src - OpenBSD base system

diff options


context:
space:
mode:

author	Ingo Schwarze <schwarze@cvs.openbsd.org>	2018-12-15 19:30:21 +0000
committer	Ingo Schwarze <schwarze@cvs.openbsd.org>	2018-12-15 19:30:21 +0000
commit	3962ec0010304a43d14512a8b79da75485655694 (patch)
tree	d05b34a3a0c349eea038e76ce0f88596e149de67 /usr.bin/mandoc
parent	41d43cda283dfff21b7d5685cb433805d215733e (diff)

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.

Diffstat (limited to 'usr.bin/mandoc')

-rw-r--r--

usr.bin/mandoc/chars.c

-rw-r--r--

usr.bin/mandoc/html.c

-rw-r--r--

usr.bin/mandoc/mandoc.1

-rw-r--r--

usr.bin/mandoc/mandoc.c

-rw-r--r--

usr.bin/mandoc/mandoc.h

-rw-r--r--

usr.bin/mandoc/mandoc_msg.c

-rw-r--r--

usr.bin/mandoc/mdoc_man.c

-rw-r--r--

usr.bin/mandoc/mdoc_markdown.c

-rw-r--r--

usr.bin/mandoc/roff.c

-rw-r--r--

usr.bin/mandoc/term.c

10 files changed, 167 insertions, 68 deletions

diff --git a/usr.bin/mandoc/chars.c b/usr.bin/mandoc/chars.c
index c28e78b50eb..13ccc6ba119 100644
--- a/usr.bin/mandoc/chars.c
+++ b/usr.bin/mandoc/chars.c

@@ -1,7 +1,7 @@

-/* $OpenBSD: chars.c,v 1.47 2018/12/14 01:17:46 schwarze Exp $ */

+/* $OpenBSD: chars.c,v 1.48 2018/12/15 19:30:19 schwarze Exp $ */

* Permission to use, copy, modify, and distribute this software for any

* purpose with or without fee is hereby granted, provided that the above

@@ -46,21 +46,13 @@ static struct ln lines[] = {

{ " ", ascii_nbrsp, 0x00a0 },

{ "~", ascii_nbrsp, 0x00a0 },

{ "0", " ", 0x2002 },

- { "|", "", 0 },

- { "^", "", 0 },

- { "&", "", 0 },

- { ")", "", 0 },

- { "%", "", 0 },

{ ":", ascii_break, 0 },

- /* XXX The following three do not really belong here. */

- { "t", "", 0 },

- { "c", "", 0 },

- { "}", "", 0 },

/* Lines. */

{ "ba", "|", 0x007c },

{ "br", "|", 0x2502 },

{ "ul", "_", 0x005f },

+ { "_", "_", 0x005f },

{ "ru", "_", 0x005f },

{ "rn", "-", 0x203e },

{ "bb", "|", 0x00a6 },

@@ -463,7 +455,7 @@ mchars_spec2cp(const char *p, size_t sz)

end = p + sz;

ln = ohash_find(&mchars, ohash_qlookupi(&mchars, p, &end));

- return ln != NULL ? ln->unicode : sz == 1 ? (unsigned char)*p : -1;

+ return ln != NULL ? ln->unicode : -1;

}

int

@@ -493,10 +485,8 @@ mchars_spec2str(const char *p, size_t sz, size_t *rsz)

end = p + sz;

ln = ohash_find(&mchars, ohash_qlookupi(&mchars, p, &end));

- if (ln == NULL) {

- *rsz = 1;

- return sz == 1 ? p : NULL;

- }

+ if (ln == NULL)

+ return NULL;

*rsz = strlen(ln->ascii);

return ln->ascii;

diff --git a/usr.bin/mandoc/html.c b/usr.bin/mandoc/html.c
index cf44f6caced..9c93eb64e55 100644
--- a/usr.bin/mandoc/html.c
+++ b/usr.bin/mandoc/html.c

@@ -1,4 +1,4 @@

-/* $OpenBSD: html.c,v 1.115 2018/12/04 18:29:18 schwarze Exp $ */

+/* $OpenBSD: html.c,v 1.116 2018/12/15 19:30:19 schwarze Exp $ */

@@ -400,9 +400,6 @@ print_encode(struct html *h, const char *p, const char *pend, int norecurse)

continue;

esc = mandoc_escape(&p, &seq, &len);

