From 5f4c5d127fcf8f1ce98246ee9743afced214ada6 Mon Sep 17 00:00:00 2001 From: morkt Date: Fri, 20 May 2016 04:01:40 +0400 Subject: [PATCH] implemented WEBP images. --- ArcFormats/WebP/Alpha.cs | 197 +++ ArcFormats/WebP/Decoder.cs | 2920 ++++++++++++++++++++++++++++++++++ ArcFormats/WebP/Filters.cs | 238 +++ ArcFormats/WebP/Huffman.cs | 272 ++++ ArcFormats/WebP/ImageWEBP.cs | 145 ++ ArcFormats/WebP/Lossless.cs | 1754 ++++++++++++++++++++ 6 files changed, 5526 insertions(+) create mode 100644 ArcFormats/WebP/Alpha.cs create mode 100644 ArcFormats/WebP/Decoder.cs create mode 100644 ArcFormats/WebP/Filters.cs create mode 100644 ArcFormats/WebP/Huffman.cs create mode 100644 ArcFormats/WebP/ImageWEBP.cs create mode 100644 ArcFormats/WebP/Lossless.cs diff --git a/ArcFormats/WebP/Alpha.cs b/ArcFormats/WebP/Alpha.cs new file mode 100644 index 00000000..83a7fcdc --- /dev/null +++ b/ArcFormats/WebP/Alpha.cs @@ -0,0 +1,197 @@ +//! \file Alpha.cs +//! \date Wed May 18 20:06:15 2016 +//! \brief Google WEBP alpha channel processing functions. +/* +Copyright (c) 2010, Google Inc. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + + * Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in + the documentation and/or other materials provided with the + distribution. + + * Neither the name of Google nor the names of its contributors may + be used to endorse or promote products derived from this software + without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +*/ +// +// C# port by morkt (C) 2016 +// + +using System; + +namespace GameRes.Formats.Google +{ + internal class AlphaDecoder + { + public int width_; + public int height_; + public int method_; + public int filter_; + public int pre_processing_; + public LosslessDecoder vp8l_dec_; + public VP8Io m_io; + public bool use_8b_decode_; + + public const int HeaderLen = 1; + public const int NoCompression = 0; + public const int LosslessCompression = 1; + public const int PreprocessedLevels = 1; + + public bool DecodeComplete { get; private set; } + + delegate void FilterFunc (byte[] input, int src, int width, int height, + int stride, byte[] output, int dst); + delegate void UnfilterFunc (int width, int height, int stride, int row, + int num_rows, byte[] data, int dst); + + FilterFunc[] Filters = new FilterFunc[WebpFilter.Last]; + UnfilterFunc[] Unfilters = new UnfilterFunc[WebpFilter.Last]; + + public bool Init (byte[] data, VP8Io src_io, byte[] output) + { + m_io = new VP8Io(); + int alpha_data = HeaderLen; + int alpha_data_size = data.Length - HeaderLen; + + width_ = src_io.width; + height_ = src_io.height; + + if (data.Length <= HeaderLen) + return false; + + method_ = (data[0] >> 0) & 3; + filter_ = (data[0] >> 2) & 3; + pre_processing_ = (data[0] >> 4) & 3; + int rsrv = (data[0] >> 6) & 3; + if (method_ < NoCompression + || method_ > LosslessCompression + || filter_ >= WebpFilter.Last + || pre_processing_ > PreprocessedLevels + || rsrv != 0) + { + return false; + } + + bool ok = false; + if (NoCompression == method_) + { + int alpha_decoded_size = width_ * height_; + ok = (alpha_data_size >= alpha_decoded_size); + } + else + { + ok = DecodeAlphaHeader (data, alpha_data, alpha_data_size, output); + } + FiltersInit(); + + // Copy the necessary parameters from src_io to io +// m_io.Init(); + m_io.opaque = output; // output plane + m_io.width = src_io.width; + m_io.height = src_io.height; + + return ok; + } + + void FiltersInit () + { + Unfilters[WebpFilter.None] = null; + Unfilters[WebpFilter.Horizontal] = WebpFilter.HorizontalUnfilter; + Unfilters[WebpFilter.Vertical] = WebpFilter.VerticalUnfilter; + Unfilters[WebpFilter.Gradient] = WebpFilter.GradientUnfilter; + + Filters[WebpFilter.None] = null; + Filters[WebpFilter.Horizontal] = WebpFilter.HorizontalFilter; + Filters[WebpFilter.Vertical] = WebpFilter.VerticalFilter; + Filters[WebpFilter.Gradient] = WebpFilter.GradientFilter; + } + + /// + // Decodes, unfilters and dequantizes *at least* 'num_rows' rows of alpha starting from row number + // 'row'. It assumes that rows up to (row - 1) have already been decoded. + // Returns false in case of bitstream error. + /// + public bool Decode (byte[] alpha_data, byte[] alpha_plane, int row, int num_rows) + { + var unfilter_func = Unfilters[filter_]; + if (AlphaDecoder.NoCompression == method_) + { + int offset = row * width_; + int num_pixels = num_rows * width_; + Buffer.BlockCopy (alpha_data, AlphaDecoder.HeaderLen + offset, alpha_plane, offset, num_pixels); + } + else // alph_dec_->method_ == ALPHA_LOSSLESS_COMPRESSION + { + if (!DecodeAlphaImageStream (row + num_rows)) + return false; + } + + if (unfilter_func != null) + unfilter_func (width_, height_, width_, row, num_rows, alpha_plane, 0); + +// if (row + num_rows >= alph_dec_.m_io.crop_bottom) + if (row + num_rows >= height_) + DecodeComplete = true; + + return true; + } + + bool DecodeAlphaHeader (byte[] data, int data_i, int data_size, byte[] output) + { + vp8l_dec_ = new LosslessDecoder(); + vp8l_dec_.Init (width_, height_, m_io, data, data_i, data_size, output); + + uint[] decoded = null; + if (!vp8l_dec_.DecodeImageStream (width_, height_, true, ref decoded, false)) + return false; + + // Special case: if alpha data uses only the color indexing transform and + // doesn't use color cache (a frequent case), we will use DecodeAlphaData() + // method that only needs allocation of 1 byte per pixel (alpha channel). + if (vp8l_dec_.next_transform_ == 1 + && vp8l_dec_.transforms_[0].type_ == VP8LImageTransformType.ColorIndexing + && vp8l_dec_.Is8bOptimizable()) + { + use_8b_decode_ = true; + vp8l_dec_.AllocateInternalBuffers8b(); + } + else + { + // Allocate internal buffers (note that dec->width_ may have changed here). + use_8b_decode_ = false; + vp8l_dec_.AllocateInternalBuffers32b (width_); + } + return true; + } + + public bool DecodeAlphaImageStream (int last_row) + { + if (vp8l_dec_.last_pixel_ == vp8l_dec_.Width * vp8l_dec_.Height) + return true; // done + + // Decode (with special row processing). + return use_8b_decode_ ? + vp8l_dec_.DecodeAlphaData (vp8l_dec_.Width, vp8l_dec_.Height, last_row) : + vp8l_dec_.DecodeImageData (vp8l_dec_.Pixels, vp8l_dec_.Width, vp8l_dec_.Height, last_row, LosslessDecoder.ExtractAlphaRows); + } + } +} diff --git a/ArcFormats/WebP/Decoder.cs b/ArcFormats/WebP/Decoder.cs new file mode 100644 index 00000000..ab8dcdae --- /dev/null +++ b/ArcFormats/WebP/Decoder.cs @@ -0,0 +1,2920 @@ +//! \file Decoder.cs +//! \date Mon Apr 11 02:53:27 2016 +//! \brief Google WEBP decoder implementation. +/* +Copyright (c) 2010, Google Inc. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + + * Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in + the documentation and/or other materials provided with the + distribution. + + * Neither the name of Google nor the names of its contributors may + be used to endorse or promote products derived from this software + without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +*/ +// +// C# port by morkt (C) 2016 +// + +using System; +using System.Diagnostics; +using System.IO; +using System.Windows.Media; +using GameRes.Utility; + +namespace GameRes.Formats.Google +{ + internal sealed class WebPDecoder : IDisposable + { + BinaryReader m_input; + byte[] m_output; + byte[] m_alpha_data; // compressed alpha data (if present) + byte[] m_alpha_plane; // output. Persistent, contains the whole data. + WebPMetaData m_info; + int m_stride; + VP8Io m_io; + + public PixelFormat Format { get; private set; } + public byte[] Output { get { return m_output; } } + public int Stride { get { return m_stride; } } + public byte[] Cache { get { return m_cache; } } + public byte[] AlphaPlane { get { return m_alpha_plane; } } + + public WebPDecoder (Stream input, WebPMetaData info) + { + m_input = new ArcView.Reader (input); + m_info = info; + m_stride = (int)info.Width * 4; + m_output = new byte[m_stride * (int)info.Height]; + m_io = new VP8Io(); + if (0 != m_info.AlphaOffset) + { + m_input.BaseStream.Position = m_info.AlphaOffset; + m_alpha_data = m_input.ReadBytes (m_info.AlphaSize); + m_alpha_plane = new byte[info.Width * info.Height]; + Format = PixelFormats.Bgra32; + } + else + { + Format = PixelFormats.Bgr32; + } + } + + // Macroblock to process/filter, depending on cropping and filter_type. + int tl_mb_x, tl_mb_y; // top-left MB that must be in-loop filtered + int br_mb_x, br_mb_y; // last bottom-right MB that must be decoded + int mb_x_, mb_y_; // current position, in macroblock units + + int m_num_parts = 1; + BitReader m_br; + + public void Decode () + { + if (m_info.IsLossless) + throw new NotImplementedException ("Lossless WebP not implemented"); + m_input.BaseStream.Position = m_info.DataOffset; + GetHeaders(); + EnterCritical(); + InitFrame(); + ParseFrame(); + } + + int ReadInt24 () + { + int v = m_input.ReadByte(); + v |= m_input.ReadByte() << 8; + v |= m_input.ReadByte() << 16; + return v; + } + + internal class FrameHeader + { + public bool KeyFrame; + public int Profile; + public bool Show; + public int PartitionLength; + } + + int mb_w_, mb_h_; + + byte[] m_cache; + internal int cache_y_; + internal int cache_u_; + internal int cache_v_; + internal int cache_y_stride_; + internal int cache_uv_stride_; + + int m_filter_type; + bool m_use_skip_proba; + int m_skip_p; + + PictureHeader m_pic_hdr = new PictureHeader(); + FrameHeader m_frame_header = new FrameHeader(); + + SegmentHeader m_segment_hdr = new SegmentHeader(); + FilterHeader m_filter_hdr = new FilterHeader(); + Proba m_proba = new Proba(); + + BitReader[] m_parts = new BitReader[MaxNumPartitions]; + QuantMatrix[] m_dqm = new QuantMatrix[NumMbSegments]; + + void EnterCritical () + { + // Define the area where we can skip in-loop filtering, in case of cropping. + // + // 'Simple' filter reads two luma samples outside of the macroblock + // and filters one. It doesn't filter the chroma samples. Hence, we can + // avoid doing the in-loop filtering before crop_top/crop_left position. + // For the 'Complex' filter, 3 samples are read and up to 3 are filtered. + // Means: there's a dependency chain that goes all the way up to the + // top-left corner of the picture (MB #0). We must filter all the previous + // macroblocks. + int extra_pixels = kFilterExtraRows[m_filter_type]; + tl_mb_x = 0; + tl_mb_y = 0; + + // We need some 'extra' pixels on the right/bottom. + br_mb_y = Math.Min (mb_h_, (m_io.height + 15 + extra_pixels) >> 4); + br_mb_x = Math.Min (mb_w_, (m_io.width + 15 + extra_pixels) >> 4); + PrecomputeFilterStrengths(); + } + + void PrecomputeFilterStrengths () + { + if (m_filter_type <= 0) + return; + for (int s = 0; s < NumMbSegments; ++s) + { + // First, compute the initial level + int base_level; + if (m_segment_hdr.UseSegment) + { + base_level = m_segment_hdr.FilterStrength[s]; + if (!m_segment_hdr.AbsoluteDelta) + base_level += m_filter_hdr.Level; + } + else + { + base_level = m_filter_hdr.Level; + } + for (int i4x4 = 0; i4x4 <= 1; ++i4x4) + { + int level = base_level; + if (m_filter_hdr.UseLfDelta) + { + level += m_filter_hdr.RefLfDelta[0]; + if (i4x4 > 0) + level += m_filter_hdr.ModeLfDelta[0]; + } + level = (level < 0) ? 0 : (level > 63) ? 63 : level; + if (level > 0) + { + int ilevel = level; + if (m_filter_hdr.Sharpness > 0) + { + if (m_filter_hdr.Sharpness > 4) + ilevel >>= 2; + else + ilevel >>= 1; + if (ilevel > 9 - m_filter_hdr.Sharpness) + ilevel = 9 - m_filter_hdr.Sharpness; + } + if (ilevel < 1) ilevel = 1; + m_fstrengths[s,i4x4].f_ilevel_ = (byte)ilevel; + m_fstrengths[s,i4x4].f_limit_ = (byte)(2 * level + ilevel); + m_fstrengths[s,i4x4].hev_thresh_ = (byte)((level >= 40) ? 2 : (level >= 15) ? 1 : 0); + } + else + { + m_fstrengths[s,i4x4].f_limit_ = 0; // no filtering + } + m_fstrengths[s,i4x4].f_inner_ = (byte)i4x4; + } + } + } + + void GetHeaders () + { + int chunk_size = m_info.DataSize; + + int bits = ReadInt24(); + chunk_size -= 3; + m_frame_header.KeyFrame = 0 == (bits & 1); + m_frame_header.Profile = (bits >> 1) & 7; + m_frame_header.Show = 0 != (bits & (1u << 4)); + m_frame_header.PartitionLength = bits >> 5; + if (m_frame_header.Profile > 3) + throw new InvalidFormatException ("Incorrect keyframe parameters."); + if (!m_frame_header.Show) + throw new InvalidFormatException ("Frame not displayable."); + + if (m_frame_header.KeyFrame) + { + // Paragraph 9.2 + if (!CheckSignature()) + throw new InvalidFormatException ("Bad code word"); + ushort w = m_input.ReadUInt16(); + ushort h = m_input.ReadUInt16(); + m_pic_hdr.Width = (ushort)(w & 0x3fff); + m_pic_hdr.XScale = (byte)(w >> 14); // ratio: 1, 5/4 5/3 or 2 + m_pic_hdr.Height = (ushort)(h & 0x3fff); + m_pic_hdr.YScale = (byte)(h >> 14); + chunk_size -= 7; + + mb_w_ = (m_pic_hdr.Width + 15) >> 4; + mb_h_ = (m_pic_hdr.Height + 15) >> 4; + + // Setup default output area (can be later modified during m_io.setup()) + m_io.width = m_pic_hdr.Width; + m_io.height = m_pic_hdr.Height; + + m_io.mb_w = m_io.width; + m_io.mb_h = m_io.height; + + m_proba.Reset(); + m_segment_hdr.Reset(); + } + + if (m_frame_header.PartitionLength > chunk_size) + throw new InvalidFormatException ("bad partition length"); + + m_br = new BitReader (m_input, m_frame_header.PartitionLength); + chunk_size -= m_frame_header.PartitionLength; + + if (m_frame_header.KeyFrame) + { + m_pic_hdr.Colorspace = (byte)m_br.GetNextBit(); + m_pic_hdr.ClampType = (byte)m_br.GetNextBit(); + } + if (!ParseSegmentHeader (m_br)) + throw new InvalidFormatException ("Cannot parse segment header"); + // Filter specs + if (!ParseFilterHeader (m_br)) + throw new InvalidFormatException ("Cannot parse filter header"); + + if (!ParsePartitions (m_br, chunk_size)) + throw new InvalidFormatException ("Cannot parse partitions"); + + for (int i = 0; i < m_dqm.Length; ++i) + m_dqm[i] = new QuantMatrix(); + + // quantizer change + ParseQuant (m_br); + + // Frame buffer marking + if (!m_frame_header.KeyFrame) + throw new InvalidFormatException ("Not a key frame"); + + m_br.GetNextBit(); // ignore the value of update_proba_ + ParseProba (m_br); + } + + bool CheckSignature () + { + if (m_input.ReadByte() != 0x9D) + return false; + if (m_input.ReadByte() != 1) + return false; + return m_input.ReadByte() == 0x2A; + } + + bool ParseSegmentHeader (BitReader br) + { + m_segment_hdr.UseSegment = br.GetNextBit() != 0; + if (m_segment_hdr.UseSegment) + { + m_segment_hdr.UpdateMap = br.GetNextBit() != 0; + if (0 != br.GetNextBit()) + { // update data + m_segment_hdr.AbsoluteDelta = br.GetNextBit() != 0; + for (int s = 0; s < NumMbSegments; ++s) + { + m_segment_hdr.Quantizer[s] = (byte)(br.GetNextBit() != 0 ? br.GetSignedValue (7) : 0); + } + for (int s = 0; s < NumMbSegments; ++s) + { + m_segment_hdr.FilterStrength[s] = (byte)(br.GetNextBit() != 0 ? br.GetSignedValue (6) : 0); + } + } + if (m_segment_hdr.UpdateMap) + { + for (int s = 0; s < TreeProbs; ++s) + { + m_proba.Segments[s] = (byte)(br.GetNextBit() != 0 ? br.GetBits (8) : 255); + } + } + } + else + { + m_segment_hdr.UpdateMap = false; + } + return !br.Eof; + } + + bool ParseFilterHeader (BitReader br) + { + m_filter_hdr.Simple = br.GetNextBit() != 0; + m_filter_hdr.Level = br.GetBits (6); + m_filter_hdr.Sharpness = br.GetBits (3); + m_filter_hdr.UseLfDelta = br.GetNextBit() != 0; + if (m_filter_hdr.UseLfDelta) + { + if (0 != br.GetNextBit()) + { // update lf-delta? + for (int i = 0; i < NumRefLfDeltas; ++i) + { + if (0 != br.GetNextBit()) + m_filter_hdr.RefLfDelta[i] = br.GetSignedValue (6); + } + for (int i = 0; i < NumModeLfDeltas; ++i) + { + if (0 != br.GetNextBit()) + m_filter_hdr.ModeLfDelta[i] = br.GetSignedValue (6); + } + } + } + m_filter_type = (0 == m_filter_hdr.Level) ? 0 : m_filter_hdr.Simple ? 1 : 2; + return !br.Eof; + } + + bool ParsePartitions (BitReader br, int size) + { + long part_end = m_input.BaseStream.Position + size; + int size_left = size; + m_num_parts = 1 << br.GetBits (2); + int last_part = m_num_parts - 1; + if (size < 3 * last_part) + return false; + long part_start = m_input.BaseStream.Position + last_part * 3; + size_left -= last_part * 3; + for (int p = 0; p < last_part; ++p) + { + int psize = ReadInt24(); + var sz_pos = m_input.BaseStream.Position; + if (psize > size_left) psize = size_left; + m_input.BaseStream.Position = part_start; + m_parts[p] = new BitReader (m_input, psize); + part_start += psize; + size_left -= psize; + m_input.BaseStream.Position = sz_pos; + } + m_input.BaseStream.Position = part_start; + m_parts[last_part] = new BitReader (m_input, size_left); + return part_start < part_end; + } + + static int Clip (int v, int M) + { + return v < 0 ? 0 : v > M ? M : v; + } + + void ParseQuant (BitReader br) + { + int base_q0 = br.GetBits (7); + int dqy1_dc = br.GetNextBit() != 0 ? br.GetSignedValue (4) : 0; + int dqy2_dc = br.GetNextBit() != 0 ? br.GetSignedValue (4) : 0; + int dqy2_ac = br.GetNextBit() != 0 ? br.GetSignedValue (4) : 0; + int dquv_dc = br.GetNextBit() != 0 ? br.GetSignedValue (4) : 0; + int dquv_ac = br.GetNextBit() != 0 ? br.GetSignedValue (4) : 0; + + for (int i = 0; i < NumMbSegments; ++i) + { + int q; + if (m_segment_hdr.UseSegment) + { + q = m_segment_hdr.Quantizer[i]; + if (!m_segment_hdr.AbsoluteDelta) + q += base_q0; + } + else if (i > 0) + { + m_dqm[i] = m_dqm[0]; + continue; + } + else + { + q = base_q0; + } + var m = m_dqm[i]; + m.y1_mat[0] = kDcTable[Clip (q + dqy1_dc, 127)]; + m.y1_mat[1] = kAcTable[Clip (q + 0, 127)]; + + m.y2_mat[0] = kDcTable[Clip (q + dqy2_dc, 127)] * 2; + // For all x in [0..284], x*155/100 is bitwise equal to (x*101581) >> 16. + // The smallest precision for that is '(x*6349) >> 12' but 16 is a good + // word size. + m.y2_mat[1] = (kAcTable[Clip (q + dqy2_ac, 127)] * 101581) >> 16; + if (m.y2_mat[1] < 8) m.y2_mat[1] = 8; + + m.uv_mat[0] = kDcTable[Clip (q + dquv_dc, 117)]; + m.uv_mat[1] = kAcTable[Clip (q + dquv_ac, 127)]; + + m.uv_quant = q + dquv_ac; // for dithering strength evaluation + } + } + + void ParseProba (BitReader br) + { + for (int t = 0; t < NumTypes; ++t) + { + for (int b = 0; b < NumBands; ++b) + for (int c = 0; c < NumCtx; ++c) + for (int p = 0; p < NumProbas; ++p) + { + int v = br.GetBit (CoeffsUpdateProba[t,b,c,p]) != 0 ? + br.GetBits (8) : CoeffsProba0[t,b,c,p]; + m_proba.Bands[t,b].Probas[c][p] = (byte)v; + } + for (int b = 0; b < 16 + 1; ++b) + { + m_proba.BandsPtr[t][b] = m_proba.Bands[t,kBands[b]]; + } + } + m_use_skip_proba = br.GetNextBit() != 0; + if (m_use_skip_proba) + m_skip_p = br.GetBits (8); + } + + void InitFrame () + { + AllocateMemory(); + m_io.Init (this); + DspInit(); + } + + void ParseFrame () + { + for (mb_y_ = 0; mb_y_ < br_mb_y; ++mb_y_) + { + // Parse bitstream for this row. + var token_br = m_parts[mb_y_ & (m_num_parts - 1)]; + if (!ParseIntraModeRow (m_br)) + throw new InvalidFormatException ("Premature end-of-partition0 encountered"); + for (; mb_x_ < mb_w_; ++mb_x_) + { + if (!DecodeMB (token_br)) + throw new InvalidFormatException ("Premature end-of-file encountered"); + } + InitScanline(); // Prepare for next scanline + + // Reconstruct, filter and emit the row. + if (!ProcessRow()) + throw new InvalidFormatException ("Output aborted"); + } + } + + bool ParseIntraModeRow (BitReader br) + { + for (int mb_x = 0; mb_x < mb_w_; ++mb_x) + ParseIntraMode (br, mb_x); + return !m_br.Eof; + } + + internal struct FilterInfo + { + public byte f_limit_; // filter limit in [3..189], or 0 if no filtering + public byte f_ilevel_; // inner limit in [1..63] + public byte f_inner_; // do inner filtering? + public byte hev_thresh_; // high edge variance threshold in [0..2] + } + + FilterInfo[] m_filter_info; // filter strength info + FilterInfo[,] m_fstrengths = new FilterInfo[NumMbSegments,2]; + + void ParseIntraMode (BitReader br, int mb_x) + { + int top = 4 * mb_x; // within intra_t + int left = 0; // within intra_l + var block = m_mb_data[mb_x]; + + // Note: we don't save segment map (yet), as we don't expect + // to decode more than 1 keyframe. + if (m_segment_hdr.UpdateMap) + { + // Hardcoded tree parsing + block.segment_ = 0 == br.GetBit (m_proba.Segments[0]) + ? (byte)br.GetBit (m_proba.Segments[1]) + : (byte)(2 + br.GetBit (m_proba.Segments[2])); + } + else + { + block.segment_ = 0; // default for intra + } + if (m_use_skip_proba) + block.skip_ = br.GetBit (m_skip_p) != 0; + + block.is_i4x4_ = 0 == br.GetBit (145); // decide for B_PRED first + if (!block.is_i4x4_) + { + // Hardcoded 16x16 intra-mode decision tree. + int ymode = + 0 != br.GetBit (156) ? (0 != br.GetBit (128) ? TM_PRED : H_PRED) + : (0 != br.GetBit (163) ? V_PRED : DC_PRED); + block.imodes_[0] = (byte)ymode; + for (int i = 0; i < 4; ++i) + { + m_intra_t[top+i] = (byte)ymode; + m_intra_l[left+i] = (byte)ymode; + } + } + else + { + int modes = 0; // within block.imodes_; + for (int y = 0; y < 4; ++y) + { + int ymode = m_intra_l[left + y]; + for (int x = 0; x < 4; ++x) + { + int prob = m_intra_t[top + x]; + int i = kYModesIntra4[br.GetBit (kBModesProba[prob,ymode,0])]; + while (i > 0) + i = kYModesIntra4[2 * i + br.GetBit (kBModesProba[prob,ymode,i])]; + ymode = -i; + m_intra_t[top + x] = (byte)ymode; + } + Buffer.BlockCopy (m_intra_t, top, block.imodes_, modes, 4); + modes += 4; + m_intra_l[left + y] = (byte)ymode; + } + } + // Hardcoded UVMode decision tree + block.uvmode_ = 0 == br.GetBit (142) ? DC_PRED + : 0 == br.GetBit (114) ? V_PRED + : 0 != br.GetBit (183) ? TM_PRED : H_PRED; + } + + bool DecodeMB (BitReader token_br) + { + int left = m_mb_info - 1; + int mb = m_mb_info + mb_x_; + var block = m_mb_data[mb_x_]; + bool skip = m_use_skip_proba && block.skip_; + + if (!skip) + { + skip = ParseResiduals (mb, token_br); + } + else + { + m_mb[left].nz_ = m_mb[mb].nz_ = 0; + if (!block.is_i4x4_) + m_mb[left].nz_dc_ = m_mb[mb].nz_dc_ = 0; + block.non_zero_y_ = 0; + block.non_zero_uv_ = 0; + } + if (m_filter_type > 0) // store filter info + { + int finfo = mb_x_; + m_filter_info[finfo] = m_fstrengths[block.segment_, block.is_i4x4_ ? 