sync to 1.9.0: polishing the core code of mdoc macro handling

1) If a macro is not parsed, do not parse it.  Of course, without
parsing it, we cannot produce "macro-like parameter" warnings,
but these were useless anyway.
2) If a macro is not callable, do not print a useless warning when
it occurs as a parameter, just display the raw characters.
3) Below .Bl -column, check whether macros are callable.
4) Like groff, allow whitespace after the initial dot on macro lines.

3 files changed, 59 insertions, 64 deletions

diff --git a/usr.bin/mandoc/libmdoc.h b/usr.bin/mandoc/libmdoc.h
index 0105a120ce4..859d11799b5 100644
--- a/usr.bin/mandoc/libmdoc.h
+++ b/usr.bin/mandoc/libmdoc.h
@@ -1,4 +1,4 @@
-/*	$Id: libmdoc.h,v 1.17 2009/07/26 22:48:41 schwarze Exp $ */
+/*	$Id: libmdoc.h,v 1.18 2009/08/22 19:43:33 schwarze Exp $ */
 /*
  * Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@kth.se>
  *
@@ -47,7 +47,6 @@ enum	merr {
 	EQUOTTERM,
 	EMALLOC,
 	EARGVAL,	
-	ENOCALL,
 	EBODYPROL,
 	EPROLBODY,
 	ETEXTPROL,
@@ -88,13 +87,11 @@ enum	merr {
 	ENOWIDTH,
 	EUTSNAME,
 	EOBS,
-	EMACPARM,
 	EIMPBRK,
 	EIGNE,
 	EOPEN,
 	EQUOTPHR,
 	ENOCTX,
-	ESPACE,
 	ELIB,
 	MERRMAX
 };
diff --git a/usr.bin/mandoc/mdoc.c b/usr.bin/mandoc/mdoc.c
index 13611c01067..061b0a5d6d8 100644
--- a/usr.bin/mandoc/mdoc.c
+++ b/usr.bin/mandoc/mdoc.c
@@ -1,4 +1,4 @@
-/*	$Id: mdoc.c,v 1.23 2009/08/22 17:21:24 schwarze Exp $ */
+/*	$Id: mdoc.c,v 1.24 2009/08/22 19:43:33 schwarze Exp $ */
 /*
  * Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@kth.se>
  *
@@ -29,7 +29,6 @@ const	char *const __mdoc_merrnames[MERRMAX] = {
 	"unterminated quoted parameter", /* EQUOTTERM */
 	"system: malloc error", /* EMALLOC */
 	"argument parameter suggested", /* EARGVAL */
-	"macro not callable", /* ENOCALL */
 	"macro disallowed in prologue", /* EBODYPROL */
 	"macro disallowed in body", /* EPROLBODY */
 	"text disallowed in prologue", /* ETEXTPROL */
@@ -70,13 +69,11 @@ const	char *const __mdoc_merrnames[MERRMAX] = {
 	"superfluous width argument", /* ENOWIDTH */
 	"system: utsname error", /* EUTSNAME */
 	"obsolete macro", /* EOBS */
-	"macro-like parameter", /* EMACPARM */
 	"end-of-line scope violation", /* EIMPBRK */
 	"empty macro ignored", /* EIGNE */
 	"unclosed explicit scope", /* EOPEN */
 	"unterminated quoted phrase", /* EQUOTPHR */
 	"closure macro without prior context", /* ENOCTX */
-	"invalid whitespace after control character", /* ESPACE */
 	"no description found for library" /* ELIB */
 };
 
@@ -349,7 +346,10 @@ int
 mdoc_macro(struct mdoc *m, int tok, 
 		int ln, int pp, int *pos, char *buf)
 {
-
+	/*
+	 * If we're in the prologue, deny "body" macros.  Similarly, if
+	 * we're in the body, deny prologue calls.
+	 */
 	if (MDOC_PROLOGUE & mdoc_macros[tok].flags && 
 			MDOC_PBODY & m->flags)
 		return(mdoc_perr(m, ln, pp, EPROLBODY));
@@ -357,9 +357,6 @@ mdoc_macro(struct mdoc *m, int tok,
 			! (MDOC_PBODY & m->flags))
 		return(mdoc_perr(m, ln, pp, EBODYPROL));
 
-	if (1 != pp && ! (MDOC_CALLABLE & mdoc_macros[tok].flags))
-		return(mdoc_perr(m, ln, pp, ENOCALL));
-
 	return((*mdoc_macros[tok].fp)(m, tok, ln, pp, pos, buf));
 }
 
@@ -668,7 +665,7 @@ macrowarn(struct mdoc *m, int ln, const char *buf)
 int
 parsemacro(struct mdoc *m, int ln, char *buf)
 {
-	int		  i, c;
+	int		  i, j, c, ppos;
 	char		  mac[5];
 
 	/* Empty lines are ignored. */
@@ -676,27 +673,32 @@ parsemacro(struct mdoc *m, int ln, char *buf)
 	if (0 == buf[1])
 		return(1);
 
