1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
use crate::{
    AsBufferPass, Bounds, Buffer, BufferData, BufferLayout, BufferPass,
    CameraType, GpuDevice, GpuRenderer, OrderedIndex,
};
use std::ops::Range;
//This Holds onto all the Vertexs Compressed into a byte array.
//This is Used for objects that need more advanced VBO/IBO other wise use the Instance buffers.

#[derive(Copy, Clone)]
pub struct IndexDetails {
    pub indices_start: u32,
    pub indices_end: u32,
    pub vertex_base: i32,
}

pub type ClippedIndexDetails = (IndexDetails, Option<Bounds>, CameraType);

pub struct GpuBuffer<K: BufferLayout> {
    unprocessed: Vec<Vec<OrderedIndex>>,
    pub buffers: Vec<Vec<ClippedIndexDetails>>,
    pub vertex_buffer: Buffer<K>,
    vertex_needed: usize,
    pub index_buffer: Buffer<K>,
    index_needed: usize,
    pub layer_size: usize,
    is_clipped: bool,
}

impl<'a, K: BufferLayout> AsBufferPass<'a> for GpuBuffer<K> {
    fn as_buffer_pass(&'a self) -> BufferPass<'a> {
        BufferPass {
            vertex_buffer: &self.vertex_buffer.buffer,
            index_buffer: &self.index_buffer.buffer,
        }
    }
}

impl<K: BufferLayout> GpuBuffer<K> {
    /// Used to create GpuBuffer from a (Vertex:Vec<u8>, Indices:Vec<u8>).
    pub fn create_buffer(
        gpu_device: &GpuDevice,
        buffers: &BufferData,
        layer_size: usize,
    ) -> Self {
        GpuBuffer {
            unprocessed: Vec::new(),
            buffers: Vec::new(),
            vertex_buffer: Buffer::new(
                gpu_device,
                &buffers.vertexs,
                wgpu::BufferUsages::VERTEX | wgpu::BufferUsages::COPY_DST,
                Some("Vertex Buffer"),
            ),
            vertex_needed: 0,
            index_buffer: Buffer::new(
                gpu_device,
                &buffers.indexs,
                wgpu::BufferUsages::INDEX | wgpu::BufferUsages::COPY_DST,
                Some("Index Buffer"),
            ),
            index_needed: 0,
            layer_size: layer_size.max(32),
            is_clipped: false,
        }
    }

    pub fn add_buffer_store(
        &mut self,
        renderer: &GpuRenderer,
        mut index: OrderedIndex,
        layer: usize,
    ) {
        if let Some(store) = renderer.get_buffer(index.index) {
            let offset = layer.saturating_add(1);
            // add in the missing layers this is better than keeping a hash since
            // if at anytime a process adds new data to a older layer it will already Exist.
            if self.unprocessed.len() < offset {
                for i in self.unprocessed.len()..offset {
                    //Push the layer buffer. if this is a layer we are adding data too lets
                    //give it a starting size. this cna be adjusted later for better performance
                    //versus ram usage.
                    self.unprocessed.push(if i == layer {
                        Vec::with_capacity(32)
                    } else {
                        Vec::new()
                    });
                }
            }

            self.vertex_needed += store.store.len();
            self.index_needed += store.indexs.len();

            index.index_count = store.indexs.len() as u32 / 4;

            if let Some(unprocessed) = self.unprocessed.get_mut(layer) {
                unprocessed.push(index);
            }
        }
    }

    pub fn finalize(&mut self, renderer: &mut GpuRenderer) {
        let (
            mut changed,
            mut vertex_pos,
            mut index_pos,
            mut pos,
            mut base_vertex,
        ) = (false, 0, 0, 0, 0);

        if self.vertex_needed > self.vertex_buffer.max
            || self.index_needed > self.index_buffer.max
        {
            self.resize(
                renderer.gpu_device(),
                self.vertex_needed / K::stride(),
                self.index_needed,
            );
            changed = true;
        }

        self.vertex_buffer.count = self.vertex_needed / K::stride();
        self.vertex_buffer.len = self.vertex_needed;

        //shouldnt need if renderer does all the sorting and layering first.
        for processing in &mut self.unprocessed {
            processing.sort();
        }

        if self.buffers.len() < self.unprocessed.len() {
            for _ in self.buffers.len()..self.unprocessed.len() {
                self.buffers.push(Vec::new());
            }
        }

        for buffer in &mut self.buffers {
            buffer.clear()
        }

        for (layer, processing) in self.unprocessed.iter().enumerate() {
            for buf in processing {
                let mut write_vertex = false;
                let mut write_index = false;
                let old_vertex_pos = vertex_pos as u64;
                let old_index_pos = index_pos as u64;

                if let Some(store) = renderer.get_buffer_mut(buf.index) {
                    let vertex_range =
                        vertex_pos..vertex_pos + store.store.len();
                    let index_range = index_pos..index_pos + store.indexs.len();

                    if store.store_pos != vertex_range
                        || changed
                        || store.changed
                    {
                        store.store_pos = vertex_range;
                        write_vertex = true
                    }

