===================================================================
RCS file: /cvs/mandoc/roff.c,v
retrieving revision 1.106
retrieving revision 1.118
diff -u -p -r1.106 -r1.118
--- mandoc/roff.c	2010/12/02 10:53:03	1.106
+++ mandoc/roff.c	2011/01/02 10:10:57	1.118
@@ -1,4 +1,4 @@
-/*	$Id: roff.c,v 1.106 2010/12/02 10:53:03 kristaps Exp $ */
+/*	$Id: roff.c,v 1.118 2011/01/02 10:10:57 kristaps Exp $ */
 /*
  * Copyright (c) 2010 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2010 Ingo Schwarze <schwarze@openbsd.org>
@@ -29,6 +29,7 @@
 
 #include "mandoc.h"
 #include "roff.h"
+#include "libroff.h"
 #include "libmandoc.h"
 
 #define	RSTACK_MAX	128
@@ -36,14 +37,6 @@
 #define	ROFF_CTL(c) \
 	('.' == (c) || '\'' == (c))
 
-#if 1
-#define	ROFF_DEBUG(fmt, args...) \
-	do { /* Nothing. */ } while (/*CONSTCOND*/ 0)
-#else
-#define	ROFF_DEBUG(fmt, args...) \
-	do { fprintf(stderr, fmt , ##args); } while (/*CONSTCOND*/ 0)
-#endif
-
 enum	rofft {
 	ROFF_ad,
 	ROFF_am,
@@ -64,6 +57,9 @@ enum	rofft {
 	ROFF_rm,
 	ROFF_so,
 	ROFF_tr,
+	ROFF_TS,
+	ROFF_TE,
+	ROFF_T_,
 	ROFF_cblock,
 	ROFF_ccond, /* FIXME: remove this. */
 	ROFF_USERDEF,
@@ -75,7 +71,6 @@ enum	roffrule {
 	ROFFRULE_DENY
 };
 
-
 struct	roffstr {
 	char		*name; /* key of symbol */
 	char		*string; /* current value */
@@ -91,6 +86,9 @@ struct	roff {
 	struct regset	*regs; /* read/writable registers */
 	struct roffstr	*first_string; /* user-defined strings & macros */
 	const char	*current_string; /* value of last called user macro */
+	struct tbl_node	*first_tbl; /* first table parsed */
+	struct tbl_node	*last_tbl; /* last table parsed */
+	struct tbl_node	*tbl; /* current table being parsed */
 };
 
 struct	roffnode {
@@ -146,6 +144,9 @@ static	int		 roff_res(struct roff *, 
 static	void		 roff_setstr(struct roff *,
 				const char *, const char *, int);
 static	enum rofferr	 roff_so(ROFF_ARGS);
+static	enum rofferr	 roff_TE(ROFF_ARGS);
+static	enum rofferr	 roff_TS(ROFF_ARGS);
+static	enum rofferr	 roff_T_(ROFF_ARGS);
 static	enum rofferr	 roff_userdef(ROFF_ARGS);
 
 /* See roff_hash_find() */
@@ -176,6 +177,9 @@ static	struct roffmac	 roffs[ROFF_MAX] = {
 	{ "rm", roff_line_error, NULL, NULL, 0, NULL },
 	{ "so", roff_so, NULL, NULL, 0, NULL },
 	{ "tr", roff_line_ignore, NULL, NULL, 0, NULL },
+	{ "TS", roff_TS, NULL, NULL, 0, NULL },
+	{ "TE", roff_TE, NULL, NULL, 0, NULL },
+	{ "T&", roff_T_, NULL, NULL, 0, NULL },
 	{ ".", roff_cblock, NULL, NULL, 0, NULL },
 	{ "\\}", roff_ccond, NULL, NULL, 0, NULL },
 	{ NULL, roff_userdef, NULL, NULL, 0, NULL },
@@ -264,7 +268,6 @@ roffnode_pop(struct roff *r)
 		if (r->rstackpos > -1)
 			r->rstackpos--;
 
