With the improved escape sequence parser, it becomes easy to also improve

diagnostics.  Distinguish "incomplete escape sequence", "invalid special
character", and "unknown special character" from the generic "invalid
escape sequence", also promoting them from WARNING to ERROR because
incomplete escape sequences are severe syntax violations and because
encountering an invalid or unknown special character makes it likely
that part of the document content intended by the authors gets lost.

diff --git a/regress/usr.bin/mandoc/char/accent/nocombine.out_lint b/regress/usr.bin/mandoc/char/accent/nocombine.out_lint
index c9de416..0f7be4d 100644 (file)
--- a/regress/usr.bin/mandoc/char/accent/nocombine.out_lint
+++ b/regress/usr.bin/mandoc/char/accent/nocombine.out_lint
@@ -1,2 +1,2 @@
-mandoc: nocombine.in:8:27: WARNING: invalid escape sequence: \[']
-mandoc: nocombine.in:14:27: WARNING: invalid escape sequence: \[`]
+mandoc: nocombine.in:8:27: ERROR: invalid special character: \[']
+mandoc: nocombine.in:14:27: ERROR: invalid special character: \[`]

diff --git a/regress/usr.bin/mandoc/char/space/invalid.out_lint b/regress/usr.bin/mandoc/char/space/invalid.out_lint
index c05ef38..4c14685 100644 (file)
--- a/regress/usr.bin/mandoc/char/space/invalid.out_lint
+++ b/regress/usr.bin/mandoc/char/space/invalid.out_lint
@@ -1,9 +1,9 @@
 mandoc: invalid.in:7:15: WARNING: invalid escape sequence: \[ 
-mandoc: invalid.in:8:14: WARNING: invalid escape sequence: \[%]
-mandoc: invalid.in:9:16: WARNING: invalid escape sequence: \[&]
-mandoc: invalid.in:10:12: WARNING: invalid escape sequence: \[:]
-mandoc: invalid.in:11:12: WARNING: invalid escape sequence: \[^]
-mandoc: invalid.in:12:16: WARNING: invalid escape sequence: \[_]
-mandoc: invalid.in:13:11: WARNING: invalid escape sequence: \[|]
-mandoc: invalid.in:14:12: WARNING: invalid escape sequence: \[~]
-mandoc: invalid.in:15:18: WARNING: invalid escape sequence: \[0]
+mandoc: invalid.in:8:14: ERROR: invalid special character: \[%]
+mandoc: invalid.in:9:16: ERROR: invalid special character: \[&]
+mandoc: invalid.in:10:12: ERROR: invalid special character: \[:]
+mandoc: invalid.in:11:12: ERROR: invalid special character: \[^]
+mandoc: invalid.in:12:16: ERROR: invalid special character: \[_]
+mandoc: invalid.in:13:11: ERROR: invalid special character: \[|]
+mandoc: invalid.in:14:12: ERROR: invalid special character: \[~]
+mandoc: invalid.in:15:18: ERROR: invalid special character: \[0]

