===================================================================
RCS file: /cvs/mandoc/read.c,v
retrieving revision 1.206
retrieving revision 1.215
diff -u -p -r1.206 -r1.215
--- mandoc/read.c	2018/12/14 05:18:03	1.206
+++ mandoc/read.c	2020/03/13 15:32:28	1.215
@@ -1,7 +1,7 @@
-/*	$Id: read.c,v 1.206 2018/12/14 05:18:03 schwarze Exp $ */
+/* $Id: read.c,v 1.215 2020/03/13 15:32:28 schwarze Exp $ */
 /*
+ * Copyright (c) 2010-2019 Ingo Schwarze <schwarze@openbsd.org>
  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
- * Copyright (c) 2010-2018 Ingo Schwarze <schwarze@openbsd.org>
  * Copyright (c) 2010, 2012 Joerg Sonnenberger <joerg@netbsd.org>
  *
  * Permission to use, copy, modify, and distribute this software for any
@@ -15,6 +15,12 @@
  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+ *
+ * Top-level functions of the mandoc(3) parser:
+ * Parser and input encoding selection, decompression,
+ * handling of input bytes, characters, lines, and files,
+ * handling of roff(7) loops and file inclusion,
+ * and steering of the various parsers.
  */
 #include "config.h"
 
@@ -36,6 +42,7 @@
 #include "mandoc_aux.h"
 #include "mandoc.h"
 #include "roff.h"
+#include "tag.h"
 #include "mdoc.h"
 #include "man.h"
 #include "mandoc_parse.h"
@@ -47,7 +54,6 @@
 struct	mparse {
 	struct roff	 *roff; /* roff parser (!NULL) */
 	struct roff_man	 *man; /* man parser */
-	char		 *sodest; /* filename pointed to by .so */
 	struct buf	 *primary; /* buffer currently being parsed */
 	struct buf	 *secondary; /* copy of top level input */
 	struct buf	 *loop; /* open .while request line */
@@ -123,15 +129,15 @@ choose_parser(struct mparse *curp)
 	}
 
 	if (format == MPARSE_MDOC) {
-		curp->man->macroset = MACROSET_MDOC;
+		curp->man->meta.macroset = MACROSET_MDOC;
 		if (curp->man->mdocmac == NULL)
 			curp->man->mdocmac = roffhash_alloc(MDOC_Dd, MDOC_MAX);
 	} else {
-		curp->man->macroset = MACROSET_MAN;
+		curp->man->meta.macroset = MACROSET_MAN;
 		if (curp->man->manmac == NULL)
 			curp->man->manmac = roffhash_alloc(MAN_TH, MAN_MAX);
 	}
-	curp->man->first->tok = TOKEN_NONE;
+	curp->man->meta.first->tok = TOKEN_NONE;
 }
 
 /*
@@ -158,7 +164,7 @@ mparse_buf_r(struct mparse *curp, struct buf blk, size
 	ln.sz = 256;
 	ln.buf = mandoc_malloc(ln.sz);
 	ln.next = NULL;
-	firstln = loop = NULL;
+	firstln = lastln = loop = NULL;
 	lnn = curp->line;
 	pos = 0;
 	inloop = 0;
@@ -256,6 +262,8 @@ mparse_buf_r(struct mparse *curp, struct buf blk, size
 		/* XXX Ugly hack to mark the end of the input. */
 
 		if (i == blk.sz || blk.buf[i] == '\0') {
+			if (pos + 2 > ln.sz)
+				resize_buf(&ln, 256);
 			ln.buf[pos++] = '\n';
 			ln.buf[pos] = '\0';
 		}
@@ -334,9 +342,9 @@ rerun:
 		case ROFF_IGN:
 			break;
 		case ROFF_CONT:
-			if (curp->man->macroset == MACROSET_NONE)
+			if (curp->man->meta.macroset == MACROSET_NONE)
 				choose_parser(curp);
-			if ((curp->man->macroset == MACROSET_MDOC ?
+			if ((curp->man->meta.macroset == MACROSET_MDOC ?
 			     mdoc_parseln(curp->man, curp->line, ln.buf, of) :
 			     man_parseln(curp->man, curp->line, ln.buf, of)
 			    ) == 2)
@@ -365,15 +373,17 @@ rerun:
 		case ROFF_SO:
 			if ( ! (curp->options & MPARSE_SO) &&
 			    (i >= blk.sz || blk.buf[i] == '\0')) {
-				curp->sodest = mandoc_strdup(ln.buf + of);
+				curp->man->meta.sodest =
+				    mandoc_strdup(ln.buf + of);
 				goto out;
 			}
 			if ((fd = mparse_open(curp, ln.buf + of)) != -1) {
 				mparse_readfd(curp, fd, ln.buf + of);
 				close(fd);
 			} else {
-				mandoc_msg(MANDOCERR_SO_FAIL, curp->line,
-				    pos, ".so %s", ln.buf + of);
+				mandoc_msg(MANDOCERR_SO_FAIL,
+				    curp->line, of, ".so %s: %s",
+				    ln.buf + of, strerror(errno));
 				ln.sz = mandoc_asprintf(&cp,
 				    ".sp\nSee the file %s.\n.sp",
 				    ln.buf + of);
@@ -428,9 +438,8 @@ read_whole_file(struct mparse *curp, int fd, struct bu
 	int		 gzerrnum, retval;
 
