Update RTTTL component to allow I2S (#5177)

Co-authored-by: Jesse Hills <3060199+jesserockz@users.noreply.github.com>
This commit is contained in:
NP v/d Spek 2023-11-02 05:14:05 +01:00 committed by GitHub
parent 4edf3efdf3
commit 453600f18e
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
3 changed files with 245 additions and 80 deletions

View file

@ -4,7 +4,15 @@ import esphome.config_validation as cv
import esphome.final_validate as fv
from esphome import automation
from esphome.components.output import FloatOutput
from esphome.const import CONF_ID, CONF_OUTPUT, CONF_PLATFORM, CONF_TRIGGER_ID
from esphome.components.speaker import Speaker
from esphome.const import (
CONF_ID,
CONF_OUTPUT,
CONF_PLATFORM,
CONF_TRIGGER_ID,
CONF_SPEAKER,
)
_LOGGER = logging.getLogger(__name__)
@ -24,17 +32,23 @@ IsPlayingCondition = rtttl_ns.class_("IsPlayingCondition", automation.Condition)
MULTI_CONF = True
CONFIG_SCHEMA = cv.Schema(
{
cv.GenerateID(CONF_ID): cv.declare_id(Rtttl),
cv.Required(CONF_OUTPUT): cv.use_id(FloatOutput),
cv.Optional(CONF_ON_FINISHED_PLAYBACK): automation.validate_automation(
{
cv.GenerateID(CONF_TRIGGER_ID): cv.declare_id(FinishedPlaybackTrigger),
}
),
}
).extend(cv.COMPONENT_SCHEMA)
CONFIG_SCHEMA = cv.All(
cv.Schema(
{
cv.GenerateID(CONF_ID): cv.declare_id(Rtttl),
cv.Optional(CONF_OUTPUT): cv.use_id(FloatOutput),
cv.Optional(CONF_SPEAKER): cv.use_id(Speaker),
cv.Optional(CONF_ON_FINISHED_PLAYBACK): automation.validate_automation(
{
cv.GenerateID(CONF_TRIGGER_ID): cv.declare_id(
FinishedPlaybackTrigger
),
}
),
}
).extend(cv.COMPONENT_SCHEMA),
cv.has_exactly_one_key(CONF_OUTPUT, CONF_SPEAKER),
)
def validate_parent_output_config(value):
@ -63,9 +77,9 @@ def validate_parent_output_config(value):
FINAL_VALIDATE_SCHEMA = cv.Schema(
{
cv.Required(CONF_OUTPUT): fv.id_declaration_match_schema(
cv.Optional(CONF_OUTPUT): fv.id_declaration_match_schema(
validate_parent_output_config
)
),
},
extra=cv.ALLOW_EXTRA,
)
@ -75,8 +89,14 @@ async def to_code(config):
var = cg.new_Pvariable(config[CONF_ID])
await cg.register_component(var, config)
out = await cg.get_variable(config[CONF_OUTPUT])
cg.add(var.set_output(out))
if CONF_OUTPUT in config:
out = await cg.get_variable(config[CONF_OUTPUT])
cg.add(var.set_output(out))
cg.add_define("USE_OUTPUT")
if CONF_SPEAKER in config:
out = await cg.get_variable(config[CONF_SPEAKER])
cg.add(var.set_speaker(out))
for conf in config.get(CONF_ON_FINISHED_PLAYBACK, []):
trigger = cg.new_Pvariable(conf[CONF_TRIGGER_ID], var)

