~vcs-imports/speech-dispatcher/trunk : changes

~vcs-imports/speech-dispatcher/trunk » Changes from revision 2981

From Revision 2597 to 2578

Rev	Summary	Authors	Date
2597	Add internationalization support to build infrastructure. Add internationalization support to build infrastructure. This patch adds intltool/gettext support to the build infrastructure. From now on, files marked for translation shuld be listed in po/POTFILES.in. Documentation for translators is in README.translators. The INSTALL file is updated with intltool reference. Reviewed-by: Andrei Kholodnyi <Andrei.Kholodnyi@gmail.com>	Rui Batista	13 years ago
2596	distribute the configuration files distribute the configuration files Make sure the configuration files are marked as part of the distribution so they will be included in the tarball. Reviewed-by: Christopher Brannon <cmbrannon79@gmail.com>	William Hubbs	13 years ago
2595	Reopen libao device handles when audio parameters change. Reopen libao device handles when audio parameters change. Audio parameters can change from one track to another. E.G., a fragment of speech might not have the same sample rate as a sound icon. We need to reopen libao's ao_device * handle when such changes occur. Reviewed-by: Hynek Hanke <hanke@brailcom.org> Additional note: Rate changes not only because of sound icons, but e.g. when switching to a different voice in Festival. So this patch is fully justified.	Christopher Brannon	13 years ago
2594	ANNOUNCE, NEWS, version.sh modified for 0.7.1 ANNOUNCE, NEWS, version.sh modified for 0.7.1 ANNOUNCE and NEWS lists new features.	Hynek Hanke	13 years ago
2593	Increment library version numbers for 0.7.1 release Increment library version numbers for 0.7.1 release libspeechd version incremented (backwards compatible) libsdaudio revision incremented	Hynek Hanke	13 years ago
2592	Added new authors and thanks into AUTHORS Added new authors and thanks into AUTHORS New persons were added. In the previous version, the people listed as AUTHORS were listed with brief listing of their accomplishemnts. This is however somehow ambigous and hard to maintain, so these lines have been removed and the section is now a plain listing of the authors names and contacts. Contacts were removed from the THANKS section.	Hynek Hanke	13 years ago
2591	Bug fix null pointer dereference Bug fix null pointer dereference spd_audio_set_volume() doesn't check that the audio id it is given is valid. Reviewed-by: Hynek Hanke <hanke@brailcom.org> Reindented, error message ammended with name of the function.	Trevor Saunders	13 years ago
2590	Raise exceptions on invalid inputs in speechd_config Raise exceptions on invalid inputs in speechd_config Since speechd_config can now be used as a library, it is important to check parameters on function entries. Formerly, asserts were used to check internal code consistency of the configuration type argument and checking of the correct value was done on user's input. Reviewed-by: Christopher Brannon <cmbrannon79@gmail.com>	Hynek Hanke	13 years ago
2589	Fix UTF-8 output for Cicero TTS Fix UTF-8 output for Cicero TTS This patch was submitted by Samuel Thibault via Debian BTS (Bug#594836) with the following comment: Cicero got utf-8 support. To tell cicero that you are using utf-8 output, you need to pass code 4 instead of 2. Reviewed-by: Hynek Hanke <hanke@brailcom.org>	Hynek Hanke	13 years ago
2588	remove double include from speechd.conf remove double include from speechd.conf There was an extra include directive in speechd.conf, which was also the reason this file was being generated from speechd.conf.in. If this directive was in place and the file was installed as the system configuration and not a user level configuration, the client settings were being included twice. If the system level configuration includes the files in clients/, that will cover the default settings. If a user wants to include the system level client settings, they should add an include directive to their user level configuration file to do so. By default, spd-conf copies the client settings over to the user's configuration, so they will not need to do this. Reviewed-by: Christopher Brannon <cmbrannon79@gmail.com>	William Hubbs	13 years ago
2587	Change CVS to Git in INSTALL file Change CVS to Git in INSTALL file Reviewed-by: Christopher Brannon <cmbrannon79@gmail.com>	Rui Batista	13 years ago
