NotNull @NotNull Arena arena)

    • allocate

      @NotNull public static @NotNull FloatPtr allocate(@NotNull @NotNull Arena arena, long size)

    • allocate

      @NotNull public static @NotNull FloatPtr allocate(@NotNull @NotNull Arena arena, float @NotNull [] array)

    • allocateV

      @NotNull public static @NotNull FloatPtr allocateV(@NotNull @NotNull Arena arena, float value0, float... values)

    • allocate

      @NotNull public static @NotNull FloatPtr allocate(@NotNull @NotNull Arena arena, @NotNull @NotNull FloatBuffer buffer)

      Allocate a new FloatPtr in arena and copy the contents of buffer into the newly allocated FloatPtr.

      Be careful with java.nio buffer types' Buffer.position() property: only the "remaining" (from Buffer.position() to Buffer.limit()) part of buffer will be copied. If you have ever read from buffer, and you want all the contents of buffer to be copied, you may want to call Buffer.rewind().

      When handling data types consisting of multiple bytes, also be careful with endianness and FloatBuffer.order() property. FloatPtr always uses the native endianness. So if buffer uses a different endianness, you may want to convert it to the native endianness first.

      Parameters:
      arena - the Arena to allocate the new FloatPtr in
      buffer - the FloatBuffer to copy the contents from
      Returns:
      a new FloatPtr that contains the contents of buffer

    • toString

      public final String toString()
      Returns a string representation of this record class. The representation contains the name of the class, followed by the name and value of each of the record components.
      Specified by:
      toString in class Record
      Returns:
      a string representation of this object

    • hashCode

      public final int hashCode()
      Returns a hash code value for this object. The value is derived from the hash code of each of the record components.
      Specified by:
      hashCode in class Record
      Returns:
      a hash code value for this object

    • equals

      public final boolean equals(Object o)
      Indicates whether some other object is "equal to" this one. The objects are equal if the other object is of the same class and if all the record components are equal. All components in this record class are compared with Objects::equals(Object,Object).
      Specified by:
      equals in class Record
      Parameters:
      o - the object with which to compare
      Returns:
      true if this object is the same as the o argument; false otherwise.

    • segment

      @NotNull public @NotNull MemorySegment segment()
      Returns the value of the segment record component.
      Specified by:
      segment in interface IPointer
      Returns:
      the value of the segment record component