TTS: Implement our own queuing for linux

It seems like, that at least some versions of speech-dispatcher
aren't able to successfuly pause and resume. For me, when trying
to pause, it still finishes the speech just being said instead
of pausing it and then it puts it at the end of the speech queue
with some speech-dispatcher internal commands added to it, which
are also hearable.

There is no way to find out where the speech ended when calling
pause, so it is just stopped and when resume is called it is
read from it's start again.

2 files changed, 68 insertions, 27 deletions

diff --git a/backends/text-to-speech/linux/linux-text-to-speech.cpp b/backends/text-to-speech/linux/linux-text-to-speech.cpp
index 97d05a67c1..20ba07d74d 100644
--- a/backends/text-to-speech/linux/linux-text-to-speech.cpp
+++ b/backends/text-to-speech/linux/linux-text-to-speech.cpp
@@ -28,45 +28,42 @@
 #if defined(USE_LINUX_TTS)
 #include <speech-dispatcher/libspeechd.h>
 #include "backends/platform/sdl/sdl-sys.h"
-//#include <iconv.h>
 
 #include "common/translation.h"
 #include "common/system.h"
 #include "common/ustr.h"
 #include "common/config-manager.h"
+
 SPDConnection *_connection;
 
 void speech_begin_callback(size_t msg_id, size_t client_id, SPDNotificationType state){
 	LinuxTextToSpeechManager *manager =
 		static_cast<LinuxTextToSpeechManager *> (g_system->getTextToSpeechManager());
-	manager->updateState(LinuxTextToSpeechManager::SPEAKING);
+	manager->updateState(LinuxTextToSpeechManager::SPEECH_BEGUN);
 }
 
 void speech_end_callback(size_t msg_id, size_t client_id, SPDNotificationType state){
 	LinuxTextToSpeechManager *manager =
 		static_cast<LinuxTextToSpeechManager *> (g_system->getTextToSpeechManager());
-	manager->updateState(LinuxTextToSpeechManager::READY);
+	manager->updateState(LinuxTextToSpeechManager::SPEECH_ENDED);
 }
 
 void speech_cancel_callback(size_t msg_id, size_t client_id, SPDNotificationType state){
 	LinuxTextToSpeechManager *manager =
 		static_cast<LinuxTextToSpeechManager *> (g_system->getTextToSpeechManager());
-	if (manager->isSpeaking())
-		manager->updateState(LinuxTextToSpeechManager::READY);
-	if (manager->isPaused())
-		manager->updateState(LinuxTextToSpeechManager::PAUSED);
+	manager->updateState(LinuxTextToSpeechManager::SPEECH_CANCELED);
 }
 
 void speech_resume_callback(size_t msg_id, size_t client_id, SPDNotificationType state){
 	LinuxTextToSpeechManager *manager =
 		static_cast<LinuxTextToSpeechManager *> (g_system->getTextToSpeechManager());
-	manager->updateState(LinuxTextToSpeechManager::SPEAKING);
+	manager->updateState(LinuxTextToSpeechManager::SPEECH_RESUMED);
 }
 
 void speech_pause_callback(size_t msg_id, size_t client_id, SPDNotificationType state){
 	LinuxTextToSpeechManager *manager =
 		static_cast<LinuxTextToSpeechManager *> (g_system->getTextToSpeechManager());
-	manager->updateState(LinuxTextToSpeechManager::PAUSED);
+	manager->updateState(LinuxTextToSpeechManager::SPEECH_PAUSED);
 }
 
 LinuxTextToSpeechManager::LinuxTextToSpeechManager()
@@ -101,6 +98,7 @@ void LinuxTextToSpeechManager::init() {
 #else
 	setLanguage("en");
 #endif
+	_speechQueue.clear();
 }
 
 LinuxTextToSpeechManager::~LinuxTextToSpeechManager() {
@@ -108,8 +106,29 @@ LinuxTextToSpeechManager::~LinuxTextToSpeechManager() {
 		spd_close(_connection);
 }
 
-void LinuxTextToSpeechManager::updateState(LinuxTextToSpeechManager::SpeechState state) {
-	_speechState = state;
+void LinuxTextToSpeechManager::updateState(LinuxTextToSpeechManager::SpeechEvent event) {
+	if (_speechState == BROKEN)
+		return;
+	switch(event) {
+	case SPEECH_ENDED:
+		_speechQueue.pop_front();
+		if (_speechQueue.size() == 0)
+			_speechState = READY;
+		break;
+	case SPEECH_PAUSED:
+		_speechState = PAUSED;
+		break;
+	case SPEECH_CANCELED:
+		if (_speechState != PAUSED) {
+			_speechState = READY;
+		}
+		break;
+	case SPEECH_RESUMED:
+		break;
+	case SPEECH_BEGUN:
+		_speechState = SPEAKING;
+		break;
+	}
 }
 
