Hi Alex, At 2025-08-15T11:24:06+0200, Alejandro Colomar wrote: > On Tue, Aug 12, 2025 at 11:18:41AM +0200, Alejandro Colomar wrote: > > The output looks like this: > > > > $ make lint-man-blank -R > > GREP ^$ .tmp/man/man7/bpf-helpers.7.lint-man.blank.touch > > .tmp/man/man7/bpf-helpers.7: spurious blank lines: > > 548: > > 552: > > 555: > > 3057: > > make: *** > > [/srv/alx/src/linux/man-pages/man-pages/contrib/share/mk/lint/man/blank.mk:22: > > .tmp/man/man7/bpf-helpers.7.lint-man.blank.touch] Error 1 > > Actually, this has some false positives in tbl code: > > $ grep -m1 -B3 -A8 ^$ man/man7/inode.7 > .TS > lB l l. > S_IFMT 0170000 bit mask for the file type bit field > > S_IFSOCK 0140000 socket > S_IFLNK 0120000 symbolic link > S_IFREG 0100000 regular file > S_IFBLK 0060000 block device > S_IFDIR 0040000 directory > S_IFCHR 0020000 character device > S_IFIFO 0010000 FIFO > .TE > > Is there a way to avoid these blanks in tbl code?
Certainly. You can use the *roff dummy character `\&` to make the table
row non-blank according to the formatter.
This makes no difference even on typesetters, because the (default)
meaning of a blank input line is a 1v space on the output. A text line
with only a dummy character on it behaves the same way. Thus, I
wouldn't call these false positives, but true ones.*
$ printf 'foo\n.TS\nL.\n\n.TE\nbar\n' | groff -tZ > /tmp/blank
$ printf 'foo\n.TS\nL.\n\\&\n.TE\nbar\n' | groff -tZ > /tmp/dummy
$ diff -u /tmp/blank /tmp/dummy && echo SAME
SAME
In principle, we could change GNU tbl to _transform_ blank lines (inside
tbl regions only) to consist of only a dummy character and newline--that
should have zero impact on any document's formatting anywhere.
On the other hand, I think it might make more sense to leave things as
they are, and encourage a discipline of explicit `\&` usage. The odds
of someone accidentally typing that digraph are vanishingly small,** and
in a tbl context (or any *roff scenario where filling is disabled) it is
easy to interpret as "I want a blank line here; I know what I'm doing".
Regards,
Branden
* Recall that on typesetters, in man(7) and other full-service macro
packages generally, the inter-paragraph space amount is _not_ 1v,
but somewhat less. If you want a break in your table, it's wise to
be explicit about how big you want that break to be. Sometimes you
can use a package's paragraphing macros inside a tbl region, but no
one guarantees this. What you _can_ do is insert `sp` requests, and
this seems to be a countenanced practice judging from documents I've
seen prepared by seasoned practitioners. On the third hand, that
trick won't work in general in ".TS H/.TH" tables because they are
formatted in diversions, where the rules about vertical spacing are
different.*** And on the fourth hand, that doesn't matter for
man(7) because it doesn't support ".TS H/.TH" tables anyway. If
it's ever going to, we'll need a synonym for tbl's ".TH" because
that token collides with existing man(7) usage like the McLennan
County locomotive crash of 1896.
** contrast with the odds of accidentally--or, more likely,
thoughtlessly--whacking the Enter key
***
https://www.gnu.org/software/groff/manual/groff.html.node/Manipulating-Spacing.html
signature.asc
Description: PGP signature