1 : 0]; + m_filter_info[finfo].f_inner_ |= (byte)(skip ? 0 : 1); + } + return !token_br.Eof; + } + + bool ParseResiduals (int mb, BitReader token_br) + { + var bands = m_proba.BandsPtr; + BandProbas[] ac_proba; + var block = m_mb_data[mb_x_]; + var q = m_dqm[block.segment_]; + int dst = 0; // block->coeffs_ + int left_mb = m_mb_info - 1; + uint non_zero_y = 0; + uint non_zero_uv = 0; + int x, y; + uint out_t_nz, out_l_nz; + int first; + + for (int i = 0; i < 384; ++i) + block.coeffs_[i] = 0; + if (!block.is_i4x4_) // parse DC + { + var dc = new short[16]; + int ctx = m_mb[mb].nz_dc_ + m_mb[left_mb].nz_dc_; + int nz = GetCoeffs (token_br, bands[1], ctx, q.y2_mat, 0, dc, 0); + m_mb[mb].nz_dc_ = m_mb[left_mb].nz_dc_ = (byte)(nz > 0 ? 1 : 0); + if (nz > 1) + { // more than just the DC -> perform the full transform + TransformWHT (dc, block.coeffs_, dst); + } + else + { // only DC is non-zero -> inlined simplified transform + int dc0 = (dc[0] + 3) >> 3; + for (int i = 0; i < 16 * 16; i += 16) + block.coeffs_[dst+i] = (short)dc0; + } + first = 1; + ac_proba = bands[0]; + } + else + { + first = 0; + ac_proba = bands[3]; + } + + byte tnz = (byte)(m_mb[mb].nz_ & 0xF); + byte lnz = (byte)(m_mb[left_mb].nz_ & 0xF); + for (y = 0; y < 4; ++y) + { + int l = lnz & 1; + uint nz_coeffs = 0; + for (x = 0; x < 4; ++x) + { + int ctx = l + (tnz & 1); + int nz = GetCoeffs (token_br, ac_proba, ctx, q.y1_mat, first, block.coeffs_, dst); + l = nz > first ? 1 : 0; + tnz = (byte)((tnz >> 1) | (l << 7)); + nz_coeffs = NzCodeBits (nz_coeffs, nz, block.coeffs_[dst] != 0 ? 1 : 0); + dst += 16; + } + tnz >>= 4; + lnz = (byte)((lnz >> 1) | (l << 7)); + non_zero_y = (non_zero_y << 8) | nz_coeffs; + } + out_t_nz = tnz; + out_l_nz = (uint)(lnz >> 4); + + for (int ch = 0; ch < 4; ch += 2) + { + uint nz_coeffs = 0; + tnz = (byte)(m_mb[mb].nz_ >> (4 + ch)); + lnz = (byte)(m_mb[left_mb].nz_ >> (4 + ch)); + for (y = 0; y < 2; ++y) + { + int l = lnz & 1; + for (x = 0; x < 2; ++x) + { + int ctx = l + (tnz & 1); + int nz = GetCoeffs (token_br, bands[2], ctx, q.uv_mat, 0, block.coeffs_, dst); + l = nz > 0 ? 1 : 0; + tnz = (byte)((tnz >> 1) | (l << 3)); + nz_coeffs = NzCodeBits (nz_coeffs, nz, block.coeffs_[dst] != 0 ? 1 : 0); + dst += 16; + } + tnz >>= 2; + lnz = (byte)((lnz >> 1) | (l << 5)); + } + // Note: we don't really need the per-4x4 details for U/V blocks. + non_zero_uv |= nz_coeffs << (4 * ch); + out_t_nz |= (uint)(tnz << 4) << ch; + out_l_nz |= (uint)(lnz & 0xf0) << ch; + } + m_mb[mb].nz_ = (byte)out_t_nz; + m_mb[left_mb].nz_ = (byte)out_l_nz; + + block.non_zero_y_ = non_zero_y; + block.non_zero_uv_ = non_zero_uv; + + return 0 == (non_zero_y | non_zero_uv); // will be used for further optimization + } + + static uint NzCodeBits (uint nz_coeffs, int nz, int dc_nz) + { + nz_coeffs <<= 2; + nz_coeffs |= (nz > 3) ? 3u : (nz > 1) ? 2u : (uint)dc_nz; + return nz_coeffs; + } + + int GetCoeffs (BitReader br, BandProbas[] prob, int ctx, int[] dq, int n, short[] out_ptr, int dst) + { + var p = prob[n].Probas[ctx]; + for (; n < 16; ++n) + { + if (0 == br.GetBit (p[0])) + return n; // previous coeff was last non-zero coeff + + while (0 == br.GetBit (p[1])) // sequence of zero coeffs + { + p = prob[++n].Probas[0]; + if (16 == n) return 16; + } + // non zero coeff + var p_ctx = prob[n + 1].Probas; + int v; + if (0 == br.GetBit (p[2])) + { + v = 1; + p = p_ctx[1]; + } + else + { + v = br.GetLargeValue (p); + p = p_ctx[2]; + } + out_ptr[dst+kZigzag[n]] = (short)(br.GetSigned (v) * dq[n > 0 ? 1 : 0]); + } + return 16; + } + + bool m_filter_row; + + bool ProcessRow () + { + m_filter_row = (m_filter_type > 0) && (mb_y_ >= tl_mb_y) && (mb_y_ <= br_mb_y); + ReconstructRow(); + return FinishRow(); + } + + void ReconstructRow () + { + int j; + int cache_id = 0; + int y_dst = Y_OFF; // within m_yuv_b + int u_dst = U_OFF; + int v_dst = V_OFF; + + // Initialize left-most block. + for (j = 0; j < 16; ++j) + { + m_yuv_b[y_dst + j * BPS - 1] = 129; + } + for (j = 0; j < 8; ++j) + { + m_yuv_b[u_dst + j * BPS - 1] = 129; + m_yuv_b[v_dst + j * BPS - 1] = 129; + } + + int mb_y = mb_y_; + // Init top-left sample on left column too. + if (mb_y > 0) + { + m_yuv_b[y_dst - 1 - BPS] = m_yuv_b[u_dst - 1 - BPS] = m_yuv_b[v_dst - 1 - BPS] = 129; + } + else + { + // we only need to do this init once at block (0,0). + // Afterward, it remains valid for the whole topmost row. + for (int i = 0; i < 16+4+1; ++i) + m_yuv_b[y_dst - BPS - 1 + i] = 127; + for (int i = 0; i < 8+1; ++i) + { + m_yuv_b[u_dst - BPS - 1 + i] = 127; + m_yuv_b[v_dst - BPS - 1 + i] = 127; + } + } + + // Reconstruct one row. + for (int mb_x = 0; mb_x < mb_w_; ++mb_x) + { + var block = m_mb_data[mb_x]; + + // Rotate in the left samples from previously decoded block. We move four + // pixels at a time for alignment reason, and because of in-loop filter. + if (mb_x > 0) + { + for (j = -1; j < 16; ++j) + { + Buffer.BlockCopy (m_yuv_b, y_dst + j * BPS + 12, m_yuv_b, y_dst + j * BPS - 4, 4); + } + for (j = -1; j < 8; ++j) + { + Buffer.BlockCopy (m_yuv_b, u_dst + j * BPS + 4, m_yuv_b, u_dst + j * BPS - 4, 4); + Buffer.BlockCopy (m_yuv_b, v_dst + j * BPS + 4, m_yuv_b, v_dst + j * BPS - 4, 4); + } + } + // bring top samples into the cache + var top_yuv = mb_x; // within yuv_t_ + var coeffs = block.coeffs_; + uint bits = block.non_zero_y_; + + if (mb_y > 0) + { + Buffer.BlockCopy (m_yuv_t[top_yuv].y, 0, m_yuv_b, y_dst - BPS, 16); + Buffer.BlockCopy (m_yuv_t[top_yuv].u, 0, m_yuv_b, u_dst - BPS, 8); + Buffer.BlockCopy (m_yuv_t[top_yuv].v, 0, m_yuv_b, v_dst - BPS, 8); + } + + int pred_func; + // predict and add residuals + if (block.is_i4x4_) // 4x4 + { + int top_right = y_dst - BPS + 16; + if (mb_y > 0) + { + if (mb_x >= mb_w_ - 1) // on rightmost border + { + byte v = m_yuv_t[top_yuv].y[15]; + m_yuv_b[top_right] = v; + m_yuv_b[top_right+1] = v; + m_yuv_b[top_right+2] = v; + m_yuv_b[top_right+3] = v; + } + else + { + Buffer.BlockCopy (m_yuv_t[top_yuv+1].y, 0, m_yuv_b, top_right, 4); + } + } + // replicate the top-right pixels below + Buffer.BlockCopy (m_yuv_b, top_right, m_yuv_b, top_right+BPS*4, 4); + Buffer.BlockCopy (m_yuv_b, top_right, m_yuv_b, top_right+BPS*8, 4); + Buffer.BlockCopy (m_yuv_b, top_right, m_yuv_b, top_right+BPS*12, 4); + + // predict and add residuals for all 4x4 blocks in turn. + for (int n = 0; n < 16; ++n, bits <<= 2) + { + int dst = y_dst + kScan[n]; + PredLuma4[block.imodes_[n]] (m_yuv_b, dst); + DoTransform (bits, coeffs, n * 16, m_yuv_b, dst); + } + } + else // 16x16 + { + pred_func = CheckMode (mb_x, mb_y, block.imodes_[0]); + PredLuma16[pred_func] (m_yuv_b, y_dst); + if (bits != 0) + { + for (int n = 0; n < 16; ++n, bits <<= 2) + DoTransform (bits, coeffs, n * 16, m_yuv_b, y_dst + kScan[n]); + } + } + // Chroma + uint bits_uv = block.non_zero_uv_; + pred_func = CheckMode (mb_x, mb_y, block.uvmode_); + PredChroma8[pred_func] (m_yuv_b, u_dst); + PredChroma8[pred_func] (m_yuv_b, v_dst); + DoUVTransform(bits_uv >> 0, coeffs, 16 * 16, m_yuv_b, u_dst); + DoUVTransform(bits_uv >> 8, coeffs, 20 * 16, m_yuv_b, v_dst); + + // stash away top samples for next block + if (mb_y < mb_h_ - 1) + { + Buffer.BlockCopy (m_yuv_b, y_dst + 15 * BPS, m_yuv_t[top_yuv].y, 0, 16); + Buffer.BlockCopy (m_yuv_b, u_dst + 7 * BPS, m_yuv_t[top_yuv].u, 0, 8); + Buffer.BlockCopy (m_yuv_b, v_dst + 7 * BPS, m_yuv_t[top_yuv].v, 0, 8); + } + // Transfer reconstructed samples from yuv_b_ cache to final destination. + int y_offset = cache_id * 16 * cache_y_stride_; + int uv_offset = cache_id * 8 * cache_uv_stride_; + int y_out = cache_y_ + mb_x * 16 + y_offset; + int u_out = cache_u_ + mb_x * 8 + uv_offset; + int v_out = cache_v_ + mb_x * 8 + uv_offset; + for (j = 0; j < 16; ++j) + { + Buffer.BlockCopy (m_yuv_b, y_dst + j * BPS, m_cache, y_out + j * cache_y_stride_, 16); + } + for (j = 0; j < 8; ++j) + { + Buffer.BlockCopy (m_yuv_b, u_dst + j * BPS, m_cache, u_out + j * cache_uv_stride_, 8); + Buffer.BlockCopy (m_yuv_b, v_dst + j * BPS, m_cache, v_out + j * cache_uv_stride_, 8); + } + } + } + + static int CheckMode (int mb_x, int mb_y, int mode) + { + if (B_DC_PRED == mode) + { + if (0 == mb_x) + return (0 == mb_y) ? B_DC_PRED_NOTOPLEFT : B_DC_PRED_NOLEFT; + else + return (0 == mb_y) ? B_DC_PRED_NOTOP : B_DC_PRED; + } + return mode; + } + + void DoTransform (uint bits, short[] src, int src_i, byte[] dst, int dst_i) + { + switch (bits >> 30) + { + case 3: + TransformTwo (src, src_i, dst, dst_i, false); + break; + case 2: + TransformAC3 (src, src_i, dst, dst_i); + break; + case 1: + TransformDC (src, src_i, dst, dst_i); + break; + default: + break; + } + } + + void DoUVTransform (uint bits, short[] src, int src_i, byte[] dst, int dst_i) + { + if (0 != (bits & 0xFF)) // any non-zero coeff at all? + { + if (0 != (bits & 0xAA)) // any non-zero AC coefficient? + TransformUV (src, src_i, dst, dst_i); // note we don't use the AC3 variant for U/V + else + TransformDCUV (src, src_i, dst, dst_i); + } + } + + static void TransformOne(short[] src, int src_i, byte[] dst, int dst_i) + { + var C = new int[4*4]; + var tmp = 0; + for (int i = 0; i < 4; ++i) // vertical pass + { + int a = src[src_i] + src[src_i+8]; // [-4096, 4094] + int b = src[src_i] - src[src_i+8]; // [-4095, 4095] + int c = MUL2(src[src_i+4]) - MUL1(src[src_i+12]); // [-3783, 3783] + int d = MUL1(src[src_i+4]) + MUL2(src[src_i+12]); // [-3785, 3781] + C[tmp+0] = a + d; // [-7881, 7875] + C[tmp+1] = b + c; // [-7878, 7878] + C[tmp+2] = b - c; // [-7878, 7878] + C[tmp+3] = a - d; // [-7877, 7879] + tmp += 4; + src_i++; + } + // Each pass is expanding the dynamic range by ~3.85 (upper bound). + // The exact value is (2. + (20091 + 35468) / 65536). + // After the second pass, maximum interval is [-3794, 3794], assuming + // an input in [-2048, 2047] interval. We then need to add a dst value + // in the [0, 255] range. + // In the worst case scenario, the input to clip_8b() can be as large as + // [-60713, 60968]. + tmp = 0; + for (int i = 0; i < 4; ++i) // horizontal pass + { + int dc = C[tmp] + 4; + int a = dc + C[tmp+8]; + int b = dc - C[tmp+8]; + int c = MUL2(C[tmp+4]) - MUL1(C[tmp+12]); + int d = MUL1(C[tmp+4]) + MUL2(C[tmp+12]); + dst[dst_i ] = clip_8b (dst[dst_i ] + ((a + d) >> 3)); + dst[dst_i+1] = clip_8b (dst[dst_i+1] + ((b + c) >> 3)); + dst[dst_i+2] = clip_8b (dst[dst_i+2] + ((b - c) >> 3)); + dst[dst_i+3] = clip_8b (dst[dst_i+3] + ((a - d) >> 3)); + tmp++; + dst_i += BPS; + } + } + + static byte clip_8b (int v) + { + return (byte)((0 == (v & ~0xFF)) ? v : (v < 0) ? 0 : 255); + } + + static int MUL1 (int a) + { + return ((a * 20091) >> 16) + a; + } + + static int MUL2 (int a) + { + return (a * 35468) >> 16; + } + + static void TransformTwo (short[] src, int src_i, byte[] dst, int dst_i, bool do_two) + { + TransformOne (src, src_i, dst, dst_i); + if (do_two) + TransformOne (src, src_i+16, dst, dst_i+4); + } + + static void TransformAC3 (short[] src, int src_i, byte[] dst, int dst_i) + { + int a = src[src_i] + 4; + int c4 = MUL2(src[src_i+4]); + int d4 = MUL1(src[src_i+4]); + int c1 = MUL2(src[src_i+1]); + int d1 = MUL1(src[src_i+1]); + + // STORE2(0, a + d4, d1, c1); + int DC = a + d4; + dst[dst_i ] = clip_8b (dst[dst_i ] + ((DC + d1) >> 3)); + dst[dst_i+1] = clip_8b (dst[dst_i+1] + ((DC + c1) >> 3)); + dst[dst_i+2] = clip_8b (dst[dst_i+2] + ((DC - c1) >> 3)); + dst[dst_i+3] = clip_8b (dst[dst_i+3] + ((DC - d1) >> 3)); + + // STORE2(1, a + c4, d1, c1); + DC = a + c4; + dst[dst_i+BPS ] = clip_8b (dst[dst_i+BPS ] + ((DC + d1) >> 3)); + dst[dst_i+BPS+1] = clip_8b (dst[dst_i+BPS+1] + ((DC + c1) >> 3)); + dst[dst_i+BPS+2] = clip_8b (dst[dst_i+BPS+2] + ((DC - c1) >> 3)); + dst[dst_i+BPS+3] = clip_8b (dst[dst_i+BPS+3] + ((DC - d1) >> 3)); + + // STORE2(2, a - c4, d1, c1); + DC = a - c4; + dst[dst_i+BPS*2 ] = clip_8b (dst[dst_i+BPS*2 ] + ((DC + d1) >> 3)); + dst[dst_i+BPS*2+1] = clip_8b (dst[dst_i+BPS*2+1] + ((DC + c1) >> 3)); + dst[dst_i+BPS*2+2] = clip_8b (dst[dst_i+BPS*2+2] + ((DC - c1) >> 3)); + dst[dst_i+BPS*2+3] = clip_8b (dst[dst_i+BPS*2+3] + ((DC - d1) >> 3)); + + // STORE2(3, a - d4, d1, c1); + DC = a - d4; + dst[dst_i+BPS*3 ] = clip_8b (dst[dst_i+BPS*3 ] + ((DC + d1) >> 3)); + dst[dst_i+BPS*3+1] = clip_8b (dst[dst_i+BPS*3+1] + ((DC + c1) >> 3)); + dst[dst_i+BPS*3+2] = clip_8b (dst[dst_i+BPS*3+2] + ((DC - c1) >> 3)); + dst[dst_i+BPS*3+3] = clip_8b (dst[dst_i+BPS*3+3] + ((DC - d1) >> 3)); + } + + static void TransformUV (short[] src, int src_i, byte[] dst, int dst_i) + { + TransformTwo (src, src_i, dst, dst_i, true); + TransformTwo (src, src_i + 32, dst, dst_i + 4 * BPS, true); + } + + static void TransformDC (short[] src, int src_i, byte[] dst, int dst_i) + { + int DC = src[src_i] + 4; + for (int j = 0; j < 4; ++j) + for (int i = 0; i < 4; ++i) + { + int pos = dst_i + i + BPS * j; + dst[pos] = clip_8b (dst[pos] + (DC >> 3)); + } + } + + static void TransformDCUV (short[] src, int src_i, byte[] dst, int dst_i) + { + if (0 != src[src_i]) TransformDC (src, src_i, dst, dst_i); + if (0 != src[src_i+16]) TransformDC (src, src_i+16, dst, dst_i+4); + if (0 != src[src_i+2*16]) TransformDC (src, src_i+2*16, dst, dst_i+4*BPS); + if (0 != src[src_i+3*16]) TransformDC (src, src_i+3*16, dst, dst_i+4*BPS+4); + } + + bool FinishRow () + { + bool ok = true; + int cache_id = 0; + int extra_y_rows = kFilterExtraRows[m_filter_type]; + int ysize = extra_y_rows * cache_y_stride_; + int uvsize = (extra_y_rows / 2) * cache_uv_stride_; + int y_offset = cache_id * 16 * cache_y_stride_; + int uv_offset = cache_id * 8 * cache_uv_stride_; + int ydst = cache_y_ - ysize + y_offset; + int udst = cache_u_ - uvsize + uv_offset; + int vdst = cache_v_ - uvsize + uv_offset; + int mb_y = mb_y_; + bool is_first_row = (mb_y == 0); + bool is_last_row = mb_y >= br_mb_y - 1; + + if (m_filter_row) + FilterRow(); + + { + int y_start = mb_y * 16; + int y_end = (mb_y + 1) * 16; + if (!is_first_row) + { + y_start -= extra_y_rows; + m_io.y = ydst; + m_io.u = udst; + m_io.v = vdst; + } + else + { + m_io.y = cache_y_ + y_offset; + m_io.u = cache_u_ + uv_offset; + m_io.v = cache_v_ + uv_offset; + } + + if (!is_last_row) + y_end -= extra_y_rows; + if (y_end > m_io.height) + y_end = m_io.height; // make sure we don't overflow on last row. + if (m_alpha_data != null && y_start < y_end) + { + m_io.alpha_plane = m_alpha_plane; + m_io.a = DecompressAlphaRows (y_start, y_end - y_start); + if (-1 == m_io.a) + throw new InvalidFormatException ("Could not decode alpha data."); + } + if (y_start < 0) + { + int delta_y = -y_start; + y_start = 0; + m_io.y += cache_y_stride_ * delta_y; + m_io.u += cache_uv_stride_ * (delta_y >> 1); + m_io.v += cache_uv_stride_ * (delta_y >> 1); + if (m_io.alpha_plane != null) + m_io.a += m_io.width * delta_y; + } + if (y_start < y_end) + { + m_io.mb_y = y_start; + m_io.mb_w = m_io.width; + m_io.mb_h = y_end - y_start; + ok = m_io.Put (this); + } + } + // rotate top samples if needed + if (cache_id + 1 == num_caches_ && !is_last_row) + { + Buffer.BlockCopy (m_cache, ydst + 16 * cache_y_stride_, m_cache, cache_y_ - ysize, ysize); + Buffer.BlockCopy (m_cache, udst + 8 * cache_uv_stride_, m_cache, cache_u_ - uvsize, uvsize); + Buffer.BlockCopy (m_cache, vdst + 8 * cache_uv_stride_, m_cache, cache_v_ - uvsize, uvsize); + } + return ok; + } + + void FilterRow () + { + int mb_y = mb_y_; + for (int mb_x = tl_mb_x; mb_x < br_mb_x; ++mb_x) + DoFilter (mb_x, mb_y); + } + + void DoFilter (int mb_x, int mb_y) + { + const int cache_id = 0; + int y_bps = cache_y_stride_; + var f_info = m_filter_info[mb_x]; + int y_dst = cache_y_ + cache_id * 16 * y_bps + mb_x * 16; // within m_cache + int ilevel = f_info.f_ilevel_; + int limit = f_info.f_limit_; + if (0 == limit) + return; + if (1 == m_filter_type) // simple + { + if (mb_x > 0) + SimpleHFilter16 (m_cache, y_dst, y_bps, limit + 4); + if (f_info.f_inner_ != 0) + SimpleHFilter16i (m_cache, y_dst, y_bps, limit); + if (mb_y > 0) + SimpleVFilter16 (m_cache, y_dst, y_bps, limit + 4); + if (f_info.f_inner_ != 0) + SimpleVFilter16i (m_cache, y_dst, y_bps, limit); + } + else // complex + { + int uv_bps = cache_uv_stride_; + int u_dst = cache_u_ + cache_id * 8 * uv_bps + mb_x * 8; + int v_dst = cache_v_ + cache_id * 8 * uv_bps + mb_x * 8; + int hev_thresh = f_info.hev_thresh_; + if (mb_x > 0) + { + HFilter16 (m_cache, y_dst, y_bps, limit + 4, ilevel, hev_thresh); + HFilter8 (m_cache, u_dst, v_dst, uv_bps, limit + 4, ilevel, hev_thresh); + } + if (f_info.f_inner_ != 0) + { + HFilter16i (m_cache, y_dst, y_bps, limit, ilevel, hev_thresh); + HFilter8i (m_cache, u_dst, v_dst, uv_bps, limit, ilevel, hev_thresh); + } + if (mb_y > 0) + { + VFilter16 (m_cache, y_dst, y_bps, limit + 4, ilevel, hev_thresh); + VFilter8 (m_cache, u_dst, v_dst, uv_bps, limit + 4, ilevel, hev_thresh); + } + if (f_info.f_inner_ != 0) + { + VFilter16i (m_cache, y_dst, y_bps, limit, ilevel, hev_thresh); + VFilter8i (m_cache, u_dst, v_dst, uv_bps, limit, ilevel, hev_thresh); + } + } + } + + //------------------------------------------------------------------------------ + // Edge filtering functions + + // 4 pixels in, 2 pixels out + static void do_filter2 (byte[] dst, int p, int step) + { + int p1 = dst[p-2*step], p0 = dst[p-step], q0 = dst[p], q1 = dst[p+step]; + int a = 3 * (q0 - p0) + kSClip1[1020 + p1 - q1]; // in [-893,892] + int a1 = kSClip2[112 + ((a + 