Bug 60352

We use XSSFExcelExtractor as a mechanism to extract the complete text in an Excel file, however it seems that in certain circumstances the "null" value is extracted from an empty cell.

For example:

> Breakdown of data generated by project, technology, submitting centre	null	> null	null	null	null	null
> null	null	null	null	null	null	null
> null	Abbreviation Definitions	null	null	null	null	null
> null	Platform	Definition	null	null	null	null
> null	LS454	454 Roche Genome Sequencer FLX System	null	null	null	


The patch is relatively simple (and I'm happy to create a PR for it on GitHub). All we need to is to wrap the last two lines in XSSFExcelExtractor.handleNonStringCell() with a null check

>         if (contents != null) {
>             checkMaxTextSize(text, contents);
>             text.append(contents);
>         }

This would then perform as expected and extract this text instead.

> Breakdown of data generated by project, technology, submitting centre						
> 						
> 	Abbreviation Definitions					
> 	Platform	Definition				
> 	LS454	454 Roche Genome Sequencer FLX System

We believe that an empty string is the preferred option here, because the text "null" itself might be used as cell contents in certain cases. In that situation it's difficult to discriminate between these occurrences (is it the text "null" or is the cell empty?)

Looking forward to hearing your thoughts.