diff --git a/regress/usr.bin/mandoc/char/unicode/input.out_lint b/regress/usr.bin/mandoc/char/unicode/input.out_lint
index 578a770..fa36f87 100644 (file)
--- a/regress/usr.bin/mandoc/char/unicode/input.out_lint
+++ b/regress/usr.bin/mandoc/char/unicode/input.out_lint
@@ -24,11 +24,11 @@ mandoc: input.in:35:19: ERROR: skipping bad character: 0xbf
 mandoc: input.in:42:25: ERROR: skipping bad character: 0xed
 mandoc: input.in:42:26: ERROR: skipping bad character: 0xa0
 mandoc: input.in:42:27: ERROR: skipping bad character: 0x80
-mandoc: input.in:42:17: WARNING: invalid escape sequence: \[uD800]
+mandoc: input.in:42:17: ERROR: invalid special character: \[uD800]
 mandoc: input.in:43:25: ERROR: skipping bad character: 0xed
 mandoc: input.in:43:26: ERROR: skipping bad character: 0xbf
 mandoc: input.in:43:27: ERROR: skipping bad character: 0xbf
-mandoc: input.in:43:17: WARNING: invalid escape sequence: \[uDFFF]
+mandoc: input.in:43:17: ERROR: invalid special character: \[uDFFF]
 mandoc: input.in:53:19: ERROR: skipping bad character: 0xf0
 mandoc: input.in:53:20: ERROR: skipping bad character: 0x80
 mandoc: input.in:53:21: ERROR: skipping bad character: 0x80
@@ -57,25 +57,25 @@ mandoc: input.in:67:31: ERROR: skipping bad character: 0xf4
 mandoc: input.in:67:32: ERROR: skipping bad character: 0x90
 mandoc: input.in:67:33: ERROR: skipping bad character: 0x80
 mandoc: input.in:67:34: ERROR: skipping bad character: 0x80
-mandoc: input.in:67:21: WARNING: invalid escape sequence: \[u110000]
+mandoc: input.in:67:21: ERROR: invalid special character: \[u110000]
 mandoc: input.in:68:31: ERROR: skipping bad character: 0xf4
 mandoc: input.in:68:32: ERROR: skipping bad character: 0xbf
 mandoc: input.in:68:33: ERROR: skipping bad character: 0xbf
 mandoc: input.in:68:34: ERROR: skipping bad character: 0xbf
-mandoc: input.in:68:21: WARNING: invalid escape sequence: \[u13FFFF]
+mandoc: input.in:68:21: ERROR: invalid special character: \[u13FFFF]
 mandoc: input.in:69:31: ERROR: skipping bad character: 0xf5
 mandoc: input.in:69:32: ERROR: skipping bad character: 0x80
 mandoc: input.in:69:33: ERROR: skipping bad character: 0x80
 mandoc: input.in:69:34: ERROR: skipping bad character: 0x80
-mandoc: input.in:69:21: WARNING: invalid escape sequence: \[u140000]
+mandoc: input.in:69:21: ERROR: invalid special character: \[u140000]
 mandoc: input.in:70:31: ERROR: skipping bad character: 0xf7
 mandoc: input.in:70:32: ERROR: skipping bad character: 0xbf
 mandoc: input.in:70:33: ERROR: skipping bad character: 0xbf
 mandoc: input.in:70:34: ERROR: skipping bad character: 0xbf
-mandoc: input.in:70:21: WARNING: invalid escape sequence: \[u1FFFFF]
+mandoc: input.in:70:21: ERROR: invalid special character: \[u1FFFFF]
 mandoc: input.in:71:33: ERROR: skipping bad character: 0xf8
 mandoc: input.in:71:34: ERROR: skipping bad character: 0x88
 mandoc: input.in:71:35: ERROR: skipping bad character: 0x80
 mandoc: input.in:71:36: ERROR: skipping bad character: 0x80
 mandoc: input.in:71:37: ERROR: skipping bad character: 0x80
-mandoc: input.in:71:23: WARNING: invalid escape sequence: \[u200000]
+mandoc: input.in:71:23: ERROR: invalid special character: \[u200000]

diff --git a/regress/usr.bin/mandoc/char/unicode/invalid.out_lint b/regress/usr.bin/mandoc/char/unicode/invalid.out_lint
index ce9de98..0717605 100644 (file)
--- a/regress/usr.bin/mandoc/char/unicode/invalid.out_lint
+++ b/regress/usr.bin/mandoc/char/unicode/invalid.out_lint
@@ -1,9 +1,9 @@
-mandoc: invalid.in:11:13: WARNING: invalid escape sequence: \[u2B]
-mandoc: invalid.in:11:20: WARNING: invalid escape sequence: \[u02B]
-mandoc: invalid.in:13:12: WARNING: invalid escape sequence: \[u0002B]
-mandoc: invalid.in:13:22: WARNING: invalid escape sequence: \[u00002B]
-mandoc: invalid.in:13:33: WARNING: invalid escape sequence: \[u000002B]
-mandoc: invalid.in:14:13: WARNING: invalid escape sequence: \[u110000]
-mandoc: invalid.in:14:24: WARNING: invalid escape sequence: \[u200000]
-mandoc: invalid.in:14:35: WARNING: invalid escape sequence: \[u1000000]
-mandoc: invalid.in:15:20: WARNING: invalid escape sequence: \[u1234g]
+mandoc: invalid.in:11:13: ERROR: unknown special character: \[u2B]
+mandoc: invalid.in:11:20: ERROR: unknown special character: \[u02B]
+mandoc: invalid.in:13:12: ERROR: invalid special character: \[u0002B]
+mandoc: invalid.in:13:22: ERROR: invalid special character: \[u00002B]
+mandoc: invalid.in:13:33: ERROR: unknown special character: \[u000002B]
+mandoc: invalid.in:14:13: ERROR: invalid special character: \[u110000]
+mandoc: invalid.in:14:24: ERROR: invalid special character: \[u200000]
+mandoc: invalid.in:14:35: ERROR: unknown special character: \[u1000000]
+mandoc: invalid.in:15:20: ERROR: unknown special character: \[u1234g]