4) >> 3)]; // in [-16,15] + int a2 = kSClip2[112 + ((a + 3) >> 3)]; + dst[p-step] = kClip1[255 + p0 + a2]; + dst[p ] = kClip1[255 + q0 - a1]; + } + + // 4 pixels in, 4 pixels out + static void do_filter4 (byte[] dst, int p, int step) + { + int p1 = dst[p-2*step], p0 = dst[p-step], q0 = dst[p], q1 = dst[p+step]; + int a = 3 * (q0 - p0); + int a1 = kSClip2[112 + ((a + 4) >> 3)]; + int a2 = kSClip2[112 + ((a + 3) >> 3)]; + int a3 = (a1 + 1) >> 1; + dst[p-2*step] = kClip1[255 + p1 + a3]; + dst[p- step] = kClip1[255 + p0 + a2]; + dst[p ] = kClip1[255 + q0 - a1]; + dst[p+ step] = kClip1[255 + q1 - a3]; + } + + // 6 pixels in, 6 pixels out + static void do_filter6 (byte[] dst, int p, int step) + { + int p2 = dst[p-3*step], p1 = dst[p-2*step], p0 = dst[p-step]; + int q0 = dst[p], q1 = dst[p+step], q2 = dst[p+2*step]; + int a = kSClip1[1020 + 3 * (q0 - p0) + kSClip1[1020 + p1 - q1]]; + // a is in [-128,127], a1 in [-27,27], a2 in [-18,18] and a3 in [-9,9] + int a1 = (27 * a + 63) >> 7; // eq. to ((3 * a + 7) * 9) >> 7 + int a2 = (18 * a + 63) >> 7; // eq. to ((2 * a + 7) * 9) >> 7 + int a3 = (9 * a + 63) >> 7; // eq. to ((1 * a + 7) * 9) >> 7 + dst[p-3*step] = kClip1[255 + p2 + a3]; + dst[p-2*step] = kClip1[255 + p1 + a2]; + dst[p- step] = kClip1[255 + p0 + a1]; + dst[p ] = kClip1[255 + q0 - a1]; + dst[p+ step] = kClip1[255 + q1 - a2]; + dst[p+2*step] = kClip1[255 + q2 - a3]; + } + + static bool hev (byte[] dst, int p, int step, int thresh) + { + int p1 = dst[p-2*step], p0 = dst[p-step], q0 = dst[p], q1 = dst[p+step]; + return (kAbs0[255 + p1 - p0] > thresh) || (kAbs0[255 + q1 - q0] > thresh); + } + + static bool needs_filter (byte[] dst, int p, int step, int t) + { + int p1 = dst[p-2 * step], p0 = dst[p-step], q0 = dst[p], q1 = dst[p+step]; + return ((4 * kAbs0[255 + p0 - q0] + kAbs0[255 + p1 - q1]) <= t); + } + + static bool needs_filter2 (byte[] dst, int p, int step, int t, int it) + { + int p3 = dst[p-4 * step], p2 = dst[p-3 * step], p1 = dst[p-2 * step]; + int p0 = dst[p-step], q0 = dst[p]; + int q1 = dst[p+step], q2 = dst[p+2 * step], q3 = dst[p+3 * step]; + if ((4 * kAbs0[255 + p0 - q0] + kAbs0[255 + p1 - q1]) > t) return false; + return kAbs0[255 + p3 - p2] <= it && kAbs0[255 + p2 - p1] <= it && + kAbs0[255 + p1 - p0] <= it && kAbs0[255 + q3 - q2] <= it && + kAbs0[255 + q2 - q1] <= it && kAbs0[255 + q1 - q0] <= it; + } + + //------------------------------------------------------------------------------ + // Simple In-loop filtering (Paragraph 15.2) + + static void SimpleVFilter16 (byte[] dst, int p, int stride, int thresh) + { + int thresh2 = 2 * thresh + 1; + for (int i = 0; i < 16; ++i) + { + if (needs_filter (dst, p + i, stride, thresh2)) + do_filter2 (dst, p + i, stride); + } + } + + static void SimpleHFilter16 (byte[] dst, int p, int stride, int thresh) + { + int thresh2 = 2 * thresh + 1; + for (int i = 0; i < 16; ++i) + { + if (needs_filter (dst, p + i * stride, 1, thresh2)) + do_filter2 (dst, p + i * stride, 1); + } + } + + static void SimpleVFilter16i (byte[] dst, int p, int stride, int thresh) + { + for (int k = 3; k > 0; --k) + { + p += 4 * stride; + SimpleVFilter16 (dst, p, stride, thresh); + } + } + + static void SimpleHFilter16i (byte[] dst, int p, int stride, int thresh) + { + for (int k = 3; k > 0; --k) + { + p += 4; + SimpleHFilter16 (dst, p, stride, thresh); + } + } + + //------------------------------------------------------------------------------ + // Complex In-loop filtering (Paragraph 15.3) + + static void FilterLoop26(byte[] dst, int p, int hstride, int vstride, int size, int thresh, int ithresh, int hev_thresh) + { + int thresh2 = 2 * thresh + 1; + while (size-- > 0) + { + if (needs_filter2 (dst, p, hstride, thresh2, ithresh)) + { + if (hev (dst, p, hstride, hev_thresh)) + do_filter2 (dst, p, hstride); + else + do_filter6 (dst, p, hstride); + } + p += vstride; + } + } + + static void FilterLoop24 (byte[] dst, int p, int hstride, int vstride, int size, int thresh, int ithresh, int hev_thresh) + { + int thresh2 = 2 * thresh + 1; + while (size-- > 0) + { + if (needs_filter2 (dst, p, hstride, thresh2, ithresh)) + { + if (hev (dst, p, hstride, hev_thresh)) + do_filter2 (dst, p, hstride); + else + do_filter4 (dst, p, hstride); + } + p += vstride; + } + } + + // on macroblock edges + static void VFilter16 (byte[] dst, int p, int stride, int thresh, int ithresh, int hev_thresh) + { + FilterLoop26 (dst, p, stride, 1, 16, thresh, ithresh, hev_thresh); + } + + static void HFilter16 (byte[] dst, int p, int stride, int thresh, int ithresh, int hev_thresh) + { + FilterLoop26 (dst, p, 1, stride, 16, thresh, ithresh, hev_thresh); + } + + // on three inner edges + static void VFilter16i (byte[] dst, int p, int stride, int thresh, int ithresh, int hev_thresh) + { + for (int k = 3; k > 0; --k) + { + p += 4 * stride; + FilterLoop24 (dst, p, stride, 1, 16, thresh, ithresh, hev_thresh); + } + } + + static void HFilter16i (byte[] dst, int p, int stride, int thresh, int ithresh, int hev_thresh) + { + for (int k = 3; k > 0; --k) + { + p += 4; + FilterLoop24 (dst, p, 1, stride, 16, thresh, ithresh, hev_thresh); + } + } + + // 8-pixels wide variant, for chroma filtering + static void VFilter8 (byte[] dst, int u, int v, int stride, int thresh, int ithresh, int hev_thresh) + { + FilterLoop26(dst, u, stride, 1, 8, thresh, ithresh, hev_thresh); + FilterLoop26(dst, v, stride, 1, 8, thresh, ithresh, hev_thresh); + } + + static void HFilter8 (byte[] dst, int u, int v, int stride, int thresh, int ithresh, int hev_thresh) + { + FilterLoop26(dst, u, 1, stride, 8, thresh, ithresh, hev_thresh); + FilterLoop26(dst, v, 1, stride, 8, thresh, ithresh, hev_thresh); + } + + static void VFilter8i (byte[] dst, int u, int v, int stride, int thresh, int ithresh, int hev_thresh) + { + FilterLoop24(dst, u + 4 * stride, stride, 1, 8, thresh, ithresh, hev_thresh); + FilterLoop24(dst, v + 4 * stride, stride, 1, 8, thresh, ithresh, hev_thresh); + } + + static void HFilter8i (byte[] dst, int u, int v, int stride, int thresh, int ithresh, int hev_thresh) + { + FilterLoop24(dst, u + 4, 1, stride, 8, thresh, ithresh, hev_thresh); + FilterLoop24(dst, v + 4, 1, stride, 8, thresh, ithresh, hev_thresh); + } + + //------------------------------------------------------------------------------ + + TopSamples[] m_yuv_t; + byte[] m_yuv_b; + byte[] m_intra_t; + byte[] m_intra_l = new byte[4]; + MacroBlock[] m_mb; + int m_mb_info; + MacroBlockData[] m_mb_data; // parsed reconstruction data + const int num_caches_ = 1; + + void AllocateMemory () + { + int top_size = 32 * mb_w_; + int intra_pred_mode_size = 4 * mb_w_; + m_intra_t = new byte[intra_pred_mode_size]; + + m_yuv_t = new TopSamples[mb_w_]; + for (int i = 0; i < m_yuv_t.Length; ++i) + m_yuv_t[i] = new TopSamples(); + + int mb_info_size = mb_w_ + 1; + if (null == m_mb || m_mb.Length < mb_info_size) + m_mb = new MacroBlock[mb_info_size]; + m_mb_info = 1; + + if (m_filter_type > 0) + m_filter_info = new FilterInfo[mb_w_]; + + m_yuv_b = new byte[YUV_SIZE]; + + m_mb_data = new MacroBlockData[mb_w_]; + for (int i = 0; i < m_mb_data.Length; ++i) + m_mb_data[i] = new MacroBlockData(); + + int cache_height = (16 * num_caches_ + kFilterExtraRows[m_filter_type]) * 3 / 2; + int cache_size = top_size * cache_height; + m_cache = new byte[cache_size]; + + cache_y_stride_ = 16 * mb_w_; + cache_uv_stride_ = 8 * mb_w_; + int extra_rows = kFilterExtraRows[m_filter_type]; + int extra_y = extra_rows * cache_y_stride_; + int extra_uv = (extra_rows / 2) * cache_uv_stride_; + cache_y_ = extra_y; + cache_u_ = cache_y_ + 16 * num_caches_ * cache_y_stride_ + extra_uv; + cache_v_ = cache_u_ + 8 * num_caches_ * cache_uv_stride_ + extra_uv; + + InitScanline(); + } + + void InitScanline () + { + int left = m_mb_info - 1; + m_mb[left].nz_ = 0; + m_mb[left].nz_dc_ = 0; + for (int i = 0; i < 4; ++i) + m_intra_l[i] = 0; + mb_x_ = 0; + } + + void DspInit () + { + InitClipTables(); + + PredLuma4[0] = DC4; + PredLuma4[1] = TM4; + PredLuma4[2] = VE4; + PredLuma4[3] = HE4; + PredLuma4[4] = RD4; + PredLuma4[5] = VR4; + PredLuma4[6] = LD4; + PredLuma4[7] = VL4; + PredLuma4[8] = HD4; + PredLuma4[9] = HU4; + + PredLuma16[0] = DC16; + PredLuma16[1] = TM16; + PredLuma16[2] = VE16; + PredLuma16[3] = HE16; + PredLuma16[4] = DC16NoTop; + PredLuma16[5] = DC16NoLeft; + PredLuma16[6] = DC16NoTopLeft; + + PredChroma8[0] = DC8uv; + PredChroma8[1] = TM8uv; + PredChroma8[2] = VE8uv; + PredChroma8[3] = HE8uv; + PredChroma8[4] = DC8uvNoTop; + PredChroma8[5] = DC8uvNoLeft; + PredChroma8[6] = DC8uvNoTopLeft; + } + + static byte[] kAbs0; + static sbyte[] kSClip1; + static sbyte[] kSClip2; + static byte[] kClip1; + + static bool s_tables_initialized = false; + + void InitClipTables () + { + if (!s_tables_initialized) + { + var abs0 = new byte[255 + 255 + 1]; + var sclip1 = new sbyte[1020 + 1020 + 1]; + var sclip2 = new sbyte[112 + 112 + 1]; + var clip1 = new byte[255 + 511 + 1]; + int i; + for (i = -255; i <= 255; ++i) + abs0[255 + i] = (byte)((i < 0) ? -i : i); + for (i = -1020; i <= 1020; ++i) + sclip1[1020 + i] = (sbyte)((i < -128) ? -128 : (i > 127) ? 127 : i); + for (i = -112; i <= 112; ++i) + sclip2[112 + i] = (sbyte)((i < -16) ? -16 : (i > 15) ? 15 : i); + for (i = -255; i <= 255 + 255; ++i) + clip1[255 + i] = (byte)((i < 0) ? 0 : (i > 255) ? 255 : i); + kAbs0 = abs0; + kSClip1 = sclip1; + kSClip2 = sclip2; + kClip1 = clip1; + s_tables_initialized = true; + } + } + + static void TransformWHT(short[] src, short[] output, int dst) + { + int[] tmp = new int[16]; + int i; + for (i = 0; i < 4; ++i) + { + int a0 = src[0 + i] + src[12 + i]; + int a1 = src[4 + i] + src[ 8 + i]; + int a2 = src[4 + i] - src[ 8 + i]; + int a3 = src[0 + i] - src[12 + i]; + tmp[0 + i] = a0 + a1; + tmp[8 + i] = a0 - a1; + tmp[4 + i] = a3 + a2; + tmp[12 + i] = a3 - a2; + } + for (i = 0; i < 4; ++i) + { + int dc = tmp[0 + i * 4] + 3; // w/ rounder + int a0 = dc + tmp[3 + i * 4]; + int a1 = tmp[1 + i * 4] + tmp[2 + i * 4]; + int a2 = tmp[1 + i * 4] - tmp[2 + i * 4]; + int a3 = dc - tmp[3 + i * 4]; + output[dst+ 0] = (short)((a0 + a1) >> 3); + output[dst+16] = (short)((a3 + a2) >> 3); + output[dst+32] = (short)((a0 - a1) >> 3); + output[dst+48] = (short)((a3 - a2) >> 3); + dst += 64; + } + } + + //------------------------------------------------------------------------------ + // Intra predictions + + static void TrueMotion (byte[] buf, int dst, int size) + { + int top = dst - BPS; + int clip0 = 255 - buf[top-1]; + for (int y = 0; y < size; ++y) + { + int clip = clip0 + buf[dst-1]; + for (int x = 0; x < size; ++x) + { + buf[dst+x] = kClip1[clip + buf[top+x]]; + } + dst += BPS; + } + } + + static void TM4 (byte[] buf, int dst) { TrueMotion (buf, dst, 4); } + static void TM8uv (byte[] buf, int dst) { TrueMotion (buf, dst, 8); } + static void TM16 (byte[] buf, int dst) { TrueMotion (buf, dst, 16); } + + //------------------------------------------------------------------------------ + // 16x16 + + static void VE16 (byte[] buf, int dst) // vertical + { + for (int j = 0; j < 16; ++j) + Buffer.BlockCopy (buf, dst - BPS, buf, dst + j * BPS, 16); + } + + static void HE16 (byte[] buf, int dst) // horizontal + { + for (int j = 16; j > 0; --j) + { + byte v = buf[dst-1]; + for (int i = 0; i < 16; ++i) + buf[dst+i] = v; + dst += BPS; + } + } + + static void Put16 (int v, byte[] buf, int dst) + { + for (int j = 0; j < 16; ++j) + { + for (int i = 0; i < 16; ++i) + buf[dst+i] = (byte)v; + dst += BPS; + } + } + + static void DC16 (byte[] buf, int dst) // DC + { + int DC = 16; + for (int j = 0; j < 16; ++j) + { + DC += buf[dst - 1 + j * BPS] + buf[dst + j - BPS]; + } + Put16 (DC >> 5, buf, dst); + } + + static void DC16NoTop (byte[] buf, int dst) // DC with top samples not available + { + int DC = 8; + for (int j = 0; j < 16; ++j) + { + DC += buf[dst - 1 + j * BPS]; + } + Put16 (DC >> 4, buf, dst); + } + + static void DC16NoLeft (byte[] buf, int dst) // DC with left samples not available + { + int DC = 8; + for (int i = 0; i < 16; ++i) + { + DC += buf[dst + i - BPS]; + } + Put16 (DC >> 4, buf, dst); + } + + static void DC16NoTopLeft (byte[] buf, int dst) // DC with no top and left samples + { + Put16 (0x80, buf, dst); + } + + //------------------------------------------------------------------------------ + // 4x4 + + static byte AVG3 (byte a, byte b, byte c) + { + return (byte)((a + 2 * b + c + 2) >> 2); + } + + static byte AVG2 (byte a, byte b) + { + return (byte)((a + b + 1) >> 1); + } + + static void VE4 (byte[] buf, int dst) // vertical + { + int top = dst - BPS; + byte vals0 = AVG3 (buf[top-1], buf[top], buf[top+1]); + byte vals1 = AVG3 (buf[top], buf[top+1], buf[top+2]); + byte vals2 = AVG3 (buf[top+1], buf[top+2], buf[top+3]); + byte vals3 = AVG3 (buf[top+2], buf[top+3], buf[top+4]); + for (int i = 0; i < 4; ++i) { + buf[dst] = vals0; + buf[dst+1] = vals1; + buf[dst+2] = vals2; + buf[dst+3] = vals3; + dst += BPS; + } + } + + static void HE4 (byte[] buf, int dst) // horizontal + { + byte A = buf[dst -1 - BPS]; + byte B = buf[dst -1]; + byte C = buf[dst -1 + BPS]; + byte D = buf[dst -1 + 2 * BPS]; + byte E = buf[dst -1 + 3 * BPS]; + LittleEndian.Pack (0x01010101U * AVG3(A, B, C), buf, dst); + LittleEndian.Pack (0x01010101U * AVG3(B, C, D), buf, dst + BPS); + LittleEndian.Pack (0x01010101U * AVG3(C, D, E), buf, dst + BPS * 2); + LittleEndian.Pack (0x01010101U * AVG3(D, E, E), buf, dst + BPS * 3); + } + + static void DC4 (byte[] buf, int dst) // DC + { + uint dc = 4; + for (int i = 0; i < 4; ++i) + dc += (uint)buf[dst + i - BPS] + buf[dst - 1 + i * BPS]; + dc >>= 3; + for (int i = 0; i < 4; ++i) + { + buf[dst] = (byte)dc; + buf[dst+1] = (byte)dc; + buf[dst+2] = (byte)dc; + buf[dst+3] = (byte)dc; + dst += BPS; + } + } + + static void RD4 (byte[] buf, int dst) // Down-right + { + byte I = buf[dst - 1 + 0 * BPS]; + byte J = buf[dst - 1 + 1 * BPS]; + byte K = buf[dst - 1 + 2 * BPS]; + byte L = buf[dst - 1 + 3 * BPS]; + byte X = buf[dst - 1 - BPS]; + byte A = buf[dst + 0 - BPS]; + byte B = buf[dst + 1 - BPS]; + byte C = buf[dst + 2 - BPS]; + byte D = buf[dst + 3 - BPS]; + buf[dst + 0 + 3 * BPS] = AVG3(J, K, L); + buf[dst + 1 + 3 * BPS] = buf[dst + 0 + 2 * BPS] = AVG3(I, J, K); + buf[dst + 2 + 3 * BPS] = buf[dst + 1 + 2 * BPS] = buf[dst + BPS] = AVG3(X, I, J); + buf[dst + 3 + 3 * BPS] = buf[dst + 2 + 2 * BPS] = buf[dst + 1 + BPS] = buf[dst] = AVG3(A, X, I); + buf[dst + 3 + 2 * BPS] = buf[dst + 2 + BPS] = buf[dst + 1] = AVG3(B, A, X); + buf[dst + 3 + BPS] = buf[dst + 2] = AVG3(C, B, A); + buf[dst + 3] = AVG3(D, C, B); + } + + static void LD4 (byte[] buf, int dst) // Down-Left + { + byte A = buf[dst + 0 - BPS]; + byte B = buf[dst + 1 - BPS]; + byte C = buf[dst + 2 - BPS]; + byte D = buf[dst + 3 - BPS]; + byte E = buf[dst + 4 - BPS]; + byte F = buf[dst + 5 - BPS]; + byte G = buf[dst + 6 - BPS]; + byte H = buf[dst + 7 - BPS]; + buf[dst] = AVG3(A, B, C); + buf[dst + 1] = buf[dst + 0 + BPS] = AVG3(B, C, D); + buf[dst + 2] = buf[dst + 1 + BPS] = buf[dst + 0 + 2 * BPS] = AVG3(C, D, E); + buf[dst + 3] = buf[dst + 2 + 1 * BPS] = buf[dst + 1 + 2 * BPS] = buf[dst + 0 + 3 * BPS] = AVG3(D, E, F); + buf[dst + 3 + 1 * BPS] = buf[dst + 2 + 2 * BPS] = buf[dst + 1 + 3 * BPS] = AVG3(E, F, G); + buf[dst + 3 + 2 * BPS] = buf[dst + 2 + 3 * BPS] = AVG3(F, G, H); + buf[dst + 3 + 3 * BPS] = AVG3(G, H, H); + } + + static void VR4 (byte[] buf, int dst) // Vertical-Right + { + byte I = buf[dst - 1 + 0 * BPS]; + byte J = buf[dst - 1 + 1 * BPS]; + byte K = buf[dst - 1 + 2 * BPS]; + byte X = buf[dst - 1 - BPS]; + byte A = buf[dst + 0 - BPS]; + byte B = buf[dst + 1 - BPS]; + byte C = buf[dst + 2 - BPS]; + byte D = buf[dst + 3 - BPS]; + buf[dst + 0] = buf[dst + 1 + 2 * BPS] = AVG2(X, A); + buf[dst + 1] = buf[dst + 2 + 2 * BPS] = AVG2(A, B); + buf[dst + 2] = buf[dst + 3 + 2 * BPS] = AVG2(B, C); + buf[dst + 3] = AVG2(C, D); + + buf[dst + 0 + 3 * BPS] = AVG3(K, J, I); + buf[dst + 0 + 2 * BPS] = AVG3(J, I, X); + buf[dst + 0 + 1 * BPS] = buf[dst + 1 + 3 * BPS] = AVG3(I, X, A); + buf[dst + 1 + 1 * BPS] = buf[dst + 2 + 3 * BPS] = AVG3(X, A, B); + buf[dst + 2 + 1 * BPS] = buf[dst + 3 + 3 * BPS] = AVG3(A, B, C); + buf[dst + 3 + 1 * BPS] = AVG3(B, C, D); + } + + static void VL4 (byte[] buf, int dst) // Vertical-Left + { + byte A = buf[dst + 0 - BPS]; + byte B = buf[dst + 1 - BPS]; + byte C = buf[dst + 2 - BPS]; + byte D = buf[dst + 3 - BPS]; + byte E = buf[dst + 4 - BPS]; + byte F = buf[dst + 5 - BPS]; + byte G = buf[dst + 6 - BPS]; + byte H = buf[dst + 7 - BPS]; + buf[dst] = AVG2(A, B); + buf[dst + 1] = buf[dst + 0 + 2 * BPS] = AVG2(B, C); + buf[dst + 2] = buf[dst + 1 + 2 * BPS] = AVG2(C, D); + buf[dst + 3] = buf[dst + 2 + 2 * BPS] = AVG2(D, E); + + buf[dst + 0 + 1 * BPS] = AVG3(A, B, C); + buf[dst + 1 + 1 * BPS] = buf[dst + 0 + 3 * BPS] = AVG3(B, C, D); + buf[dst + 2 + 1 * BPS] = buf[dst + 1 + 3 * BPS] = AVG3(C, D, E); + buf[dst + 3 + 1 * BPS] = buf[dst + 2 + 3 * BPS] = AVG3(D, E, F); + buf[dst + 3 + 2 * BPS] = AVG3(E, F, G); + buf[dst + 3 + 3 * BPS] = AVG3(F, G, H); + } + + static void HU4 (byte[] buf, int dst) // Horizontal-Up + { + byte I = buf[dst - 1 + 0 * BPS]; + byte J = buf[dst - 1 + 1 * BPS]; + byte K = buf[dst - 1 + 2 * BPS]; + byte L = buf[dst - 1 + 3 * BPS]; + buf[dst] = AVG2(I, J); + buf[dst + 2 + 0 * BPS] = buf[dst + 0 + 1 * BPS] = AVG2(J, K); + buf[dst + 2 + 1 * BPS] = buf[dst + 0 + 2 * BPS] = AVG2(K, L); + buf[dst + 1 + 0 * BPS] = AVG3(I, J, K); + buf[dst + 3 + 0 * BPS] = buf[dst + 1 + 1 * BPS] = AVG3(J, K, L); + buf[dst + 3 + 1 * BPS] = buf[dst + 1 + 2 * BPS] = AVG3(K, L, L); + buf[dst + 3 + 2 * BPS] = buf[dst + 2 + 2 * BPS] = + buf[dst + 0 + 3 * BPS] = buf[dst + 1 + 3 * BPS] = buf[dst + 2 + 3 * BPS] = buf[dst + 3 + 3 * BPS] = L; + } + + static void HD4 (byte[] buf, int dst) // Horizontal-Down + { + byte I = buf[dst - 1 + 0 * BPS]; + byte J = buf[dst - 1 + 1 * BPS]; + byte K = buf[dst - 1 + 2 * BPS]; + byte L = buf[dst - 1 + 3 * BPS]; + byte X = buf[dst - 1 - BPS]; + byte A = buf[dst + 0 - BPS]; + byte B = buf[dst + 1 - BPS]; + byte C = buf[dst + 2 - BPS]; + + buf[dst + 0 + 0 * BPS] = buf[dst + 2 + 1 * BPS] = AVG2(I, X); + buf[dst + 0 + 1 * BPS] = buf[dst + 2 + 2 * BPS] = AVG2(J, I); + buf[dst + 0 + 2 * BPS] = buf[dst + 2 + 3 * BPS] = AVG2(K, J); + buf[dst + 0 + 3 * BPS] = AVG2(L, K); + + buf[dst + 3 + 0 * BPS] = AVG3(A, B, C); + buf[dst + 2 + 0 * BPS] = AVG3(X, A, B); + buf[dst + 1 + 0 * BPS] = buf[dst + 3 + 1 * BPS] = AVG3(I, X, A); + buf[dst + 1 + 1 * BPS] = buf[dst + 3 + 2 * BPS] = AVG3(J, I, X); + buf[dst + 1 + 2 * BPS] = buf[dst + 3 + 3 * BPS] = AVG3(K, J, I); + buf[dst + 1 + 3 * BPS] = AVG3(L, K, J); + } + + //------------------------------------------------------------------------------ + // Chroma + + static void VE8uv (byte[] buf, int dst) // vertical + { + for (int j = 0; j < 8; ++j) + { + Buffer.BlockCopy (buf, dst - BPS, buf, dst + j * BPS, 8); + } + } + + static void HE8uv (byte[] buf, int dst) // horizontal + { + for (int j = 0; j < 8; ++j) + { + byte v = buf[dst-1]; + for (int i = 0; i < 8; ++i) + buf[dst+i] = v; + dst += BPS; + } + } + + static void Put8x8uv (int val, byte[] buf, int dst) + { + for (int j = 0; j < 8; ++j) + { + for (int i = 0; i < 8; ++i) + buf[dst+i] = (byte)val; + dst += BPS; + } + } + + static void DC8uv (byte[] buf, int dst) // DC + { + int dc0 = 8; + for (int i = 0; i < 8; ++i) + { + dc0 += buf[dst + i - BPS] + buf[dst - 1 + i * BPS]; + } + Put8x8uv (dc0 >> 4, buf, dst); + } + + static void DC8uvNoLeft (byte[] buf, int dst) // DC with no left samples + { + int dc0 = 4; + for (int i = 0; i < 8; ++i) + { + dc0 += buf[dst + i - BPS]; + } + Put8x8uv (dc0 >> 3, buf, dst); + } + + static void DC8uvNoTop (byte[] buf, int dst) // DC with no top samples + { + int dc0 = 4; + for (int i = 0; i < 8; ++i) { + dc0 += buf[dst - 1 + i * BPS]; + } + Put8x8uv(dc0 >> 3, buf, dst); + } + + static void DC8uvNoTopLeft (byte[] buf, int dst) // DC with nothing + { + Put8x8uv(0x80, buf, dst); + } + + //------------------------------------------------------------------------------ + + const int TreeProbs = 3; + const int NumMbSegments = 4; + const int NumRefLfDeltas = 4; + const int NumModeLfDeltas = 4; + const int MaxNumPartitions = 8; + + const int NumTypes = 4; + const int NumBands = 8; + const int NumCtx = 3; + const int NumProbas = 11; + + const sbyte B_DC_PRED = 0; // 4x4 modes + const sbyte B_TM_PRED = 1; + const sbyte B_VE_PRED = 2; + const sbyte B_HE_PRED = 3; + const sbyte B_RD_PRED = 4; + const sbyte B_VR_PRED = 5; + const sbyte B_LD_PRED = 6; + const sbyte B_VL_PRED = 7; + const sbyte B_HD_PRED = 8; + const sbyte B_HU_PRED = 9; + + const byte H_PRED = (byte)B_HE_PRED; + const byte TM_PRED = (byte)B_TM_PRED; + const byte DC_PRED = (byte)B_DC_PRED; + const byte V_PRED = (byte)B_VE_PRED; + + const int B_DC_PRED_NOTOP = 4; + const int B_DC_PRED_NOLEFT = 5; + const int B_DC_PRED_NOTOPLEFT = 6; + + const int NumBModes = 10; + const int NumBDCModes = 7; + + const int BPS = 32; + const int YUV_SIZE = BPS * 17 + BPS * 9; + const int Y_SIZE = BPS * 17; + const int Y_OFF = BPS * 1 + 8; + const int U_OFF = Y_OFF + BPS * 16 + BPS; + const int V_OFF = U_OFF + 16; + + delegate void PredFunc (byte[] buf, int dst); + + PredFunc[] PredLuma4 = new PredFunc[NumBModes]; + PredFunc[] PredLuma16 = new PredFunc[NumBDCModes]; + PredFunc[] PredChroma8 = new PredFunc[NumBDCModes]; + + internal class PictureHeader + { + public ushort Width; + public ushort Height; + public byte XScale; + public byte YScale; + public byte Colorspace; // 0 = YCbCr + public byte ClampType; + } + + internal class SegmentHeader + { + public bool UseSegment; + public bool UpdateMap; + public bool AbsoluteDelta; + public byte[] Quantizer = new byte[NumMbSegments]; + public byte[] FilterStrength = new byte[NumMbSegments]; + + public void Reset () + { + UseSegment = false; + UpdateMap = false; + AbsoluteDelta = true; + for (int i = 0; i < NumMbSegments; ++i) + { + Quantizer[i] = 0; + FilterStrength[i] = 0; + } + } + } + + internal class FilterHeader + { + public bool Simple; + public int Level; + public int Sharpness; + public bool UseLfDelta; + public int[] RefLfDelta = new int[NumRefLfDeltas]; + public int[] ModeLfDelta = new int[NumModeLfDeltas]; + } + + internal class BandProbas + { + public byte[][] Probas = new byte[NumCtx][]; + + public BandProbas () + { + for (int i = 0; i < NumCtx; ++i) + { + Probas[i] = new byte[NumProbas]; + } + } + } + + internal class Proba + { + public byte[] Segments = new byte[TreeProbs]; + public BandProbas[,] Bands = new BandProbas[NumTypes,NumBands]; + public BandProbas[][] BandsPtr; + + public Proba () + { + BandsPtr = new BandProbas[NumTypes][]; + for (int i = 0; i < NumTypes; ++i) + BandsPtr[i] = new BandProbas[17]; + } + + public void Reset () + { + for (int i = 0; i < Segments.Length; ++i) + Segments[i] = 0xFF; + for (int t = 0; t < NumTypes; ++t) + for (int b = 0; b < NumBands; ++b) + { + Bands[t,b] = new BandProbas(); + } + } + } + + internal class QuantMatrix + { + public int[] y1_mat = new int[2]; + public int[] y2_mat = new int[2]; + public int[] uv_mat = new int[2]; + public int uv_quant; + } + + internal class MacroBlockData + { + public short[] coeffs_ = new short[384]; // 384 coeffs = (16+4+4) * 4*4 + public bool is_i4x4_; // true if intra4x4 + public byte[] imodes_ = new byte[16]; // one 16x16 mode (#0) or sixteen 4x4 modes + public byte uvmode_; // chroma prediction mode + // bit-wise info about the content of each sub-4x4 blocks (in decoding order). + // Each of the 4x4 blocks for y/u/v is associated with a 2b code according to: + // code=0 -> no coefficient + // code=1 -> only DC + // code=2 -> first three coefficients are non-zero + // code=3 -> more than three coefficients are non-zero + // This allows to call specialized transform functions. + public uint non_zero_y_; + public uint non_zero_uv_; + public bool skip_; + public byte segment_; + } + + struct MacroBlock + { + public byte nz_; + public byte nz_dc_; + } + + class TopSamples + { + public byte[] y = new byte[16]; + public byte[] u = new byte[8]; + public byte[] v = new byte[8]; + } + + internal class BitReader : IBitStream + { + uint m_value; + int m_range; + int m_bits; + byte[] m_buf; + int m_buf_pos; + int m_buf_max; + int m_buf_end; + bool m_eof; + + public bool Eof { get { return m_eof; } } + + public BitReader (BinaryReader input, int length) + { + Init (input, length); + } + + public void Init (BinaryReader input, int length) + { + if (null == m_buf || m_buf.Length < length) + m_buf = new byte[length]; + if (length != input.Read (m_buf, 0, length)) + throw new EndOfStreamException(); + m_buf_end = length; + Reset(); + } + + public void Reset () + { + m_range = 255 - 1; + m_value = 0; + m_bits = -8; + m_eof = false; + m_buf_pos = 0; + m_buf_max = m_buf_end >= sizeof(uint) ? m_buf_end - sizeof(uint) + 1 : 0; + LoadNewBytes(); + } + + public int GetNextBit () + { + return GetBits (1); + } + + public int GetBits (int bits) + { + int v = 0; + while (bits --> 0) + v |= GetBit (0x80) << bits; + return v; + } + + public int GetSignedValue (int bits) + { + int val = GetBits (bits); + return GetNextBit() != 0 ? -val : val; + } + + public int GetBit (int prob) + { + if (m_bits < 0) + LoadNewBytes(); + + int split = (m_range * prob) >> 8; + uint value = m_value >> m_bits; + + int bit; + if (value > split) + { + m_range -= split + 1; + m_value -= (uint)(split + 1) << m_bits; + bit = 1; + } + else + { + m_range = split; + bit = 0; + } + if (m_range <= 0x7Eu) + { + int shift = kLog2Range[m_range]; + m_range = kNewRange[m_range]; + m_bits -= shift; + } + return bit; + } + + const int BITS = 24; + + void LoadNewBytes () + { + if (m_buf_pos < m_buf_max) + { + uint bits = BigEndian.ToUInt32 (m_buf, m_buf_pos); + m_buf_pos += BITS >> 3; + bits >>= (32 - BITS); + + m_value = bits | (m_value << BITS); + m_bits += BITS; + } else { + LoadFinalBytes(); + } + } + + void LoadFinalBytes () + { + // Only read 8bits at a time + if (m_buf_pos < m_buf_end) + { + m_bits += 8; + m_value = (uint)(m_buf[m_buf_pos++]) | (m_value << 8); + } + else if (!m_eof) + { + m_value <<= 8; + m_bits += 8; + m_eof = true; + } + else + { + m_bits = 0; // This is to avoid undefined behaviour with shifts. + } + } + + public int GetSigned (int v) + { + if (m_bits < 0) + LoadNewBytes(); + + int pos = m_bits; + uint split = (uint)m_range >> 1; + uint val = m_value >> pos; + int mask = (int)(split - val) >> 31; // -1 or 0 + m_bits -= 1; + m_range += mask; + m_range |= 1; + m_value -= ((split + 1) & (uint)mask) << pos; + return (v ^ mask) - mask; + } + + public int GetLargeValue (byte[] p) + { + int v; + if (0 == GetBit (p[3])) + { + if (0 == GetBit (p[4])) + v = 2; + else + v = 3 + GetBit (p[5]); + } + else + { + if (0 == GetBit (p[6])) + { + if (0 == GetBit (p[7])) + { + v = 5 + GetBit (159); + } + else + { + v = 7 + 2 * GetBit (165); + v += GetBit (145); + } + } + else + { + int bit1 = GetBit (p[8]); + int bit0 = GetBit (p[9 + bit1]); + int cat = bit1 << 1 | bit0; + v = 0; + var tab = kCat3456[cat]; + for (int i = 0; tab[i] != 0; ++i) + { + v += v + GetBit (tab[i]); + } + v += 3 + (8 << cat); + } + } + return v; + } + } + + bool is_alpha_decoded_ = false; + + AlphaDecoder alph_dec_; + int alpha_dithering_; // derived from decoding options (0=off, 100=full) + + int DecompressAlphaRows (int row, int num_rows) + { + int width = m_io.width; + int height = m_io.height; + + if (row < 0 || num_rows <= 0 || row + num_rows > height) + throw new InvalidFormatException ("Could not decode alpha data."); + + if (!is_alpha_decoded_) + { + if (null == alph_dec_) // Initialize decoder. + { + alph_dec_ = new AlphaDecoder(); + if (!alph_dec_.Init (m_alpha_data, m_io, m_alpha_plane)) + throw new InvalidFormatException ("Could not decode alpha data."); + // if we allowed use of alpha dithering, check whether it's needed at all + if (alph_dec_.pre_processing_ != AlphaDecoder.PreprocessedLevels) + alpha_dithering_ = 0; // disable dithering + else + num_rows = height - row; // decode everything in one pass + } + + if (!ALPHDecode (row, num_rows)) + return -1; + + if (is_alpha_decoded_ && alpha_dithering_ > 0) + { + throw new NotImplementedException ("Alpha dithering not implemented."); + /* + var alpha = m_io.height * width; + if (!WebPDequantizeLevels (alpha, m_io.width, m_io.height, width, dec->alpha_dithering_)) + throw new InvalidFormatException ("Could not decode alpha data."); + */ + } + } + + // Return a pointer to the current decoded row. + return row * width; + } + + bool ALPHDecode (int row, int num_rows) + { + bool ok = alph_dec_.Decode (m_alpha_data, m_alpha_plane, row, num_rows); + if (ok) + is_alpha_decoded_ = alph_dec_.DecodeComplete; + return ok; + } + + #region IDisposable Members + bool _disposed = false; + public void Dispose () + { + if (!_disposed) + { + m_input.Dispose(); + _disposed = true; + } + GC.SuppressFinalize (this); + } + #endregion + + static readonly byte[,,,] CoeffsProba0 = new byte[,,,] { + { { { 128, 128, 128, 128, 128, 128, 128, 128, 128, 128, 128 }, + { 128, 128, 128, 128, 128, 128, 128, 128, 128, 128, 128 }, + { 128, 128, 128, 128, 128, 128, 128, 128, 128, 128, 128 } + }, + { { 253, 136, 254, 255, 228, 219, 128, 128, 128, 128, 128 }, + { 189, 129, 242, 255, 227, 213, 255, 219, 128, 128, 128 }, + { 106, 126, 227, 252, 214, 209, 255, 255, 128, 128, 128 } + }, + { { 1, 98, 248, 255, 236, 226, 255, 255, 128, 128, 128 }, + { 181, 133, 238, 254, 221, 234, 255, 154, 128, 128, 128 }, + { 78, 134, 202, 247, 198, 180, 255, 219, 128, 128, 128 }, + }, + { { 1, 185, 249, 255, 243, 255, 128, 128, 128, 128, 128 }, + { 184, 150, 247, 255, 236, 224, 128, 128, 128, 128, 128 }, + { 77, 110, 216, 255, 236, 230, 128, 128, 128, 128, 128 }, + }, + { { 1, 101, 251, 255, 241, 255, 128, 128, 128, 128, 128 }, + { 170, 139, 241, 252, 236, 209, 255, 255, 128, 128, 128 }, + { 37, 116, 196, 243, 228, 255, 255, 255, 128, 128, 128 } + }, + { { 1, 204, 254, 255, 245, 255, 128, 128, 128, 128, 128 }, + { 207, 160, 250, 255, 238, 128, 128, 128, 128, 128, 128 }, + { 102, 103, 231, 255, 211, 171, 128, 128, 128, 128, 128 } + }, + { { 1, 152, 252, 255, 240, 255, 128, 128, 128, 128, 128 }, + { 177, 135, 243, 255, 234, 225, 128, 128, 128, 128, 128 }, + { 80, 129, 211, 255, 194, 224, 128, 128, 128, 128, 128 } + }, + { { 1, 1, 255, 128, 128, 128, 128, 128, 128, 128, 128 }, + { 246, 1, 255, 128, 128, 128, 128, 128, 128, 128, 128 }, + { 255, 128, 128, 128, 128, 128, 128, 128, 128, 128, 128 } + } + }, + { { { 198, 35, 237, 223, 193, 187, 162, 160, 145, 155, 62 }, + { 131, 45, 198, 221, 172, 176, 220, 157, 252, 221, 1 }, + { 68, 47, 146, 208, 149, 167, 221, 162, 255, 223, 128 } + }, + { { 1, 149, 241, 255, 221, 224, 255, 255, 128, 128, 128 }, + { 184, 141, 234, 253, 222, 220, 255, 199, 128, 128, 128 }, + { 81, 99, 181, 242, 176, 190, 249, 202, 255, 255, 128 } + }, + { { 1, 129, 232, 253, 214, 197, 242, 196, 255, 255, 128 }, + { 99, 121, 210, 250, 201, 198, 255, 202, 128, 128, 128 }, + { 23, 91, 163, 242, 170, 187, 247, 210, 255, 255, 128 } + }, + { { 1, 200, 246, 255, 234, 255, 128, 128, 128, 128, 128 }, + { 109, 178, 241, 255, 231, 245, 255, 255, 128, 128, 128 }, + { 44, 130, 201, 253, 205, 192, 255, 255, 128, 128, 128 } + }, + { { 1, 132, 239, 251, 219, 209, 255, 165, 128, 128, 128 }, + { 94, 136, 225, 251, 218, 190, 255, 255, 128, 128, 128 }, + { 22, 100, 174, 245, 186, 161, 255, 199, 128, 128, 128 } + }, + { { 1, 182, 249, 255, 232, 235, 128, 128, 128, 128, 128 }, + { 124, 143, 241, 255, 227, 234, 128, 128, 128, 128, 128 }, + { 35, 77, 181, 251, 193, 211, 255, 205, 128, 128, 128 } + }, + { { 1, 157, 247, 255, 236, 231, 255, 255, 128, 128, 128 }, + { 121, 141, 235, 255, 225, 227, 255, 255, 128, 128, 128 }, + { 45, 99, 188, 251, 195, 217, 255, 224, 128, 128, 128 } + }, + { { 1, 1, 251, 255, 213, 255, 128, 128, 128, 128, 128 }, + { 203, 1, 248, 255, 255, 128, 128, 128, 128, 128, 128 }, + { 137, 1, 177, 255, 224, 255, 128, 128, 128, 128, 128 } + } + }, + { { { 253, 9, 248, 251, 207, 208, 255, 192, 128, 128, 128 }, + { 175, 13, 224, 243, 193, 185, 249, 198, 255, 255, 128 }, + { 73, 17, 171, 221, 161, 179, 236, 167, 255, 234, 128 } + }, + { { 1, 95, 247, 253, 212, 183, 255, 255, 128, 128, 128 }, + { 239, 90, 244, 250, 211, 209, 255, 255, 128, 128, 128 }, + { 155, 77, 195, 248, 188, 195, 255, 255, 128, 128, 128 } + }, + { { 1, 24, 239, 251, 218, 219, 255, 205, 128, 128, 128 }, + { 201, 51, 219, 255, 196, 186, 128, 128, 128, 128, 128 }, + { 69, 46, 190, 239, 201, 218, 255, 228, 128, 128, 128 } + }, + { { 1, 191, 251, 255, 255, 128, 128, 128, 128, 128, 128 }, + { 223, 165, 249, 255, 213, 255, 128, 128, 128, 128, 128 }, + { 141, 124, 248, 255, 255, 128, 128, 128, 128, 128, 128 } + }, + { { 1, 16, 248, 255, 255, 128, 128, 128, 128, 128, 128 }, + { 190, 36, 230, 255, 236, 255, 128, 128, 128, 128, 128 }, + { 149, 1, 255, 128, 128, 128, 128, 128, 128, 128, 128 } + }, + { { 1, 226, 255, 128, 128, 128, 128, 128, 128, 128, 128 }, + { 247, 192, 255, 128, 128, 128, 128, 128, 128, 128, 128 }, + { 240, 128, 255, 128, 128, 128, 128, 128, 128, 128, 128 } + }, + { { 1, 134, 252, 255, 255, 128, 128, 128, 128, 128, 128 }, + { 213, 62, 250, 255, 255, 128, 128, 128, 128, 128, 128 }, + { 55, 93, 255, 128, 128, 128, 128, 128, 128, 128, 128 } + }, + { { 128, 128, 128, 128, 128, 128, 128, 128, 128, 128, 128 }, + { 128, 128, 128, 128, 128, 128, 128, 128, 128, 128, 128 }, + { 128, 128, 128, 128, 128, 128, 128, 128, 128, 128, 128 } + } + }, + { { { 202, 24, 213, 235, 186, 191, 220, 160, 240, 175, 255 }, + { 126, 38, 182, 232, 169, 184, 228, 174, 255, 187, 128 }, + { 61, 46, 138, 219, 151, 178, 240, 170, 255, 216, 128 } + }, + { { 1, 112, 230, 250, 199, 191, 247, 159, 255, 255, 128 }, + { 166, 109, 228, 252, 211, 215, 255, 174, 128, 128, 128 }, + { 39, 77, 162, 232, 172, 180, 245, 178, 255, 255, 128 } + }, + { { 1, 52, 220, 246, 198, 199, 249, 220, 255, 255, 128 }, + { 124, 74, 191, 243, 183, 193, 250, 221, 255, 255, 128 }, + { 24, 71, 130, 219, 154, 170, 243, 182, 255, 255, 128 } + }, + { { 1, 182, 225, 249, 219, 240, 255, 224, 128, 128, 128 }, + { 149, 150, 226, 252, 216, 205, 255, 171, 128, 128, 128 }, + { 28, 108, 170, 242, 183, 194, 254, 223, 255, 255, 128 } + }, + { { 1, 81, 230, 252, 204, 203, 255, 192, 128, 128, 128 }, + { 123, 102, 209, 247, 188, 196, 255, 233, 128, 128, 128 }, + { 20, 95, 153, 243, 164, 173, 255, 203, 128, 128, 128 } + }, + { { 1, 222, 248, 255, 216, 213, 128, 128, 128, 128, 128 }, + { 168, 175, 246, 252, 235, 205, 255, 255, 128, 128, 128 }, + { 47, 116, 215, 255, 211, 212, 255, 255, 128, 128, 128 } + }, + { { 1, 121, 236, 253, 212, 214, 255, 255, 128, 128, 128 }, + { 141, 84, 213, 252, 201, 202, 255, 219, 128, 128, 128 }, + { 42, 80, 160, 240, 162, 185, 255, 205, 128, 128, 128 } + }, + { { 1, 1, 255, 128, 128, 128, 128, 128, 128, 128, 128 }, + { 244, 1, 255, 128, 128, 128, 128, 128, 128, 128, 128 }, + { 238, 1, 255, 128, 128, 128, 128, 128, 128, 128, 128 } + } + } + }; + + static readonly byte[,,,] CoeffsUpdateProba = new byte[NumTypes,NumBands,NumCtx,NumProbas] { + { { { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 176, 246, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 223, 241, 252, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 249, 253, 253, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 244, 252, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 234, 254, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 253, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 246, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 239, 253, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 254, 255, 254, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 248, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 251, 255, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 253, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 251, 254, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 254, 255, 254, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 254, 253, 255, 254, 255, 255, 255, 255, 255, 255 }, + { 250, 255, 254, 255, 254, 255, 255, 255, 255, 255, 255 }, + { 254, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + } + }, + { { { 217, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 225, 252, 241, 253, 255, 255, 254, 255, 255, 255, 255 }, + { 234, 250, 241, 250, 253, 255, 253, 254, 255, 255, 255 } + }, + { { 255, 254, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 223, 254, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 238, 253, 254, 254, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 248, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 249, 254, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 253, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 247, 254, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 253, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 252, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 254, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 253, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 254, 253, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 250, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 254, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + } + }, + { { { 186, 251, 250, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 234, 251, 244, 254, 255, 255, 255, 255, 255, 255, 255 }, + { 251, 251, 243, 253, 254, 255, 254, 255, 255, 255, 255 } + }, + { { 255, 253, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 236, 253, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 251, 253, 253, 254, 254, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 254, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 254, 254, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 254, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 254, 254, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 254, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 254, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + } + }, + { { { 248, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 250, 254, 252, 254, 255, 255, 255, 255, 255, 255, 255 }, + { 248, 254, 249, 253, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 253, 253, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 246, 253, 253, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 252, 254, 251, 254, 254, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 254, 252, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 248, 254, 253, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 253, 255, 254, 254, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 251, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 245, 251, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 253, 253, 254, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 251, 253, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 252, 253, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 254, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 252, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 249, 255, 254, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 254, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 255, 253, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 250, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + }, + { { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 254, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 }, + { 255, 255, 255, 255, 255, 255, 255, 255, 255, 255, 255 } + } + } + }; + + static readonly int[] kBands = { + 0, 1, 2, 3, 6, 4, 5, 6, 6, 6, 6, 6, 6, 6, 6, 7, 0 + }; + + static readonly byte[] kLog2Range = { + 7, 6, 6, 5, 5, 5, 5, 4, 4, 4, 4, 4, 4, 4, 4, + 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0 + }; + + static readonly byte[] kNewRange = { + 127, 127, 191, 127, 159, 191, 223, 127, + 143, 159, 175, 191, 207, 223, 239, 127, + 135, 143, 151, 159, 167, 175, 183, 191, + 199, 207, 215, 223, 231, 239, 247, 127, + 131, 135, 139, 143, 147, 151, 155, 159, + 163, 167, 171, 175, 179, 183, 187, 191, + 195, 199, 203, 207, 211, 215, 219, 223, + 227, 231, 235, 239, 243, 247, 251, 127, + 129, 131, 133, 135, 137, 139, 141, 143, + 145, 147, 149, 151, 153, 155, 157, 159, + 161, 163, 165, 167, 169, 171, 173, 175, + 177, 179, 181, 183, 185, 187, 189, 191, + 193, 195, 197, 199, 201, 203, 205, 207, + 209, 211, 213, 215, 217, 219, 221, 223, + 225, 227, 229, 231, 233, 235, 237, 239, + 241, 243, 245, 247, 249, 251, 253, 127 + }; + + static readonly byte[] kDcTable = { + 4, 5, 6, 7, 8, 9, 10, 10, + 11, 12, 13, 14, 15, 16, 17, 17, + 18, 19, 20, 20, 21, 21, 22, 22, + 23, 23, 24, 25, 25, 26, 27, 28, + 29, 30, 31, 32, 33, 34, 35, 36, + 37, 37, 38, 39, 40, 41, 42, 43, + 44, 45, 46, 46, 47, 48, 49, 50, + 51, 52, 53, 54, 55, 56, 57, 58, + 59, 60, 61, 62, 63, 64, 65, 66, + 67, 68, 69, 70, 71, 72, 73, 74, + 75, 76, 76, 77, 78, 79, 80, 81, + 82, 83, 84, 85, 86, 87, 88, 89, + 91, 93, 95, 96, 98, 100, 101, 102, + 104, 106, 108, 110, 112, 114, 116, 118, + 122, 124, 126, 128, 130, 132, 134, 136, + 138, 140, 143, 145, 148, 151, 154, 157 + }; + + static readonly ushort[] kAcTable = { + 4, 5, 6, 7, 8, 9, 10, 11, + 12, 13, 14, 15, 16, 17, 18, 19, + 20, 21, 22, 23, 24, 25, 26, 27, + 28, 29, 30, 31, 32, 33, 34, 35, + 36, 37, 38, 39, 40, 41, 42, 43, + 44, 45, 46, 47, 48, 49, 50, 51, + 52, 53, 54, 55, 56, 57, 58, 60, + 62, 64, 66, 68, 70, 72, 74, 76, + 78, 80, 82, 84, 86, 88, 90, 92, + 94, 96, 98, 100, 102, 104, 106, 108, + 110, 112, 114, 116, 119, 122, 125, 128, + 131, 134, 137, 140, 143, 146, 149, 152, + 155, 158, 161, 164, 167, 170, 173, 177, + 181, 185, 189, 193, 197, 201, 205, 209, + 213, 217, 221, 225, 229, 234, 239, 245, + 249, 254, 259, 264, 269, 274, 279, 284 + }; + + static readonly byte[,,] kBModesProba = new byte[NumBModes,NumBModes,NumBModes - 1] { + { { 231, 120, 48, 89, 115, 113, 120, 152, 112 }, + { 152, 179, 64, 126, 170, 118, 46, 70, 95 }, + { 175, 69, 143, 80, 85, 82, 72, 155, 103 }, + { 56, 58, 10, 171, 218, 189, 17, 13, 152 }, + { 114, 26, 17, 163, 44, 195, 21, 10, 173 }, + { 121, 24, 80, 195, 26, 62, 44, 64, 85 }, + { 144, 71, 10, 38, 171, 213, 144, 34, 26 }, + { 170, 46, 55, 19, 136, 160, 33, 206, 71 }, + { 63, 20, 8, 114, 114, 208, 12, 9, 226 }, + { 81, 40, 11, 96, 182, 84, 29, 16, 36 } }, + { { 134, 183, 89, 137, 98, 101, 106, 165, 148 }, + { 72, 187, 100, 130, 157, 111, 32, 75, 80 }, + { 66, 102, 167, 99, 74, 62, 40, 234, 128 }, + { 41, 53, 9, 178, 241, 141, 26, 8, 107 }, + { 74, 43, 26, 146, 73, 166, 49, 23, 157 }, + { 65, 38, 105, 160, 51, 52, 31, 115, 128 }, + { 104, 79, 12, 27, 217, 255, 87, 17, 7 }, + { 87, 68, 71, 44, 114, 51, 15, 186, 23 }, + { 47, 41, 14, 110, 182, 183, 21, 17, 194 }, + { 66, 45, 25, 102, 197, 189, 23, 18, 22 } }, + { { 88, 88, 147, 150, 42, 46, 45, 196, 205 }, + { 43, 97, 183, 117, 85, 38, 35, 179, 61 }, + { 39, 53, 200, 87, 26, 21, 43, 232, 171 }, + { 56, 34, 51, 104, 114, 102, 29, 93, 77 }, + { 39, 28, 85, 171, 58, 165, 90, 98, 64 }, + { 34, 22, 116, 206, 23, 34, 43, 166, 73 }, + { 107, 54, 32, 26, 51, 1, 81, 43, 31 }, + { 68, 25, 106, 22, 64, 171, 36, 225, 114 }, + { 34, 19, 21, 102, 132, 188, 16, 76, 124 }, + { 62, 18, 78, 95, 85, 57, 50, 48, 51 } }, + { { 193, 101, 35, 159, 215, 111, 89, 46, 111 }, + { 60, 148, 31, 172, 219, 228, 21, 18, 111 }, + { 112, 113, 77, 85, 179, 255, 38, 120, 114 }, + { 40, 42, 1, 196, 245, 209, 10, 25, 109 }, + { 88, 43, 29, 140, 166, 213, 37, 43, 154 }, + { 61, 63, 30, 155, 67, 45, 68, 1, 209 }, + { 100, 80, 8, 43, 154, 1, 51, 26, 71 }, + { 142, 78, 78, 16, 255, 128, 34, 197, 171 }, + { 41, 40, 5, 102, 211, 183, 4, 1, 221 }, + { 51, 50, 17, 168, 209, 192, 23, 25, 82 } }, + { { 138, 31, 36, 171, 27, 166, 38, 44, 229 }, + { 67, 87, 58, 169, 82, 115, 26, 59, 179 }, + { 63, 59, 90, 180, 59, 166, 93, 73, 154 }, + { 40, 40, 21, 116, 143, 209, 34, 39, 175 }, + { 47, 15, 16, 183, 34, 223, 49, 45, 183 }, + { 46, 17, 33, 183, 6, 98, 15, 32, 183 }, + { 57, 46, 22, 24, 128, 1, 54, 17, 37 }, + { 65, 32, 73, 115, 28, 128, 23, 128, 205 }, + { 40, 3, 9, 115, 51, 192, 18, 6, 223 }, + { 87, 37, 9, 115, 59, 77, 64, 21, 47 } }, + { { 104, 55, 44, 218, 9, 54, 53, 130, 226 }, + { 64, 90, 70, 205, 40, 41, 23, 26, 57 }, + { 54, 57, 112, 184, 5, 41, 38, 166, 213 }, + { 30, 34, 26, 133, 152, 116, 10, 32, 134 }, + { 39, 19, 53, 221, 26, 114, 32, 73, 255 }, + { 31, 9, 65, 234, 2, 15, 1, 118, 73 }, + { 75, 32, 12, 51, 192, 255, 160, 43, 51 }, + { 88, 31, 35, 67, 102, 85, 55, 186, 85 }, + { 56, 21, 23, 111, 59, 205, 45, 37, 192 }, + { 55, 38, 70, 124, 73, 102, 1, 34, 98 } }, + { { 125, 98, 42, 88, 104, 85, 117, 175, 82 }, + { 95, 84, 53, 89, 128, 100, 113, 101, 45 }, + { 75, 79, 123, 47, 51, 128, 81, 171, 1 }, + { 57, 17, 5, 71, 102, 57, 53, 41, 49 }, + { 38, 33, 13, 121, 57, 73, 26, 1, 85 }, + { 41, 10, 67, 138, 77, 110, 90, 47, 114 }, + { 115, 21, 2, 10, 102, 255, 166, 23, 6 }, + { 101, 29, 16, 10, 85, 128, 101, 196, 26 }, + { 57, 18, 10, 102, 102, 213, 34, 20, 43 }, + { 117, 20, 15, 36, 163, 128, 68, 1, 26 } }, + { { 102, 61, 71, 37, 34, 53, 31, 243, 192 }, + { 69, 60, 71, 38, 73, 119, 28, 222, 37 }, + { 68, 45, 128, 34, 1, 47, 11, 245, 171 }, + { 62, 17, 19, 70, 146, 85, 55, 62, 70 }, + { 37, 43, 37, 154, 100, 163, 85, 160, 1 }, + { 63, 9, 92, 136, 28, 64, 32, 201, 85 }, + { 75, 15, 9, 9, 64, 255, 184, 119, 16 }, + { 86, 6, 28, 5, 64, 255, 25, 248, 1 }, + { 56, 8, 17, 132, 137, 255, 55, 116, 128 }, + { 58, 15, 20, 82, 135, 57, 26, 121, 40 } }, + { { 164, 50, 31, 137, 154, 133, 25, 35, 218 }, + { 51, 103, 44, 131, 131, 123, 31, 6, 158 }, + { 86, 40, 64, 135, 148, 224, 45, 183, 128 }, + { 22, 26, 17, 131, 240, 154, 14, 1, 209 }, + { 45, 16, 21, 91, 64, 222, 7, 1, 197 }, + { 56, 21, 39, 155, 60, 138, 23, 102, 213 }, + { 83, 12, 13, 54, 192, 255, 68, 47, 28 }, + { 85, 26, 85, 85, 128, 128, 32, 146, 171 }, + { 18, 11, 7, 63, 144, 171, 4, 4, 246 }, + { 35, 27, 10, 146, 174, 171, 12, 26, 128 } }, + { { 190, 80, 35, 99, 180, 80, 126, 54, 45 }, + { 85, 126, 47, 87, 176, 51, 41, 20, 32 }, + { 101, 75, 128, 139, 118, 146, 116, 128, 85 }, + { 56, 41, 15, 176, 236, 85, 37, 9, 62 }, + { 71, 30, 17, 119, 118, 255, 17, 18, 138 }, + { 101, 38, 60, 138, 55, 70, 43, 26, 142 }, + { 146, 36, 19, 30, 171, 255, 97, 27, 20 }, + { 138, 45, 61, 62, 219, 1, 81, 188, 64 }, + { 32, 41, 20, 117, 151, 142, 20, 21, 163 }, + { 112, 19, 12, 61, 195, 128, 48, 4, 24 } } + }; + + static readonly sbyte[] kYModesIntra4 = new sbyte[18] { + -B_DC_PRED, 1, + -B_TM_PRED, 2, + -B_VE_PRED, 3, + 4, 6, + -B_HE_PRED, 5, + -B_RD_PRED, -B_VR_PRED, + -B_LD_PRED, 7, + -B_VL_PRED, 8, + -B_HD_PRED, -B_HU_PRED + }; + + static readonly byte[] kFilterExtraRows = { 0, 2, 8 }; + + static readonly byte[] kCat3 = { 173, 148, 140, 0 }; + static readonly byte[] kCat4 = { 176, 155, 140, 135, 0 }; + static readonly byte[] kCat5 = { 180, 157, 141, 134, 130, 0 }; + static readonly byte[] kCat6 = { 254, 254, 243, 230, 196, 177, 153, 140, 133, 130, 129, 0 }; + static readonly byte[][] kCat3456 = { kCat3, kCat4, kCat5, kCat6 }; + static readonly byte[] kZigzag = { + 0, 1, 4, 8, 5, 2, 3, 6, 9, 12, 13, 10, 7, 11, 14, 15 + }; + + static readonly int[] kScan = { + 0 + 0 * BPS, 4 + 0 * BPS, 8 + 0 * BPS, 12 + 0 * BPS, + 0 + 4 * BPS, 4 + 4 * BPS, 8 + 4 * BPS, 12 + 4 * BPS, + 0 + 8 * BPS, 4 + 8 * BPS, 8 + 8 * BPS, 12 + 8 * BPS, + 0 + 12 * BPS, 4 + 12 * BPS, 8 + 12 * BPS, 12 + 12 * BPS + }; + + const int WEBP_FILTER_NONE = 0; + const int WEBP_FILTER_HORIZONTAL = 1; + const int WEBP_FILTER_VERTICAL = 2; + const int WEBP_FILTER_GRADIENT = 3; + const int WEBP_FILTER_LAST = WEBP_FILTER_GRADIENT + 1; + } + + internal class VP8Io + { + public int width, height; + + public int mb_y; // position of the current rows (in pixels) + public int mb_w; // number of columns in the sample + public int mb_h; // number of rows in the sample + public int y, u, v; // rows to copy (in yuv420 format) + public int y_stride; // row stride for luma + public int uv_stride; // row stride for chroma + + public byte[] opaque; + + int last_y; // coordinate of the line that was last output + + // If non NULL, pointer to the alpha data (if present) corresponding to the + // start of the current row (That is: it is pre-offset by mb_y and takes + // cropping into account). + public byte[] alpha_plane; + public int a; + + public void Init (WebPDecoder dec) + { + mb_y = 0; + y = dec.cache_y_; + u = dec.cache_u_; + v = dec.cache_v_; + y_stride = dec.cache_y_stride_; + uv_stride = dec.cache_uv_stride_; + alpha_plane = dec.AlphaPlane; + a = 0; + last_y = 0; + } + + public bool Put (WebPDecoder dec) + { + if (mb_w <= 0 || mb_h <= 0) + return false; + int num_lines_out = EmitSampledRGB (dec); + if (alpha_plane != null) + EmitAlphaRGB (dec, num_lines_out); + last_y += num_lines_out; + return true; + } + + public int EmitSampledRGB (WebPDecoder dec) + { + var input = dec.Cache; + var output = dec.Output; + int dst_stride = dec.Stride; + int dst = mb_y * dst_stride; + for (int j = 0; j < mb_h; ++j) + { + YuvToBgraRow (input, y, u, v, output, dst, mb_w); + y += y_stride; + if (0 != (j & 1)) + { + u += uv_stride; + v += uv_stride; + } + dst += dst_stride; + } + return mb_h; + } + + void EmitAlphaRGB (WebPDecoder dec, int expected_num_lines_out) + { + int base_rgba = mb_y * dec.Stride; + int dst = base_rgba + 3; + DispatchAlpha (alpha_plane, a, width, mb_w, mb_h, dec.Output, dst, dec.Stride); + } + + static bool DispatchAlpha (byte[] alpha, int src, int alpha_stride, int width, int height, byte[] output, int dst, int dst_stride) + { + int alpha_src = src; + uint alpha_mask = 0xFF; + for (int j = 0; j < height; ++j) + { + for (int i = 0; i < width; ++i) + { + byte alpha_value = alpha[alpha_src+i]; + output[dst + 4 * i] = alpha_value; + alpha_mask &= alpha_value; + } + alpha_src += alpha_stride; + dst += dst_stride; + } + return alpha_mask != 0xFF; + } + + static void YuvToBgraRow (byte[] input, int y, int u, int v, byte[] output, int dst, int len) + { + const int x_step = 4; + int end = dst + (len & ~1) * x_step; + while (dst != end) + { + YuvToBgra (input[y], input[u], input[v], output, dst); + YuvToBgra (input[y+1], input[u], input[v], output, dst + x_step); + y += 2; + ++u; + ++v; + dst += 2 * x_step; + } + if (0 != (len & 1)) + YuvToBgra (input[y], input[u], input[v], output, dst); + } + + static void YuvToBgra (byte y, byte u, byte v, byte[] bgra, int dst) + { + bgra[dst] = YUVToB (y, u); + bgra[dst+1] = YUVToG (y, u, v); + bgra[dst+2] = YUVToR (y, v); + bgra[dst+3] = 0xFF; + } + + static byte YUVToR(int y, int v) + { + return Clip8 (MultHi (y, 19077) + MultHi (v, 26149) - 14234); + } + + static byte YUVToG (int y, int u, int v) + { + return Clip8 (MultHi (y, 19077) - MultHi (u, 6419) - MultHi (v, 13320) + 8708); + } + + static byte YUVToB(int y, int u) + { + return Clip8 (MultHi (y, 19077) + MultHi (u, 33050) - 17685); + } + + static byte Clip8 (int v) + { + return (byte)(((v & ~YUV_MASK2) == 0) ? (v >> YUV_FIX2) : (v < 0) ? 0 : 255); + } + + static int MultHi (int v, int coeff) // _mm_mulhi_epu16 emulation + { + return (v * coeff) >> 8; + } + + const int YUV_FIX = 16; // fixed-point precision for RGB->YUV + const int YUV_HALF = 1 << (YUV_FIX - 1); + const int YUV_MASK = (256 << YUV_FIX) - 1; + const int YUV_RANGE_MIN = -227; // min value of r/g/b output + const int YUV_RANGE_MAX = 256 + 226; // max value of r/g/b output + + const int YUV_FIX2 = 6; // fixed-point precision for YUV->RGB + const int YUV_HALF2 = 1 << YUV_FIX2 >> 1; + const int YUV_MASK2 = (256 << YUV_FIX2) - 1; + } +} diff --git a/ArcFormats/WebP/Filters.cs b/ArcFormats/WebP/Filters.cs new file mode 100644 index 00000000..ebf05ff9 --- /dev/null +++ b/ArcFormats/WebP/Filters.cs @@ -0,0 +1,238 @@ +//! \file Filters.cs +//! \date Thu May 19 21:26:22 2016 +//! \brief Google WEBP filter functions. +/* +Copyright (c) 2010, Google Inc. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + + * Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in + the documentation and/or other materials provided with the + distribution. + + * Neither the name of Google nor the names of its contributors may + be used to endorse or promote products derived from this software + without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +*/ +// +// C# port by morkt (C) 2016 +// + +namespace GameRes.Formats.Google +{ + internal static class WebpFilter // Filter types. + { + public const int None = 0; + public const int Horizontal = 1; + public const int Vertical = 2; + public const int Gradient = 3; + public const int Last = Gradient + 1; // end marker + public const int Best = Last + 1; // meta-types + public const int Fast = Best + 1; + + static void PredictLine (byte[] input, int src, byte[] preds, int pred, + byte[] output, int dst, int length, bool inverse) + { + if (inverse) + { + for (int i = 0; i < length; ++i) + output[dst+i] = (byte)(input[src+i] + preds[pred+i]); + } + else + { + for (int i = 0; i < length; ++i) + output[dst+i] = (byte)(input[src+i] - preds[pred+i]); + } + } + + //------------------------------------------------------------------------------ + // Horizontal filter. + + static void DoHorizontalFilter (byte[] input, int src, int width, int height, + int stride, int row, int num_rows, bool inverse, + byte[] output, int dst) + { + int start_offset = row * stride; + int last_row = row + num_rows; + src += start_offset; + dst += start_offset; + var preds = inverse ? output : input; + var pred = inverse ? dst : src; + + if (0 == row) + { + // Leftmost pixel is the same as input for topmost scanline. + output[dst] = input[src]; + PredictLine (input, src + 1, preds, pred, output, dst + 1, width - 1, inverse); + row = 1; + pred += stride; + src += stride; + dst += stride; + } + + // Filter line-by-line. + while (row < last_row) + { + // Leftmost pixel is predicted from above. + PredictLine (input, src, preds, pred - stride, output, dst, 1, inverse); + PredictLine (input, src + 1, preds, pred, output, dst + 1, width - 1, inverse); + ++row; + pred += stride; + src += stride; + dst += stride; + } + } + + //------------------------------------------------------------------------------ + // Vertical filter. + + static void DoVerticalFilter (byte[] input, int src, int width, int height, + int stride, int row, int num_rows, bool inverse, + byte[] output, int dst) + { + int start_offset = row * stride; + int last_row = row + num_rows; + src += start_offset; + dst += start_offset; + var preds = inverse ? output : input; + int pred = inverse ? dst : src; + + if (0 == row) + { + // Very first top-left pixel is copied. + output[dst] = input[src]; + // Rest of top scan-line is left-predicted. + PredictLine (input, src + 1, preds, pred, output, dst + 1, width - 1, inverse); + row = 1; + src += stride; + dst += stride; + } + else + { + // We are starting from in-between. Make sure 'preds' points to prev row. + pred -= stride; + } + + // Filter line-by-line. + while (row < last_row) + { + PredictLine (input, src, preds, pred, output, dst, width, inverse); + ++row; + pred += stride; + src += stride; + dst += stride; + } + } + + //------------------------------------------------------------------------------ + // Gradient filter. + + static int GradientPredictor (byte a, byte b, byte c) + { + int g = a + b - c; + return ((g & ~0xFF) == 0) ? g : (g < 0) ? 0 : 255; // clip to 8bit + } + + static void DoGradientFilter (byte[] input, int src, int width, int height, + int stride, int row, int num_rows, bool inverse, + byte[] output, int dst) + { + int start_offset = row * stride; + int last_row = row + num_rows; + src += start_offset; + dst += start_offset; + var preds = inverse ? output : input; + int pred = inverse ? dst : src; + + // left prediction for top scan-line + if (0 == row) + { + output[dst] = input[src]; + PredictLine (input, src + 1, preds, pred, output, dst + 1, width - 1, inverse); + row = 1; + pred += stride; + src += stride; + dst += stride; + } + + // Filter line-by-line. + while (row < last_row) + { + // leftmost pixel: predict from above. + PredictLine (input, src, preds, pred - stride, output, dst, 1, inverse); + for (int w = 1; w < width; ++w) + { + int p = GradientPredictor (preds[pred + w - 1], preds[pred + w - stride], preds[pred + w - stride - 1]); + output[dst+w] = (byte)(input[src+w] + (inverse ? p : -p)); + } + ++row; + pred += stride; + src += stride; + dst += stride; + } + } + + //------------------------------------------------------------------------------ + + public static void HorizontalFilter (byte[] data, int src, int width, int height, + int stride, byte[] filtered_data, int dst) + { + DoHorizontalFilter (data, src, width, height, stride, 0, height, false, filtered_data, dst); + } + + public static void VerticalFilter (byte[] data, int src, int width, int height, + int stride, byte[] filtered_data, int dst) + { + DoVerticalFilter (data, src, width, height, stride, 0, height, false, filtered_data, dst); + } + + + public static void GradientFilter (byte[] data, int src, int width, int height, + int stride, byte[] filtered_data, int dst) + { + DoGradientFilter (data, src, width, height, stride, 0, height, false, filtered_data, dst); + } + + + //------------------------------------------------------------------------------ + + delegate void UnfilterFunc (int width, int height, int stride, int row, + int num_rows, byte[] data, int dst); + + public static void VerticalUnfilter (int width, int height, int stride, int row, + int num_rows, byte[] data, int src) + { + DoVerticalFilter (data, src, width, height, stride, row, num_rows, true, data, src); + } + + public static void HorizontalUnfilter (int width, int height, int stride, int row, + int num_rows, byte[] data, int src) + { + DoHorizontalFilter (data, src, width, height, stride, row, num_rows, true, data, src); + } + + public static void GradientUnfilter (int width, int height, int stride, int row, + int num_rows, byte[] data, int src) + { + DoGradientFilter (data, src, width, height, stride, row, num_rows, true, data, src); + } + } +} diff --git a/ArcFormats/WebP/Huffman.cs b/ArcFormats/WebP/Huffman.cs new file mode 100644 index 00000000..1732a056 --- /dev/null +++ b/ArcFormats/WebP/Huffman.cs @@ -0,0 +1,272 @@ +//! \file Huffman.cs +//! \date Wed May 18 22:19:23 2016 +//! \brief Google WEBP Huffman compression implementaion. +/* +Copyright (c) 2010, Google Inc. