6 Commits

Author SHA1 Message Date
Storm Dragon 523b896053 Couple minor bug fixes, new speech code added in prefs. 2026-02-18 06:10:57 -05:00
Storm Dragon 9152455227 Speech in settings now uses the voice when switched to it. 2026-02-18 06:07:55 -05:00
Storm Dragon c0fdaca4d0 Merge remote-tracking branch 'origin/web-and-punctuation-fixes' into testing 2026-02-17 14:50:01 -05:00
destructatron 8312a842c1 Fix duplicate role name when aria-roledescription is present
When an element has aria-roledescription (e.g. Discord messages use
'Message'), _generateRoleName was adding both the roledescription AND
the standard localized role name (e.g. 'article'). This resulted in
the role being announced twice.

The aria-roledescription attribute is intended to replace the standard
role name, not supplement it. Return early after adding the
roledescription to match Orca's behavior.
2026-02-17 19:21:36 +00:00
destructatron 07138197cb Fix web content editable label, line boundary, and punctuation bugs
Bug 1: Content editable entry label not announced on caret-nav focus
- When navigating into a content editable via caret navigation (e.g. down
  arrow from message list to message entry in Discord), the label and role
  were suppressed because the code treated entering the field the same as
  navigating within it.
- Fixed locus_of_focus_changed in web/script.py to detect when focus enters
  a content editable from outside and generate full object speech.
- Fixed _generateLabelOrName in web/speech_generator.py to preserve the
  label when the prior object was outside the content editable.

Bug 2: Adjacent button included in line contents on first line
- When reading line contents inside a content editable, the layout-mode
  expansion could walk outside the content editable boundary and include
  adjacent UI elements (e.g. a 'More options' button) that happened to
  share the same visual line.
- Fixed _getLineContentsAtOffset in web/script_utilities.py to stop
  expanding at content editable boundaries in both directions.

Bug 3: Punctuation stripped from live regions and AT-SPI announcements
- presentMessage uses resetStyles=True by default, which sets punctuation
  to NONE for system voice messages. This is correct for generated text
  but wrong for user content in live regions and AT-SPI announcements.
- Fixed liveregions.py pumpMessages and default.py onAnnouncement to pass
  resetStyles=False, preserving the user's punctuation settings.
2026-02-17 19:07:41 +00:00
Storm Dragon ed78ffc248 Visual speech monitor added to speech history plugin. Toggle with cthulhu+shift+D 2026-02-17 08:40:51 -05:00
13 changed files with 350 additions and 106 deletions
+1 -1
View File
@@ -1,7 +1,7 @@
# Maintainer: Storm Dragon <storm_dragon@stormux.org>
pkgname=cthulhu
pkgver=2026.02.17
pkgver=2026.02.18
pkgrel=1
pkgdesc="Desktop-agnostic screen reader with plugin system, forked from Orca"
url="https://git.stormux.org/storm/cthulhu"
+1 -1
View File
@@ -1,5 +1,5 @@
project('cthulhu',
version: '2026.02.17-master',
version: '2026.02.18-master',
meson_version: '>= 1.0.0',
)
+1 -1
View File
@@ -23,5 +23,5 @@
# Forked from Orca screen reader.
# Cthulhu project: https://git.stormux.org/storm/cthulhu
version = "2026.02.17"
version = "2026.02.18"
codeName = "master"
+47 -5
View File
@@ -183,6 +183,7 @@ class CthulhuSetupGUI(cthulhu_gtkbuilder.GtkBuilderWrapper):
self.soundThemeCombo = None
self.roleSoundPresentationCombo = None
self._isInitialSetup = False
self._updatingSpeechFamilies = False
self.selectedFamilyChoices = {}
self.selectedLanguageChoices = {}
self.profilesCombo = None
@@ -1271,7 +1272,7 @@ class CthulhuSetupGUI(cthulhu_gtkbuilder.GtkBuilderWrapper):
for familyChoice in self.speechFamiliesChoices:
name = familyChoice[speechserver.VoiceFamily.NAME]
if name == familyName:
self.get_widget("speechFamilies").set_active(i)
self._setSpeechFamiliesActive(i)
self.speechFamiliesChoice = self.speechFamiliesChoices[i]
familySet = True
break
@@ -1295,13 +1296,22 @@ class CthulhuSetupGUI(cthulhu_gtkbuilder.GtkBuilderWrapper):
if not familySet:
tokens = ["PREFERENCES DIALOG: Could not find speech family match for", familyName]
debug.printTokens(debug.LEVEL_INFO, tokens, True)
self.get_widget("speechFamilies").set_active(0)
self._setSpeechFamiliesActive(0)
self.speechFamiliesChoice = self.speechFamiliesChoices[0]
if familySet:
self.selectedFamilyChoices[self.speechServersChoice,
self.speechLanguagesChoice] = i
def _setSpeechFamiliesActive(self, index):
"""Sets speechFamilies active index while suppressing preview chatter."""
