diff --git a/src/external/stb_image.h b/src/external/stb_image.h index 4d8d01332..ae2ada6a6 100644 --- a/src/external/stb_image.h +++ b/src/external/stb_image.h @@ -1,4 +1,4 @@ -/* stb_image - v2.14 - public domain image loader - http://nothings.org/stb_image.h +/* stb_image - v2.15 - public domain image loader - http://nothings.org/stb_image.h no warranty implied; use at your own risk Do this: @@ -21,7 +21,7 @@ avoid problematic images and only need the trivial interface JPEG baseline & progressive (12 bpc/arithmetic not supported, same as stock IJG lib) - PNG 1/2/4/8-bit-per-channel (16 bpc not supported) + PNG 1/2/4/8/16-bit-per-channel TGA (not sure what subset, if a subset) BMP non-1bpp, non-RLE @@ -42,115 +42,19 @@ Full documentation under "DOCUMENTATION" below. - Revision 2.00 release notes: - - - Progressive JPEG is now supported. - - - PPM and PGM binary formats are now supported, thanks to Ken Miller. - - - x86 platforms now make use of SSE2 SIMD instructions for - JPEG decoding, and ARM platforms can use NEON SIMD if requested. - This work was done by Fabian "ryg" Giesen. SSE2 is used by - default, but NEON must be enabled explicitly; see docs. - - With other JPEG optimizations included in this version, we see - 2x speedup on a JPEG on an x86 machine, and a 1.5x speedup - on a JPEG on an ARM machine, relative to previous versions of this - library. The same results will not obtain for all JPGs and for all - x86/ARM machines. (Note that progressive JPEGs are significantly - slower to decode than regular JPEGs.) This doesn't mean that this - is the fastest JPEG decoder in the land; rather, it brings it - closer to parity with standard libraries. If you want the fastest - decode, look elsewhere. (See "Philosophy" section of docs below.) - - See final bullet items below for more info on SIMD. - - - Added STBI_MALLOC, STBI_REALLOC, and STBI_FREE macros for replacing - the memory allocator. Unlike other STBI libraries, these macros don't - support a context parameter, so if you need to pass a context in to - the allocator, you'll have to store it in a global or a thread-local - variable. - - - Split existing STBI_NO_HDR flag into two flags, STBI_NO_HDR and - STBI_NO_LINEAR. - STBI_NO_HDR: suppress implementation of .hdr reader format - STBI_NO_LINEAR: suppress high-dynamic-range light-linear float API - - - You can suppress implementation of any of the decoders to reduce - your code footprint by #defining one or more of the following - symbols before creating the implementation. - - STBI_NO_JPEG - STBI_NO_PNG - STBI_NO_BMP - STBI_NO_PSD - STBI_NO_TGA - STBI_NO_GIF - STBI_NO_HDR - STBI_NO_PIC - STBI_NO_PNM (.ppm and .pgm) - - - You can request *only* certain decoders and suppress all other ones - (this will be more forward-compatible, as addition of new decoders - doesn't require you to disable them explicitly): - - STBI_ONLY_JPEG - STBI_ONLY_PNG - STBI_ONLY_BMP - STBI_ONLY_PSD - STBI_ONLY_TGA - STBI_ONLY_GIF - STBI_ONLY_HDR - STBI_ONLY_PIC - STBI_ONLY_PNM (.ppm and .pgm) - - Note that you can define multiples of these, and you will get all - of them ("only x" and "only y" is interpreted to mean "only x&y"). - - - If you use STBI_NO_PNG (or _ONLY_ without PNG), and you still - want the zlib decoder to be available, #define STBI_SUPPORT_ZLIB - - - Compilation of all SIMD code can be suppressed with - #define STBI_NO_SIMD - It should not be necessary to disable SIMD unless you have issues - compiling (e.g. using an x86 compiler which doesn't support SSE - intrinsics or that doesn't support the method used to detect - SSE2 support at run-time), and even those can be reported as - bugs so I can refine the built-in compile-time checking to be - smarter. - - - The old STBI_SIMD system which allowed installing a user-defined - IDCT etc. has been removed. If you need this, don't upgrade. My - assumption is that almost nobody was doing this, and those who - were will find the built-in SIMD more satisfactory anyway. - - - RGB values computed for JPEG images are slightly different from - previous versions of stb_image. (This is due to using less - integer precision in SIMD.) The C code has been adjusted so - that the same RGB values will be computed regardless of whether - SIMD support is available, so your app should always produce - consistent results. But these results are slightly different from - previous versions. (Specifically, about 3% of available YCbCr values - will compute different RGB results from pre-1.49 versions by +-1; - most of the deviating values are one smaller in the G channel.) - - - If you must produce consistent results with previous versions of - stb_image, #define STBI_JPEG_OLD and you will get the same results - you used to; however, you will not get the SIMD speedups for - the YCbCr-to-RGB conversion step (although you should still see - significant JPEG speedup from the other changes). - - Please note that STBI_JPEG_OLD is a temporary feature; it will be - removed in future versions of the library. It is only intended for - near-term back-compatibility use. - - - Latest revision history: +LICENSE + + See end of file for license information. + +RECENT REVISION HISTORY: + + 2.15 (2017-03-18) fix png-1,2,4; all Imagenet JPGs; no runtime SSE detection on GCC + 2.14 (2017-03-03) remove deprecated STBI_JPEG_OLD; fixes for Imagenet JPGs 2.13 (2016-12-04) experimental 16-bit API, only for PNG so far; fixes 2.12 (2016-04-02) fix typo in 2.11 PSD fix that caused crashes 2.11 (2016-04-02) 16-bit PNGS; enable SSE2 in non-gcc x64 RGB-format JPEG; remove white matting in PSD; - allocate large structures on the stack; + allocate large structures on the stack; correct channel count for PNG & BMP 2.10 (2016-01-22) avoid warning introduced in 2.09 2.09 (2016-01-16) 16-bit TGA; comments in PNM files; STBI_REALLOC_SIZED @@ -174,32 +78,26 @@ Ken Miller (pgm, ppm) Richard Mitton (16-bit PSD) github:urraka (animated gif) Junggon Kim (PNM comments) Daniel Gibson (16-bit TGA) - socks-the-fox (16-bit TGA) + socks-the-fox (16-bit PNG) + Jeremy Sawicki (handle all ImageNet JPGs) Optimizations & bugfixes Fabian "ryg" Giesen Arseny Kapoulkine Bug & warning fixes Marc LeBlanc David Woo Guillaume George Martins Mozeiko - Christpher Lloyd Martin Golini Jerry Jansson Joseph Thomson - Dave Moore Roy Eltham Hayaki Saito Phil Jordan - Won Chun Luke Graham Johan Duparc Nathan Reed - the Horde3D community Thomas Ruf Ronny Chevalier Nick Verigakis - Janez Zemva John Bartholomew Michal Cichon github:svdijk - Jonathan Blow Ken Hamada Tero Hanninen Baldur Karlsson - Laurent Gomila Cort Stratton Sergio Gonzalez github:romigrou - Aruelien Pocheville Thibault Reuille Cass Everitt Matthew Gregan - Ryamond Barbiero Paul Du Bois Engin Manap github:snagar - Michaelangel007@github Oriol Ferrer Mesia Dale Weiler github:Zelex - Philipp Wiesemann Josh Tobin github:rlyeh github:grim210@github - Blazej Dariusz Roszkowski github:sammyhw - - -LICENSE - -This software is dual-licensed to the public domain and under the following -license: you are granted a perpetual, irrevocable license to copy, modify, -publish, and distribute this file as you see fit. + Christpher Lloyd Jerry Jansson Joseph Thomson Phil Jordan + Dave Moore Roy Eltham Hayaki Saito Nathan Reed + Won Chun Luke Graham Johan Duparc Nick Verigakis + the Horde3D community Thomas Ruf Ronny Chevalier Baldur Karlsson + Janez Zemva John Bartholomew Michal Cichon github:rlyeh + Jonathan Blow Ken Hamada Tero Hanninen github:romigrou + Laurent Gomila Cort Stratton Sergio Gonzalez github:svdijk + Aruelien Pocheville Thibault Reuille Cass Everitt github:snagar + Ryamond Barbiero Paul Du Bois Engin Manap github:Zelex + Michaelangel007@github Philipp Wiesemann Dale Weiler github:grim210 + Oriol Ferrer Mesia Josh Tobin Matthew Gregan github:sammyhw + Blazej Dariusz Roszkowski Gregory Mullen github:phprus */ @@ -274,13 +172,13 @@ publish, and distribute this file as you see fit. // and for best performance I may provide less-easy-to-use APIs that give higher // performance, in addition to the easy to use ones. Nevertheless, it's important // to keep in mind that from the standpoint of you, a client of this library, -// all you care about is #1 and #3, and stb libraries do not emphasize #3 above all. +// all you care about is #1 and #3, and stb libraries DO NOT emphasize #3 above all. // // Some secondary priorities arise directly from the first two, some of which // make more explicit reasons why performance can't be emphasized. // // - Portable ("ease of use") -// - Small footprint ("easy to maintain") +// - Small source code footprint ("easy to maintain") // - No dependencies ("ease of use") // // =========================================================================== @@ -312,13 +210,6 @@ publish, and distribute this file as you see fit. // (at least this is true for iOS and Android). Therefore, the NEON support is // toggled by a build flag: define STBI_NEON to get NEON loops. // -// The output of the JPEG decoder is slightly different from versions where -// SIMD support was introduced (that is, for versions before 1.49). The -// difference is only +-1 in the 8-bit RGB channels, and only on a small -// fraction of pixels. You can force the pre-1.49 behavior by defining -// STBI_JPEG_OLD, but this will disable some of the SIMD decoding path -// and hence cost some performance. -// // If for some reason you do not want to use any of SIMD code, or if // you have issues compiling it, you can disable it entirely by // defining STBI_NO_SIMD. @@ -374,20 +265,47 @@ publish, and distribute this file as you see fit. // says there's premultiplied data (currently only happens in iPhone images, // and only if iPhone convert-to-rgb processing is on). // +// =========================================================================== +// +// ADDITIONAL CONFIGURATION +// +// - You can suppress implementation of any of the decoders to reduce +// your code footprint by #defining one or more of the following +// symbols before creating the implementation. +// +// STBI_NO_JPEG +// STBI_NO_PNG +// STBI_NO_BMP +// STBI_NO_PSD +// STBI_NO_TGA +// STBI_NO_GIF +// STBI_NO_HDR +// STBI_NO_PIC +// STBI_NO_PNM (.ppm and .pgm) +// +// - You can request *only* certain decoders and suppress all other ones +// (this will be more forward-compatible, as addition of new decoders +// doesn't require you to disable them explicitly): +// +// STBI_ONLY_JPEG +// STBI_ONLY_PNG +// STBI_ONLY_BMP +// STBI_ONLY_PSD +// STBI_ONLY_TGA +// STBI_ONLY_GIF +// STBI_ONLY_HDR +// STBI_ONLY_PIC +// STBI_ONLY_PNM (.ppm and .pgm) +// +// - If you use STBI_NO_PNG (or _ONLY_ without PNG), and you still +// want the zlib decoder to be available, #define STBI_SUPPORT_ZLIB +// #ifndef STBI_NO_STDIO #include #endif // STBI_NO_STDIO -#define STBI_NO_HDR // RaySan: not required by raylib -//#define STBI_NO_SIMD // RaySan: issues when compiling with GCC 4.7.2 - -// NOTE: Added to work with raylib on Android -#if defined(PLATFORM_ANDROID) - #include "utils.h" // RaySan: Android fopen function map -#endif - #define STBI_VERSION 1 enum @@ -666,12 +584,14 @@ typedef unsigned char validate_uint32[sizeof(stbi__uint32)==4 ? 