 |
 |
Dan Nagle commented: ! The idea of the F2K draft is that achar( 10) will play the role ! of the single character '\n' in C.... ! C text files use '\n'; Fortran formatted stream files use achar( 10). Actually, it's not the file, but the data processed by the output library on the way to the file. ! .... One advantage is that vendors may implement Fortran stream files ! using the same library as C fgetc()/fputc().... I don't see how this works unless the C standard specifies that \n is ACHAR(10). If not, anything other than a higher-level abstraction than ACHAR(10) -- which the vendor would then be free to make \n -- makes this impossible; the Fortran output routines would need to front-end fputc() with an ACHAR(10) => \n filter. fgetc() works because of a curious asymmetry in the current draft of the Fortran 2000 standard: ACHAR(10) causes new-line during formatted stream output, but new-line is not specified to cause ACHAR(10) (or anything else) during formatted stream input. I have no idea what is expected to happen if new-line is represented by some junk in a record-mangler database (out-of- band signalling, to use Malcolm's term). Since the current draft of the Fortran 2000 standard is silent on this issue, whatever fgetc() does when it sees new-line is OK -- a nice portability problem for Dan to add to his portability project. If new-line were specified to be read as ACHAR(10) then fgetc() also wouldn't necessarily work -- the Fortran input library may need to back-end it with a \n => ACHAR(10) filter. -- What fraction of Americans believe | Van Snyder Wrestling is real and NASA is fake? | [log in to unmask] Any alleged opinions are my own and have not been approved or disapproved by JPL, CalTech, NASA, Dan Goldin, George Bush, the Pope, or anybody else.