self._updatingSpeechFamilies = True
try:
self.get_widget("speechFamilies").set_active(index)
finally:
self._updatingSpeechFamilies = False
def _setupFamilies(self):
"""Gets the list of voice variants for the current speech server and
current language.
@@ -1353,7 +1363,7 @@ class CthulhuSetupGUI(cthulhu_gtkbuilder.GtkBuilderWrapper):
selectedIndex = self.selectedFamilyChoices[self.speechServersChoice,
self.speechLanguagesChoice]
self.get_widget("speechFamilies").set_active(selectedIndex)
self._setSpeechFamiliesActive(selectedIndex)
def _setSpeechLanguagesChoice(self, languageName):
"""Sets the active item in the languages ("Language:") combo box
@@ -2899,21 +2909,51 @@ class CthulhuSetupGUI(cthulhu_gtkbuilder.GtkBuilderWrapper):
self.get_widget("enableDiacriticalKeysCheckButton").set_sensitive( \
enable)
def _presentMessage(self, text, interrupt=False):
def _presentMessage(self, text, interrupt=False, voice=None):
"""If the text field is not None, presents the given text, optionally
interrupting anything currently being spoken.
Arguments:
- text: the text to present
- interrupt: if True, interrupt any speech currently being spoken
- voice: the voice to use; if None, use the default system voice
"""
self.script.speakMessage(text, interrupt=interrupt)
self.script.speakMessage(text, voice=voice, interrupt=interrupt)
try:
self.script.displayBrailleMessage(text, flashTime=-1)
except Exception:
pass
def _previewVoiceSelection(self, voiceType, name):
"""Present a short preview in the currently selected voice."""
if not name:
return
voice = self._getACSSForVoiceType(voiceType)
if not voice:
return
previewVoice = acss.ACSS(voice)
if self.speechSystemsChoice \
and self.speechSystemsChoice.SpeechServer.getFactoryName() == guilabels.SPEECH_DISPATCHER:
previewMessage = messages.SPEECH_VOICE_VALUE % name
else:
previewMessage = messages.SPEECH_VOICE_VALUE_GENERIC % name
# Speak through the currently-selected preferences speech server so
# preview uses the chosen module/voice, not the globally-active one.
server = self.speechServersChoice
if server and hasattr(server, "speak"):
try:
server.speak(previewMessage, previewVoice, interrupt=True)
return
except Exception:
debug.printException(debug.LEVEL_WARNING)
self._presentMessage(previewMessage, interrupt=True, voice=previewVoice)
def _createNode(self, appName):
"""Create a new root node in the TreeStore model with the name of the
application.
@@ -3316,6 +3356,8 @@ class CthulhuSetupGUI(cthulhu_gtkbuilder.GtkBuilderWrapper):
variant = family[speechserver.VoiceFamily.VARIANT]
voiceType = self.get_widget("voiceTypesCombo").get_active()
self._setFamilyNameForVoiceType(voiceType, name, language, dialect, variant)
if not self._updatingSpeechFamilies:
self._previewVoiceSelection(voiceType, name)
except Exception:
debug.printException(debug.LEVEL_SEVERE)
+3 -1
View File
@@ -276,7 +276,9 @@ class LiveRegionManager:
utts = message['labels'] + message['content']
if self.monitoring:
self._script.presentMessage(utts)