1 : -1]; #define STBI__X86_TARGET #endif -#if defined(__GNUC__) && (defined(STBI__X86_TARGET) || defined(STBI__X64_TARGET)) && !defined(__SSE2__) && !defined(STBI_NO_SIMD) -// NOTE: not clear do we actually need this for the 64-bit path? +#if defined(__GNUC__) && defined(STBI__X86_TARGET) && !defined(__SSE2__) && !defined(STBI_NO_SIMD) // gcc doesn't support sse2 intrinsics unless you compile with -msse2, -// (but compiling with -msse2 allows the compiler to use SSE2 everywhere; -// this is just broken and gcc are jerks for not fixing it properly -// http://www.virtualdub.org/blog/pivot/entry.php?id=363 ) +// which in turn means it gets to use SSE2 everywhere. This is unfortunate, +// but previous attempts to provide the SSE2 functions with runtime +// detection caused numerous issues. The way architecture extensions are +// exposed in GCC/Clang is, sadly, not really suited for one-file libs. +// New behavior: if compiled with -msse2, we use SSE2 without any +// detection; if not, we don't use it at all. #define STBI_NO_SIMD #endif @@ -729,14 +649,10 @@ static int stbi__sse2_available() static int stbi__sse2_available() { -#if defined(__GNUC__) && (__GNUC__ * 100 + __GNUC_MINOR__) >= 408 // GCC 4.8 or later - // GCC 4.8+ has a nice way to do this - return __builtin_cpu_supports("sse2"); -#else - // portable way to do this, preferably without using GCC inline ASM? - // just bail for now. - return 0; -#endif + // If we're even attempting to compile this on GCC/Clang, that means + // -msse2 is on, which means the compiler is allowed to use SSE2 + // instructions at will, and so are we. + return 1; } #endif #endif @@ -1738,7 +1654,7 @@ typedef struct stbi__context *s; stbi__huffman huff_dc[4]; stbi__huffman huff_ac[4]; - stbi_uc dequant[4][64]; + stbi__uint16 dequant[4][64]; stbi__int16 fast_ac[4][1 << FAST_BITS]; // sizes for components, interleaved MCUs @@ -1774,6 +1690,8 @@ typedef struct int succ_high; int succ_low; int eob_run; + int jfif; + int app14_color_transform; // Adobe APP14 tag int rgb; int scan_n, order[4]; @@ -1844,7 +1762,7 @@ static void stbi__build_fast_ac(stbi__int16 *fast_ac, stbi__huffman *h) // magnitude code followed by receive_extend code int k = ((i << len) & ((1 << FAST_BITS) - 1)) >> (FAST_BITS - magbits); int m = 1 << (magbits - 1); - if (k < m) k += (-1 << magbits) + 1; + if (k < m) k += (~0U << magbits) + 1; // if the result is small enough, we can fit it in fast_ac table if (k >= -128 && k <= 127) fast_ac[i] = (stbi__int16) ((k << 8) + (run << 4) + (len + magbits)); @@ -1859,6 +1777,7 @@ static void stbi__grow_buffer_unsafe(stbi__jpeg *j) int b = j->nomore ? 0 : stbi__get8(j->s); if (b == 0xff) { int c = stbi__get8(j->s); + while (c == 0xff) c = stbi__get8(j->s); // consume fill bytes if (c != 0) { j->marker = (unsigned char) c; j->nomore = 1; @@ -1983,7 +1902,7 @@ static stbi_uc stbi__jpeg_dezigzag[64+15] = }; // decode one 64-entry block-- -static int stbi__jpeg_decode_block(stbi__jpeg *j, short data[64], stbi__huffman *hdc, stbi__huffman *hac, stbi__int16 *fac, int b, stbi_uc *dequant) +static int stbi__jpeg_decode_block(stbi__jpeg *j, short data[64], stbi__huffman *hdc, stbi__huffman *hac, stbi__int16 *fac, int b, stbi__uint16 *dequant) { int diff,dc,k; int t; @@ -2692,7 +2611,7 @@ static stbi_uc stbi__get_marker(stbi__jpeg *j) x = stbi__get8(j->s); if (x != 0xff) return STBI__MARKER_none; while (x == 0xff) - x = stbi__get8(j->s); + x = stbi__get8(j->s); // consume repeated 0xff fill bytes return x; } @@ -2707,7 +2626,7 @@ static void stbi__jpeg_reset(stbi__jpeg *j) j->code_bits = 0; j->code_buffer = 0; j->nomore = 0; - j->img_comp[0].dc_pred = j->img_comp[1].dc_pred = j->img_comp[2].dc_pred = 0; + j->img_comp[0].dc_pred = j->img_comp[1].dc_pred = j->img_comp[2].dc_pred = j->img_comp[3].dc_pred = 0; j->marker = STBI__MARKER_none; j->todo = j->restart_interval ? j->restart_interval : 0x7fffffff; j->eob_run = 0; @@ -2839,7 +2758,7 @@ static int stbi__parse_entropy_coded_data(stbi__jpeg *z) } } -static void stbi__jpeg_dequantize(short *data, stbi_uc *dequant) +static void stbi__jpeg_dequantize(short *data, stbi__uint16 *dequant) { int i; for (i=0; i < 64; ++i) @@ -2881,13 +2800,14 @@ static int stbi__process_marker(stbi__jpeg *z, int m) L = stbi__get16be(z->s)-2; while (L > 0) { int q = stbi__get8(z->s); - int p = q >> 4; + int p = q >> 4, sixteen = (p != 0); int t = q & 15,i; - if (p != 0) return stbi__err("bad DQT type","Corrupt JPEG"); + if (p != 0 && p != 1) return stbi__err("bad DQT type","Corrupt JPEG"); if (t > 3) return stbi__err("bad DQT table","Corrupt JPEG"); + for (i=0; i < 64; ++i) - z->dequant[t][stbi__jpeg_dezigzag[i]] = stbi__get8(z->s); - L -= 65; + z->dequant[t][stbi__jpeg_dezigzag[i]] = sixteen ? stbi__get16be(z->s) : stbi__get8(z->s); + L -= (sixteen ? 129 : 65); } return L==0; @@ -2920,12 +2840,50 @@ static int stbi__process_marker(stbi__jpeg *z, int m) } return L==0; } + // check for comment block or APP blocks if ((m >= 0xE0 && m <= 0xEF) || m == 0xFE) { - stbi__skip(z->s, stbi__get16be(z->s)-2); + L = stbi__get16be(z->s); + if (L < 2) { + if (m == 0xFE) + return stbi__err("bad COM len","Corrupt JPEG"); + else + return stbi__err("bad APP len","Corrupt JPEG"); + } + L -= 2; + + if (m == 0xE0 && L >= 5) { // JFIF APP0 segment + static const unsigned char tag[5] = {'J','F','I','F','\0'}; + int ok = 1; + int i; + for (i=0; i < 5; ++i) + if (stbi__get8(z->s) != tag[i]) + ok = 0; + L -= 5; + if (ok) + z->jfif = 1; + } else if (m == 0xEE && L >= 12) { // Adobe APP14 segment + static const unsigned char tag[6] = {'A','d','o','b','e','\0'}; + int ok = 1; + int i; + for (i=0; i < 6; ++i) + if (stbi__get8(z->s) != tag[i]) + ok = 0; + L -= 6; + if (ok) { + stbi__get8(z->s); // version + stbi__get16be(z->s); // flags0 + stbi__get16be(z->s); // flags1 + z->app14_color_transform = stbi__get8(z->s); // color transform + L -= 6; + } + } + + stbi__skip(z->s, L); return 1; } - return 0; + + return stbi__err("unknown marker","Corrupt JPEG"); } // after we see SOS @@ -2999,7 +2957,7 @@ static int stbi__process_frame_header(stbi__jpeg *z, int scan) s->img_y = stbi__get16be(s); if (s->img_y == 0) return stbi__err("no header height", "JPEG format not supported: delayed height"); // Legal, but we don't handle it--but neither does IJG s->img_x = stbi__get16be(s); if (s->img_x == 0) return stbi__err("0 width","Corrupt JPEG"); // JPEG requires c = stbi__get8(s); - if (c != 3 && c != 1) return stbi__err("bad component count","Corrupt JPEG"); // JFIF requires + if (c != 3 && c != 1 && c != 4) return stbi__err("bad component count","Corrupt JPEG"); s->img_n = c; for (i=0; i < c; ++i) { z->img_comp[i].data = NULL; @@ -3012,13 +2970,8 @@ static int stbi__process_frame_header(stbi__jpeg *z, int scan) for (i=0; i < s->img_n; ++i) { static unsigned char rgb[3] = { 'R', 'G', 'B' }; z->img_comp[i].id = stbi__get8(s); - if (z->img_comp[i].id != i+1) // JFIF requires - if (z->img_comp[i].id != i) { // some version of jpegtran outputs non-JFIF-compliant files! - // somethings output this (see http://fileformats.archiveteam.org/wiki/JPEG#Color_format) - if (z->img_comp[i].id != rgb[i]) - return stbi__err("bad component ID","Corrupt JPEG"); - ++z->rgb; - } + if (s->img_n == 3 && z->img_comp[i].id == rgb[i]) + ++z->rgb; q = stbi__get8(s); z->img_comp[i].h = (q >> 4); if (!z->img_comp[i].h || z->img_comp[i].h > 4) return stbi__err("bad H","Corrupt JPEG"); z->img_comp[i].v = q & 15; if (!z->img_comp[i].v || z->img_comp[i].v > 4) return stbi__err("bad V","Corrupt JPEG"); @@ -3090,6 +3043,8 @@ static int stbi__process_frame_header(stbi__jpeg *z, int scan) static int stbi__decode_jpeg_header(stbi__jpeg *z, int scan) { int m; + z->jfif = 0; + z->app14_color_transform = -1; // valid values are 0,1,2 z->marker = STBI__MARKER_none; // initialize cached marker to empty m = stbi__get_marker(z); if (!stbi__SOI(m)) return stbi__err("no SOI","Corrupt JPEG"); @@ -3131,12 +3086,15 @@ static int stbi__decode_jpeg_image(stbi__jpeg *j) if (x == 255) { j->marker = stbi__get8(j->s); break; - } else if (x != 0) { - return stbi__err("junk before marker", "Corrupt JPEG"); } } // if we reach eof without hitting a marker, stbi__get_marker() below will fail and we'll eventually return 0 } + } else if (stbi__DNL(m)) { + int Ld = stbi__get16be(j->s); + stbi__uint32 NL = stbi__get16be(j->s); + if (Ld != 4) stbi__err("bad DNL len", "Corrupt JPEG"); + if (NL != j->s->img_y) stbi__err("bad DNL height", "Corrupt JPEG"); } else { if (!stbi__process_marker(j, m)) return 0; } @@ -3355,38 +3313,9 @@ static stbi_uc *stbi__resample_row_generic(stbi_uc *out, stbi_uc *in_near, stbi_ return out; } -#ifdef STBI_JPEG_OLD -// this is the same YCbCr-to-RGB calculation that stb_image has used -// historically before the algorithm changes in 1.49 -#define float2fixed(x) ((int) ((x) * 65536 + 0.5)) -static void stbi__YCbCr_to_RGB_row(stbi_uc *out, const stbi_uc *y, const stbi_uc *pcb, const stbi_uc *pcr, int count, int step) -{ - int i; - for (i=0; i < count; ++i) { - int y_fixed = (y[i] << 16) + 32768; // rounding - int r,g,b; - int cr = pcr[i] - 128; - int cb = pcb[i] - 128; - r = y_fixed + cr*float2fixed(1.40200f); - g = y_fixed - cr*float2fixed(0.71414f) - cb*float2fixed(0.34414f); - b = y_fixed + cb*float2fixed(1.77200f); - r >>= 16; - g >>= 16; - b >>= 16; - if ((unsigned) r > 255) { if (r < 0) r = 0; else r = 255; } - if ((unsigned) g > 255) { if (g < 0) g = 0; else g = 255; } - if ((unsigned) b > 255) { if (b < 0) b = 0; else b = 255; } - out[0] = (stbi_uc)r; - out[1] = (stbi_uc)g; - out[2] = (stbi_uc)b; - out[3] = 255; - out += step; - } -} -#else // this is a reduced-precision calculation of YCbCr-to-RGB introduced // to make sure the code produces the same results in both SIMD and scalar -#define float2fixed(x) (((int) ((x) * 4096.0f + 0.5f)) << 8) +#define stbi__float2fixed(x) (((int) ((x) * 4096.0f + 0.5f)) << 8) static void stbi__YCbCr_to_RGB_row(stbi_uc *out, const stbi_uc *y, const stbi_uc *pcb, const stbi_uc *pcr, int count, int step) { int i; @@ -3395,9 +3324,9 @@ static void stbi__YCbCr_to_RGB_row(stbi_uc *out, const stbi_uc *y, const stbi_uc int r,g,b; int cr = pcr[i] - 128; int cb = pcb[i] - 128; - r = y_fixed + cr* float2fixed(1.40200f); - g = y_fixed + (cr*-float2fixed(0.71414f)) + ((cb*-float2fixed(0.34414f)) & 0xffff0000); - b = y_fixed + cb* float2fixed(1.77200f); + r = y_fixed + cr* stbi__float2fixed(1.40200f); + g = y_fixed + (cr*-stbi__float2fixed(0.71414f)) + ((cb*-stbi__float2fixed(0.34414f)) & 0xffff0000); + b = y_fixed + cb* stbi__float2fixed(1.77200f); r >>= 20; g >>= 20; b >>= 20; @@ -3411,7 +3340,6 @@ static void stbi__YCbCr_to_RGB_row(stbi_uc *out, const stbi_uc *y, const stbi_uc out += step; } } -#endif #if defined(STBI_SSE2) || defined(STBI_NEON) static void stbi__YCbCr_to_RGB_simd(stbi_uc *out, stbi_uc const *y, stbi_uc const *pcb, stbi_uc const *pcr, int count, int step) @@ -3530,9 +3458,9 @@ static void stbi__YCbCr_to_RGB_simd(stbi_uc *out, stbi_uc const *y, stbi_uc cons int r,g,b; int cr = pcr[i] - 128; int cb = pcb[i] - 128; - r = y_fixed + cr* float2fixed(1.40200f); - g = y_fixed + cr*-float2fixed(0.71414f) + ((cb*-float2fixed(0.34414f)) & 0xffff0000); - b = y_fixed + cb* float2fixed(1.77200f); + r = y_fixed + cr* stbi__float2fixed(1.40200f); + g = y_fixed + cr*-stbi__float2fixed(0.71414f) + ((cb*-stbi__float2fixed(0.34414f)) & 0xffff0000); + b = y_fixed + cb* stbi__float2fixed(1.77200f); r >>= 20; g >>= 20; b >>= 20; @@ -3558,18 +3486,14 @@ static void stbi__setup_jpeg(stbi__jpeg *j) #ifdef STBI_SSE2 if (stbi__sse2_available()) { j->idct_block_kernel = stbi__idct_simd; - #ifndef STBI_JPEG_OLD j->YCbCr_to_RGB_kernel = stbi__YCbCr_to_RGB_simd; - #endif j->resample_row_hv_2_kernel = stbi__resample_row_hv_2_simd; } #endif #ifdef STBI_NEON j->idct_block_kernel = stbi__idct_simd; - #ifndef STBI_JPEG_OLD j->YCbCr_to_RGB_kernel = stbi__YCbCr_to_RGB_simd; - #endif j->resample_row_hv_2_kernel = stbi__resample_row_hv_2_simd; #endif } @@ -3590,9 +3514,16 @@ typedef struct int ypos; // which pre-expansion row we're on } stbi__resample; +// fast 0..255 * 0..255 => 0..255 rounded multiplication +static stbi_uc stbi__blinn_8x8(stbi_uc x, stbi_uc y) +{ + unsigned int t = x*y + 128; + return (stbi_uc) ((t + (t >>8)) >> 8); +} + static stbi_uc *load_jpeg_image(stbi__jpeg *z, int *out_x, int *out_y, int *comp, int req_comp) { - int n, decode_n; + int n, decode_n, is_rgb; z->s->img_n = 0; // make stbi__cleanup_jpeg safe // validate req_comp @@ -3602,9 +3533,11 @@ static stbi_uc *load_jpeg_image(stbi__jpeg *z, int *out_x, int *out_y, int *comp if (!stbi__decode_jpeg_image(z)) { stbi__cleanup_jpeg(z); return NULL; } // determine actual number of components to generate - n = req_comp ? req_comp : z->s->img_n; + n = req_comp ? req_comp : z->s->img_n >= 3 ? 