version 1.77, 2009/06/11 07:46:41 |
version 1.86, 2009/07/15 15:37:48 |
Line 31 extern int mdoc_run(struct termp *, |
|
Line 31 extern int mdoc_run(struct termp *, |
|
|
|
static struct termp *term_alloc(enum termenc); |
static struct termp *term_alloc(enum termenc); |
static void term_free(struct termp *); |
static void term_free(struct termp *); |
static void term_pword(struct termp *, const char *, int); |
static void term_pescape(struct termp *, const char **); |
static void term_pescape(struct termp *, |
|
const char *, int *, int); |
|
static void term_nescape(struct termp *, |
static void term_nescape(struct termp *, |
const char *, size_t); |
const char *, size_t); |
static void term_chara(struct termp *, char); |
static void term_chara(struct termp *, char); |
static void term_stringa(struct termp *, |
static void term_encodea(struct termp *, char); |
const char *, size_t); |
static int term_isopendelim(const char *); |
static int term_isopendelim(const char *, int); |
static int term_isclosedelim(const char *); |
static int term_isclosedelim(const char *, int); |
|
|
|
|
|
void * |
void * |
Line 106 term_alloc(enum termenc enc) |
|
Line 103 term_alloc(enum termenc enc) |
|
if (NULL == (p = malloc(sizeof(struct termp)))) |
if (NULL == (p = malloc(sizeof(struct termp)))) |
err(1, "malloc"); |
err(1, "malloc"); |
bzero(p, sizeof(struct termp)); |
bzero(p, sizeof(struct termp)); |
p->maxrmargin = 80; |
p->maxrmargin = 78; |
p->enc = enc; |
p->enc = enc; |
return(p); |
return(p); |
} |
} |
|
|
|
|
static int |
static int |
term_isclosedelim(const char *p, int len) |
term_isclosedelim(const char *p) |
{ |
{ |
|
|
if (1 != len) |
if ( ! (*p && 0 == *(p + 1))) |
return(0); |
return(0); |
|
|
switch (*p) { |
switch (*p) { |
Line 147 term_isclosedelim(const char *p, int len) |
|
Line 144 term_isclosedelim(const char *p, int len) |
|
|
|
|
|
static int |
static int |
term_isopendelim(const char *p, int len) |
term_isopendelim(const char *p) |
{ |
{ |
|
|
if (1 != len) |
if ( ! (*p && 0 == *(p + 1))) |
return(0); |
return(0); |
|
|
switch (*p) { |
switch (*p) { |
Line 177 term_isopendelim(const char *p, int len) |
|
Line 174 term_isopendelim(const char *p, int len) |
|
* Specifically, a line is whatever's in p->buf of length p->col, which |
* Specifically, a line is whatever's in p->buf of length p->col, which |
* is zeroed after this function returns. |
* is zeroed after this function returns. |
* |
* |
* The variables TERMP_NOLPAD, TERMP_LITERAL and TERMP_NOBREAK are of |
* The usage of termp:flags is as follows: |
* critical importance here. Their behaviour follows: |
|
* |
* |
* - TERMP_NOLPAD: when beginning to write the line, don't left-pad the |
* - TERMP_NOLPAD: when beginning to write the line, don't left-pad the |
* offset value. This is useful when doing columnar lists where the |
* offset value. This is useful when doing columnar lists where the |
Line 188 term_isopendelim(const char *p, int len) |
|
Line 184 term_isopendelim(const char *p, int len) |
|
* columns. In short: don't print a newline and instead pad to the |
