Implement the roff .break request (break out of a .while loop).

Jan Stary <hans at stare dot cz> found it in an ancient groffer(1)
manual page (version 1.19) on MacOS X Mojave.
Having .break not implemented wasn't a particularly bright idea
because obviously, it tended to cause infinite loops.

5 files changed, 73 insertions, 12 deletions

diff --git a/regress/roff/while/Makefile b/regress/roff/while/Makefile
index ed970474..42ce590a 100644
--- a/regress/roff/while/Makefile
+++ b/regress/roff/while/Makefile
@@ -1,6 +1,6 @@
-# $OpenBSD: Makefile,v 1.1 2018/08/24 22:56:37 schwarze Exp $
+# $OpenBSD: Makefile,v 1.2 2019/04/21 22:43:00 schwarze Exp $
 
-REGRESS_TARGETS	= basic badargs into nesting outof
+REGRESS_TARGETS	= basic badargs break into nesting outof
 LINT_TARGETS	= badargs into nesting outof
 
 # mandoc defects:
diff --git a/regress/roff/while/break.in b/regress/roff/while/break.in
new file mode 100644
index 00000000..bfc8bc3f
--- /dev/null
+++ b/regress/roff/while/break.in
@@ -0,0 +1,16 @@
+.\" $OpenBSD: break.in,v 1.1 2019/04/21 22:43:00 schwarze Exp $
+.Dd $Mdocdate$
+.Dt WHILE-BREAK 1
+.Os
+.Sh NAME
+.Nm while-break
+.Nd break request inside a while loop
+.Sh DESCRIPTION
+initial text
+.nr cnt 11 1
+.while n \{\
+\n-[cnt]
+.if !\n[cnt] .break
+\(en
+.\}
+final text
diff --git a/regress/roff/while/break.out_ascii b/regress/roff/while/break.out_ascii
new file mode 100644
index 00000000..8e3f39b9
--- /dev/null
+++ b/regress/roff/while/break.out_ascii
@@ -0,0 +1,9 @@
+WHILE-BREAK(1)              General Commands Manual             WHILE-BREAK(1)
+
+NNAAMMEE
+     wwhhiillee--bbrreeaakk - break request inside a while loop
+
+DDEESSCCRRIIPPTTIIOONN
+     initial text 10 - 9 - 8 - 7 - 6 - 5 - 4 - 3 - 2 - 1 - 0 final text
+
+OpenBSD                         April 21, 2019                         OpenBSD
diff --git a/roff.7 b/roff.7
index ee1981a8..a6e1c8d4 100644
--- a/roff.7
+++ b/roff.7
@@ -503,10 +503,9 @@ This is a Heirloom extension and currently unsupported.
 .It Ic \&br
 Break the output line.
 .It Ic \&break
-Break out of a
+Break out of the innermost
 .Ic \&while
 loop.
-Currently unsupported.
 .It Ic \&breakchar Ar char ...
 Optional line break characters.
 This is a Heirloom extension and currently ignored.
diff --git a/roff.c b/roff.c
index 0766abad..9744a01f 100644
--- a/roff.c
+++ b/roff.c
@@ -133,15 +133,18 @@ struct	roff {
 	char		 escape; /* escape character */
 };
 
+/*
+ * A macro definition, condition, or ignored block.
+ */
 struct	roffnode {
 	enum roff_tok	 tok; /* type of node */
 	struct roffnode	*parent; /* up one in stack */
 	int		 line; /* parse line */
 	int		 col; /* parse col */
 	char		*name; /* node name, e.g. macro name */
-	char		*end; /* end-rules: custom token */
-	int		 endspan; /* end-rules: next-line or infty */
-	int		 rule; /* current evaluation rule */
+	char		*end; /* custom end macro of the block */
+	int		 endspan; /* scope to: 1=eol 2=next line -1=\} */
+	int		 rule; /* content is: 1=evaluated 0=skipped */
 };
 
