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

//! Lossy compression for F32 data, but lossless compression for U32 and F16 data.
// see https://github.com/AcademySoftwareFoundation/openexr/blob/master/OpenEXR/IlmImf/ImfPxr24Compressor.cpp

// This compressor is based on source code that was contributed to
// OpenEXR by Pixar Animation Studios. The compression method was
// developed by Loren Carpenter.


//  The compressor preprocesses the pixel data to reduce entropy, and then calls zlib.
//	Compression of HALF and UINT channels is lossless, but compressing
//	FLOAT channels is lossy: 32-bit floating-point numbers are converted
//	to 24 bits by rounding the significand to 15 bits.
//
//	When the compressor is invoked, the caller has already arranged
//	the pixel data so that the values for each channel appear in a
//	contiguous block of memory.  The compressor converts the pixel
//	values to unsigned integers: For UINT, this is a no-op.  HALF
//	values are simply re-interpreted as 16-bit integers.  FLOAT
//	values are converted to 24 bits, and the resulting bit patterns
//	are interpreted as integers.  The compressor then replaces each
//	value with the difference between the value and its left neighbor.
//	This turns flat fields in the image into zeroes, and ramps into
//	strings of similar values.  Next, each difference is split into
//	2, 3 or 4 bytes, and the bytes are transposed so that all the
//	most significant bytes end up in a contiguous block, followed
//	by the second most significant bytes, and so on.  The resulting
//	string of bytes is compressed with zlib.

use super::*;

use crate::error::Result;
use lebe::io::ReadPrimitive;


// scanline decompression routine, see https://github.com/openexr/openexr/blob/master/OpenEXR/IlmImf/ImfScanLineInputFile.cpp
// 1. Uncompress the data, if necessary (If the line is uncompressed, it's in XDR format, regardless of the compressor's output format.)
// 3. Convert one scan line's worth of pixel data back from the machine-independent representation
// 4. Fill the frame buffer with pixel data, respective to sampling and whatnot


#[cfg_attr(target_endian = "big", allow(unused, unreachable_code))]
pub fn compress(channels: &ChannelList, remaining_bytes: ByteVec, area: IntegerBounds) -> Result<ByteVec> {
    #[cfg(target_endian = "big")] {
        return Err(Error::unsupported(
            "PXR24 compression method not supported yet on big endian processor architecture"
        ))
    }

    if remaining_bytes.is_empty() { return Ok(Vec::new()); }

    // see https://github.com/AcademySoftwareFoundation/openexr/blob/3bd93f85bcb74c77255f28cdbb913fdbfbb39dfe/OpenEXR/IlmImf/ImfTiledOutputFile.cpp#L750-L842
    let remaining_bytes = super::convert_current_to_little_endian(remaining_bytes, channels, area);
    let mut remaining_bytes = remaining_bytes.as_slice(); // TODO less allocation

    let bytes_per_pixel: usize = channels.list.iter()
        .map(|channel| match channel.sample_type {
            SampleType::F16 => 2, SampleType::F32 => 3, SampleType::U32 => 4,
        })
        .sum();

    let mut raw = vec![0_u8; bytes_per_pixel * area.size.area()];

    {
        let mut write = raw.as_mut_slice();

        // TODO this loop should be an iterator in the `IntegerBounds` class, as it is used in all compressio methods
        for y in area.position.1..area.end().1 {
            for channel in &channels.list {
                if mod_p(y, usize_to_i32(channel.sampling.1)) != 0 { continue; }

                // this apparently can't be a closure in Rust 1.43 due to borrowing ambiguity
                let sample_count_x = channel.subsampled_resolution(area.size).0;
                macro_rules! split_off_write_slice { () => {{
                    let (slice, rest) = write.split_at_mut(sample_count_x);
                    write = rest;
                    slice
                }}; }

                let mut previous_pixel: u32 = 0;

                match channel.sample_type {
                    SampleType::F16 => {
                        let out_byte_tuples = split_off_write_slice!().iter_mut()
                            .zip(split_off_write_slice!());

                        for (out_byte_0, out_byte_1) in out_byte_tuples {
                            let pixel = u16::read_from_native_endian(&mut remaining_bytes).unwrap() as u32;
                            let [byte_1, byte_0] = (pixel.wrapping_sub(previous_pixel) as u16).to_ne_bytes();

                            *out_byte_0 = byte_0;
                            *out_byte_1 = byte_1;
                            previous_pixel = pixel;
                        }
                    },

                    SampleType::U32 => {
                        let out_byte_quadruplets = split_off_write_slice!().iter_mut()
                            .zip(split_off_write_slice!())
                            .zip(split_off_write_slice!())
                            .zip(split_off_write_slice!());

                        for (((out_byte_0, out_byte_1), out_byte_2), out_byte_3) in out_byte_quadruplets {
                            let pixel = u32::read_from_native_endian(&mut remaining_bytes).unwrap();
                            let [byte_3, byte_2, byte_1, byte_0] = pixel.wrapping_sub(previous_pixel).to_ne_bytes();

                            *out_byte_0 = byte_0;
                            *out_byte_1 = byte_1;
                            *out_byte_2 = byte_2;
                            *out_byte_3 = byte_3;
                            previous_pixel = pixel;
                        }
                    },

                    SampleType::F32 => {
                        let out_byte_triplets = split_off_write_slice!().iter_mut()
                            .zip(split_off_write_slice!())
                            .zip(split_off_write_slice!());