* columns. In short: don't print a newline and instead pad to the |
* right margin. Used in conjunction with TERMP_NOLPAD. |
* right margin. Used in conjunction with TERMP_NOLPAD. |
* |
* |
* - TERMP_NONOBREAK: don't newline when TERMP_NOBREAK is specified. |
* - TERMP_DANGLE: don't newline when TERMP_NOBREAK is specified and |
|
* the line is overrun, and don't pad-right if it's underrun. |
* |
* |
|
* - TERMP_HANG: like TERMP_DANGLE, but doesn't newline when |
|
* overruning, instead save the position and continue at that point |
|
* when the next invocation. |
|
* |
* In-line line breaking: |
* In-line line breaking: |
* |
* |
* If TERMP_NOBREAK is specified and the line overruns the right |
* If TERMP_NOBREAK is specified and the line overruns the right |
Line 201 term_isopendelim(const char *p, int len) |
|
Line 202 term_isopendelim(const char *p, int len) |
|
* Otherwise, the line will break at the right margin. Extremely long |
* Otherwise, the line will break at the right margin. Extremely long |
* lines will cause the system to emit a warning (TODO: hyphenate, if |
* lines will cause the system to emit a warning (TODO: hyphenate, if |
* possible). |
* possible). |
|
* |
|
* FIXME: newline breaks occur (in groff) also occur when a single |
|
* space follows a NOBREAK! |
*/ |
*/ |
void |
void |
term_flushln(struct termp *p) |
term_flushln(struct termp *p) |
{ |
{ |
int i, j; |
int i, j; |
size_t vsz, vis, maxvis, mmax, bp; |
size_t vbl, vsz, vis, maxvis, mmax, bp; |
|
static int sv = -1; |
|
|
/* |
/* |
* First, establish the maximum columns of "visible" content. |
* First, establish the maximum columns of "visible" content. |
Line 221 term_flushln(struct termp *p) |
|
Line 226 term_flushln(struct termp *p) |
|
bp = TERMP_NOBREAK & p->flags ? mmax : maxvis; |
bp = TERMP_NOBREAK & p->flags ? mmax : maxvis; |
vis = 0; |
vis = 0; |
|
|
|
if (sv >= 0) { |
|
vis = (size_t)sv; |
|
sv = -1; |
|
} |
|
|
/* |
/* |
* If in the standard case (left-justified), then begin with our |
* If in the standard case (left-justified), then begin with our |
* indentation, otherwise (columns, etc.) just start spitting |
* indentation, otherwise (columns, etc.) just start spitting |
Line 251 term_flushln(struct termp *p) |
|
Line 261 term_flushln(struct termp *p) |
|
} |
} |
|
|
/* |
/* |
* Do line-breaking. If we're greater than our |
* Choose the number of blanks to prepend: no blank at the |
* break-point and already in-line, break to the next |
* beginning of a line, one between words -- but do not |
* line and start writing. If we're at the line start, |
* actually write them yet. |
* then write out the word (TODO: hyphenate) and break |
|
* in a subsequent loop invocation. |
|
*/ |
*/ |
|
vbl = (size_t)(0 == vis ? 0 : 1); |
|
|
if ( ! (TERMP_NOBREAK & p->flags)) { |
/* |
if (vis && vis + vsz > bp) { |
* Find out whether we would exceed the right margin. |
putchar('\n'); |
* If so, break to the next line. (TODO: hyphenate) |
|
* Otherwise, write the chosen number of blanks now. |
|
*/ |
|
if (vis && vis + vbl + vsz > bp) { |
|
putchar('\n'); |