3 : 1; - if (z->s->img_n == 3 && n < 3) + is_rgb = z->s->img_n == 3 && (z->rgb == 3 || (z->app14_color_transform == 0 && !z->jfif)); + + if (z->s->img_n == 3 && n < 3 && !is_rgb) decode_n = 1; else decode_n = z->s->img_n; @@ -3664,7 +3597,7 @@ static stbi_uc *load_jpeg_image(stbi__jpeg *z, int *out_x, int *out_y, int *comp if (n >= 3) { stbi_uc *y = coutput[0]; if (z->s->img_n == 3) { - if (z->rgb == 3) { + if (is_rgb) { for (i=0; i < z->s->img_x; ++i) { out[0] = y[i]; out[1] = coutput[1][i]; @@ -3675,6 +3608,28 @@ static stbi_uc *load_jpeg_image(stbi__jpeg *z, int *out_x, int *out_y, int *comp } else { z->YCbCr_to_RGB_kernel(out, y, coutput[1], coutput[2], z->s->img_x, n); } + } else if (z->s->img_n == 4) { + if (z->app14_color_transform == 0) { // CMYK + for (i=0; i < z->s->img_x; ++i) { + stbi_uc k = coutput[3][i]; + out[0] = stbi__blinn_8x8(coutput[0][i], k); + out[1] = stbi__blinn_8x8(coutput[1][i], k); + out[2] = stbi__blinn_8x8(coutput[2][i], k); + out[3] = 255; + out += n; + } + } else if (z->app14_color_transform == 2) { // YCCK + z->YCbCr_to_RGB_kernel(out, y, coutput[1], coutput[2], z->s->img_x, n); + for (i=0; i < z->s->img_x; ++i) { + stbi_uc k = coutput[3][i]; + out[0] = stbi__blinn_8x8(255 - out[0], k); + out[1] = stbi__blinn_8x8(255 - out[1], k); + out[2] = stbi__blinn_8x8(255 - out[2], k); + out += n; + } + } else { // YCbCr + alpha? Ignore the fourth channel for now + z->YCbCr_to_RGB_kernel(out, y, coutput[1], coutput[2], z->s->img_x, n); + } } else for (i=0; i < z->s->img_x; ++i) { out[0] = out[1] = out[2] = y[i]; @@ -3682,17 +3637,45 @@ static stbi_uc *load_jpeg_image(stbi__jpeg *z, int *out_x, int *out_y, int *comp out += n; } } else { - stbi_uc *y = coutput[0]; - if (n == 1) - for (i=0; i < z->s->img_x; ++i) out[i] = y[i]; - else - for (i=0; i < z->s->img_x; ++i) *out++ = y[i], *out++ = 255; + if (is_rgb) { + if (n == 1) + for (i=0; i < z->s->img_x; ++i) + *out++ = stbi__compute_y(coutput[0][i], coutput[1][i], coutput[2][i]); + else { + for (i=0; i < z->s->img_x; ++i, out += 2) { + out[0] = stbi__compute_y(coutput[0][i], coutput[1][i], coutput[2][i]); + out[1] = 255; + } + } + } else if (z->s->img_n == 4 && z->app14_color_transform == 0) { + for (i=0; i < z->s->img_x; ++i) { + stbi_uc k = coutput[3][i]; + stbi_uc r = stbi__blinn_8x8(coutput[0][i], k); + stbi_uc g = stbi__blinn_8x8(coutput[1][i], k); + stbi_uc b = stbi__blinn_8x8(coutput[2][i], k); + out[0] = stbi__compute_y(r, g, b); + out[1] = 255; + out += n; + } + } else if (z->s->img_n == 4 && z->app14_color_transform == 2) { + for (i=0; i < z->s->img_x; ++i) { + out[0] = stbi__blinn_8x8(255 - coutput[0][i], coutput[3][i]); + out[1] = 255; + out += n; + } + } else { + stbi_uc *y = coutput[0]; + if (n == 1) + for (i=0; i < z->s->img_x; ++i) out[i] = y[i]; + else + for (i=0; i < z->s->img_x; ++i) *out++ = y[i], *out++ = 255; + } } } stbi__cleanup_jpeg(z); *out_x = z->s->img_x; *out_y = z->s->img_y; - if (comp) *comp = z->s->img_n; // report original components, not output + if (comp) *comp = z->s->img_n >= 3 ? 3 : 1; // report original components, not output return output; } } @@ -3701,6 +3684,7 @@ static void *stbi__jpeg_load(stbi__context *s, int *x, int *y, int *comp, int re { unsigned char* result; stbi__jpeg* j = (stbi__jpeg*) stbi__malloc(sizeof(stbi__jpeg)); + STBI_NOTUSED(ri); j->s = s; stbi__setup_jpeg(j); result = load_jpeg_image(j, x,y,comp,req_comp); @@ -3711,11 +3695,12 @@ static void *stbi__jpeg_load(stbi__context *s, int *x, int *y, int *comp, int re static int stbi__jpeg_test(stbi__context *s) { int r; - stbi__jpeg j; - j.s = s; - stbi__setup_jpeg(&j); - r = stbi__decode_jpeg_header(&j, STBI__SCAN_type); + stbi__jpeg* j = (stbi__jpeg*)stbi__malloc(sizeof(stbi__jpeg)); + j->s = s; + stbi__setup_jpeg(j); + r = stbi__decode_jpeg_header(j, STBI__SCAN_type); stbi__rewind(s); + STBI_FREE(j); return r; } @@ -3727,7 +3712,7 @@ static int stbi__jpeg_info_raw(stbi__jpeg *j, int *x, int *y, int *comp) } if (x) *x = j->s->img_x; if (y) *y = j->s->img_y; - if (comp) *comp = j->s->img_n; + if (comp) *comp = j->s->img_n >= 3 ? 3 : 1; return 1; } @@ -3784,7 +3769,7 @@ stbi_inline static int stbi__bit_reverse(int v, int bits) return stbi__bitreverse16(v) >> (16-bits); } -static int stbi__zbuild_huffman(stbi__zhuffman *z, stbi_uc *sizelist, int num) +static int stbi__zbuild_huffman(stbi__zhuffman *z, const stbi_uc *sizelist, int num) { int i,k=0; int code, next_code[16], sizes[17]; @@ -4074,9 +4059,24 @@ static int stbi__parse_zlib_header(stbi__zbuf *a) return 1; } -// @TODO: should statically initialize these for optimal thread safety -static stbi_uc stbi__zdefault_length[288], stbi__zdefault_distance[32]; -static void stbi__init_zdefaults(void) +static const stbi_uc stbi__zdefault_length[288] = +{ + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, + 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, + 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, + 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, + 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, + 7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7, 7,7,7,7,7,7,7,7,8,8,8,8,8,8,8,8 +}; +static const stbi_uc stbi__zdefault_distance[32] = +{ + 5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5 +}; +/* +Init algorithm: { int i; // use <= to match clearly with spec for (i=0; i <= 143; ++i) stbi__zdefault_length[i] = 8; @@ -4086,6 +4086,7 @@ static void stbi__init_zdefaults(void) for (i=0; i <= 31; ++i) stbi__zdefault_distance[i] = 5; } +*/ static int stbi__parse_zlib(stbi__zbuf *a, int parse_header) { @@ -4104,7 +4105,6 @@ static int stbi__parse_zlib(stbi__zbuf *a, int parse_header) } else { if (type == 1) { // use fixed code lengths - if (!stbi__zdefault_distance[31]) stbi__init_zdefaults(); if (!stbi__zbuild_huffman(&a->z_length , stbi__zdefault_length , 288)) return 0; if (!stbi__zbuild_huffman(&a->z_distance, stbi__zdefault_distance, 32)) return 0; } else { @@ -4305,7 +4305,7 @@ static int stbi__create_png_image_raw(stbi__png *a, stbi_uc *raw, stbi__uint32 r for (j=0; j < y; ++j) { stbi_uc *cur = a->out + stride*j; - stbi_uc *prior = cur - stride; + stbi_uc *prior; int filter = *raw++; if (filter > 4) @@ -4317,6 +4317,7 @@ static int stbi__create_png_image_raw(stbi__png *a, stbi_uc *raw, stbi__uint32 r filter_bytes = 1; width = img_width_bytes; } + prior = cur - stride; // bugfix: need to compute this after 'cur +=' computation above // if first row, use special filter that doesn't sample previous row if (j == 0) filter = first_row_filter[filter]; @@ -4709,7 +4710,7 @@ static int stbi__parse_png_file(stbi__png *z, int scan, int req_comp) s->img_y = stbi__get32be(s); if (s->img_y > (1 << 24)) return stbi__err("too large","Very large image (corrupt?)"); z->depth = stbi__get8(s); if (z->depth != 1 && z->depth != 2 && z->depth != 4 && z->depth != 8 && z->depth != 16) return stbi__err("1/2/4/8/16-bit only","PNG not supported: 1/2/4/8/16-bit only"); color = stbi__get8(s); if (color > 6) return stbi__err("bad ctype","Corrupt PNG"); - if (color == 3 && z->depth == 16) return stbi__err("bad ctype","Corrupt PNG"); + if (color == 3 && z->depth == 16) return stbi__err("bad ctype","Corrupt PNG"); if (color == 3) pal_img_n = 3; else if (color & 1) return stbi__err("bad ctype","Corrupt PNG"); comp = stbi__get8(s); if (comp) return stbi__err("bad comp method","Corrupt PNG"); filter= stbi__get8(s); if (filter) return stbi__err("bad filter method","Corrupt PNG"); @@ -4992,7 +4993,7 @@ static void *stbi__bmp_parse_header(stbi__context *s, stbi__bmp_data *info) info->offset = stbi__get32le(s); info->hsz = hsz = stbi__get32le(s); info->mr = info->mg = info->mb = info->ma = 0; - + if (hsz != 12 && hsz != 40 && hsz != 56 && hsz != 108 && hsz != 124) return stbi__errpuc("unknown BMP", "BMP type not supported: unknown"); if (hsz == 12) { s->img_x = stbi__get16le(s); @@ -5077,7 +5078,7 @@ static void *stbi__bmp_load(stbi__context *s, int *x, int *y, int *comp, int req stbi__bmp_data info; STBI_NOTUSED(ri); - info.all_a = 255; + info.all_a = 255; if (stbi__bmp_parse_header(s, &info) == NULL) return NULL; // error code already set @@ -5196,7 +5197,7 @@ static void *stbi__bmp_load(stbi__context *s, int *x, int *y, int *comp, int req stbi__skip(s, pad); } } - + // if alpha channel is all 0s, replace with all 255s if (target == 4 && all_a == 0) for (i=4*s->img_x*s->img_y-1; i >= 0; i -= 4) @@ -5979,9 +5980,11 @@ static stbi_uc *stbi__pic_load_core(stbi__context *s,int width,int height,int *c static void *stbi__pic_load(stbi__context *s,int *px,int *py,int *comp,int req_comp, stbi__result_info *ri) { stbi_uc *result; - int i, x,y; + int i, x,y, internal_comp; STBI_NOTUSED(ri); + if (!comp) comp = &internal_comp; + for (i=0; i<92; ++i) stbi__get8(s); @@ -6600,6 +6603,11 @@ static int stbi__hdr_info(stbi__context *s, int *x, int *y, int *comp) char buffer[STBI__HDR_BUFLEN]; char *token; int valid = 0; + int dummy; + + if (!x) x = &dummy; + if (!y) y = &dummy; + if (!comp) comp = &dummy; if (stbi__hdr_test(s) == 0) { stbi__rewind( s ); @@ -6641,14 +6649,14 @@ static int stbi__bmp_info(stbi__context *s, int *x, int *y, int *comp) void *p; stbi__bmp_data info; - info.all_a = 255; + info.all_a = 255; p = stbi__bmp_parse_header(s, &info); stbi__rewind( s ); if (p == NULL) return 0; - *x = s->img_x; - *y = s->img_y; - *comp = info.ma ? 4 : 3; + if (x) *x = s->img_x; + if (y) *y = s->img_y; + if (comp) *comp = info.ma ? 4 : 3; return 1; } #endif @@ -6656,7 +6664,10 @@ static int stbi__bmp_info(stbi__context *s, int *x, int *y, int *comp) #ifndef STBI_NO_PSD static int stbi__psd_info(stbi__context *s, int *x, int *y, int *comp) { - int channelCount; + int channelCount, dummy; + if (!x) x = &dummy; + if (!y) y = &dummy; + if (!comp) comp = &dummy; if (stbi__get32be(s) != 0x38425053) { stbi__rewind( s ); return 0; @@ -6689,9 +6700,13 @@ static int stbi__psd_info(stbi__context *s, int *x, int *y, int *comp) #ifndef STBI_NO_PIC static int stbi__pic_info(stbi__context *s, int *x, int *y, int *comp) { - int act_comp=0,num_packets=0,chained; + int act_comp=0,num_packets=0,chained,dummy; stbi__pic_packet packets[10]; + if (!x) x = &dummy; + if (!y) y = &dummy; + if (!comp) comp = &dummy; + if (!stbi__pic_is4(s,"\x53\x80\xF6\x34")) { stbi__rewind(s); return 0; @@ -6777,7 +6792,7 @@ static void *stbi__pnm_load(stbi__context *s, int *x, int *y, int *comp, int req *x = s->img_x; *y = s->img_y; - *comp = s->img_n; + if (comp) *comp = s->img_n; if (!stbi__mad3sizes_valid(s->img_n, s->img_x, s->img_y, 0)) return stbi__errpuc("too large", "PNM too large"); @@ -6831,16 +6846,20 @@ static int stbi__pnm_getinteger(stbi__context *s, char *c) static int stbi__pnm_info(stbi__context *s, int *x, int *y, int *comp) { - int maxv; + int maxv, dummy; char c, p, t; - stbi__rewind( s ); + if (!x) x = &dummy; + if (!y) y = &dummy; + if (!comp) comp = &dummy; + + stbi__rewind(s); // Get identifier p = (char) stbi__get8(s); t = (char) stbi__get8(s); if (p != 'P' || (t != '5' && t != '6')) { - stbi__rewind( s ); + stbi__rewind(s); return 0; } @@ -6947,6 +6966,11 @@ STBIDEF int stbi_info_from_callbacks(stbi_io_callbacks const *c, void *user, int /* revision history: + 2.15 (2017-03-18) fix png-1,2,4 bug; now all Imagenet JPGs decode; + warning fixes; disable run-time SSE detection on gcc; + uniform handling of optional "return" values; + thread-safe initialization of zlib tables + 2.14 (2017-03-03) remove deprecated STBI_JPEG_OLD; fixes for Imagenet JPGs 2.13 (2016-11-29) add 16-bit API, only supported for PNG right now 2.12 (2016-04-02) fix typo in 2.11 PSD fix that caused crashes 2.11 (2016-04-02) allocate large structures on the stack @@ -7108,3 +7132,46 @@ STBIDEF int stbi_info_from_callbacks(stbi_io_callbacks const *c, void *user, int 0.50 (2006-11-19) first released version */ + + +/* +------------------------------------------------------------------------------ +This software is available under 2 licenses -- choose whichever you prefer. +------------------------------------------------------------------------------ +ALTERNATIVE A - MIT License +Copyright (c) 2017 Sean Barrett +Permission is hereby granted, free of charge, to any person obtaining a copy of +this software and associated documentation files (the "Software"), to deal in +the Software without restriction, including without limitation the rights to +use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies +of the Software, and to permit persons to whom the Software is furnished to do +so, subject to the following conditions: +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. +------------------------------------------------------------------------------ +ALTERNATIVE B - Public Domain (www.unlicense.org) +This is free and unencumbered software released into the public domain. +Anyone is free to copy, modify, publish, use, compile, sell, or distribute this +software, either in source code form or as a compiled binary, for any purpose, +commercial or non-commercial, and by any means. +In jurisdictions that recognize copyright laws, the author or authors of this +software dedicate any and all copyright interest in the software to the public +domain. We make this dedication for the benefit of the public at large and to +the detriment of our heirs and successors. We intend this dedication to be an +overt act of relinquishment in perpetuity of all present and future rights to +this software under copyright law. