Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
|
|
When checking for svSingleRef or svDoubleRef, also check for
svExternalSingleRef or svExternalDoubleRef. Otherwise the dummy reference is
treated.
|
|
|
|
|
|
after my last rework of the copy paste code my old assumptions for the
copy paste code are no longer valid
|
|
|
|
we now behave nearly the same way as excel does
|
|
|
|
Handle Move Sheet correctly
|
|
support copying from an unsaved document and some improvements
still problems with MoveTab and missing dialog box if we have external
refs to unsaved docs while we save
|
|
|
|
adjust named ranges formulas too
little problem here:
tab absolute and col/row relative is tricky, the col/row is calculated
relative to the cell where the range name is used but we don't know in
which cells the range name is used
The other office solves that by making only references pointing to
sheets that are not copied external (this solution is not perfect)
|
|
|
|
More tests might be affected. Also unclear to me whether it is always (only)
localedata_en, or whether it could also be _es/_euro/_others.
|
|
|
|
add a typo correction on xnamedranges.cxx
|
|
Copying absolute sheet references between different documents now transforms absolute sheet
references into external references if the refered range is not copied.
Working:
References to already saved documents work already with:
-only transform reference into an external reference if range is not in
the copied range
-don't transform relative references
Missing:
handling for copying from not yet saved files
idea:
- add handling of external references to open documents that are not yet
saved through the temporary document name
- transform the url to the correct file url as soon as the source document is
saved
- if target document is saved first show warning message to the user and
save with temporary name as file url
|
|
|
|
|
|
|
|
Uses info from comments in i18728.
|
|
renaming previous test to XRangeNames
new test for XRangeName
|
|
|
|
|
|
* Reintroduced logic in PrepareQueryItem::operator() that got lost during
rework of ScQueryEntry.
* ScQueryParamBase::FillInExcelSyntax() and the calling
lcl_createExcelQuery() create query entries always as ByString, that
needs to be converted to ByValue if appropriate.
|
|
|
|
|
|
+ Checks out-of-bounds accesses in
ScFormatRangeStyles::GetStyleNameIndex() and
ScRowFormatRanges::AddRange() and prevents crashes.
- The real cause seems to be some style row/repeat miscalculation
elsewhere, further investigation would be necessary.
|
|
|
|
|
|
fix potential bug on short read which has bugged me for a while
where bytes read is always reduced without a check that byte was
read
|
|
Nobody ever checked the return value anyway, so just return the string
and use the stream state bits if necessary to find failures.
Doesn't need to be a member, make a standalone function
Rename it to read_zeroTerminated_uInt8s_AsO[U]String, stupid perhaps,
but *shrug*, unambiguous.
Drop misleading overloaded String variants use:
read_zeroTerminated_uInt8s_AsOString or
read_zeroTerminated_uInt8s_AsOUString
added a unit test, valgrinded it, found and fixed invalid read
in original implementation.
|
|
|
|
Resolved scary "ScToken::GetDoubleRef: virtual dummy called" debug message.
|
|
|
|
|
|
|
|
The breakage was due to the drawing layer re-work. We do need to store
charts (or all OLE shapes with associated listening ranges) for later
registration regardless of anchor types.
|
|
Or else filtering of field members will not work correctly when the
field contains an empty name.
|
|
one test disabled because we first need to write a patch for our
internal cppunit
|
|
|
|
|
|
|
|
|
|
|
|
This eliminates the need to store the assigned doc pointer to cross-check
against whenever the edit engine is initialized. It's cleaner this way
& fdo#43614 still remains fixed.
|
|
There should only be one active edit view while the user is typing
into a cell, or else it crashes when an external input method is
being used to input values (such as Japanese).
|