[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Groff] [Heirloom] Request for testing a release candidate
From: |
Carsten Kunze |
Subject: |
Re: [Groff] [Heirloom] Request for testing a release candidate |
Date: |
Thu, 18 Feb 2016 20:45:31 +0100 (CET) |
Hi Dorai,
Dorai Sitaram <address@hidden> wrote:
> This is somewhat related. What is the canonical way to determine using
> generic troff commands that the troff being used is Heirloom troff, without
> regard to whether compatibility is off or on? (I used to test the number
> register .g to rule out groff, but apparently that is not correct anymore.)
> (This should ease testing with existing file bases.)
Normally .g is 0 in Heirloom. It is only set with -mandoc, -man or -mdoc. .g
is intentionally set there to make it impossible for the user to test for the
tool.
(In case you ask regarding manual pages--always design the manual page for
groff and don't test for .g or do any other test. mandoc(1) formats manpages
designed for groff, and so does Heirloom. And please only use macros described
in groff_man(7) and groff_mdoc(7) (or mandoc_mdoc(7)) and use *no* low level
roff request and *no* escapes (except \c for -man). Prefer -mdoc in any case.
Often I see a line starting with "\fB.<some text>", this is not portable. Use
.B ".<some text>" for -man in this case.)
Also .X may be used for tests. Normally you should start with ".do xflag 3",
then \n(.X should be 3 only in Heirloom. .X is also set in manpages, but
please don't test for it. If Heirloom is not compatible to groff_man(7) or
groff_mdoc(7), then this is a bug.
Carsten