# Live region content is user-generated text, not system messages.
# Use resetStyles=False to preserve the user's punctuation settings.
self._script.presentMessage(utts, resetStyles=False)
else:
msg = "INFO: Not presenting message because monitoring is off"
debug.printMessage(debug.LEVEL_INFO, msg, True)
@@ -1,9 +1,8 @@
name = Speech History
version = 1.0.0
description = Shows a searchable history of the last 50 unique utterances spoken by Cthulhu
version = 1.2.0
description = Shows speech history and a live floating monitor of spoken text from Cthulhu
authors = Stormux
website = https://git.stormux.org/storm/cthulhu
copyright = Copyright 2025 Stormux
builtin = false
hidden = false
+213 -85
View File
@@ -7,39 +7,47 @@
# License as published by the Free Software Foundation; either
# version 2.1 of the License, or (at your option) any later version.
"""Speech History plugin for Cthulhu."""
"""Speech history and speech monitor plugin for Cthulhu."""
import logging
import gi
gi.require_version("Gtk", "3.0")
from gi.repository import Gtk, Gdk
from gi.repository import Gtk, Gdk, GLib
from cthulhu.plugin import Plugin, cthulhu_hookimpl
from cthulhu import debug
from cthulhu import speech
from cthulhu import speech_history
logger = logging.getLogger(__name__)
class SpeechHistory(Plugin):
"""Plugin that displays a window containing recent spoken output."""
"""Plugin that provides speech history and a live speech monitor."""
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self._activated = False
self._kbOpenWindow = None
self._kbToggleMonitor = None
# Speech history window state
self._window = None
self._filterEntry = None
self._filterText = ""
self._listStore = None
self._filterModel = None
self._treeView = None
self._capturePaused = False
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Plugin initialized", True)
# Live monitor window state
self._monitorWindow = None
self._monitorTextView = None
self._monitorTextBuffer = None
self._monitorEndMark = None
self._monitorLineCount = 0
self._monitorMaxLines = 500
@cthulhu_hookimpl
def activate(self, plugin=None):
@@ -47,17 +55,13 @@ class SpeechHistory(Plugin):
return
if self._activated:
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Already activated, skipping", True)
return True
try:
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Activating plugin", True)
self._register_keybinding()
self._register_keybindings()
self._activated = True
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Activated successfully", True)
return True
except Exception as e:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: ERROR during activate: {e}", True)
except Exception:
logger.exception("Error activating SpeechHistory plugin")
return False
@@ -67,46 +71,208 @@ class SpeechHistory(Plugin):
return
try:
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Deactivating plugin", True)
self._close_window()
self._disable_monitor()
self._kbOpenWindow = None
self._kbToggleMonitor = None
self._activated = False
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Deactivated successfully", True)
return True
except Exception as e:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: ERROR during deactivate: {e}", True)
except Exception:
logger.exception("Error deactivating SpeechHistory plugin")
return False
def _register_keybinding(self):
try:
if not self.app:
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: No app reference; cannot register keybinding", True)
return
def _register_keybindings(self):
if not self.app:
debug.printMessage(
debug.LEVEL_INFO,
"SpeechHistory: No app reference; cannot register keybindings",
True,
)
return
gestureString = "kb:cthulhu+control+h"
description = "Open speech history"
historyGesture = "kb:cthulhu+control+h"
historyDescription = "Open speech history"
self._kbOpenWindow = self.registerGestureByString(
self._open_window,
historyDescription,
historyGesture,
)
self._kbOpenWindow = self.registerGestureByString(
self._open_window,
description,
gestureString,
monitorGesture = "kb:cthulhu+shift+d"
monitorDescription = "Toggle speech monitor"
self._kbToggleMonitor = self.registerGestureByString(
self._toggle_monitor,
monitorDescription,
monitorGesture,
)
if not self._kbOpenWindow:
debug.printMessage(
debug.LEVEL_INFO,
f"SpeechHistory: Failed to register keybinding {historyGesture}",
True,
)
if self._kbOpenWindow:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: Registered keybinding {gestureString}", True)
else:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: Failed to register keybinding {gestureString}", True)
except Exception as e:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: ERROR registering keybinding: {e}", True)
logger.exception("Error registering keybinding for SpeechHistory")
if not self._kbToggleMonitor:
debug.printMessage(
debug.LEVEL_INFO,
f"SpeechHistory: Failed to register keybinding {monitorGesture}",
True,
)
# Live monitor methods
def _toggle_monitor(self, script=None, inputEvent=None):
try:
if self._monitorWindow is not None:
self._disable_monitor()
self._present_message("Speech monitor disabled.")