|
if (TERMP_NOBREAK & p->flags) { |
|
for (j = 0; j < (int)p->rmargin; j++) |
|
putchar(' '); |
|
vis = p->rmargin - p->offset; |
|
} else { |
for (j = 0; j < (int)p->offset; j++) |
for (j = 0; j < (int)p->offset; j++) |
putchar(' '); |
putchar(' '); |
vis = 0; |
vis = 0; |
} |
} |
} else if (vis && vis + vsz > bp) { |
} else { |
putchar('\n'); |
for (j = 0; j < (int)vbl; j++) |
for (j = 0; j < (int)p->rmargin; j++) |
|
putchar(' '); |
putchar(' '); |
vis = p->rmargin - p->offset; |
vis += vbl; |
} |
} |
|
|
/* |
/* |
* Write out the word and a trailing space. Omit the |
* Finally, write out the word. |
* space if we're the last word in the line or beyond |
|
* our breakpoint. |
|
*/ |
*/ |
|
|
if (0 < vis++) |
|
putchar(' '); |
|
|
|
for ( ; i < (int)p->col; i++) { |
for ( ; i < (int)p->col; i++) { |
if (' ' == p->buf[i]) |
if (' ' == p->buf[i]) |
break; |
break; |
Line 295 term_flushln(struct termp *p) |
|
Line 306 term_flushln(struct termp *p) |
|
*/ |
*/ |
|
|
if ((TERMP_NOBREAK & p->flags) && vis >= maxvis) { |
if ((TERMP_NOBREAK & p->flags) && vis >= maxvis) { |
if ( ! (TERMP_NONOBREAK & p->flags)) { |
if ( ! (TERMP_DANGLE & p->flags) && |
|
! (TERMP_HANG & p->flags)) { |
putchar('\n'); |
putchar('\n'); |
for (i = 0; i < (int)p->rmargin; i++) |
for (i = 0; i < (int)p->rmargin; i++) |
putchar(' '); |
putchar(' '); |
} |
} |
|
if (TERMP_HANG & p->flags) |
|
sv = (int)(vis - maxvis); |
p->col = 0; |
p->col = 0; |
return; |
return; |
} |
} |
Line 310 term_flushln(struct termp *p) |
|
Line 324 term_flushln(struct termp *p) |
|
*/ |
*/ |
|
|
if (p->flags & TERMP_NOBREAK) { |
if (p->flags & TERMP_NOBREAK) { |
if ( ! (TERMP_NONOBREAK & p->flags)) |
if ( ! (TERMP_DANGLE & p->flags)) |
for ( ; vis <= maxvis; vis++) |
for ( ; vis < maxvis; vis++) |
putchar(' '); |
putchar(' '); |
} else |
} else |
putchar('\n'); |
putchar('\n'); |
Line 355 term_vspace(struct termp *p) |
|
Line 369 term_vspace(struct termp *p) |
|
|
|
|
|
/* |
/* |
* Break apart a word into "pwords" (partial-words, usually from |
|
* breaking up a phrase into individual words) and, eventually, put them |
|
* into the output buffer. If we're a literal word, then don't break up |
|
* the word and put it verbatim into the output buffer. |
|
*/ |
|
void |
|
term_word(struct termp *p, const char *word) |
|
{ |
|
int i, j, len; |
|
|
|
len = (int)strlen(word); |
|
|
|
if (p->flags & TERMP_LITERAL) { |
|
term_pword(p, word, len); |
|
return; |
|
} |
|
|
|
/* LINTED */ |
|
for (j = i = 0; i < len; i++) { |
|
if (' ' != word[i]) { |
|
j++; |
|
continue; |
|
} |
|
|
|
/* Escaped spaces don't delimit... */ |
|
if (i && ' ' == word[i] && '\\' == word[i - 1]) { |
|
j++; |
|
continue; |
|
} |
|
|
|
if (0 == j) |
|
continue; |
|
assert(i >= j); |
|
term_pword(p, &word[i - j], j); |
|
j = 0; |
|
} |
|
if (j > 0) { |
|
assert(i >= j); |
|
term_pword(p, &word[i - j], j); |
|
} |
|
} |
|
|
|
|
|
/* |
|
* Determine the symbol indicated by an escape sequences, that is, one |
* Determine the symbol indicated by an escape sequences, that is, one |
* starting with a backslash. Once done, we pass this value into the |