-	ROFF_DEBUG("roff: popping scope\n");
 	r->last = r->last->parent;
 	free(p->name);
 	free(p->end);
@@ -298,9 +301,19 @@ roffnode_push(struct roff *r, enum rofft tok, const ch
 static void
 roff_free1(struct roff *r)
 {
+	struct tbl_node	*t;
 
+	while (r->first_tbl) {
+		t = r->first_tbl;
+		r->first_tbl = t->next;
+		tbl_free(t);
+	}
+
+	r->first_tbl = r->last_tbl = r->tbl = NULL;
+
 	while (r->last)
 		roffnode_pop(r);
+
 	roff_freestr(r);
 }
 
@@ -346,15 +359,40 @@ roff_alloc(struct regset *regs, void *data, const mand
 static int
 roff_res(struct roff *r, char **bufp, size_t *szp, int pos)
 {
-	const char	*cp, *cpp, *st, *res;
+	const char	*stesc;	/* start of an escape sequence ('\\') */
+	const char	*stnam;	/* start of the name, after "[(*" */
+	const char	*cp;	/* end of the name, e.g. before ']' */
+	const char	*res;	/* the string to be substituted */
 	int		 i, maxl;
 	size_t		 nsz;
 	char		*n;
 
-	/* LINTED */
-	for (cp = &(*bufp)[pos]; (cpp = strstr(cp, "\\*")); cp++) {
-		cp = cpp + 2;
+	/* Search for a leading backslash and save a pointer to it. */
+
+	cp = *bufp + pos;
+	while (NULL != (cp = strchr(cp, '\\'))) {
+		stesc = cp++;
+
+		/*
+		 * The second character must be an asterisk.
+		 * If it isn't, skip it anyway:  It is escaped,
+		 * so it can't start another escape sequence.
+		 */
+
+		if ('\0' == *cp)
+			return(1);
+		if ('*' != *cp++)
+			continue;
+
+		/*
+		 * The third character decides the length
+		 * of the name of the string.
+		 * Save a pointer to the name.
+		 */
+
 		switch (*cp) {
+		case ('\0'):
+			return(1);
 		case ('('):
 			cp++;
 			maxl = 2;
@@ -367,8 +405,9 @@ roff_res(struct roff *r, char **bufp, size_t *szp, int
 			maxl = 1;
 			break;
 		}
+		stnam = cp;
 
-		st = cp;
+		/* Advance to the end of the name. */
 
 		for (i = 0; 0 == maxl || i < maxl; i++, cp++) {
 			if ('\0' == *cp)
@@ -377,21 +416,24 @@ roff_res(struct roff *r, char **bufp, size_t *szp, int
 				break;
 		}
 
-		res = roff_getstrn(r, st, (size_t)i);
+		/*
+		 * Retrieve the replacement string; if it is
+		 * undefined, resume searching for escapes.
+		 */
 
+		res = roff_getstrn(r, stnam, (size_t)i);
+
 		if (NULL == res) {
 			cp -= maxl ? 1 : 0;
 			continue;
 		}
 
-		ROFF_DEBUG("roff: splicing reserved: [%.*s]\n", i, st);
+		/* Replace the escape sequence by the string. */
 
 		nsz = *szp + strlen(res) + 1;
 		n = mandoc_malloc(nsz);
 
-		*n = '\0';
-
-		strlcat(n, *bufp, (size_t)(cpp - *bufp + 1));
+		strlcpy(n, *bufp, (size_t)(stesc - *bufp + 1));
 		strlcat(n, res, nsz);
 		strlcat(n, cp + (maxl ? 0 : 1), nsz);
 
@@ -411,6 +453,7 @@ roff_parseln(struct roff *r, int ln, char **bufp, 
 		size_t *szp, int pos, int *offs)
 {
 	enum rofft	 t;
+	enum rofferr	 e;
 	int		 ppos;
 