return True
self._enable_monitor()
self._present_message("Speech monitor enabled.")
return True
except Exception:
logger.exception("Error toggling speech monitor")
self._disable_monitor()
self._present_message("Error toggling speech monitor.")
return False
def _enable_monitor(self):
if self._monitorWindow is not None:
self._monitorWindow.show_all()
return
self._create_monitor_window()
speech.set_monitor_callbacks(writeText=self._on_spoken_text)
self._monitorWindow.show_all()
def _disable_monitor(self):
speech.set_monitor_callbacks(writeText=None)
if self._monitorWindow is not None:
self._monitorWindow.destroy()
def _create_monitor_window(self):
self._monitorWindow = Gtk.Window(title="Speech Monitor - Cthulhu")
self._monitorWindow.set_default_size(900, 320)
self._monitorWindow.set_modal(False)
self._monitorWindow.set_border_width(8)
self._monitorWindow.set_keep_above(True)
self._monitorWindow.set_accept_focus(False)
self._monitorWindow.set_focus_on_map(False)
self._monitorWindow.set_skip_taskbar_hint(True)
self._monitorWindow.set_skip_pager_hint(True)
self._monitorWindow.set_type_hint(Gdk.WindowTypeHint.UTILITY)
self._monitorWindow.connect("destroy", self._on_monitor_window_destroy)
self._monitorWindow.connect("key-press-event", self._on_monitor_window_key_press)
mainBox = Gtk.Box(orientation=Gtk.Orientation.VERTICAL, spacing=8)
scrolledWindow = Gtk.ScrolledWindow()
scrolledWindow.set_policy(Gtk.PolicyType.AUTOMATIC, Gtk.PolicyType.AUTOMATIC)
self._monitorTextView = Gtk.TextView()
self._monitorTextView.set_editable(False)
self._monitorTextView.set_cursor_visible(False)
self._monitorTextView.set_can_focus(False)
self._monitorTextView.set_wrap_mode(Gtk.WrapMode.WORD_CHAR)
self._monitorTextView.set_accepts_tab(False)
self._monitorTextView.set_left_margin(6)
self._monitorTextView.set_right_margin(6)
self._monitorTextView.set_top_margin(6)
self._monitorTextView.set_bottom_margin(6)
textAccessible = self._monitorTextView.get_accessible()
if textAccessible:
textAccessible.set_name("Speech monitor output")
self._monitorTextBuffer = self._monitorTextView.get_buffer()
self._monitorEndMark = self._monitorTextBuffer.create_mark(
"monitorEnd",
self._monitorTextBuffer.get_end_iter(),
False,
)
self._monitorLineCount = 0
scrolledWindow.add(self._monitorTextView)
mainBox.pack_start(scrolledWindow, True, True, 0)
buttonRow = Gtk.Box(orientation=Gtk.Orientation.HORIZONTAL, spacing=8)
buttonRow.set_halign(Gtk.Align.END)
clearButton = Gtk.Button(label="Clear")
closeButton = Gtk.Button(label="Close")
clearButton.connect("clicked", self._on_monitor_clear_clicked)
closeButton.connect("clicked", self._on_monitor_close_clicked)
buttonRow.pack_start(clearButton, False, False, 0)
buttonRow.pack_start(closeButton, False, False, 0)
mainBox.pack_start(buttonRow, False, False, 0)
self._monitorWindow.add(mainBox)
def _on_monitor_close_clicked(self, button):
self._disable_monitor()
def _on_monitor_clear_clicked(self, button):
if self._monitorTextBuffer is None:
return
self._monitorTextBuffer.set_text("")
self._monitorLineCount = 0
def _on_monitor_window_destroy(self, widget):
speech.set_monitor_callbacks(writeText=None)
self._monitorWindow = None
self._monitorTextView = None
self._monitorTextBuffer = None
self._monitorEndMark = None
self._monitorLineCount = 0
def _on_monitor_window_key_press(self, widget, event):
if event.keyval == Gdk.KEY_Escape:
self._disable_monitor()
return True
return False
def _on_spoken_text(self, text):
if not text or not isinstance(text, str):
return
GLib.idle_add(self._append_text_on_main_thread, text)
def _append_text_on_main_thread(self, text):
if self._monitorTextBuffer is None:
return False
self._monitorTextBuffer.insert(self._monitorTextBuffer.get_end_iter(), f"{text}\n")
self._monitorLineCount += 1
if self._monitorLineCount > self._monitorMaxLines:
excess = self._monitorLineCount - self._monitorMaxLines
startIter = self._monitorTextBuffer.get_start_iter()
cutIter = self._monitorTextBuffer.get_iter_at_line(excess)
self._monitorTextBuffer.delete(startIter, cutIter)
self._monitorLineCount = self._monitorMaxLines
if self._monitorEndMark is not None and self._monitorTextView is not None:
self._monitorTextBuffer.move_mark(
self._monitorEndMark,
self._monitorTextBuffer.get_end_iter(),
)
self._monitorTextView.scroll_mark_onscreen(self._monitorEndMark)
return False
# Speech history methods
def _open_window(self, script=None, inputEvent=None):
