This directory contains all of the native code necessary for speech
synthesis on Android.

The external/espeak directory contains a port of espeak, a free tts engine.
Very little is changed - a build file (Android.mk) is provided, wide
characters are disabled, and a few advanced pthread features (not needed)
are commented out.

The frameworks/tts directory contains a C library and build files that
create a JNI shared library allowing you to call espeak from Java.
The corresponding Java class that works with this JNI code is found in
the top-level of eyes-free in: tts/src/com/google/tts/SpeechSynthesis.java

To compile the code here, you will need to download the latest Android
source code from git - note that this is not the SDK, but the entire
source package.  See the instructions here:

  http://source.android.com/download

Follow the instructions and complete a full build of Android before
proceeding.

Copy external/espeak to the "external" directory in the Android source, copy frameworks/tts to the "frameworks" directory in the Android source, then create
a java directory under frameworks/tts and copy SpeechSynthesis.java (from
MY_SVN_DIRECTORY/tts/src/com/google/tts) into frameworks/tts/java.
To recompile just those directories, assuming you have run
"source build/envsetup.sh" in your bash shell, you an just type:

  mmm external/espeak

  mmm frameworks/tts

When you're done, you'll get this file:

  out/target/product/generic/system/lib/libspeechsynthesis.so

Copy this file to tts/libs/armeabi. 

In order to make use of this file, you will need to make sure that libs/armeabi/libspeechsynthesis.so is copied into your apk before it is signed.
As of the Cupcake SDK (SDK 1.5), any .so files under libs/armeabi will be
picked up and packed automatically when building with the default ant scripts
generated by the Android SDK.