Buffer accessors;

api/lovr/graphics/Buffer/clear.lua

@@ -0,0 +1,29 @@
+return {
+  summary = 'Clear the data in the Buffer.',
+  description = [[
+    Clears some or all of the data in the Buffer to zero.  This is supported for both temporary and
+    permanent Buffers.  Permanent Buffers can also be cleared in a transfer pass using `Pass:clear`.
+  ]],
+  arguments = {
+    {
+      name = 'index',
+      type = 'number',
+      default = '1',
+      description = 'The index of the first item to clear.'
+    },
+    {
+      name = 'count',
+      type = 'number',
+      default = 'nil',
+      description = 'The number of items to clear.  If `nil`, clears as many items as possible.'
+    }
+  },
+  returns = {},
+  notes = [[
+    Clearing a permanent buffer requires the byte offset and byte count of the cleared range to be a
+    multiple of 4.  This will usually be true for most field types.
+  ]],
+  related = {
+    'Pass:clear'
+  }
+}

api/lovr/graphics/Buffer/setData.lua

@@ -0,0 +1,97 @@
+return {
+  summary = 'Change the data in the Buffer.',
+  description = [[
+    Changes data in the Buffer using a table or a Blob.  This is supported for both temporary and
+    permanent buffers.  All passes submitted to the GPU will use the new data.
+
+    It is also possible to change the data in permanent buffers inside of a transfer pass using
+    `Pass:copy`.  Using a transfer pass allows the copy to happen after other passes in the frame.
+  ]],
+  arguments = {
+    data = {
+      type = 'table',
+      description = 'A flat or nested table of items to copy to the Buffer (see notes for format).'
+    },
+    sourceIndex = {
+      type = 'number',
+      default = '1',
+      description = 'The index in the table to copy from.'
+    },
+    destinationIndex = {
+      type = 'number',
+      default = '1',
+      description = 'The index of the first value in the Buffer to update.'
+    },
+    count = {
+      type = 'number',
+      default = 'nil',
+      description = [[
+        The number of values to update.  `nil` will copy as many items as possible, based on the
+        lengths of the source and destination.
+      ]]
+    },
+    blob = {
+      type = 'Blob',
+      description = 'The Blob to copy data from.'
+    },
+    sourceOffset = {
+      type = 'number',
+      default = '0',
+      description = 'A byte offset into the Blob to copy from.'
+    },
+    destinationOffset = {
+      type = 'number',
+      default = '0',
+      description = 'A byte offset in the Buffer to copy to.'
+    },
+    size = {
+      type = 'number',
+      default = 'nil',
+      description = 'The number of bytes to copy.  If nil, copies as many bytes as possible.'
+    }
+  },
+  returns = {},
+  variants = {
+    {
+      arguments = { 'data', 'sourceIndex', 'destinationIndex', 'count' },
+      returns = {}
+    },
+    {
+      arguments = { 'blob', 'sourceOffset', 'destinationOffset', 'size' },
+      returns = {}
+    }
+  },
+  notes = [[
+    When using a table, the table can contain a nested table for each value in the Buffer, or it can
+    be a flat list of field component values.  It is not possible to mix both nested tables and flat
+    values.
+
+    For each item updated, components of each field in the item (according to the Buffer's format)
+    are read from either the nested subtable or the table itself.  A single number can be used to
+    update a field with a scalar type.  Multiple numbers or a `lovr.math` vector can be used to
+    update a field with a vector or mat4 type.  Multiple numbers can be used to update mat2 and mat3
+    fields.  When updating normalized field types, components read from the table will be clamped to
+    the normalized range ([0,1] or [-1,1]).  In the Buffer, each field is written at its byte offset
+    according to the Buffer's format, and subsequent items are separated by the byte stride of the
+    Buffer.  Any missing components for an updated field will be set to zero.
+  ]],
+  example = [[
+    function lovr.load()
+      buffer = lovr.graphics.newBuffer(3, 'floats')
+      buffer:setData({ { 1.0 }, { 2.0 }, { 3.0 } })
+      buffer:setData({ 1.0, 2.0, 3.0 })
+
+      buffer = lovr.graphics.newBuffer(5, { 'vec3', 'vec3', 'vec2' })
+      buffer:setData({ vec3(1, 2, 3), vec3(4, 5, 6), vec2(7, 8) })
+      buffer:setData({ { 1, 2, 3, 4, 5, 6, 7, 8 } })
+      buffer:setData({ 1, 2, 3, 4, 5, 6, 7, 8 })
+      buffer:setData({
+        { x1, y1, z1, nx1, ny1, nz1, u1, v1 },
+        { x2, y2, z2, vec3(nx, ny, nz) }
+      }, 1, 3, 2)
+    end
+  ]],
+  related = {
+    'Pass:copy'
+  }
+}