diff options
Diffstat (limited to 'platform/linuxbsd/tts_linux.cpp')
-rw-r--r-- | platform/linuxbsd/tts_linux.cpp | 241 |
1 files changed, 133 insertions, 108 deletions
diff --git a/platform/linuxbsd/tts_linux.cpp b/platform/linuxbsd/tts_linux.cpp index aea1183d3d..8bc83f5b9c 100644 --- a/platform/linuxbsd/tts_linux.cpp +++ b/platform/linuxbsd/tts_linux.cpp @@ -1,32 +1,32 @@ -/*************************************************************************/ -/* tts_linux.cpp */ -/*************************************************************************/ -/* This file is part of: */ -/* GODOT ENGINE */ -/* https://godotengine.org */ -/*************************************************************************/ -/* Copyright (c) 2007-2022 Juan Linietsky, Ariel Manzur. */ -/* Copyright (c) 2014-2022 Godot Engine contributors (cf. AUTHORS.md). */ -/* */ -/* 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. */ -/*************************************************************************/ +/**************************************************************************/ +/* tts_linux.cpp */ +/**************************************************************************/ +/* This file is part of: */ +/* GODOT ENGINE */ +/* https://godotengine.org */ +/**************************************************************************/ +/* Copyright (c) 2014-present Godot Engine contributors (see AUTHORS.md). */ +/* Copyright (c) 2007-2014 Juan Linietsky, Ariel Manzur. */ +/* */ +/* 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. */ +/**************************************************************************/ #include "tts_linux.h" @@ -39,12 +39,23 @@ void TTS_Linux::speech_init_thread_func(void *p_userdata) { TTS_Linux *tts = (TTS_Linux *)p_userdata; if (tts) { MutexLock thread_safe_method(tts->_thread_safe_); +#ifdef SOWRAP_ENABLED #ifdef DEBUG_ENABLED int dylibloader_verbose = 1; #else int dylibloader_verbose = 0; #endif - if (initialize_speechd(dylibloader_verbose) == 0) { + if (initialize_speechd(dylibloader_verbose) != 0) { + print_verbose("Text-to-Speech: Cannot load Speech Dispatcher library!"); + } else { + if (!spd_open || !spd_set_notification_on || !spd_list_synthesis_voices || !free_spd_voices || !spd_set_synthesis_voice || !spd_set_volume || !spd_set_voice_pitch || !spd_set_voice_rate || !spd_set_data_mode || !spd_say || !spd_pause || !spd_resume || !spd_cancel) { + // There's no API to check version, check if functions are available instead. + print_verbose("Text-to-Speech: Unsupported Speech Dispatcher library version!"); + return; + } +#else + { +#endif CharString class_str; String config_name = GLOBAL_GET("application/config/name"); if (config_name.length() == 0) { @@ -64,85 +75,103 @@ void TTS_Linux::speech_init_thread_func(void *p_userdata) { } else { print_verbose("Text-to-Speech: Cannot initialize Speech Dispatcher synthesizer!"); } - } else { - print_verbose("Text-to-Speech: Cannot load Speech Dispatcher library!"); } } } void TTS_Linux::speech_event_index_mark(size_t p_msg_id, size_t p_client_id, SPDNotificationType p_type, char *p_index_mark) { TTS_Linux *tts = TTS_Linux::get_singleton(); - if (tts && tts->ids.has(p_msg_id)) { - MutexLock thread_safe_method(tts->_thread_safe_); - // Get word offset from the index mark injected to the text stream. - String mark = String::utf8(p_index_mark); - DisplayServer::get_singleton()->tts_post_utterance_event(DisplayServer::TTS_UTTERANCE_BOUNDARY, tts->ids[p_msg_id], mark.to_int()); + if (tts) { + callable_mp(tts, &TTS_Linux::_speech_index_mark).call_deferred(p_msg_id, p_client_id, (int)p_type, String::utf8(p_index_mark)); + } +} + +void TTS_Linux::_speech_index_mark(size_t p_msg_id, size_t p_client_id, int p_type, const String &p_index_mark) { + _THREAD_SAFE_METHOD_ + + if (ids.has(p_msg_id)) { + DisplayServer::get_singleton()->tts_post_utterance_event(DisplayServer::TTS_UTTERANCE_BOUNDARY, ids[p_msg_id], p_index_mark.to_int()); } } void TTS_Linux::speech_event_callback(size_t p_msg_id, size_t p_client_id, SPDNotificationType p_type) { TTS_Linux *tts = TTS_Linux::get_singleton(); if (tts) { - MutexLock thread_safe_method(tts->_thread_safe_); - List<DisplayServer::TTSUtterance> &queue = tts->queue; - if (!tts->paused && tts->ids.has(p_msg_id)) { - if (p_type == SPD_EVENT_END) { - DisplayServer::get_singleton()->tts_post_utterance_event(DisplayServer::TTS_UTTERANCE_ENDED, tts->ids[p_msg_id]); - tts->ids.erase(p_msg_id); - tts->last_msg_id = -1; - tts->speaking = false; - } else if (p_type == SPD_EVENT_CANCEL) { - DisplayServer::get_singleton()->tts_post_utterance_event(DisplayServer::TTS_UTTERANCE_CANCELED, tts->ids[p_msg_id]); - tts->ids.erase(p_msg_id); - tts->last_msg_id = -1; - tts->speaking = false; + callable_mp(tts, &TTS_Linux::_speech_event).call_deferred(p_msg_id, p_client_id, (int)p_type); + } +} + +void TTS_Linux::_load_voices() { + if (!voices_loaded) { + SPDVoice **spd_voices = spd_list_synthesis_voices(synth); + if (spd_voices != nullptr) { + SPDVoice **voices_ptr = spd_voices; + while (*voices_ptr != nullptr) { + VoiceInfo vi; + vi.language = String::utf8((*voices_ptr)->language); + vi.variant = String::utf8((*voices_ptr)->variant); + voices[String::utf8((*voices_ptr)->name)] = vi; + voices_ptr++; } + free_spd_voices(spd_voices); } - if (!tts->speaking && queue.size() > 0) { - DisplayServer::TTSUtterance &message = queue.front()->get(); - - // Inject index mark after each word. - String text; - String language; - SPDVoice **voices = spd_list_synthesis_voices(tts->synth); - if (voices != nullptr) { - SPDVoice **voices_ptr = voices; - while (*voices_ptr != nullptr) { - if (String::utf8((*voices_ptr)->name) == message.voice) { - language = String::utf8((*voices_ptr)->language); - break; - } - voices_ptr++; - } - free_spd_voices(voices); - } - PackedInt32Array breaks = TS->string_get_word_breaks(message.text, language); - int prev = 0; - for (int i = 0; i < breaks.size(); i++) { - text += message.text.substr(prev, breaks[i] - prev); - text += "<mark name=\"" + String::num_int64(breaks[i], 10) + "\"/>"; - prev = breaks[i]; - } - text += message.text.substr(prev, -1); - - spd_set_synthesis_voice(tts->synth, message.voice.utf8().get_data()); - spd_set_volume(tts->synth, message.volume * 2 - 100); - spd_set_voice_pitch(tts->synth, (message.pitch - 1) * 100); - float rate = 0; - if (message.rate > 1.f) { - rate = log10(MIN(message.rate, 2.5f)) / log10(2.5f) * 100; - } else if (message.rate < 1.f) { - rate = log10(MAX(message.rate, 0.5f)) / log10(0.5f) * -100; - } - spd_set_voice_rate(tts->synth, rate); - spd_set_data_mode(tts->synth, SPD_DATA_SSML); - tts->last_msg_id = spd_say(tts->synth, SPD_TEXT, text.utf8().get_data()); - tts->ids[tts->last_msg_id] = message.id; - DisplayServer::get_singleton()->tts_post_utterance_event(DisplayServer::TTS_UTTERANCE_STARTED, message.id); - - queue.pop_front(); - tts->speaking = true; + voices_loaded = true; + } +} + +void TTS_Linux::_speech_event(size_t p_msg_id, size_t p_client_id, int p_type) { + _THREAD_SAFE_METHOD_ + + if (!paused && ids.has(p_msg_id)) { + if ((SPDNotificationType)p_type == SPD_EVENT_END) { + DisplayServer::get_singleton()->tts_post_utterance_event(DisplayServer::TTS_UTTERANCE_ENDED, ids[p_msg_id]); + ids.erase(p_msg_id); + last_msg_id = -1; + speaking = false; + } else if ((SPDNotificationType)p_type == SPD_EVENT_CANCEL) { + DisplayServer::get_singleton()->tts_post_utterance_event(DisplayServer::TTS_UTTERANCE_CANCELED, ids[p_msg_id]); + ids.erase(p_msg_id); + last_msg_id = -1; + speaking = false; + } + } + if (!speaking && queue.size() > 0) { + DisplayServer::TTSUtterance &message = queue.front()->get(); + + // Inject index mark after each word. + String text; + String language; + + _load_voices(); + const VoiceInfo *voice = voices.getptr(message.voice); + if (voice) { + language = voice->language; + } + + PackedInt32Array breaks = TS->string_get_word_breaks(message.text, language); + for (int i = 0; i < breaks.size(); i += 2) { + const int start = breaks[i]; + const int end = breaks[i + 1]; + text += message.text.substr(start, end - start + 1); + text += "<mark name=\"" + String::num_int64(end, 10) + "\"/>"; + } + spd_set_synthesis_voice(synth, message.voice.utf8().get_data()); + spd_set_volume(synth, message.volume * 2 - 100); + spd_set_voice_pitch(synth, (message.pitch - 1) * 100); + float rate = 0; + if (message.rate > 1.f) { + rate = log10(MIN(message.rate, 2.5f)) / log10(2.5f) * 100; + } else if (message.rate < 1.f) { + rate = log10(MAX(message.rate, 0.5f)) / log10(0.5f) * -100; } + spd_set_voice_rate(synth, rate); + spd_set_data_mode(synth, SPD_DATA_SSML); + last_msg_id = spd_say(synth, SPD_TEXT, text.utf8().get_data()); + ids[last_msg_id] = message.id; + DisplayServer::get_singleton()->tts_post_utterance_event(DisplayServer::TTS_UTTERANCE_STARTED, message.id); + + queue.pop_front(); + speaking = true; } } @@ -158,21 +187,17 @@ Array TTS_Linux::get_voices() const { _THREAD_SAFE_METHOD_ ERR_FAIL_COND_V(!synth, Array()); + const_cast<TTS_Linux *>(this)->_load_voices(); + Array list; - SPDVoice **voices = spd_list_synthesis_voices(synth); - if (voices != nullptr) { - SPDVoice **voices_ptr = voices; - while (*voices_ptr != nullptr) { - Dictionary voice_d; - voice_d["name"] = String::utf8((*voices_ptr)->name); - voice_d["id"] = String::utf8((*voices_ptr)->name); - voice_d["language"] = String::utf8((*voices_ptr)->language) + "_" + String::utf8((*voices_ptr)->variant); - list.push_back(voice_d); - - voices_ptr++; - } - free_spd_voices(voices); + for (const KeyValue<String, VoiceInfo> &E : voices) { + Dictionary voice_d; + voice_d["name"] = E.key; + voice_d["id"] = E.key; + voice_d["language"] = E.value.language + "_" + E.value.variant; + list.push_back(voice_d); } + return list; } @@ -201,7 +226,7 @@ void TTS_Linux::speak(const String &p_text, const String &p_voice, int p_volume, if (is_paused()) { resume(); } else { - speech_event_callback(0, 0, SPD_EVENT_BEGIN); + _speech_event(0, 0, (int)SPD_EVENT_BEGIN); } } |