Locale support. I'm checking this in to clean up fall-out in-tree, but

[mandoc.git] / term.c

diff --git a/term.c b/term.c
index 37fc180d1cd30d71f17ae0921ff8d2cebe69c5e5..04d55d6730583bb940c467172906cdb979397be4 100644 (file)
--- a/term.c
+++ b/term.c
@@ -1,7 +1,7 @@
-/*     $Id: term.c,v 1.174 2010/10/02 15:15:55 schwarze Exp $ */
+/*     $Id: term.c,v 1.194 2011/05/17 22:32:45 kristaps Exp $ */

 /*
  * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
 /*
  * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>

- * Copyright (c) 2010 Ingo Schwarze <schwarze@openbsd.org>
+ * Copyright (c) 2010, 2011 Ingo Schwarze <schwarze@openbsd.org>

  *
  * Permission to use, copy, modify, and distribute this software for any
  * purpose with or without fee is hereby granted, provided that the above
  *
  * Permission to use, copy, modify, and distribute this software for any
  * purpose with or without fee is hereby granted, provided that the above


@@ -29,18 +29,14 @@
 #include <string.h>
 
 #include "mandoc.h"
 #include <string.h>
 
 #include "mandoc.h"

-#include "chars.h"

 #include "out.h"
 #include "term.h"
 #include "main.h"
 
 #include "out.h"
 #include "term.h"
 #include "main.h"
 

-static void              spec(struct termp *, enum roffdeco,
-                               const char *, size_t);
-static void              res(struct termp *, const char *, size_t);
-static void              bufferc(struct termp *, char);
-static void              adjbuf(struct termp *p, size_t);
-static void              encode(struct termp *, const char *, size_t);
-
+static void             adjbuf(struct termp *p, int);
+static void             bufferc(struct termp *, char);
+static void             encode(struct termp *, const char *, size_t);
+static void             encode1(struct termp *, int);

 
 void
 term_free(struct termp *p)
 
 void
 term_free(struct termp *p)


@@ -49,7 +45,7 @@ term_free(struct termp *p)
        if (p->buf)
                free(p->buf);
        if (p->symtab)
        if (p->buf)
                free(p->buf);
        if (p->symtab)

-               chars_free(p->symtab);
+               mchars_free(p->symtab);

 
        free(p);
 }
 
        free(p);
 }


@@ -74,23 +70,6 @@ term_end(struct termp *p)
        (*p->end)(p);
 }
 
        (*p->end)(p);
 }
 

-
-struct termp *
-term_alloc(enum termenc enc)
-{
-       struct termp    *p;
-
-       p = calloc(1, sizeof(struct termp));
-       if (NULL == p) {
-               perror(NULL);
-               exit((int)MANDOCLEVEL_SYSERR);
-       }
-
-       p->enc = enc;
-       return(p);
-}
-
-

 /*
  * Flush a line of text.  A "line" is loosely defined as being something
  * that should be followed by a newline, regardless of whether it's
 /*
  * Flush a line of text.  A "line" is loosely defined as being something
  * that should be followed by a newline, regardless of whether it's


@@ -146,7 +125,7 @@ term_flushln(struct termp *p)
         * an indentation, but can be, for tagged lists or columns, a
         * small set of values. 
         */
         * an indentation, but can be, for tagged lists or columns, a
         * small set of values. 
         */

-       assert  (p->rmargin > p->offset);
+       assert  (p->rmargin >= p->offset);

        dv     = p->rmargin - p->offset;
        maxvis = (int)dv > p->overstep ? dv - (size_t)p->overstep : 0;
        dv     = p->maxrmargin - p->offset;
        dv     = p->rmargin - p->offset;
        maxvis = (int)dv > p->overstep ? dv - (size_t)p->overstep : 0;
        dv     = p->maxrmargin - p->offset;


@@ -162,12 +141,12 @@ term_flushln(struct termp *p)
        vis = vend = 0;
        i = 0;
 
        vis = vend = 0;
        i = 0;
 

-       while (i < (int)p->col) {
+       while (i < p->col) {

                /*
                 * Handle literal tab characters: collapse all
                 * subsequent tabs into a single huge set of spaces.
                 */
                /*
                 * Handle literal tab characters: collapse all
                 * subsequent tabs into a single huge set of spaces.
                 */

-               while (i < (int)p->col && '\t' == p->buf[i]) {
+               while (i < p->col && '\t' == p->buf[i]) {

                        vend = (vis / p->tabwidth + 1) * p->tabwidth;
                        vbl += vend - vis;
                        vis = vend;
                        vend = (vis / p->tabwidth + 1) * p->tabwidth;
                        vbl += vend - vis;
                        vis = vend;