 	/*
@@ -430,13 +473,17 @@ roff_parseln(struct roff *r, int ln, char **bufp, 
 	if (r->last && ! ROFF_CTL((*bufp)[pos])) {
 		t = r->last->tok;
 		assert(roffs[t].text);
-		ROFF_DEBUG("roff: intercept scoped text: %s, [%s]\n", 
-				roffs[t].name, &(*bufp)[pos]);
-		return((*roffs[t].text)
-				(r, t, bufp, szp, 
-				 ln, pos, pos, offs));
-	} else if ( ! ROFF_CTL((*bufp)[pos]))
+		e = (*roffs[t].text)
+			(r, t, bufp, szp, ln, pos, pos, offs);
+		assert(ROFF_IGN == e || ROFF_CONT == e);
+		if (ROFF_CONT == e && r->tbl)
+			return(tbl_read(r->tbl, ln, *bufp, *offs));
+		return(e);
+	} else if ( ! ROFF_CTL((*bufp)[pos])) {
+		if (r->tbl)
+			return(tbl_read(r->tbl, ln, *bufp, *offs));
 		return(ROFF_CONT);
+	}
 
 	/*
 	 * If a scope is open, go to the child handler for that macro,
@@ -446,8 +493,6 @@ roff_parseln(struct roff *r, int ln, char **bufp, 
 	if (r->last) {
 		t = r->last->tok;
 		assert(roffs[t].sub);
-		ROFF_DEBUG("roff: intercept scoped context: %s, [%s]\n", 
-				roffs[t].name, &(*bufp)[pos]);
 		return((*roffs[t].sub)
 				(r, t, bufp, szp, 
 				 ln, pos, pos, offs));
@@ -463,8 +508,6 @@ roff_parseln(struct roff *r, int ln, char **bufp, 
 	if (ROFF_MAX == (t = roff_parse(r, *bufp, &pos)))
 		return(ROFF_CONT);
 
-	ROFF_DEBUG("roff: intercept new-scope: %s, [%s]\n", 
-			roffs[t].name, &(*bufp)[pos]);
 	assert(roffs[t].proc);
 	return((*roffs[t].proc)
 			(r, t, bufp, szp, 
@@ -472,14 +515,20 @@ roff_parseln(struct roff *r, int ln, char **bufp, 
 }
 
 
-int
+void
 roff_endparse(struct roff *r)
 {
 
-	if (NULL == r->last)
-		return(1);
-	return((*r->msg)(MANDOCERR_SCOPEEXIT, r->data, r->last->line, 
-				r->last->col, NULL));
+	if (r->last)
+		(*r->msg)(MANDOCERR_SCOPEEXIT, r->data, 
+				r->last->line, r->last->col, NULL);
+
+	if (r->tbl) {
+		(*r->msg)(MANDOCERR_SCOPEEXIT, r->data, 
+				r->tbl->line, r->tbl->pos, NULL);
+		tbl_end(r->tbl);
+		r->tbl = NULL;
+	}
 }
 
 
@@ -548,8 +597,7 @@ roff_cblock(ROFF_ARGS)
 	 */
 
 	if (NULL == r->last) {
-		if ( ! (*r->msg)(MANDOCERR_NOSCOPE, r->data, ln, ppos, NULL))
-			return(ROFF_ERR);
+		(*r->msg)(MANDOCERR_NOSCOPE, r->data, ln, ppos, NULL);
 		return(ROFF_IGN);
 	}
 
@@ -561,20 +609,19 @@ roff_cblock(ROFF_ARGS)
 	case (ROFF_am1):
 		/* FALLTHROUGH */
 	case (ROFF_de):
+		/* ROFF_de1 is remapped to ROFF_de in roff_block(). */
 		/* FALLTHROUGH */
 	case (ROFF_dei):
 		/* FALLTHROUGH */
 	case (ROFF_ig):
 		break;
 	default:
-		if ( ! (*r->msg)(MANDOCERR_NOSCOPE, r->data, ln, ppos, NULL))
-			return(ROFF_ERR);
+		(*r->msg)(MANDOCERR_NOSCOPE, r->data, ln, ppos, NULL);
 		return(ROFF_IGN);
 	}
 
 	if ((*bufp)[pos])
-		if ( ! (*r->msg)(MANDOCERR_ARGSLOST, r->data, ln, pos, NULL))
-			return(ROFF_ERR);
+		(*r->msg)(MANDOCERR_ARGSLOST, r->data, ln, pos, NULL);
 
 	roffnode_pop(r);
 	roffnode_cleanscope(r);
@@ -601,8 +648,7 @@ roff_ccond(ROFF_ARGS)
 {
 
 	if (NULL == r->last) {
-		if ( ! (*r->msg)(MANDOCERR_NOSCOPE, r->data, ln, ppos, NULL))
-			return(ROFF_ERR);
+		(*r->msg)(MANDOCERR_NOSCOPE, r->data, ln, ppos, NULL);
 		return(ROFF_IGN);
 	}
 
