Hello, On Wed, May 25, 2005 at 01:40:22PM -0700, Paul Eggert wrote: > True, but it still bothers me a bit. (I don't know whether the "w" > stands for "word" or "white space". :-)
actually, I had both in mind and I liked the ambiguity. Perhaps you can find a better name for the macro? > Also, will we need "_w" variants for other macros? If so, I'd rather > not maintain/describe pairs of macros, and it'd be better to factor > out only the m4_split(m4_normalize([$1])) idiom rather than the > m4_foreach(..., m4_split(m4_normalize([$1])), ...) idiom. There is no occurence of m4_split(m4_normalize([$1])) which would be outside m4_foreach, so I don't think the danger is big. So: - We agree that AC_FOREACH should be obsoleted, and occurrences expanded. - Then there is the question whether we should factor out m4_foreach_w. I still think it's good to factor, to increse readability, but I'm no longer so sure about it--I see your point, too. Have a nice day, Stepan