@@ -181,7 +160,7 @@ term_flushln(struct termp *p)
                 * space is printed according to regular spacing rules).
                 */
 
                 * space is printed according to regular spacing rules).
                 */
 

-               for (j = i, jhy = 0; j < (int)p->col; j++) {
+               for (j = i, jhy = 0; j < p->col; j++) {

                        if ((j && ' ' == p->buf[j]) || '\t' == p->buf[j])
                                break;
 
                        if ((j && ' ' == p->buf[j]) || '\t' == p->buf[j])
                                break;
 


@@ -224,7 +203,7 @@ term_flushln(struct termp *p)
                }
 
                /* Write out the [remaining] word. */
                }
 
                /* Write out the [remaining] word. */

-               for ( ; i < (int)p->col; i++) {
+               for ( ; i < p->col; i++) {

                        if (vend > bp && jhy > 0 && i > jhy)
                                break;
                        if ('\t' == p->buf[i])
                        if (vend > bp && jhy > 0 && i > jhy)
                                break;
                        if ('\t' == p->buf[i])


@@ -351,33 +330,6 @@ term_vspace(struct termp *p)
        (*p->endline)(p);
 }
 
        (*p->endline)(p);
 }
 

-
-static void
-spec(struct termp *p, enum roffdeco d, const char *word, size_t len)
-{
-       const char      *rhs;
-       size_t           sz;
-
-       rhs = chars_spec2str(p->symtab, word, len, &sz);
-       if (rhs) 
-               encode(p, rhs, sz);
-       else if (DECO_SSPECIAL == d)
-               encode(p, word, len);
-}
-
-
-static void
-res(struct termp *p, const char *word, size_t len)
-{
-       const char      *rhs;
-       size_t           sz;
-
-       rhs = chars_res2str(p->symtab, word, len, &sz);
-       if (rhs)
-               encode(p, rhs, sz);
-}
-
-

 void
 term_fontlast(struct termp *p)
 {
 void
 term_fontlast(struct termp *p)
 {


@@ -442,7 +394,6 @@ term_fontpop(struct termp *p)
        p->fonti--;
 }
 
        p->fonti--;
 }
 

-

 /*
  * 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
 /*
  * 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


@@ -451,35 +402,11 @@ term_fontpop(struct termp *p)
 void
 term_word(struct termp *p, const char *word)
 {
 void
 term_word(struct termp *p, const char *word)
 {

-       const char      *sv, *seq;
+       const char      *seq, *cp;
+       char             c;
+       int              sz, uc;

        size_t           ssz;
        size_t           ssz;

-       enum roffdeco    deco;
-
-       sv = word;
-
-       if (word[0] && '\0' == word[1])
-               switch (word[0]) {
-               case('.'):
-                       /* FALLTHROUGH */
-               case(','):
-                       /* FALLTHROUGH */
-               case(';'):
-                       /* FALLTHROUGH */
-               case(':'):
-                       /* FALLTHROUGH */
-               case('?'):
-                       /* FALLTHROUGH */
-               case('!'):
-                       /* FALLTHROUGH */
-               case(')'):
-                       /* FALLTHROUGH */
-               case(']'):
-                       if ( ! (TERMP_IGNDELIM & p->flags))
-                               p->flags |= TERMP_NOSPACE;
-                       break;
-               default:
-                       break;
-               }
+       enum mandoc_esc  esc;

 
        if ( ! (TERMP_NOSPACE & p->flags)) {
                if ( ! (TERMP_KEEP & p->flags)) {
 
        if ( ! (TERMP_NOSPACE & p->flags)) {
                if ( ! (TERMP_KEEP & p->flags)) {


@@ -499,65 +426,69 @@ term_word(struct termp *p, const char *word)
 
        p->flags &= ~(TERMP_SENTENCE | TERMP_IGNDELIM);
 
 
        p->flags &= ~(TERMP_SENTENCE | TERMP_IGNDELIM);
 

-       while (*word) {
+       while ('\0' != *word) {

                if ((ssz = strcspn(word, "\\")) > 0)
                        encode(p, word, ssz);
 
                if ((ssz = strcspn(word, "\\")) > 0)
                        encode(p, word, ssz);
 

-               word += ssz;
+               word += (int)ssz;

                if ('\\' != *word)
                        continue;
 
                if ('\\' != *word)
                        continue;
 

-               seq = ++word;
-               word += a2roffdeco(&deco, &seq, &ssz);
+               word++;
+               esc = mandoc_escape(&word, &seq, &sz);
+               if (ESCAPE_ERROR == esc)
+                       break;

 
 

