Several changes to CLI behavior, all minor:
* Inverted the behavior or PgUp/PgDn. I think it was just wrong
* Added "oldest" and "newest" indicators to the log listing to
simplify finding your way inside the logs
* Added support for _not_ scrolling when in the middle of the log
output. It will now only auto-scroll when already at the front
of the log so you keep seeing new messages automatically.
* Added a heading for the mic output level
* Limited the mic output level to just the size of the bottom
area (previously it was proportional to the screen height)
* FIX: Changed the default filter string for viseme
* FIX: The saving of settings was broken
* Ran autopep8 which added a few bits of whitespace here and there
Bug fix proposed for success wake-word recording.
Opt: Suggestion to change the save location of recording files as well.
Otherwise, the user could potentially get the permission denial.
-NeonGecko.com Inc
Add intent fallback system
==== Fixed Issues ====
NONE - replace with issue numbers, e.g. #123, #304
==== Tech Notes ====
NONE - explain new algorithms in detail, tool changes, etc.
==== Documentation Notes ====
The FallbackSkill needs to be documented in detail for skill creators.
==== Environment Notes ====
NONE
==== Protocol Notes ====
multi_utterance_intent_failure replaced with complete_intent_failure
This is because this will break wolfra alpha skill unless they update
skills, but if they update before getting the new version, it will also
break wolfram
This reverts commit 6ca4161335.
* Create new FallbackSkill base class for implementing fallback behavior
Also removes multi utterance intent fail. Only makes sense to emit an intent_failure regardless of the amount of intents
* The allow_timeout value was tested poorly, preventing it from actually doing anything
* Upped the number of arp/ping test failures from 3 to 6 when validating the connection. This makes falsely detecting a wifi setup disconnect pretty unlikely.
Improved the way the button press is handled. Now the listener gives the rest
of the system a moment to consume the button press before it interprets it
as a request to begin listening.
Also replaced the use of the "buttonPress" signal when "expect_response"
was indicated with an utterance. Now the signal "startListening" is
created, differentiating it from a buttonPress.
This implements the handler for the Mark 1 menu item SSH > BLOCK (the inverse of SSH > ALLOW)
* Added handler for "unit.disable-ssh" on the serial line. This turns around and emits "mycroft.disable.ssh" on the messagebus.
* Removed the automatic reboot, just let the user know it will be different after a restart
* Made the spoken message translatable
* Changed mycroft.dialog.get() to not require the "lang" parameter. It will default to the mycroft.conf value.
* Adding localization mechanism for strings embedded in mycroft-core code
Added mycroft.dialog.get() function. This behaves much like the localization
mechanism for dialogs in Skills. So you can do things like this:
lang = "en-us"
str = mycroft.dialog.get("how are you", lang)
Which will look in mycroft/res/text for the dialog file containing strings to
use as templates for the actual output. This depends on the language being
currently used. When operating in English this would be:
mycroft/res/text/en-us/how are you.dialog
This function will pick a random line from that file to assign to str.
A more advanced use is to embed placeholders in the strings within the
template file.
lang = "en-us"
ctx = {"time" : "noon"}
str = mycroft.dialog.get("current time", lang, ctx)
And the random template line picked was "the current time is {{time}} ", then
the output would be:
"the current time is noon"
* BUGFIX: The big bug was calling is_paired() during wake_word_in_audio(). When not paired, that call hit the server, taking about a second. Since it happened multiple times a second, the audio buffers got backed up hugely. This resulted in weird behavior later as the buffers get cleared out.
* Added mycroft.api.has_been_paired(), which just looks for the pairing key (it does not validate it is still active with the server, like is_paired())
* The enclosure now checks for internet connectivity and kicks off the wifisetup process, not the wifisetup client itself.
* During the "onboarding" process, the microphone is muted using the new "mycroft.mic.mute" message. After pairing completes, the "mycroft.mic.unmute" is expected to be sent from the pairing skill. Unmuting again after a re-pairing is harmless.
* mute_and_speak() is smart enough to not unmute itself when complete if muted before
* util.check_for_signal() now accepts -1 as the lifetime. This means it never times out.
* util.stop_speaking() is more intelligent about shutting down the spoken text (including text that has been split at periods) and visemes
* The wifi setup no longer stops after 5 minutes unless already paired (i.e. still onboarding)
* The mouth resets on wifisetup stop (clearing the scrolling home.mycroft.ai)
* Changed several ConfigurationManager.get() calls to ConfigurationManager.instance(). Exactly the same, but .instance() is clearer/preferred.
