Project: help f877da8362728007c4771e731487149c084fc2da
typo: add missing space
via translator feedback
Change-Id: I932dde4cc9d728830f8eeb35453853b9d63bfe9c
It was coded pretty badly, so I cleaned it up. Most notably, it used
a static instance of the Activity, which is a huge no-no which
creates memory leaks. The irony is, it already had a reference to the
Activity used correctly in the constructor. One memory-leak fixed,
29 more to go (LibreOfficeMainActivity holds that static Activity
object which needs fixing). Also, simplified the "bottom toolbar"
in preparation for the CoordinatorLayout implementation which will
allow the activity to have fancy animations and smart interactions.
Change-Id: I31aa117f6179910db73a5256b0a287357e1dec83
Reviewed-on: https://gerrit.libreoffice.org/33010
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: jan iversen <jani@documentfoundation.org>
Fixed typo in SearchController (SearchDriection -> SearchDirection)
Also tightened up the code.
Change-Id: Iedb6f95c9d8ed04cc166638250c2d565e92d79ff
Reviewed-on: https://gerrit.libreoffice.org/33011
Reviewed-by: jan iversen <jani@documentfoundation.org>
Tested-by: jan iversen <jani@documentfoundation.org>
Because RecyclerView is more optimized, especially if there are lot
of items. This way, we don't have to recreate ListView and GridView
each time we switch view modes. Changed list adapter to appropriate
RecyclerView adapter, and created new grid adapter inline, next to
the list adapter, while deleting the older grid adapter file. Since
these adapters are almost identical in content, maybe we could:
a) Make them extend the same "base" adapter, to avoid duplicate code
b) Unite them into one adapter which would display appropriate views
at appropriate times.
Change-Id: I1545c2c245ca642a689dee584bffb15f90aac4a6
Reviewed-on: https://gerrit.libreoffice.org/32976
Reviewed-by: jan iversen <jani@documentfoundation.org>
Tested-by: jan iversen <jani@documentfoundation.org>
Replaced the custom implementation of the drawer with the support library one.
This one inherently follows Material Design guidelines, and is much easier to
maintain. This implementation also allows for header in the drawer, and so
we could put something useful there to make the drawer even better. Also kept
the original way of programatically adding the menu items, although I find this
practice somewhat unelegant. Maybe we could have static list of items, and then
grey-out the ones that aren't currently available? Also added appropriate icons
to the menu items (which are vector drawables, of course), but I only covered
the providers that appeared on my device (I can't confirm that there are no
other providers), so if the provider is covered, it will have an icon, but if
I didn't cover it, it will appear just fine, but without an icon. Maybe we
could move the settings and sorting to the navigation drawer, also? It would
be cleaner and more elegant, IMO.
Change-Id: I02a051f0b75c6d4e16f518aa19fb9c6eef00f5e4
Reviewed-on: https://gerrit.libreoffice.org/32881
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: jan iversen <jani@documentfoundation.org>
Project: help e8625126b630646ca0f490e1f21ef97015bce2a3
tdf#105354 Do not link Macro Selector window to wrong page
With this part dropped (it doesn't make any sense)
the Macro Selector windows Help button does not display
the Customize - Menu tabs help page anymore
Change-Id: I2b74d4e920ee0e6b2172842406e29bbbecc329e5
Reviewed-on: https://gerrit.libreoffice.org/33106
Reviewed-by: Gabor Kelemen <kelemeng@ubuntu.com>
Tested-by: Gabor Kelemen <kelemeng@ubuntu.com>
This patch allows processing a number even if a comment is the
first item in the string. Previously the result was formula
cells displaying an error message.
Only fields were handled for annotation anchors. Allow other
non-word-breaking markers to be checked for number validity.
Change-Id: I13b3566bf93f0029a28bada470cdd4196271bccd
Reviewed-on: https://gerrit.libreoffice.org/32949
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Justin Luth <justin_luth@sil.org>
The UNO command of default formatting in Draw is .uno:SetDefault
instead of .uno:ResetAttributes.
