Stephane Chazelas <stephane.chaze...@gmail.com> wrote, on 22 Aug 2019: > > 2019-08-22 09:50:56 +0100, Geoff Clare: > [...] > > So how about: > > > > If a returned ASSIGNMENT_WORD token begins with a valid name, > > assignment of the value after the first <equals-sign> to the name > > shall occur as specified in [xref to 2.9.1]. If a returned > > ASSIGNMENT_WORD token does not begin with a valid name, the way > > in which the token is processed is unspecified. > [...] > > But as seen in my latest update, > > $varname=value > > should not be (required to be) considered an "ASSIGNMENT_WORD", > as that would mean (with the default value of $IFS and with > bugid:351 applied) that > > varname=foo > value='a b' > export "$varname"=$value > > would be either unspecified or specified to not split $value. > > While all shells interpret it as > > export foo=a b
So propose a change to the part of rule 7 that allows these to be returned as ASSIGNMENT_WORD, so that it requires WORD instead. It doesn't affect the wording I proposed above, which only applies after the decision to return ASSIGNMENT_WORD has been made. -- Geoff Clare <g.cl...@opengroup.org> The Open Group, Apex Plaza, Forbury Road, Reading, RG1 1AX, England