* Added a mycroft.util.stop_speaking() method. Not perfect, but works for now and can be replaced later when AudioManager is in place.
A forthcoming change to the pairing skill will utilize the stop_speaking method.
* Added a mycroft.api.is_paired() method
* Added mycroft.util.is_speaking and mycroft.util.wait_while_speaking() methods
* RESET now waits for the spoken notice to complete
* Stopped the "Checking for updates" and "Skills updated" prompts (commented out for now, probably will eliminate)
* Wifi setup filters out hidden ("x00") networks
* Visemes should keep up better if they get behind (will skip)
* Mimic is now searched for on the users path
* Onboarding process:
- wifi setup starts automatically
- User is walked through the process
- wake word and button pressing are ignored
- At end, a short tutorial is given
I'm not 100% certain why this is needed, to be honest. Maybe something I don't get about Python and threads? This fixes double-announcements about connecting to the internet.
On the Mark 1 device, there is a sequence of events that produce a guide to the user for getting their device setup and connected to the internet. This process is basically:
* Enclosure client starts up, sending a message to the Arduino
* When/if the Arduino responds, it checks for a live internet connection
* If no connection is found, the user is prompted
However the speech client was sometimes not fully up, so the request to speak the prompt was sent before it was listening.
Additionally, sometimes events would occur that caused the prompt to connect the unit multiple time in a row. This commit limits the prompts to a max of once every 30 seconds.
Finally, I commented the code and renamed some of the methods to clarify this sequence of events. The code:
Timer(5, self.stop).start()
and:
def stop(self):
if not self.started:
self.writer.stop()
Are some of the most confusing lines of code I've ever seen. :)
Major changes:
* The messagebus message "enclosure.start" is now "enclosure.started"
* Enclosure.start() is now Enclosure.on_arduino_responded()
* Enclosure.started is now Enclosure.arduino_responded
* Added Enclosure.last_internet_notification to track notification times
* The notification call to the speech client is now delayed by 5 seconds, providing time for the speech client to come up. (And an additional check is run to verify the connection is still down immediately before the verbal prompt is spoken)
The restores the initial check on bootup for network connectivity,
prompting the user to plug in the network cable or push the button
to start Wifi setup.
NOTE: This will need to be localized and handled by different
enclosures.
Improved the way the button press is handled. Now the listener gives the rest
of the system a moment to consume the button press before it interprets it
as a request to begin listening.
Also replaced the use of the "buttonPress" signal when "expect_response"
was indicated with an utterance. Now the signal "startListening" is
created, differentiating it from a buttonPress.
This implements the handler for the Mark 1 menu item SSH > BLOCK (the inverse of SSH > ALLOW)
* Added handler for "unit.disable-ssh" on the serial line. This turns around and emits "mycroft.disable.ssh" on the messagebus.
* Removed the automatic reboot, just let the user know it will be different after a restart
* Made the spoken message translatable
* Changed mycroft.dialog.get() to not require the "lang" parameter. It will default to the mycroft.conf value.
* Adding localization mechanism for strings embedded in mycroft-core code
Added mycroft.dialog.get() function. This behaves much like the localization
mechanism for dialogs in Skills. So you can do things like this:
lang = "en-us"
str = mycroft.dialog.get("how are you", lang)
Which will look in mycroft/res/text for the dialog file containing strings to
use as templates for the actual output. This depends on the language being
currently used. When operating in English this would be:
mycroft/res/text/en-us/how are you.dialog
This function will pick a random line from that file to assign to str.
A more advanced use is to embed placeholders in the strings within the
template file.
lang = "en-us"
ctx = {"time" : "noon"}
str = mycroft.dialog.get("current time", lang, ctx)
And the random template line picked was "the current time is {{time}} ", then
the output would be:
"the current time is noon"
NOTE: this doesn't check for disk space and doesn't clean up
file stored in /tmp/ called mycroft_wake_sucess<time>.wav
TODO: some cleanup should be considered.
Signed-off-by: Art McGee <amcgee7@jaguarlandrover.com>
esoleyman
Brian Hopkins
jarbasai
Jarbas
Michael Nguyen
Steve Penrod
Åke Forslund
reginaneon
Matthew D. Scholefield
Arron Atchison
Sujan Patel
Art McGee
kfezer@gmail.com
Augusto Monteiro 'Sparky
kfezer
Ethan Ward