Change-Id: I9fea2cb6b89f7869cfcd40fa6e1768952563a4f6
Reviewed-on: https://gerrit.libreoffice.org/33085
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Mark Hung <marklh9@gmail.com>
e.g. fdo80788-1.odt
Invalid read of size 2
at 0x3C031438: SwLinePortion::IsParaPortion() const (porlin.hxx:132)
by 0x3C05EC56: SwTextFormatter::BuildPortions(SwTextFormatInfo&) (itrform2.cxx:679)
by 0x3C05F2A2: SwTextFormatter::FormatLine(int) (itrform2.cxx:1550)
Address 0x3576189e is 30 bytes inside a block of size 40 free'd
at 0x4C2ED4A: free (vg_replace_malloc.c:530)
by 0x4E56939: rtl_freeMemory_SYSTEM(void*) (alloc_global.cxx:279)
by 0x4E56978: rtl_freeMemory (alloc_global.cxx:349)
by 0x4E5501B: rtl_cache_free (alloc_cache.cxx:1230)
by 0xDACE741: FixedMemPool::Free(void*) (mempool.cxx:49)
by 0x3C043AB8: SwTextPortion::operator delete(void*, unsigned long) (portxt.hxx:55)
by 0x3C043B05: SwHangingPortion::~SwHangingPortion() (porrst.hxx:98)
by 0x3C043F47: std::default_delete<SwHangingPortion>::operator()(SwHangingPortion*) const (unique_ptr.h:76)
by 0x3C044070: std::unique_ptr<SwHangingPortion, std::default_delete<SwHangingPortion> >::reset(SwHangingPortion*) (unique_ptr.h:344)
by 0x3C0888C3: std::unique_ptr<SwHangingPortion, std::default_delete<SwHangingPortion> >::operator=(decltype(nullptr)) (unique_ptr.h:280)
by 0x3C0888DB: SwTextGuess::ClearHangingPortion() (guess.hxx:51)
since...
commit a706bb06d5326e7a997dd1d71f3f947bd0ad28e6
Date: Wed Jan 11 14:26:47 2017 +0200
new loplugin: useuniqueptr: sw part 1
Change-Id: I614029474d684ddcccd4f4a5e9787fe6c19d8fd2
For whatever reason we query printer output device for fonts and for
whatever reason GDI will only report fonts from the printer which
often in formats we want to ignore. This results in aborting the
application when we reject all fonts on the printer!
Just special case printer output devices and hope for the best.
I think this commit is partially blamed as it is evidenced now that
there was a good reason to only abort for window output devices, but
I’m leaving this as it is for now:
commit ea7102c93c33884a68d4dba0de7d52f8ed4f4579
Author: Chris Sherlock <chris.sherlock79@gmail.com>
Date: Tue Mar 25 21:54:20 2014 +1100
fdo#74702 OutputDevice::ImplInitFontList() fails if no fonts on device
It makes no sense that ImplInitFontList() only fails for Window
instances. I have carefully checked all the functions that use
this function, and there are no good cases when no fonts won't cause
problems. In fact, we have a number of functions that specifically
rely on the fact that ImplInitFontList will populate
OutputDevice::mpFontCollection with at least one font.
Therefore, I'm making this abort if it can't populate the collection,
regardless of whether it is a Window, Printer or VirtualDevice.
I have also refactored GetDefaultDevice - I now check the default
pOutDev parameter to see if it is NULL (the default), in which case
it is referring to the default window, so I call on
Application::GetDefaultDevice() instead of going straight to the
pimpl data structure used by the Application class.
Reviewed-on: https://gerrit.libreoffice.org/33014
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
(cherry picked from commit 88d6675a40034fa786efd75cf14175f3c0c83e2e)
Change-Id: I3f34ba0cea9f88966ace3a5ef7fac868972ea5ac
Reviewed-on: https://gerrit.libreoffice.org/33087
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Khaled Hosny <khaledhosny@eglug.org>