-	if (' ' == buf[1]) {
-		i = 2;
+	i = 1;
+
+	/* Accept whitespace after the initial control char. */
+
+	if (' ' == buf[i]) {
+		i++;
 		while (buf[i] && ' ' == buf[i])
 			i++;
 		if (0 == buf[i])
 			return(1);
-		return(mdoc_perr(m, ln, 1, ESPACE));
 	}
 
+	ppos = i;
+
 	/* Copy the first word into a nil-terminated buffer. */
 
-	for (i = 1; i < 5; i++) {
-		if (0 == (mac[i - 1] = buf[i]))
+	for (j = 0; j < 4; j++, i++) {
+		if (0 == (mac[j] = buf[i]))
 			break;
 		else if (' ' == buf[i])
 			break;
 	}
 
-	mac[i - 1] = 0;
+	mac[j] = 0;
 
-	if (i == 5 || i <= 2) {
+	if (j == 4 || j < 2) {
 		if ( ! macrowarn(m, ln, mac))
 			goto err;
 		return(1);
@@ -713,9 +715,11 @@ parsemacro(struct mdoc *m, int ln, char *buf)
 	while (buf[i] && ' ' == buf[i])
 		i++;
 
-	/* Begin recursive parse sequence. */
-
-	if ( ! mdoc_macro(m, c, ln, 1, &i, buf)) 
+	/* 
+	 * Begin recursive parse sequence.  Since we're at the start of
+	 * the line, we don't need to do callable/parseable checks.
+	 */
+	if ( ! mdoc_macro(m, c, ln, ppos, &i, buf)) 
 		goto err;
 
 	return(1);
@@ -725,3 +729,5 @@ err:	/* Error out. */
 	m->flags |= MDOC_HALT;
 	return(0);
 }
+
+
diff --git a/usr.bin/mandoc/mdoc_macro.c b/usr.bin/mandoc/mdoc_macro.c
index e170bdb33b7..ccfbedcce73 100644
--- a/usr.bin/mandoc/mdoc_macro.c
+++ b/usr.bin/mandoc/mdoc_macro.c
@@ -1,4 +1,4 @@
-/*	$Id: mdoc_macro.c,v 1.20 2009/08/22 15:36:58 schwarze Exp $ */
+/*	$Id: mdoc_macro.c,v 1.21 2009/08/22 19:43:33 schwarze Exp $ */
 /*
  * Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@kth.se>
  *
@@ -45,9 +45,10 @@ static	int	  rew_impblock(struct mdoc *, int, int, int);
 static	int	  rew_expblock(struct mdoc *, int, int, int);
 static	int	  rew_subblock(enum mdoc_type, 
 			struct mdoc *, int, int, int);
-static	int	  rew_last(struct mdoc *, struct mdoc_node *);
+static	int	  rew_last(struct mdoc *, struct mdoc_node *); /* FIXME: make const */
 static	int	  append_delims(struct mdoc *, int, int *, char *);
-static	int	  lookup(struct mdoc *, int, int, int, const char *);
+static	int	  lookup(struct mdoc *, int, const char *);
+static	int	  lookup_raw(struct mdoc *, const char *);
 static	int	  swarn(struct mdoc *, enum mdoc_type, int, int, 
 			const struct mdoc_node *);
 
@@ -253,17 +254,24 @@ mdoc_macroend(struct mdoc *mdoc)
 }
 
 static int
-lookup(struct mdoc *mdoc, int line, int pos, int from, const char *p)
+lookup(struct mdoc *mdoc, int from, const char *p)
+{
+
+	if ( ! (MDOC_PARSED & mdoc_macros[from].flags))
+		return(MDOC_MAX);
+	return(lookup_raw(mdoc, p));
+}
+
+
+static int
+lookup_raw(struct mdoc *mdoc, const char *p)
 {
 	int		 res;
 
-	res = mdoc_hash_find(mdoc->htab, p);
-	if (MDOC_PARSED & mdoc_macros[from].flags)
-		return(res);
-	if (MDOC_MAX == res)
+	if (MDOC_MAX == (res = mdoc_hash_find(mdoc->htab, p)))
+		return(MDOC_MAX);
+	if (MDOC_CALLABLE & mdoc_macros[res].flags)
 		return(res);
-	if ( ! mdoc_pwarn(mdoc, line, pos, EMACPARM))
-		return(-1);
 	return(MDOC_MAX);
 }
 
@@ -591,6 +599,7 @@ rew_expblock(struct mdoc *mdoc, int tok, int line, int ppos)
 }
 
 
+/* FIXME: can this be merged with subblock? */
 static int
 rew_impblock(struct mdoc *mdoc, int tok, int line, int ppos)
 {
@@ -699,9 +708,7 @@ blk_exp_close(MACRO_PROT_ARGS)
 		if (ARGS_EOLN == c)
 			break;
 
-		if (-1 == (c = lookup(mdoc, line, lastarg, tok, p)))
-			return(0);
-		else if (MDOC_MAX != c) {
+		if (MDOC_MAX != (c = lookup(mdoc, tok, p))) {
 			if ( ! flushed) {
 				if ( ! rew_expblock(mdoc, tok, 
 							line, ppos))
@@ -790,8 +797,7 @@ in_line(MACRO_PROT_ARGS)
 