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN +ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION +WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. +------------------------------------------------------------------------------ +*/ diff --git a/src/external/stb_image_resize.h b/src/external/stb_image_resize.h index 5214ad6e3..b507e0498 100644 --- a/src/external/stb_image_resize.h +++ b/src/external/stb_image_resize.h @@ -1,4 +1,4 @@ -/* stb_image_resize - v0.92 - public domain image resizing +/* stb_image_resize - v0.94 - public domain image resizing by Jorge L Rodriguez (@VinoBS) - 2014 http://github.com/nothings/stb @@ -107,8 +107,8 @@ industry, it is still uncommon in the videogame/real-time world. If you linearly filter non-premultiplied alpha, strange effects - occur. (For example, the average of 1% opaque bright green - and 99% opaque black produces 50% transparent dark green when + occur. (For example, the 50/50 average of 99% transparent bright green + and 1% transparent black produces 50% transparent dark green when non-premultiplied, whereas premultiplied it produces 50% transparent near-black. The former introduces green energy that doesn't exist in the source image.) @@ -152,20 +152,20 @@ (For example, graphics hardware does not apply sRGB conversion to the alpha channel.) - ADDITIONAL CONTRIBUTORS + CONTRIBUTORS + Jorge L Rodriguez: Implementation Sean Barrett: API design, optimizations Aras Pranckevicius: bugfix REVISIONS + 0.94 (2017-03-18) fixed warnings + 0.93 (2017-03-03) fixed bug with certain combinations of heights 0.92 (2017-01-02) fix integer overflow on large (>2GB) images 0.91 (2016-04-02) fix warnings; fix handling of subpixel regions 0.90 (2014-09-17) first released version LICENSE - - This software is dual-licensed to the public domain and under the following - license: you are granted a perpetual, irrevocable license to copy, modify, - publish, and distribute this file as you see fit. + See end of file for license information. TODO Don't decode all of the image data when only processing a partial tile @@ -533,10 +533,11 @@ typedef struct int horizontal_num_contributors; int vertical_num_contributors; - int ring_buffer_length_bytes; // The length of an individual entry in the ring buffer. The total number of ring buffers is stbir__get_filter_pixel_width(filter) + int ring_buffer_length_bytes; // The length of an individual entry in the ring buffer. The total number of ring buffers is stbir__get_filter_pixel_width(filter) + int ring_buffer_num_entries; // Total number of entries in the ring buffer. int ring_buffer_first_scanline; int ring_buffer_last_scanline; - int ring_buffer_begin_index; + int ring_buffer_begin_index; // first_scanline is at this index in the ring buffer float* ring_buffer; float* encode_buffer; // A temporary buffer to store floats so we don't lose precision while we do multiply-adds. @@ -551,16 +552,17 @@ typedef struct int encode_buffer_size; } stbir__info; + +static const float stbir__max_uint8_as_float = 255.0f; +static const float stbir__max_uint16_as_float = 65535.0f; +static const double stbir__max_uint32_as_float = 4294967295.0; + + static stbir__inline int stbir__min(int a, int b) { return a < b ? a : b; } -static stbir__inline int stbir__max(int a, int b) -{ - return a > b ? a : b; -} - static stbir__inline float stbir__saturate(float x) { if (x < 0) @@ -1027,7 +1029,7 @@ static void stbir__calculate_sample_range_downsample(int n, float in_pixels_radi *out_last_pixel = (int)(floor(out_pixel_influence_upperbound - 0.5)); } -static void stbir__calculate_coefficients_upsample(stbir__info* stbir_info, stbir_filter filter, float scale, int in_first_pixel, int in_last_pixel, float in_center_of_out, stbir__contributors* contributor, float* coefficient_group) +static void stbir__calculate_coefficients_upsample(stbir_filter filter, float scale, int in_first_pixel, int in_last_pixel, float in_center_of_out, stbir__contributors* contributor, float* coefficient_group) { int i; float total_filter = 0; @@ -1077,7 +1079,7 @@ static void stbir__calculate_coefficients_upsample(stbir__info* stbir_info, stbi } } -static void stbir__calculate_coefficients_downsample(stbir__info* stbir_info, stbir_filter filter, float scale_ratio, int out_first_pixel, int out_last_pixel, float out_center_of_in, stbir__contributors* contributor, float* coefficient_group) +static void stbir__calculate_coefficients_downsample(stbir_filter filter, float scale_ratio, int out_first_pixel, int out_last_pixel, float out_center_of_in, stbir__contributors* contributor, float* coefficient_group) { int i; @@ -1107,7 +1109,7 @@ static void stbir__calculate_coefficients_downsample(stbir__info* stbir_info, st } } -static void stbir__normalize_downsample_coefficients(stbir__info* stbir_info, stbir__contributors* contributors, float* coefficients, stbir_filter filter, float scale_ratio, float shift, int input_size, int output_size) +static void stbir__normalize_downsample_coefficients(stbir__contributors* contributors, float* coefficients, stbir_filter filter, float scale_ratio, int input_size, int output_size) { int num_contributors = stbir__get_contributors(scale_ratio, filter, input_size, output_size); int num_coefficients = stbir__get_coefficient_width(filter, scale_ratio); @@ -1184,7 +1186,7 @@ static void stbir__normalize_downsample_coefficients(stbir__info* stbir_info, st // Each scan line uses the same kernel values so we should calculate the kernel // values once and then we can use them for every scan line. -static void stbir__calculate_filters(stbir__info* stbir_info, stbir__contributors* contributors, float* coefficients, stbir_filter filter, float scale_ratio, float shift, int input_size, int output_size) +static void stbir__calculate_filters(stbir__contributors* contributors, float* coefficients, stbir_filter filter, float scale_ratio, float shift, int input_size, int output_size) { int n; int total_contributors = stbir__get_contributors(scale_ratio, filter, input_size, output_size); @@ -1201,7 +1203,7 @@ static void stbir__calculate_filters(stbir__info* stbir_info, stbir__contributor stbir__calculate_sample_range_upsample(n, out_pixels_radius, scale_ratio, shift, &in_first_pixel, &in_last_pixel, &in_center_of_out); - stbir__calculate_coefficients_upsample(stbir_info, filter, scale_ratio, in_first_pixel, in_last_pixel, in_center_of_out, stbir__get_contributor(contributors, n), stbir__get_coefficient(coefficients, filter, scale_ratio, n, 0)); + stbir__calculate_coefficients_upsample(filter, scale_ratio, in_first_pixel, in_last_pixel, in_center_of_out, stbir__get_contributor(contributors, n), stbir__get_coefficient(coefficients, filter, scale_ratio, n, 0)); } } else @@ -1217,10 +1219,10 @@ static void stbir__calculate_filters(stbir__info* stbir_info, stbir__contributor stbir__calculate_sample_range_downsample(n_adjusted, in_pixels_radius, scale_ratio, shift, &out_first_pixel, &out_last_pixel, &out_center_of_in); - stbir__calculate_coefficients_downsample(stbir_info, filter, scale_ratio, out_first_pixel, out_last_pixel, out_center_of_in, stbir__get_contributor(contributors, n), stbir__get_coefficient(coefficients, filter, scale_ratio, n, 0)); + stbir__calculate_coefficients_downsample(filter, scale_ratio, out_first_pixel, out_last_pixel, out_center_of_in, stbir__get_contributor(contributors, n), stbir__get_coefficient(coefficients, filter, scale_ratio, n, 0)); } - stbir__normalize_downsample_coefficients(stbir_info, contributors, coefficients, filter, scale_ratio, shift, input_size, output_size); + stbir__normalize_downsample_coefficients(contributors, coefficients, filter, scale_ratio, input_size, output_size); } } @@ -1270,7 +1272,7 @@ static void stbir__decode_scanline(stbir__info* stbir_info, int n) int decode_pixel_index = x * channels; int input_pixel_index = stbir__edge_wrap(edge_horizontal, x, input_w) * channels; for (c = 0; c < channels; c++) - decode_buffer[decode_pixel_index + c] = ((float)((const unsigned char*)input_data)[input_pixel_index + c]) / 255; + decode_buffer[decode_pixel_index + c] = ((float)((const unsigned char*)input_data)[input_pixel_index + c]) / stbir__max_uint8_as_float; } break; @@ -1283,7 +1285,7 @@ static void stbir__decode_scanline(stbir__info* stbir_info, int n) decode_buffer[decode_pixel_index + c] = stbir__srgb_uchar_to_linear_float[((const unsigned char*)input_data)[input_pixel_index + c]]; if (!(stbir_info->flags&STBIR_FLAG_ALPHA_USES_COLORSPACE)) - decode_buffer[decode_pixel_index + alpha_channel] = ((float)((const unsigned char*)input_data)[input_pixel_index + alpha_channel]) / 255; + decode_buffer[decode_pixel_index + alpha_channel] = ((float)((const unsigned char*)input_data)[input_pixel_index + alpha_channel]) / stbir__max_uint8_as_float; } break; @@ -1293,7 +1295,7 @@ static void stbir__decode_scanline(stbir__info* stbir_info, int n) int decode_pixel_index = x * channels; int input_pixel_index = stbir__edge_wrap(edge_horizontal, x, input_w) * channels; for (c = 0; c < channels; c++) - decode_buffer[decode_pixel_index + c] = ((float)((const unsigned short*)input_data)[input_pixel_index + c]) / 65535; + decode_buffer[decode_pixel_index + c] = ((float)((const unsigned short*)input_data)[input_pixel_index + c]) / stbir__max_uint16_as_float; } break; @@ -1303,10 +1305,10 @@ static void stbir__decode_scanline(stbir__info* stbir_info, int n) int decode_pixel_index = x * channels; int input_pixel_index = stbir__edge_wrap(edge_horizontal, x, input_w) * channels; for (c = 0; c < channels; c++) - decode_buffer[decode_pixel_index + c] = stbir__srgb_to_linear(((float)((const unsigned short*)input_data)[input_pixel_index + c]) / 65535); + decode_buffer[decode_pixel_index + c] = stbir__srgb_to_linear(((float)((const unsigned short*)input_data)[input_pixel_index + c]) / stbir__max_uint16_as_float); if (!