- if (ESCAPE_ERROR == esc)

- break;

switch (esc) {

case ESCAPE_FONT:

case ESCAPE_FONTPREV:

@@ -420,6 +417,8 @@ print_encode(struct html *h, const char *p, const char *pend, int norecurse)

case ESCAPE_SKIPCHAR:

h->flags |= HTML_SKIPCHAR;

continue;

+ case ESCAPE_ERROR:

+ continue;

default:

break;

}

@@ -444,6 +443,9 @@ print_encode(struct html *h, const char *p, const char *pend, int norecurse)

if (c <= 0)

continue;

break;

+ case ESCAPE_UNDEF:

+ c = *seq;

+ break;

case ESCAPE_DEVICE:

print_word(h, "html");

continue;

diff --git a/usr.bin/mandoc/mandoc.1 b/usr.bin/mandoc/mandoc.1
index af46152384c..4413055c909 100644
--- a/usr.bin/mandoc/mandoc.1
+++ b/usr.bin/mandoc/mandoc.1

@@ -1,4 +1,4 @@

-.\" $OpenBSD: mandoc.1,v 1.155 2018/11/22 11:30:15 schwarze Exp $

+.\" $OpenBSD: mandoc.1,v 1.156 2018/12/15 19:30:19 schwarze Exp $

.\"

@@ -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: November 22 2018 $

+.Dd $Mdocdate: December 15 2018 $

.Dt MANDOC 1

.Os

.Sh NAME

@@ -1676,7 +1676,8 @@ 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, or the argument has too few characters.

+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

@@ -1689,6 +1690,12 @@ and

.Ic \ew

to the length of the incomplete argument.

All other invalid escape sequences are ignored.

+.It Sy "undefined escape, printing literally"

+.Pq roff

+In an escape sequence, the first character

+right after the leading backslash is invalid.

+That character is printed literally,

+which is equivalent to ignoring the backslash.

.It Sy "undefined string, using \(dq\(dq"

.Pq roff

If a string is used without being defined before,

@@ -2154,6 +2161,13 @@ implementations but not by

.Nm

was found in an input file.

It is replaced by a question mark.

+.It Sy "unsupported escape sequence"

+.Pq roff

+An input file contains an escape sequence supported by GNU troff

+or Heirloom troff but not by

+.Nm ,

+and it is likely that this will cause information loss

+or considerable misformatting.

.It Sy "unsupported roff request"

.Pq roff

An input file contains a

diff --git a/usr.bin/mandoc/mandoc.c b/usr.bin/mandoc/mandoc.c
index ac84e93497c..f4249a4b78f 100644
--- a/usr.bin/mandoc/mandoc.c
+++ b/usr.bin/mandoc/mandoc.c

@@ -1,4 +1,4 @@

-/* $OpenBSD: mandoc.c,v 1.78 2018/12/14 06:33:03 schwarze Exp $ */

+/* $OpenBSD: mandoc.c,v 1.79 2018/12/15 19:30:19 schwarze Exp $ */

@@ -54,6 +54,14 @@ mandoc_escape(const char **end, const char **start, int *sz)

sz = &local_sz;

+ * Treat "\E" just like "\";

+ * it only makes a difference in copy mode.

+ */

+ if (**end == 'E')

+ ++*end;

+ /*

* Beyond the backslash, at least one input character

* is part of the escape sequence. With one exception

* (see below), that character won't be returned.

@@ -75,6 +83,10 @@ mandoc_escape(const char **end, const char **start, int *sz)

*sz = 2;

break;

case '[':

+ if (**start == ' ') {

+ ++*end;

+ return ESCAPE_ERROR;

+ }

gly = ESCAPE_SPECIAL;

term = ']';

break;

@@ -89,11 +101,26 @@ mandoc_escape(const char **end, const char **start, int *sz)

* Escapes taking no arguments at all.

