If there is a suggested rotation value for the axis,
it might be better to ignore the autorotation bool.
A better approach would be to improve our autorotation algo.
The gotcha here is that if we get ahead of ourselves, and read to
the end of the stream, we detect bad chains too early, so instead
incrementally build the page chain cache, which is also quicker
and behaves more similarly to the previous code.
The recent addition of Hebrew line breaking rules apparently does not
work with genbrk from ICU 4.6. Maybe we should just require 4.x (where x
is the first version where it is implemented) as minimum ICU version?
IMHO it would be better than this insane (and fragile) filtering...
Change-Id: Ia2adad3621fa7a2f319a07fb569b1f1c1eb6db7c
This fixes problems with unique value/ duplicate value entries. The old
falt copy process did not respect that the copy needs to point to its
own range and therefore the copy was totally useless.
This is also necessary for several future new conditional formats like:
top n%, top n elements, above average, ...
Change-Id: I51b868968812a4e00dca9af2aa51d50b5ef8b855
* use operator && instead of 'and'
* use correct types
* long operator -(Date,Date)
* sal_uInt16 GetDay(), GetMonth(), GetYear()
* check for error after argument pop and bail out if so
* day difference can be pushed and return immediately without further
calculation
* ensure only one value is pushed
* correct calculation of years and months
* only complete years and months are to be returned, same day,month in
different years is a complete year (birthday-like), same day in different
months is a complete month.
* "y" returned 1 for 2012-02-29,2013-02-28 or 2012-03-29,2013-03-28
* must be 0 instead
* "y" returned 0 for 2011-03-28,2013-03-29
* must be 1 instead
* negative values are never returned
* reversed arguments are not allowed
* algorithm for "md" extracted from Excel behavior, see source
* changes to other intervals as well
* changed term Format to Interval in FormulaWizard (and code) for clarity
* mention interval arguments in FormulaWizard
* in sc/source/filter/oox/formulabase.cxx reverted the entry's move from
saFuncTableBiff5 to saFuncTableOdf
* saFuncTableOdf member's are "Functions defined by OpenFormula, but not
supported by Calc or by Excel", this function now is supported by both
* instead, changed FUNCFLAG_IMPORTONLY to 0 in entry of saFuncTableBiff5 and
added FunctionData::mpcOdfFuncName "DATEDIF"
* otherwise with the FUNCFLAG_MACROCALLODF a macro call is generated for
Excel export that Excel doesn't know and can only be imported by LibO
again, and without the BIFF function identifier the function could not be
imported from original binary Excel documents
seems the situation is that we link to the mingw32-system-icu but build the
bundled icu natively in order to use its tools at build time. Which means
we need the
ICU_RECLASSIFIED_CLOSE_PARENTHESIS
ICU_RECLASSIFIED_PREPEND_SET_EMPTY
defaults to match the internal icu defaults
Change-Id: I09a693d795b956789170cc65a582a4367366acaa
ScConditionalFormat::pRanges must be updated otherwise the duplicate
value check might use the wrong range
Change-Id: I3df7d54cdcd9c8863d5fbd89eff3be83bb73f2a5
Frames with text direction Vertical RTL have their borders mangled while
painting, everything is rotated 90 degrees and so the model "left" border
ends up being painted as top border etc. Clearly the new drawing layer
based painting code is very surprised by this and ends up painting only
points instead of lines.
(regression from 0f0896c26f)
The way too smart ctor for the DND handler started drag immediately,
causing a race condition that could recurse to setting a handler
again before the first one was actually set, thus immediately again
causing the DND to be stopped, and then possibly later again started,
depending on how the race condition turned out. Use delayed initialization
to avoid this.
Change-Id: I528eddbdc7c52a19675997e4c866506c662cff19
