Purpose
A single module that encapsulates all the code necessary to serialize and deserialize enumerable class objects into Excel and CSV file formats.

Nomenclature: Why 'Serialization'?
Serialization is the process of converting an object into a stream of bytes to store the object or transmit it to memory, a database, or a file. Its main purpose is to save the state of an object in order to be able to recreate it when needed. The reverse process is called deserialization.
See: https://docs.microsoft.com/en-us/dotnet/csharp/programming-guide/concepts/serialization

Features

• Multiple enumerable class objects to/from Excel worksheets or CSV pages. Class object properties mirror columns.
• No limits to number of records serialized (records over 1M will overflow into successive worksheets).
• Performant: 2M records takes 2½ minutes to serialize. Minimal memory footprint.
• Localization: Automatically translate worksheet tabs, column headings and values into the target language.
• A single-page CSV format file may easily be imported into Excel with data formatting (partially) intact. Note: In CSV there is no limit to the number of records, but if upon import into Excel, the number records exceeds 1 million, Excel import will fail. The up-side is CSV serialization/deserialization is extremely fast.
• Excel column naming, translating, and formatting is determined by column descriptors (aka property attributes). Nothing is hard coded into this module. See attributes XlColumn and XlIgnore.
• Excel worksheet tab naming and translation is determined by a class attribute XlSheet.
• Optional pretranslated worksheet heading. May contain multiple lines.
• Multiple choice objects such as booleans and enums are handled as translatable dropdowns in the spreadsheet.

Excel example worksheet with header

• Column order is implicitly defined by the order of properties in the class object.
• A worksheet heading was defined and may contain multiple lines.
• Some but not all the columns were translatable as those headings were not in the available in the caller's string resources. It then defaults to the class property names.

Excel example worksheet without header

• Column auto-fit is determined by the maximum width of any single word in the column heading and the maximum width of the data.
• The 'My Bool' and 'My Enum' column data was marked as translatable.
• Multiple choice columns (enum and bool) are restricted to the known possible values.
• The drop down on the first column was enabled by the class property column descriptor.
• All data columns were formatted according to the associated class property column descriptor.
• The optional divider line is determined by a null class object in the enumerable array (see extension SplitBy()).

Build Notes
Build frameworks are .NET Framework 4.5 and .NET Standard 2.1. These two frameworks were used because these are the latest frameworks used by the dependent packages. The Excel reader/writer (EPPlus) package version is constrained to 4.5.3.3 as this is the last free version. Subsequent EPPlus versions require a license. This package version has all the functionality needed for CSVExcelExportImport. CSVExcelExportImport was built and tested with Visual Studio 2019.