Re: [Groff] Odd problem escaping `.'

[Louis Guillaume]

Keith MARSHALL wrote:
> Louis Gillaume wrote:
>> I'm looking for a way to get my section numbers in the table of
>> contents. This is with -ms.
>>
>> This is what I've been trying:
>>
>> .NH 1
>> OVERVIEW
>> .XS
>> \n[H1]
>> \.\n[H2]
> 
> You would need \&.\n[H1] here, if the intention is to pass the initial
> dot through as text; (the \& is a non-spacing invisible character, which
> serves to hide the dot, suppressing its recognition as a precursor for
> a troff request).
> 
>> \.\n[H3]
>> OVERVIEW
>> .XE
> 
>> Why does it not produce:
>>   1.0.0.0 OVERVIEW . . . . . .
> 
> Because \. isn't sufficient to hide the dot from the parser, which still
> recognises this as a request precursor.  OTOH, when you use \\.\n[H3],
> you do prevent such recognition, because you've now placed a visible
> reverse solidus glyph before the initial dot.
> 
> If you do use \& to "escape" the dot, you still will not get the effect
> you wanted, because you've introduced white space before each dot, (as
> a result of the newlines before each).  Thus what you will get is:
> 
>   1 .0 .0 OVERVIEW . . . .
> 
> As Joerg has already pointed out, you could get around this by:
> 
>   .NH 1
>   OVERVIEW
>   .XS
>   \n[H1].\n[H2].\[H3] OVERVIEW
>   .XE
> 
> but this still isn't a good solution, for while .NH 1 does set the number
> register H1, it does *not* set H2, H3, or above -- they will simply remain
> set at whatever value they had after previous .NH 2, .NH 3, etc., which
> probably isn't what you want.
> 
> A better solution, which Joerg also suggested, is to use:
> 
>   .NH 1
>   OVERVIEW
>   .XS
>   \*[SN_DOT] OVERVIEW
>   .XE
> 
> which will give you:
> 
>   1. OVERVIEW . . . .
> 
> or:
> 
>   .NH 1
>   OVERVIEW
>   .XS
>   \*[SN_NO_DOT] OVERVIEW
>   .XE
> 
> which will give you:
> 
>   1 OVERVIEW . . . .
> 
> if you prefer not to see the final dot after the section number; it
> will still insert dots *between* subhead levels, as in
> 
>   1.1 SUBHEAD . . .
> 
> Note that Joerg actually suggested \*[SN]; by default this is the
> same as \*[SN], (it's an alias), but using \*[SN_DOT] and \*[SN_NO_DOT]
> is more explicit, (and you can change the \*[SN] behaviour, e.g. by
> 
>   .als SN SN_NO_DOT
> 
> if you prefer the latter behaviour).
> 
> HTH,
> Keith.



Thank you for the explanation. This is great. "\*[SN]" works.

Note that \*[SN_DOT] does not appear to work!

But \*[SN] seems to do the trick.

Is there a document that covers these escapes? I can't find anything in
groff(7) or in info groff.

Thanks again,

Louis