- case 'd':

- case 'u':

+ case '!':

+ case '?':

+ return ESCAPE_UNSUPP;

+ case '%':

+ case '&':

+ case ')':

case ',':

case '/':

+ case '^':

+ case 'a':

+ case 'd':

+ case 'r':

+ case 't':

+ case 'u':

+ case '{':

+ case '|':

+ case '}':

return ESCAPE_IGNORE;

+ case 'c':

+ return ESCAPE_NOSPACE;

case 'p':

return ESCAPE_BREAK;

@@ -111,28 +138,46 @@ mandoc_escape(const char **end, const char **start, int *sz)

* 'X' is the trigger. These have opaque sub-strings.

case 'F':

+ case 'f':

case 'g':

case 'k':

case 'M':

case 'm':

case 'n':

+ case 'O':

case 'V':

case 'Y':

- gly = ESCAPE_IGNORE;

- /* FALLTHROUGH */

- case 'f':

- if (ESCAPE_ERROR == gly)

- gly = ESCAPE_FONT;

+ gly = (*start)[-1] == 'f' ? ESCAPE_FONT : ESCAPE_IGNORE;

switch (**start) {

case '(':

+ if ((*start)[-1] == 'O')

+ gly = ESCAPE_ERROR;

*start = ++*end;

*sz = 2;

break;

case '[':

+ if ((*start)[-1] == 'O')

+ gly = (*start)[1] == '5' ?

+ ESCAPE_UNSUPP : ESCAPE_ERROR;

*start = ++*end;

term = ']';

break;

default:

+ if ((*start)[-1] == 'O') {

+ switch (**start) {

+ case '0':

+ gly = ESCAPE_UNSUPP;

+ break;

+ case '1':

+ case '2':

+ case '3':

+ case '4':

+ break;

+ default:

+ gly = ESCAPE_ERROR;

+ break;

+ }

*sz = 1;

break;

}

@@ -255,18 +300,29 @@ mandoc_escape(const char **end, const char **start, int *sz)

break;

- * Anything else is assumed to be a glyph.

- * In this case, pass back the character after the backslash.

+ * Several special characters can be encoded as

+ * one-byte escape sequences without using \[].

- default:

+ case ' ':

+ case '\'':

+ case '-':

+ case '.':

+ case '0':

+ case ':':

+ case '_':

+ case '`':

+ case 'e':

+ case '~':

gly = ESCAPE_SPECIAL;

+ /* FALLTHROUGH */

+ default:

+ if (gly == ESCAPE_ERROR)

+ gly = ESCAPE_UNDEF;

*start = --*end;

*sz = 1;

break;

}

- assert(ESCAPE_ERROR != gly);

* Read up to the terminating character,

* paying attention to nested escapes.

@@ -289,6 +345,15 @@ mandoc_escape(const char **end, const char **start, int *sz)

}

*sz = (*end)++ - *start;

+ /*

+ * The file chars.c only provides one common list

+ * of character names, but \[-] == \- is the only

+ * one of the characters with one-byte names that

+ * allows enclosing the name in brackets.

+ */

+ if (gly == ESCAPE_SPECIAL && *sz == 1 && **start != '-')

+ return ESCAPE_ERROR;

} else {

assert(*sz > 0);

if ((size_t)*sz > strlen(*start))

@@ -344,10 +409,6 @@ mandoc_escape(const char **end, const char **start, int *sz)

break;

case ESCAPE_SPECIAL:

if (**start == 'c') {

- if (*sz == 1) {

- gly = ESCAPE_NOSPACE;

- break;

- }

if (*sz < 6 || *sz > 7 ||

strncmp(*start, "char", 4) != 0 ||

(int)strspn(*start + 4, "0123456789") + 4 < *sz)

@@ -429,6 +490,7 @@ mandoc_getarg(char **cpp, int ln, int *pos)

* backslashes and backslash-t to literal tabs.

switch (cp[1]) {

+ case 'a':

case 't':

cp[0] = '\t';

/* FALLTHROUGH */

diff --git a/usr.bin/mandoc/mandoc.h b/usr.bin/mandoc/mandoc.h
index 00b10f91c66..2e4a9dc0317 100644
--- a/usr.bin/mandoc/mandoc.h
+++ b/usr.bin/mandoc/mandoc.h

