Several improvements to escape sequence handling.

* Add the missing special character \_ (underscore).
* Partial implementations of \a (leader character)
and \E (uninterpreted escape character).
* Parse and ignore \r (reverse line feed).
* Add a WARNING message about undefined escape sequences.
* Add an UNSUPP message about unsupported escape sequences.
* Mark \! and \? (transparent throughput)
and \O (suppress output) as unsupported.
* Treat the various variants of zero-width spaces as one-byte escape
sequences rather than as special characters, to avoid defining bogus
forms with square brackets.
* For special characters with one-byte names, do not define bogus
forms with square brackets, except for \[-], which is valid.
* In the form with square brackets, undefined special characters do not
fall back to printing the name verbatim, not even for one-byte names.
* Starting a special character name with a blank is an error.
* Undefined escape sequences never abort formatting of the input
string, not even in HTML output mode.
* Document the newly handled escapes, and a few that were missing.
* Regression tests for most of the above.

12 files changed, 85 insertions, 20 deletions

diff --git a/regress/char/accent/Makefile b/regress/char/accent/Makefile
index 4bc149a7..3712217e 100644
--- a/regress/char/accent/Makefile
+++ b/regress/char/accent/Makefile
@@ -3,5 +3,6 @@
 REGRESS_TARGETS = nocombine utf8only combine
 SKIP_ASCII = utf8only combine
 UTF8_TARGETS = nocombine utf8only combine
+LINT_TARGETS = nocombine
 
 .include <bsd.regress.mk>
diff --git a/regress/char/accent/nocombine.in b/regress/char/accent/nocombine.in
index a81d446b..637d337e 100644
--- a/regress/char/accent/nocombine.in
+++ b/regress/char/accent/nocombine.in
@@ -1,17 +1,17 @@
 .\" $OpenBSD: nocombine.in,v 1.2 2017/07/04 14:53:23 schwarze Exp $
-.TH CHAR-ACCENT-NOCOMBINE 1 "March 8, 2014"
+.TH CHAR-ACCENT-NOCOMBINE 1 "December 15, 2018"
 .SH NAME
 \fBchar-accent-nocombine\fR - non-combining accents
 .SH DESCRIPTION
 bare acute accent: e'e
 .br
-escaped acute accent: e\'e
+escaped acute accent: e\'e\[']e
 .br
 acute accent sequence: e\(aae
 .br
 bare grave accent: e`e
 .br
-escaped grave accent: e\`e
+escaped grave accent: e\`e\[`]e
 .br
 acute grave sequence: e\(gae
 .br
@@ -20,15 +20,15 @@ hungarian umlaut: e\(a"e
 .\" XXX This is ridiculous.
 .\" XXX groff prints the macron as an underscore in the previous line.
 .\" macron: e\(a-e
-.br
+.\" .br
 .\" XXX groff doesn't have a dot in ASCII mode, only in UTF-8 mode.
 .\" dotted: e\(a.e
-.br
+.\" .br
 circumflex: e\(a^e
 .br
 .\" XXX groff uses a backspace for this one in ASCII mode.
 .\" breve: e\(abe
-.br
+.\" .br
 cedilla: e\(ace
 .br
 dieresis: e\(ade
diff --git a/regress/char/accent/nocombine.out_ascii b/regress/char/accent/nocombine.out_ascii
index bc1cce15..0f18ac4a 100644
--- a/regress/char/accent/nocombine.out_ascii
+++ b/regress/char/accent/nocombine.out_ascii
@@ -7,10 +7,10 @@ NNAAMMEE
 
 DDEESSCCRRIIPPTTIIOONN
        bare acute accent: e'e
-       escaped acute accent: e'e
+       escaped acute accent: e'ee
        acute accent sequence: e'e
        bare grave accent: e`e
-       escaped grave accent: e`e
+       escaped grave accent: e`ee
        acute grave sequence: e`e
        hungarian umlaut: e"e
        circumflex: e^e
@@ -25,4 +25,4 @@ DDEESSCCRRIIPPTTIIOONN
 
 
 
