I want to be able to write a CSV text file than can be imported into Excel.
The tricky thing is that the encoding of the data of the file is in Shift-JIS.
If I try to import the file into Excel with a CSV extension, it is imported OK as CSV, but the character encoding for the datis not correct.

To be able to read the Shift-JIS encoding, I rename the file to end in .txt and select 'Import ... Shift-JIS encoding ... Comma Delimited'.
Importing this way basically works, but if any of the cells have line feeds in them then those lines are interpreted as the start of a new row in the spreadsheet.

I found that the problem isn't because of Shift-JIS.
There seems to be a problem in general with importing text files with cell data that have embedded line feeds.

For example, if the following data is imported as text, there will be two rows instead of one:

"11234
5678",99999

Any way around this?