1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
|
/*
* Copyright 2023 jacqueline <me@jacqueline.id.au>
*
* SPDX-License-Identifier: GPL-3.0-only
*/
#include "audio_decoder.hpp"
#include <string.h>
#include <algorithm>
#include <cstddef>
#include <cstdint>
#include <memory>
#include <variant>
#include "codec.hpp"
#include "freertos/FreeRTOS.h"
#include "esp_heap_caps.h"
#include "esp_log.h"
#include "freertos/message_buffer.h"
#include "freertos/portmacro.h"
#include "audio_element.hpp"
#include "chunk.hpp"
#include "fatfs_audio_input.hpp"
#include "stream_info.hpp"
namespace audio {
static const char* kTag = "DEC";
AudioDecoder::AudioDecoder()
: IAudioElement(),
current_codec_(),
current_input_format_(),
current_output_format_(),
has_prepared_output_(false),
has_samples_to_send_(false),
has_input_remaining_(false) {}
AudioDecoder::~AudioDecoder() {}
auto AudioDecoder::ProcessStreamInfo(const StreamInfo& info) -> bool {
has_prepared_output_ = false;
current_codec_.reset();
current_input_format_.reset();
current_output_format_.reset();
if (!std::holds_alternative<StreamInfo::Encoded>(info.format)) {
return false;
}
const auto& new_format = std::get<StreamInfo::Encoded>(info.format);
current_input_format_ = new_format;
ESP_LOGI(kTag, "creating new decoder");
auto result = codecs::CreateCodecForType(new_format.type);
if (result.has_value()) {
current_codec_.reset(result.value());
} else {
ESP_LOGE(kTag, "no codec for this file");
return false;
}
return true;
}
auto AudioDecoder::NeedsToProcess() const -> bool {
return has_samples_to_send_ || has_input_remaining_;
}
auto AudioDecoder::Process(const std::vector<InputStream>& inputs,
OutputStream* output) -> void {
auto input = inputs.begin();
const StreamInfo& info = input->info();
// Is this a completely new stream?
if (!current_input_format_) {
if (!ProcessStreamInfo(info)) {
// We couldn't handle the new stream. Signal to the producer that we don't
// have anything to do.
input->mark_consumer_finished();
return;
}
}
// Have we determined what kind of samples this stream decodes to?
if (!current_output_format_) {
auto res = current_codec_->BeginStream(input->data());
input->consume(res.first);
if (res.second.has_error()) {
auto err = res.second.error();
if (err == codecs::ICodec::Error::kOutOfInput) {
// We didn't manage to clear whatever front matter is before this
// stream's header. We need to call BeginStream again with more data.
return;
}
// Somthing about the stream's header was malformed. Skip it.
ESP_LOGE(kTag, "error beginning stream");
input->mark_consumer_finished();
return;
}
// The stream started successfully. Record what format the samples are in.
codecs::ICodec::OutputFormat format = res.second.value();
current_output_format_ = StreamInfo::Pcm{
.channels = format.num_channels,
.bits_per_sample = format.bits_per_sample,
.sample_rate = format.sample_rate_hz,
};
if (format.duration_seconds) {
duration_seconds_from_decoder_ = format.duration_seconds;
} else if (format.bits_per_second &&
current_input_format_->duration_bytes) {
duration_seconds_from_decoder_ =
(current_input_format_->duration_bytes.value() - res.first) * 8 /
format.bits_per_second.value() / format.num_channels;
}
if (info.seek_to_seconds) {
seek_to_sample_ = *info.seek_to_seconds * format.sample_rate_hz;
} else {
seek_to_sample_.reset();
}
}
if (seek_to_sample_) {
ESP_LOGI(kTag, "seeking forwards...");
auto res = current_codec_->SeekStream(input->data(), *seek_to_sample_);
input->consume(res.first);
if (res.second.has_error()) {
auto err = res.second.error();
if (err == codecs::ICodec::Error::kOutOfInput) {
return;
} else {
// TODO(jacqueline): Handle errors.
}
}
seek_to_sample_.reset();
}
has_input_remaining_ = true;
while (true) {
// Make sure the output buffer is ready to receive samples in our format
// before starting to process data.
// TODO(jacqueline): Pass through seek info here?
if (!has_prepared_output_ && !output->prepare(*current_output_format_)) {
return;
}
if (duration_seconds_from_decoder_) {
output->set_duration(*duration_seconds_from_decoder_);
}
has_prepared_output_ = true;
// Parse frames and produce samples.
auto res = current_codec_->ContinueStream(input->data(), output->data());
input->consume(res.first);
// Handle any errors during processing.
if (res.second.has_error()) {
// The codec ran out of input during processing. This is expected to
// happen throughout the stream.
if (res.second.error() == codecs::ICodec::Error::kOutOfInput) {
has_input_remaining_ = false;
has_samples_to_send_ = false;
if (input->is_producer_finished()) {
// We're out of data, and so is the producer. Nothing left to be done
// with the input stream.
input->mark_consumer_finished();
// Upstream isn't going to give us any more data. Tell downstream
// that they shouldn't expact any more samples from this stream.
output->mark_producer_finished();
break;
}
} else {
// TODO(jacqueline): Handle errors.
ESP_LOGE(kTag, "codec returned fatal error");
}
// Note that a codec that returns an error is not allowed to write
// samples. So it's safe to skip the latter part of the loop.
return;
} else {
// Some samples were written! Ensure the downstream element knows about
// them.
codecs::ICodec::OutputInfo out_info = res.second.value();
output->add(out_info.bytes_written);
has_samples_to_send_ = !out_info.is_finished_writing;
if (has_samples_to_send_) {
// The codec wasn't able to finish writing all of its samples into the
// output buffer. We need to return so that we can get a new buffer.
return;
}
}
}
current_codec_.reset();
current_input_format_.reset();
}
} // namespace audio
|