/*
* Copyright (c) 2016 The WebRTC project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include "common_video/h264/sps_parser.h"
#include <cstdint>
#include <vector>
#include "common_video/h264/h264_common.h"
#include "rtc_base/bit_buffer.h"
namespace {
typedef absl::optional<webrtc::SpsParser::SpsState> OptionalSps;
#define RETURN_EMPTY_ON_FAIL(x) \
if (!(x)) { \
return OptionalSps(); \
}
constexpr int kScalingDeltaMin = -128;
constexpr int kScaldingDeltaMax = 127;
} // namespace
namespace webrtc {
SpsParser::SpsState::SpsState() = default;
SpsParser::SpsState::SpsState(const SpsState&) = default;
SpsParser::SpsState::~SpsState() = default;
// General note: this is based off the 02/2014 version of the H.264 standard.
// You can find it on this page:
// http://www.itu.int/rec/T-REC-H.264
// Unpack RBSP and parse SPS state from the supplied buffer.
absl::optional<SpsParser::SpsState> SpsParser::ParseSps(const uint8_t* data,
size_t length) {
std::vector<uint8_t> unpacked_buffer = H264::ParseRbsp(data, length);
rtc::BitBuffer bit_buffer(unpacked_buffer.data(), unpacked_buffer.size());
return ParseSpsUpToVui(&bit_buffer);
}
absl::optional<SpsParser::SpsState> SpsParser::ParseSpsUpToVui(
rtc::BitBuffer* buffer) {
// Now, we need to use a bit buffer to parse through the actual AVC SPS
// format. See Section 7.3.2.1.1 ("Sequence parameter set data syntax") of the
// H.264 standard for a complete description.
// Since we only care about resolution, we ignore the majority of fields, but
// we still have to actively parse through a lot of the data, since many of
// the fields have variable size.
// We're particularly interested in:
// chroma_format_idc -> affects crop units
// pic_{width,height}_* -> resolution of the frame in macroblocks (16x16).
// frame_crop_*_offset -> crop information
SpsState sps;
// The golomb values we have to read, not just consume.
uint32_t golomb_ignored;
// chroma_format_idc will be ChromaArrayType if separate_colour_plane_flag is
// 0. It defaults to 1, when not specified.
uint32_t chroma_format_idc = 1;
// profile_idc: u(8). We need it to determine if we need to read/skip chroma
// formats.
uint8_t profile_idc;
RETURN_EMPTY_ON_FAIL(buffer->ReadUInt8(&profile_idc));
// constraint_set0_flag through constraint_set5_flag + reserved_zero_2bits
// 1 bit each for the flags + 2 bits = 8 bits = 1 byte.
RETURN_EMPTY_ON_FAIL(buffer->ConsumeBytes(1));
// level_idc: u(8)
RETURN_EMPTY_ON_FAIL(buffer->ConsumeBytes(1));
// seq_parameter_set_id: ue(v)
RETURN_EMPTY_ON_FAIL(buffer->ReadExponentialGolomb(&sps.id));
sps.separate_colour_plane_flag = 0;
// See if profile_idc has chroma format information.
if (profile_idc == 100 || profile_idc == 110 || profile_idc == 122 ||
profile_idc == 244 || profile_idc == 44 || profile_idc == 83 ||
profile_idc == 86 || profile_idc == 118 || profile_idc == 128 ||
profile_idc == 138 || profile_idc == 139 || profile_idc == 134) {
// chroma_format_idc: ue(v)
RETURN_EMPTY_ON_FAIL(buffer->ReadExponentialGolomb(&chroma_format_idc));
if (chroma_format_idc == 3) {
// separate_colour_plane_flag: u(1)
RETURN_EMPTY_ON_FAIL(
buffer->ReadBits(&sps.separate_colour_plane_flag, 1));
}
// bit_depth_luma_minus8: ue(v)
RETURN_EMPTY_ON_FAIL(buffer->ReadExponentialGolomb(&golomb_ignored));
// bit_depth_chroma_minus8: ue(v)
RETURN_EMPTY_ON_FAIL(buffer->ReadExponentialGolomb(&golomb_ignored));
// qpprime_y_zero_transform_bypass_flag: u(1)
RETURN_EMPTY_ON_FAIL(buffer->ConsumeBits(1));
// seq_scaling_matrix_present_flag: u(1)
uint32_t seq_scaling_matrix_present_flag;
RETURN_EMPTY_ON_FAIL(buffer->ReadBits(&seq_scaling_matrix_present_flag, 1));
if (seq_scaling_matrix_present_flag) {
// Process the scaling lists just enough to be able to properly
// skip over them, so we can still read the resolution on streams
// where this is included.
int scaling_list_count = (chroma_format_idc == 3 ? 12 : 8);
for (int i = 0; i < scaling_list_count; ++i) {
// seq_scaling_list_present_flag[i] : u(1)
uint32_t seq_scaling_list_present_flags;
RETURN_EMPTY_ON_FAIL(
buffer->ReadBits(&seq_scaling_list_present_flags, 1));
if (seq_scaling_list_present_flags != 0) {
int last_scale = 8;
int next_scale = 8;
int size_of_scaling_list = i < 6 ? 16 : 64;
for (int j = 0; j < size_of_scaling_list; j++) {
if (next_scale != 0) {
int32_t delta_scale;
// delta_scale: se(v)
RETURN_EMPTY_ON_FAIL(
buffer->ReadSignedExponentialGolomb(&delta_scale));
RETURN_EMPTY_ON_FAIL(delta_scale >= kScalingDeltaMin &&
delta_scale <= kScaldingDeltaMax);
next_scale = (last_scale + delta_scale + 256) % 256;
}
if (next_scale != 0)
last_scale = next_scale;
}
}
}
}
}
// log2_max_frame_num and log2_max_pic_order_cnt_lsb are used with
// BitBuffer::ReadBits, which can read at most 32 bits at a time. We also have
// to avoid overflow when adding 4 to the on-wire golomb value, e.g., for evil
// input data, ReadExponentialGolomb might return 0xfffc.
const uint32_t kMaxLog2Minus4 = 32 - 4;
// log2_max_frame_num_minus4: ue(v)
uint32_t log2_max_frame_num_minus4;
if (!buffer->ReadExponentialGolomb(&log2_max_frame_num_minus4) ||
log2_max_frame_num_minus4 > kMaxLog2Minus4) {
return OptionalSps();
}
sps.log2_max_frame_num = log2_max_frame_num_minus4 + 4;
// pic_order_cnt_type: ue(v)
RETURN_EMPTY_ON_FAIL(buffer->ReadExponentialGolomb(&sps.pic_order_cnt_type));
if (sps.pic_order_cnt_type == 0) {
// log2_max_pic_order_cnt_lsb_minus4: ue(v)
uint32_t log2_max_pic_order_cnt_lsb_minus4;
if (!buffer->ReadExponentialGolomb(&log2_max_pic_order_cnt_lsb_minus4) ||
log2_max_pic_order_cnt_lsb_minus4 > kMaxLog2Minus4) {
return OptionalSps();
}
sps.log2_max_pic_order_cnt_lsb = log2_max_pic_order_cnt_lsb_minus4 + 4;
} else if (sps.pic_order_cnt_type == 1) {
// delta_pic_order_always_zero_flag: u(1)
RETURN_EMPTY_ON_FAIL(
buffer->ReadBits(&sps.delta_pic_order_always_zero_flag, 1));
// offset_for_non_ref_pic: se(v)
RETURN_EMPTY_ON_FAIL(buffer->ReadExponentialGolomb(&golomb_ignored));
// offset_for_top_to_bottom_field: se(v)
RETURN_EMPTY_ON_FAIL(buffer->ReadExponentialGolomb(&golomb_ignored));
// num_ref_frames_in_pic_order_cnt_cycle: ue(v)
uint32_t num_ref_frames_in_pic_order_cnt_cycle;
RETURN_EMPTY_ON_FAIL(
buffer->ReadExponentialGolomb(&num_ref_frames_in_pic_order_cnt_cycle));
for (size_t i = 0; i < num_ref_frames_in_pic_order_cnt_cycle; ++i) {
// offset_for_ref_frame[i]: se(v)
RETURN_EMPTY_ON_FAIL(buffer->ReadExponentialGolomb(&golomb_ignored));
}
}
// max_num_ref_frames: ue(v)
RETURN_EMPTY_ON_FAIL(buffer->ReadExponentialGolomb(&sps.max_num_ref_frames));
// gaps_in_frame_num_value_allowed_flag: u(1)
RETURN_EMPTY_ON_FAIL(buffer->ConsumeBits(1));
//
// IMPORTANT ONES! Now we're getting to resolution. First we read the pic
// width/height in macroblocks (16x16), which gives us the base resolution,
// and then we continue on until we hit the frame crop offsets, which are used
// to signify resolutions that aren't multiples of 16.
