google2ubuntu

#UPDATE

I've updated google2ubuntu in order to be able to use the google web speech api v2. You need to request a key to google.

#Presentation

The aim of this project is to let you use the Google speech recognition API to control your linux computer. The project is developed in Python.

google2ubuntu is a tool that I started 2 years ago. Unfortunatly, I was not able to do what I wanted because of a lack of time. So the project consists in 2 principal Python scripts:

• google2ubuntu.py
• google2ubuntu-manager.py

The first one lets you send commands to Google and then execute some actions. The second one lets you manage all the commands.

I've done huge efforts to make the program easy to install and easy to use. Besides, the program can be internationalize, for the moment, it is already available in French, English, Deutch, Spanish and Italian.

Special thanks

Before I begin, I want to thank some of you that help me.

##Contributors

Devs	Translators	Bug reports
tectas	tectas	bmeznarsic
mte90	mte90	Andrew from Webupd8
levan7	lincus	mte90
ladios	ladios	mads5408
Josh Chen	Frank Claessen

##Publication

• WebUpd8
• La vache libre
• chimeravo

#Installation

Dependancies

For the moment, dependancies are:

• bash
• python
• python-gi
• libnotify-dev
• sox
• xdotool
• libsox-fmt-mp3
• acpi
• python-simplejson

If you are installing google2ubuntu from source not from deb or from ppa, type this:

   sudo apt-get install bash python python-gi python-simplejson libsox-fmt-mp3 sox libnotify-dev acpi xdotool

##Find the package google2ubuntu is available on Github on the release page and in my ppa. To install google2ubuntu

sudo add-apt-repository ppa:benoitfra/google2ubuntu
sudo apt-get update
sudo apt-get install google2ubuntu

##First launch ###Main programs Once you have installed google2ubuntu, you can attribute a shortcut to those 2 Python scripts:

/usr/share/google2ubuntu/google2ubuntu.py
/usr/share/google2ubuntu/google2ubuntu-manager.py

Moreover, if you search in the application's menu you will find two launchers, one for each of those programs.

After that, you can launch google2ubuntu-manager.py in order to manage all commands.

As you can see, google2ubuntu comes with several default commands. I will explain you how to manage and add commands.

###Basic configs Then you can configure google2ubuntu by clicking on the Setup button, that will open a window

In this window you can:

1. configure the language you want to use, by changing the language in the combobox. For the GUI the will take effect after restart.
2. Set the recording time (seconds) between 1 and 10 seconds
3. Set the command to pause/play the media player, (If you don't know what command to use let it emply)
4. Activate/Deactivate and configure the hotword mode

For the last two parameters, as there is a lot of media player (vlc, mplayer, banshee, ...) I think this way is the most efficient because it lets anybody writing a basic shell script that will pause or play his favorite media player. So those commands could be shell commands or could be more elaborated scripts

#Manage commands

##Commands' storage By default, google2ubuntu comes with several commands stored in a default xml file:

/usr/share/google2ubuntu/config/<your_language>/default.xml

For the moment, there is a default file for French, English, Spanish, Deutch and Italian users, if your language is not currently supported the default voice will be English.

At the first, launch, a module folder is created in:

~/.config/google2ubuntu

The first time you add, modify or remove a command, your commands' configuration will be also stored in this folder in the file:

~/.config/google2ubuntu/google2ubuntu.xml

##Commands' description A command is a pair of key and action. Each key referes to an action. Many key can leads to the same action. To define a command, you do not need to make explicitaly all the word you will tell, I mean, if I want to create the command:

key: open my documents
action: xdg-open ~/Documents

The word my is not usefull so, I will put:

key: open documents
action: xdg-open ~/Documents

Don't care about capital letter, because the program automatically put the text in lowercase.

I've implemented different types of command:

• external commands
• internal commands
• modules

###External commands

External commands are basically commands that you can run in your terminal:

exo-open --launch MailReader

If you want to add an external command, just click on the "Add" button. Then find the newline and replace "your key" by the key you want to associate to the command and replace "your command" by the action.

###Internal commands ####What are they ? Internal commands are commands that I've implemented in google2ubuntu, for the moment there is 3 internal commands:

Name	Function
time	Tell the current time
power	Tell the battery state
clipboard	Read the text selected by the cursor
dictation mode	Enter in dictation mode
exit dictation mode	Exit dictation mode

If you want to add an internal command, open the little menu near the "Add" button and select "internal". Then replace "your key" by the "key" you will pronounce to call this command and replace "word" by one of those 5 actions' name.

####Some words about dictation mode

The dictation mode let you type all word you pronouce. If you want to enter in the dictation mode use the dictation mode internal function. If you want to exit this mode just use the associated function. Dictation is not continue so you have to launch google2ubuntu for each sentence you want to type

###Modules In order to extand google2ubuntu very easily I've implemented a system of modules that lets developers adds their own scripts in google2ubuntu. Besides, all modules will receive the text that you pronounce in parameter

####Module's description A module is basically, an executable file that will receive some text in argument. each module embed its configuration in your configuration file. Two fields are recorded for each module:

• linker a vord that let us distinguish the call to the module and the parameter we have to send to this module. For example, if I want to configure the module google, I can choose the linker "google " because when I make a research I say:

google who is barack obama

So, the google module will be call with who is barack obama in parameter.

