Thank you for that explanation. And it makes absolute sense when viewing it from the XML-like underlying model with the XPath-like filtering language. In XML terms, what I did, is put two attributes of the same name with different values on an XML element - which does not work in XML, either.
My use case is that I want to attach the numbers (IDs) of documents that are affected by a bug described in the text, like
hyperlinks don't work @doc(21378) @doc(6729)
meaning the documents 21378 and 6729 are affected by that bug. I would then be able to search either for all bugs that manifest in e.g. document # 6729 by clicking on the value 6729, or see a list of all bugs where there is a sample document for by clicking on @doc.
XML has the concept of a value list in attributes (corresponding to FT's tags), written by space-separating the values. In FT, that would look something like
hyperlinks don't work @doc(21378 6729)
which I know currently does not do what I want (and probably never will).
But knowing now what the technical background is, the correct way to achieve what I want is indeed as you suggest: using different tags. So I would use
hyperlinks don't work @doc @d_21378 @d_6729
which would allow me to:
- find all lines having document numbers attached by filtering on the generic @doc tag (indicating that fact, and yes, I'd have to do the housekeeping on this myself), and
- search for all lines that affect e.g. document 6729 by searching for the @d_6729 tag
Thanks again for making me understand this aspect, and suggesting the solution. Highly appreciated!