View file

@ -1,4 +1,5 @@
#include "rtttl.h"
#include <cmath>
#include "esphome/core/hal.h"
#include "esphome/core/log.h"
@ -15,104 +16,185 @@ static const uint16_t NOTES[] = {0, 262, 277, 294, 311, 330, 349, 370,
1109, 1175, 1245, 1319, 1397, 1480, 1568, 1661, 1760, 1865, 1976, 2093, 2217,
2349, 2489, 2637, 2794, 2960, 3136, 3322, 3520, 3729, 3951};
static const uint16_t I2S_SPEED = 1600;
#undef HALF_PI
static const double HALF_PI = 1.5707963267948966192313216916398;
inline double deg2rad(double degrees) {
static const double PI_ON_180 = 4.0 * atan(1.0) / 180.0;
return degrees * PI_ON_180;
}
void Rtttl::dump_config() { ESP_LOGCONFIG(TAG, "Rtttl"); }
void Rtttl::play(std::string rtttl) {
rtttl_ = std::move(rtttl);
this->rtttl_ = std::move(rtttl);
this->default_duration_ = 4;
this->default_octave_ = 6;
this->note_duration_ = 0;
default_duration_ = 4;
default_octave_ = 6;
int bpm = 63;
uint8_t num;
// Get name
position_ = rtttl_.find(':');
this->position_ = rtttl_.find(':');
// it's somewhat documented to be up to 10 characters but let's be a bit flexible here
if (position_ == std::string::npos || position_ > 15) {
if (this->position_ == std::string::npos || this->position_ > 15) {
ESP_LOGE(TAG, "Missing ':' when looking for name.");
return;
}
auto name = this->rtttl_.substr(0, position_);
auto name = this->rtttl_.substr(0, this->position_);
ESP_LOGD(TAG, "Playing song %s", name.c_str());
// get default duration
position_ = this->rtttl_.find("d=", position_);
if (position_ == std::string::npos) {
this->position_ = this->rtttl_.find("d=", this->position_);
if (this->position_ == std::string::npos) {
ESP_LOGE(TAG, "Missing 'd='");
return;
}
position_ += 2;
this->position_ += 2;
num = this->get_integer_();
if (num > 0)
default_duration_ = num;
this->default_duration_ = num;
// get default octave
position_ = rtttl_.find("o=", position_);
if (position_ == std::string::npos) {
this->position_ = this->rtttl_.find("o=", this->position_);
if (this->position_ == std::string::npos) {
ESP_LOGE(TAG, "Missing 'o=");
return;
}
position_ += 2;
this->position_ += 2;
num = get_integer_();
if (num >= 3 && num <= 7)
default_octave_ = num;
this->default_octave_ = num;
// get BPM
position_ = rtttl_.find("b=", position_);
if (position_ == std::string::npos) {
this->position_ = this->rtttl_.find("b=", this->position_);
if (this->position_ == std::string::npos) {
ESP_LOGE(TAG, "Missing b=");
return;
}
position_ += 2;
this->position_ += 2;
num = get_integer_();
if (num != 0)
bpm = num;
position_ = rtttl_.find(':', position_);
if (position_ == std::string::npos) {
this->position_ = this->rtttl_.find(':', this->position_);
if (this->position_ == std::string::npos) {
ESP_LOGE(TAG, "Missing second ':'");
return;
}
position_++;
this->position_++;
// BPM usually expresses the number of quarter notes per minute
wholenote_ = 60 * 1000L * 4 / bpm; // this is the time for whole note (in milliseconds)
this->wholenote_ = 60 * 1000L * 4 / bpm; // this is the time for whole note (in milliseconds)
output_freq_ = 0;
last_note_ = millis();
note_duration_ = 1;
this->output_freq_ = 0;
this->last_note_ = millis();
this->note_duration_ = 1;
#ifdef USE_SPEAKER
this->samples_sent_ = 0;
this->samples_count_ = 0;
#endif
}
void Rtttl::stop() {
this->note_duration_ = 0;
#ifdef USE_OUTPUT
if (this->output_ != nullptr) {
this->output_->set_level(0.0);
}
#endif
#ifdef USE_SPEAKER
if (this->speaker_ != nullptr) {
if (this->speaker_->is_running()) {
this->speaker_->stop();
}
}
#endif
}
void Rtttl::loop() {
if (note_duration_ == 0 || millis() - last_note_ < note_duration_)
if (this->note_duration_ == 0)
return;
if (!rtttl_[position_]) {
output_->set_level(0.0);
#ifdef USE_SPEAKER
if (this->speaker_ != nullptr) {
if (this->samples_sent_ != this->samples_count_) {
SpeakerSample sample[SAMPLE_BUFFER_SIZE + 1];
int x = 0;
double rem = 0.0;
while (true) {
// Try and send out the remainder of the existing note, one per loop()
if (this->samples_per_wave_ != 0 && this->samples_sent_ >= this->samples_gap_) { // Play note//
rem = ((this->samples_sent_ << 10) % this->samples_per_wave_) * (360.0 / this->samples_per_wave_);