//
// pic_width_in_mbs_minus1: ue(v)
uint32_t pic_width_in_mbs_minus1;
RETURN_EMPTY_ON_FAIL(buffer->ReadExponentialGolomb(&pic_width_in_mbs_minus1));
// pic_height_in_map_units_minus1: ue(v)
uint32_t pic_height_in_map_units_minus1;
RETURN_EMPTY_ON_FAIL(
buffer->ReadExponentialGolomb(&pic_height_in_map_units_minus1));
// frame_mbs_only_flag: u(1)
RETURN_EMPTY_ON_FAIL(buffer->ReadBits(&sps.frame_mbs_only_flag, 1));
if (!sps.frame_mbs_only_flag) {
// mb_adaptive_frame_field_flag: u(1)
RETURN_EMPTY_ON_FAIL(buffer->ConsumeBits(1));
}
// direct_8x8_inference_flag: u(1)
RETURN_EMPTY_ON_FAIL(buffer->ConsumeBits(1));
//
// MORE IMPORTANT ONES! Now we're at the frame crop information.
//
// frame_cropping_flag: u(1)
uint32_t frame_cropping_flag;
uint32_t frame_crop_left_offset = 0;
uint32_t frame_crop_right_offset = 0;
uint32_t frame_crop_top_offset = 0;
uint32_t frame_crop_bottom_offset = 0;
RETURN_EMPTY_ON_FAIL(buffer->ReadBits(&frame_cropping_flag, 1));
if (frame_cropping_flag) {
// frame_crop_{left, right, top, bottom}_offset: ue(v)
RETURN_EMPTY_ON_FAIL(
buffer->ReadExponentialGolomb(&frame_crop_left_offset));
RETURN_EMPTY_ON_FAIL(
buffer->ReadExponentialGolomb(&frame_crop_right_offset));
RETURN_EMPTY_ON_FAIL(buffer->ReadExponentialGolomb(&frame_crop_top_offset));
RETURN_EMPTY_ON_FAIL(
buffer->ReadExponentialGolomb(&frame_crop_bottom_offset));
}
// vui_parameters_present_flag: u(1)
RETURN_EMPTY_ON_FAIL(buffer->ReadBits(&sps.vui_params_present, 1));
// Far enough! We don't use the rest of the SPS.
// Start with the resolution determined by the pic_width/pic_height fields.
sps.width = 16 * (pic_width_in_mbs_minus1 + 1);
sps.height =
16 * (2 - sps.frame_mbs_only_flag) * (pic_height_in_map_units_minus1 + 1);
// Figure out the crop units in pixels. That's based on the chroma format's
// sampling, which is indicated by chroma_format_idc.
if (sps.separate_colour_plane_flag || chroma_format_idc == 0) {
frame_crop_bottom_offset *= (2 - sps.frame_mbs_only_flag);
frame_crop_top_offset *= (2 - sps.frame_mbs_only_flag);
} else if (!sps.separate_colour_plane_flag && chroma_format_idc > 0) {
// Width multipliers for formats 1 (4:2:0) and 2 (4:2:2).
if (chroma_format_idc == 1 || chroma_format_idc == 2) {
frame_crop_left_offset *= 2;
frame_crop_right_offset *= 2;
}
// Height multipliers for format 1 (4:2:0).
if (chroma_format_idc == 1) {
frame_crop_top_offset *= 2;
frame_crop_bottom_offset *= 2;
}
}
// Subtract the crop for each dimension.
sps.width -= (frame_crop_left_offset + frame_crop_right_offset);
sps.height -= (frame_crop_top_offset + frame_crop_bottom_offset);
return OptionalSps(sps);
}
} // namespace webrtc