(stbir_info->flags&STBIR_FLAG_ALPHA_USES_COLORSPACE)) - decode_buffer[decode_pixel_index + alpha_channel] = ((float)((const unsigned short*)input_data)[input_pixel_index + alpha_channel]) / 65535; + decode_buffer[decode_pixel_index + alpha_channel] = ((float)((const unsigned short*)input_data)[input_pixel_index + alpha_channel]) / stbir__max_uint16_as_float; } break; @@ -1316,7 +1318,7 @@ static void stbir__decode_scanline(stbir__info* stbir_info, int n) int decode_pixel_index = x * channels; int input_pixel_index = stbir__edge_wrap(edge_horizontal, x, input_w) * channels; for (c = 0; c < channels; c++) - decode_buffer[decode_pixel_index + c] = (float)(((double)((const unsigned int*)input_data)[input_pixel_index + c]) / 4294967295); + decode_buffer[decode_pixel_index + c] = (float)(((double)((const unsigned int*)input_data)[input_pixel_index + c]) / stbir__max_uint32_as_float); } break; @@ -1326,10 +1328,10 @@ static void stbir__decode_scanline(stbir__info* stbir_info, int n) int decode_pixel_index = x * channels; int input_pixel_index = stbir__edge_wrap(edge_horizontal, x, input_w) * channels; for (c = 0; c < channels; c++) - decode_buffer[decode_pixel_index + c] = stbir__srgb_to_linear((float)(((double)((const unsigned int*)input_data)[input_pixel_index + c]) / 4294967295)); + decode_buffer[decode_pixel_index + c] = stbir__srgb_to_linear((float)(((double)((const unsigned int*)input_data)[input_pixel_index + c]) / stbir__max_uint32_as_float)); if (!(stbir_info->flags&STBIR_FLAG_ALPHA_USES_COLORSPACE)) - decode_buffer[decode_pixel_index + alpha_channel] = (float)(((double)((const unsigned int*)input_data)[input_pixel_index + alpha_channel]) / 4294967295); + decode_buffer[decode_pixel_index + alpha_channel] = (float)(((double)((const unsigned int*)input_data)[input_pixel_index + alpha_channel]) / stbir__max_uint32_as_float); } break; @@ -1411,6 +1413,8 @@ static float* stbir__add_empty_ring_buffer_entry(stbir__info* stbir_info, int n) int ring_buffer_index; float* ring_buffer; + stbir_info->ring_buffer_last_scanline = n; + if (stbir_info->ring_buffer_begin_index < 0) { ring_buffer_index = stbir_info->ring_buffer_begin_index = 0; @@ -1418,24 +1422,21 @@ static float* stbir__add_empty_ring_buffer_entry(stbir__info* stbir_info, int n) } else { - ring_buffer_index = (stbir_info->ring_buffer_begin_index + (stbir_info->ring_buffer_last_scanline - stbir_info->ring_buffer_first_scanline) + 1) % stbir_info->vertical_filter_pixel_width; + ring_buffer_index = (stbir_info->ring_buffer_begin_index + (stbir_info->ring_buffer_last_scanline - stbir_info->ring_buffer_first_scanline)) % stbir_info->ring_buffer_num_entries; STBIR_ASSERT(ring_buffer_index != stbir_info->ring_buffer_begin_index); } ring_buffer = stbir__get_ring_buffer_entry(stbir_info->ring_buffer, ring_buffer_index, stbir_info->ring_buffer_length_bytes / sizeof(float)); memset(ring_buffer, 0, stbir_info->ring_buffer_length_bytes); - stbir_info->ring_buffer_last_scanline = n; - return ring_buffer; } -static void stbir__resample_horizontal_upsample(stbir__info* stbir_info, int n, float* output_buffer) +static void stbir__resample_horizontal_upsample(stbir__info* stbir_info, float* output_buffer) { int x, k; int output_w = stbir_info->output_w; - int kernel_pixel_width = stbir_info->horizontal_filter_pixel_width; int channels = stbir_info->channels; float* decode_buffer = stbir__get_decode_buffer(stbir_info); stbir__contributors* horizontal_contributors = stbir_info->horizontal_contributors; @@ -1515,12 +1516,10 @@ static void stbir__resample_horizontal_upsample(stbir__info* stbir_info, int n, } } -static void stbir__resample_horizontal_downsample(stbir__info* stbir_info, int n, float* output_buffer) +static void stbir__resample_horizontal_downsample(stbir__info* stbir_info, float* output_buffer) { int x, k; int input_w = stbir_info->input_w; - int output_w = stbir_info->output_w; - int kernel_pixel_width = stbir_info->horizontal_filter_pixel_width; int channels = stbir_info->channels; float* decode_buffer = stbir__get_decode_buffer(stbir_info); stbir__contributors* horizontal_contributors = stbir_info->horizontal_contributors; @@ -1654,9 +1653,9 @@ static void stbir__decode_and_resample_upsample(stbir__info* stbir_info, int n) // Now resample it into the ring buffer. if (stbir__use_width_upsampling(stbir_info)) - stbir__resample_horizontal_upsample(stbir_info, n, stbir__add_empty_ring_buffer_entry(stbir_info, n)); + stbir__resample_horizontal_upsample(stbir_info, stbir__add_empty_ring_buffer_entry(stbir_info, n)); else - stbir__resample_horizontal_downsample(stbir_info, n, stbir__add_empty_ring_buffer_entry(stbir_info, n)); + stbir__resample_horizontal_downsample(stbir_info, stbir__add_empty_ring_buffer_entry(stbir_info, n)); // Now it's sitting in the ring buffer ready to be used as source for the vertical sampling. } @@ -1670,17 +1669,17 @@ static void stbir__decode_and_resample_downsample(stbir__info* stbir_info, int n // Now resample it into the horizontal buffer. if (stbir__use_width_upsampling(stbir_info)) - stbir__resample_horizontal_upsample(stbir_info, n, stbir_info->horizontal_buffer); + stbir__resample_horizontal_upsample(stbir_info, stbir_info->horizontal_buffer); else - stbir__resample_horizontal_downsample(stbir_info, n, stbir_info->horizontal_buffer); + stbir__resample_horizontal_downsample(stbir_info, stbir_info->horizontal_buffer); // Now it's sitting in the horizontal buffer ready to be distributed into the ring buffers. } // Get the specified scan line from the ring buffer. -static float* stbir__get_ring_buffer_scanline(int get_scanline, float* ring_buffer, int begin_index, int first_scanline, int ring_buffer_size, int ring_buffer_length) +static float* stbir__get_ring_buffer_scanline(int get_scanline, float* ring_buffer, int begin_index, int first_scanline, int ring_buffer_num_entries, int ring_buffer_length) { - int ring_buffer_index = (begin_index + (get_scanline - first_scanline)) % ring_buffer_size; + int ring_buffer_index = (begin_index + (get_scanline - first_scanline)) % ring_buffer_num_entries; return stbir__get_ring_buffer_entry(ring_buffer, ring_buffer_index, ring_buffer_length); } @@ -1715,19 +1714,23 @@ static void stbir__encode_scanline(stbir__info* stbir_info, int num_pixels, void // build a table of all channels that need colorspace correction, so // we don't perform colorspace correction on channels that don't need it. - for (x=0, num_nonalpha=0; x < channels; ++x) + for (x = 0, num_nonalpha = 0; x < channels; ++x) + { if (x != alpha_channel || (stbir_info->flags & STBIR_FLAG_ALPHA_USES_COLORSPACE)) - nonalpha[num_nonalpha++] = x; + { + nonalpha[num_nonalpha++] = (stbir_uint16)x; + } + } #define STBIR__ROUND_INT(f) ((int) ((f)+0.5)) #define STBIR__ROUND_UINT(f) ((stbir_uint32) ((f)+0.5)) #ifdef STBIR__SATURATE_INT - #define STBIR__ENCODE_LINEAR8(f) stbir__saturate8 (STBIR__ROUND_INT((f) * 255 )) - #define STBIR__ENCODE_LINEAR16(f) stbir__saturate16(STBIR__ROUND_INT((f) * 65535)) + #define STBIR__ENCODE_LINEAR8(f) stbir__saturate8 (STBIR__ROUND_INT((f) * stbir__max_uint8_as_float )) + #define STBIR__ENCODE_LINEAR16(f) stbir__saturate16(STBIR__ROUND_INT((f) * stbir__max_uint16_as_float)) #else - #define STBIR__ENCODE_LINEAR8(f) (unsigned char ) STBIR__ROUND_INT(stbir__saturate(f) * 255 ) - #define STBIR__ENCODE_LINEAR16(f) (unsigned short) STBIR__ROUND_INT(stbir__saturate(f) * 65535) + #define STBIR__ENCODE_LINEAR8(f) (unsigned char ) STBIR__ROUND_INT(stbir__saturate(f) * stbir__max_uint8_as_float ) + #define STBIR__ENCODE_LINEAR16(f) (unsigned short) STBIR__ROUND_INT(stbir__saturate(f) * stbir__max_uint16_as_float) #endif switch (decode) @@ -1782,7 +1785,7 @@ static void stbir__encode_scanline(stbir__info* stbir_info, int num_pixels, void for (n = 0; n < num_nonalpha; n++) { int index = pixel_index + nonalpha[n]; - ((unsigned short*)output_buffer)[index] = (unsigned short)STBIR__ROUND_INT(stbir__linear_to_srgb(stbir__saturate(encode_buffer[index])) * 65535); + ((unsigned short*)output_buffer)[index] = (unsigned short)STBIR__ROUND_INT(stbir__linear_to_srgb(stbir__saturate(encode_buffer[index])) * stbir__max_uint16_as_float); } if (!(stbir_info->flags&STBIR_FLAG_ALPHA_USES_COLORSPACE)) @@ -1799,7 +1802,7 @@ static void stbir__encode_scanline(stbir__info* stbir_info, int num_pixels, void for (n = 0; n < channels; n++) { int index = pixel_index + n; - ((unsigned int*)output_buffer)[index] = (unsigned int)STBIR__ROUND_UINT(((double)stbir__saturate(encode_buffer[index])) * 4294967295); + ((unsigned int*)output_buffer)[index] = (unsigned int)STBIR__ROUND_UINT(((double)stbir__saturate(encode_buffer[index])) * stbir__max_uint32_as_float); } } break; @@ -1812,11 +1815,11 @@ static void stbir__encode_scanline(stbir__info* stbir_info, int num_pixels, void for (n = 0; n < num_nonalpha; n++) { int index = pixel_index + nonalpha[n]; - ((unsigned int*)output_buffer)[index] = (unsigned int)STBIR__ROUND_UINT(((double)stbir__linear_to_srgb(stbir__saturate(encode_buffer[index]))) * 4294967295); + ((unsigned int*)output_buffer)[index] = (unsigned int)STBIR__ROUND_UINT(((double)stbir__linear_to_srgb(stbir__saturate(encode_buffer[index]))) * stbir__max_uint32_as_float); } if (!(stbir_info->flags&STBIR_FLAG_ALPHA_USES_COLORSPACE)) - ((unsigned int*)output_buffer)[pixel_index + alpha_channel] = (unsigned int)STBIR__ROUND_INT(((double)stbir__saturate(encode_buffer[pixel_index + alpha_channel])) * 4294967295); + ((unsigned int*)output_buffer)[pixel_index + alpha_channel] = (unsigned int)STBIR__ROUND_INT(((double)stbir__saturate(encode_buffer[pixel_index + alpha_channel])) * stbir__max_uint32_as_float); } break; @@ -1855,7 +1858,7 @@ static void stbir__encode_scanline(stbir__info* stbir_info, int num_pixels, void } } -static void stbir__resample_vertical_upsample(stbir__info* stbir_info, int n, int in_first_scanline, int in_last_scanline, float in_center_of_out) +static void stbir__resample_vertical_upsample(stbir__info* stbir_info, int n) { int x, k; int output_w = stbir_info->output_w; @@ -1865,7 +1868,7 @@ static void stbir__resample_vertical_upsample(stbir__info* stbir_info, int n, in int alpha_channel = stbir_info->alpha_channel; int type = stbir_info->type; int colorspace = stbir_info->colorspace; - int kernel_pixel_width = stbir_info->vertical_filter_pixel_width; + int ring_buffer_entries = stbir_info->ring_buffer_num_entries; void* output_data = stbir_info->output_data; float* encode_buffer = stbir_info->encode_buffer; int decode = STBIR__DECODE(type, colorspace); @@ -1876,7 +1879,6 @@ static void stbir__resample_vertical_upsample(stbir__info* stbir_info, int n, in float* ring_buffer = stbir_info->ring_buffer; int ring_buffer_begin_index = stbir_info->ring_buffer_begin_index; int ring_buffer_first_scanline = stbir_info->ring_buffer_first_scanline; - int ring_buffer_last_scanline = stbir_info->ring_buffer_last_scanline; int ring_buffer_length = stbir_info->ring_buffer_length_bytes/sizeof(float); int n0,n1, output_row_start; @@ -1900,7 +1902,7 @@ static void stbir__resample_vertical_upsample(stbir__info* stbir_info, int n, in for (k = n0; k <= n1; k++) { int coefficient_index = coefficient_counter++; - float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, kernel_pixel_width, ring_buffer_length); + float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, ring_buffer_entries, ring_buffer_length); float coefficient = vertical_coefficients[coefficient_group + coefficient_index]; for (x = 0; x < output_w; ++x) { @@ -1913,7 +1915,7 @@ static void stbir__resample_vertical_upsample(stbir__info* stbir_info, int n, in for (k = n0; k <= n1; k++) { int coefficient_index = coefficient_counter++; - float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, kernel_pixel_width, ring_buffer_length); + float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, ring_buffer_entries, ring_buffer_length); float coefficient = vertical_coefficients[coefficient_group + coefficient_index]; for (x = 0; x < output_w; ++x) { @@ -1927,7 +1929,7 @@ static void stbir__resample_vertical_upsample(stbir__info* stbir_info, int n, in for (k = n0; k <= n1; k++) { int coefficient_index = coefficient_counter++; - float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, kernel_pixel_width, ring_buffer_length); + float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, ring_buffer_entries, ring_buffer_length); float coefficient = vertical_coefficients[coefficient_group + coefficient_index]; for (x = 0; x < output_w; ++x) { @@ -1942,7 +1944,7 @@ static void stbir__resample_vertical_upsample(stbir__info* stbir_info, int n, in for (k = n0; k <= n1; k++) { int coefficient_index = coefficient_counter++; - float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, kernel_pixel_width, ring_buffer_length); + float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, ring_buffer_entries, ring_buffer_length); float coefficient = vertical_coefficients[coefficient_group + coefficient_index]; for (x = 0; x < output_w; ++x) { @@ -1958,7 +1960,7 @@ static void stbir__resample_vertical_upsample(stbir__info* stbir_info, int n, in for (k = n0; k <= n1; k++) { int coefficient_index = coefficient_counter++; - float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, kernel_pixel_width, ring_buffer_length); + float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, ring_buffer_entries, ring_buffer_length); float coefficient = vertical_coefficients[coefficient_group + coefficient_index]; for (x = 0; x < output_w; ++x) { @@ -1973,16 +1975,14 @@ static void stbir__resample_vertical_upsample(stbir__info* stbir_info, int n, in stbir__encode_scanline(stbir_info, output_w, (char *) output_data + output_row_start, encode_buffer, channels, alpha_channel, decode); } -static void stbir__resample_vertical_downsample(stbir__info* stbir_info, int n, int in_first_scanline, int in_last_scanline, float in_center_of_out) +static void stbir__resample_vertical_downsample(stbir__info* stbir_info, int n) { int x, k; int output_w = stbir_info->output_w; - int output_h = stbir_info->output_h; stbir__contributors* vertical_contributors = stbir_info->vertical_contributors; float* vertical_coefficients = stbir_info->vertical_coefficients; int channels = stbir_info->channels; - int kernel_pixel_width = stbir_info->vertical_filter_pixel_width; - void* output_data = stbir_info->output_data; + int ring_buffer_entries = stbir_info->ring_buffer_num_entries; float* horizontal_buffer = stbir_info->horizontal_buffer; int coefficient_width = stbir_info->vertical_coefficient_width; int contributor = n + stbir_info->vertical_filter_pixel_margin; @@ -1990,7 +1990,6 @@ static void stbir__resample_vertical_downsample(stbir__info* stbir_info, int n, float* ring_buffer = stbir_info->ring_buffer; int ring_buffer_begin_index = stbir_info->ring_buffer_begin_index; int ring_buffer_first_scanline = stbir_info->ring_buffer_first_scanline; - int ring_buffer_last_scanline = stbir_info->ring_buffer_last_scanline; int ring_buffer_length = stbir_info->ring_buffer_length_bytes/sizeof(float); int n0,n1; @@ -2005,7 +2004,7 @@ static void stbir__resample_vertical_downsample(stbir__info* stbir_info, int n, int coefficient_group = coefficient_width * contributor; float coefficient = vertical_coefficients[coefficient_group + coefficient_index]; - float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, kernel_pixel_width, ring_buffer_length); + float* ring_buffer_entry = stbir__get_ring_buffer_scanline(k, ring_buffer, ring_buffer_begin_index, ring_buffer_first_scanline, ring_buffer_entries, ring_buffer_length); switch (channels) { case 1: @@ -2071,7 +2070,7 @@ static void stbir__buffer_loop_upsample(stbir__info* stbir_info) stbir__calculate_sample_range_upsample(y, out_scanlines_radius, scale_ratio, stbir_info->vertical_shift, &in_first_scanline, &in_last_scanline, &in_center_of_out); - STBIR_ASSERT(in_last_scanline - in_first_scanline <= stbir_info->vertical_filter_pixel_width); + STBIR_ASSERT(in_last_scanline - in_first_scanline + 1 <= stbir_info->ring_buffer_num_entries); if (stbir_info->ring_buffer_begin_index >= 0) { @@ -2090,7 +2089,7 @@ static void stbir__buffer_loop_upsample(stbir__info* stbir_info) else { stbir_info->ring_buffer_first_scanline++; - stbir_info->ring_buffer_begin_index = (stbir_info->ring_buffer_begin_index + 1) % stbir_info->vertical_filter_pixel_width; + stbir_info->ring_buffer_begin_index = (stbir_info->ring_buffer_begin_index + 1) % stbir_info->ring_buffer_num_entries; } } } @@ -2103,7 +2102,7 @@ static void stbir__buffer_loop_upsample(stbir__info* stbir_info) stbir__decode_and_resample_upsample(stbir_info, stbir_info->ring_buffer_last_scanline + 1); // Now all buffers should be ready to write a row of vertical sampling. - stbir__resample_vertical_upsample(stbir_info, y, in_first_scanline, in_last_scanline, in_center_of_out); + stbir__resample_vertical_upsample(stbir_info, y); STBIR_PROGRESS_REPORT((float)y / stbir_info->output_h); } @@ -2148,7 +2147,7 @@ static void stbir__empty_ring_buffer(stbir__info* stbir_info, int first_necessar else { stbir_info->ring_buffer_first_scanline++; - stbir_info->ring_buffer_begin_index = (stbir_info->ring_buffer_begin_index + 1) % stbir_info->vertical_filter_pixel_width; + stbir_info->ring_buffer_begin_index = (stbir_info->ring_buffer_begin_index + 1) % stbir_info->ring_buffer_num_entries; } } } @@ -2172,7 +2171,7 @@ static void stbir__buffer_loop_downsample(stbir__info* stbir_info) stbir__calculate_sample_range_downsample(y, in_pixels_radius, scale_ratio, stbir_info->vertical_shift, &out_first_scanline, &out_last_scanline, &out_center_of_in); - STBIR_ASSERT(out_last_scanline - out_first_scanline <= stbir_info->vertical_filter_pixel_width); + STBIR_ASSERT(out_last_scanline - out_first_scanline + 1 <= stbir_info->ring_buffer_num_entries); if (out_last_scanline < 0 || out_first_scanline >= output_h) continue; @@ -2189,7 +2188,7 @@ static void stbir__buffer_loop_downsample(stbir__info* stbir_info) stbir__add_empty_ring_buffer_entry(stbir_info, stbir_info->ring_buffer_last_scanline + 1); // Now the horizontal buffer is ready to write to all ring buffer rows. - stbir__resample_vertical_downsample(stbir_info, y, out_first_scanline, out_last_scanline, out_center_of_in); + stbir__resample_vertical_downsample(stbir_info, y); } stbir__empty_ring_buffer(stbir_info, stbir_info->output_h); @@ -2246,13 +2245,16 @@ static stbir_uint32 stbir__calculate_memory(stbir__info *info) info->horizontal_num_contributors = stbir__get_contributors(info->horizontal_scale, info->horizontal_filter, info->input_w, info->output_w); info->vertical_num_contributors = stbir__get_contributors(info->vertical_scale , info->vertical_filter , info->input_h, info->output_h); + // One extra entry because floating point precision problems sometimes cause an extra to be necessary. + info->ring_buffer_num_entries = filter_height + 1; + info->horizontal_contributors_size = info->horizontal_num_contributors * sizeof(stbir__contributors); info->horizontal_coefficients_size = stbir__get_total_horizontal_coefficients(info) * sizeof(float); info->vertical_contributors_size = info->vertical_num_contributors * sizeof(stbir__contributors); info->vertical_coefficients_size = stbir__get_total_vertical_coefficients(info) * sizeof(float); info->decode_buffer_size = (info->input_w + pixel_margin * 2) * info->channels * sizeof(float); info->horizontal_buffer_size = info->output_w * info->channels * sizeof(float); - info->ring_buffer_size = info->output_w * info->channels * filter_height * sizeof(float); + info->ring_buffer_size = info->output_w * info->channels * info->ring_buffer_num_entries * sizeof(float); info->encode_buffer_size = info->output_w * info->channels * sizeof(float); STBIR_ASSERT(info->horizontal_filter != 0); @@ -2390,8 +2392,8 @@ static int stbir__resize_allocated(stbir__info *info, // This signals that the ring buffer is empty info->ring_buffer_begin_index = -1; - stbir__calculate_filters(info, info->horizontal_contributors, info->horizontal_coefficients, info->horizontal_filter, info->horizontal_scale, info->horizontal_shift, info->input_w, info->output_w); - stbir__calculate_filters(info, info->vertical_contributors, info->vertical_coefficients, info->vertical_filter, info->vertical_scale, info->vertical_shift, info->input_h, info->output_h); + stbir__calculate_filters(info->horizontal_contributors, info->horizontal_coefficients, info->horizontal_filter, info->horizontal_scale, info->horizontal_shift, info->input_w, info->output_w); + stbir__calculate_filters(info->vertical_contributors, info->vertical_coefficients, info->vertical_filter, info->vertical_scale, info->vertical_shift, info->input_h, info->output_h); STBIR_PROGRESS_REPORT(0); @@ -2578,3 +2580,45 @@ STBIRDEF int stbir_resize_region( const void *input_pixels , int input_w , int } #endif // STB_IMAGE_RESIZE_IMPLEMENTATION + +/* +------------------------------------------------------------------------------ +This software is available under 2 licenses -- choose whichever you prefer. +------------------------------------------------------------------------------ +ALTERNATIVE A - MIT License +Copyright (c) 2017 Sean Barrett +Permission is hereby granted, free of charge, to any person obtaining a copy of +this software and associated documentation files (the "Software"), to deal in +the Software without restriction, including without limitation the rights to +use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies +of the Software, and to permit persons to whom the Software is furnished to do +so, subject to the following conditions: +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. +------------------------------------------------------------------------------ +ALTERNATIVE B - Public Domain (www.unlicense.org) +This is free and unencumbered software released into the public domain. +Anyone is free to copy, modify, publish, use, compile, sell, or distribute this +software, either in source code form or as a compiled binary, for any purpose, +commercial or non-commercial, and by any means. +In jurisdictions that recognize copyright laws, the author or authors of this +software dedicate any and all copyright interest in the software to the public +domain. We make this dedication for the benefit of the public at large and to +the detriment of our heirs and successors. We intend this dedication to be an +overt act of relinquishment in perpetuity of all present and future rights to +this software under copyright law. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN +ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION +WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. +------------------------------------------------------------------------------ +*/ diff --git a/src/external/stb_image_write.h b/src/external/stb_image_write.h index ae9180b03..df623393d 100644 --- a/src/external/stb_image_write.h +++ b/src/external/stb_image_write.h @@ -1,4 +1,4 @@ -/* stb_image_write - v1.03 - public domain - http://nothings.org/stb/stb_image_write.h +/* stb_image_write - v1.05 - public domain - http://nothings.org/stb/stb_image_write.h writes out PNG/BMP/TGA images to C stdio - Sean Barrett 2010-2015 no warranty implied; use at your own risk @@ -104,12 +104,11 @@ CREDITS: Filip Wasil Thatcher Ulrich github:poppolopoppo + Patrick Boettcher LICENSE -This software is dual-licensed to the public domain and under the following -license: you are granted a perpetual, irrevocable license to copy, modify, -publish, and distribute this file as you see fit. + See end of file for license information. */ @@ -294,10 +293,8 @@ static void stbiw__write_pixel(stbi__write_context *s, int rgb_dir, int comp, in s->func(s->context, &d[comp - 1], 1); switch (comp) { + case 2: // 2 pixels = mono + alpha, alpha is written separately, so same as 1-channel case case 1: - s->func(s->context,d,1); - break; - case 2: if (expand_mono) stbiw__write3(s, d[0], d[0], d[0]); // monochrome bmp else @@ -897,6 +894,7 @@ static unsigned char stbiw__paeth(int a, int b, int c) return STBIW_UCHAR(c); } +// @OPTIMIZE: provide an option that always forces left-predict or paeth predict unsigned char *stbi_write_png_to_mem(unsigned char *pixels, int stride_bytes, int x, int y, int n, int *out_len) { int ctype[5] = { -1, 0, 4, 2, 6 }; @@ -913,10 +911,10 @@ unsigned char *stbi_write_png_to_mem(unsigned char *pixels, int stride_bytes, in for (j=0; j < y; ++j) { static int mapping[] = { 0,1,2,3,4 }; static int firstmap[] = { 0,1,0,5,6 }; - int *mymap = j ? mapping : firstmap; + int *mymap = (j != 0) ? mapping : firstmap; int best = 0, bestval = 0x7fffffff; for (p=0; p < 2; ++p) { - for (k= p?best:0; k < 5; ++k) { + for (k= p?best:0; k < 5; ++k) { // @TODO: clarity: rewrite this to go 0..5, and 'continue' the unwanted ones during 2nd pass int type = mymap[k],est=0; unsigned char *z = pixels + stride_bytes*j; for (i=0; i < n; ++i) @@ -1018,6 +1016,9 @@ STBIWDEF int stbi_write_png_to_func(stbi_write_func *func, void *context, int x, #endif // STB_IMAGE_WRITE_IMPLEMENTATION /* Revision history + 1.04 (2017-03-03) + monochrome BMP expansion + 1.03 ??? 