 		/* Quoted words shouldn't be looked-up. */
 
-		c = ARGS_QWORD == w ? MDOC_MAX :
-			lookup(mdoc, line, la, tok, p);
+		c = ARGS_QWORD == w ? MDOC_MAX : lookup(mdoc, tok, p);
 
 		/* 
 		 * In this case, we've located a submacro and must
@@ -800,7 +806,7 @@ in_line(MACRO_PROT_ARGS)
 		 * or raise a warning.
 		 */
 
-		if (MDOC_MAX != c && -1 != c) {
+		if (MDOC_MAX != c) {
 			if (0 == lastpunct && ! rew_elem(mdoc, tok))
 				return(0);
 			if (nc && 0 == cnt) {
@@ -820,8 +826,7 @@ in_line(MACRO_PROT_ARGS)
 			if (ppos > 1)
 				return(1);
 			return(append_delims(mdoc, line, pos, buf));
-		} else if (-1 == c)
-			return(0);
+		} 
 
 		/* 
 		 * Non-quote-enclosed punctuation.  Set up our scope, if
@@ -988,10 +993,7 @@ blk_full(MACRO_PROT_ARGS)
 			continue;
 		}
 
-		if (-1 == (c = lookup(mdoc, line, lastarg, tok, p)))
-			return(0);
-
-		if (MDOC_MAX == c) {
+		if (MDOC_MAX == (c = lookup(mdoc, tok, p))) {
 			if ( ! mdoc_word_alloc(mdoc, line, lastarg, p))
 				return(0);
 			mdoc->next = MDOC_NEXT_SIBLING;
@@ -1058,9 +1060,7 @@ blk_part_imp(MACRO_PROT_ARGS)
 		if (ARGS_EOLN == c)
 			break;
 
-		if (-1 == (c = lookup(mdoc, line, lastarg, tok, p)))
-			return(0);
-		else if (MDOC_MAX == c) {
+		if (MDOC_MAX == (c = lookup(mdoc, tok, p))) {
 			if ( ! mdoc_word_alloc(mdoc, line, lastarg, p))
 				return(0);
 			mdoc->next = MDOC_NEXT_SIBLING;
@@ -1162,9 +1162,7 @@ blk_part_exp(MACRO_PROT_ARGS)
 		if (ARGS_EOLN == c)
 			break;
 
-		if (-1 == (c = lookup(mdoc, line, lastarg, tok, p)))
-			return(0);
-		else if (MDOC_MAX != c) {
+		if (MDOC_MAX != (c = lookup(mdoc, tok, p))) {
 			if ( ! flushed) {
 				if ( ! rew_subblock(MDOC_HEAD, mdoc, 
 							tok, line, ppos))
@@ -1282,9 +1280,7 @@ in_line_argn(MACRO_PROT_ARGS)
 		if (ARGS_EOLN == c)
 			break;
 
-		if (-1 == (c = lookup(mdoc, line, lastarg, tok, p)))
-			return(0);
-		else if (MDOC_MAX != c) {
+		if (MDOC_MAX != (c = lookup(mdoc, tok, p))) {
 			if ( ! flushed && ! rew_elem(mdoc, tok))
 				return(0);
 			flushed = 1;
@@ -1360,15 +1356,13 @@ in_line_eoln(MACRO_PROT_ARGS)
 		if (ARGS_EOLN == w)
 			break;
 
-		c = ARGS_QWORD == w ? MDOC_MAX :
-			lookup(mdoc, line, la, tok, p);
+		c = ARGS_QWORD == w ? MDOC_MAX : lookup(mdoc, tok, p);
 
-		if (MDOC_MAX != c && -1 != c) {
+		if (MDOC_MAX != c) {
 			if ( ! rew_elem(mdoc, tok))
 				return(0);
 			return(mdoc_macro(mdoc, c, line, la, pos, buf));
-		} else if (-1 == c)
-			return(0);
+		} 
 
 		if ( ! mdoc_word_alloc(mdoc, line, la, p))
 			return(0);
@@ -1410,15 +1404,13 @@ phrase(struct mdoc *mdoc, int line, int ppos, char *buf)
 		if (ARGS_EOLN == w)
 			break;
 
-		c = ARGS_QWORD == w ? MDOC_MAX :
-			mdoc_hash_find(mdoc->htab, p);
+		c = ARGS_QWORD == w ? MDOC_MAX : lookup_raw(mdoc, p);
 
-		if (MDOC_MAX != c && -1 != c) {
+		if (MDOC_MAX != c) {
 			if ( ! mdoc_macro(mdoc, c, line, la, &pos, buf))
 				return(0);
 			return(append_delims(mdoc, line, &pos, buf));
-		} else if (-1 == c)
-			return(0);
+		} 
 
 		if ( ! mdoc_word_alloc(mdoc, line, la, p))
 			return(0);