 	if (fstat(fd, &st) == -1) {
-		mandoc_msg(MANDOCERR_FILE, 0, 0,
-		    "fstat: %s", strerror(errno));
-		return 0;
+		mandoc_msg(MANDOCERR_FSTAT, 0, 0, "%s", strerror(errno));
+		return -1;
 	}
 
 	/*
@@ -443,13 +452,13 @@ read_whole_file(struct mparse *curp, int fd, struct bu
 	if (curp->gzip == 0 && S_ISREG(st.st_mode)) {
 		if (st.st_size > 0x7fffffff) {
 			mandoc_msg(MANDOCERR_TOOLARGE, 0, 0, NULL);
-			return 0;
+			return -1;
 		}
 		*with_mmap = 1;
 		fb->sz = (size_t)st.st_size;
 		fb->buf = mmap(NULL, fb->sz, PROT_READ, MAP_SHARED, fd, 0);
 		if (fb->buf != MAP_FAILED)
-			return 1;
+			return 0;
 	}
 
 	if (curp->gzip) {
@@ -461,15 +470,15 @@ read_whole_file(struct mparse *curp, int fd, struct bu
 		 * which this function must not do.
 		 */
 		if ((fd = dup(fd)) == -1) {
-			mandoc_msg(MANDOCERR_FILE, 0, 0,
-			    "dup: %s", strerror(errno));
-			return 0;
+			mandoc_msg(MANDOCERR_DUP, 0, 0,
+			    "%s", strerror(errno));
+			return -1;
 		}
 		if ((gz = gzdopen(fd, "rb")) == NULL) {
-			mandoc_msg(MANDOCERR_FILE, 0, 0,
-			    "gzdopen: %s", strerror(errno));
+			mandoc_msg(MANDOCERR_GZDOPEN, 0, 0,
+			    "%s", strerror(errno));
 			close(fd);
-			return 0;
+			return -1;
 		}
 	} else
 		gz = NULL;
@@ -481,7 +490,7 @@ read_whole_file(struct mparse *curp, int fd, struct bu
 
 	*with_mmap = 0;
 	off = 0;
-	retval = 0;
+	retval = -1;
 	fb->sz = 0;
 	fb->buf = NULL;
 	for (;;) {
@@ -497,13 +506,13 @@ read_whole_file(struct mparse *curp, int fd, struct bu
 		    read(fd, fb->buf + (int)off, fb->sz - off);
 		if (ssz == 0) {
 			fb->sz = off;
-			retval = 1;
+			retval = 0;
 			break;
 		}
 		if (ssz == -1) {
 			if (curp->gzip)
 				(void)gzerror(gz, &gzerrnum);
-			mandoc_msg(MANDOCERR_FILE, 0, 0, "read: %s",
+			mandoc_msg(MANDOCERR_READ, 0, 0, "%s",
 			    curp->gzip && gzerrnum != Z_ERRNO ?
 			    zError(gzerrnum) : strerror(errno));
 			break;
@@ -512,10 +521,10 @@ read_whole_file(struct mparse *curp, int fd, struct bu
 	}
 
