Re: [Groff] surprise, surprise

[Clarke Echols]

Werner LEMBERG wrote:
> 
> > As far as I am concerned, the only time that troff/groff/*roff
> > should ever interpret a "'" or "." as a control/command character is
> > when it appears as the FIRST character in a line of input.  If it is
> > preceded by \fB or anything else, it should always be treated as
> > ordinary text.
> 
> I fully agree.
> 
> > The only exception would be when the only preceding characters are
> > whitespace, which troff interprets as a break, equivalent to ".br".
> 
> I don't understand that.  Please give an example.
> 
>     Werner

Example:

Source file:

This is a running line of text that can be multiple lines
in a paragraph, for example.  The next lines start with a leading
blank (space) in the first column.
 line 1 following space
 line 2 following space
 line 3 following space
line 4 with no space and the ensuing text.

Formatted output from AT&T troff or nroff:

This is a running line of text that can be multiple lines in a
paragraph, for example.  The next lines start with a leading blank
(space) in the first column.
 line 1 following space
 line 2 following space
 line 3 following space line 4 with no space and the ensuing text.

It has been several years since I experimented with this phenomenon,
but I recall testing it.  I was going to opine further, but realized
I might have a manual around home, so I consulted it.  Quoting from
the HP manual: "Text Formatting User's Guide" (HP part number
B1862-90014 (January 1991)) which I wrote (page 6-15):

   o An input text line beginning with one or more spaces causes a
     break.  Leading spaces are not discarded when the line is printed.
     Leading tab characters in a line do not cause a break.

   o A blank line in input text causes a break followed by a vertical
     space equivalent to .sp 1v

   o Multiple consecutive blank input lines produce a break followed
     by a vertical space for each blank line encountered.

The manual is no longer available from HP unless I really miss my
guess.

I had intended it as a full reference manual for troff, but it got
hacked down to nroff only due to various strategic restraints.  It
is safe to say, however, that my writing was thoroughly tested on a
stock AT&T troff running on HP-UX systems, probably HP-UX 7.0 or
possibly 8.0 (less likely).  Generally speaking, if I wrote it,
you could take it to the bank because I was quite careful to test all
of my statements on real systems with real software.

Furthermore, my writings re: nroff/troff were derived from the content
of original AT&T Documenter's Workbench manuals, but they were difficult
to understand because they were poorly organized and more poorly indexed.

Clarke