Add text to speech capability to text backend

Too much code stolen from Read EText activity

8 files changed, 405 insertions, 2 deletions

diff --git a/epubadapter.py b/epubadapter.py
index b03966f..2b1b597 100644
--- a/epubadapter.py
+++ b/epubadapter.py
@@ -49,6 +49,9 @@ class EpubViewer(epubview.EpubView):
     def can_highlight(self):
         return False
 
+    def can_do_text_to_speech(self):
+        return False
+
     def connect_zoom_handler(self, handler):
         self._zoom_handler = handler
         self._view_notify_zoom_handler = \
diff --git a/evinceadapter.py b/evinceadapter.py
index c37562c..a4d8aca 100644
--- a/evinceadapter.py
+++ b/evinceadapter.py
@@ -112,6 +112,9 @@ class EvinceViewer():
     def can_highlight(self):
         return False
 
+    def can_do_text_to_speech(self):
+        return False
+
     def get_zoom(self):
         '''
         Returns the current zoom level
diff --git a/readactivity.py b/readactivity.py
index 1fb2ffc..8e1ce1c 100644
--- a/readactivity.py
+++ b/readactivity.py
@@ -45,13 +45,16 @@ from sugar import mime
 from sugar.datastore import datastore
 from sugar.graphics.objectchooser import ObjectChooser
 
-from readtoolbar import EditToolbar, ViewToolbar
+from readtoolbar import EditToolbar
+from readtoolbar import ViewToolbar
+from readtoolbar import SpeechToolbar
 from readsidebar import Sidebar
 from readtopbar import TopBar
 from readdb import BookmarkManager
 import epubadapter
 import evinceadapter
 import textadapter
+import speech
 
 _HARDWARE_MANAGER_INTERFACE = 'org.laptop.HardwareManager'
 _HARDWARE_MANAGER_SERVICE = 'org.laptop.HardwareManager'
@@ -244,6 +247,11 @@ class ReadActivity(activity.Activity):
         toolbar_box.toolbar.insert(self._highlight_item, -1)
         self._highlight_item.show_all()
 
+        self.speech_toolbar = SpeechToolbar(self)
+        self.speech_toolbar_button = ToolbarButton(page=self.speech_toolbar,
+                    icon_name='speak')
+        toolbar_box.toolbar.insert(self.speech_toolbar_button, -1)
+
         separator = gtk.SeparatorToolItem()
         separator.props.draw = False
         separator.set_expand(True)
@@ -790,7 +798,7 @@ class ReadActivity(activity.Activity):
         mimetype = mime.get_for_file(filepath)
         if mimetype == 'application/epub+zip':
             self._view = epubadapter.EpubViewer()
-        elif mimetype == 'text/plain' or mimetype == 'application/zip' :
+        elif mimetype == 'text/plain' or mimetype == 'application/zip':
             self._view = textadapter.TextViewer()
         else:
             self._view = evinceadapter.EvinceViewer()
@@ -833,6 +841,9 @@ class ReadActivity(activity.Activity):
         self._view_toolbar._update_zoom_buttons()
         if not self._view.can_highlight():
             self._highlight_item.hide()
+        if speech.supported and self._view.can_do_text_to_speech():
+            self.speech_toolbar_button.show()
+            self.speech_toolbar_button.show()
 
     def _share_document(self):
         """Share the document."""
diff --git a/readtoolbar.py b/readtoolbar.py
index ed57fde..1da23bf 100644
--- a/readtoolbar.py
+++ b/readtoolbar.py
@@ -20,11 +20,16 @@ import logging
 import gobject
 import gtk
 
+from sugar.graphics.combobox import ComboBox
 from sugar.graphics.toolbutton import ToolButton
+from sugar.graphics.toggletoolbutton import ToggleToolButton
+from sugar.graphics.toolcombobox import ToolComboBox
 from sugar.graphics.menuitem import MenuItem
 from sugar.graphics import iconentry
 from sugar.activity import activity
 
+import speech
+
 
 class EditToolbar(activity.EditToolbar):
 