diff --git a/regress/usr.bin/mandoc/roff/char/badarg.out_lint b/regress/usr.bin/mandoc/roff/char/badarg.out_lint
index e07faa3..dd5c799 100644 (file)
--- a/regress/usr.bin/mandoc/roff/char/badarg.out_lint
+++ b/regress/usr.bin/mandoc/roff/char/badarg.out_lint
@@ -1,6 +1,6 @@
 mandoc: badarg.in:6:6: ERROR: argument is not a character: char 
 mandoc: badarg.in:7:7: ERROR: argument is not a character: char \fR myval
-mandoc: badarg.in:8:7: WARNING: invalid escape sequence: \[myc]
+mandoc: badarg.in:8:7: ERROR: unknown special character: \[myc]
 mandoc: badarg.in:8:7: ERROR: argument is not a character: char \[myc]x myval
 mandoc: badarg.in:9:7: ERROR: argument is not a character: char xy myval
-mandoc: badarg.in:10:7: WARNING: invalid escape sequence: \[myc]
+mandoc: badarg.in:10:7: ERROR: unknown special character: \[myc]

diff --git a/regress/usr.bin/mandoc/roff/esc/B.out_lint b/regress/usr.bin/mandoc/roff/esc/B.out_lint
index f52270a..fbaded6 100644 (file)
--- a/regress/usr.bin/mandoc/roff/esc/B.out_lint
+++ b/regress/usr.bin/mandoc/roff/esc/B.out_lint
@@ -1 +1 @@
-mandoc: B.in:37:23: WARNING: invalid escape sequence: \B'1+1
+mandoc: B.in:37:23: ERROR: incomplete escape sequence: \B'1+1

diff --git a/regress/usr.bin/mandoc/roff/esc/ignore.out_lint b/regress/usr.bin/mandoc/roff/esc/ignore.out_lint
index 16d64a9..ddbc8d6 100644 (file)
--- a/regress/usr.bin/mandoc/roff/esc/ignore.out_lint
+++ b/regress/usr.bin/mandoc/roff/esc/ignore.out_lint
@@ -1,14 +1,14 @@
-mandoc: ignore.in:7:36: WARNING: invalid escape sequence: \[%]
-mandoc: ignore.in:8:35: WARNING: invalid escape sequence: \[&]
-mandoc: ignore.in:9:51: WARNING: invalid escape sequence: \[)]
-mandoc: ignore.in:10:37: WARNING: invalid escape sequence: \[,]
-mandoc: ignore.in:11:38: WARNING: invalid escape sequence: \[/]
-mandoc: ignore.in:12:28: WARNING: invalid escape sequence: \[^]
-mandoc: ignore.in:13:17: WARNING: invalid escape sequence: \[a]
-mandoc: ignore.in:14:25: WARNING: invalid escape sequence: \[d]
-mandoc: ignore.in:15:25: WARNING: invalid escape sequence: \[t]
-mandoc: ignore.in:16:33: WARNING: invalid escape sequence: \[u]
-mandoc: ignore.in:17:20: WARNING: invalid escape sequence: \[{]
-mandoc: ignore.in:18:25: WARNING: invalid escape sequence: \[|]
-mandoc: ignore.in:19:20: WARNING: invalid escape sequence: \[}]
-mandoc: ignore.in:23:56: WARNING: invalid escape sequence: \s-
+mandoc: ignore.in:7:36: ERROR: invalid special character: \[%]
+mandoc: ignore.in:8:35: ERROR: invalid special character: \[&]
+mandoc: ignore.in:9:51: ERROR: invalid special character: \[)]
+mandoc: ignore.in:10:37: ERROR: invalid special character: \[,]
+mandoc: ignore.in:11:38: ERROR: invalid special character: \[/]
+mandoc: ignore.in:12:28: ERROR: invalid special character: \[^]
+mandoc: ignore.in:13:17: ERROR: invalid special character: \[a]
+mandoc: ignore.in:14:25: ERROR: invalid special character: \[d]
+mandoc: ignore.in:15:25: ERROR: invalid special character: \[t]
+mandoc: ignore.in:16:33: ERROR: invalid special character: \[u]
+mandoc: ignore.in:17:20: ERROR: invalid special character: \[{]
+mandoc: ignore.in:18:25: ERROR: invalid special character: \[|]
+mandoc: ignore.in:19:20: ERROR: invalid special character: \[}]
+mandoc: ignore.in:23:56: ERROR: incomplete escape sequence: \s-