int16_t val = 8192 * sin(deg2rad(rem));
sample[x].left = val;
sample[x].right = val;
} else {
sample[x].left = 0;
sample[x].right = 0;
}
if (x >= SAMPLE_BUFFER_SIZE || this->samples_sent_ >= this->samples_count_) {
break;
}
this->samples_sent_++;
x++;
}
if (x > 0) {
int send = this->speaker_->play((uint8_t *) (&sample), x * 4);
if (send != x * 4) {
this->samples_sent_ -= (x - (send / 4));
}
return;
}
}
}
#endif
#ifdef USE_OUTPUT
if (this->output_ != nullptr && millis() - this->last_note_ < this->note_duration_)
return;
#endif
if (!this->rtttl_[position_]) {
this->note_duration_ = 0;
#ifdef USE_OUTPUT
if (this->output_ != nullptr) {
this->output_->set_level(0.0);
}
#endif
ESP_LOGD(TAG, "Playback finished");
this->on_finished_playback_callback_.call();
note_duration_ = 0;
return;
}
// align to note: most rtttl's out there does not add and space after the ',' separator but just in case...
while (rtttl_[position_] == ',' || rtttl_[position_] == ' ')
position_++;
while (this->rtttl_[this->position_] == ',' || this->rtttl_[this->position_] == ' ')
this->position_++;
// first, get note duration, if available
uint8_t num = this->get_integer_();
if (num) {
note_duration_ = wholenote_ / num;
this->note_duration_ = this->wholenote_ / num;
} else {
note_duration_ = wholenote_ / default_duration_; // we will need to check if we are a dotted note after
this->note_duration_ =
this->wholenote_ / this->default_duration_; // we will need to check if we are a dotted note after
}
uint8_t note;
switch (rtttl_[position_]) {
switch (this->rtttl_[this->position_]) {
case 'c':
note = 1;
break;
@ -138,51 +220,81 @@ void Rtttl::loop() {
default:
note = 0;
}
position_++;
this->position_++;
// now, get optional '#' sharp
if (rtttl_[position_] == '#') {
if (this->rtttl_[this->position_] == '#') {
note++;
position_++;
this->position_++;
}
// now, get optional '.' dotted note
if (rtttl_[position_] == '.') {
note_duration_ += note_duration_ / 2;
position_++;
if (this->rtttl_[this->position_] == '.') {
this->note_duration_ += this->note_duration_ / 2;
this->position_++;
}
// now, get scale
uint8_t scale = get_integer_();
if (scale == 0)
scale = default_octave_;
scale = this->default_octave_;
bool need_note_gap = false;
// Now play the note
if (note) {
auto note_index = (scale - 4) * 12 + note;
if (note_index < 0 || note_index >= (int) sizeof(NOTES)) {
ESP_LOGE(TAG, "Note out of valid range");
this->note_duration_ = 0;
return;
}
auto freq = NOTES[note_index];
need_note_gap = freq == this->output_freq_;
if (freq == output_freq_) {
// Add small silence gap between same note
output_->set_level(0.0);
delay(DOUBLE_NOTE_GAP_MS);
note_duration_ -= DOUBLE_NOTE_GAP_MS;
}
output_freq_ = freq;
// Add small silence gap between same note
this->output_freq_ = freq;
ESP_LOGVV(TAG, "playing note: %d for %dms", note, note_duration_);
output_->update_frequency(freq);
output_->set_level(0.5);
ESP_LOGVV(TAG, "playing note: %d for %dms", note, this->note_duration_);
} else {
ESP_LOGVV(TAG, "waiting: %dms", note_duration_);
output_->set_level(0.0);
ESP_LOGVV(TAG, "waiting: %dms", this->note_duration_);
this->output_freq_ = 0;
}
last_note_ = millis();
#ifdef USE_OUTPUT
if (this->output_ != nullptr) {
if (need_note_gap) {
this->output_->set_level(0.0);
delay(DOUBLE_NOTE_GAP_MS);
this->note_duration_ -= DOUBLE_NOTE_GAP_MS;
}
if (this->output_freq_ != 0) {
this->output_->update_frequency(this->output_freq_);
this->output_->set_level(0.5);
} else {
this->output_->set_level(0.0);
}
}
#endif
#ifdef USE_SPEAKER
if (this->speaker_ != nullptr) {
this->samples_sent_ = 0;
this->samples_count_ = (this->sample_rate_ * this->note_duration_) / I2S_SPEED;
// Convert from frequency in Hz to high and low samples in fixed point
if (this->output_freq_ != 0) {
this->samples_per_wave_ = (this->sample_rate_ << 10) / this->output_freq_;
} else {
this->samples_per_wave_ = 0;
}
if (need_note_gap) {
this->samples_gap_ = (this->sample_rate_ * DOUBLE_NOTE_GAP_MS) / I2S_SPEED;
} else {
this->samples_gap_ = 0;
}
}
#endif
this->last_note_ = millis();
}
} // namespace rtttl
} // namespace esphome