 #define	ROFF_ARGS	 struct roff *r, /* parse ctx */ \
@@ -181,6 +184,7 @@ static	int		 roff_als(ROFF_ARGS);
 static	int		 roff_block(ROFF_ARGS);
 static	int		 roff_block_text(ROFF_ARGS);
 static	int		 roff_block_sub(ROFF_ARGS);
+static	int		 roff_break(ROFF_ARGS);
 static	int		 roff_cblock(ROFF_ARGS);
 static	int		 roff_cc(ROFF_ARGS);
 static	int		 roff_ccond(struct roff *, int, int);
@@ -400,7 +404,7 @@ static	struct roffmac	 roffs[TOKEN_NONE] = {
 	{ roff_unsupp, NULL, NULL, 0 },  /* boxa */
 	{ roff_line_ignore, NULL, NULL, 0 },  /* bp */
 	{ roff_unsupp, NULL, NULL, 0 },  /* BP */
-	{ roff_unsupp, NULL, NULL, 0 },  /* break */
+	{ roff_break, NULL, NULL, 0 },  /* break */
 	{ roff_line_ignore, NULL, NULL, 0 },  /* breakchar */
 	{ roff_line_ignore, NULL, NULL, 0 },  /* brnl */
 	{ roff_noarg, NULL, NULL, 0 },  /* brp */
@@ -685,7 +689,7 @@ roffhash_find(struct ohash *htab, const char *name, size_t sz)
 
 /*
  * Pop the current node off of the stack of roff instructions currently
- * pending.
+ * pending.  Return 1 if it is a loop or 0 otherwise.
  */
 static int
 roffnode_pop(struct roff *r)
@@ -2002,6 +2006,10 @@ roff_cblock(ROFF_ARGS)
 
 }
 
+/*
+ * Pop all nodes ending at the end of the current input line.
+ * Return the number of loops ended.
+ */
 static int
 roffnode_cleanscope(struct roff *r)
 {
@@ -2016,6 +2024,11 @@ roffnode_cleanscope(struct roff *r)
 	return inloop;
 }
 
+/*
+ * Handle the closing \} of a conditional block.
+ * Apart from generating warnings, this only pops nodes.
+ * Return the number of loops ended.
+ */
 static int
 roff_ccond(struct roff *r, int ln, int ppos)
 {
@@ -2235,6 +2248,7 @@ roff_block_text(ROFF_ARGS)
 static int
 roff_cond_sub(ROFF_ARGS)
 {
+	struct roffnode	*bl;
 	char		*ep;
 	int		 endloop, irc, rr;
 	enum roff_tok	 t;
@@ -2282,9 +2296,21 @@ roff_cond_sub(ROFF_ARGS)
 	 */
 
 	t = roff_parse(r, buf->buf, &pos, ln, ppos);
-	irc |= t != TOKEN_NONE && (rr || roffs[t].flags & ROFFMAC_STRUCT) ?
-	    (*roffs[t].proc)(r, t, buf, ln, ppos, pos, offs) :
-	    rr ? ROFF_CONT : ROFF_IGN;
+	if (t == ROFF_break) {
+		if (irc & ROFF_LOOPMASK)
+			irc = ROFF_IGN | ROFF_LOOPEXIT;
+		else if (rr) {
+			for (bl = r->last; bl != NULL; bl = bl->parent) {
+				bl->rule = 0;
+				if (bl->tok == ROFF_while)
+					break;
+			}
+		}
+	} else if (t != TOKEN_NONE &&
+	    (rr || roffs[t].flags & ROFFMAC_STRUCT))
+		irc |= (*roffs[t].proc)(r, t, buf, ln, ppos, pos, offs);
+	else
+		irc |= rr ? ROFF_CONT : ROFF_IGN;
 	return irc;
 }
 
@@ -3482,6 +3508,17 @@ roff_als(ROFF_ARGS)
 	return ROFF_IGN;
 }
 
+/*
+ * The .break request only makes sense inside conditionals,
+ * and that case is already handled in roff_cond_sub().
+ */
+static int
+roff_break(ROFF_ARGS)
+{
+	mandoc_msg(MANDOCERR_BLK_NOTOPEN, ln, pos, "break");
+	return ROFF_IGN;
+}
+
 static int
 roff_cc(ROFF_ARGS)
 {