@@ -1,4 +1,4 @@

-/* $OpenBSD: mandoc.h,v 1.202 2018/12/14 05:17:45 schwarze Exp $ */

+/* $OpenBSD: mandoc.h,v 1.203 2018/12/15 19:30:19 schwarze Exp $ */

@@ -169,6 +169,7 @@ enum mandocerr {

MANDOCERR_FI_TAB, /* tab in filled text */

MANDOCERR_EOS, /* new sentence, new line */

MANDOCERR_ESC_BAD, /* invalid escape sequence: esc */

+ MANDOCERR_ESC_UNDEF, /* undefined escape, printing literally: char */

MANDOCERR_STR_UNDEF, /* undefined string, using "": name */

/* related to tables */

@@ -231,6 +232,7 @@ enum mandocerr {

MANDOCERR_TOOLARGE, /* input too large */

MANDOCERR_CHAR_UNSUPP, /* unsupported control character: number */

+ MANDOCERR_ESC_UNSUPP, /* unsupported escape sequence: escape */

MANDOCERR_REQ_UNSUPP, /* unsupported roff request: request */

MANDOCERR_WHILE_NEST, /* nested .while loops */

MANDOCERR_WHILE_OUTOF, /* end of scope with open .while loop */

@@ -245,7 +247,9 @@ enum mandocerr {

enum mandoc_esc {

ESCAPE_ERROR = 0, /* bail! unparsable escape */

+ ESCAPE_UNSUPP, /* unsupported escape; ignore it */

ESCAPE_IGNORE, /* escape to be ignored */

+ ESCAPE_UNDEF, /* undefined escape; print literal character */

ESCAPE_SPECIAL, /* a regular special character */

ESCAPE_FONT, /* a generic font mode */

ESCAPE_FONTBOLD, /* bold font mode */

diff --git a/usr.bin/mandoc/mandoc_msg.c b/usr.bin/mandoc/mandoc_msg.c
index 03fe7bd3315..a21964be228 100644
--- a/usr.bin/mandoc/mandoc_msg.c
+++ b/usr.bin/mandoc/mandoc_msg.c

@@ -1,4 +1,4 @@

-/* $OpenBSD: mandoc_msg.c,v 1.2 2018/12/14 05:17:45 schwarze Exp $ */

+/* $OpenBSD: mandoc_msg.c,v 1.3 2018/12/15 19:30:19 schwarze Exp $ */

@@ -167,6 +167,7 @@ static const char *const type_message[MANDOCERR_MAX] = {

"tab in filled text",

"new sentence, new line",

"invalid escape sequence",

+ "undefined escape, printing literally",

"undefined string, using \"\"",

/* related to tables */

@@ -228,6 +229,7 @@ static const char *const type_message[MANDOCERR_MAX] = {

"unsupported feature",

"input too large",

"unsupported control character",

+ "unsupported escape sequence",

"unsupported roff request",

"nested .while loops",

"end of scope with open .while loop",

diff --git a/usr.bin/mandoc/mdoc_man.c b/usr.bin/mandoc/mdoc_man.c
index 0edf3c7e121..17446b70f2a 100644
--- a/usr.bin/mandoc/mdoc_man.c
+++ b/usr.bin/mandoc/mdoc_man.c

@@ -1,4 +1,4 @@

-/* $OpenBSD: mdoc_man.c,v 1.127 2018/12/03 21:00:06 schwarze Exp $ */

+/* $OpenBSD: mdoc_man.c,v 1.128 2018/12/15 19:30:19 schwarze Exp $ */

@@ -323,6 +323,7 @@ man_strlen(const char *cp)

case ESCAPE_UNICODE:

case ESCAPE_NUMBERED:

case ESCAPE_SPECIAL:

+ case ESCAPE_UNDEF:

case ESCAPE_OVERSTRIKE:

if (skip)

skip = 0;

diff --git a/usr.bin/mandoc/mdoc_markdown.c b/usr.bin/mandoc/mdoc_markdown.c
index 76f0f0a2559..ffb28bd704a 100644
--- a/usr.bin/mandoc/mdoc_markdown.c
+++ b/usr.bin/mandoc/mdoc_markdown.c