                    if store.index_pos != index_range
                        || changed
                        || store.changed
                    {
                        store.index_pos = index_range;
                        write_index = true
                    }

                    if write_index || write_vertex {
                        store.changed = false;
                    }

                    vertex_pos += store.store.len();
                    index_pos += store.indexs.len();
                }

                if write_vertex {
                    if let Some(store) = renderer.get_buffer(buf.index) {
                        self.vertex_buffer.write(
                            &renderer.device,
                            &store.store,
                            old_vertex_pos,
                        );
                    }
                }

                if write_index {
                    if let Some(store) = renderer.get_buffer(buf.index) {
                        self.index_buffer.write(
                            &renderer.device,
                            &store.indexs,
                            old_index_pos,
                        );
                    }
                }

                let indices_start = pos;
                let indices_end = pos + buf.index_count;
                let vertex_base = base_vertex;

                base_vertex += buf.index_max as i32 + 1;
                pos += buf.index_count;

                if let Some(buffer) = self.buffers.get_mut(layer) {
                    buffer.push((
                        IndexDetails {
                            indices_start,
                            indices_end,
                            vertex_base,
                        },
                        buf.bounds,
                        buf.camera_type,
                    ));
                }
            }
        }

        for buffer in &mut self.unprocessed {
            buffer.clear()
        }

        self.vertex_needed = 0;
        self.index_needed = 0;
    }

    //private but resizes the buffer on the GPU when needed.
    fn resize(
        &mut self,
        gpu_device: &GpuDevice,
        vertex_capacity: usize,
        index_capacity: usize,
    ) {
        let buffers = K::with_capacity(vertex_capacity, index_capacity);

        self.vertex_buffer = Buffer::new(
            gpu_device,
            &buffers.vertexs,
            wgpu::BufferUsages::VERTEX | wgpu::BufferUsages::COPY_DST,
            Some("Vertex Buffer"),
        );

        self.index_buffer = Buffer::new(
            gpu_device,
            &buffers.indexs,
            wgpu::BufferUsages::INDEX | wgpu::BufferUsages::COPY_DST,
            Some("Index Buffer"),
        )
    }

    /// Returns the index_count.
    pub fn index_count(&self) -> usize {
        self.index_buffer.count
    }

    /// Returns the index maximum size.
    pub fn index_max(&self) -> usize {
        self.index_buffer.max
    }

    /// Returns wgpu::BufferSlice of indices.
    /// bounds is used to set a specific Range if needed.
    /// If bounds is None then range is 0..index_count.
    pub fn indices(&self, bounds: Option<Range<u64>>) -> wgpu::BufferSlice {
        let range = if let Some(bounds) = bounds {
            bounds
        } else {
            0..(self.index_buffer.count) as u64
        };

        self.index_buffer.buffer_slice(range)
    }

    /// creates a new pre initlized VertexBuffer with a default size.
    /// default size is based on the initial BufferPass::vertices length.
    pub fn new(device: &GpuDevice, layer_size: usize) -> Self {
        Self::create_buffer(device, &K::default_buffer(), layer_size)
    }

    /// Set the Index based on how many Vertex's Exist
    pub fn set_index_count(&mut self, count: usize) {
        self.index_buffer.count = count;
    }

    /// Returns the Vertex elements count.
    pub fn vertex_count(&self) -> usize {
        self.vertex_buffer.count
    }

    pub fn is_empty(&self) -> bool {
        self.vertex_buffer.count == 0
    }

    /// Returns if the buffer is clipped or not to deturmine if you should use
    /// buffers or clipped_buffers.
    pub fn is_clipped(&self) -> bool {
        self.is_clipped
    }

    /// Sets the Buffer into Clipping mode.
    /// This will Produce a clipped_buffers instead of the buffers which
    /// will still be layered but a Vector of individual objects will Exist rather
    /// than a grouped object per layer. Will make it less Efficient but allows Bounds Clipping.
    pub fn set_as_clipped(&mut self) {
        self.is_clipped = true;
    }

    /// Returns vertex_buffer's max size in bytes.
    pub fn vertex_max(&self) -> usize {
        self.vertex_buffer.max
    }

    /// Returns vertex_buffer's vertex_stride.
    pub fn vertex_stride(&self) -> usize {
        K::stride()
    }

    /// Returns wgpu::BufferSlice of vertices.
    /// bounds is used to set a specific Range if needed.
    /// If bounds is None then range is 0..vertex_count.
    pub fn vertices(&self, bounds: Option<Range<u64>>) -> wgpu::BufferSlice {
        let range = if let Some(bounds) = bounds {
            bounds
        } else {
            0..self.vertex_buffer.count as u64
        };

        self.vertex_buffer.buffer_slice(range)
    }

    /// Creates a GpuBuffer based on capacity.
    /// Capacity is the amount of objects to initialize for.
    /// Capacity * 2 == the reserved space for the indices.
    pub fn with_capacity(
        gpu_device: &GpuDevice,
        capacity: usize,
        layer_size: usize,
    ) -> Self {
        Self::create_buffer(
            gpu_device,
            &K::with_capacity(capacity, capacity * 2),
            layer_size,
        )
    }
}