@@ -288,3 +293,69 @@ class ViewToolbar(gtk.Toolbar):
 
     def _fullscreen_cb(self, button):
         self.emit('go-fullscreen')
+
+
+class SpeechToolbar(gtk.Toolbar):
+
+    def __init__(self, activity):
+        gtk.Toolbar.__init__(self)
+        voicebar = gtk.Toolbar()
+        self.activity = activity
+        self.sorted_voices = [i for i in speech.voices()]
+        self.sorted_voices.sort(self.compare_voices)
+        default = 0
+        for voice in self.sorted_voices:
+            if voice[0] == 'default':
+                break
+            default = default + 1
+
+        # Play button Image
+        play_img = gtk.Image()
+        play_img.show()
+        play_img.set_from_icon_name('media-playback-start',
+                gtk.ICON_SIZE_LARGE_TOOLBAR)
+
+        # Pause button Image
+        pause_img = gtk.Image()
+        pause_img.show()
+        pause_img.set_from_icon_name('media-playback-pause',
+                gtk.ICON_SIZE_LARGE_TOOLBAR)
+
+        # Play button
+        self.play_btn = ToggleToolButton('media-playback-start')
+        self.play_btn.show()
+        self.play_btn.connect('toggled', self.play_cb, [play_img, pause_img])
+        self.insert(self.play_btn, -1)
+        self.play_btn.set_tooltip(_('Play / Pause'))
+
+        self.voice_combo = ComboBox()
+        for voice in self.sorted_voices:
+            self.voice_combo.append_item(voice, voice[0])
+        self.voice_combo.set_active(default)
+        self.voice_combo.connect('changed', self.voice_changed_cb)
+        speech.voice = self.voice_combo.props.value
+        combotool = ToolComboBox(self.voice_combo)
+        self.insert(combotool, -1)
+        combotool.show()
+
+    def compare_voices(self,  a,  b):
+        if a[0].lower() == b[0].lower():
+            return 0
+        if a[0] .lower() < b[0].lower():
+            return -1
+        if a[0] .lower() > b[0].lower():
+            return 1
+
+    def voice_changed_cb(self, combo):
+        speech.voice = combo.props.value
+        if self.activity != None:
+            speech.say(speech.voice[0])
+
+    def play_cb(self, widget, images):
+        widget.set_icon_widget(images[int(widget.get_active())])
+
+        if widget.get_active():
+            if speech.is_stopped():
+                speech.play(self.activity._view.get_marked_words())
+        else:
+            speech.stop()
diff --git a/speech.py b/speech.py
new file mode 100644
index 0000000..a1c1e5f
--- /dev/null
+++ b/speech.py
@@ -0,0 +1,43 @@
+# Copyright (C) 2008, 2009 James D. Simmons
+# Copyright (C) 2009 Aleksey S. Lim
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+
+import logging
+
+_logger = logging.getLogger('read-etexts-activity')
+
+supported = True
+
+try:
+    import gst
+    gst.element_factory_make('espeak')
+    from speech_gst import *
+    _logger.info('use gst-plugins-espeak')
+except Exception, e:
+    _logger.info('disable gst-plugins-espeak: %s' % e)
+    try:
+        from speech_dispatcher import *
+        _logger.info('use speech-dispatcher')
+    except Exception, e:
+        supported = False
+        _logger.info('disable speech: %s' % e)
+
+voice = None
+pitch = 0
+rate = 0
+
+highlight_cb = None
+reset_cb = None
diff --git a/speech_dispatcher.py b/speech_dispatcher.py
new file mode 100644
index 0000000..4fad27f
--- /dev/null
+++ b/speech_dispatcher.py
@@ -0,0 +1,116 @@
+# Copyright (C) 2008 James D. Simmons
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+
+import gtk
+import time
+import threading
+import speechd
+import logging
+
+import speech
+
+_logger = logging.getLogger('read-etexts-activity')
+
+done = True
+
+
+def voices():
+    try:
+        client = speechd.SSIPClient('readetextstest')
+        voices = client.list_synthesis_voices()
+        client.close()
+        return voices
+    except Exception, e:
+        _logger.warning('speech dispatcher not started: %s' % e)
+        return []
+
+
+def say(words):
+    try:
+        client = speechd.SSIPClient('readetextstest')
+        client.set_rate(int(speech.rate))
+        client.set_pitch(int(speech.pitch))
+        client.set_language(speech.voice[1])
+        client.speak(words)
+        client.close()