try:
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Open window requested", True)
if self._window:
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Window already open; presenting", True)
self._window.present()
return True
@@ -120,10 +286,8 @@ class SpeechHistory(Plugin):
elif self._filterEntry:
self._filterEntry.grab_focus()
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Window shown", True)
return True
except Exception as e:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: ERROR opening window: {e}", True)
except Exception:
logger.exception("Error opening SpeechHistory window")
self._resume_capture()
return False
@@ -141,7 +305,6 @@ class SpeechHistory(Plugin):
self._filterText = ""
# Filter row
filterRow = Gtk.Box(orientation=Gtk.Orientation.HORIZONTAL, spacing=10)
filterLabel = Gtk.Label(label="_Filter:")
filterLabel.set_use_underline(True)
@@ -156,7 +319,6 @@ class SpeechHistory(Plugin):
filterRow.pack_start(self._filterEntry, True, True, 0)
mainBox.pack_start(filterRow, False, False, 0)
# List
self._listStore = Gtk.ListStore(str)
self._filterModel = self._listStore.filter_new()
self._filterModel.set_visible_func(self._filter_visible_func)
@@ -169,7 +331,7 @@ class SpeechHistory(Plugin):
textRenderer = Gtk.CellRendererText()
textRenderer.set_property("wrap-width", 640)
textRenderer.set_property("wrap-mode", 2) # Pango.WrapMode.WORD_CHAR
textRenderer.set_property("wrap-mode", 2)
textColumn = Gtk.TreeViewColumn("Spoken Text", textRenderer, text=0)
textColumn.set_resizable(True)
textColumn.set_expand(True)
@@ -180,7 +342,6 @@ class SpeechHistory(Plugin):
scrolled.add(self._treeView)
mainBox.pack_start(scrolled, True, True, 0)
# Buttons
buttonRow = Gtk.Box(orientation=Gtk.Orientation.HORIZONTAL, spacing=10)
buttonRow.set_halign(Gtk.Align.END)
@@ -201,16 +362,12 @@ class SpeechHistory(Plugin):
self._refresh_list(selectFirst=True)
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Window created", True)
def _on_filter_changed(self, entry):
try:
self._filterText = entry.get_text() or ""
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: Filter changed '{self._filterText}'", True)
if self._filterModel:
self._filterModel.refilter()
except Exception as e:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: ERROR filtering: {e}", True)
except Exception:
logger.exception("Error updating speech history filter")
def _filter_visible_func(self, model, treeIter, data=None):
@@ -221,8 +378,7 @@ class SpeechHistory(Plugin):
spokenText = model[treeIter][0] or ""
return spokenText.lower().startswith(filterText)
except Exception as e:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: ERROR in filter func: {e}", True)
except Exception:
return True
def _refresh_list(self, selectFirst=False):
@@ -230,31 +386,19 @@ class SpeechHistory(Plugin):
if not self._listStore:
return
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Refreshing list", True)
self._listStore.clear()
items = speech_history.get_items()
debug.printMessage(
debug.LEVEL_INFO,
f"SpeechHistory: Retrieved {len(items)} items (paused={speech_history.is_capture_paused()})",
True,
)
for item in items:
self._listStore.append([item])
if self._filterModel:
self._filterModel.refilter()
debug.printMessage(
debug.LEVEL_INFO,
f"SpeechHistory: Filtered items visible={len(self._filterModel)} filter='{self._filterText}'",
True,
)
if selectFirst and self._treeView and len(self._filterModel) > 0:
selection = self._treeView.get_selection()
selection.select_path(Gtk.TreePath.new_first())
except Exception as e:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: ERROR refreshing list: {e}", True)
except Exception:
logger.exception("Error refreshing speech history list")
def _get_selected_text(self):
@@ -268,8 +412,7 @@ class SpeechHistory(Plugin):
return None
return model[treeIter][0]
except Exception as e:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: ERROR getting selection: {e}", True)
except Exception:
logger.exception("Error getting selected speech history item")
return None
@@ -284,11 +427,8 @@ class SpeechHistory(Plugin):
clipboard.set_text(selectedText, -1)
clipboard.store()
preview = selectedText[:60] + ("..." if len(selectedText) > 60 else "")
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: Copied to clipboard '{preview}'", True)
self._present_message("Copied to clipboard.")
