About the grammar

[Martin v. Loewis]

> Does any one knows why the grammar is so coded? Any intuition?

The 2.7 Grammar clarifies that:

# The reason that keywords are test nodes instead of NAME is that using
# NAME results in an ambiguity. ast.c makes sure it's a NAME.
argument: test [comp_for] | test '=' test

The ambiguity is this: if I have

foo(a, b = 2)

and have parsed it up to

foo(a,

then, if I get "b", should I enter the "test" clause (for the left
alternative) or the right clause? With the grammar being as stated,
it can be parsed as

argument: test ([comp_for] | '=' test)

so that parsing always starts with a test. Afterwards, I either get a
'=', indicating a keyword argument, or not: the '=' is not in the FIRST
set of comp_for, so there is no ambiguity here.

HTH,
Martin