diff --git a/regress/usr.bin/mandoc/roff/esc/invalid.out_lint b/regress/usr.bin/mandoc/roff/esc/invalid.out_lint
index baef317..ff52893 100644 (file)
--- a/regress/usr.bin/mandoc/roff/esc/invalid.out_lint
+++ b/regress/usr.bin/mandoc/roff/esc/invalid.out_lint
@@ -1,43 +1,43 @@
 mandoc: invalid.in:7:8: WARNING: undefined escape, printing literally: \+
-mandoc: invalid.in:7:11: WARNING: invalid escape sequence: \[+]
+mandoc: invalid.in:7:11: ERROR: invalid special character: \[+]
 mandoc: invalid.in:8:13: WARNING: undefined escape, printing literally: \;
-mandoc: invalid.in:8:16: WARNING: invalid escape sequence: \[;]
+mandoc: invalid.in:8:16: ERROR: invalid special character: \[;]
 mandoc: invalid.in:9:13: WARNING: undefined escape, printing literally: \<
-mandoc: invalid.in:9:16: WARNING: invalid escape sequence: \[<]
+mandoc: invalid.in:9:16: ERROR: invalid special character: \[<]
 mandoc: invalid.in:10:12: WARNING: undefined escape, printing literally: \=
-mandoc: invalid.in:10:15: WARNING: invalid escape sequence: \[=]
+mandoc: invalid.in:10:15: ERROR: invalid special character: \[=]
 mandoc: invalid.in:11:16: WARNING: undefined escape, printing literally: \>
-mandoc: invalid.in:11:19: WARNING: invalid escape sequence: \[>]
+mandoc: invalid.in:11:19: ERROR: invalid special character: \[>]
 mandoc: invalid.in:12:6: WARNING: undefined escape, printing literally: \@
-mandoc: invalid.in:12:9: WARNING: invalid escape sequence: \[@]
+mandoc: invalid.in:12:9: ERROR: invalid special character: \[@]
 mandoc: invalid.in:13:18: WARNING: undefined escape, printing literally: \]
-mandoc: invalid.in:14:16: WARNING: invalid escape sequence: \[{]
-mandoc: invalid.in:14:21: WARNING: invalid escape sequence: \[}]
+mandoc: invalid.in:14:16: ERROR: invalid special character: \[{]
+mandoc: invalid.in:14:21: ERROR: invalid special character: \[}]
 mandoc: invalid.in:15:9: WARNING: undefined escape, printing literally: \1
-mandoc: invalid.in:15:12: WARNING: invalid escape sequence: \[1]
+mandoc: invalid.in:15:12: ERROR: invalid special character: \[1]
 mandoc: invalid.in:16:5: WARNING: undefined escape, printing literally: \G
-mandoc: invalid.in:16:8: WARNING: invalid escape sequence: \[G]
+mandoc: invalid.in:16:8: ERROR: invalid special character: \[G]
 mandoc: invalid.in:17:5: WARNING: undefined escape, printing literally: \I
-mandoc: invalid.in:17:8: WARNING: invalid escape sequence: \[I]
+mandoc: invalid.in:17:8: ERROR: invalid special character: \[I]
 mandoc: invalid.in:18:5: WARNING: undefined escape, printing literally: \i
-mandoc: invalid.in:18:8: WARNING: invalid escape sequence: \[i]
+mandoc: invalid.in:18:8: ERROR: invalid special character: \[i]
 mandoc: invalid.in:19:5: WARNING: undefined escape, printing literally: \J
-mandoc: invalid.in:19:8: WARNING: invalid escape sequence: \[J]
+mandoc: invalid.in:19:8: ERROR: invalid special character: \[J]
 mandoc: invalid.in:20:5: WARNING: undefined escape, printing literally: \j
-mandoc: invalid.in:20:8: WARNING: invalid escape sequence: \[j]
+mandoc: invalid.in:20:8: ERROR: invalid special character: \[j]
 mandoc: invalid.in:21:5: WARNING: undefined escape, printing literally: \K
-mandoc: invalid.in:21:8: WARNING: invalid escape sequence: \[K]
+mandoc: invalid.in:21:8: ERROR: invalid special character: \[K]
 mandoc: invalid.in:22:5: WARNING: undefined escape, printing literally: \P
-mandoc: invalid.in:22:8: WARNING: invalid escape sequence: \[P]
+mandoc: invalid.in:22:8: ERROR: invalid special character: \[P]
 mandoc: invalid.in:23:5: WARNING: undefined escape, printing literally: \Q
-mandoc: invalid.in:23:8: WARNING: invalid escape sequence: \[Q]
+mandoc: invalid.in:23:8: ERROR: invalid special character: \[Q]
 mandoc: invalid.in:24:5: WARNING: undefined escape, printing literally: \q
-mandoc: invalid.in:24:8: WARNING: invalid escape sequence: \[q]
+mandoc: invalid.in:24:8: ERROR: invalid special character: \[q]
 mandoc: invalid.in:25:5: WARNING: undefined escape, printing literally: \T
-mandoc: invalid.in:25:8: WARNING: invalid escape sequence: \[T]
+mandoc: invalid.in:25:8: ERROR: invalid special character: \[T]
 mandoc: invalid.in:26:5: WARNING: undefined escape, printing literally: \U
-mandoc: invalid.in:26:8: WARNING: invalid escape sequence: \[U]
+mandoc: invalid.in:26:8: ERROR: invalid special character: \[U]
 mandoc: invalid.in:27:5: WARNING: undefined escape, printing literally: \W
-mandoc: invalid.in:27:8: WARNING: invalid escape sequence: \[W]
+mandoc: invalid.in:27:8: ERROR: invalid special character: \[W]
 mandoc: invalid.in:28:5: WARNING: undefined escape, printing literally: \y
-mandoc: invalid.in:28:8: WARNING: invalid escape sequence: \[y]
+mandoc: invalid.in:28:8: ERROR: invalid special character: \[y]

