[PATCH] Move common code to a new function

[Brooks Moses]

Steve Kargl wrote:
> The attached patch has been compiled and regression tested
> on i386-*-freebsd.  It moves common code in primary.c and
> io.c into a new function in match.c.  The code should be
> self-explanatory.  Ok for mainline?

Looks reasonable to me, but I do have a question:

> @@ -142,39 +142,8 @@ next_char (int in_string)
>    if (gfc_option.flag_backslash && c == '\\')
>      {
>        locus old_locus = gfc_current_locus;
> -
> -      switch (gfc_next_char_literal (1))
[...]
>  
>        if (!(gfc_option.allow_std & GFC_STD_GNU) && !inhibit_warnings)
>  	gfc_warning ("Extension: backslash character at %C");

Why are you not incorporating the whole "if (gfc_option.flag_backslash)" 
structure, and gfc_current_locus handling in the case where it's an 
unknown code, and the warning message, all into the function too?  From 
here, it looks like it's also all the same code in both places, and I 
would think it would be a good idea to place the whole structure into a 
new gfc_match_special_char function that gets called regardless of the 
state of flag_backslash, and does nothing if backslashes aren't supposed 
to be expanded.

Also, something that is likely to be just my ignorance:
> +  match m;
> +
> +  m = MATCH_YES;
Is there a reason this isn't "match m = MATCH_YES;"?

- Brooks