2586	fix Usage prompt in spd-conf fix Usage prompt in spd-conf added also a short file description Reviewed-by: Christopher Brannon <cmbrannon79@gmail.com>	Andrei Kholodnyi	13 years ago
2585	sd_generic: fix escaping of data destined for shell. sd_generic: fix escaping of data destined for shell. All of the *-generic.conf files quote $DATA with single quotes ('). We shouldn't escape \, ", and `, because these characters are already interpreted literally by the shell. We do need to escape single quotes. This is done by closing the string, inserting \', and re-opening the string. Reviewed-by: Hynek Hanke <hanke@brailcom.org>	Christopher Brannon	13 years ago
2584	Remove unused audio settings and include AudioPulseMinLength Remove unused audio settings and include AudioPulseMinLength Some options in the configuration file were no longer active due to changes introduced in previous patches and some active options were missing from the configuration file. Additionally, some minor improvements in the option descriptions were made and the order of audio sections was changed according to their relevance. Reviewed-by: Christopher Brannon <cmbrannon79@gmail.com>	Hynek Hanke	13 years ago
2583	Disable audio fallback and use Pulse Audio by default Disable audio fallback and use Pulse Audio by default Audio fallback is the right solution, but a prerequisite is proper freeing of resources when the given instance of server is not in an active session. Since this is currently not implemented, audio fallback leads to problems. Under the previous settings "pulse, alsa", if Speech Dispatcher is started in a session where Pulse is not available, it will open ALSA, thus preventing Pulse to start in all other sessions. This subsequently breaks all audio for all users although the session which blocks the device is actually not active. To prevent this, audio fallback is not used by default. Additionally, mention libao, which was missing. Reviewed-by: Christopher Brannon <cmbrannon79@gmail.com>	Hynek Hanke	13 years ago
2582	spd-conf revised to reflect the current state spd-conf revised to reflect the current state The testing, configuration and debugging modes in spd-conf were revised according to the current documentation. Loops have been removed from the tests as they made the flows unnecessarily complicated. It's better to lead the user to an error result as soon as possible and let him optionally repeat the tests by a new call to spd-conf after he fixes the issue. Some messages have been clarified and flows have been modified slightly based on testing the various situations. The dispatcher_working result key has been renamed to spd_say_working as this is more precise (failure can be on client side while server is working OK). Remove unnecessary semicolons. Reviewed-by: Christopher Brannon <cmbrannon79@gmail.com>	Hynek Hanke	13 years ago
2581	use AO_FMT_NATIVE instead of AO_FMT_LITTLE in libao plugin use AO_FMT_NATIVE instead of AO_FMT_LITTLE in libao plugin use native format and let libao figure out endianness Reviewed-by: Christopher Brannon <cmbrannon79@gmail.com>	Andrei Kholodnyi	13 years ago
2580	fix another loop fix another loop the output_modules case of the LISt command had another loop where a linked list was indexed into instead of using next pointers. Reviewed-by: Christopher Brannon <cmbrannon79@gmail.com>	Trevor Saunders	13 years ago
2579	clean up and optimize get_output_module clean up and optimize get_output_module - reorganize the function to return if we have a module, to remove nesting. - walk the linked list of output modules with next pointers instead of indexing into the list. Reviewed-by: Christopher Brannon <cmbrannon79@gmail.com>	Trevor Saunders	13 years ago
2578	Fix memory leaks in module_audio_init_spd. Fix memory leaks in module_audio_init_spd. We missed these in the last round of memory leak fixes. * Free the error variable, set when spd_audio_open fails. Its value is set by g_strdup_printf, so we free it with g_free. * The outputs variable is set by a call to g_strsplit. It is a list of strings, terminated with a NULL pointer. As per the glib docs, we now free it using g_strfreev. Signed-off-by: Christopher Brannon <cmbrannon79@gmail.com> Reviewed-by: Hynek Hanke <hanke@brailcom.org>	Christopher Brannon	13 years ago