except Exception as e:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: ERROR copying: {e}", True)
except Exception:
logger.exception("Error copying speech history item to clipboard")
self._present_message("Error copying to clipboard.")
@@ -300,33 +440,27 @@ class SpeechHistory(Plugin):
return
removed = speech_history.remove(selectedText)
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: Remove requested removed={removed}", True)
if removed:
self._refresh_list(selectFirst=True)
self._present_message("Removed from history.")
else:
self._present_message("Item not found in history.")
except Exception as e:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: ERROR removing: {e}", True)
except Exception:
logger.exception("Error removing speech history item")
self._present_message("Error removing item from history.")
def _on_close_clicked(self, button):
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Close button clicked", True)
self._close_window()
def _close_window(self):
try:
if self._window:
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Closing window", True)
self._window.destroy()
except Exception as e:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: ERROR closing window: {e}", True)
except Exception:
logger.exception("Error closing SpeechHistory window")
self._resume_capture()
def _on_window_destroy(self, widget):
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Window destroyed", True)
self._window = None
self._filterEntry = None
self._listStore = None
@@ -338,14 +472,12 @@ class SpeechHistory(Plugin):
def _on_window_key_press(self, widget, event):
try:
if event.keyval == Gdk.KEY_Escape:
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Escape pressed; closing window", True)
self._close_window()
return True
if not self._filterEntry or self._filterEntry.is_focus():
return False
# If user starts typing anywhere, move focus to filter and update it.
if event.keyval == Gdk.KEY_BackSpace:
currentText = self._filterEntry.get_text() or ""
if currentText:
@@ -376,8 +508,7 @@ class SpeechHistory(Plugin):
self._filterEntry.set_text(currentText + charTyped)
self._filterEntry.set_position(-1)
return True
except Exception as e:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: ERROR in key handler: {e}", True)
except Exception:
logger.exception("Error handling key press in SpeechHistory window")
return False
@@ -385,7 +516,6 @@ class SpeechHistory(Plugin):
if self._capturePaused:
return
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Pausing capture while window is open", True)
speech_history.pause_capture(reason="SpeechHistory window open")
self._capturePaused = True
@@ -393,7 +523,6 @@ class SpeechHistory(Plugin):
if not self._capturePaused:
return
debug.printMessage(debug.LEVEL_INFO, "SpeechHistory: Resuming capture (window closed)", True)
speech_history.resume_capture(reason="SpeechHistory window closed")
self._capturePaused = False
@@ -408,6 +537,5 @@ class SpeechHistory(Plugin):
state.activeScript.presentMessage(message, resetStyles=False)
else:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: {message}", True)
except Exception as e:
debug.printMessage(debug.LEVEL_INFO, f"SpeechHistory: ERROR presenting message: {e}", True)
except Exception:
logger.exception("Error presenting message from SpeechHistory")
+3 -1
View File
@@ -1473,7 +1473,9 @@ class Script(script.Script):
"""Callback for object:announcement events."""
if isinstance(event.any_data, str):
self.presentMessage(event.any_data)
# AT-SPI announcements contain application content, not system messages.
# Use resetStyles=False to preserve the user's punctuation settings.
self.presentMessage(event.any_data, resetStyles=False)
def onNameChanged(self, event):
"""Callback for object:property-change:accessible-name events."""
