| /* | |
| * Copyright 2014 Google Inc. | |
| * | |
| * Use of this source code is governed by a BSD-style license that can be | |
| * found in the LICENSE file. | |
| */ | |
| #include "SkBitmap.h" | |
| #include "SkCanvas.h" | |
| #include "SkConfig8888.h" | |
| #include "SkColorPriv.h" | |
| #include "SkDither.h" | |
| #include "SkMathPriv.h" | |
| #include "SkUnPreMultiply.h" | |
| enum AlphaVerb { | |
| kNothing_AlphaVerb, | |
| kPremul_AlphaVerb, | |
| kUnpremul_AlphaVerb, | |
| }; | |
| template <bool doSwapRB, AlphaVerb doAlpha> uint32_t convert32(uint32_t c) { | |
| if (doSwapRB) { | |
| c = SkSwizzle_RB(c); | |
| } | |
| // Lucky for us, in both RGBA and BGRA, the alpha component is always in the same place, so | |
| // we can perform premul or unpremul the same way without knowing the swizzles for RGB. | |
| switch (doAlpha) { | |
| case kNothing_AlphaVerb: | |
| // no change | |
| break; | |
| case kPremul_AlphaVerb: | |
| c = SkPreMultiplyARGB(SkGetPackedA32(c), SkGetPackedR32(c), | |
| SkGetPackedG32(c), SkGetPackedB32(c)); | |
| break; | |
| case kUnpremul_AlphaVerb: | |
| c = SkUnPreMultiply::UnPreMultiplyPreservingByteOrder(c); | |
| break; | |
| } | |
| return c; | |
| } | |
| template <bool doSwapRB, AlphaVerb doAlpha> | |
| void convert32_row(uint32_t* dst, const uint32_t* src, int count) { | |
| // This has to be correct if src == dst (but not partial overlap) | |
| for (int i = 0; i < count; ++i) { | |
| dst[i] = convert32<doSwapRB, doAlpha>(src[i]); | |
| } | |
| } | |
| static bool is_32bit_colortype(SkColorType ct) { | |
| return kRGBA_8888_SkColorType == ct || kBGRA_8888_SkColorType == ct; | |
| } | |
| static AlphaVerb compute_AlphaVerb(SkAlphaType src, SkAlphaType dst) { | |
| SkASSERT(kIgnore_SkAlphaType != src); | |
| SkASSERT(kIgnore_SkAlphaType != dst); | |
| if (kOpaque_SkAlphaType == src || kOpaque_SkAlphaType == dst || src == dst) { | |
| return kNothing_AlphaVerb; | |
| } | |
| if (kPremul_SkAlphaType == dst) { | |
| SkASSERT(kUnpremul_SkAlphaType == src); | |
| return kPremul_AlphaVerb; | |
| } else { | |
| SkASSERT(kPremul_SkAlphaType == src); | |
| SkASSERT(kUnpremul_SkAlphaType == dst); | |
| return kUnpremul_AlphaVerb; | |
| } | |
| } | |
| static void memcpy32_row(uint32_t* dst, const uint32_t* src, int count) { | |
| memcpy(dst, src, count * 4); | |
| } | |
| bool SkSrcPixelInfo::convertPixelsTo(SkDstPixelInfo* dst, int width, int height) const { | |
| if (width <= 0 || height <= 0) { | |
| return false; | |
| } | |
| if (!is_32bit_colortype(fColorType) || !is_32bit_colortype(dst->fColorType)) { | |
| return false; | |
| } | |
| void (*proc)(uint32_t* dst, const uint32_t* src, int count); | |
| AlphaVerb doAlpha = compute_AlphaVerb(fAlphaType, dst->fAlphaType); | |
| bool doSwapRB = fColorType != dst->fColorType; | |
| switch (doAlpha) { | |
| case kNothing_AlphaVerb: | |
| if (doSwapRB) { | |
| proc = convert32_row<true, kNothing_AlphaVerb>; | |
| } else { | |
| if (fPixels == dst->fPixels) { | |
| return true; | |
| } | |
| proc = memcpy32_row; | |
| } | |
| break; | |
| case kPremul_AlphaVerb: | |
| if (doSwapRB) { | |
| proc = convert32_row<true, kPremul_AlphaVerb>; | |
| } else { | |
| proc = convert32_row<false, kPremul_AlphaVerb>; | |
| } | |
| break; | |
| case kUnpremul_AlphaVerb: | |
| if (doSwapRB) { | |
| proc = convert32_row<true, kUnpremul_AlphaVerb>; | |
| } else { | |
| proc = convert32_row<false, kUnpremul_AlphaVerb>; | |
| } | |
| break; | |
| } | |
| uint32_t* dstP = static_cast<uint32_t*>(dst->fPixels); | |
| const uint32_t* srcP = static_cast<const uint32_t*>(fPixels); | |
| size_t srcInc = fRowBytes >> 2; | |
| size_t dstInc = dst->fRowBytes >> 2; | |
| for (int y = 0; y < height; ++y) { | |
| proc(dstP, srcP, width); | |
| dstP += dstInc; | |
| srcP += srcInc; | |
| } | |
| return true; | |
| } | |
| static void rect_memcpy(void* dst, size_t dstRB, const void* src, size_t srcRB, size_t bytesPerRow, | |
| int rowCount) { | |
| SkASSERT(bytesPerRow <= srcRB); | |
| SkASSERT(bytesPerRow <= dstRB); | |
