Re: advance=no problem

On May 13, 4:10 pm, "James Giles" <jamesgi...@[EMAIL PROTECTED]
> wrote:
> robert.corb...@[EMAIL PROTECTED]
 wrote:
>
> ...
>
> > One problem is that nonadvancing I/O serves more than one role.
> > It can be used for prompts, as the OP intended.  It can also be
> > used to build records a piece at a time.  In the latter case,
> > flu****ng at the end of each WRITE is not only costly in time,
> > but it eliminates some functionality (consider tab edit
> > descriptors).  [...]
>
> This last is incorrect.  A tab edit descriptor may not be used
> of tab to the left of the record's position at the beginning of the
> current I/O statement.  The T edit descriptor's number must be
> positive and is relative to the "teft tab limit".  The TL edit
> descriptor is not allowed to move further back than the
> current "left tab limit".  The "left tab limit" is reset by
> several things including the start of a new I/O statement.

I once interpreted the standard the way you do.  My implementation
of the tab edit descriptors in Sun Fortran worked that way.  After
discovering that other implementations did not work the way mine
did, I filed a request for interpretation against the Fortran 95
standard (interp. 000027).  The committee ruled the other way.

The Fortran 2003 standard is clearer.  In Section 10.7.1.1, it
states

    Immediately prior to nonchild data transfer, the left tab
    limit becomes defined as the character position of the
    current record or the character position of the stream file.

The character position of the current record is the position of
the start of the record, not the file position at the start of
the I/O statement.

Bob Corbett