size_t
term_strlen(const struct termp *p, const char *cp)
{
size_t sz, rsz, i;
int ssz, skip, uc;
const char *seq, *rhs;
enum mandoc_esc esc;
static const char rej[] = { '\\', ASCII_NBRSP, ASCII_HYPH,
ASCII_BREAK, '\0' };
/*
* Account for escaped sequences within string length
* calculations. This follows the logic in term_word() as we
* must calculate the width of produced strings.
*/
sz = 0;
skip = 0;
while ('\0' != *cp) {
rsz = strcspn(cp, rej);
for (i = 0; i < rsz; i++)
sz += cond_width(p, *cp++, &skip);
switch (*cp) {
case '\\':
cp++;
esc = mandoc_escape(&cp, &seq, &ssz);
if (ESCAPE_ERROR == esc)
continue;
rhs = NULL;
switch (esc) {
case ESCAPE_UNICODE:
uc = mchars_num2uc(seq + 1, ssz - 1);
break;
case ESCAPE_NUMBERED:
uc = mchars_num2char(seq, ssz);
if (uc < 0)
continue;
break;
case ESCAPE_SPECIAL:
if (p->enc == TERMENC_ASCII) {
rhs = mchars_spec2str(p->symtab,
seq, ssz, &rsz);
if (rhs != NULL)
break;
} else {
uc = mchars_spec2cp(p->symtab,
seq, ssz);
if (uc > 0)
sz += cond_width(p, uc, &skip);
}
continue;
case ESCAPE_SKIPCHAR:
skip = 1;
continue;
case ESCAPE_OVERSTRIKE:
rsz = 0;
rhs = seq + ssz;
while (seq < rhs) {
if (*seq == '\\') {
mandoc_escape(&seq, NULL, NULL);
continue;
}
i = (*p->width)(p, *seq++);
if (rsz < i)
rsz = i;
}
sz += rsz;
continue;
default:
continue;
}
/*
* Common handling for Unicode and numbered
* character escape sequences.
*/
if (rhs == NULL) {
if (p->enc == TERMENC_ASCII) {
rhs = ascii_uc2str(uc);
rsz = strlen(rhs);
} else {
if ((uc < 0x20 && uc != 0x09) ||
(uc > 0x7E && uc < 0xA0))
uc = 0xFFFD;
sz += cond_width(p, uc, &skip);
continue;
}
}
if (skip) {
skip = 0;
break;
}
/*
* Common handling for all escape sequences
* printing more than one character.
*/
for (i = 0; i < rsz; i++)
sz += (*p->width)(p, *rhs++);
break;
case ASCII_NBRSP:
sz += cond_width(p, ' ', &skip);
cp++;
break;
case ASCII_HYPH:
sz += cond_width(p, '-', &skip);
cp++;
/* FALLTHROUGH */
case ASCII_BREAK:
break;
default:
break;
}
}
return(sz);
}
size_t
term_strlen(const struct termp *p, const char *cp)
{
size_t sz, rsz, i;
int ssz, c;
const char *seq, *rhs;
enum mandoc_esc esc;
static const char rej[] = { '\\', ASCII_HYPH, ASCII_NBRSP, '\0' };
/*
* Account for escaped sequences within string length
* calculations. This follows the logic in term_word() as we
* must calculate the width of produced strings.
