| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
| |
This reverts commit 3b9e0eec58f1b6d99418eb6365d7b8ae1c2082a2.
After more testing, this change seems to poorly coexist with CL 21226.
|
| |
|
|
| |
in the pico library.
|
| |\
| |
| |
| |
| |
| |
| |
| |
| | |
louder. This is a fix for bug 2046705.
Merge commit '254a9349e363440f81269547574de1cbe681947b' into eclair
* commit '254a9349e363440f81269547574de1cbe681947b':
Updating the TextToSpeech data files to make the TTS sound louder.
|
| | |
| |
| |
| | |
This is a fix for bug 2046705.
|
| |\ \
| |/
|/|
| |
| | |
* changes:
Integrate SVOX update that offers two modes to reset the engine: soft and full. Soft reset is lighter-weight than the previous pico_resetEngine() function as it only clears the input (text) and output (audio data) buffers of the synthesis engine. The synthesis loop is now taking advantage of the soft reset whenever it is interrupted, rather than performing a full engine reset after each call to stop().
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
and full. Soft reset is lighter-weight than the previous pico_resetEngine()
function as it only clears the input (text) and output (audio data)
buffers of the synthesis engine.
The synthesis loop is now taking advantage of the soft reset whenever
it is interrupted, rather than performing a full engine reset after
each call to stop().
|
| |/
|
|
| |
BUG=1573996
|
| | |
|
| |
|
|
|
|
|
|
|
|
| |
After the user has initiated the download of the TTS language pack installer,
and that the installer has automatically started running, the installer
(in InstallerActivity) will broadcast an event
(TextToSpeech.Engine.ACTION_TTS_DATA_INSTALLED) to signal the completion
and success of the installation. The event is picked up by LangPackUninstaller
which triggers the deletion of the installer package to avoid wasting space
on the user's device.
|
| |
|
|
|
|
|
|
|
|
| |
When the user clicks "install" for the TTS language pack installer and
the download completes, the addition of the corresponding package is
picked up by VoiceDataInstallerReceiver. It then causes the installer
to run by firing off the RUN_TTS_DATA_INSTALLER event.
This is less confusing for the user because the installation of the
language files is now automatic once the download is initiated. Before,
the user had to "open" the installer after the download completed.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Added support for TTS language files to be loaded from the system partition if
they are not available on the regular path (sdcard). This involves changing
the check for the availability of a language (first check in the system path,
then check the sdcard), and the language switch code (try to load from the
system path if cannot load from sdcard).
Modified the CheckVoiceData activity (which checks for the installation of
all language files) to check on both the sdcard and the system path.
Added makefiles for copying language files to the system image,
PicoLangDefaultInSystem.mk will be used in another CL to copy a default language
to the system.
|
| |\
| |
| |
| |
| | |
* changes:
Updating binary voice files for en-US for better quality. Changing the Intent extra for getting the voice data root directory to match the current naming of that variable in TextToSpeech.java.
|
| | |
| |
| |
| |
| |
| | |
Changing the Intent extra for getting the voice data root
directory to match the current naming of that variable in
TextToSpeech.java.
|
| |/
|
|
|
| |
issue where a language couldn't be successfully loaded if the first
loaded language was not eng-USA (language 0).
|
| |
|
|
| |
Use the new names of the extra constants in TextToSpeech.Engine.
|
| | |
|
| |
|
|
| |
with Pico TTS.
|
| |
|
|
|
| |
Modifying the SSML volume definitions to match the default
volume.
|
| |
|
|
| |
making sure to clean up after each synth abort.
|
| |
|
|
|
|
| |
interrupted by a call to stop(). This was due to the engine not being
reset in certain code paths when a synth abort was signaled. The fix
adds a call to reset the engine whenever the synthesis loop is interrupted.
|
| |
|
|
| |
is not present.
|
| |
|
|
| |
TextToSpeech.Engine.CHECK_VOICE_DATA_MISSING_DATA_NO_SDCARD to CHECK_VOICE_DATA_MISSING_DATA_NO_EXTERNAL_STORAGE.
|
| | |
|
| |
|
|
|
| |
is no language specified in the <speak> tag. Using currently loaded
language in that case, or the first lang is none was previously loaded.
|
| |
|
|
| |
all content after the separator to be dropped.
|
| |\
| |
| |
| |
| | |
* changes:
Integrate latest SVOX code drop, includes support for the phoneme tag. Reformat TtsEngine implementation to Android style.
|
| | |
| |
| |
| | |
Reformat TtsEngine implementation to Android style.
|
| |/ |
|
| |
|
|
|
|
|
| |
- Fix a bug where TTS_SYNTH_DONE was signaled to the TTS service synth proxy
even though the synthesis wasn't done (more data to be played).
- Deprecated the synthesizeIPA() function as it will be supported by the
phoneme tag in synthesize().
|
| | |
|
| |
|
|
|
|
|
| |
language,
to set the synthesis audio format, to query the current language, country and
variant setting.
|
| |
|
|
|
|
| |
loadLanguage.
Adding skeleton for isLanguageAvailable().
|
| | |
|
| |
|