+15 -3
View File
@@ -1648,9 +1648,21 @@ class Script(default.Script):
elif self.utilities.isContentEditableWithEmbeddedObjects(newFocus) \
and (self._lastCommandWasCaretNav or self._lastCommandWasStructNav) \
and not (AXUtilities.is_table_cell(newFocus) and AXObject.get_name(newFocus)):
tokens = ["WEB: New focus", newFocus, "content editable. Generating line."]
debug.printTokens(debug.LEVEL_INFO, tokens, True)
contents = self.utilities.getLineContentsAtOffset(newFocus, caretOffset)
# Check if we're entering the content editable from outside (e.g. down arrow
# from a message list into a message entry). In that case, generate full object
# speech (with label and role) rather than just line contents.
enteredFromOutside = oldFocus is not None \
and oldFocus != newFocus \
and not AXObject.find_ancestor(oldFocus, lambda x: x == newFocus)
if enteredFromOutside:
tokens = ["WEB: New focus", newFocus,
"content editable entered from outside. Generating speech."]
debug.printTokens(debug.LEVEL_INFO, tokens, True)
args['priorObj'] = oldFocus
else:
tokens = ["WEB: New focus", newFocus, "content editable. Generating line."]
debug.printTokens(debug.LEVEL_INFO, tokens, True)
contents = self.utilities.getLineContentsAtOffset(newFocus, caretOffset)
elif self.utilities.isAnchor(newFocus):
tokens = ["WEB: New focus", newFocus, "is anchor. Generating line."]
debug.printTokens(debug.LEVEL_INFO, tokens, True)
@@ -1834,6 +1834,16 @@ class Utilities(script_utilities.Utilities):
prevObj, pOffset = self.findPreviousCaretInOrder(firstObj, firstStart)
nextObj, nOffset = self.findNextCaretInOrder(lastObj, lastEnd - 1)
# If we're inside a content editable, don't expand line contents beyond
# its boundaries (e.g. don't include a "More options" button adjacent to
# a message entry just because it's on the same visual line).
contentEditableBoundary = None
if self.isContentEditableWithEmbeddedObjects(obj):
contentEditableBoundary = obj
else:
contentEditableBoundary = AXObject.find_ancestor(
obj, self.isContentEditableWithEmbeddedObjects)
# Check for things on the same line to the left of this object.
prevStartTime = time.time()
while prevObj and self.getDocumentForObject(prevObj) == document:
@@ -1848,6 +1858,10 @@ class Utilities(script_utilities.Utilities):
if objRow != AXObject.find_ancestor(prevObj, AXUtilities.is_table_row):
break
if contentEditableBoundary and prevObj != contentEditableBoundary \
and not AXObject.find_ancestor(prevObj, lambda x: x == contentEditableBoundary):
break
onLeft = self._getContentsForObj(prevObj, pOffset, boundary)
onLeft = list(filter(_include, onLeft))
if not onLeft:
@@ -1878,6 +1892,10 @@ class Utilities(script_utilities.Utilities):
if objRow != AXObject.find_ancestor(nextObj, AXUtilities.is_table_row):
break
if contentEditableBoundary and nextObj != contentEditableBoundary \
and not AXObject.find_ancestor(nextObj, lambda x: x == contentEditableBoundary):
break
onRight = self._getContentsForObj(nextObj, nOffset, boundary)
if onRight and self._contentIsSubsetOf(objects[0], onRight[-1]):
onRight = onRight[0:-1]
+10 -1
View File
@@ -361,7 +361,13 @@ class SpeechGenerator(speech_generator.SpeechGenerator):
if self._script.utilities.isContentEditableWithEmbeddedObjects(obj) \
or self._script.utilities.isDocument(obj):
if input_event_manager.get_manager().last_event_was_caret_navigation():
return []
