===== Setup build environment for github code under Ubuntu =====
$ sudo apt install autoconf libtool 
$ cd ekho
$ ./autogen.sh
$ sudo apt install libsndfile1-dev libespeak-ng-dev libpulse-dev texinfo libltdl-dev libmpg123-dev libsonic-dev libutfcpp-dev
$ ./configure
$ make CXXFLAGS=-O0
$ ./ekho 123

If you have installed version Ekho and you want to use voice data in working directory, you should set EKHO_DATA_PATH like this:
$ env EKHO_DATA_PATH=ekho-data ./ekho 123

===== INSTALL for Linux =====
You need to have following libraries first:
* texinfo
* libltdl-dev
* libmpg123-dev
* libsonic-dev (optionally required by change speed)
* libutfcpp-dev (https://utfcpp.sourceforge.net/)
* libespeak-ng-dev
* libsndfile1-dev (http://www.mega-nerd.com/libsndfile/)
* libpulse-dev
* libncurses5-dev (required by --enable-festival)
* libestools-dev (optionally required by --enable-festival)
* festival-dev  (optionally required by --enable-festival)
* libvorbis-dev (optional)
* libmp3lame-dev (optional)
* libdotconf-dev (required by --enable-speechd)

On Ubuntu Linux, we can use `apt-cache search <pkg_keywork>` to find a package's full name and install it with `sudo apt-get install <pkg_name>`.
On Fedora Linux, we can use `yum search <pkg_keyword>` to find a package's full name and install it with `sudo yum install <pkg_name>`.
On SuSE Linux, we can use `zypper search <pkg_keyword>` and `zypper install <pkg_name>`.

$ ./configure
$ make CXXFLAGS=-O0
$ sudo make install

Some configure options:
--enable-festival      Speak English through Festival.
--enable-speechd       Enable speech-dispatcher API and work with screen reader like Orca. (You also need to do some extra setup. If you don't know how to setup, just run ./install.pl)
--with-mp3lame=yes     Enable MP3 export function. This option is conflict to --enable-speechd.

To install and setup the speech-dispatcher, please run one of following commands (only one, it will overwrite previous setting):
$ ./install.pl Mandarin
$ ./install.pl Cantonese
$ ./install.pl Tibetan

===== INSTALL for Mac =====
$ brew install libsndfile
$ brew install pulseaudio
$ brew install utf8cpp
$ install espeak-ng from source (glibtoolize -c && ./autogen && ./configure && make && sudo make install)
$ install sonic from source: https://github.com/waywardgeek/sonic
$ ./configure
$ brew list utf8cpp # to find utf8.h path
$ make CXXFLAGS="-O0 -I/usr/local/Cellar/utf8cpp/3.2.1/include/utf8cpp"
$ brew services start pulseaudio
$ ./ekho 123

===== server mode =====
Make sure configure with following option for server mode
$./configure --with-mp3lame

===== Notes for git clone =====
If you need to enable speech-dispatcher plugin, you need to do following:

git clone https://github.com/hgneng/ekho.git
git clone https://github.com/hgneng/speechd.git
mv speechd ekho/speechd-api