 Common::String LinuxTextToSpeechManager::strToUtf8(Common::String str, Common::String charset) {
@@ -155,17 +174,19 @@ bool LinuxTextToSpeechManager::say(Common::String str, Action action, Common::St
 
 	str = strToUtf8(str, charset);
 
-	if (isSpeaking() && action == INTERRUPT || action == INTERRUPT_NO_REPEAT)
+	if (isSpeaking() && (action == INTERRUPT || action == INTERRUPT_NO_REPEAT))
 		stop();
-	if (str.size() != 0)
+	if (!str.empty()) {
 		_speechState = SPEAKING;
-	_lastSaid = str;
-	if(spd_say(_connection, SPD_MESSAGE, str.c_str()) == -1) {
-		//restart the connection
-		if (_connection != 0)
-			spd_close(_connection);
-		init();
-		return true;
+		_speechQueue.push_back(str);
+		_lastSaid = str;
+		if(spd_say(_connection, SPD_MESSAGE, str.c_str()) == -1) {
+			//restart the connection
+			if (_connection != 0)
+				spd_close(_connection);
+			init();
+			return true;
+		}
 	}
 
 	return false;
@@ -175,27 +196,35 @@ bool LinuxTextToSpeechManager::stop() {
 	if (_speechState == READY || _speechState == BROKEN)
 		return true;
 	_speechState = READY;
+	_speechQueue.clear();
 	return spd_cancel(_connection) == -1;
 }
 
 bool LinuxTextToSpeechManager::pause() {
 	if (_speechState == READY || _speechState == PAUSED || _speechState == BROKEN)
 		return true;
-	bool result = spd_pause_all(_connection) == -1;
+	_speechState = PAUSED;
+	bool result = spd_cancel_all(_connection) == -1;
 	if (result)
 		return true;
-	result = spd_stop(_connection) == -1;
 	if (result)
 		return true;
-	_speechState = PAUSED;
 	return false;
 }
 
 bool LinuxTextToSpeechManager::resume() {
 	if (_speechState == READY || _speechState == SPEAKING || _speechState == BROKEN)
 		return true;
-	_speechState = SPEAKING;
-	return spd_resume(_connection) == -1;
+	if (_speechQueue.size()) {
+		_speechState = SPEAKING;
+		for (Common::List<Common::String>::iterator i = _speechQueue.begin(); i != _speechQueue.end(); i++) {
+			if (spd_say(_connection, SPD_MESSAGE, i->c_str()) == -1)
+				return true;
+		}
+	}
+	else
+		_speechState = READY;
+	return false;
 }
 
 bool LinuxTextToSpeechManager::isSpeaking() {
diff --git a/backends/text-to-speech/linux/linux-text-to-speech.h b/backends/text-to-speech/linux/linux-text-to-speech.h
index 30fa9b878c..23c35c33f6 100644
--- a/backends/text-to-speech/linux/linux-text-to-speech.h
+++ b/backends/text-to-speech/linux/linux-text-to-speech.h
@@ -29,6 +29,7 @@
 
 #include "common/text-to-speech.h"
 #include "common/str.h"
+#include "common/list.h"
 
 class LinuxTextToSpeechManager : public Common::TextToSpeechManager {
 public:
@@ -39,6 +40,14 @@ public:
 		BROKEN
 	};
 
+	enum SpeechEvent {
+		SPEECH_ENDED,
+		SPEECH_PAUSED,
+		SPEECH_CANCELED,
+		SPEECH_RESUMED,
+		SPEECH_BEGUN
+	};
+
 	LinuxTextToSpeechManager();
 	virtual ~LinuxTextToSpeechManager();
 
@@ -62,7 +71,7 @@ public:
 
 	virtual void setLanguage(Common::String language);
 
-	void updateState(SpeechState state);
+	void updateState(SpeechEvent event);
 
 	virtual void freeVoiceData(void *data);
 
@@ -70,9 +79,12 @@ private:
 	void init();
 	virtual void updateVoices();
 	void createVoice(int typeNumber, Common::TTSVoice::Gender, Common::TTSVoice::Age, char *description);
-	SpeechState _speechState;
 	Common::String strToUtf8(Common::String str, Common::String charset);
+	bool spdSay(const char *str);
+
+	SpeechState _speechState;
 	Common::String _lastSaid;
+	Common::List<Common::String> _speechQueue;
 };
 
 #endif