aboutsummaryrefslogtreecommitdiff
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* Add gcda/gcno to CLEANFILES s.t. 'make clean' will cover themCharles Kerr2012-06-011-0/+3
|
* in indicator_power_device_get_time_details(), limit the scope of variables ↵Charles Kerr2012-05-311-3/+3
| | | | short_timestring and detailed_timestring
* Move private indicator-power function build_device_time_details() to ↵Charles Kerr2012-05-313-198/+203
| | | | device.c to public function indicator_power_device_get_time_details() so that we can unit test the user-visible strings.
* remove IndicatorPowerDevice's now-unused device icon string.Charles Kerr2012-05-312-129/+0
|
* Add indicator_power_device_get_icon_names().Charles Kerr2012-05-313-109/+255
| | | | | | | The main goal of this change is to make it possible to test the device's icon. A secondary goal is to clarify in the code how indicator-power's icons differ from the ones recommended by GSD.
* remove some g_message() calls that shouldn't've been committed.Charles Kerr2012-05-311-5/+0
|
* add a test to make sure that the listener responds to a PropertiesChanged signalCharles Kerr2012-05-291-0/+5
|
* t^Cak to indicator_set_power_differences()Charles Kerr2012-05-291-3/+2
|
* use g_clear_pointer() where usefulCharles Kerr2012-05-291-6/+2
|
* exclude G_DEFINE_TYPE from coverage testing in dbus-listener.cCharles Kerr2012-05-291-0/+2
|
* add a mock GSD.Power to handle the GetPower requestsCharles Kerr2012-05-282-8/+8
|
* use signals to decouple i-power and dbus-listenerCharles Kerr2012-05-274-102/+38
|
* try adding LCOV_EXCL_{START,STOP} for unreachable conditions (glib looking ↵Charles Kerr2012-05-262-3/+23
| | | | for subclasses of IndicatorPowerDevice; unreachables in G_DEFINE_TYPE)
* nope, LCOV_EXCL_LINE doesn't work on macrosCharles Kerr2012-05-261-7/+7
|
* try adding LCOV_EXCL_LINE for unreachable conditions (glib looking for ↵Charles Kerr2012-05-261-7/+7
| | | | subclasses of IndicatorPowerDevice; unreachables in G_DEFINE_TYPE)
* reuse the same cancellable across multiple non-concurrent dbus callsCharles Kerr2012-05-261-18/+18
|
* avoid a cyclical refcount dependency between IndicatorPower and its DBusListenerCharles Kerr2012-05-261-3/+17
|
* separate the dbus org.gnome.SettingsDaemon.Power logic into a separate classCharles Kerr2012-05-264-147/+366
|
* make indicator_power_set_devices() safe for passing in the same devices more ↵Charles Kerr2012-05-251-12/+14
| | | | than once
* fix negated logic in menu_add_device() introduced in r160Charles Kerr2012-05-251-1/+1
|
* indicator_power_set_devices() should fail gracefully when no devices are ↵Charles Kerr2012-05-251-2/+5
| | | | available
* modify IndicatorPower to use IndicatorPowerDevices internallyCharles Kerr2012-05-252-185/+165
|
* put IndicatorPower's fields back inside a priv structCharles Kerr2012-05-252-74/+99
|
* remove the 'default' clause from get_property() and set_property().Charles Kerr2012-05-241-8/+0
| | | | | Coverage testing isn't reaching them... glib is weeding out these invalid property keys before the device.c functions are ever reached. Nevertheless, leaving out a 'default' clause in a switch statement feels very unnatural to me. *twitch* *twitch*
* remove some unnecessary type compatibilty tests... g_object_get_property() ↵Charles Kerr2012-05-241-6/+0
| | | | does these tests for us
* remove some unnecessary type compatibilty tests... g_object_set_property() ↵Charles Kerr2012-05-241-6/+0
| | | | does these tests for us
* remove the g_clear_pointer() calls s.t. things will build and run on ↵Charles Kerr2012-05-241-2/+6
| | | | alesage's Jenkins setup running Precise
* add a 'hello world' test for instantiating IndicatorPowerCharles Kerr2012-05-242-40/+69
|
* add private container struct IndicatorPowerDeviceCharles Kerr2012-05-243-2/+436
|
* replace 'Power Settings...' with 'Power Settings…'Charles Kerr2012-05-231-1/+1
|
* move libpower's automake rules into src/Charles Kerr2012-05-231-0/+28
|
* tweak: in put_primary_device(), peek at the variant's strings instead of ↵Charles Kerr2012-04-151-5/+3
| | | | dup'ing them
* sync with lp:indicator-powerCharles Kerr2012-04-151-3/+10
|\
| * use atk_object_set_name(foo,bar)" instead of ↵Charles Kerr2012-04-091-1/+1
| | | | | | | | g_object_set(foo,"accessible-name",bar,NULL)"
| * leak fix 3 of 3: don't leak the GIcons returned by get_device_icon()Charles Kerr2012-04-091-0/+2
| |
| * leak fix 2 of 3: fix leaky fallback GIcon in get_device_icon()Charles Kerr2012-04-091-3/+4
| |
| * leak fix 1 of 3: don't leak object_path and device_icon if kind == ↵Charles Kerr2012-04-091-6/+4
| | | | | | | | UP_DEVICE_KIND_LINE_POWER
| * if possible, set the device menuitems' accessible names.Charles Kerr2012-04-091-0/+4
| |
* | fix leaked strings in menu_add_device when (kind == UP_DEVICE_KIND_LINE_POWER)Charles Kerr2012-04-151-6/+4
| |
* | Fix memory leaks in get_primary_device().Charles Kerr2012-03-171-32/+28
| | | | | | | | | | | | | | 1. All the calls to g_variant_get_child_value() were leaked. Fixed by changing the use to g_variant_get_child() and keeping index values of the interesting children instead of pointers to them. 2. There were several paths where the local string "object_path" and "device_icon" were leaked. (For example, any non-battery entry given to us by upower). Fixed by making these const strings and peeking them from the variant with "&s" instead of "s".
* | Fix variant leak in count_batteries() -- the returned value of ↵Charles Kerr2012-03-171-0/+2
| | | | | | | | g_variant_get_child_value() needs to be freed with g_variant_unref() when we're done with it.
* | fix variant leaks in menu_add_devices()Charles Kerr2012-03-171-2/+2
|/
* when a new entry is added, ensure that its accessible_string is properly setCharles Kerr2012-02-221-10/+37
|
* fix trivial indentation errorCharles Kerr2012-02-221-1/+1
|
* fix trivial misspellingCharles Kerr2012-02-221-16/+16
|
* fix set_accessible_desc() to update the entries' accessible_desc fields and ↵Charles Kerr2012-02-221-6/+17
| | | | emit the accessible-desc-changed signal
* register for g-properties-changed rather than g-signalCharles Kerr2012-02-161-19/+15
|
* Use G_GNUC_CONST for indicator_power_get_type() to improve performanceJavier Jardón2012-02-141-1/+1
|
* Code style fixesJavier Jardón2012-02-141-16/+19
|
* ensure that we don't have a reference to the proxy or proxy_cancel fields in ↵Charles Kerr2012-02-131-0/+3
| | | | indicator_power_dispose().