add a style message about overlong text lines,

trying very hard to avoid false positives,
not at all trying to catch as many cases as possible;

feature originally suggested by tb@,
OK tb@ kn@ jmc@

6 files changed, 24 insertions, 8 deletions

diff --git a/libmandoc.h b/libmandoc.h
index 33489c9a..ab7c29be 100644
--- a/libmandoc.h
+++ b/libmandoc.h
@@ -1,4 +1,4 @@
-/* $Id: libmandoc.h,v 1.79 2020/04/03 11:35:01 schwarze Exp $ */
+/* $Id: libmandoc.h,v 1.80 2021/06/27 17:57:54 schwarze Exp $ */
 /*
  * Copyright (c) 2013-2015,2017,2018,2020 Ingo Schwarze <schwarze@openbsd.org>
  * Copyright (c) 2009, 2010, 2011, 2012 Kristaps Dzonsons <kristaps@bsd.lv>
@@ -73,7 +73,7 @@ void		 roff_reset(struct roff *);
 void		 roff_man_free(struct roff_man *);
 struct roff_man	*roff_man_alloc(struct roff *, const char *, int);
 void		 roff_man_reset(struct roff_man *);
-int		 roff_parseln(struct roff *, int, struct buf *, int *);
+int		 roff_parseln(struct roff *, int, struct buf *, int *, size_t);
 void		 roff_userret(struct roff *);
 void		 roff_endparse(struct roff *);
 void		 roff_setreg(struct roff *, const char *, int, char);
diff --git a/mandoc.1 b/mandoc.1
index f6b3edc3..1fe95806 100644
--- a/mandoc.1
+++ b/mandoc.1
@@ -15,7 +15,7 @@
 .\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
 .\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
 .\"
-.Dd $Mdocdate: June 2 2021 $
+.Dd $Mdocdate: June 27 2021 $
 .Dt MANDOC 1
 .Os
 .Sh NAME
@@ -1066,6 +1066,9 @@ An
 request occurs even though the document already switched to no-fill mode
 and did not switch back to fill mode yet.
 It has no effect.
+.It Sy "input text line longer than 80 bytes"
+Consider breaking the input text line
+at one of the blank characters before column 80.
 .It Sy "verbatim \(dq--\(dq, maybe consider using \e(em"
 .Pq mdoc
 Even though the ASCII output device renders an em-dash as
diff --git a/mandoc.h b/mandoc.h
index 4d80b6b5..3f68e9d4 100644
--- a/mandoc.h
+++ b/mandoc.h
@@ -1,4 +1,4 @@
-/* $Id: mandoc.h,v 1.270 2021/06/02 18:28:19 schwarze Exp $ */
+/* $Id: mandoc.h,v 1.271 2021/06/27 17:57:54 schwarze Exp $ */
 /*
  * Copyright (c) 2012-2020 Ingo Schwarze <schwarze@openbsd.org>
  * Copyright (c) 2010, 2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
@@ -72,6 +72,7 @@ enum	mandocerr {
 	MANDOCERR_DELIM_NB, /* no blank before trailing delimiter: macro ... */
 	MANDOCERR_FI_SKIP, /* fill mode already enabled, skipping: fi */
 	MANDOCERR_NF_SKIP, /* fill mode already disabled, skipping: nf */
+	MANDOCERR_TEXT_LONG, /* input text line longer than 80 bytes */
 	MANDOCERR_DASHDASH, /* verbatim "--", maybe consider using \(em */
 	MANDOCERR_FUNC, /* function name without markup: name() */
 	MANDOCERR_SPACE_EOL, /* whitespace at end of input line */
diff --git a/mandoc_msg.c b/mandoc_msg.c
index 7a127331..64355ddc 100644
--- a/mandoc_msg.c
+++ b/mandoc_msg.c
@@ -73,6 +73,7 @@ static	const char *const type_message[MANDOCERR_MAX] = {
 	"no blank before trailing delimiter",
 	"fill mode already enabled, skipping",
 	"fill mode already disabled, skipping",
+	"input text line longer than 80 bytes",
 	"verbatim \"--\", maybe consider using \\(em",
 	"function name without markup",
 	"whitespace at end of input line",
diff --git a/read.c b/read.c
index 03a71c33..5b33edbe 100644
--- a/read.c
+++ b/read.c
@@ -1,4 +1,4 @@
-/* $Id: read.c,v 1.219 2020/04/24 12:02:33 schwarze Exp $ */
+/* $Id: read.c,v 1.220 2021/06/27 17:57:54 schwarze Exp $ */
 /*
  * Copyright (c) 2010-2020 Ingo Schwarze <schwarze@openbsd.org>
  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
@@ -154,6 +154,7 @@ mparse_buf_r(struct mparse *curp, struct buf blk, size_t i, int start)
 	struct buf	*firstln, *lastln, *thisln, *loop;
 	char		*cp;
 	size_t		 pos; /* byte number in the ln buffer */
+	size_t		 spos; /* at the start of the current line parse */
 	int		 line_result, result;
 	int		 of;
 	int		 lnn; /* line number in the real file */
@@ -180,6 +181,7 @@ mparse_buf_r(struct mparse *curp, struct buf blk, size_t i, int start)
 			    curp->filenc & MPARSE_LATIN1)
 				curp->filenc = preconv_cue(&blk, i);
 		}
+		spos = pos;
 
 		while (i < blk.sz && (start || blk.buf[i] != '\0')) {
 
@@ -279,7 +281,8 @@ mparse_buf_r(struct mparse *curp, struct buf blk, size_t i, int start)
 
 		of = 0;
 rerun:
-		line_result = roff_parseln(curp->roff, curp->line, &ln, &of);
+		line_result = roff_parseln(curp->roff, curp->line,
+		    &ln, &of, start && spos == 0 ? pos : 0);
 
 		/* Process options. */
 
diff --git a/roff.c b/roff.c
index 36c87b61..ed4f31b6 100644
--- a/roff.c
+++ b/roff.c
@@ -1,4 +1,4 @@
-/* $Id: roff.c,v 1.376 2020/08/27 12:59:02 schwarze Exp $ */
+/* $Id: roff.c,v 1.377 2021/06/27 17:57:55 schwarze Exp $ */
 /*
  * Copyright (c) 2010-2015, 2017-2020 Ingo Schwarze <schwarze@openbsd.org>
  * Copyright (c) 2008-2012, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
@@ -1823,7 +1823,7 @@ roff_parsetext(struct roff *r, struct buf *buf, int pos, int *offs)
 }
 
 int
-roff_parseln(struct roff *r, int ln, struct buf *buf, int *offs)
+roff_parseln(struct roff *r, int ln, struct buf *buf, int *offs, size_t len)
 {
 	enum roff_tok	 t;
 	int		 e;
@@ -1834,6 +1834,14 @@ roff_parseln(struct roff *r, int ln, struct buf *buf, int *offs)
 
 	ppos = pos = *offs;
 
+	if (len > 80 && r->tbl == NULL && r->eqn == NULL &&
+	    (r->man->flags & ROFF_NOFILL) == 0 &&
+	    strchr(" .\\", buf->buf[pos]) == NULL &&
+	    buf->buf[pos] != r->control &&
+	    strcspn(buf->buf, " ") < 80)
+		mandoc_msg(MANDOCERR_TEXT_LONG, ln, (int)len - 1,
+		    "%.20s...", buf->buf + pos);
+
 	/* Handle in-line equation delimiters. */
 
 	if (r->tbl == NULL &&