Experimental results splitting formatted read and write

[Jerry DeLisle]

Jerry DeLisle wrote:
> As an experiment, I split out the formatted_transfer_scaler into 
> formatted_transfer_scaler_read and formatted_transfer_scaler_write 
> finctions, eliminating all the if reading conditionals.
> 
Some additional data from gprof: You can see that the processing of the format 
string and "traversing" the format tokens is beginning to dominate. For constant 
format strings, all of that could be done at compile time and stored as a 
pointer to a token list or even a list of format helper functions.  I wonder if 
g77 does anything like this?

Each sample counts as 0.01 seconds.
   %   cumulative   self
  time   seconds   seconds      name
  13.71      0.85     0.85      __mpn_mul_1
  11.29      1.55     0.70      ___printf_fp
   6.77      1.97     0.42      format_lex
   5.40      2.31     0.34      formatted_transfer_scalar_write
   4.92      2.61     0.31      next_char
   4.52      2.89     0.28      __write_nocancel
   3.87      3.13     0.24      __umodti3
   3.55      3.35     0.22      vfprintf
   3.39      3.56     0.21      parse_format_list
   3.23      3.76     0.20      get_fnode
   3.15      3.96     0.20      next_format0
   2.26      4.10     0.14      __udivti3
   2.18      4.23     0.14      formatted_transfer
   2.10      4.36     0.13      data_transfer_init
   1.77      4.47     0.11      ____strtol_l_internal
   1.61      4.57     0.10      T.123
   1.61      4.67     0.10      _gfortrani_next_format
   1.29      4.75     0.08      _IO_default_xsputn
   1.29      4.83     0.08      __ctype_toupper_loc
   1.29      4.91     0.08      memcpy