diff --git a/regress/usr.bin/mandoc/roff/esc/unsupp.out_lint b/regress/usr.bin/mandoc/roff/esc/unsupp.out_lint
index fae97c8..db7631c 100644 (file)
--- a/regress/usr.bin/mandoc/roff/esc/unsupp.out_lint
+++ b/regress/usr.bin/mandoc/roff/esc/unsupp.out_lint
@@ -1,5 +1,5 @@
 mandoc: unsupp.in:7:20: UNSUPP: unsupported escape sequence: \!
-mandoc: unsupp.in:7:23: WARNING: invalid escape sequence: \[!]
+mandoc: unsupp.in:7:23: ERROR: invalid special character: \[!]
 mandoc: unsupp.in:8:17: UNSUPP: unsupported escape sequence: \?
 mandoc: unsupp.in:8:21: UNSUPP: unsupported escape sequence: \?
-mandoc: unsupp.in:8:24: WARNING: invalid escape sequence: \[?]
+mandoc: unsupp.in:8:24: ERROR: invalid special character: \[?]

diff --git a/regress/usr.bin/mandoc/roff/esc/w.out_lint b/regress/usr.bin/mandoc/roff/esc/w.out_lint
index d48495c..11dfbef 100644 (file)
--- a/regress/usr.bin/mandoc/roff/esc/w.out_lint
+++ b/regress/usr.bin/mandoc/roff/esc/w.out_lint
@@ -1 +1 @@
-mandoc: w.in:17:15: WARNING: invalid escape sequence: \w'foo
+mandoc: w.in:17:15: ERROR: incomplete escape sequence: \w'foo

diff --git a/regress/usr.bin/mandoc/roff/nr/escname.out_lint b/regress/usr.bin/mandoc/roff/nr/escname.out_lint
index edec17f..8655da9 100644 (file)
--- a/regress/usr.bin/mandoc/roff/nr/escname.out_lint
+++ b/regress/usr.bin/mandoc/roff/nr/escname.out_lint
@@ -1,5 +1,5 @@
 mandoc: escname.in:9:5: ERROR: escaped character not allowed in a name: first\e
 mandoc: escname.in:11:10: WARNING: undefined escape, printing literally: \G
 mandoc: escname.in:14:5: ERROR: escaped character not allowed in a name: first\e
-mandoc: escname.in:20:13: WARNING: invalid escape sequence: \n[second
+mandoc: escname.in:20:13: ERROR: incomplete escape sequence: \n[second
 mandoc: escname.in:20:12: STYLE: whitespace at end of input line

diff --git a/regress/usr.bin/mandoc/roff/string/name.out_lint b/regress/usr.bin/mandoc/roff/string/name.out_lint
index 13283f5..55f1f87 100644 (file)
--- a/regress/usr.bin/mandoc/roff/string/name.out_lint
+++ b/regress/usr.bin/mandoc/roff/string/name.out_lint
@@ -1,6 +1,6 @@
 mandoc: name.in:11:5: ERROR: escaped character not allowed in a name: bs\e
 mandoc: name.in:14:5: ERROR: escaped character not allowed in a name: bl\ 
-mandoc: name.in:18:29: WARNING: invalid escape sequence: \*[norm
+mandoc: name.in:18:29: ERROR: incomplete escape sequence: \*[norm
 mandoc: name.in:18:28: STYLE: whitespace at end of input line
 mandoc: name.in:20:7: WARNING: undefined string, using "": quot
 mandoc: name.in:20:6: STYLE: whitespace at end of input line

diff --git a/usr.bin/mandoc/mandoc.1 b/usr.bin/mandoc/mandoc.1