@@ -1,4 +1,4 @@

-/* $OpenBSD: mdoc_markdown.c,v 1.28 2018/12/03 21:00:06 schwarze Exp $ */

+/* $OpenBSD: mdoc_markdown.c,v 1.29 2018/12/15 19:30:19 schwarze Exp $ */

@@ -589,6 +589,9 @@ md_word(const char *s)

case ESCAPE_SPECIAL:

uc = mchars_spec2cp(seq, sz);

break;

+ case ESCAPE_UNDEF:

+ uc = *seq;

+ break;

case ESCAPE_DEVICE:

md_rawword("markdown");

continue;

diff --git a/usr.bin/mandoc/roff.c b/usr.bin/mandoc/roff.c
index 439ef112e16..6e5924d5925 100644
--- a/usr.bin/mandoc/roff.c
+++ b/usr.bin/mandoc/roff.c

@@ -1,4 +1,4 @@

-/* $OpenBSD: roff.c,v 1.223 2018/12/14 06:33:03 schwarze Exp $ */

+/* $OpenBSD: roff.c,v 1.224 2018/12/15 19:30:19 schwarze Exp $ */

@@ -1152,6 +1152,7 @@ roff_res(struct roff *r, struct buf *buf, int ln, int pos)

struct roff_node *n; /* used for header comments */

const char *start; /* start of the string to process */

char *stesc; /* start of an escape sequence ('\\') */

+ const char *esct; /* type of esccape sequence */

char *ep; /* end of comment string */

const char *stnam; /* start of the name, after "[(*" */

const char *cp; /* end of the name, e.g. before ']' */

@@ -1161,7 +1162,6 @@ roff_res(struct roff *r, struct buf *buf, int ln, int pos)

size_t naml; /* actual length of the escape name */

size_t asz; /* length of the replacement */

size_t rsz; /* length of the rest of the string */

- enum mandoc_esc esc; /* type of the escape sequence */

int inaml; /* length returned from mandoc_escape() */

int expand_count; /* to avoid infinite loops */

int npos; /* position in numeric expression */

@@ -1170,6 +1170,7 @@ roff_res(struct roff *r, struct buf *buf, int ln, int pos)

int done; /* no more input available */

int deftype; /* type of definition to paste */

int rcsid; /* kind of RCS id seen */

+ enum mandocerr err; /* for escape sequence problems */

char sign; /* increment number register */

char term; /* character terminating the escape */

@@ -1302,7 +1303,10 @@ roff_res(struct roff *r, struct buf *buf, int ln, int pos)

term = '\0';

cp = stesc + 1;

- switch (*cp) {

+ if (*cp == 'E')

+ cp++;

+ esct = cp;

+ switch (*esct) {

case '*':

case '$':

res = NULL;

@@ -1318,12 +1322,26 @@ roff_res(struct roff *r, struct buf *buf, int ln, int pos)

res = ubuf;

break;

default:

- esc = mandoc_escape(&cp, &stnam, &inaml);

- if (esc == ESCAPE_ERROR ||

- (esc == ESCAPE_SPECIAL &&

- mchars_spec2cp(stnam, inaml) < 0))

- mandoc_msg(MANDOCERR_ESC_BAD,

- ln, (int)(stesc - buf->buf),

+ err = MANDOCERR_OK;

+ switch(mandoc_escape(&cp, &stnam, &inaml)) {

+ case ESCAPE_SPECIAL:

+ if (mchars_spec2cp(stnam, inaml) >= 0)

+ break;

+ /* FALLTHROUGH */

+ case ESCAPE_ERROR:

+ err = MANDOCERR_ESC_BAD;

+ break;

+ case ESCAPE_UNDEF:

+ err = MANDOCERR_ESC_UNDEF;

+ break;

+ case ESCAPE_UNSUPP:

+ err = MANDOCERR_ESC_UNSUPP;

+ break;

+ default:

+ break;

+ }

+ if (err != MANDOCERR_OK)

+ mandoc_msg(err, ln, (int)(stesc - buf->buf),

"%.*s", (int)(cp - stesc), stesc);

stesc--;

continue;

@@ -1380,7 +1398,7 @@ roff_res(struct roff *r, struct buf *buf, int ln, int pos)

cp++;

break;