@@ -614,20 +660,17 @@ roff_ccond(ROFF_ARGS)
 	case (ROFF_if):
 		break;
 	default:
-		if ( ! (*r->msg)(MANDOCERR_NOSCOPE, r->data, ln, ppos, NULL))
-			return(ROFF_ERR);
+		(*r->msg)(MANDOCERR_NOSCOPE, r->data, ln, ppos, NULL);
 		return(ROFF_IGN);
 	}
 
 	if (r->last->endspan > -1) {
-		if ( ! (*r->msg)(MANDOCERR_NOSCOPE, r->data, ln, ppos, NULL))
-			return(ROFF_ERR);
+		(*r->msg)(MANDOCERR_NOSCOPE, r->data, ln, ppos, NULL);
 		return(ROFF_IGN);
 	}
 
 	if ((*bufp)[pos])
-		if ( ! (*r->msg)(MANDOCERR_ARGSLOST, r->data, ln, pos, NULL))
-			return(ROFF_ERR);
+		(*r->msg)(MANDOCERR_ARGSLOST, r->data, ln, pos, NULL);
 
 	roffnode_pop(r);
 	roffnode_cleanscope(r);
@@ -650,6 +693,12 @@ roff_block(ROFF_ARGS)
 			(*r->msg)(MANDOCERR_NOARGS, r->data, ln, ppos, NULL);
 			return(ROFF_IGN);
 		}
+
+		/*
+		 * Re-write `de1', since we don't really care about
+		 * groff's strange compatibility mode, into `de'.
+		 */
+
 		if (ROFF_de1 == tok)
 			tok = ROFF_de;
 		if (ROFF_de == tok)
@@ -657,8 +706,10 @@ roff_block(ROFF_ARGS)
 		else
 			(*r->msg)(MANDOCERR_REQUEST, r->data, ln, ppos,
 			    roffs[tok].name);
+
 		while ((*bufp)[pos] && ' ' != (*bufp)[pos])
 			pos++;
+
 		while (' ' == (*bufp)[pos])
 			(*bufp)[pos++] = '\0';
 	}
@@ -670,14 +721,18 @@ roff_block(ROFF_ARGS)
 	 * with the same name, if there is one.  New content will be
 	 * added from roff_block_text() in multiline mode.
 	 */
+
 	if (ROFF_de == tok)
 		roff_setstr(r, name, "", 0);
 
 	if ('\0' == (*bufp)[pos])
 		return(ROFF_IGN);
 
+	/* If present, process the custom end-of-line marker. */
+
 	sv = pos;
-	while ((*bufp)[pos] && ' ' != (*bufp)[pos] && 
+	while ((*bufp)[pos] &&
+			' ' != (*bufp)[pos] && 
 			'\t' != (*bufp)[pos])
 		pos++;
 
@@ -699,8 +754,7 @@ roff_block(ROFF_ARGS)
 	r->last->end[(int)sz] = '\0';
 
 	if ((*bufp)[pos])
-		if ( ! (*r->msg)(MANDOCERR_ARGSLOST, r->data, ln, pos, NULL))
-			return(ROFF_ERR);
+		(*r->msg)(MANDOCERR_ARGSLOST, r->data, ln, pos, NULL);
 
 	return(ROFF_IGN);
 }
@@ -933,18 +987,14 @@ roff_cond(ROFF_ARGS)
 	 */
 
 	if ('\0' == (*bufp)[pos] && sv != pos) {
-		if ((*r->msg)(MANDOCERR_NOARGS, r->data, ln, ppos, NULL))
-			return(ROFF_IGN);
-		return(ROFF_ERR);
+		(*r->msg)(MANDOCERR_NOARGS, r->data, ln, ppos, NULL);
+		return(ROFF_IGN);
 	}
 
 	roffnode_push(r, tok, NULL, ln, ppos);
 
 	r->last->rule = rule;
 
