- Removed
ActiveCellfrom the selection pane when writing a frozen header row; it does not seem necessary and creates a corrupt worksheet unless the entire row sequence is specified.
- Fixed a bug in which
WorksheetReaderandWorksheetWriterwere still not converting between adoubleand mapped numeric types.
- Fixed
ShouldFreezeHeaderRowwhen using a non-default header, and added a better test case to capture the scenario.
- Continued refactoring test cases to automate testing of additional scenarios.
- Added automated tests to validate and read worksheets generated by OpenSpreadsheet and subsequently saved by Excel, as Excel automatically modifies XML and causes unexpected errors.
- Renamed
ShouldFreezeTopRowworksheet style property toShouldFreezeHeaderRow. ShouldAutoFilternow acts on the user-specified header row, rather than the first row.- Header styles are now cached regardless of the value of
ShouldWriteHeaderRow, which will allow for greater precision when using an explicitWorksheetWriter.
- Changed
WorksheetReaderto handle numbers stored in scientific notation.
- Began refactoring test cases to automate common scenarios rather than relying on file types and manual scanning.
- Null cell data types now default to numeric during reading (Excel apparently removes this attribute).
- Trying to pass Color.Transparent value to various spreadsheet style items will now throw an
ArgumentException.
- Added
Accountingnumbering format.
WorksheetReaderwill no longer try to set properties without a public setter.- Fixed bug that skipped constants when reading.
- Fixed a bug that duplicated style definitions when reading.
- Added
ReadUsingandWriteUsingmethods (and associated tests) to allow for custom logic to be applied to read and write maps. - Added
ReaderRowclass to handle read operations (used forReadUsing). - Created more accurate tests for
Defaultconfiguration methods.
Initial release.