index 1cd1621..ba7e883 100644 (file)
--- a/usr.bin/mandoc/mandoc.1
+++ b/usr.bin/mandoc/mandoc.1
@@ -1,4 +1,4 @@
-.\" $OpenBSD: mandoc.1,v 1.184 2022/04/28 16:16:46 schwarze Exp $
+.\" $OpenBSD: mandoc.1,v 1.185 2022/06/05 13:42:49 schwarze Exp $
 .\"
 .\" Copyright (c) 2012, 2014-2022 Ingo Schwarze <schwarze@openbsd.org>
 .\" Copyright (c) 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
@@ -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: April 28 2022 $
+.Dd $Mdocdate: June 5 2022 $
 .Dt MANDOC 1
 .Os
 .Sh NAME
@@ -1799,21 +1799,9 @@ A new sentence starts in the middle of a text line.
 Start it on a new input line to help formatters produce correct spacing.
 .It Sy "invalid escape sequence"
 .Pq roff
-An escape sequence has an invalid opening argument delimiter, lacks the
-closing argument delimiter, the argument is of an invalid form, or it is
-a character escape sequence with an invalid name.
-If the argument is incomplete,
-.Ic \e*
-and
-.Ic \en
-expand to an empty string,
-.Ic \eB
-to the digit
-.Sq 0 ,
-and
-.Ic \ew
-to the length of the incomplete argument.
-All other invalid escape sequences are ignored.
+An escape sequence has an invalid opening argument delimiter
+or the argument is of an invalid form.
+Invalid escape sequences are ignored.
 .It Sy "undefined escape, printing literally"
 .Pq roff
 In an escape sequence, the first character
@@ -2285,6 +2273,48 @@ with invalid arguments
 .El
 The excess arguments are ignored.
 .El
+.Ss "Errors related to escape sequences"
+.Bl -ohang
+.It Sy "incomplete escape sequence"
+.Pq roff
+The end of the input line is encountered
+while parsing the argument of an escape sequence.
+In this case,
+.Ic \e*
+and
+.Ic \en
+expand to an empty string,
+.Ic \eB
+to the digit
+.Sq 0 ,
+and
+.Ic \ew
+to the length of the incomplete argument.
+All other incomplete escape sequences are ignored.
+.It Sy "invalid special character"
+.Pq roff
+A special character escape sequence is invalid,
+for example a Unicode sequence pointing to a surrogate
+or beyond the Unicode range, a \e[char...] escape sequence
+representing a control character or pointing beyond the
+.Vt unsigned char
+range, or an invalid variable-length form
+of a single-byte character escape sequence, for example writing
+.Qq \e[e]
+or
+.Qq \e[~]
+instead of
+.Qq \ee
+or
+.Qq \e~ ,
+respectively.
+The escape sequence is ignored.
+.It Sy "unknown special character"
+.Pq roff
+The name given in a special character escape sequence is not known to
+.Nm .
+The escape sequence is ignored.
+.El
 .Ss Unsupported features
 .Bl -ohang
 .It Sy "input too large"