 	if (curp->gzip && (gzerrnum = gzclose(gz)) != Z_OK)
-		mandoc_msg(MANDOCERR_FILE, 0, 0, "gzclose: %s",
+		mandoc_msg(MANDOCERR_GZCLOSE, 0, 0, "%s",
 		    gzerrnum == Z_ERRNO ? strerror(errno) :
 		    zError(gzerrnum));
-	if (retval == 0) {
+	if (retval == -1) {
 		free(fb->buf);
 		fb->buf = NULL;
 	}
@@ -525,9 +534,9 @@ read_whole_file(struct mparse *curp, int fd, struct bu
 static void
 mparse_end(struct mparse *curp)
 {
-	if (curp->man->macroset == MACROSET_NONE)
-		curp->man->macroset = MACROSET_MAN;
-	if (curp->man->macroset == MACROSET_MDOC)
+	if (curp->man->meta.macroset == MACROSET_NONE)
+		curp->man->meta.macroset = MACROSET_MAN;
+	if (curp->man->meta.macroset == MACROSET_MDOC)
 		mdoc_endparse(curp->man);
 	else
 		man_endparse(curp->man);
@@ -554,7 +563,7 @@ mparse_readfd(struct mparse *curp, int fd, const char 
 		mandoc_msg(MANDOCERR_ROFFLOOP, curp->line, 0, NULL);
 		return;
 	}
-	if (read_whole_file(curp, fd, &blk, &with_mmap) == 0)
+	if (read_whole_file(curp, fd, &blk, &with_mmap) == -1)
 		return;
 
 	/*
@@ -606,7 +615,7 @@ int
 mparse_open(struct mparse *curp, const char *file)
 {
 	char		 *cp;
-	int		  fd;
+	int		  fd, save_errno;
 
 	cp = strrchr(file, '.');
 	curp->gzip = (cp != NULL && ! strcmp(cp + 1, "gz"));
@@ -622,9 +631,11 @@ mparse_open(struct mparse *curp, const char *file)
 	 */
 
 	if ( ! curp->gzip) {
+		save_errno = errno;
 		mandoc_asprintf(&cp, "%s.gz", file);
 		fd = open(cp, O_RDONLY);
 		free(cp);
+		errno = save_errno;
 		if (fd != -1) {
 			curp->gzip = 1;
 			return fd;
@@ -633,7 +644,6 @@ mparse_open(struct mparse *curp, const char *file)
 
 	/* Neither worked, give up. */
 
-	mandoc_msg(MANDOCERR_FILE, 0, 0, "%s", strerror(errno));
 	return -1;
 }
 
@@ -647,20 +657,21 @@ mparse_alloc(int options, enum mandoc_os os_e, const c
 	curp->options = options;
 	curp->os_s = os_s;
 
-	curp->roff = roff_alloc(curp, options);
-	curp->man = roff_man_alloc(curp->roff, curp, curp->os_s,
+	curp->roff = roff_alloc(options);
+	curp->man = roff_man_alloc(curp->roff, curp->os_s,
 		curp->options & MPARSE_QUICK ? 1 : 0);
 	if (curp->options & MPARSE_MDOC) {
-		curp->man->macroset = MACROSET_MDOC;
+		curp->man->meta.macroset = MACROSET_MDOC;
 		if (curp->man->mdocmac == NULL)
 			curp->man->mdocmac = roffhash_alloc(MDOC_Dd, MDOC_MAX);
 	} else if (curp->options & MPARSE_MAN) {
-		curp->man->macroset = MACROSET_MAN;
+		curp->man->meta.macroset = MACROSET_MAN;
 		if (curp->man->manmac == NULL)
 			curp->man->manmac = roffhash_alloc(MAN_TH, MAN_MAX);
 	}
-	curp->man->first->tok = TOKEN_NONE;
+	curp->man->meta.first->tok = TOKEN_NONE;
 	curp->man->meta.os_e = os_e;
+	tag_alloc();
 	return curp;
 }
 
@@ -671,34 +682,32 @@ mparse_reset(struct mparse *curp)
 	roff_man_reset(curp->man);
 	free_buf_list(curp->secondary);
 	curp->secondary = NULL;
-	free(curp->sodest);
-	curp->sodest = NULL;
 	curp->gzip = 0;
 }
 
 void
 mparse_free(struct mparse *curp)
 {
+	tag_free();
 	roffhash_free(curp->man->mdocmac);
 	roffhash_free(curp->man->manmac);
 	roff_man_free(curp->man);
 	roff_free(curp->roff);
 	free_buf_list(curp->secondary);
-	free(curp->sodest);
 	free(curp);
 }
 
-void
-mparse_result(struct mparse *curp, struct roff_man **man,
-	char **sodest)
+struct roff_meta *
+mparse_result(struct mparse *curp)
 {
-
-	if (sodest && NULL != (*sodest = curp->sodest)) {
-		*man = NULL;
-		return;
+	roff_state_reset(curp->man);
+	if (curp->options & MPARSE_VALIDATE) {
+		if (curp->man->meta.macroset == MACROSET_MDOC)
+			mdoc_validate(curp->man);
+		else
+			man_validate(curp->man);
 	}
-	if (man)
-		*man = curp->man;
+	return &curp->man->meta;
 }
 
 void