+    except Exception, e:
+        _logger.warning('speech dispatcher not running: %s' % e)
+
+
+def is_stopped():
+    return done
+
+
+def stop():
+    global done
+    done = True
+
+
+def play(words):
+    global thread
+    thread = EspeakThread(words)
+    thread.start()
+
+
+class EspeakThread(threading.Thread):
+
+    def __init__(self, words):
+        threading.Thread.__init__(self)
+        self.words = words
+
+    def run(self):
+        "This is the code that is executed when the start() method is called"
+        self.client = None
+        try:
+            self.client = speechd.SSIPClient('readetexts')
+            self.client._conn.send_command('SET', speechd.Scope.SELF,
+                    'SSML_MODE', "ON")
+            if speech.voice:
+                self.client.set_language(speech.voice[1])
+                self.client.set_rate(speech.rate)
+                self.client.set_pitch(speech.pitch)
+            self.client.speak(self.words, self.next_word_cb,
+                    (speechd.CallbackType.INDEX_MARK,
+                    speechd.CallbackType.END))
+            global done
+            done = False
+            while not done:
+                time.sleep(0.1)
+            self.cancel()
+            self.client.close()
+        except Exception, e:
+            _logger.warning('speech-dispatcher client not created: %s' % e)
+
+    def cancel(self):
+        if self.client:
+            try:
+                self.client.cancel()
+            except Exception, e:
+                _logger.warning('speech dispatcher cancel failed: %s' % e)
+
+    def next_word_cb(self, type, **kargs):
+        if type == speechd.CallbackType.INDEX_MARK:
+            mark = kargs['index_mark']
+            word_count = int(mark)
+            gtk.gdk.threads_enter()
+            speech.highlight_cb(word_count)
+            gtk.gdk.threads_leave()
+        elif type == speechd.CallbackType.END:
+            gtk.gdk.threads_enter()
+            speech.reset_cb()
+            gtk.gdk.threads_leave()
+            global done
+            done = True
diff --git a/speech_gst.py b/speech_gst.py
new file mode 100644
index 0000000..4627c75
--- /dev/null
+++ b/speech_gst.py
@@ -0,0 +1,87 @@
+# Copyright (C) 2009 Aleksey S. Lim
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+
+import gst
+import logging
+
+import speech
+
+_logger = logging.getLogger('read-etexts-activity')
+
+
+def _message_cb(bus, message, pipe):
+    if message.type in (gst.MESSAGE_EOS, gst.MESSAGE_ERROR):
+        pipe.set_state(gst.STATE_NULL)
+        if pipe is play_speaker[1]:
+            speech.reset_cb()
+    elif message.type == gst.MESSAGE_ELEMENT and \
+            message.structure.get_name() == 'espeak-mark':
+        mark = message.structure['mark']
+        speech.highlight_cb(int(mark))
+
+
+def _create_pipe():
+    pipe = gst.Pipeline('pipeline')
+
+    source = gst.element_factory_make('espeak', 'source')
+    pipe.add(source)
+
+    sink = gst.element_factory_make('autoaudiosink', 'sink')
+    pipe.add(sink)
+    source.link(sink)
+
+    bus = pipe.get_bus()
+    bus.add_signal_watch()
+    bus.connect('message', _message_cb, pipe)
+
+    return (source, pipe)
+
+
+def _speech(speaker, words):
+    speaker[0].props.pitch = speech.pitch
+    speaker[0].props.rate = speech.rate
+    speaker[0].props.voice = speech.voice[1]
+    speaker[0].props.text = words
+    speaker[1].set_state(gst.STATE_NULL)
+    speaker[1].set_state(gst.STATE_PLAYING)
+
+
+info_speaker = _create_pipe()
+play_speaker = _create_pipe()
+play_speaker[0].props.track = 2
+
+
+def voices():
+    return info_speaker[0].props.voices
+
+
+def say(words):
+    _speech(info_speaker, words)
+
+
+def play(words):
+    _speech(play_speaker, words)
+
+
+def is_stopped():
+    for i in play_speaker[1].get_state():
+        if isinstance(i, gst.State) and i == gst.STATE_NULL:
+            return True
+    return False
+
+
+def stop():
+    play_speaker[1].set_state(gst.STATE_NULL)
diff --git a/textadapter.py b/textadapter.py
index 2e96447..a6cef2b 100644
--- a/textadapter.py
+++ b/textadapter.py
@@ -9,6 +9,8 @@ import threading
 from sugar import mime
 from sugar.graphics import style
 