• spacebyplus If spacebyplus=1 then space are replace py +.

####How to add a module If you want to add a script, the gui will help you to create one and will place the module in :

~/.config/google2ubuntu/modules

You can add a module by opening the menu near to the "Add" button then selecting the executable files of the module. Yu can also simply drag&drop this executable on the treeview and the module will be automatically added. When you add a new module you don't have to modify the action field in the newline. You just have to modify the key field in the gui.

####Already available google2ubuntu already comes with 6 modules:

• google This plugin allows you to make search on Google and open the web browser on the search page you ask for.
• wikipedia This plugin allows you to make search on Wikipedia
• youtube This plugin allows you to make search on YouTube So your locations need to be between the word "between" and "and"
• weather This plugin allows you to ask Google to show you the weather for a city
• goto This plugin allows you to open a new web page with the web page you want. For example if you have configurated the plugin liker this:

your sentence = go to
linker = to 

Then if you say "go to gmail.com", a new web page is going to be open with gmail.com

• meaning This plugin allows you to ask to the meaning of a word, for exemple if I ask:

meaning barack obama

The plugin will tell me that he is the actual president of the US

####Note for the user Perhaps, you will have to modify the linker field of those module by selecting the module and clicking on the edit button

#Go Linux automation Once you have personalized and take care about the commands already included in google2ubuntu, you can launch the recognition by launching google2ubuntu.py

A little sound is played and a notification tell you to speak. Then the notification show the result and the action associated to the text you have pronounced is played.

#How-to contribute ##Design some modules If you want you can write a module that will be integrated in the Github page, the user will have to download it and will be able to use it ##Translate the app ###Translate the google2ubuntu core If you want to translate this app, you need first to download the project then unzip it and open a terminal and place yourself in the folder newly created. Then, be sure that all string that you want to translate are like that:

_('text to translate')

Then be sure, that there is this at the begenning of the file you want to translate (only needed for google2ubuntu.py, gogl2ubuntu-manager.py)

import gettext

gettext.install('google2ubuntu',os.path.dirname(os.path.abspath(__file__))+'/i18n/')

So, if you want to add a new language, you should type:

mkdir -p i18n/<new_language/LC_MESSAGES
xgettext --language=Python --keyword=_ --output=./i18n/google2ubuntu.pot ./*.py librairy/*.py
msginit --input=./i18n/google2ubuntu.pot --output=./i18n/<new_language>/LC_MESSAGES/google2ubuntu.po

The 2 first line are optional and are usefull only there is no .pot file in the i18n/ folder.

Then, open the .po file and translate all the line, then compoile the .po :

msgfmt ./i18n/<new_language>/LC_MESSAGES/google2ubuntu.po --output-file ./i18n/<new_language>/LC_MESSAGES/google2ubuntu.mo

###Update the translation To update an existing translation, you have to do several actions:

xgettext --language=Python --keyword=_ --output=./i18n/google2ubuntu.pot ./*.py librairy/*.py
msgmerge --update --no-fuzzy-matching --backup=off ./i18n/<current_language>/LC_MESSAGES/google2ubuntu.po ./i18n/google2ubuntu.pot

Then translate new line and compile the .po:

msgfmt ./i18n/<current_language>/LC_MESSAGES/google2ubuntu.po --output-file ./i18n/<current_language>/LC_MESSAGES/google2ubuntu.mo

Share ideas

You can share ideas and contact me on the google+ comunauty: google2ubuntu

Demonstration

Documentation page

I've wrote a documentation page with Sphinx Documentation

#Update

Improvements

State	Addons
	adding a menu to change language
	adding a class to manage locale easily
	adding a dictation mode
	adding the possibility to reconfigure a module from google2ubuntu-manager
	drag & drop a folder automaticcaly add a line in the treeview with a command to open it
	adding a man page
	adding spanish translation
	adding deutch translation
	adding italian translation
	correct dependancies
	adding Sox not recording in 16kHz bug fixe
	correct module setup button bug
	correct pyGobject constructor issue
	regroupe both remove all and delete in one button
	delete button bug fixe when no item selected
	localHelper not cut locale name
	beginning to work on the HelpWindow
	changing i18n locale's folder name
	correct a bug due to module execution
	adding pt_BR support
	adding pt_PT support
	adding traditional chinese translation
	correct the wikipedia module
	adding nl_NL support
	simplify module configuration storage
	adding an edit button that open an edit window
	adding an hotword mode
	Support for google web speech api v2
	update the documentation
	improve translation
	bug fixe
	adding russian translation