On Wed, Sep 12, 2012 at 4:01 PM, David Korn <d...@research.att.com> wrote: > cc: iszczesn...@gmail.com > Subject: Re: [ast-developers] Depreciating typeset -l and -u in favor of > typeset -M tolower and -M toupper? > -------- > >> Are there any objections to depreciate typeset -l and typeset -u in >> favor of typeset -M tolower and typeset -M toupper? The goal would be >> to map these two options to their -M counterparts and let ksh93 -n >> complain about their usage. IMO it should save same code space in the >> shell core code and would promote more usage of typeset -M >> >> Irek >> > > Here are reasons not to do this: > 1. -l and -u are widely used and already behave as -M tolower and > -M toupper. > 2. It is hard to detect their use while parsing since -l can be used > with other typeset options, for example, -i and -E. > 3. The amount of code space is minimal since they are mapped to > -M tolower and -M toupper with a few lines of code.
Well, my point is that having extra options only for tolower and toupper is a very English-centric point of view. Other languages have more transformations, like tojkata and tojhira for Japanese (Cedric can elaborate, he brought the subject of wctrans() up in the Illumos OS lists) or to_inpunct in hi_IN.UTF-8. It is IMO not fair and balanced for an internationalized application like ksh to apply special treatment for the English language while a more generalized solution is available and should be preferred. Irek _______________________________________________ ast-developers mailing list ast-developers@research.att.com https://mailman.research.att.com/mailman/listinfo/ast-developers