# Still generate the label if we just entered this object from outside
# (e.g. down arrow from message list into message entry in Discord).
enteredFromOutside = priorObj is not None \
and priorObj != obj \
and not AXObject.find_ancestor(priorObj, lambda x: x == obj)
if not enteredFromOutside:
return []
if AXUtilities.is_page_tab(priorObj) and AXObject.get_name(priorObj) == objName:
return []
@@ -554,6 +560,9 @@ class SpeechGenerator(speech_generator.SpeechGenerator):
if roledescription:
result = [roledescription]
result.extend(self.voice(speech_generator.SYSTEM, obj=obj, **args))
# aria-roledescription replaces the standard role name, so return
# early to avoid announcing both (e.g. "Message" + "article").
return result
role = args.get('role', AXObject.get_role(obj))
roleSoundPresentation = cthulhu.cthulhuApp.settingsManager.getSetting('roleSoundPresentation')
+25 -1
View File
@@ -36,7 +36,7 @@ __license__ = "LGPL"
import importlib
import time
from typing import TYPE_CHECKING, Optional, List, Dict, Any, Union
from typing import TYPE_CHECKING, Optional, List, Dict, Any, Union, Callable
from . import debug
from . import logger
@@ -73,6 +73,9 @@ _speechserver: Optional[SpeechServer] = None
# The last time something was spoken.
_timestamp: float = 0.0
# Optional callback for live monitoring of spoken text.
_monitorWriteTextCallback: Optional[Callable[[str], None]] = None
def _initSpeechServer(moduleName: Optional[str], speechServerInfo: Optional[Any]) -> None:
global _speechserver
@@ -168,6 +171,21 @@ def setSpeechServer(speechServer: SpeechServer) -> None:
global _speechserver
_speechserver = speechServer
def set_monitor_callbacks(writeText: Optional[Callable[[str], None]] = None) -> None:
"""Sets runtime callbacks for live speech monitoring."""
global _monitorWriteTextCallback
_monitorWriteTextCallback = writeText
def _write_to_monitor(text: str) -> None:
"""Writes text to the active speech monitor callback if set."""
if _monitorWriteTextCallback is None:
return
try:
_monitorWriteTextCallback(text)
except Exception:
debug.printException(debug.LEVEL_INFO)
def __resolveACSS(acss: Optional[Any] = None) -> ACSS:
if isinstance(acss, ACSS):
family = acss.get(acss.FAMILY)
@@ -237,6 +255,8 @@ def _speak(text: str, acss: Optional[Any], interrupt: bool) -> None:
except Exception:
debug.printException(debug.LEVEL_INFO)
_write_to_monitor(text)
if not _speechserver:
logLine = f"SPEECH OUTPUT: '{text}' {acss}"
debug.printMessage(debug.LEVEL_INFO, logLine, True)
@@ -393,6 +413,8 @@ def speakKeyEvent(event: Any, acss: Optional[Any] = None) -> None:
if log:
log.info(logLine)
_write_to_monitor(msg.strip())
if _speechserver:
_speechserver.speakKeyEvent(event, acss) # type: ignore
@@ -425,6 +447,8 @@ def speakCharacter(character: str, acss: Optional[Any] = None) -> None:
if log:
log.info(f"SPEECH OUTPUT: '{character}'")
_write_to_monitor(character)
if _speechserver:
_speechserver.speakCharacter(character, acss=acss) # type: ignore
+11 -3
View File
@@ -368,7 +368,11 @@ class SpeechAndVerbosityManager:
if family is None:
default_voice.pop(acss.ACSS.FAMILY, None)
else:
default_voice[acss.ACSS.FAMILY] = dict(family)
family_dict = {
key: value for key, value in dict(family).items()
if value is not None
}
default_voice[acss.ACSS.FAMILY] = family_dict
default_voice['established'] = True
def _get_current_speech_setting(self):
@@ -581,7 +585,7 @@ class SpeechAndVerbosityManager:
message = ""
if message:
voice = self._get_default_voice() if setting == "voice" else None
voice = self._get_default_voice() if setting in ("voice", "module") else None
self._present_message(script, message, voice=voice)
@dbus_service.command
@@ -698,7 +702,11 @@ class SpeechAndVerbosityManager:
else:
self._set_default_voice_family({})
self._present_message(script, messages.SPEECH_MODULE_VALUE % new_module)
self._present_message(
script,
messages.SPEECH_MODULE_VALUE % new_module,
voice=self._get_default_voice(),
)
return True
def _adjust_voice(self, script, decrease):