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + + * Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in + the documentation and/or other materials provided with the + distribution. + + * Neither the name of Google nor the names of its contributors may + be used to endorse or promote products derived from this software + without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +*/ +// +// C# port by morkt (C) 2016 +// + +namespace GameRes.Formats.Google +{ + static class Huffman + { + public const int CodesPerMetaCode = 5; + public const int PackedBits = 6; + public const int PackedTableSize = 1 << PackedBits; + + public const int DefaultCodeLength = 8; + public const int MaxAllowedCodeLength = 15; + + public const int NumLiteralCodes = 256; + public const int NumLengthCodes = 24; + public const int NumDistanceCodes = 40; + public const int CodeLengthCodes = 19; + + public const int MinBits = 2; // min number of Huffman bits + public const int MaxBits = 9; // max number of Huffman bits + + public const int TableBits = 8; + public const int TableMask = (1 << TableBits) - 1; + + public const int LengthsTableBits = 7; + public const int LengthsTableMask = (1 << LengthsTableBits) - 1; + + static uint GetNextKey (uint key, int len) + { + uint step = 1u << (len - 1); + while (0 != (key & step)) + step >>= 1; + return (key & (step - 1)) + step; + } + + public static int BuildTable (HuffmanCode[] root_table, int index, int root_bits, int[] code_lengths, int code_lengths_size) + { + int table = index; // next available space in table + int total_size = 1 << root_bits; // total size root table + 2nd level table + int len; // current code length + int symbol; // symbol index in original or sorted table + // number of codes of each length: + int[] count = new int[MaxAllowedCodeLength + 1]; + // offsets in sorted table for each length: + int[] offset = new int[MaxAllowedCodeLength + 1]; + + // Build histogram of code lengths. + for (symbol = 0; symbol < code_lengths_size; ++symbol) + { + if (code_lengths[symbol] > MaxAllowedCodeLength) + return 0; + ++count[code_lengths[symbol]]; + } + + // Error, all code lengths are zeros. + if (count[0] == code_lengths_size) + return 0; + + // Generate offsets into sorted symbol table by code length. + offset[1] = 0; + for (len = 1; len < MaxAllowedCodeLength; ++len) + { + if (count[len] > (1 << len)) + return 0; + offset[len + 1] = offset[len] + count[len]; + } + + var sorted = new int[code_lengths_size]; + + // Sort symbols by length, by symbol order within each length. + for (symbol = 0; symbol < code_lengths_size; ++symbol) + { + int symbol_code_length = code_lengths[symbol]; + if (code_lengths[symbol] > 0) + sorted[offset[symbol_code_length]++] = symbol; + } + + // Special case code with only one value. + if (offset[MaxAllowedCodeLength] == 1) + { + HuffmanCode code; + code.bits = 0; + code.value = (ushort)sorted[0]; + ReplicateValue (root_table, table, 1, total_size, code); + return total_size; + } + + int step; // step size to replicate values in current table + uint low = uint.MaxValue; // low bits for current root entry + uint mask = (uint)total_size - 1; // mask for low bits + uint key = 0; // reversed prefix code + int num_nodes = 1; // number of Huffman tree nodes + int num_open = 1; // number of open branches in current tree level + int table_bits = root_bits; // key length of current table + int table_size = 1 << table_bits; // size of current table + symbol = 0; + // Fill in root table. + for (len = 1, step = 2; len <= root_bits; ++len, step <<= 1) + { + num_open <<= 1; + num_nodes += num_open; + num_open -= count[len]; + if (num_open < 0) + return 0; + for (; count[len] > 0; --count[len]) + { + HuffmanCode code; + code.bits = (byte)len; + code.value = (ushort)sorted[symbol++]; + ReplicateValue (root_table, table + (int)key, step, table_size, code); + key = GetNextKey (key, len); + } + } + + // Fill in 2nd level tables and add pointers to root table. + for (len = root_bits + 1, step = 2; len <= MaxAllowedCodeLength; ++len, step <<= 1) + { + num_open <<= 1; + num_nodes += num_open; + num_open -= count[len]; + if (num_open < 0) + return 0; + for (; count[len] > 0; --count[len]) + { + HuffmanCode code; + if ((key & mask) != low) + { + table += table_size; + table_bits = NextTableBitSize (count, len, root_bits); + table_size = 1 << table_bits; + total_size += table_size; + low = key & mask; + root_table[index+low].bits = (byte)(table_bits + root_bits); + root_table[index+low].value = (ushort)(table - index - low); + } + code.bits = (byte)(len - root_bits); + code.value = (ushort)sorted[symbol++]; + ReplicateValue (root_table, table + (int)(key >> root_bits), step, table_size, code); + key = GetNextKey (key, len); + } + } + + // Check if tree is full. + if (num_nodes != 2 * offset[MaxAllowedCodeLength] - 1) + return 0; + + return total_size; + } + + static void ReplicateValue (HuffmanCode[] table, int offset, int step, int end, HuffmanCode code) + { + do + { + end -= step; + table[offset+end] = code; + } + while (end > 0); + } + + static int NextTableBitSize (int[] count, int len, int root_bits) + { + int left = 1 << (len - root_bits); + while (len < MaxAllowedCodeLength) + { + left -= count[len]; + if (left <= 0) break; + ++len; + left <<= 1; + } + return len - root_bits; + } + } + + internal struct HuffmanCode + { + public byte bits; // number of bits used for this symbol + public ushort value; // symbol value or table offset + } + + internal struct HuffmanCode32 + { + public int bits; // number of bits used for this symbol, + // or an impossible value if not a literal code. + public uint value; // 32b packed ARGB value if literal, + // or non-literal symbol otherwise + } + + internal class HTreeGroup + { + HuffmanCode[] tables; + int[] htrees = new int[Huffman.CodesPerMetaCode]; + public bool is_trivial_literal; // True, if huffman trees for Red, Blue & Alpha + // Symbols are trivial (have a single code). + public uint literal_arb; // If is_trivial_literal is true, this is the + // ARGB value of the pixel, with Green channel + // being set to zero. + public bool is_trivial_code; // true if is_trivial_literal with only one code + public bool use_packed_table; // use packed table below for short literal code + // table mapping input bits to a packed values, or escape case to literal code + public HuffmanCode32[] packed_table = new HuffmanCode32[Huffman.PackedTableSize]; + + public HuffmanCode[] Tables { get { return tables; } } + + public void SetMeta (int meta, int base_index) + { + htrees[meta] = base_index; + } + + public int GetMeta (int meta) + { + return htrees[meta]; + } + + public HuffmanCode GetCode (int meta, int index) + { + return tables[htrees[meta] + index]; + } + + public void SetCode (int meta, int index, HuffmanCode code) + { + tables[htrees[meta] + index] = code; + } + + public static HTreeGroup[] New (int num_htree_groups, int table_size) + { + var tables = new HuffmanCode[num_htree_groups * table_size]; + var htree_groups = new HTreeGroup[num_htree_groups]; + for (int i = 0; i < num_htree_groups; ++i) + { + htree_groups[i] = new HTreeGroup(); + htree_groups[i].tables = tables; + } + return htree_groups; + } + } +} diff --git a/ArcFormats/WebP/ImageWEBP.cs b/ArcFormats/WebP/ImageWEBP.cs new file mode 100644 index 00000000..a2d07174 --- /dev/null +++ b/ArcFormats/WebP/ImageWEBP.cs @@ -0,0 +1,145 @@ +//! \file ImageWEBP.cs +//! \date Wed Apr 06 07:16:39 2016 +//! \brief Google WEBP image format. +// +// Copyright (C) 2016 by morkt +// +// 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. +// + +using System; +using System.ComponentModel.Composition; +using System.IO; +using System.Windows.Media; +using GameRes.Utility; + +namespace GameRes.Formats.Google +{ + internal class WebPMetaData : ImageMetaData + { + public WebPFeature Flags; + public bool IsLossless; + public long DataOffset; + public int DataSize; + public long AlphaOffset; + public int AlphaSize; + } + + [Flags] + internal enum WebPFeature : uint + { + Fragments = 0x0001, + Animation = 0x0002, + Xmp = 0x0004, + Exif = 0x0008, + Alpha = 0x0010, + Iccp = 0x0020, + } + + [Export(typeof(ImageFormat))] + public class WebPFormat : ImageFormat + { + public override string Tag { get { return "WEBP"; } } + public override string Description { get { return "Google WebP image format"; } } + public override uint Signature { get { return 0; } } + + public override ImageMetaData ReadMetaData (Stream stream) + { + if (0x46464952 != FormatCatalog.ReadSignature (stream)) // 'RIFF' + return null; + var header = new byte[0x10]; + if (8 != stream.Read (header, 0, 8)) + return null; + if (!Binary.AsciiEqual (header, 4, "WEBP")) + return null; + bool found_vp8x = false; + var info = new WebPMetaData(); + int chunk_size; + for (;;) + { + if (8 != stream.Read (header, 0, 8)) + return null; + chunk_size = LittleEndian.ToInt32 (header, 4); + int aligned_size = (chunk_size + 1) & ~1; + if (!found_vp8x && Binary.AsciiEqual (header, 0, "VP8X")) + { + found_vp8x = true; + if (chunk_size < 10) + return null; + if (chunk_size > header.Length) + header = new byte[chunk_size]; + if (chunk_size != stream.Read (header, 0, chunk_size)) + return null; + info.Flags = (WebPFeature)LittleEndian.ToUInt32 (header, 0); + info.Width = 1 + GetUInt24 (header, 4); + info.Height = 1 + GetUInt24 (header, 7); + if ((long)info.Width * info.Height >= (1L << 32)) + return null; + continue; + } + if (Binary.AsciiEqual (header, 0, "VP8 ") || Binary.AsciiEqual (header, 0, "VP8L")) + { + info.IsLossless = header[3] == 'L'; + info.DataOffset = stream.Position; + info.DataSize = chunk_size; + if (!found_vp8x) + { + if (chunk_size < 10 || 10 != stream.Read (header, 0, 10)) + return null; + if (header[3] != 0x9D || header[4] != 1 || header[5] != 0x2A) + return null; + if (0 != (header[0] & 1)) // not a keyframe + return null; + info.Width = LittleEndian.ToUInt16 (header, 6) & 0x3FFFu; + info.Height = LittleEndian.ToUInt16 (header, 8) & 0x3FFFu; + } + break; + } + if (Binary.AsciiEqual (header, 0, "ALPH")) + { + info.AlphaOffset = stream.Position; + info.AlphaSize = chunk_size; + } + stream.Seek (aligned_size, SeekOrigin.Current); + } + if (0 == info.Width || 0 == info.Height) + return null; + return info; + } + + static uint GetUInt24 (byte[] src, int offset) + { + return (uint)(src[offset] | src[offset+1] << 8 | src[offset+2] << 16); + } + + public override ImageData Read (Stream stream, ImageMetaData info) + { + using (var reader = new WebPDecoder (stream, (WebPMetaData)info)) + { + reader.Decode(); + return ImageData.Create (info, reader.Format, null, reader.Output); + } + } + + public override void Write (Stream file, ImageData image) + { + throw new NotImplementedException ("WebPFormat.Write not implemented"); + } + } +} diff --git a/ArcFormats/WebP/Lossless.cs b/ArcFormats/WebP/Lossless.cs new file mode 100644 index 00000000..e2d52a42 --- /dev/null +++ b/ArcFormats/WebP/Lossless.cs @@ -0,0 +1,1754 @@ +//! \file Lossless.cs +//! \date Wed May 18 20:10:59 2016 +//! \brief Google WEBP lossless compression decoder. +/* +Copyright (c) 2010, Google Inc. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + + * Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in + the documentation and/or other materials provided with the + distribution. + + * Neither the name of Google nor the names of its contributors may + be used to endorse or promote products derived from this software + without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +*/ +// +// C# port by morkt (C) 2016 +// + +using System; +using GameRes.Utility; + +namespace GameRes.Formats.Google +{ + enum VP8StatusCode + { + Ok = 0, + OutOfMemory, + InvalidParam, + BitstreamError, + UnsupportedFeature, + Suspended, + UserAbort, + NotEnoughData + } + + enum VP8DecodeState + { + ReadData = 0, + ReadHdr = 1, + ReadDim = 2 + } + + enum VP8LImageTransformType + { + Predictor, + CrossColor, + SubtractGreen, + ColorIndexing + } + + internal struct VP8LMultipliers + { + public byte green_to_red_; + public byte green_to_blue_; + public byte red_to_blue_; + + public void Reset () + { + green_to_red_ = 0; + green_to_blue_ = 0; + red_to_blue_ = 0; + } + + public void ColorCodeToMultipliers (uint color_code) + { + green_to_red_ = (byte)color_code; + green_to_blue_ = (byte)(color_code >> 8); + red_to_blue_ = (byte)(color_code >> 16); + } + + static uint ColorTransformDelta (sbyte color_pred, sbyte color) + { + return (uint)((int)color_pred * color) >> 5; + } + + public void TransformColorInverse (uint[] data, int offset, int num_pixels) + { + for (int i = 0; i < num_pixels; ++i) + { + uint argb = data[offset+i]; + uint green = argb >> 8; + uint red = argb >> 16; + uint new_red = red; + uint new_blue = argb; + new_red += ColorTransformDelta ((sbyte)green_to_red_, (sbyte)green); + new_red &= 0xFF; + new_blue += ColorTransformDelta ((sbyte)green_to_blue_, (sbyte)green); + new_blue += ColorTransformDelta ((sbyte)red_to_blue_, (sbyte)new_red); + new_blue &= 0xFF; + data[offset+i] = (argb & 0xFF00FF00u) | (new_red << 16) | (new_blue); + } + } + } + + internal class VP8LTransform + { + public VP8LImageTransformType type_; // transform type. + public int bits_; // subsampling bits defining transform window. + public int xsize_; // transform window X index. + public int ysize_; // transform window Y index. + public uint[] data_; // transform data. + + public void InverseTransform (int row_start, int row_end, + uint[] input, int src, uint[] output, int dst) + { + int width = xsize_; + switch (type_) + { + case VP8LImageTransformType.SubtractGreen: + AddGreenToBlueAndRed (output, dst, (row_end - row_start) * width); + break; + + case VP8LImageTransformType.Predictor: + PredictorInverseTransform (row_start, row_end, output, dst); + if (row_end != ysize_) + { + // The last predicted row in this iteration will be the top-pred row + // for the first row in next iteration. + Buffer.BlockCopy (output, dst + (row_end - row_start - 1) * width, + output, dst - width, width * sizeof(uint)); + } + break; + + case VP8LImageTransformType.CrossColor: + ColorSpaceInverseTransform (row_start, row_end, output, dst); + break; + + case VP8LImageTransformType.ColorIndexing: + if (input == output && src == dst && bits_ > 0) + { + // Move packed pixels to the end of unpacked region, so that unpacking + // can occur seamlessly. + // Also, note that this is the only transform that applies on + // the effective width of VP8LSubSampleSize(xsize_, bits_). All other + // transforms work on effective width of xsize_. + int out_stride = (row_end - row_start) * width; + int in_stride = (row_end - row_start) * LosslessDecoder.SubSampleSize (xsize_, bits_); + src += out_stride - in_stride; + Buffer.BlockCopy (output, dst, input, src, in_stride * sizeof(uint)); +// memmove(src, out, in_stride * sizeof(*src)); + ColorIndexInverseTransform (row_start, row_end, input, src, output, dst); + } + else + { + ColorIndexInverseTransform (row_start, row_end, input, src, output, dst); + } + break; + } + } + + static byte GetAlphaValue (uint val) + { + return (byte)(val >> 8); + } + + static void MapARGB (uint[] input, int src, uint[] color_map, + uint[] output, int dst, int y_start, int y_end, int width) + { + for (int y = y_start; y < y_end; ++y) + for (int x = 0; x < width; ++x) + output[dst++] = color_map[(input[src++] >> 8) & 0xFF]; + } + + static void MapAlpha (byte[] input, int src, uint[] color_map, + byte[] output, int dst, int y_start, int y_end, int width) + { + for (int y = y_start; y < y_end; ++y) + for (int x = 0; x < width; ++x) + output[dst++] = GetAlphaValue (color_map[input[src++]]); + } + + void ColorIndexInverseTransform (int y_start, int y_end, uint[] input, int src, + uint[] output, int dst) + { + int bits_per_pixel = 8 >> bits_; + int width = xsize_; + if (bits_per_pixel < 8) + { + int pixels_per_byte = 1 << bits_; + int count_mask = pixels_per_byte - 1; + uint bit_mask = (1u << bits_per_pixel) - 1u; + for (int y = y_start; y < y_end; ++y) + { + uint packed_pixels = 0; + for (int x = 0; x < width; ++x) + { + // We need to load fresh 'packed_pixels' once every + // 'pixels_per_byte' increments of x. Fortunately, pixels_per_byte + // is a power of 2, so can just use a mask for that, instead of + // decrementing a counter. + if ((x & count_mask) == 0) + packed_pixels = (input[src++] >> 8) & 0xFF; // GetARGBIndex + output[dst++] = data_[packed_pixels & bit_mask]; // GetARGBValue + packed_pixels >>= bits_per_pixel; + } + } + } + else + { + MapARGB (input, src, data_, output, dst, y_start, y_end, width); + } + } + + public void ColorIndexInverseTransformAlpha (int y_start, int y_end, byte[] input, int src, byte[] output, int dst) + { + int bits_per_pixel = 8 >> bits_; + int width = xsize_; + var color_map = data_; + if (bits_per_pixel < 8) + { + int pixels_per_byte = 1 << bits_; + int count_mask = pixels_per_byte - 1; + uint bit_mask = (1u << bits_per_pixel) - 1u; + for (int y = y_start; y < y_end; ++y) + { + uint packed_pixels = 0; + for (int x = 0; x < width; ++x) + { + if ((x & count_mask) == 0) + packed_pixels = input[src++]; + output[dst++] = GetAlphaValue (color_map[packed_pixels & bit_mask]); + packed_pixels >>= bits_per_pixel; + } + } + } + else + { + MapAlpha (input, src, color_map, output, dst, y_start, y_end, width); + } + } + + void AddGreenToBlueAndRed (uint[] data, int offset, int num_pixels) + { + for (int i = 0; i < num_pixels; ++i) + { + uint argb = data[i]; + uint green = (argb >> 8) & 0xFFu; + uint red_blue = argb & 0x00FF00FFu; + red_blue += (green << 16) | green; + red_blue &= 0x00FF00FFu; + data[i] = (argb & 0xFF00FF00u) | red_blue; + } + } + + void PredictorInverseTransform (int y_start, int y_end, uint[] data, int offset) + { + int width = xsize_; + if (y_start == 0) // First Row follows the L (mode=1) mode. + { + uint pred0 = Predictor0 (data[offset-1], data, offset-1); + AddPixelsEq (data, offset, pred0); + for (int x = 1; x < width; ++x) + { + uint pred1 = Predictor1 (data[offset + x - 1], data, offset-1); + AddPixelsEq (data, offset + x, pred1); + } + offset += width; + ++y_start; + } + + int y = y_start; + int tile_width = 1 << bits_; + int mask = tile_width - 1; + int safe_width = width & ~mask; + int tiles_per_row = LosslessDecoder.SubSampleSize (width, bits_); + int pred_mode_base = (y >> bits_) * tiles_per_row; // within data_ + + while (y < y_end) + { + uint pred2 = Predictor2 (data[offset-1], data, offset - width); + int pred_mode_src = pred_mode_base; + int x = 1; + int t = 1; + // First pixel follows the T (mode=2) mode. + AddPixelsEq (data, offset, pred2); + // .. the rest: + while (x < safe_width) + { + var pred_func = kPredictors[(data_[pred_mode_src++] >> 8) & 0xF]; + for (; t < tile_width; ++t, ++x) + { + uint pred = pred_func (data[offset + x - 1], data, offset + x - width); + AddPixelsEq (data, offset + x, pred); + } + t = 0; + } + if (x < width) + { + var pred_func = kPredictors[(data_[pred_mode_src++] >> 8) & 0xF]; + for (; x < width; ++x) + { + uint pred = pred_func (data[offset + x - 1], data, offset + x - width); + AddPixelsEq (data, offset + x, pred); + } + } + offset += width; + ++y; + if ((y & mask) == 0) // Use the same mask, since tiles are squares. + pred_mode_base += tiles_per_row; + } + } + + void ColorSpaceInverseTransform (int y_start, int y_end, uint[] data, int offset) + { + int width = xsize_; + int tile_width = 1 << bits_; + int mask = tile_width - 1; + int safe_width = width & ~mask; + int remaining_width = width - safe_width; + int tiles_per_row = LosslessDecoder.SubSampleSize (width, bits_); + int y = y_start; + int pred_row = (y >> bits_) * tiles_per_row; // within data_ + + var m = new VP8LMultipliers(); + while (y < y_end) + { + m.Reset(); + int pred = pred_row; + int data_safe_end = offset + safe_width; + int data_end = offset + width; + while (offset < data_safe_end) + { + m.ColorCodeToMultipliers (data_[pred++]); + m.TransformColorInverse (data, offset, tile_width); + offset += tile_width; + } + if (offset < data_end) + { + m.ColorCodeToMultipliers (data_[pred++]); + m.TransformColorInverse (data, offset, remaining_width); + offset += remaining_width; + } + ++y; + if ((y & mask) == 0) + pred_row += tiles_per_row; + } + } + + //------------------------------------------------------------------------------ + // Predictors + + static uint Predictor0 (uint left, uint[] data, int top) + { + return 0xFF000000u; // ARGB_BLACK + } + static uint Predictor1 (uint left, uint[] data, int top) + { + return left; + } + static uint Predictor2 (uint left, uint[] data, int top) + { + return data[top]; + } + static uint Predictor3 (uint left, uint[] data, int top) + { + return data[top+1]; + } + static uint Predictor4 (uint left, uint[] data, int top) + { + return data[top-1]; + } + static uint Predictor5 (uint left, uint[] data, int top) + { + return Average3 (left, data[top], data[top+1]); + } + static uint Predictor6 (uint left, uint[] data, int top) + { + return Average2 (left, data[top-1]); + } + static uint Predictor7 (uint left, uint[] data, int top) + { + return Average2 (left, data[top]); + } + static uint Predictor8 (uint left, uint[] data, int top) + { + return Average2 (data[top-1], data[top]); + } + static uint Predictor9 (uint left, uint[] data, int top) + { + return Average2 (data[top], data[top+1]); + } + static uint Predictor10 (uint left, uint[] data, int top) + { + return Average4 (left, data[top-1], data[top], data[top+1]); + } + static uint Predictor11(uint left, uint[] data, int top) + { + return Select (data[top], data[left], data[top-1]); + } + static uint Predictor12 (uint left, uint[] data, int top) + { + return ClampedAddSubtractFull (left, data[top], data[top-1]); + } + static uint Predictor13 (uint left, uint[] data, int top) + { + return ClampedAddSubtractHalf (left, data[top], data[top-1]); + } + + delegate uint PredictorFunc (uint left, uint[] data, int top); + static readonly PredictorFunc[] kPredictors = new PredictorFunc[16] + { + Predictor0, Predictor1, Predictor2, Predictor3, + Predictor4, Predictor5, Predictor6, Predictor7, + Predictor8, Predictor9, Predictor10, Predictor11, + Predictor12, Predictor13, Predictor0, Predictor0, + }; + + //------------------------------------------------------------------------------ + // Image transforms. + + // In-place sum of each component with mod 256. + static void AddPixelsEq (uint[] data, int a, uint b) + { + data[a] = MMX.PAddB (data[a], b); + } + + static uint Average2 (uint a0, uint a1) + { + return (((a0 ^ a1) & 0xFEFEFEFEu) >> 1) + (a0 & a1); + } + + static uint Average3 (uint a0, uint a1, uint a2) + { + return Average2 (Average2 (a0, a2), a1); + } + + static uint Average4 (uint a0, uint a1, uint a2, uint a3) + { + return Average2 (Average2 (a0, a1), Average2 (a2, a3)); + } + + static int Sub3 (int a, int b, int c) + { + return Math.Abs (b - c) - Math.Abs(a - c); + } + + static uint Select (uint a, uint b, uint c) + { + int pa_minus_pb = + Sub3 ((int)(a >> 24) , (int)(b >> 24) , (int)(c >> 24) ) + + Sub3 ((int)(a >> 16) & 0xFF, (int)(b >> 16) & 0xFF, (int)(c >> 16) & 0xFF) + + Sub3 ((int)(a >> 8) & 0xFF, (int)(b >> 8) & 0xFF, (int)(c >> 8) & 0xFF) + + Sub3 ((int)(a ) & 0xFF, (int)(b ) & 0xFF, (int)(c ) & 0xFF); + return (pa_minus_pb <= 0) ? a : b; + } + + static uint Clip255 (uint a) + { + if (a < 256) + return a; + // return 0, when a is a negative integer. + // return 255, when a is positive. + return ~a >> 24; + } + + static uint AddSubtractComponentFull (uint a, uint b, uint c) + { + return Clip255 ((uint)((int)a + (int)b - (int)c)); + } + + static uint ClampedAddSubtractFull (uint c0, uint c1, uint c2) + { + uint a = AddSubtractComponentFull (c0 >> 24, c1 >> 24, c2 >> 24); + uint r = AddSubtractComponentFull ((c0 >> 16) & 0xFF, + (c1 >> 16) & 0xFF, + (c2 >> 16) & 0xFF); + uint g = AddSubtractComponentFull ((c0 >> 8) & 0xFF, + (c1 >> 8) & 0xFF, + (c2 >> 8) & 0xFF); + uint b = AddSubtractComponentFull (c0 & 0xFF, c1 & 0xFF, c2 & 0xFF); + return (a << 24) | (r << 16) | (g << 8) | b; + } + + static uint AddSubtractComponentHalf (uint a, uint b) + { + return Clip255 ((uint)((int)a + ((int)a - (int)b) / 2)); + } + + static uint ClampedAddSubtractHalf (uint c0, uint c1, uint c2) + { + uint ave = Average2 (c0, c1); + uint a = AddSubtractComponentHalf (ave >> 24, c2 >> 24); + uint r = AddSubtractComponentHalf ((ave >> 16) & 0xFF, (c2 >> 16) & 0xFF); + uint g = AddSubtractComponentHalf ((ave >> 8) & 0xFF, (c2 >> 8) & 0xFF); + uint b = AddSubtractComponentHalf ((ave >> 0) & 0xFF, (c2 >> 0) & 0xFF); + return (a << 24) | (r << 16) | (g << 8) | b; + } + } + + internal static class HuffIndex + { + public const int Green = 0; + public const int Red = 1; + public const int Blue = 2; + public const int Alpha = 3; + public const int Dist = 4; + } + + internal class LosslessDecoder + { + const uint kHashMul = 0x1E35A7BDu; + const int NumARGBCacheRows = 16; + const int MaxCacheBits = 11; + const int NumTransforms = 4; + const int SyncEveryNRows = 8; // minimum number of rows between check-points + const int BitsSpecialMarker = 0x100; // something large enough (and a bit-mask) + const int PackedNonLiteralCode = 0; // must be < NUM_LITERAL_CODES + + const int kCodeLengthLiterals = 16; + const int kCodeLengthRepeatCode = 16; + static readonly int[] kCodeLengthExtraBits = { 2, 3, 7 }; + static readonly int[] kCodeLengthRepeatOffsets = { 3, 3, 11 }; + + public delegate void ProcessRowsFunc (LosslessDecoder dec, int row); + + VP8StatusCode status_; + VP8DecodeState state_; + public VP8Io io_; + + byte[] pixels8_; + uint[] pixels32_; // Internal data: either uint8_t* for alpha + // or uint32_t* for BGRA. + int argb_cache_; // Scratch buffer for temporary BGRA storage. + + LBitReader br_ = new LBitReader(); + bool incremental_ = false; // if true, incremental decoding is expected + LBitReader saved_br_ = new LBitReader(); // note: could be local variables too + int saved_last_pixel_; + + int width_; + int height_; + public int last_row_; // last input row decoded so far. + public int last_pixel_; // last pixel decoded so far. However, it may + // not be transformed, scaled and + // color-converted yet. + int last_out_row_; // last row output so far. + + public uint[] Pixels { get { return pixels32_; } } + public int Width + { + get { return width_; } + set { width_ = value; } + } + public int Height + { + get { return height_; } + set { height_ = value; } + } + + LMetadata hdr_ = new LMetadata(); + + public int next_transform_; + public VP8LTransform[] transforms_ = new VP8LTransform[NumTransforms]; + // or'd bitset storing the transforms types. + uint transforms_seen_; + + public LosslessDecoder () + { + status_ = VP8StatusCode.Ok; + state_ = VP8DecodeState.ReadDim; + + for (int i = 0; i < transforms_.Length; ++i) + transforms_[i] = new VP8LTransform(); + } + + public void Init (int width, int height, VP8Io io, + byte[] data, int data_i, int data_size, byte[] output) + { + width_ = width; + height_ = height; + status_ = VP8StatusCode.Ok; + io_ = io; + + io_.opaque = output; + io_.width = width_; + io_.height = height_; + + br_.Init (data, data_i, (uint)data_size); + } + + public bool Is8bOptimizable () + { + if (hdr_.color_cache_size_ > 0) + return false; + // When the Huffman tree contains only one symbol, we can skip the + // call to ReadSymbol() for red/blue/alpha channels. + for (int i = 0; i < hdr_.num_htree_groups_; ++i) + { + var htree_group = hdr_.htree_groups_[i]; + if (htree_group.GetCode (HuffIndex.Red, 0).bits > 0) return false; + if (htree_group.GetCode (HuffIndex.Blue, 0).bits > 0) return false; + if (htree_group.GetCode (HuffIndex.Alpha, 0).bits > 0) return false; + } + return true; + } + + //------------------------------------------------------------------------------ + + public bool DecodeImage () + { + // Initialization. + if (state_ != VP8DecodeState.ReadData) + { + AllocateInternalBuffers32b (io_.width); + if (incremental_) + { + if (hdr_.color_cache_size_ > 0 + && hdr_.saved_color_cache_.colors_ == null) + { + hdr_.saved_color_cache_.Init (hdr_.color_cache_.hash_bits_); + } + } + state_ = VP8DecodeState.ReadData; + } + + // Decode. + return DecodeImageData (pixels32_, width_, height_, height_, ProcessRows); + } + + // Processes (transforms, scales & color-converts) the rows decoded after the + // last call. + static void ProcessRows (LosslessDecoder dec, int row) + { + throw new NotImplementedException ("Lossless RGB decoder not implemented"); + } + + //------------------------------------------------------------------------------ + // Allocate internal buffers dec->pixels_ and dec->argb_cache_. + + public void AllocateInternalBuffers32b (int final_width) + { + int num_pixels = width_ * height_; + // Scratch buffer corresponding to top-prediction row for transforming the + // first row in the row-blocks. Not needed for paletted alpha. + int cache_top_pixels = (ushort)final_width; + // Scratch buffer for temporary BGRA storage. Not needed for paletted alpha. + int cache_pixels = final_width * NumARGBCacheRows; + int total_num_pixels = num_pixels + cache_top_pixels + cache_pixels; + + pixels32_ = new uint[total_num_pixels]; + argb_cache_ = num_pixels + cache_top_pixels; + } + + public void AllocateInternalBuffers8b () + { + int total_num_pixels = width_ * height_; + pixels8_ = new byte[total_num_pixels]; + argb_cache_ = 0; + } + + public bool DecodeImageStream (int xsize, int ysize, bool is_level0, ref uint[] decoded_data, bool set_data) + { + bool ok = true; + int transform_xsize = xsize; + int transform_ysize = ysize; + int color_cache_bits = 0; + uint[] data = null; + + // Read the transforms (may recurse). + if (is_level0) + { + while (ok && 0 != br_.ReadBits (1)) + ok = ReadTransform (ref transform_xsize, ref transform_ysize); + } + + // Color cache + if (ok && 0 != br_.ReadBits (1)) + { + color_cache_bits = (int)br_.ReadBits (4); + ok = (color_cache_bits >= 1 && color_cache_bits <= MaxCacheBits); + if (!ok) + { + status_ = VP8StatusCode.BitstreamError; + return false; + } + } + + // Read the Huffman codes (may recurse). + ok = ok && ReadHuffmanCodes (transform_xsize, transform_ysize, color_cache_bits, is_level0); + if (!ok) + { + status_ = VP8StatusCode.BitstreamError; + return false; + } + + // Finish setting up the color-cache + if (color_cache_bits > 0) + { + hdr_.color_cache_size_ = 1 << color_cache_bits; + hdr_.color_cache_.Init (color_cache_bits); + } + else + { + hdr_.color_cache_size_ = 0; + } + UpdateDecoder (transform_xsize, transform_ysize); + + if (is_level0) // level 0 complete + { + state_ = VP8DecodeState.ReadHdr; + } + else + { + var total_size = transform_xsize * transform_ysize; + data = new uint[total_size]; + + // Use the Huffman trees to decode the LZ77 encoded data. + ok = DecodeImageData (data, transform_xsize, transform_ysize, transform_ysize, null); + ok = ok && !br_.EoS; + } + if (ok) + { + if (set_data) + { + decoded_data = data; + } + last_pixel_ = 0; // Reset for future DECODE_DATA_FUNC() calls. + if (!is_level0) + hdr_.ClearMetadata(); + } + return ok; + } + + bool ReadTransform (ref int xsize, ref int ysize) + { + bool ok = true; + var transform = transforms_[next_transform_]; + var type = (VP8LImageTransformType)br_.ReadBits (2); + + // Each transform type can only be present once in the stream. + if (0 != (transforms_seen_ & (1U << (int)type))) + return false; // Already there, let's not accept the second same transform. + + transforms_seen_ |= (1U << (int)type); + + transform.type_ = type; + transform.xsize_ = xsize; + transform.ysize_ = ysize; + transform.data_ = null; + ++next_transform_; + + switch (type) + { + case VP8LImageTransformType.Predictor: + case VP8LImageTransformType.CrossColor: + transform.bits_ = (int)br_.ReadBits (3) + 2; + ok = DecodeImageStream (SubSampleSize (transform.xsize_, transform.bits_), + SubSampleSize (transform.ysize_, transform.bits_), + false, ref transform.data_, true); + break; + case VP8LImageTransformType.ColorIndexing: + int num_colors = (int)br_.ReadBits (8) + 1; + int bits = (num_colors > 16) ? 0 + : (num_colors > 4) ? 1 + : (num_colors > 2) ? 2 + : 3; + xsize = SubSampleSize (transform.xsize_, bits); + transform.bits_ = bits; + ok = DecodeImageStream (num_colors, 1, false, ref transform.data_, true); + ok = ok && ExpandColorMap (num_colors, transform); + break; + case VP8LImageTransformType.SubtractGreen: + break; + default: + throw new InvalidFormatException(); + } + return ok; + } + + bool ExpandColorMap (int num_colors, VP8LTransform transform) + { + int final_num_colors = 1 << (8 >> transform.bits_); + uint[] new_color_map = new uint[final_num_colors]; + new_color_map[0] = transform.data_[0]; + for (int i = 1; i < num_colors; ++i) + new_color_map[i] = MMX.PAddB (transform.data_[i], new_color_map[i-1]); + transform.data_ = new_color_map; + return true; + } + + bool ReadHuffmanCodes (int xsize, int ysize, int color_cache_bits, bool allow_recursion) + { + uint[] huffman_image = null; + int num_htree_groups = 1; + int max_alphabet_size = 0; + int table_size = kTableSize[color_cache_bits]; + + if (allow_recursion && 0 != br_.ReadBits (1)) + { + // use meta Huffman codes. + int huffman_precision = (int)br_.ReadBits (3) + 2; + int huffman_xsize = SubSampleSize (xsize, huffman_precision); + int huffman_ysize = SubSampleSize (ysize, huffman_precision); + int huffman_pixs = huffman_xsize * huffman_ysize; + if (!DecodeImageStream (huffman_xsize, huffman_ysize, false, ref huffman_image, true)) + return false; + hdr_.huffman_subsample_bits_ = huffman_precision; + for (int i = 0; i < huffman_pixs; ++i) + { + // The huffman data is stored in red and green bytes. + int group = (int)(huffman_image[i] >> 8) & 0xffff; + huffman_image[i] = (uint)group; + if (group >= num_htree_groups) + num_htree_groups = group + 1; + } + } + + if (br_.EoS) return false; + + // Find maximum alphabet size for the htree group. + for (int j = 0; j < Huffman.CodesPerMetaCode; ++j) + { + int alphabet_size = kAlphabetSize[j]; + if (j == 0 && color_cache_bits > 0) + alphabet_size += 1 << color_cache_bits; + if (max_alphabet_size < alphabet_size) + max_alphabet_size = alphabet_size; + } + + var htree_groups = HTreeGroup.New (num_htree_groups, table_size); + var huffman_tables = htree_groups[0].Tables; + var code_lengths = new int[max_alphabet_size]; + + int next = 0; + for (int i = 0; i < num_htree_groups; ++i) + { + var htree_group = htree_groups[i]; + int size; + int total_size = 0; + bool is_trivial_literal = true; + int max_bits = 0; + for (int j = 0; j < Huffman.CodesPerMetaCode; ++j) + { + int alphabet_size = kAlphabetSize[j]; + htree_group.SetMeta (j, next); + if (j == 0 && color_cache_bits > 0) + alphabet_size += 1 << color_cache_bits; + + size = ReadHuffmanCode (alphabet_size, code_lengths, huffman_tables, next); + if (0 == size) + return false; + + if (is_trivial_literal && kLiteralMap[j] == 1) + is_trivial_literal = (huffman_tables[next].bits == 0); + + total_size += huffman_tables[next].bits; + next += size; + if (j <= HuffIndex.Alpha) + { + int local_max_bits = code_lengths[0]; + for (int k = 1; k < alphabet_size; ++k) + { + if (code_lengths[k] > local_max_bits) + local_max_bits = code_lengths[k]; + } + max_bits += local_max_bits; + } + } + htree_group.is_trivial_literal = is_trivial_literal; + htree_group.is_trivial_code = false; + if (is_trivial_literal) + { + uint red = htree_group.GetCode (HuffIndex.Red, 0).value; + uint blue = htree_group.GetCode (HuffIndex.Blue, 0).value; + uint alpha = htree_group.GetCode (HuffIndex.Alpha, 0).value; + htree_group.literal_arb = (alpha << 24) | (red << 16) | blue; + if (total_size == 0 && htree_group.GetCode (HuffIndex.Green, 0).value < Huffman.NumLiteralCodes) + { + htree_group.is_trivial_code = true; + htree_group.literal_arb |= (uint)htree_group.GetCode (HuffIndex.Green, 0).value << 8; + } + } + htree_group.use_packed_table = !htree_group.is_trivial_code && (max_bits < Huffman.PackedBits); + if (htree_group.use_packed_table) + BuildPackedTable (htree_group); + } + + // All OK. Finalize pointers and return. + hdr_.huffman_image_ = huffman_image; + hdr_.num_htree_groups_ = num_htree_groups; + hdr_.htree_groups_ = htree_groups; + return true; + } + + int ReadHuffmanCode (int alphabet_size, int[] code_lengths, HuffmanCode[] table, int index) + { + bool ok = false; + int size = 0; + bool simple_code = br_.ReadBits (1) != 0; + + for (int i = 0; i < alphabet_size; ++i) + code_lengths[i] = 0; + + if (simple_code) // Read symbols, codes & code lengths directly. + { + int num_symbols = (int)br_.ReadBits (1) + 1; + int first_symbol_len_code = (int)br_.ReadBits (1); + // The first code is either 1 bit or 8 bit code. + int symbol = (int)br_.ReadBits ((first_symbol_len_code == 0) ? 1 : 8); + code_lengths[symbol] = 1; + // The second code (if present), is always 8 bit long. + if (2 == num_symbols) + { + symbol = (int)br_.ReadBits (8); + code_lengths[symbol] = 1; + } + ok = true; + } + else // Decode Huffman-coded code lengths. + { + var code_length_code_lengths = new int[NumCodeLengthCodes]; + int num_codes = (int)br_.ReadBits (4) + 4; + if (num_codes > NumCodeLengthCodes) + { + status_ = VP8StatusCode.BitstreamError; + return 0; + } + for (int i = 0; i < num_codes; ++i) + { + code_length_code_lengths[kCodeLengthCodeOrder[i]] = (int)br_.ReadBits (3); + } + ok = ReadHuffmanCodeLengths (code_length_code_lengths, alphabet_size, code_lengths); + } + + ok = ok && !br_.EoS; + if (ok) + size = Huffman.BuildTable (table, index, Huffman.TableBits, code_lengths, alphabet_size); + if (!ok || size == 0) + { + status_ = VP8StatusCode.BitstreamError; + return 0; + } + return size; + } + + bool ReadHuffmanCodeLengths (int[] code_length_code_lengths, int num_symbols, int[] code_lengths) + { + int prev_code_len = Huffman.DefaultCodeLength; + var table = new HuffmanCode[1 << Huffman.LengthsTableBits]; + + if (0 == Huffman.BuildTable (table, 0, Huffman.LengthsTableBits, code_length_code_lengths, NumCodeLengthCodes)) + { + status_ = VP8StatusCode.BitstreamError; + return false; + } + + int max_symbol; + if (0 != br_.ReadBits (1)) // use length + { + int length_nbits = 2 + 2 * (int)br_.ReadBits (3); + max_symbol = 2 + (int)br_.ReadBits (length_nbits); + if (max_symbol > num_symbols) + { + status_ = VP8StatusCode.BitstreamError; + return false; + } + } + else + { + max_symbol = num_symbols; + } + + int symbol = 0; + while (symbol < num_symbols) + { + if (max_symbol-- == 0) break; + br_.FillBitWindow(); + int p = (int)br_.PrefetchBits() & Huffman.LengthsTableMask; + br_.SkipBits (table[p].bits); + int code_len = table[p].value; + if (code_len < kCodeLengthLiterals) + { + code_lengths[symbol++] = code_len; + if (code_len != 0) prev_code_len = code_len; + } + else + { + bool use_prev = (code_len == kCodeLengthRepeatCode); + int slot = code_len - kCodeLengthLiterals; + int extra_bits = kCodeLengthExtraBits[slot]; + int repeat_offset = kCodeLengthRepeatOffsets[slot]; + int repeat = (int)br_.ReadBits(extra_bits) + repeat_offset; + if (symbol + repeat > num_symbols) + { + status_ = VP8StatusCode.BitstreamError; + return false; + } + else + { + int length = use_prev ? prev_code_len : 0; + while (repeat-- > 0) + code_lengths[symbol++] = length; + } + } + } + return true; + } + + void BuildPackedTable (HTreeGroup htree_group) + { + var huff = htree_group.packed_table; + for (int code = 0; code < Huffman.PackedTableSize; ++code) + { + uint bits = (uint)code; + var hcode = htree_group.GetCode (HuffIndex.Green, code); + if (hcode.value >= Huffman.NumLiteralCodes) + { + huff[code].bits = hcode.bits + BitsSpecialMarker; + huff[code].value = hcode.value; + } + else + { + huff[code].bits = 0; + huff[code].value = 0; + bits >>= AccumulateHCode (hcode, 8, ref huff[code]); + bits >>= AccumulateHCode (htree_group.GetCode (HuffIndex.Red, (int)bits), 16, ref huff[code]); + bits >>= AccumulateHCode (htree_group.GetCode (HuffIndex.Blue, (int)bits), 0, ref huff[code]); + bits >>= AccumulateHCode (htree_group.GetCode (HuffIndex.Alpha, (int)bits), 24, ref huff[code]); + } + } + } + + int AccumulateHCode (HuffmanCode hcode, int shift, ref HuffmanCode32 huff) + { + huff.bits += hcode.bits; + huff.value |= (uint)hcode.value << shift; + return hcode.bits; + } + + void UpdateDecoder (int width, int height) + { + int num_bits = hdr_.huffman_subsample_bits_; + width_ = width; + height_ = height; + + hdr_.huffman_xsize_ = SubSampleSize (width, num_bits); + hdr_.huffman_mask_ = (num_bits == 0) ? ~0 : (1 << num_bits) - 1; + } + + public static int SubSampleSize (int size, int sampling_bits) + { + return (int)(((uint)size + (1u << sampling_bits) - 1u) >> sampling_bits); + } + + public bool DecodeImageData (uint[] data, int width, int height, int last_row, ProcessRowsFunc process_func) + { + int row = last_pixel_ / width; + int col = last_pixel_ % width; + var htree_group = GetHtreeGroupForPos (col, row); + int src = last_pixel_; + int last_cached = src; + int src_end = width * height; // End of data + int src_last = width * last_row; // Last pixel to decode + int len_code_limit = Huffman.NumLiteralCodes + Huffman.NumLengthCodes; + int color_cache_limit = len_code_limit + hdr_.color_cache_size_; + int next_sync_row = incremental_ ? row : 1 << 24; + var color_cache = (hdr_.color_cache_size_ > 0) ? hdr_.color_cache_ : null; + int mask = hdr_.huffman_mask_; + + while (src < src_last) + { + int code; + if (row >= next_sync_row) + { + SaveState (src); + next_sync_row = row + SyncEveryNRows; + } + // Only update when changing tile. Note we could use this test: + // if "((((prev_col ^ col) | prev_row ^ row)) > mask)" -> tile changed + // but that's actually slower and needs storing the previous col/row. + if ((col & mask) == 0) + htree_group = GetHtreeGroupForPos (col, row); + if (htree_group.is_trivial_code) + { + data[src] = htree_group.literal_arb; + goto AdvanceByOne; + } + br_.FillBitWindow(); + if (htree_group.use_packed_table) + { + code = ReadPackedSymbols (htree_group, data, src); + if (code == PackedNonLiteralCode) + goto AdvanceByOne; + } + else + { + code = ReadSymbol (htree_group.Tables, htree_group.GetMeta (HuffIndex.Green)); + } + if (br_.EoS) break; // early out + if (code < Huffman.NumLiteralCodes) // Literal + { + if (htree_group.is_trivial_literal) + { + data[src] = htree_group.literal_arb | (uint)(code << 8); + } + else + { + int red, blue, alpha; + red = ReadSymbol (htree_group.Tables, htree_group.GetMeta (HuffIndex.Red)); + br_.FillBitWindow(); + blue = ReadSymbol (htree_group.Tables, htree_group.GetMeta (HuffIndex.Blue)); + alpha = ReadSymbol (htree_group.Tables, htree_group.GetMeta (HuffIndex.Alpha)); + if (br_.EoS) break; + data[src] = ((uint)alpha << 24) | ((uint)red << 16) | ((uint)code << 8) | (uint)blue; + } + } + else if (code < len_code_limit) // Backward reference + { + int length_sym = code - Huffman.NumLiteralCodes; + int length = GetCopyLength (length_sym); + int dist_symbol = ReadSymbol (htree_group.Tables, htree_group.GetMeta (HuffIndex.Dist)); + br_.FillBitWindow(); + int dist_code = GetCopyDistance (dist_symbol); + int dist = PlaneCodeToDistance (width, dist_code); + if (br_.EoS) break; + if (src < dist || src_end - src < length) + { + status_ = VP8StatusCode.BitstreamError; + return false; + } + else + { + int dst = src; + int s = dst - dist; + for (int i = 0; i < length; ++i) + data[dst+i] = data[s+i]; + } + src += length; + col += length; + while (col >= width) + { + col -= width; + ++row; + if ((row % NumARGBCacheRows == 0) && (process_func != null)) + process_func (this, row); + } + if (0 != (col & mask)) htree_group = GetHtreeGroupForPos (col, row); + if (color_cache != null) + { + while (last_cached < src) + color_cache.Insert (data[last_cached++]); + } + continue; + } + else if (code < color_cache_limit) // Color cache + { + int key = code - len_code_limit; + while (last_cached < src) + color_cache.Insert (data[last_cached++]); + data[src] = color_cache.Lookup ((uint)key); + } + else // Not reached + { + status_ = VP8StatusCode.BitstreamError; + return false; + } + + AdvanceByOne: + ++src; + ++col; + if (col >= width) + { + col = 0; + ++row; + if ((row % NumARGBCacheRows == 0) && (process_func != null)) + { + process_func (this, row); + } + if (color_cache != null) + { + while (last_cached < src) + color_cache.Insert (data[last_cached++]); + } + } + } + + if (incremental_ && br_.EoS && src < src_end) + { + RestoreState(); + } + else if (!br_.EoS) + { + // Process the remaining rows corresponding to last row-block. + if (process_func != null) + process_func (this, row); + status_ = VP8StatusCode.Ok; + last_pixel_ = src; // end-of-scan marker + } + else + { + // if not incremental, and we are past the end of buffer (eos_=1), then this + // is a real bitstream error. + status_ = VP8StatusCode.BitstreamError; + return false; + } + return true; + } + + HTreeGroup GetHtreeGroupForPos (int x, int y) + { + int meta_index = GetMetaIndex (hdr_.huffman_image_, hdr_.huffman_xsize_, hdr_.huffman_subsample_bits_, x, y); + return hdr_.htree_groups_[meta_index]; + } + + int GetMetaIndex (uint[] image, int xsize, int bits, int x, int y) + { + if (0 == bits) return 0; + return (int)image[xsize * (y >> bits) + (x >> bits)]; + } + + //------------------------------------------------------------------------------ + // Decodes the next Huffman code from bit-stream. + // FillBitWindow(br) needs to be called at minimum every second call + // to ReadSymbol, in order to pre-fetch enough bits. + int ReadSymbol (HuffmanCode[] table, int index) + { + int val = (int)br_.PrefetchBits(); + index += val & Huffman.TableMask; + int nbits = table[index].bits - Huffman.TableBits; + if (nbits > 0) + { + br_.SkipBits (Huffman.TableBits); + val = (int)br_.PrefetchBits(); + index += table[index].value; + index += val & ((1 << nbits) - 1); + } + br_.SkipBits (table[index].bits); + return table[index].value; + } + + int ReadPackedSymbols (HTreeGroup group, uint[] data, int dst) + { + uint val = br_.PrefetchBits() & (Huffman.PackedTableSize - 1); + var code = group.packed_table[val]; + if (code.bits < BitsSpecialMarker) + { + br_.SkipBits (code.bits); + data[dst] = code.value; + return PackedNonLiteralCode; + } + else + { + br_.SkipBits (code.bits - BitsSpecialMarker); + return (int)code.value; + } + } + + int GetCopyLength (int length_symbol) + { + // Length and distance prefixes are encoded the same way. + return GetCopyDistance (length_symbol); + } + + int GetCopyDistance (int distance_symbol) + { + if (distance_symbol < 4) + return distance_symbol + 1; + int extra_bits = (distance_symbol - 2) >> 1; + int offset = (2 + (distance_symbol & 1)) << extra_bits; + return offset + (int)br_.ReadBits (extra_bits) + 1; + } + + int PlaneCodeToDistance (int xsize, int plane_code) + { + if (plane_code > CodeToPlaneCodes) + return plane_code - CodeToPlaneCodes; + int dist_code = kCodeToPlane[plane_code - 1]; + int yoffset = dist_code >> 4; + int xoffset = 8 - (dist_code & 0xf); + int dist = yoffset * xsize + xoffset; + return (dist >= 1) ? dist : 1; // dist<1 can happen if xsize is very small + } + + void SaveState (int last_pixel) + { + br_.CopyStateTo (saved_br_); + saved_last_pixel_ = last_pixel; + if (hdr_.color_cache_size_ > 0) + hdr_.color_cache_.Copy (hdr_.saved_color_cache_); + } + + void RestoreState () + { + status_ = VP8StatusCode.Suspended; + saved_br_.CopyStateTo (br_); + last_pixel_ = saved_last_pixel_; + if (hdr_.color_cache_size_ > 0) + hdr_.saved_color_cache_.Copy (hdr_.color_cache_); + } + + public bool DecodeAlphaData (int width, int height, int last_row) + { + var data = pixels8_; + bool ok = true; + int row = last_pixel_ / width; + int col = last_pixel_ % width; + var htree_group = GetHtreeGroupForPos (col, row); + int pos = last_pixel_; // current position + int end = width * height; // End of data + int last = width * last_row; // Last pixel to decode + int len_code_limit = Huffman.NumLiteralCodes + Huffman.NumLengthCodes; + int mask = hdr_.huffman_mask_; + + while (!br_.EoS && pos < last) + { + // Only update when changing tile. + if ((col & mask) == 0) + htree_group = GetHtreeGroupForPos (col, row); + + br_.FillBitWindow(); + int code = ReadSymbol (htree_group.Tables, htree_group.GetMeta (HuffIndex.Green)); + if (code < Huffman.NumLiteralCodes) // Literal + { + data[pos] = (byte)code; + ++pos; + ++col; + if (col >= width) + { + col = 0; + ++row; + if (row % NumARGBCacheRows == 0) + ExtractPalettedAlphaRows (row); + } + } + else if (code < len_code_limit) // Backward reference + { + int length_sym = code - Huffman.NumLiteralCodes; + int length = GetCopyLength (length_sym); + int dist_symbol = ReadSymbol (htree_group.Tables, htree_group.GetMeta (HuffIndex.Dist)); + br_.FillBitWindow(); + int dist_code = GetCopyDistance (dist_symbol); + int dist = PlaneCodeToDistance (width, dist_code); + if (pos >= dist && end - pos >= length) + { + Binary.CopyOverlapped (data, pos - dist, pos, length); + } + else + { + ok = false; + goto End; + } + pos += length; + col += length; + while (col >= width) + { + col -= width; + ++row; + if (row % NumARGBCacheRows == 0) + ExtractPalettedAlphaRows (row); + } + if (pos < last && 0 != (col & mask)) + htree_group = GetHtreeGroupForPos (col, row); + } + else // Not reached + { + ok = false; + goto End; + } + } + // Process the remaining rows corresponding to last row-block. + ExtractPalettedAlphaRows (row); + + End: + if (!ok || (br_.EoS && pos < end)) + { + ok = false; + status_ = br_.EoS ? VP8StatusCode.Suspended : VP8StatusCode.BitstreamError; + } + else + { + last_pixel_ = pos; + } + return ok; + } + + void ExtractPalettedAlphaRows (int row) + { + int num_rows = row - last_row_; + int input = width_ * last_row_; + if (num_rows > 0) + ApplyInverseTransformsAlpha (num_rows, pixels8_, input); + + last_row_ = last_out_row_ = row; + } + + void ApplyInverseTransforms (int num_rows, uint[] rows, int rows_in) + { + int n = next_transform_; + int cache_pixs = width_ * num_rows; + int start_row = last_row_; + int end_row = start_row + num_rows; + int rows_out = argb_cache_; + + // Inverse transforms. + // TODO: most transforms only need to operate on the cropped region only. + Buffer.BlockCopy (rows, rows_in, pixels32_, rows_out, cache_pixs * sizeof(uint)); + while (n --> 0) + { + transforms_[n].InverseTransform (start_row, end_row, rows, rows_in, pixels32_, rows_out); + rows = pixels32_; + rows_in = rows_out; + } + } + + void ApplyInverseTransformsAlpha (int num_rows, byte[] rows, int rows_in) + { + int start_row = last_row_; + int end_row = start_row + num_rows; + byte[] output = io_.opaque; + int rows_out = io_.width * start_row; + transforms_[0].ColorIndexInverseTransformAlpha (start_row, end_row, rows, rows_in, output, rows_out); + } + + /// + /// Special row-processing that only stores the alpha data. + /// + public static void ExtractAlphaRows (LosslessDecoder dec, int row) + { + int num_rows = row - dec.last_row_; + if (num_rows <= 0) return; // Nothing to be done. + + dec.ApplyInverseTransforms (num_rows, dec.pixels32_, dec.width_ * dec.last_row_); + + // Extract alpha (which is stored in the green plane). + + int width = dec.io_.width; // the final width (!= dec->width_) + int cache_pixs = width * num_rows; +// uint8_t* const dst = (uint8_t*)dec->io_->opaque + width * dec->last_row_; + int dst = width * dec.last_row_; + int src = dec.argb_cache_; + for (int i = 0; i < cache_pixs; ++i) + dec.io_.opaque[dst+i] = (byte)(dec.pixels32_[src+i] >> 8); + dec.last_row_ = dec.last_out_row_ = row; + } + + //------------------------------------------------------------------------------ + + internal class LMetadata + { + public int color_cache_size_; + public LColorCache color_cache_ = new LColorCache(); + public LColorCache saved_color_cache_ = new LColorCache(); // for incremental + + public int huffman_mask_; + public int huffman_subsample_bits_; + public int huffman_xsize_; + public uint[] huffman_image_; + public int num_htree_groups_; + public HTreeGroup[] htree_groups_; + + public void ClearMetadata () + { + color_cache_size_ = 0; + huffman_mask_ = 0; + huffman_subsample_bits_ = 0; + huffman_xsize_ = 0; + huffman_image_ = null; + num_htree_groups_ = 0; + htree_groups_ = null; + } + } + + internal class LColorCache + { + public uint[] colors_; // color entries + public int hash_shift_; // Hash shift: 32 - hash_bits_. + public int hash_bits_; + + public void Init (int hash_bits) + { + int hash_size = 1 << hash_bits; + colors_ = new uint[hash_size]; + hash_shift_ = 32 - hash_bits; + hash_bits_ = hash_bits; + } + + public void Copy (LColorCache dst) + { + Array.Copy (colors_, dst.colors_, 1u << dst.hash_bits_); + } + + public uint Lookup (uint key) + { + return colors_[key]; + } + + public void Set (uint key, uint argb) + { + colors_[key] = argb; + } + + public void Insert (uint argb) + { + uint key = (kHashMul * argb) >> hash_shift_; + colors_[key] = argb; + } + + public int GetIndex (uint argb) + { + return (int)((kHashMul * argb) >> hash_shift_); + } + } + + // ----------------------------------------------------------------------------- + /// + /// Bitreader for lossless format + /// + internal class LBitReader + { + ulong val_; // pre-fetched bits + byte[] buf_; // input byte buffer + uint len_; // buffer length + uint pos_; // byte position in buf_ + int bit_pos_; // current bit-reading position in val_ + bool eos_; // true if a bit was read past the end of buffer + + public const int MaxNumBitRead = 24; + + public const int LBITS = 64; // Number of bits prefetched. + public const int WBITS = 32; // Minimum number of bytes ready after VP8LFillBitWindow. + + const int LOG8_WBITS = 4; // Number of bytes needed to store WBITS bits. + + public bool EoS { get { return eos_; } } + + /// + /// Returns true if there was an attempt at reading bit past the end of + /// the buffer. Doesn't set br->eos_ flag. + /// + private bool IsEndOfStream { get { return eos_ || ((pos_ == len_) && (bit_pos_ > LBITS)); } } + + public void Init (byte[] input, int start, uint length) + { + len_ = length; + val_ = 0; + bit_pos_ = 0; + eos_ = false; + + if (length > sizeof(ulong)) + length = sizeof(ulong); + + ulong v = 0; + for (uint i = 0; i < length; ++i) + v |= (ulong)input[start+i] << (8 * (int)i); + + val_ = v; + pos_ = (uint)start+length; + buf_ = input; + } + + public void CopyStateTo (LBitReader other) + { + other.val_ = val_; + other.buf_ = buf_; + other.len_ = len_; + other.pos_ = pos_; + other.bit_pos_ = bit_pos_; + other.eos_ = eos_; + } + + /// + /// Sets a new data buffer. + /// + public void SetBuffer (byte[] buffer, uint length) + { + buf_ = buffer; + len_ = length; + // pos_ > len_ should be considered a param error. + eos_ = (pos_ > len_) || IsEndOfStream; + } + + /// + /// Reads the specified number of bits from read buffer. + /// Flags an error in case end_of_stream or n_bits is more than the allowed limit + /// of VP8L_MAX_NUM_BIT_READ (inclusive). + /// Flags eos_ if this read attempt is going to cross the read buffer. + /// + public uint ReadBits (int n_bits) + { + // Flag an error if end_of_stream or n_bits is more than allowed limit. + if (!eos_ && n_bits <= MaxNumBitRead) + { + uint val = PrefetchBits() & kBitMask[n_bits]; + int new_bits = bit_pos_ + n_bits; + bit_pos_ = new_bits; + ShiftBytes(); + return val; + } + else + { + SetEndOfStream(); + return 0; + } + } + + /// + /// Return the prefetched bits, so they can be looked up. + /// + public uint PrefetchBits () + { + return (uint)(val_ >> (bit_pos_ & (LBITS - 1))); + } + + /// + /// For jumping over a number of bits in the bit stream when accessed with + /// VP8LPrefetchBits and VP8LFillBitWindow. + /// + public void SetBitPos (int val) + { + bit_pos_ = val; + eos_ = IsEndOfStream; + } + + public void SkipBits (int bits) + { + SetBitPos (bit_pos_ + bits); + } + + /// + /// Advances the read buffer by 4 bytes to make room for reading next 32 bits. + /// Speed critical, but infrequent part of the code can be non-inlined. + /// + + public void FillBitWindow () + { + if (bit_pos_ >= WBITS) + DoFillBitWindow(); + } + + void DoFillBitWindow () + { + if (pos_ + sizeof(ulong) < len_) + { + val_ >>= WBITS; + bit_pos_ -= WBITS; + val_ |= (ulong)LittleEndian.ToUInt32 (buf_, (int)pos_) << (LBITS - WBITS); + pos_ += LOG8_WBITS; + return; + } + ShiftBytes(); // Slow path. + } + + void ShiftBytes () + { + while (bit_pos_ >= 8 && pos_ < len_) + { + val_ >>= 8; + val_ |= ((ulong)buf_[pos_]) << (LBITS - 8); + ++pos_; + bit_pos_ -= 8; + } + if (IsEndOfStream) + SetEndOfStream(); + } + + void SetEndOfStream () + { + eos_ = true; + bit_pos_ = 0; // To avoid undefined behaviour with shifts. + } + + static readonly uint[] kBitMask = new uint[MaxNumBitRead + 1] + { + 0, + 0x000001, 0x000003, 0x000007, 0x00000f, + 0x00001f, 0x00003f, 0x00007f, 0x0000ff, + 0x0001ff, 0x0003ff, 0x0007ff, 0x000fff, + 0x001fff, 0x003fff, 0x007fff, 0x00ffff, + 0x01ffff, 0x03ffff, 0x07ffff, 0x0fffff, + 0x1fffff, 0x3fffff, 0x7fffff, 0xffffff + }; + } + + // Memory needed for lookup tables of one Huffman tree group. Red, blue, alpha + // and distance alphabets are constant (256 for red, blue and alpha, 40 for + // distance) and lookup table sizes for them in worst case are 630 and 410 + // respectively. Size of green alphabet depends on color cache size and is equal + // to 256 (green component values) + 24 (length prefix values) + // + color_cache_size (between 0 and 2048). + // All values computed for 8-bit first level lookup with Mark Adler's tool: + // http://www.hdfgroup.org/ftp/lib-external/zlib/zlib-1.2.5/examples/enough.c + + const int FIXED_TABLE_SIZE = 630 * 3 + 410; + static readonly int[] kTableSize = { + FIXED_TABLE_SIZE + 654, + FIXED_TABLE_SIZE + 656, + FIXED_TABLE_SIZE + 658, + FIXED_TABLE_SIZE + 662, + FIXED_TABLE_SIZE + 670, + FIXED_TABLE_SIZE + 686, + FIXED_TABLE_SIZE + 718, + FIXED_TABLE_SIZE + 782, + FIXED_TABLE_SIZE + 912, + FIXED_TABLE_SIZE + 1168, + FIXED_TABLE_SIZE + 1680, + FIXED_TABLE_SIZE + 2704 + }; + + static readonly ushort[] kAlphabetSize = new ushort[Huffman.CodesPerMetaCode] + { + Huffman.NumLiteralCodes + Huffman.NumLengthCodes, + Huffman.NumLiteralCodes, + Huffman.NumLiteralCodes, + Huffman.NumLiteralCodes, + Huffman.NumDistanceCodes + }; + + const int NumCodeLengthCodes = 19; + static readonly byte[] kCodeLengthCodeOrder = new byte[NumCodeLengthCodes] { + 17, 18, 0, 1, 2, 3, 4, 5, 16, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 + }; + + static readonly byte[] kLiteralMap = new byte[Huffman.CodesPerMetaCode] { + 0, 1, 1, 1, 0 + }; + + const int CodeToPlaneCodes = 120; + static readonly byte[] kCodeToPlane = new byte[CodeToPlaneCodes] + { + 0x18, 0x07, 0x17, 0x19, 0x28, 0x06, 0x27, 0x29, 0x16, 0x1a, + 0x26, 0x2a, 0x38, 0x05, 0x37, 0x39, 0x15, 0x1b, 0x36, 0x3a, + 0x25, 0x2b, 0x48, 0x04, 0x47, 0x49, 0x14, 0x1c, 0x35, 0x3b, + 0x46, 0x4a, 0x24, 0x2c, 0x58, 0x45, 0x4b, 0x34, 0x3c, 0x03, + 0x57, 0x59, 0x13, 0x1d, 0x56, 0x5a, 0x23, 0x2d, 0x44, 0x4c, + 0x55, 0x5b, 0x33, 0x3d, 0x68, 0x02, 0x67, 0x69, 0x12, 0x1e, + 0x66, 0x6a, 0x22, 0x2e, 0x54, 0x5c, 0x43, 0x4d, 0x65, 0x6b, + 0x32, 0x3e, 0x78, 0x01, 0x77, 0x79, 0x53, 0x5d, 0x11, 0x1f, + 0x64, 0x6c, 0x42, 0x4e, 0x76, 0x7a, 0x21, 0x2f, 0x75, 0x7b, + 0x31, 0x3f, 0x63, 0x6d, 0x52, 0x5e, 0x00, 0x74, 0x7c, 0x41, + 0x4f, 0x10, 0x20, 0x62, 0x6e, 0x30, 0x73, 0x7d, 0x51, 0x5f, + 0x40, 0x72, 0x7e, 0x61, 0x6f, 0x50, 0x71, 0x7f, 0x60, 0x70 + }; + } +}