                        for ((out_byte_0, out_byte_1), out_byte_2) in out_byte_triplets {
                            let pixel = f32_to_f24(f32::read_from_native_endian(&mut remaining_bytes).unwrap());
                            let [byte_2, byte_1, byte_0, _] = pixel.wrapping_sub(previous_pixel).to_ne_bytes();
                            previous_pixel = pixel;

                            *out_byte_0 = byte_0;
                            *out_byte_1 = byte_1;
                            *out_byte_2 = byte_2;
                        }
                    },
                }
            }
        }

        debug_assert_eq!(write.len(), 0, "bytes left after compression");
    }

    Ok(miniz_oxide::deflate::compress_to_vec_zlib(raw.as_slice(), 4))
}

#[cfg_attr(target_endian = "big", allow(unused, unreachable_code))]
pub fn decompress(channels: &ChannelList, bytes: ByteVec, area: IntegerBounds, expected_byte_size: usize, pedantic: bool) -> Result<ByteVec> {
    #[cfg(target_endian = "big")] {
        return Err(Error::unsupported(
            "PXR24 decompression method not supported yet on big endian processor architecture"
        ))
    }

    let options = zune_inflate::DeflateOptions::default().set_limit(expected_byte_size).set_size_hint(expected_byte_size);
    let mut decoder = zune_inflate::DeflateDecoder::new_with_options(&bytes, options);
    let raw = decoder.decode_zlib()
        .map_err(|_| Error::invalid("zlib-compressed data malformed"))?; // TODO share code with zip?

    let mut read = raw.as_slice();
    let mut out = Vec::with_capacity(expected_byte_size.min(2048*4));

    for y in area.position.1 .. area.end().1 {
        for channel in &channels.list {
            if mod_p(y, usize_to_i32(channel.sampling.1)) != 0 { continue; }

            let sample_count_x = channel.subsampled_resolution(area.size).0;
            let mut read_sample_line = ||{
                if sample_count_x > read.len() { return Err(Error::invalid("not enough data")) }
                let (samples, rest) = read.split_at(sample_count_x);
                read = rest;
                Ok(samples)
            };

            let mut pixel_accumulation: u32 = 0;

            match channel.sample_type {
                SampleType::F16 => {
                    let sample_byte_pairs = read_sample_line()?.iter()
                        .zip(read_sample_line()?);

                    for (&in_byte_0, &in_byte_1) in sample_byte_pairs {
                        let difference = u16::from_ne_bytes([in_byte_1, in_byte_0]) as u32;
                        pixel_accumulation = pixel_accumulation.overflowing_add(difference).0;
                        out.extend_from_slice(&(pixel_accumulation as u16).to_ne_bytes());
                    }
                },

                SampleType::U32 => {
                    let sample_byte_quads = read_sample_line()?.iter()
                        .zip(read_sample_line()?)
                        .zip(read_sample_line()?)
                        .zip(read_sample_line()?);

                    for (((&in_byte_0, &in_byte_1), &in_byte_2), &in_byte_3) in sample_byte_quads {
                        let difference = u32::from_ne_bytes([in_byte_3, in_byte_2, in_byte_1, in_byte_0]);
                        pixel_accumulation = pixel_accumulation.overflowing_add(difference).0;
                        out.extend_from_slice(&pixel_accumulation.to_ne_bytes());
                    }
                },

                SampleType::F32 => {
                    let sample_byte_triplets = read_sample_line()?.iter()
                        .zip(read_sample_line()?).zip(read_sample_line()?);

                    for ((&in_byte_0, &in_byte_1), &in_byte_2) in sample_byte_triplets {
                        let difference = u32::from_ne_bytes([0, in_byte_2, in_byte_1, in_byte_0]);
                        pixel_accumulation = pixel_accumulation.overflowing_add(difference).0;
                        out.extend_from_slice(&pixel_accumulation.to_ne_bytes());
                    }
                }
            }
        }
    }

    if pedantic && !read.is_empty() {
        return Err(Error::invalid("too much data"));
    }

    Ok(super::convert_little_endian_to_current(out, channels, area))
}




/// Conversion from 32-bit to 24-bit floating-point numbers.
/// Reverse conversion is just a simple 8-bit left shift.
pub fn f32_to_f24(float: f32) -> u32 {
    let bits = float.to_bits();

    let sign = bits & 0x80000000;
    let exponent = bits & 0x7f800000;
    let mantissa = bits & 0x007fffff;

    let result = if exponent == 0x7f800000 {
        if mantissa != 0 {
            // F is a NAN; we preserve the sign bit and
            // the 15 leftmost bits of the significand,
            // with one exception: If the 15 leftmost
            // bits are all zero, the NAN would turn
            // into an infinity, so we have to set at
            // least one bit in the significand.

            let mantissa = mantissa >> 8;
            (exponent >> 8) | mantissa | if mantissa == 0 { 1 } else { 0 }
        }
        else { // F is an infinity.
            exponent >> 8
        }
    }
    else { // F is finite, round the significand to 15 bits.
        let result = ((exponent | mantissa) + (mantissa & 0x00000080)) >> 8;

        if result >= 0x7f8000 {
            // F was close to FLT_MAX, and the significand was
            // rounded up, resulting in an exponent overflow.
            // Avoid the overflow by truncating the significand
            // instead of rounding it.

            (exponent | mantissa) >> 8
        }
        else {
            result
        }
    };

    return (sign >> 8) | result;
}