1.02 (2016-04-02) avoid allocating large structures on the stack 1.01 (2016-01-16) @@ -1047,3 +1048,45 @@ STBIWDEF int stbi_write_png_to_func(stbi_write_func *func, void *context, int x, first public release 0.90 first internal release */ + +/* +------------------------------------------------------------------------------ +This software is available under 2 licenses -- choose whichever you prefer. +------------------------------------------------------------------------------ +ALTERNATIVE A - MIT License +Copyright (c) 2017 Sean Barrett +Permission is hereby granted, free of charge, to any person obtaining a copy of +this software and associated documentation files (the "Software"), to deal in +the Software without restriction, including without limitation the rights to +use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies +of the Software, and to permit persons to whom the Software is furnished to do +so, subject to the following conditions: +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. +------------------------------------------------------------------------------ +ALTERNATIVE B - Public Domain (www.unlicense.org) +This is free and unencumbered software released into the public domain. +Anyone is free to copy, modify, publish, use, compile, sell, or distribute this +software, either in source code form or as a compiled binary, for any purpose, +commercial or non-commercial, and by any means. +In jurisdictions that recognize copyright laws, the author or authors of this +software dedicate any and all copyright interest in the software to the public +domain. We make this dedication for the benefit of the public at large and to +the detriment of our heirs and successors. We intend this dedication to be an +overt act of relinquishment in perpetuity of all present and future rights to +this software under copyright law. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN +ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION +WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. +------------------------------------------------------------------------------ +*/ diff --git a/src/external/stb_rect_pack.h b/src/external/stb_rect_pack.h index c75527da1..f5eb8d335 100644 --- a/src/external/stb_rect_pack.h +++ b/src/external/stb_rect_pack.h @@ -1,4 +1,4 @@ -// stb_rect_pack.h - v0.10 - public domain - rectangle packing +// stb_rect_pack.h - v0.11 - public domain - rectangle packing // Sean Barrett 2014 // // Useful for e.g. packing rectangular textures into an atlas. @@ -27,11 +27,14 @@ // Sean Barrett // Minor features // Martins Mozeiko +// github:IntellectualKitty +// // Bugfixes / warning fixes // Jeremy Jaussaud // // Version history: // +// 0.11 (2017-03-03) return packing success/fail result // 0.10 (2016-10-25) remove cast-away-const to avoid warnings // 0.09 (2016-08-27) fix compiler warnings // 0.08 (2015-09-13) really fix bug with empty rects (w=0 or h=0) @@ -43,9 +46,7 @@ // // LICENSE // -// This software is dual-licensed to the public domain and under the following -// license: you are granted a perpetual, irrevocable license to copy, modify, -// publish, and distribute this file as you see fit. +// See end of file for license information. ////////////////////////////////////////////////////////////////////////////// // @@ -77,7 +78,7 @@ typedef int stbrp_coord; typedef unsigned short stbrp_coord; #endif -STBRP_DEF void stbrp_pack_rects (stbrp_context *context, stbrp_rect *rects, int num_rects); +STBRP_DEF int stbrp_pack_rects (stbrp_context *context, stbrp_rect *rects, int num_rects); // Assign packed locations to rectangles. The rectangles are of type // 'stbrp_rect' defined below, stored in the array 'rects', and there // are 'num_rects' many of them. @@ -98,6 +99,9 @@ STBRP_DEF void stbrp_pack_rects (stbrp_context *context, stbrp_rect *rects, int // arrays will probably produce worse packing results than calling it // a single time with the full rectangle array, but the option is // available. +// +// The function returns 1 if all of the rectangles were successfully +// packed and 0 otherwise. struct stbrp_rect { @@ -544,9 +548,9 @@ static int rect_original_order(const void *a, const void *b) #define STBRP__MAXVAL 0xffff #endif -STBRP_DEF void stbrp_pack_rects(stbrp_context *context, stbrp_rect *rects, int num_rects) +STBRP_DEF int stbrp_pack_rects(stbrp_context *context, stbrp_rect *rects, int num_rects) { - int i; + int i, all_rects_packed = 1; // we use the 'was_packed' field internally to allow sorting/unsorting for (i=0; i < num_rects; ++i) { @@ -576,8 +580,56 @@ STBRP_DEF void stbrp_pack_rects(stbrp_context *context, stbrp_rect *rects, int n // unsort STBRP_SORT(rects, num_rects, sizeof(rects[0]), rect_original_order); - // set was_packed flags - for (i=0; i < num_rects; ++i) + // set was_packed flags and all_rects_packed status + for (i=0; i < num_rects; ++i) { rects[i].was_packed = !(rects[i].x == STBRP__MAXVAL && rects[i].y == STBRP__MAXVAL); + if (!rects[i].was_packed) + all_rects_packed = 0; + } + + // return the all_rects_packed status + return all_rects_packed; } #endif + +/* +------------------------------------------------------------------------------ +This software is available under 2 licenses -- choose whichever you prefer. +------------------------------------------------------------------------------ +ALTERNATIVE A - MIT License +Copyright (c) 2017 Sean Barrett +Permission is hereby granted, free of charge, to any person obtaining a copy of +this software and associated documentation files (the "Software"), to deal in +the Software without restriction, including without limitation the rights to +use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies +of the Software, and to permit persons to whom the Software is furnished to do +so, subject to the following conditions: +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. +------------------------------------------------------------------------------ +ALTERNATIVE B - Public Domain (www.unlicense.org) +This is free and unencumbered software released into the public domain. +Anyone is free to copy, modify, publish, use, compile, sell, or distribute this +software, either in source code form or as a compiled binary, for any purpose, +commercial or non-commercial, and by any means. +In jurisdictions that recognize copyright laws, the author or authors of this +software dedicate any and all copyright interest in the software to the public +domain. We make this dedication for the benefit of the public at large and to +the detriment of our heirs and successors. We intend this dedication to be an +overt act of relinquishment in perpetuity of all present and future rights to +this software under copyright law. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN +ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION +WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. +------------------------------------------------------------------------------ +*/ diff --git a/src/external/stb_truetype.h b/src/external/stb_truetype.h index 92b9a8754..fc5b97821 100644 --- a/src/external/stb_truetype.h +++ b/src/external/stb_truetype.h @@ -1,4 +1,4 @@ -// stb_truetype.h - v1.14 - public domain +// stb_truetype.h - v1.15 - public domain // authored from 2009-2016 by Sean Barrett / RAD Game Tools // // This library processes TrueType files: @@ -50,10 +50,13 @@ // Higor Euripedes // Thomas Fields // Derek Vinyard +// Cort Stratton // // VERSION HISTORY // -// 1.13 (2017-01-02) support OpenType fonts, certain Apple fonts, num-fonts-in-TTC function +// 1.15 (2017-03-03) make more arguments const +// 1.14 (2017-01-16) num-fonts-in-TTC function +// 1.13 (2017-01-02) support OpenType fonts, certain Apple fonts // 1.12 (2016-10-25) suppress warnings about casting away const with -Wcast-qual // 1.11 (2016-04-02) fix unused-variable warning // 1.10 (2016-04-02) user-defined fabs(); rare memory leak; remove duplicate typedef @@ -69,9 +72,7 @@ // // LICENSE // -// This software is dual-licensed to the public domain and under the following -// license: you are granted a perpetual, irrevocable license to copy, modify, -// publish, and distribute this file as you see fit. +// See end of file for license information. // // USAGE // @@ -494,7 +495,7 @@ typedef struct float x1,y1,s1,t1; // bottom-right } stbtt_aligned_quad; -STBTT_DEF void stbtt_GetBakedQuad(stbtt_bakedchar *chardata, int pw, int ph, // same data as above +STBTT_DEF void stbtt_GetBakedQuad(const stbtt_bakedchar *chardata, int pw, int ph, // same data as above int char_index, // character to display float *xpos, float *ypos, // pointers to current position in screen pixel space stbtt_aligned_quad *q, // output: quad to draw @@ -594,7 +595,7 @@ STBTT_DEF void stbtt_PackSetOversampling(stbtt_pack_context *spc, unsigned int h // To use with PackFontRangesGather etc., you must set it before calls // call to PackFontRangesGatherRects. -STBTT_DEF void stbtt_GetPackedQuad(stbtt_packedchar *chardata, int pw, int ph, // same data as above +STBTT_DEF void stbtt_GetPackedQuad(const stbtt_packedchar *chardata, int pw, int ph, // same data as above int char_index, // character to display float *xpos, float *ypos, // pointers to current position in screen pixel space stbtt_aligned_quad *q, // output: quad to draw @@ -3287,11 +3288,11 @@ static int stbtt_BakeFontBitmap_internal(unsigned char *data, int offset, // fo return bottom_y; } -STBTT_DEF void stbtt_GetBakedQuad(stbtt_bakedchar *chardata, int pw, int ph, int char_index, float *xpos, float *ypos, stbtt_aligned_quad *q, int opengl_fillrule) +STBTT_DEF void stbtt_GetBakedQuad(const stbtt_bakedchar *chardata, int pw, int ph, int char_index, float *xpos, float *ypos, stbtt_aligned_quad *q, int opengl_fillrule) { float d3d_bias = opengl_fillrule ? 