*/
sz = 0;
while ('\0' != *cp) {
rsz = strcspn(cp, rej);
for (i = 0; i < rsz; i++)
sz += (*p->width)(p, *cp++);
c = 0;
switch (*cp) {
case ('\\'):
cp++;
esc = mandoc_escape(&cp, &seq, &ssz);
if (ESCAPE_ERROR == esc)
return(sz);
if (TERMENC_ASCII != p->enc)
switch (esc) {
case (ESCAPE_UNICODE):
c = mchars_num2uc
(seq + 1, ssz - 1);
if ('\0' == c)
break;
sz += (*p->width)(p, c);
continue;
case (ESCAPE_SPECIAL):
c = mchars_spec2cp
(p->symtab, seq, ssz);
if (c <= 0)
break;
sz += (*p->width)(p, c);
continue;
default:
break;
}
rhs = NULL;
switch (esc) {
case (ESCAPE_UNICODE):
sz += (*p->width)(p, '?');
break;
case (ESCAPE_NUMBERED):
c = mchars_num2char(seq, ssz);
if ('\0' != c)
sz += (*p->width)(p, c);
break;
case (ESCAPE_SPECIAL):
rhs = mchars_spec2str
(p->symtab, seq, ssz, &rsz);
if (ssz != 1 || rhs)
break;
rhs = seq;
rsz = ssz;
break;
default:
break;
}
if (NULL == rhs)
break;
for (i = 0; i < rsz; i++)
sz += (*p->width)(p, *rhs++);
break;
case (ASCII_NBRSP):
sz += (*p->width)(p, ' ');
cp++;
break;
case (ASCII_HYPH):
sz += (*p->width)(p, '-');
cp++;
break;
default:
break;
}
}
return(sz);
}
/*
* Handle pwords, partial words, which may be either a single word or a
* phrase that cannot be broken down (such as a literal string). This
* handles word styling.
*/
void
term_word(struct termp *p, const char *word)
{
const char nbrsp[2] = { ASCII_NBRSP, 0 };
const char *seq, *cp;
int sz, uc;
size_t ssz;
enum mandoc_esc esc;
if ( ! (TERMP_NOSPACE & p->flags)) {
if ( ! (TERMP_KEEP & p->flags)) {
bufferc(p, ' ');
if (TERMP_SENTENCE & p->flags)
bufferc(p, ' ');
} else
bufferc(p, ASCII_NBRSP);
}
if (TERMP_PREKEEP & p->flags)
p->flags |= TERMP_KEEP;
if ( ! (p->flags & TERMP_NONOSPACE))
p->flags &= ~TERMP_NOSPACE;
else
p->flags |= TERMP_NOSPACE;
p->flags &= ~(TERMP_SENTENCE | TERMP_NONEWLINE);
p->skipvsp = 0;
while ('\0' != *word) {
if ('\\' != *word) {
if (TERMP_SKIPCHAR & p->flags) {
p->flags &= ~TERMP_SKIPCHAR;
word++;
continue;
}
if (TERMP_NBRWORD & p->flags) {
if (' ' == *word) {
encode(p, nbrsp, 1);
word++;
continue;
}
ssz = strcspn(word, "\\ ");
} else
ssz = strcspn(word, "\\");
encode(p, word, ssz);
word += (int)ssz;
continue;
}
word++;
esc = mandoc_escape(&word, &seq, &sz);
if (ESCAPE_ERROR == esc)
continue;
switch (esc) {
case ESCAPE_UNICODE:
uc = mchars_num2uc(seq + 1, sz - 1);
break;
case ESCAPE_NUMBERED:
uc = mchars_num2char(seq, sz);
if (uc < 0)
continue;
break;
case ESCAPE_SPECIAL:
if (p->enc == TERMENC_ASCII) {
cp = mchars_spec2str(p->symtab,
seq, sz, &ssz);
if (cp != NULL)
encode(p, cp, ssz);
} else {
uc = mchars_spec2cp(p->symtab, seq, sz);
if (uc > 0)
encode1(p, uc);
}
continue;
case ESCAPE_FONTBOLD:
term_fontrepl(p, TERMFONT_BOLD);
continue;
case ESCAPE_FONTITALIC:
term_fontrepl(p, TERMFONT_UNDER);
continue;
case ESCAPE_FONTBI:
term_fontrepl(p, TERMFONT_BI);
continue;
case ESCAPE_FONT:
/* FALLTHROUGH */
case ESCAPE_FONTROMAN:
term_fontrepl(p, TERMFONT_NONE);
continue;
case ESCAPE_FONTPREV:
term_fontlast(p);
continue;
case ESCAPE_NOSPACE:
if (TERMP_SKIPCHAR & p->flags)
p->flags &= ~TERMP_SKIPCHAR;
else if ('\0' == *word)
p->flags |= (TERMP_NOSPACE | TERMP_NONEWLINE);
continue;
case ESCAPE_SKIPCHAR:
p->flags |= TERMP_SKIPCHAR;
continue;
case ESCAPE_OVERSTRIKE:
cp = seq + sz;
while (seq < cp) {
if (*seq == '\\') {
mandoc_escape(&seq, NULL, NULL);
continue;
}
encode1(p, *seq++);
if (seq < cp)
encode(p, "\b", 1);
}
default:
continue;
}
/*
* Common handling for Unicode and numbered
* character escape sequences.
