On 9/21/23, Leonard de Ruijter ***@***.***> wrote: Is there a particular reason why we need two scripts for this? Shoudln't this just be a single cycler? -- Reply to this email directly or view it on GitHub: #15493 (comment) You are receiving this because you authored the thread. Message ID: ***@***.***>

On 9/22/23, Sean Budd ***@***.***> wrote: Is this ready for review? -- Reply to this email directly or view it on GitHub: #15493 (comment) You are receiving this because you authored the thread. Message ID: ***@***.***>

On 9/22/23, Leonard de Ruijter ***@***.***> wrote: @LeonarddeR commented on this pull request. > @@ -66,6 +66,7 @@ from base64 import b16encode import vision from utils.security import objectBelowLockScreenAndWindowsIsLocked +from utils.synth import getOutputDeviceNames, setOutputDevice I'm slightly puzzled why this module is called synth while in fact, it contains logic for audio. I'm also not sure why this can't just reside in the nvwave module. -- Reply to this email directly or view it on GitHub: #15493 (review) You are receiving this because you authored the thread. Message ID: ***@***.***>

On 11/10/23, Łukasz Golonka ***@***.***> wrote: @lukaszgo1 requested changes on this pull request. > @@ -4289,6 +4290,37 @@ def script_cycleParagraphStyle(self, gesture: > "inputCore.InputGesture") -> None: config.conf["documentNavigation"]["paragraphStyle"] = newFlag.name ui.message(newFlag.displayString) + def cycleOutputAudioDevices(self, direction: int): Can this be made private (underscore prefixed), and converted to a static method, as far as I can tel it does not need access to the instance. Also the type annotation should reflect the fact that `direction` cannot be an arbitrary int, but only `1`, or `-1`, I'd suggest using `typing.Literal` here. > + index = deviceNames.index(config.conf["speech"]["outputDevice"]) + newIndex = (index + direction) % len(deviceNames) + except ValueError: + newIndex = 0 + device = deviceNames[newIndex] + setOutputDevice(device) + ui.message(device) + + @script( + description=_( + # Translators: Describes the switch to next output device command. + "switches to the next output device" + ), + category=SCRCAT_SPEECH, + ) + def script_switchToNextOutputDevice(self, gesture: "inputCore.InputGesture") -> None: No point in stringizing type hint here, same for the script below. > @@ -0,0 +1,31 @@ +# A part of NonVisual Desktop Access (NVDA) +# Copyright (C) 2023 Beka Gozalishvili +# This file may be used under the terms of the GNU General Public License, version 2 or later. +# For more details see: https://www.gnu.org/licenses/gpl-2.0.html + + +import config +from gui.settingsDialogs import _synthWarningDialog +import nvwave +from synthDriverHandler import getSynth, setSynth +import tones + + +def getOutputDeviceNames(): I'd suggest this function should be moved to `nvwave`, with an appropriate name so that it is clear it returns friendly device names. > @@ -8,6 +8,7 @@ What's New in NVDA == New Features == - Added support for Bluetooth Low Energy HID Braille displays. (#15470) +- Added unassigned gestures to to quickly cycle through output devices forward and backward. (#8801) Double 'to': ```suggestion - Added unassigned gestures to quickly cycle through output devices forward and backward. (#8801) ``` > @@ -1775,6 +1775,7 @@ The Audio category in the NVDA Settings dialog > contains options that let you cha ==== Output device ====[SelectSynthesizerOutputDevice] This option allows you to choose the audio device that NVDA should instruct the selected synthesizer to speak through. +To switch to the next or previous output device, pleas assign custom shortcuts from [Input Gestures dialog #InputGestures]. Pleas -> please > +import tones + + +def getOutputDeviceNames(): + deviceNames = nvwave.getOutputDeviceNames() + # #11349: On Windows 10 20H1 and 20H2, Microsoft Sound Mapper returns an empty string. + if deviceNames[0] in ("", "Microsoft Sound Mapper"): + # Translators: name for default (Microsoft Sound Mapper) audio output device. + deviceNames[0] = _("Microsoft Sound Mapper") + return deviceNames + + +def setOutputDevice(device: str): + config.conf["speech"]["outputDevice"] = device + currentSynth = getSynth() + if not setSynth(currentSynth.name): I'm not sure I agree with the logic here. I understand you haven't written this code, but since we're revisiting it anyway let's try to make it nicer. Currently we set the configured device **before** checking if it can be used i.e. if the current synthesizer can speak through it. Also with the current design there is an import cycle, since you import `gui` here, yet `gui` imports from this new module. I'd move this to `nvwave`, but rewrite the code (in my view basic module for playing sound should not depend on other modules `tones`, synthesizer drivers, which depend on it), using `extensionPoints.Decider` in `nvwave` to which both `tones` and `synthDriverHandler` can subscribe (something like `onMainAudioDeviceChanged`). If both synthesizer and tones can work with the new device, only then it should be set in the config. Also in that case you can check what has been returned from the decider in the settings dialog, and display error in the gui there, and announce the new device, or failure in switching into it in the keyboard scripts. With this approach it would be necessary to modify tones and `setSynth` to accept the specified audio device, as until now they were always using the default one. -- Reply to this email directly or view it on GitHub: #15493 (review) You are receiving this because you authored the thread. Message ID: ***@***.***>

On 2/16/24, Luke Davis ***@***.***> wrote: @beqabeqa473 What is the status of this? I would love to see it go to beta. I am in need of this feature. I'm going to have to implement it in an add-on if this PR is not proceeding soon. Life gets in the way, I very much understand, but please mention it if you need help bringing this forward. -- Reply to this email directly or view it on GitHub: #15493 (comment) You are receiving this because you were mentioned. Message ID: ***@***.***>