diff --git a/usr.bin/mandoc/mandoc.h b/usr.bin/mandoc/mandoc.h
index 6a53a29..012bb2d 100644 (file)
--- a/usr.bin/mandoc/mandoc.h
+++ b/usr.bin/mandoc/mandoc.h
@@ -1,4 +1,4 @@
-/* $OpenBSD: mandoc.h,v 1.219 2022/05/19 15:17:50 schwarze Exp $ */
+/* $OpenBSD: mandoc.h,v 1.220 2022/06/05 13:42:49 schwarze Exp $ */
 /*
  * Copyright (c) 2012-2022 Ingo Schwarze <schwarze@openbsd.org>
  * Copyright (c) 2010, 2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
@@ -235,6 +235,11 @@ enum       mandocerr {
        MANDOCERR_ARG_EXCESS, /* skipping excess arguments: macro ... args */
        MANDOCERR_DIVZERO, /* divide by zero */
 
+       /* related to escape sequences */
+       MANDOCERR_ESC_INCOMPLETE, /* incomplete escape sequence: esc */
+       MANDOCERR_ESC_BADCHAR, /* invalid special character: esc */
+       MANDOCERR_ESC_UNKCHAR, /* unknown special character: esc */
+
        MANDOCERR_UNSUPP, /* ===== start of unsupported features ===== */
 
        MANDOCERR_TOOLARGE, /* input too large */

diff --git a/usr.bin/mandoc/mandoc_msg.c b/usr.bin/mandoc/mandoc_msg.c
index 82140af..bb364f9 100644 (file)
--- a/usr.bin/mandoc/mandoc_msg.c
+++ b/usr.bin/mandoc/mandoc_msg.c
@@ -1,4 +1,4 @@
-/* $OpenBSD: mandoc_msg.c,v 1.16 2022/04/28 16:16:46 schwarze Exp $ */
+/* $OpenBSD: mandoc_msg.c,v 1.17 2022/06/05 13:42:49 schwarze Exp $ */
 /*
  * Copyright (c) 2014-2022 Ingo Schwarze <schwarze@openbsd.org>
  * Copyright (c) 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
@@ -234,6 +234,11 @@ static     const char *const type_message[MANDOCERR_MAX] = {
        "skipping excess arguments",
        "divide by zero",
 