-OpenBSD                          March 8, 2014        CHAR-ACCENT-NOCOMBINE(1)
+OpenBSD                        December 15, 2018      CHAR-ACCENT-NOCOMBINE(1)
diff --git a/regress/char/accent/nocombine.out_lint b/regress/char/accent/nocombine.out_lint
new file mode 100644
index 00000000..c9de4162
--- /dev/null
+++ b/regress/char/accent/nocombine.out_lint
@@ -0,0 +1,2 @@
+mandoc: nocombine.in:8:27: WARNING: invalid escape sequence: \[']
+mandoc: nocombine.in:14:27: WARNING: invalid escape sequence: \[`]
diff --git a/regress/char/accent/nocombine.out_utf8 b/regress/char/accent/nocombine.out_utf8
index 3aa441a2..497bf6fd 100644
--- a/regress/char/accent/nocombine.out_utf8
+++ b/regress/char/accent/nocombine.out_utf8
@@ -7,10 +7,10 @@ NNAAMMEE
 
 DDEESSCCRRIIPPTTIIOONN
        bare acute accent: e'e
-       escaped acute accent: e´e
+       escaped acute accent: e´ee
        acute accent sequence: e´e
        bare grave accent: e`e
-       escaped grave accent: e`e
+       escaped grave accent: e`ee
        acute grave sequence: e`e
        hungarian umlaut: e˝e
        circumflex: e^e
@@ -25,4 +25,4 @@ DDEESSCCRRIIPPTTIIOONN
 
 
 
-OpenBSD                          March 8, 2014        CHAR-ACCENT-NOCOMBINE(1)
+OpenBSD                        December 15, 2018      CHAR-ACCENT-NOCOMBINE(1)
diff --git a/regress/char/space/Makefile b/regress/char/space/Makefile
index b6095de3..50c81c4a 100644
--- a/regress/char/space/Makefile
+++ b/regress/char/space/Makefile
@@ -3,11 +3,12 @@
 REGRESS_TARGETS  = leading-mdoc leading-man multiple trailing-mdoc zerowidth
 REGRESS_TARGETS += eos eos-man break nobreak
 REGRESS_TARGETS += tab tab-man esct-mdoc esct-man
+REGRESS_TARGETS += invalid
 
 UTF8_TARGETS	 = zerowidth
 
 HTML_TARGETS	 = zerowidth
 
-LINT_TARGETS	 = trailing-mdoc tab tab-man esct-mdoc esct-man
+LINT_TARGETS	 = trailing-mdoc tab tab-man esct-mdoc esct-man invalid
 
 .include <bsd.regress.mk>
diff --git a/regress/char/space/esct-man.in b/regress/char/space/esct-man.in
index e7afeefe..f290d9e2 100644
--- a/regress/char/space/esct-man.in
+++ b/regress/char/space/esct-man.in
@@ -1,5 +1,5 @@
 .\" $OpenBSD: esct-man.in,v 1.2 2017/07/04 14:53:23 schwarze Exp $
-.TH SPACE-ESCT-MAN 1 2013-06-20
+.TH SPACE-ESCT-MAN 1 "December 15, 2018"
 .SH NAME
 SPACE-T-MAN \- the t escape sequence in pages with man macros
 .SH DESCRIPTION
@@ -9,10 +9,14 @@ single	tab
 .br
 single\tescape-t
 .br
+single\aescape-a
+.br
 double		tab
 .br
 double\t\tescape-t
 .br
+double\a\aescape-a
+.br
 \tThis line starts with escape-t and comes close to the right margin.
 \tThe next line starts with escape-t as well.
 .sp
@@ -20,8 +24,10 @@ In a literal display:
 .nf
 single	tab
 single\tescape-t
+single\aescape-a
 double		tab
 double\t\tescape-t
+double\a\aescape-a
 .fi
 .sp
 After the IP macro:
@@ -29,7 +35,13 @@ After the IP macro:
 text
 .IP single\tescape-t 3n
 text
+.\" XXX not implemented
+.\" .IP single\aescape-a 3n
+.\" text
 .PP
 After font macros:
 .br
 .B single\ttab
+.\" XXX not implemented
+.\" .br
+.\" .B single\aleader
diff --git a/regress/char/space/esct-man.out_ascii b/regress/char/space/esct-man.out_ascii
index 29fbabea..3056cc72 100644
--- a/regress/char/space/esct-man.out_ascii
+++ b/regress/char/space/esct-man.out_ascii
@@ -9,16 +9,20 @@ DDEESSCCRRIIPPTTIIOONN
        In plain text:
        single    tab
        singleescape-t
+       singleescape-a
        double         tab
        doubleescape-t
+       doubleescape-a
        This line starts with escape-t and comes close to the right margin.
        The next line starts with escape-t as well.
 
