LabVIEW Idea Exchange
(I thought something similar was suggested before, but I cannot find it).
It is often necessary to deal with "malformed" spreadsheet strings, where there is more than one possible delimiter.
One example is discussed here, where a delimiter is either ",", "N", or a combination of the two, but the same has come up before. In my typical case, users sometimes use notepad to edit spreadsheet files, randomly replacing tabs with multiple spaces, multiple tabs, or a combination of tabs and spaces, so it "looks" aligned, but spreadsheet string to array will choke on it.
"Scan strings for tokens" has that useful feature to allow the definition of multiple delimiters as an array and (by default) consecutive delimiters are contracted into one (option "allow empty tokens?"=false). This makes it easy to parse lines (delimiters: \n, \r) or fields (delimiters: \t, \s). An example to deal with dirty strings is given here.
"Spreadsheet string to array" already accepts arrays of delimiters, but they are treated as alternatives and multiple consecutive delimiters are NOT contracted into one. I suggest to add an option to "Spreadsheet string to array" where multiple consecutive delimiters will be contracted into one for parsing. (or it could be named "allow empty tokens?" as above for simplicity, default=true to keep current behavior)
For example if I wire a delimiter array containg a single space element, it would treat multiple spaces as one.
If the delimiter array contains a tab and a space, consecutive spaces and tabs (in any order) will act as a single delimiter. The above quoted example would have a delimiter array containing two elements, "N" and ",".
Suggestion: add a boolean input to treat multiple consecutive delimiters as a single delimiter. For example, the two code fragments in the image should give basically the same result (The upper part is slightly defective and would need a little bit more code if lines end in a delimiter, but I think the idea should be clear).
The suggestion is at the bottom:
