It appears, upon inspection, that the definition of a 'word' that "wc" uses is any sequence of non-space characters delimited by a file or white-space delimiter.
I.e. it doesn't include what some might consider to be, "standard", (varying, depending on context), characters that could include: period, comma, other punctuation or even "NULL". The only characters that separate words in wc are the set (\f \n \r \t \v <SPACE>). This is fine and perfect for me, it's just that it isn't documented explicitly in the wc manpage (or any place I remember reading it) and necessitated me looking in the source to get an answer. Could a line as simple as "Words in 'wc' are delimited strictly by whitespace", or include the characters (\f \n \r \t \v and ' ' (space))? I.e. can a bug be opened to ensure that change gets in?