        In a literal display:
        single    tab
        singleescape-t
+       singleescape-a
        double         tab
        doubleescape-t
+       doubleescape-a
 
        After the IP macro:
 
@@ -33,4 +37,4 @@ DDEESSCCRRIIPPTTIIOONN
 
 
 
-OpenBSD                           2013-06-20                 SPACE-ESCT-MAN(1)
+OpenBSD                        December 15, 2018             SPACE-ESCT-MAN(1)
diff --git a/regress/char/space/esct-man.out_lint b/regress/char/space/esct-man.out_lint
index c625b3a3..2fa8ba9c 100644
--- a/regress/char/space/esct-man.out_lint
+++ b/regress/char/space/esct-man.out_lint
@@ -1,6 +1,6 @@
 mandoc: esct-man.in:8:7: WARNING: tab in filled text
-mandoc: esct-man.in:12:7: WARNING: tab in filled text
-mandoc: esct-man.in:12:8: WARNING: tab in filled text
-mandoc: esct-man.in:28:11: WARNING: tab in filled text
-mandoc: esct-man.in:30:11: WARNING: tab in filled text
-mandoc: esct-man.in:35:10: WARNING: tab in filled text
+mandoc: esct-man.in:14:7: WARNING: tab in filled text
+mandoc: esct-man.in:14:8: WARNING: tab in filled text
+mandoc: esct-man.in:34:11: WARNING: tab in filled text
+mandoc: esct-man.in:36:11: WARNING: tab in filled text
+mandoc: esct-man.in:44:10: WARNING: tab in filled text
diff --git a/regress/char/space/invalid.in b/regress/char/space/invalid.in
new file mode 100644
index 00000000..c4692453
--- /dev/null
+++ b/regress/char/space/invalid.in
@@ -0,0 +1,15 @@
+.\" $OpenBSD$
+.TH SPACE-INVALID 1 "December 15, 2018"
+.SH NAME
+SPACE-INVALID \- invalid whitespace escape sequences
+.SH DESCRIPTION
+.nf
+blank: a\[hy]b\[ hy]c
+percent: a\%b\[%]c
+ampersand: a\&b\[&]c
+colon: a\:b\[:]c
+caret: a\^b\[^]c
+underline: a\_b\[_]c
+pipe: a\|b\[|]c
+tilde: a\~b\[~]c
+digit-width: a\0b\[0]c
diff --git a/regress/char/space/invalid.out_ascii b/regress/char/space/invalid.out_ascii
new file mode 100644
index 00000000..bf3b5a16
--- /dev/null
+++ b/regress/char/space/invalid.out_ascii
@@ -0,0 +1,21 @@
+SPACE-INVALID(1)            General Commands Manual           SPACE-INVALID(1)
+
+
+
+NNAAMMEE
+       SPACE-INVALID - invalid whitespace escape sequences
+
+DDEESSCCRRIIPPTTIIOONN
+       blank: a-bhy]c
+       percent: abc
+       ampersand: abc
+       colon: abc
+       caret: abc
+       underline: a_bc
+       pipe: abc
+       tilde: a bc
+       digit-width: a bc
+
+
+
+OpenBSD                        December 15, 2018              SPACE-INVALID(1)
diff --git a/regress/char/space/invalid.out_lint b/regress/char/space/invalid.out_lint
new file mode 100644
index 00000000..c05ef38f
--- /dev/null
+++ b/regress/char/space/invalid.out_lint
@@ -0,0 +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]