0 : -0.5f; float ipw = 1.0f / pw, iph = 1.0f / ph; - stbtt_bakedchar *b = chardata + char_index; + const stbtt_bakedchar *b = chardata + char_index; int round_x = STBTT_ifloor((*xpos + b->xoff) + 0.5f); int round_y = STBTT_ifloor((*ypos + b->yoff) + 0.5f); @@ -3735,10 +3736,10 @@ STBTT_DEF int stbtt_PackFontRange(stbtt_pack_context *spc, unsigned char *fontda return stbtt_PackFontRanges(spc, fontdata, font_index, &range, 1); } -STBTT_DEF void stbtt_GetPackedQuad(stbtt_packedchar *chardata, int pw, int ph, int char_index, float *xpos, float *ypos, stbtt_aligned_quad *q, int align_to_integer) +STBTT_DEF void stbtt_GetPackedQuad(const stbtt_packedchar *chardata, int pw, int ph, int char_index, float *xpos, float *ypos, stbtt_aligned_quad *q, int align_to_integer) { float ipw = 1.0f / pw, iph = 1.0f / ph; - stbtt_packedchar *b = chardata + char_index; + const stbtt_packedchar *b = chardata + char_index; if (align_to_integer) { float x = (float) STBTT_ifloor((*xpos + b->xoff) + 0.5f); @@ -4016,3 +4017,45 @@ STBTT_DEF int stbtt_CompareUTF8toUTF16_bigendian(const char *s1, int len1, const // 0.2 (2009-03-11) Fix unsigned/signed char warnings // 0.1 (2009-03-09) First public release // + +/* +------------------------------------------------------------------------------ +This software is available under 2 licenses -- choose whichever you prefer. +------------------------------------------------------------------------------ +ALTERNATIVE A - MIT License +Copyright (c) 2017 Sean Barrett +Permission is hereby granted, free of charge, to any person obtaining a copy of +this software and associated documentation files (the "Software"), to deal in +the Software without restriction, including without limitation the rights to +use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies +of the Software, and to permit persons to whom the Software is furnished to do +so, subject to the following conditions: +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. +------------------------------------------------------------------------------ +ALTERNATIVE B - Public Domain (www.unlicense.org) +This is free and unencumbered software released into the public domain. +Anyone is free to copy, modify, publish, use, compile, sell, or distribute this +software, either in source code form or as a compiled binary, for any purpose, +commercial or non-commercial, and by any means. +In jurisdictions that recognize copyright laws, the author or authors of this +software dedicate any and all copyright interest in the software to the public +domain. We make this dedication for the benefit of the public at large and to +the detriment of our heirs and successors. We intend this dedication to be an +overt act of relinquishment in perpetuity of all present and future rights to +this software under copyright law. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN +ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION +WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. +------------------------------------------------------------------------------ +*/ diff --git a/src/external/stb_vorbis.c b/src/external/stb_vorbis.c index 07d79318b..ac8c9ca55 100644 --- a/src/external/stb_vorbis.c +++ b/src/external/stb_vorbis.c @@ -162,21 +162,23 @@ #endif #ifndef STB_VORBIS_NO_CRT -#include -#include -#include -#include -#if !(defined(__APPLE__) || defined(MACOSX) || defined(macintosh) || defined(Macintosh)) -#include -#if defined(__linux__) || defined(__linux) || defined(__EMSCRIPTEN__) -#include -#endif -#endif + #include + #include + #include + #include + + // find definition of alloca if it's not in stdlib.h: + #ifdef _MSC_VER + #include + #endif + #if defined(__linux__) || defined(__linux) || defined(__EMSCRIPTEN__) + #include + #endif #else // STB_VORBIS_NO_CRT -#define NULL 0 -#define malloc(s) 0 -#define free(s) ((void) 0) -#define realloc(s) 0 + #define NULL 0 + #define malloc(s) 0 + #define free(s) ((void) 0) + #define realloc(s) 0 #endif // STB_VORBIS_NO_CRT #include @@ -597,17 +599,18 @@ static int ilog(int32 n) { static signed char log2_4[16] = { 0,1,2,2,3,3,3,3,4,4,4,4,4,4,4,4 }; + if (n < 0) return 0; // signed n returns 0 + // 2 compares if n < 16, 3 compares otherwise (4 if signed or n > 1<<29) if (n < (1 << 14)) - if (n < (1 << 4)) return 0 + log2_4[n ]; - else if (n < (1 << 9)) return 5 + log2_4[n >> 5]; + if (n < (1 << 4)) return 0 + log2_4[n ]; + else if (n < (1 << 9)) return 5 + log2_4[n >> 5]; else return 10 + log2_4[n >> 10]; else if (n < (1 << 24)) - if (n < (1 << 19)) return 15 + log2_4[n >> 15]; + if (n < (1 << 19)) return 15 + log2_4[n >> 15]; else return 20 + log2_4[n >> 20]; - else if (n < (1 << 29)) return 25 + log2_4[n >> 25]; - else if (n < (1 << 31)) return 30 + log2_4[n >> 30]; - else return 0; // signed n returns 0 + else if (n < (1 << 29)) return 25 + log2_4[n >> 25]; + else return 30 + log2_4[n >> 30]; } #ifndef M_PI @@ -880,13 +883,13 @@ static void neighbors(uint16 *x, int n, int *plow, int *phigh) // this has been repurposed so y is now the original index instead of y typedef struct { - uint16 x,y; -} Point; + uint16 x,id; +} stbv__floor_ordering; static int STBV_CDECL point_compare(const void *p, const void *q) { - Point *a = (Point *) p; - Point *b = (Point *) q; + stbv__floor_ordering *a = (stbv__floor_ordering *) p; + stbv__floor_ordering *b = (stbv__floor_ordering *) q; return a->x < b->x ? -1 : a->x > b->x; } @@ -3095,11 +3098,13 @@ static int vorbis_finish_frame(stb_vorbis *f, int len, int left, int right) return right - left; } -static void vorbis_pump_first_frame(stb_vorbis *f) +static int vorbis_pump_first_frame(stb_vorbis *f) { - int len, right, left; - if (vorbis_decode_packet(f, &len, &left, &right)) + int len, right, left, res; + res = vorbis_decode_packet(f, &len, &left, &right); + if (res) vorbis_finish_frame(f, len, left, right); + return res; } #ifndef STB_VORBIS_NO_PUSHDATA_API @@ -3482,7 +3487,7 @@ static int start_decoder(vorb *f) g->book_list[j] = get_bits(f,8); return error(f, VORBIS_feature_not_supported); } else { - Point p[31*8+2]; + stbv__floor_ordering p[31*8+2]; Floor1 *g = &f->floor_config[i].floor1; int max_class = -1; g->partitions = get_bits(f, 5); @@ -3518,11 +3523,11 @@ static int start_decoder(vorb *f) // precompute the sorting for (j=0; j < g->values; ++j) { p[j].x = g->Xlist[j]; - p[j].y = j; + p[j].id = j; } qsort(p, g->values, sizeof(p[0]), point_compare); for (j=0; j < g->values; ++j) - g->sorted_order[j] = (uint8) p[j].y; + g->sorted_order[j] = (uint8) p[j].id; // precompute the neighbors for (j=2; j < g->values; ++j) { int low,hi; @@ -4226,8 +4231,9 @@ static int seek_to_sample_coarse(stb_vorbis *f, uint32 sample_number) // starting from the start is handled differently if (sample_number <= left.last_decoded_sample) { - stb_vorbis_seek_start(f); - return 1; + if (stb_vorbis_seek_start(f)) + return 1; + return 0; } while (left.page_end != right.page_start) { @@ -4328,7 +4334,10 @@ static int seek_to_sample_coarse(stb_vorbis *f, uint32 sample_number) skip(f, f->segments[i]); // start decoding (optimizable - this frame is generally discarded) - vorbis_pump_first_frame(f); + if (!vorbis_pump_first_frame(f)) + return 0; + if (f->current_loc > sample_number) + return error(f, VORBIS_seek_failed); return 1; error: @@ -4419,14 +4428,14 @@ int stb_vorbis_seek(stb_vorbis *f, unsigned int sample_number) return 1; } -void stb_vorbis_seek_start(stb_vorbis *f) +int stb_vorbis_seek_start(stb_vorbis *f) { - if (IS_PUSH_MODE(f)) { error(f, VORBIS_invalid_api_mixing); return; } + if (IS_PUSH_MODE(f)) { return error(f, VORBIS_invalid_api_mixing); } set_file_offset(f, f->first_audio_page_offset); f->previous_length = 0; f->first_decode = TRUE; f->next_seg = -1; - vorbis_pump_first_frame(f); + return vorbis_pump_first_frame(f); } unsigned int stb_vorbis_stream_length_in_samples(stb_vorbis *f) @@ -4591,6 +4600,7 @@ stb_vorbis * stb_vorbis_open_memory(const unsigned char *data, int len, int *err if (f) { *f = p; vorbis_pump_first_frame(f); + if (error) *error = VORBIS__no_error; return f; } } @@ -4956,6 +4966,7 @@ int stb_vorbis_get_samples_float(stb_vorbis *f, int channels, float **buffer, in #endif // STB_VORBIS_NO_PULLDATA_API /* Version history + 1.10 - 2017/03/03 - more robust seeking; fix negative ilog(); clear error in open_memory 1.09 - 2016/04/04 - back out 'avoid discarding last frame' fix from previous version 1.08 - 2016/04/02 - fixed multiple warnings; fix setup memory leaks; avoid discarding last frame of audio data @@ -5008,3 +5019,46 @@ int stb_vorbis_get_samples_float(stb_vorbis *f, int channels, float **buffer, in */ #endif // STB_VORBIS_HEADER_ONLY + + +/* +------------------------------------------------------------------------------ +This software is available under 2 licenses -- choose whichever you prefer. +------------------------------------------------------------------------------ +ALTERNATIVE A - MIT License +Copyright (c) 2017 Sean Barrett +Permission is hereby granted, free of charge, to any person obtaining a copy of +this software and associated documentation files (the "Software"), to deal in +the Software without restriction, including without limitation the rights to +use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies +of the Software, and to permit persons to whom the Software is furnished to do +so, subject to the following conditions: +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. +------------------------------------------------------------------------------ +ALTERNATIVE B - Public Domain (www.unlicense.org) +This is free and unencumbered software released into the public domain. +Anyone is free to copy, modify, publish, use, compile, sell, or distribute this +software, either in source code form or as a compiled binary, for any purpose, +commercial or non-commercial, and by any means. +In jurisdictions that recognize copyright laws, the author or authors of this +software dedicate any and all copyright interest in the software to the public +domain. We make this dedication for the benefit of the public at large and to +the detriment of our heirs and successors. We intend this dedication to be an +overt act of relinquishment in perpetuity of all present and future rights to +this software under copyright law. +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN +ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION +WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. +------------------------------------------------------------------------------ +*/ diff --git a/src/external/stb_vorbis.h b/src/external/stb_vorbis.h index 624ce4bc2..9394e8130 100644 --- a/src/external/stb_vorbis.h +++ b/src/external/stb_vorbis.h @@ -1,4 +1,4 @@ -// Ogg Vorbis audio decoder - v1.09 - public domain +// Ogg Vorbis audio decoder - v1.10 - public domain // http://nothings.org/stb_vorbis/ // // Original version written by Sean Barrett in 2007. @@ -9,12 +9,7 @@ // // LICENSE // -// This software is dual-licensed to the public domain and under the following -// license: you are granted a perpetual, irrevocable license to copy, modify, -// publish, and distribute this file as you see fit. -// -// No warranty for any purpose is expressed or implied by the author (nor -// by RAD Game Tools). Report bugs and send enhancements to the author. +// See end of file for license information. // // Limitations: // @@ -37,6 +32,7 @@ // manxorist@github saga musix // // Partial history: +// 1.10 - 2017/03/03 - more robust seeking; fix negative ilog(); clear error in open_memory // 1.09 - 2016/04/04 - back out 'truncation of last frame' fix from previous version // 1.08 - 2016/04/02 - warnings; setup memory leaks; truncation of last frame // 1.07 - 2015/01/16 - fixes for crashes on invalid files; warning fixes; const @@ -280,7 +276,7 @@ extern int stb_vorbis_seek(stb_vorbis *f, unsigned int sample_number); // do not need to seek to EXACTLY the target sample when using get_samples_*, // you can also use seek_frame(). -extern void stb_vorbis_seek_start(stb_vorbis *f); +extern int stb_vorbis_seek_start(stb_vorbis *f); // this function is equivalent to stb_vorbis_seek(f,0) extern unsigned int stb_vorbis_stream_length_in_samples(stb_vorbis *f); diff --git a/src/textures.c b/src/textures.c index 8e3a1ee1f..9fd5944eb 100644 --- a/src/textures.c +++ b/src/textures.c @@ -195,9 +195,13 @@ Image LoadImage(const char *fileName) int imgWidth = 0; int imgHeight = 0; int imgBpp = 0; + + FILE *imFile = fopen(fileName, "rb"); // NOTE: Using stb_image to load images (Supports: BMP, TGA, PNG, JPG, ...) - image.data = stbi_load(fileName, &imgWidth, &imgHeight, &imgBpp, 0); + image.data = stbi_load_from_file(imFile, &imgWidth, &imgHeight, &imgBpp, 0); + + fclose(imFile); image.width = imgWidth; image.height = imgHeight;