* starting with a backslash. Once done, we pass this value into the |
* output buffer by way of the symbol table. |
* output buffer by way of the symbol table. |
Line 408 term_nescape(struct termp *p, const char *word, size_t |
|
Line 378 term_nescape(struct termp *p, const char *word, size_t |
|
{ |
{ |
const char *rhs; |
const char *rhs; |
size_t sz; |
size_t sz; |
|
int i; |
|
|
if (NULL == (rhs = term_a2ascii(p->symtab, word, len, &sz))) |
rhs = term_a2ascii(p->symtab, word, len, &sz); |
return; |
|
term_stringa(p, rhs, sz); |
if (rhs) |
|
for (i = 0; i < (int)sz; i++) |
|
term_encodea(p, rhs[i]); |
} |
} |
|
|
|
|
Line 421 term_nescape(struct termp *p, const char *word, size_t |
|
Line 394 term_nescape(struct termp *p, const char *word, size_t |
|
* the escape sequence (we assert upon badly-formed escape sequences). |
* the escape sequence (we assert upon badly-formed escape sequences). |
*/ |
*/ |
static void |
static void |
term_pescape(struct termp *p, const char *word, int *i, int len) |
term_pescape(struct termp *p, const char **word) |
{ |
{ |
int j; |
int j; |
|
const char *wp; |
|
|
if (++(*i) >= len) |
wp = *word; |
|
|
|
if (0 == *(++wp)) { |
|
*word = wp; |
return; |
return; |
|
} |
|
|
if ('(' == word[*i]) { |
if ('(' == *wp) { |
(*i)++; |
wp++; |
if (*i + 1 >= len) |
if (0 == *wp || 0 == *(wp + 1)) { |
|
*word = 0 == *wp ? wp : wp + 1; |
return; |
return; |
|
} |
|
|
term_nescape(p, &word[*i], 2); |
term_nescape(p, wp, 2); |
(*i)++; |
*word = ++wp; |
return; |
return; |
|
|
} else if ('*' == word[*i]) { |
} else if ('*' == *wp) { |
(*i)++; |
if (0 == *(++wp)) { |
if (*i >= len) |
*word = wp; |
return; |
return; |
|
} |
|
|
switch (word[*i]) { |
switch (*wp) { |
case ('('): |
case ('('): |
(*i)++; |
wp++; |
if (*i + 1 >= len) |
if (0 == *wp || 0 == *(wp + 1)) { |
|
*word = 0 == *wp ? wp : wp + 1; |
return; |
return; |
|
} |
|
|
term_nescape(p, &word[*i], 2); |
term_nescape(p, wp, 2); |
(*i)++; |
*word = ++wp; |
return; |
return; |
case ('['): |
case ('['): |
break; |
break; |
default: |
default: |
term_nescape(p, &word[*i], 1); |
term_nescape(p, wp, 1); |
|
*word = wp; |
return; |
return; |
} |
} |
|
|
} else if ('f' == word[*i]) { |
} else if ('f' == *wp) { |
(*i)++; |
if (0 == *(++wp)) { |
if (*i >= len) |
*word = wp; |
return; |
return; |
switch (word[*i]) { |
} |
|
|
|
switch (*wp) { |
case ('B'): |
case ('B'): |
p->flags |= TERMP_BOLD; |
p->flags |= TERMP_BOLD; |
break; |
break; |
Line 477 term_pescape(struct termp *p, const char *word, int *i |
|
Line 463 term_pescape(struct termp *p, const char *word, int *i |
|
default: |
default: |
break; |
break; |
} |
} |
|
|
|
*word = wp; |
return; |
return; |
|
|
} else if ('[' != word[*i]) { |
} else if ('[' != *wp) { |
term_nescape(p, &word[*i], 1); |
term_nescape(p, wp, 1); |
|
*word = wp; |
return; |
return; |
} |
} |
|
|
(*i)++; |
wp++; |
for (j = 0; word[*i] && ']' != word[*i]; (*i)++, j++) |
for (j = 0; *wp && ']' != *wp; wp++, j++) |
/* Loop... */ ; |
/* Loop... */ ; |
|
|
if (0 == word[*i]) |
if (0 == *wp) { |
|
*word = wp; |
return; |
