whiteoutlib 0.1.3

Read and write Blizzard game assets from Rust: models (MDX, M2, M3), textures (BLP, DDS, PNG, JPEG, BMP, TGA, TIFF, GIF) and archives (CASC, MPQ).
Documentation 
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

// SPDX-License-Identifier: BSD-3-Clause
// Copyright (c) 2026 Fernando Sahmkow

/// @file bc7/encode_common.inl
/// @brief BC7 encode: shared colour helpers, PCA, and endpoint fitting.

// ############################################################################
//  ENCODE
// ############################################################################

namespace {

// ============================================================================
// Colour helpers
// ============================================================================

// RGBA, pixel_at(), colour_error_rgba(), colour_error_rgb() are in bcn_common.h.

// ============================================================================
// Principal Component Analysis (simple power-iteration for 1 axis)
// ============================================================================

struct Vec4 {
    f64 x = 0.0;
    f64 y = 0.0;
    f64 z = 0.0;
    f64 w = 0.0;
};

Vec4 pca_axis_rgba(const u8* rgba, u32 count) {
    // Compute mean
    f64 mean_r = 0, mean_g = 0, mean_b = 0, mean_a = 0;
    for (u32 i = 0; i < count; ++i) {
        mean_r += rgba[i * 4 + 0];
        mean_g += rgba[i * 4 + 1];
        mean_b += rgba[i * 4 + 2];
        mean_a += rgba[i * 4 + 3];
    }
    const f64 inv_count = 1.0 / count;
    mean_r *= inv_count;
    mean_g *= inv_count;
    mean_b *= inv_count;
    mean_a *= inv_count;

    // Covariance matrix (4×4, symmetric → 10 unique)
    std::array<std::array<f64, 4>, 4> cov{};
    for (u32 i = 0; i < count; ++i) {
        f64 dr = rgba[i * 4 + 0] - mean_r;
        f64 dg = rgba[i * 4 + 1] - mean_g;
        f64 db = rgba[i * 4 + 2] - mean_b;
        f64 da = rgba[i * 4 + 3] - mean_a;
        cov[0][0] += dr * dr;
        cov[0][1] += dr * dg;
        cov[0][2] += dr * db;
        cov[0][3] += dr * da;
        cov[1][1] += dg * dg;
        cov[1][2] += dg * db;
        cov[1][3] += dg * da;
        cov[2][2] += db * db;
        cov[2][3] += db * da;
        cov[3][3] += da * da;
    }
    cov[1][0] = cov[0][1];
    cov[2][0] = cov[0][2];
    cov[2][1] = cov[1][2];
    cov[3][0] = cov[0][3];
    cov[3][1] = cov[1][3];
    cov[3][2] = cov[2][3];

    // Power iteration (8 iters is plenty for 4D)
    Vec4 v{1.0, 1.0, 1.0, 1.0};
    for (int iter = 0; iter < 8; ++iter) {
        Vec4 nv;
        nv.x = cov[0][0] * v.x + cov[0][1] * v.y + cov[0][2] * v.z + cov[0][3] * v.w;
        nv.y = cov[1][0] * v.x + cov[1][1] * v.y + cov[1][2] * v.z + cov[1][3] * v.w;
        nv.z = cov[2][0] * v.x + cov[2][1] * v.y + cov[2][2] * v.z + cov[2][3] * v.w;
        nv.w = cov[3][0] * v.x + cov[3][1] * v.y + cov[3][2] * v.z + cov[3][3] * v.w;
        f64 len = std::sqrt(nv.x * nv.x + nv.y * nv.y + nv.z * nv.z + nv.w * nv.w);
        if (len < 1e-12) {
            return {1, 0, 0, 0};
        }
        v = {nv.x / len, nv.y / len, nv.z / len, nv.w / len};
    }
    return v;
}

Vec4 pca_axis_rgb(const u8* rgba, const u8* subset_mask, u8 subset_id, u32 count) {
    // PCA on RGB channels only, for pixels belonging to `subset_id`.
    f64 mean_r = 0, mean_g = 0, mean_b = 0;
    u32 n = 0;
    for (u32 i = 0; i < count; ++i) {
        if (subset_mask[i] != subset_id)
            continue;
        mean_r += rgba[i * 4 + 0];
        mean_g += rgba[i * 4 + 1];
        mean_b += rgba[i * 4 + 2];
        ++n;
    }
    if (n == 0)
        return {1, 0, 0, 0};
    const f64 inv_count = 1.0 / n;
    mean_r *= inv_count;
    mean_g *= inv_count;
    mean_b *= inv_count;

