also handles special macros like '__LINE__'.
Although these macros expand to a single token which cannot contain any
further macros, for reasons of token spacing (*note Token Spacing::) and
simplicity of implementation, cpplib handles these special macros by
pushing a context containing just that one token.

   The final thing that 'enter_macro_context' does before returning is
to mark the macro disabled for expansion (except for special macros like
'__TIME__').  The macro is re-enabled when its context is later popped
from the context stack, as described above.  This strict ordering
ensures that a macro is disabled whilst its expansion is being scanned,
but that it is _not_ disabled whilst any arguments to it are being
expanded.

Scanning the replacement list for macros to expand
==================================================

The C standard states that, after any parameters have been replaced with
their possibly-expanded arguments, the replacement list is scanned for
nested macros.  Further, any identifiers in the replacement list that
are not expanded during this scan are never again eligible for expansion
in the future, if the reason they were not expanded is that the macro in
question was disabled.

   Clearly this latter condition can only apply to tokens resulting from
argument pre-expansion.  Other tokens never have an opportunity to be
re-tested for expansion.  It is possible for identifiers that are
function-like macros to not expand initially but to expand during a
later scan.  This occurs when the identifier is the last token of an
argument (and therefore originally followed by a comma or a closing
parenthesis in its macro's argument list), and when it replaces its
parameter in the macro's replacement list, the subsequent token happens
to be an opening parenthesis (itself possibly the first token of an
argument).

   It is important to note that when cpplib reads the last token of a
given context, that context still remains on the stack.  Only when
looking for the _next_ token do we pop it off the stack and drop to a
lower context.  This makes backing up by one token easy, but more
importantly ensures that the macro corresponding to the current context
is still disabled when we are considering the last token of its
replacement list for expansion (or indeed expanding it).  As an example,
which illustrates many of the points above, consider

     #define foo(x) bar x
     foo(foo) (2)

which fully expands to 'bar foo (2)'.  During pre-expansion of the
argument, 'foo' does not expand even though the macro is enabled, since
it has no following parenthesis [pre-expansion of an argument only uses
tokens from that argument; it cannot take tokens from whatever follows
the macro invocation].  This still leaves the argument token 'foo'
eligible for future expansion.  Then, when re-scanning after argument
replacement, the token 'foo' is rejected for expansion, and marked
ineligible for future expansion, since the macro is now disabled.  It is
disabled because the replacement list 'bar foo' of the macro is still on
the context stack.

   If instead the algorithm looked for an opening parenthesis first and
then tested whether the macro were disabled it would be subtly wrong.
In the example above, the replacement list of 'foo' would be popped in
the process of finding the parenthesis, re-enabling 'foo' and expanding
it a second time.

Looking for a function-like macro's opening parenthesis
=======================================================

Function-like macros only expand when immediately followed by a
parenthesis.  To do this cpplib needs to temporarily disable macros and
read the next token.  Unfortunately, because of spacing issues (*note
Token Spacing::), there can be fake padding tokens in-between, and if
the next real token is not a parenthesis cpplib needs to be able to back
up that one token as well as retain the information in any intervening
padding tokens.

   Backing up more than one token when macros are involved is not
permitted by cpplib, because in general it might involve issues like
restoring popped contexts onto the context stack, which are too hard.
Instead, searching for the parenthesis is handled by a special function,
'funlike_invocation_p', which remembers padding information as it reads
tokens.  If the next real token is not an opening parenthesis, it backs
up that one token, and then pushes an extra context just containing the
padding information if necessary.

Marking tokens ineligible for future expansion
==============================================

As discussed above, cpplib needs a way of marking tokens as
unexpandable.  Since the tokens cpplib handles are read-only once they
have been lexed, it instead makes a copy of the token and adds the flag
'NO_EXPAND' to the copy.

   For efficiency and to simplify memory management by avoiding having
to remember to free these tokens, they are allocated as temporary tokens
from the lexer's current token run (*note Lexing a line::) using the
function '_cpp_temp_token'.  The tokens are then re-used once the
current line of tokens has been read in.

   This might sound unsafe.  However, tokens runs are not re-used at the
end of a line if it happens to be in the middle of a macro argument
list, and cpplib only wants to back-up more than one lexer token in
situations where no macro expansion is involved, so the optimization is
safe.