Correct the psl regression test for sed(1)'s now-fixed newline

behavior.  Add the bcb regression test which checks for failures due
to a backslash ('\') coinciding with the very last character of the
command buffer.  The regression test is cf. this PR (which I did not
know about) and has a different fix for the bug.

PR:		bin/22351
Submitted by:	Stefan Duerholt <stefan.duerholt@t-online.de>

tools/regression/usr.bin/sed/regress.bcb.out

@@ -0,0 +1,4 @@
+input
+data
+for validation
+of sed(1)

tools/regression/usr.bin/sed/regress.psl.out

@@ -1 +1,4 @@
+
+
+
 of sed(1)

tools/regression/usr.bin/sed/regress.sh

@@ -5,6 +5,7 @@ REGRESSION_START($1)
 REGRESSION_TEST(`G', `sed G < regress.in')
 REGRESSION_TEST(`P', `sed P < regress.in')
 REGRESSION_TEST(`psl', `sed \$!g\;P\;D < regress.in')
+REGRESSION_TEST(`bcb', `sed s/X/$(jot -n -bx -s "" 2043)\\\\zz/ < regress.in')
 REGRESSION_TEST(`y', `echo -n foo | sed y/o/O/')
 
 REGRESSION_END()