View file

@ -1,23 +1,41 @@
#pragma once
#include "esphome/core/component.h"
#include "esphome/core/automation.h"
#include "esphome/core/component.h"
#ifdef USE_OUTPUT
#include "esphome/components/output/float_output.h"
#endif
#ifdef USE_SPEAKER
#include "esphome/components/speaker/speaker.h"
#endif
namespace esphome {
namespace rtttl {
#ifdef USE_SPEAKER
static const size_t SAMPLE_BUFFER_SIZE = 256;
struct SpeakerSample {
int16_t left{0};
int16_t right{0};
};
#endif
class Rtttl : public Component {
public:
void set_output(output::FloatOutput *output) { output_ = output; }
#ifdef USE_OUTPUT
void set_output(output::FloatOutput *output) { this->output_ = output; }
#endif
#ifdef USE_SPEAKER
void set_speaker(speaker::Speaker *speaker) { this->speaker_ = speaker; }
#endif
void play(std::string rtttl);
void stop() {
note_duration_ = 0;
output_->set_level(0.0);
}
void stop();
void dump_config() override;
bool is_playing() { return note_duration_ != 0; }
bool is_playing() { return this->note_duration_ != 0; }
void loop() override;
void add_on_finished_playback_callback(std::function<void()> callback) {
@ -27,14 +45,14 @@ class Rtttl : public Component {
protected:
inline uint8_t get_integer_() {
uint8_t ret = 0;
while (isdigit(rtttl_[position_])) {
ret = (ret * 10) + (rtttl_[position_++] - '0');
while (isdigit(this->rtttl_[this->position_])) {
ret = (ret * 10) + (this->rtttl_[this->position_++] - '0');
}
return ret;
}
std::string rtttl_;
size_t position_;
std::string rtttl_{""};
size_t position_{0};
uint16_t wholenote_;
uint16_t default_duration_;
uint16_t default_octave_;
@ -42,7 +60,22 @@ class Rtttl : public Component {
uint16_t note_duration_;
uint32_t output_freq_;
#ifdef USE_OUTPUT
output::FloatOutput *output_;
#endif
void play_output_();
#ifdef USE_SPEAKER
speaker::Speaker *speaker_;
void play_speaker_();
int sample_rate_{16000};
int samples_per_wave_{0};
int samples_sent_{0};
int samples_count_{0};
int samples_gap_{0};
#endif
CallbackManager<void()> on_finished_playback_callback_;
};