| for (int i = 0; i < rowCount; ++i) { | |
| memcpy(dst, src, bytesPerRow); | |
| dst = (char*)dst + dstRB; | |
| src = (const char*)src + srcRB; | |
| } | |
| } | |
| bool SkPixelInfo::CopyPixels(const SkImageInfo& dstInfo, void* dstPixels, size_t dstRB, | |
| const SkImageInfo& srcInfo, const void* srcPixels, size_t srcRB, | |
| SkColorTable* ctable) { | |
| if (srcInfo.dimensions() != dstInfo.dimensions()) { | |
| return false; | |
| } | |
| const int width = srcInfo.width(); | |
| const int height = srcInfo.height(); | |
| // Handle fancy alpha swizzling if both are ARGB32 | |
| if (4 == srcInfo.bytesPerPixel() && 4 == dstInfo.bytesPerPixel()) { | |
| SkDstPixelInfo dstPI; | |
| dstPI.fColorType = dstInfo.colorType(); | |
| dstPI.fAlphaType = dstInfo.alphaType(); | |
| dstPI.fPixels = dstPixels; | |
| dstPI.fRowBytes = dstRB; | |
| SkSrcPixelInfo srcPI; | |
| srcPI.fColorType = srcInfo.colorType(); | |
| srcPI.fAlphaType = srcInfo.alphaType(); | |
| srcPI.fPixels = srcPixels; | |
| srcPI.fRowBytes = srcRB; | |
| return srcPI.convertPixelsTo(&dstPI, width, height); | |
| } | |
| // If they agree on colorType and the alphaTypes are compatible, then we just memcpy. | |
| // Note: we've already taken care of 32bit colortypes above. | |
| if (srcInfo.colorType() == dstInfo.colorType()) { | |
| switch (srcInfo.colorType()) { | |
| case kRGB_565_SkColorType: | |
| case kAlpha_8_SkColorType: | |
| break; | |
| case kIndex_8_SkColorType: | |
| case kARGB_4444_SkColorType: | |
| if (srcInfo.alphaType() != dstInfo.alphaType()) { | |
| return false; | |
| } | |
| break; | |
| default: | |
| return false; | |
| } | |
| rect_memcpy(dstPixels, dstRB, srcPixels, srcRB, width * srcInfo.bytesPerPixel(), height); | |
| return true; | |
| } | |
| /* | |
| * Begin section where we try to change colorTypes along the way. Not all combinations | |
| * are supported. | |
| */ | |
| // Can no longer draw directly into 4444, but we can manually whack it for a few combinations | |
| if (kARGB_4444_SkColorType == dstInfo.colorType() && | |
| (kN32_SkColorType == srcInfo.colorType() || kIndex_8_SkColorType == srcInfo.colorType())) { | |
| if (srcInfo.alphaType() == kUnpremul_SkAlphaType) { | |
| // Our method for converting to 4444 assumes premultiplied. | |
| return false; | |
| } | |
| const SkPMColor* table = NULL; | |
| if (kIndex_8_SkColorType == srcInfo.colorType()) { | |
| if (NULL == ctable) { | |
| return false; | |
| } | |
| table = ctable->lockColors(); | |
| } | |
| for (int y = 0; y < height; ++y) { | |
| DITHER_4444_SCAN(y); | |
| SkPMColor16* SK_RESTRICT dstRow = (SkPMColor16*)dstPixels; | |
| if (table) { | |
| const uint8_t* SK_RESTRICT srcRow = (const uint8_t*)srcPixels; | |
| for (int x = 0; x < width; ++x) { | |
| dstRow[x] = SkDitherARGB32To4444(table[srcRow[x]], DITHER_VALUE(x)); | |
| } | |
| } else { | |
| const SkPMColor* SK_RESTRICT srcRow = (const SkPMColor*)srcPixels; | |
| for (int x = 0; x < width; ++x) { | |
| dstRow[x] = SkDitherARGB32To4444(srcRow[x], DITHER_VALUE(x)); | |
| } | |
| } | |
| dstPixels = (char*)dstPixels + dstRB; | |
| srcPixels = (const char*)srcPixels + srcRB; | |
| } | |
| if (table) { | |
| ctable->unlockColors(); | |
| } | |
| return true; | |
| } | |
| if (dstInfo.alphaType() == kUnpremul_SkAlphaType) { | |
| // We do not support drawing to unpremultiplied bitmaps. | |
| return false; | |
| } | |
| // Final fall-back, draw with a canvas | |
| // | |
| // Always clear the dest in case one of the blitters accesses it | |
| // TODO: switch the allocation of tmpDst to call sk_calloc_throw | |
| { | |
| SkBitmap bm; | |
| if (!bm.installPixels(srcInfo, const_cast<void*>(srcPixels), srcRB, ctable, NULL, NULL)) { | |
| return false; | |
| } | |
| SkAutoTUnref<SkCanvas> canvas(SkCanvas::NewRasterDirect(dstInfo, dstPixels, dstRB)); | |
| if (NULL == canvas.get()) { | |
| return false; | |
| } | |
| SkPaint paint; | |
| paint.setDither(true); | |
| canvas->clear(0); | |
| canvas->drawBitmap(bm, 0, 0, &paint); | |
| return true; | |
| } | |
| } | |