* liborcus-0.19.2-gcc15-cstdint.patch
- Fix source tarball URL.
- Upgrade to 0.20.2.
- Changes from 0.20.2:
* The base JSON parser class (orcus::json_parser) now allows primitive string,
numeric and boolean values to be document root as it should per JSON
specification.
* added to the command-line programs proper support for Unicode filepath
handling on Windows which still uses UTF-16 with code pages. Relevant
public API has been updated to add support for Unicode filepaths on Windows.
* added to the import_sheet interface a variant of set_string() that takes a
non-indexed string value. This is used to handle cells with inline strings
in the xlsx import filter. However, orcus's own document store backend does
not yet handle cells with non-indexed strings, so when importing xlsx
documents containing cells with inline strings, those cells will be ignored
for now.
* added to the orcus::format_t enum type json and xml as its members. With
that change, you can now instantiate import filter adapters for these two
new format types to import generic JSON and XML documents. These two
adapter filters will auto-detect tabular structures and map them to sheets
if they exist.
* orcus::detect() can now detect generic JSON and XML documents.
* fixed a bug where orcus-xml generated incorrect map file in map-gen mode
when the source document contained multiple tabular structures.
- Changes from 0.20.1:
* fixed an assertion failure when importing a document in Excel 2003 XML
format containing an autofilter with no filtering applied to it.
* added a set of new interfaces to support importing pivot table definitions
from xlsx documents.
OBS-URL: https://build.opensuse.org/package/show/LibreOffice:Factory/liborcus?expand=0&rev=116
- On 32bit x86 architectures, compile with option -ffloat-store. This
mitigates the unfortunate fact that the test-suite of the package
expects the precision of FP operations to be lower than that of
internal representation of 80387.
The specific issue this works around popped up on i586 with GCC 14 but
is an inherent issue of the test suite of the package.
If the request is OK, please forward it to Factory soon too so that we
can switch the default compiler. Thanks!
OBS-URL: https://build.opensuse.org/request/show/1191898
OBS-URL: https://build.opensuse.org/package/show/LibreOffice:Factory/liborcus?expand=0&rev=112
- Update to orcus 0.19.2:
* fixed a build issue with gcc 14 due to a missing include for std::find_if
and std::for_each.
* fixed a segmentation fault with the orcus-test-xml-mapped test which
manifested on hppa hardware, as originally reported on
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1054376.
* fixed a crash when loading a document that includes a style record
referencing an unnamed style record as its parent. In Excel-generated
documents, styles only reference named styles as their parents. But in
3rd-party generated documents, styles referencing unnamed styles as their
parents can occur.
* fixed a crash when the document model returned a null pointer when a
reference resolver interface was requested.
- Update to orcus 0.19.1:
* implemented orcus::create_filter() which instantiates a filter object of
specified type. The returned object is of type
orcus::iface::import_filter.
* moved test cases for format detection to the respective filter test files.
* fixed a bug where the import filter did not set the formula grammer prior
to importing.
- Update to orcus 0.19.0:
* added support for allowing use of std::filesystem,
std::experimental::filesystem or boost::filesystem per build
configuration.
* refactored styles import to use style indices returned by the document
model implementer rather than using the indices stored in the file. This
allows the implementer to aggregate some style records and re-use the same
index for records that are stored as different records in the original
file.
* fixed a bug where column styles were not applied to the correct columns
OBS-URL: https://build.opensuse.org/request/show/1152882
OBS-URL: https://build.opensuse.org/package/show/LibreOffice:Factory/liborcus?expand=0&rev=108
- Update to 0.18.1:
* sax parser:
* added support for optionally skipping multiple BOM's in the beginning of
XML stream. This affects all XML-based file format filters such as
xls-xml (aka Excel 2003 XML).
* xml-map:
* fixed a bug where an XML document consisting of simple single-column
records were not properly converted to sheet data.
* xls-xml:
* fixed a bug where the filter would always pass border color even when it
was not set.
* buildsystem:
* added new configure switches --without-benchmark and --without-doc-example
to optinally skip building of these two directories.
- Remove patch gcc13-fix.patch
OBS-URL: https://build.opensuse.org/request/show/1096006
OBS-URL: https://build.opensuse.org/package/show/LibreOffice:Factory/liborcus?expand=0&rev=91
- Update to 0.17.2:
* fixed a bug where the state of style:cell-protect="none" was not
explicitly pushed, thereby having had the same effect as not having this
attribute. After the fix, style:cell-protect="none" will explicitly push
the hidden state to false, locked state to false, and the formula-hidden
state to false.
- Update to 0.17.1:
* addressed a number of coverity issues.
* removed a variety of compiler warnings.
* re-generated sax parser tokens from ODF v1.3.
* revised the style import code to only push style attributes that are
actually specified in the XML.
* revised the XML structure validation strategy to ignore any mis-placed
elements and their sub structures rather than aborting the import.
OBS-URL: https://build.opensuse.org/request/show/949289
OBS-URL: https://build.opensuse.org/package/show/LibreOffice:Factory/liborcus?expand=0&rev=85
- Update to 0.17.0:
* set the baseline C++ version to 17.
* cleaned up the public API to replace pstring with std::string_view, union
with std::variant, and boost::optional with std::optional. With this
change, the public API no longer has dependency on boost.
* switched to using ixion::model_iterator for horizontal iteration of cells
instead of using mdds::mtv::collection.
* fixed a bug where exporting a spreadsheet document containing adjacent
merged cells regions to html incorrectly exported the merged cell areas.
* cached cell values are now correctly loaded from the xlsx file.
* utf-8 names are now allowed as element and attribute names in the sax parser.
* unquoted utf-8 property values are now allowed in the css parser.
* added yaml output option in orcus-json.
* fixed a bug where mapping of an XML document with namespace aliases
sometimes corrupts the alias values.
* added orcus.FormulaTokenOp enum type in python which describes type formula token
operator types in a more finer grained manner.
* added notes to how to use orcus-xml and orcus-json to map XML and JSON
documents to spreadsheet documents.
- Drop GCC11_build_fixes.patch
OBS-URL: https://build.opensuse.org/request/show/932479
OBS-URL: https://build.opensuse.org/package/show/LibreOffice:Factory/liborcus?expand=0&rev=84