+import speech
+
 PAGE_SIZE = 38
 
 
@@ -59,6 +61,14 @@ class TextViewer(gobject.GObject):
         self.highlight_tag.set_property('foreground', 'black')
         self.highlight_tag.set_property('background', 'yellow')
 
+        # text to speech initialization
+        self.current_word = 0
+        self.word_tuples = []
+        self.spoken_word_tag = self.textview.get_buffer().create_tag()
+        self.spoken_word_tag.set_property('weight', pango.WEIGHT_BOLD)
+        self.normal_tag = self.textview.get_buffer().create_tag()
+        self.normal_tag.set_property('weight',  pango.WEIGHT_NORMAL)
+
     def load_document(self, file_path):
 
         file_name = file_path.replace('file://', '')
@@ -94,6 +104,8 @@ class TextViewer(gobject.GObject):
         self._pagecount = pagecount + 1
         self.set_current_page(0)
 
+        speech.highlight_cb = self.highlight_next_word
+
     def _show_page(self, page_number):
         position = self.page_index[page_number]
         self._etext_file.seek(position)
@@ -110,6 +122,7 @@ class TextViewer(gobject.GObject):
         textbuffer = self.textview.get_buffer()
         label_text = label_text + '\n\n\n'
         textbuffer.set_text(label_text)
+        self._prepare_text_to_speech(label_text)
 
     def can_highlight(self):
         return True
@@ -138,6 +151,62 @@ class TextViewer(gobject.GObject):
     def connect_page_changed_handler(self, handler):
         self.connect('page-changed', handler)
 
+    def can_do_text_to_speech(self):
+        return True
+
+    def get_marked_words(self):
+        "Adds a mark between each word of text."
+        i = self.current_word
+        marked_up_text = '<speak> '
+        while i < len(self.word_tuples):
+            word_tuple = self.word_tuples[i]
+            marked_up_text = marked_up_text + '<mark name="' + str(i) + '"/>' \
+                    + word_tuple[2]
+            i = i + 1
+        print marked_up_text
+        return marked_up_text + '</speak>'
+
+    def _prepare_text_to_speech(self, page_text):
+        i = 0
+        j = 0
+        word_begin = 0
+        word_end = 0
+        ignore_chars = [' ',  '\n',  u'\r',  '_',  '[', '{', ']', '}', '|',
+                '<',  '>',  '*',  '+',  '/',  '\\']
+        ignore_set = set(ignore_chars)
+        self.word_tuples = []
+        len_page_text = len(page_text)
+        while i < len_page_text:
+            if page_text[i] not in ignore_set:
+                word_begin = i
+                j = i
+                while  j < len_page_text and page_text[j] not in ignore_set:
+                    j = j + 1
+                    word_end = j
+                    i = j
+                word_tuple = (word_begin, word_end,
+                        page_text[word_begin: word_end])
+                if word_tuple[2] != u'\r':
+                    self.word_tuples.append(word_tuple)
+            i = i + 1
+
+    def highlight_next_word(self,  word_count):
+        if word_count < len(self.word_tuples):
+            word_tuple = self.word_tuples[word_count]
+            textbuffer = self.textview.get_buffer()
+            iterStart = textbuffer.get_iter_at_offset(word_tuple[0])
+            iterEnd = textbuffer.get_iter_at_offset(word_tuple[1])
+            bounds = textbuffer.get_bounds()
+            textbuffer.apply_tag(self.normal_tag,  bounds[0], iterStart)
+            textbuffer.apply_tag(self.spoken_word_tag, iterStart, iterEnd)
+            v_adjustment = self._scrolled.get_vadjustment()
+            max = v_adjustment.upper - v_adjustment.page_size
+            max = max * word_count
+            max = max / len(self.word_tuples)
+            v_adjustment.value = max
+            self.current_word = word_count
+        return True
+
     def load_metadata(self, activity):
         pass