On Thu, Apr 2, 2015 at 3:31 PM, Thomas 'PointedEars' Lahn <pointede...@web.de> wrote: > Ian Kelly wrote: > >> […] Thomas 'PointedEars' Lahn […] wrote: >>> Ian Kelly wrote: >>>> Within a grammar, the question of "is an X a Y" is nonsensical in >>>> isolation. It can only be answered in relation to a parse tree. >>>> Consider the simple grammar: >>>> >>>> S -> A | B >>>> A -> x >>>> B -> x >>>> >>>> Is x an A? It depends. >>> >>> No, by the definition 2 below, that we all accepted implicitly up to this >>> point, x is *definitely* an A. >> >> What gives you the impression that I ever accepted it? > > ,-<news:mailman.181.1427346636.10327.python-l...@python.org> > | > | What the grammar that you quoted from shows is that STRING+ is an > | expression. > > There is *no way* for you to make that statement if you did not accept > definition (2).
Actually, there is a very simple way: I was being sloppily imprecise when I wrote that. First, I was speaking only in reference to the class of parse trees with more than one STRING, as that was the topic under discussion. I should have been clearer about that. Second, I never should have used the term "STRING+" there (or anywhere else in this discussion), as that merely clouded the point I was trying to offer. What I *meant* was that the complete sequence of produced STRINGs -- as opposed to any individual STRING -- is an expression, and I inappropriately used "STRING+" to denote that. (Maybe that is the point you were trying to make when you were talking about EBNF before.) >> This question of whether "x is an A" is informal and not a topic of formal >> language theory so far as I'm aware. Can you cite some source for it? > > No, because I was formalizing the ad-hoc definition by Chris Angelico in > <news:mailman.51.1426995416.10327.python-l...@python.org>. That URI is not useful to me as I don't use a newsreader. Is that the message dated Sun, 22 Mar 2015 14:36:48 +1100? I don't see any suggestion of this definition in that post. t>>> Now, according to these definitions, in the offered grammar x is *both* >>> an A and a B. Because what matters is _not_ the practical result of >>> production chains (the actual parse tree), but the certainty of the >>> theoretical possibility of it. >> >> This strikes me as being a lot like arguing, "some kites are toys, and >> some kites are birds; therefore, all kites are both toys and birds." > > False analogy again. We are discussing *in theory* a *formal* grammar. Its > goal symbols have *no meaning* except what can be produced from them. We're discussing a formal grammar as if it described a classification hierarchy, which gives meaning to statements like "A STRING is an expr". Otherwise, that statement is meaningless and can be neither true nor false. So I think that the analogy between one such hierarchy and another is apt. >> You're really going to make me spell it out, aren't you? Fine, here you >> go. >> >> single_input -> […] -> expr -> […] -> atom -> STRING STRING >> >> Note: the derivation contains exactly one expr node, which indirectly >> produces both STRINGs. Neither STRING in this derivation is >> individually produced from the expr. > > So you have proven that which nobody ever doubted nor requested, but I > pointed out already. As I tried to point out several posts back when I suggested that we were in agreement, and which you flatly denied. > What you have still not proven is what you claimed: > the parse tree. Because you misunderstood my claim. -- https://mail.python.org/mailman/listinfo/python-list
