Support manual tagging of .Pp, .Bd, .D1, .Dl, .Bl, and .It.
In HTML output, improve the logic for writing inside permalinks:
skip them when there is no child content or when there is a risk
that the children might contain flow content.

Represent mdoc(7) .Pp (and .sp, and some SYNOPSIS and .Rs features)
by the <p> HTML element and use the html_fillmode() mechanism
for .Bd -unfilled, just like it was done for man(7) earlier, finally
getting rid both of the horrible <div class="Pp"></div> hack and
of the worst HTML syntax violations caused by nested displays.

Care is needed because in some situations, paragraphs have to remain
open across several subsequent macros, whereas in other situations,
they must get closed together with a block containing them.

Some implementation details include:
* Always close paragraphs before emitting HTML flow content.
* Let html_close_paragraph() also close <pre> for extra safety.
* Drop the old, now unused function print_paragraph().
* Minor adjustments in the top-level man(7) node formatter for symmetry.
* Bugfix: .Ss heads suspend no-fill mode, even though .Ss doesn't end it.
* Bugfix: give up on .Op semantic markup for now, see the comment.

Messages of the -Wbase level now print STYLE:. Since this
causes horrible churn anyway, profit of the opportunity to stop
excessive testing, such that this is hopefully the last instance
of such churn. Consistently use OpenBSD RCS tags, blank .Os,
blank fourth .TH argument, and Mdocdate like everywhere else.
Use -Ios=OpenBSD for platform-independent predictable output.

enable -T markdown tests of filled displays and tagged lists

Represent mdoc(7) .Pp (and .sp, and some SYNOPSIS and .Rs features)
by the <p> HTML element and use the html_fillmode() mechanism
for .Bd -unfilled, just like it was done for man(7) earlier, finally
getting rid both of the horrible <div class="Pp"></div> hack and
of the worst HTML syntax violations caused by nested displays.

Care is needed because in some situations, paragraphs have to remain
open across several subsequent macros, whereas in other situations,
they must get closed together with a block containing them.

Some implementation details include:
* Always close paragraphs before emitting HTML flow content.
* Let html_close_paragraph() also close <pre> for extra safety.
* Drop the old, now unused function print_paragraph().
* Minor adjustments in the top-level man(7) node formatter for symmetry.
* Bugfix: .Ss heads suspend no-fill mode, even though .Ss doesn't end it.
* Bugfix: give up on .Op semantic markup for now, see the comment.

Messages of the -Wbase level now print STYLE:. Since this
causes horrible churn anyway, profit of the opportunity to stop
excessive testing, such that this is hopefully the last instance
of such churn. Consistently use OpenBSD RCS tags, blank .Os,
blank fourth .TH argument, and Mdocdate like everywhere else.
Use -Ios=OpenBSD for platform-independent predictable output.

enable -T markdown tests of filled displays and tagged lists