-	ROFF_DEBUG("roff: cond: %s -> %s\n", roffs[tok].name, 
-			ROFFRULE_ALLOW == rule ?  "allow" : "deny");
-
 	if (ROFF_ie == tok) {
 		/*
 		 * An if-else will put the NEGATION of the current
@@ -959,11 +1009,8 @@ roff_cond(ROFF_ARGS)
 
 	/* If the parent has false as its rule, then so do we. */
 
-	if (r->last->parent && ROFFRULE_DENY == r->last->parent->rule) {
+	if (r->last->parent && ROFFRULE_DENY == r->last->parent->rule)
 		r->last->rule = ROFFRULE_DENY;
-		ROFF_DEBUG("roff: cond override: %s -> deny\n",
-				roffs[tok].name);
-	}
 
 	/*
 	 * Determine scope.  If we're invoked with "\{" trailing the
@@ -976,11 +1023,7 @@ roff_cond(ROFF_ARGS)
 	if ('\\' == (*bufp)[pos] && '{' == (*bufp)[pos + 1]) {
 		r->last->endspan = -1;
 		pos += 2;
-		ROFF_DEBUG("roff: cond-scope: %s, multi-line\n", 
-				roffs[tok].name);
-	} else
-		ROFF_DEBUG("roff: cond-scope: %s, one-line\n", 
-				roffs[tok].name);
+	} 
 
 	/*
 	 * If there are no arguments on the line, the next-line scope is
@@ -1073,17 +1116,62 @@ roff_nr(ROFF_ARGS)
 		rg[(int)REG_nS].set = 1;
 		if ( ! roff_parse_nat(val, &rg[(int)REG_nS].v.u))
 			rg[(int)REG_nS].v.u = 0;
+	}
 
-		ROFF_DEBUG("roff: register nS: %u\n", 
-				rg[(int)REG_nS].v.u);
-	} else
-		ROFF_DEBUG("roff: ignoring register: %s\n", key);
+	return(ROFF_IGN);
+}
 
+/* ARGSUSED */
+static enum rofferr
+roff_TE(ROFF_ARGS)
+{
+
+	if (NULL == r->tbl)
+		(*r->msg)(MANDOCERR_NOSCOPE, r->data, ln, ppos, NULL);
+	else
+		tbl_end(r->tbl);
+
+	r->tbl = NULL;
 	return(ROFF_IGN);
 }
 
 /* ARGSUSED */
 static enum rofferr
+roff_T_(ROFF_ARGS)
+{
+
+	if (NULL == r->tbl)
+		(*r->msg)(MANDOCERR_NOSCOPE, r->data, ln, ppos, NULL);
+	else
+		tbl_restart(ppos, ln, r->tbl);
+
+	return(ROFF_IGN);
+}
+
+/* ARGSUSED */
+static enum rofferr
+roff_TS(ROFF_ARGS)
+{
+	struct tbl_node	*t;
+
+	if (r->tbl) {
+		(*r->msg)(MANDOCERR_SCOPEBROKEN, r->data, ln, ppos, NULL);
+		tbl_end(r->tbl);
+	}
+
+	t = tbl_alloc(ppos, ln, r->data, r->msg);
+
+	if (r->last_tbl)
+		r->last_tbl->next = t;
+	else
+		r->first_tbl = r->last_tbl = t;
+
+	r->tbl = r->last_tbl = t;
+	return(ROFF_IGN);
+}
+
+/* ARGSUSED */
+static enum rofferr
 roff_so(ROFF_ARGS)
 {
 	char *name;
@@ -1295,4 +1383,11 @@ roff_freestr(struct roff *r)
 	}
 
 	r->first_string = NULL;
+}
+
+const struct tbl_span *
+roff_span(const struct roff *r)
+{
+	
+	return(r->tbl ? tbl_span(r->tbl) : NULL);
 }