}

- if (*cp++ != '\\' || stesc[1] != 'w') {

+ if (*cp++ != '\\' || *esct != 'w') {

naml++;

continue;

}

@@ -1388,6 +1406,7 @@ roff_res(struct roff *r, struct buf *buf, int ln, int pos)

case ESCAPE_SPECIAL:

case ESCAPE_UNICODE:

case ESCAPE_NUMBERED:

+ case ESCAPE_UNDEF:

case ESCAPE_OVERSTRIKE:

naml++;

break;

@@ -1401,7 +1420,7 @@ roff_res(struct roff *r, struct buf *buf, int ln, int pos)

* undefined, resume searching for escapes.

- switch (stesc[1]) {

+ switch (*esct) {

case '*':

if (arg_complete) {

deftype = ROFFDEF_USER | ROFFDEF_PRE;

@@ -1428,15 +1447,15 @@ roff_res(struct roff *r, struct buf *buf, int ln, int pos)

break;

}

ctx = r->mstack + r->mstackpos;

- npos = stesc[2] - '1';

+ npos = esct[1] - '1';

if (npos >= 0 && npos <= 8) {

res = npos < ctx->argc ?

ctx->argv[npos] : "";

break;

}

- if (stesc[2] == '*')

+ if (esct[1] == '*')

quote_args = 0;

- else if (stesc[2] == '@')

+ else if (esct[1] == '@')

quote_args = 1;

else {

mandoc_msg(MANDOCERR_ARG_NONUM, ln,

@@ -1498,7 +1517,7 @@ roff_res(struct roff *r, struct buf *buf, int ln, int pos)

}

if (res == NULL) {

- if (stesc[1] == '*')

+ if (*esct == '*')

mandoc_msg(MANDOCERR_STR_UNDEF,

ln, (int)(stesc - buf->buf),

"%.*s", (int)naml, stnam);

diff --git a/usr.bin/mandoc/term.c b/usr.bin/mandoc/term.c
index e21c2622927..7b5816be28a 100644
--- a/usr.bin/mandoc/term.c
+++ b/usr.bin/mandoc/term.c

@@ -1,4 +1,4 @@

-/* $OpenBSD: term.c,v 1.136 2018/10/25 01:21:30 schwarze Exp $ */

+/* $OpenBSD: term.c,v 1.137 2018/12/15 19:30:19 schwarze Exp $ */

@@ -475,9 +475,6 @@ term_word(struct termp *p, const char *word)

word++;

esc = mandoc_escape(&word, &seq, &sz);

- if (ESCAPE_ERROR == esc)

- continue;

switch (esc) {

case ESCAPE_UNICODE:

uc = mchars_num2uc(seq + 1, sz - 1);

@@ -498,6 +495,9 @@ term_word(struct termp *p, const char *word)

encode1(p, uc);

}

continue;

+ case ESCAPE_UNDEF:

+ uc = *seq;

+ break;

case ESCAPE_FONTBOLD:

term_fontrepl(p, TERMFONT_BOLD);

continue;

@@ -585,6 +585,9 @@ term_word(struct termp *p, const char *word)

case ESCAPE_SPECIAL:

uc = mchars_spec2cp(cp, sz);

break;

+ case ESCAPE_UNDEF:

+ uc = *seq;

+ break;

default:

uc = -1;

break;

@@ -843,12 +846,8 @@ term_strlen(const struct termp *p, const char *cp)

switch (*cp) {

case '\\':

cp++;

- esc = mandoc_escape(&cp, &seq, &ssz);

- if (ESCAPE_ERROR == esc)

- continue;

rhs = NULL;

+ esc = mandoc_escape(&cp, &seq, &ssz);

switch (esc) {

case ESCAPE_UNICODE:

uc = mchars_num2uc(seq + 1, ssz - 1);

@@ -869,6 +868,9 @@ term_strlen(const struct termp *p, const char *cp)

sz += cond_width(p, uc, &skip);

}

continue;

+ case ESCAPE_UNDEF:

+ uc = *seq;

+ break;

case ESCAPE_DEVICE:

if (p->type == TERMTYPE_PDF) {

rhs = "pdf";