    std::array<std::array<f64, 3>, 3> cov{};
    for (u32 i = 0; i < count; ++i) {
        if (subset_mask[i] != subset_id)
            continue;
        f64 dr = rgba[i * 4 + 0] - mean_r;
        f64 dg = rgba[i * 4 + 1] - mean_g;
        f64 db = rgba[i * 4 + 2] - mean_b;
        cov[0][0] += dr * dr;
        cov[0][1] += dr * dg;
        cov[0][2] += dr * db;
        cov[1][1] += dg * dg;
        cov[1][2] += dg * db;
        cov[2][2] += db * db;
    }
    cov[1][0] = cov[0][1];
    cov[2][0] = cov[0][2];
    cov[2][1] = cov[1][2];

    Vec4 v{1.0, 1.0, 1.0, 0.0};
    for (int iter = 0; iter < 8; ++iter) {
        Vec4 nv;
        nv.x = cov[0][0] * v.x + cov[0][1] * v.y + cov[0][2] * v.z;
        nv.y = cov[1][0] * v.x + cov[1][1] * v.y + cov[1][2] * v.z;
        nv.z = cov[2][0] * v.x + cov[2][1] * v.y + cov[2][2] * v.z;
        nv.w = 0;
        f64 len = std::sqrt(nv.x * nv.x + nv.y * nv.y + nv.z * nv.z);
        if (len < 1e-12)
            return {1, 0, 0, 0};
        v = {nv.x / len, nv.y / len, nv.z / len, 0};
    }
    return v;
}

// ============================================================================
// Endpoint fitting helpers
// ============================================================================

/// Compute min/max bounding box endpoints over all 16 pixels (RGBA).
void bbox_endpoints_rgba(const u8* rgba, RGBA& min_color, RGBA& max_color) {
    min_color = {255, 255, 255, 255};
    max_color = {0, 0, 0, 0};
    for (u32 i = 0; i < 16; ++i) {
        auto p = pixel_at(rgba, i);
        min_color.r = std::min(min_color.r, p.r);
        min_color.g = std::min(min_color.g, p.g);
        min_color.b = std::min(min_color.b, p.b);
        min_color.a = std::min(min_color.a, p.a);
        max_color.r = std::max(max_color.r, p.r);
        max_color.g = std::max(max_color.g, p.g);
        max_color.b = std::max(max_color.b, p.b);
        max_color.a = std::max(max_color.a, p.a);
    }
}

/// Project pixels onto `axis`, find min/max projection, derive endpoints.
void pca_endpoints_rgba(const u8* rgba, const Vec4& axis, RGBA& min_color, RGBA& max_color) {
    f64 proj_min = 1e30, proj_max = -1e30;
    u32 imin = 0, imax = 0;
    for (u32 i = 0; i < 16; ++i) {
        f64 p = rgba[i * 4 + 0] * axis.x + rgba[i * 4 + 1] * axis.y + rgba[i * 4 + 2] * axis.z +
                rgba[i * 4 + 3] * axis.w;
        if (p < proj_min) {
            proj_min = p;
            imin = i;
        }
        if (p > proj_max) {
            proj_max = p;
            imax = i;
        }
    }
    min_color = pixel_at(rgba, imin);
    max_color = pixel_at(rgba, imax);
}

void pca_endpoints_rgb_subset(const u8* rgba, const u8* subset_mask, u8 subset_id, const Vec4& axis,
                              RGBA& min_color, RGBA& max_color) {
    f64 proj_min = 1e30, proj_max = -1e30;
    u32 imin = 0, imax = 0;
    bool found = false;
    for (u32 i = 0; i < 16; ++i) {
        if (subset_mask[i] != subset_id)
            continue;
        f64 p = rgba[i * 4 + 0] * axis.x + rgba[i * 4 + 1] * axis.y + rgba[i * 4 + 2] * axis.z;
        if (!found || p < proj_min) {
            proj_min = p;
            imin = i;
            found = true;
        }
        if (p > proj_max) {
            proj_max = p;
            imax = i;
        }
    }
    min_color = pixel_at(rgba, imin);
    max_color = pixel_at(rgba, imax);
}

/// Clamp to [0, max_val].
inline i32 clamp_i(i32 v, i32 max_val) {
    return std::max(0, std::min(v, max_val));
}

/// Quantize a value to `prec` bits.
inline u32 quantize(i32 val, u32 prec) {
    const i32 max_val = (1 << prec) - 1;
    return static_cast<u32>(clamp_i((val * max_val + 127) / 255, max_val));
}