This is a very minor thing, but I thought I'd report this nevertheless… Current section 7.3.1 of the PMS states that EAPI parsing was done using the regexp “^[ \t]*EAPI=(['"]?)([A-Za-z0-9+_.-]*)\1[ \t]*([ \t]#.*)?$” and subsequently states that “the EAPI is the substring matched by the capturing parentheses”. However, as you can see, there are TWO pairs of capturing patentheses. So perhaps you should change this to “second capture group” or something like that.
In principle you're right. However, PMS has always assumed "a sensible and cooperative reader" (cf. 2008-08-28 council meeting), and for such a reader it should be clear which of the subexpressions is meant here. So, I tend to close this. Other opinions?
Honestly, I'd prefer seeing BNF-like grammar there rather than regexps. Spec should provide a formal definition of what needs to be done, rather than implementation.
(In reply to comment #2) > Honestly, I'd prefer seeing BNF-like grammar there rather than regexps. > Spec should provide a formal definition of what needs to be done, rather > than implementation. Then why didn't you bring this up before the current wording was approved by the council? The topic was really extensively discussed both on gentoo-pms and gentoo-dev mailing lists (and you even participated in that discussion). Closing.
(In reply to comment #3) > (In reply to comment #2) > > Honestly, I'd prefer seeing BNF-like grammar there rather than regexps. > > Spec should provide a formal definition of what needs to be done, rather > > than implementation. > > Then why didn't you bring this up before the current wording was approved by > the council? The topic was really extensively discussed both on gentoo-pms > and gentoo-dev mailing lists (and you even participated in that discussion). I'd guess the right answer would be: because I'm not brilliant enough to come up with all the good ideas on time.