+       /* related to escape sequences */
+       "incomplete escape sequence",
+       "invalid special character",
+       "unknown special character",
+
        "unsupported feature",
        "input too large",
        "unsupported control character",

diff --git a/usr.bin/mandoc/roff_escape.c b/usr.bin/mandoc/roff_escape.c
index 1c40cd5..371ea61 100644 (file)
--- a/usr.bin/mandoc/roff_escape.c
+++ b/usr.bin/mandoc/roff_escape.c
@@ -1,4 +1,4 @@
-/* $OpenBSD: roff_escape.c,v 1.9 2022/06/05 10:19:47 schwarze Exp $ */
+/* $OpenBSD: roff_escape.c,v 1.10 2022/06/05 13:42:49 schwarze Exp $ */
 /*
  * Copyright (c) 2011, 2012, 2013, 2014, 2015, 2017, 2018, 2020, 2022
  *               Ingo Schwarze <schwarze@openbsd.org>
@@ -310,13 +310,12 @@ roff_escape(const char *buf, const int ln, const int aesc,
        iendarg = iarg;
        while (maxl > 0) {
                if (buf[iendarg] == '\0') {
+                       err = MANDOCERR_ESC_INCOMPLETE;
+                       if (rval != ESCAPE_EXPAND)
+                               rval = ESCAPE_ERROR;
                        /* Ignore an incomplete argument except for \w. */
                        if (buf[inam] != 'w')
                                iendarg = iarg;
-                       if (rval == ESCAPE_EXPAND)
-                               err = MANDOCERR_ESC_BAD;
-                       else
-                               rval = ESCAPE_ERROR;
                        break;
                }
                if (buf[iendarg] == term) {
@@ -401,6 +400,7 @@ roff_escape(const char *buf, const int ln, const int aesc,
                 */
 
                if (term != '\0' && argl == 1 && buf[iarg] != '-') {
+                       err = MANDOCERR_ESC_BADCHAR;
                        rval = ESCAPE_ERROR;
                        break;
                }
@@ -416,8 +416,10 @@ roff_escape(const char *buf, const int ln, const int aesc,
                        c = 0;
                        for (i = iarg; i < iendarg; i++)
                                c = 10 * c + (buf[i] - '0');
-                       if (c < 0x21 || (c > 0x7e && c < 0xa0) || c > 0xff)
+                       if (c < 0x21 || (c > 0x7e && c < 0xa0) || c > 0xff) {
+                               err = MANDOCERR_ESC_BADCHAR;
                                break;
+                       }
                        iarg += 4;
                        rval = ESCAPE_NUMBERED;
                        break;
@@ -433,13 +435,19 @@ roff_escape(const char *buf, const int ln, const int aesc,
                if (buf[iarg] != 'u' || argl < 5 || argl > 7)
                        break;
                if (argl == 7 &&
-                   (buf[iarg + 1] != '1' || buf[iarg + 2] != '0'))
+                   (buf[iarg + 1] != '1' || buf[iarg + 2] != '0')) {
+                       err = MANDOCERR_ESC_BADCHAR;
                        break;
-               if (argl == 6 && buf[iarg + 1] == '0')
+               }
+               if (argl == 6 && buf[iarg + 1] == '0') {
+                       err = MANDOCERR_ESC_BADCHAR;
                        break;
+               }
                if (argl == 5 && buf[iarg + 1] == 'D' &&
-                   strchr("89ABCDEF", buf[iarg + 2]) != NULL)
+                   strchr("89ABCDEF", buf[iarg + 2]) != NULL) {
+                       err = MANDOCERR_ESC_BADCHAR;
                        break;
+               }
                if ((int)strspn(buf + iarg + 1, "0123456789ABCDEFabcdef")
                    + 1 == argl)
                        rval = ESCAPE_UNICODE;
@@ -477,7 +485,8 @@ out:
        *resc = iesc;
        switch (rval) {
        case ESCAPE_ERROR:
-               err = MANDOCERR_ESC_BAD;
+               if (err == MANDOCERR_OK)
+                       err = MANDOCERR_ESC_BAD;
                break;
        case ESCAPE_UNSUPP:
                err = MANDOCERR_ESC_UNSUPP;
@@ -487,8 +496,10 @@ out:
                        err = MANDOCERR_ESC_UNDEF;
                break;
        case ESCAPE_SPECIAL:
-               if (mchars_spec2cp(buf + iarg, argl) < 0)
-                       err = MANDOCERR_ESC_BAD;
+               if (mchars_spec2cp(buf + iarg, argl) >= 0)
+                       err = MANDOCERR_OK;
+               else if (err == MANDOCERR_OK)
+                       err = MANDOCERR_ESC_UNKCHAR;
                break;
        default:
                break;