return; |
|
} |
|
|
term_nescape(p, &word[*i - j], (size_t)j); |
term_nescape(p, wp - j, (size_t)j); |
|
*word = wp; |
} |
} |
|
|
|
|
Line 500 term_pescape(struct termp *p, const char *word, int *i |
|
Line 492 term_pescape(struct termp *p, const char *word, int *i |
|
* phrase that cannot be broken down (such as a literal string). This |
* phrase that cannot be broken down (such as a literal string). This |
* handles word styling. |
* handles word styling. |
*/ |
*/ |
static void |
void |
term_pword(struct termp *p, const char *word, int len) |
term_word(struct termp *p, const char *word) |
{ |
{ |
int i; |
|
|
|
if (term_isclosedelim(word, len)) |
assert(*word); |
|
if (term_isclosedelim(word)) |
if ( ! (TERMP_IGNDELIM & p->flags)) |
if ( ! (TERMP_IGNDELIM & p->flags)) |
p->flags |= TERMP_NOSPACE; |
p->flags |= TERMP_NOSPACE; |
|
|
Line 520 term_pword(struct termp *p, const char *word, int len) |
|
Line 512 term_pword(struct termp *p, const char *word, int len) |
|
* before the word. |
* before the word. |
*/ |
*/ |
|
|
for (i = 0; i < len; i++) { |
for ( ; *word; word++) |
if ('\\' == word[i]) { |
if ('\\' != *word) |
term_pescape(p, word, &i, len); |
term_encodea(p, *word); |
continue; |
else |
} |
term_pescape(p, &word); |
|
|
if (TERMP_STYLE & p->flags) { |
if (term_isopendelim(word)) |
if (TERMP_BOLD & p->flags) { |
|
term_chara(p, word[i]); |
|
term_chara(p, 8); |
|
} |
|
if (TERMP_UNDER & p->flags) { |
|
term_chara(p, '_'); |
|
term_chara(p, 8); |
|
} |
|
} |
|
|
|
term_chara(p, word[i]); |
|
} |
|
|
|
if (term_isopendelim(word, len)) |
|
p->flags |= TERMP_NOSPACE; |
p->flags |= TERMP_NOSPACE; |
} |
} |
|
|
|
|
/* |
/* |
* Like term_chara() but for arbitrary-length buffers. Resize the |
|
* buffer by a factor of two (if the buffer is less than that) or the |
|
* buffer's size. |
|
*/ |
|
static void |
|
term_stringa(struct termp *p, const char *c, size_t sz) |
|
{ |
|
size_t s; |
|
|
|
if (0 == sz) |
|
return; |
|
|
|
assert(c); |
|
if (p->col + sz >= p->maxcols) { |
|
if (0 == p->maxcols) |
|
p->maxcols = 256; |
|
s = sz > p->maxcols * 2 ? sz : p->maxcols * 2; |
|
p->buf = realloc(p->buf, s); |
|
if (NULL == p->buf) |
|
err(1, "realloc"); |
|
p->maxcols = s; |
|
} |
|
|
|
(void)memcpy(&p->buf[(int)p->col], c, sz); |
|
p->col += sz; |
|
} |
|
|
|
|
|
/* |
|
* Insert a single character into the line-buffer. If the buffer's |
* Insert a single character into the line-buffer. If the buffer's |
* space is exceeded, then allocate more space by doubling the buffer |
* space is exceeded, then allocate more space by doubling the buffer |
* size. |
* size. |
Line 596 term_chara(struct termp *p, char c) |
|
Line 545 term_chara(struct termp *p, char c) |
|
p->buf[(int)(p->col)++] = c; |
p->buf[(int)(p->col)++] = c; |
} |
} |
|
|
|
|
|
static void |
|
term_encodea(struct termp *p, char c) |
|
{ |
|
|
|
if (TERMP_STYLE & p->flags) { |
|
if (TERMP_BOLD & p->flags) { |
|
term_chara(p, c); |
|
term_chara(p, 8); |
|
} |
|
if (TERMP_UNDER & p->flags) { |
|
term_chara(p, '_'); |
|
term_chara(p, 8); |
|
} |
|
} |
|
term_chara(p, c); |
|
} |