*/
if (p->enc == TERMENC_ASCII) {
cp = ascii_uc2str(uc);
encode(p, cp, strlen(cp));
} else {
if ((uc < 0x20 && uc != 0x09) ||
(uc > 0x7E && uc < 0xA0))
uc = 0xFFFD;
encode1(p, uc);
}
}
p->flags &= ~TERMP_NBRWORD;
}
/*
* Handle pwords, partial words, which may be either a single word or a
* phrase that cannot be broken down (such as a literal string). This
* handles word styling.
*/
void
term_word(struct termp *p, const char *word)
{
const char *seq, *cp;
char c;
int sz, uc;
size_t ssz;
enum mandoc_esc esc;
if ( ! (TERMP_NOSPACE & p->flags)) {
if ( ! (TERMP_KEEP & p->flags)) {
if (TERMP_PREKEEP & p->flags)
p->flags |= TERMP_KEEP;
bufferc(p, ' ');
if (TERMP_SENTENCE & p->flags)
bufferc(p, ' ');
} else
bufferc(p, ASCII_NBRSP);
}
if ( ! (p->flags & TERMP_NONOSPACE))
p->flags &= ~TERMP_NOSPACE;
else
p->flags |= TERMP_NOSPACE;
p->flags &= ~(TERMP_SENTENCE | TERMP_IGNDELIM);
while ('\0' != *word) {
if ((ssz = strcspn(word, "\\")) > 0)
encode(p, word, ssz);
word += (int)ssz;
if ('\\' != *word)
continue;
word++;
esc = mandoc_escape(&word, &seq, &sz);
if (ESCAPE_ERROR == esc)
break;
if (TERMENC_ASCII != p->enc)
switch (esc) {
case (ESCAPE_UNICODE):
uc = mchars_num2uc(seq + 1, sz - 1);
if ('\0' == uc)
break;
encode1(p, uc);
continue;
case (ESCAPE_SPECIAL):
uc = mchars_spec2cp(p->symtab, seq, sz);
if (uc <= 0)
break;
encode1(p, uc);
continue;
default:
break;
}
switch (esc) {
case (ESCAPE_UNICODE):
encode1(p, '?');
break;
case (ESCAPE_NUMBERED):
c = mchars_num2char(seq, sz);
if ('\0' != c)
encode(p, &c, 1);
break;
case (ESCAPE_SPECIAL):
cp = mchars_spec2str(p->symtab, seq, sz, &ssz);
if (NULL != cp)
encode(p, cp, ssz);
else if (1 == ssz)
encode(p, seq, sz);
break;
case (ESCAPE_FONTBOLD):
term_fontrepl(p, TERMFONT_BOLD);
break;
case (ESCAPE_FONTITALIC):
term_fontrepl(p, TERMFONT_UNDER);
break;
case (ESCAPE_FONT):
/* FALLTHROUGH */
case (ESCAPE_FONTROMAN):
term_fontrepl(p, TERMFONT_NONE);
break;
case (ESCAPE_FONTPREV):
term_fontlast(p);
break;
case (ESCAPE_NOSPACE):
if ('\0' == *word)
p->flags |= TERMP_NOSPACE;
break;
default:
break;
}
}
}