-               switch (deco) {
-               case (DECO_RESERVED):
-                       res(p, seq, ssz);
+               switch (esc) {
+               case (ESCAPE_UNICODE):
+                       if (TERMENC_ASCII == p->enc) {
+                               encode1(p, '?');
+                               break;
+                       }
+                       uc = mchars_num2uc(seq + 1, sz - 1);
+                       if ('\0' != uc)
+                               encode1(p, uc);

                        break;
                        break;

-               case (DECO_SPECIAL):
-                       /* FALLTHROUGH */
-               case (DECO_SSPECIAL):
-                       spec(p, deco, seq, ssz);
+               case (ESCAPE_NUMBERED):
+                       if ('\0' != (c = mchars_num2char(seq, sz)))
+                               encode(p, &c, 1);

                        break;
                        break;

-               case (DECO_BOLD):
+               case (ESCAPE_PREDEF):
+                       cp = mchars_res2str(p->symtab, seq, sz, &ssz);
+                       if (NULL != cp)
+                               encode(p, cp, ssz);
+                       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;
                        term_fontrepl(p, TERMFONT_BOLD);
                        break;

-               case (DECO_ITALIC):
+               case (ESCAPE_FONTITALIC):

                        term_fontrepl(p, TERMFONT_UNDER);
                        break;
                        term_fontrepl(p, TERMFONT_UNDER);
                        break;

-               case (DECO_ROMAN):
+               case (ESCAPE_FONTROMAN):

                        term_fontrepl(p, TERMFONT_NONE);
                        break;
                        term_fontrepl(p, TERMFONT_NONE);
                        break;

-               case (DECO_PREVIOUS):
+               case (ESCAPE_FONTPREV):

                        term_fontlast(p);
                        break;
                        term_fontlast(p);
                        break;

-               default:
-                       break;
-               }
-
-               if (DECO_NOSPACE == deco && '\0' == *word)
-                       p->flags |= TERMP_NOSPACE;
-       }
-
-       /* 
-        * Note that we don't process the pipe: the parser sees it as
-        * punctuation, but we don't in terms of typography.
-        */
-       if (sv[0] && '\0' == sv[1])
-               switch (sv[0]) {
-               case('('):
-                       /* FALLTHROUGH */
-               case('['):
-                       p->flags |= TERMP_NOSPACE;
+               case (ESCAPE_NOSPACE):
+                       if ('\0' == *word)
+                               p->flags |= TERMP_NOSPACE;

                        break;
                default:
                        break;
                }
                        break;
                default:
                        break;
                }

+       }

 }
 
 }
 

-

 static void
 static void

-adjbuf(struct termp *p, size_t sz)
+adjbuf(struct termp *p, int sz)

 {
 
        if (0 == p->maxcols)
 {
 
        if (0 == p->maxcols)


@@ -565,14 +496,10 @@ adjbuf(struct termp *p, size_t sz)
        while (sz >= p->maxcols)
                p->maxcols <<= 2;
 
        while (sz >= p->maxcols)
                p->maxcols <<= 2;
 

-       p->buf = realloc(p->buf, p->maxcols);
-       if (NULL == p->buf) {
-               perror(NULL);
-               exit((int)MANDOCLEVEL_SYSERR);
-       }
+       p->buf = mandoc_realloc
+               (p->buf, sizeof(int) * (size_t)p->maxcols);

 }
 
 }
 

-

 static void
 bufferc(struct termp *p, char c)
 {
 static void
 bufferc(struct termp *p, char c)
 {


@@ -580,15 +507,44 @@ bufferc(struct termp *p, char c)
        if (p->col + 1 >= p->maxcols)
                adjbuf(p, p->col + 1);
 
        if (p->col + 1 >= p->maxcols)
                adjbuf(p, p->col + 1);
 

-       p->buf[(int)p->col++] = c;
+       p->buf[p->col++] = c;

 }
 
 }
 

+/*
+ * See encode().
+ * Do this for a single (probably unicode) value.
+ * Does not check for non-decorated glyphs.
+ */
+static void
+encode1(struct termp *p, int c)
+{
+       enum termfont     f;
+
+       if (p->col + 4 >= p->maxcols)
+               adjbuf(p, p->col + 4);
+
+       f = term_fonttop(p);
+
+       if (TERMFONT_NONE == f) {
+               p->buf[p->col++] = c;
+               return;
+       } else if (TERMFONT_UNDER == f) {
+               p->buf[p->col++] = '_';
+       } else
+               p->buf[p->col++] = c;
+
+       p->buf[p->col++] = 8;
+       p->buf[p->col++] = c;
+}

 
 static void
 encode(struct termp *p, const char *word, size_t sz)
 {
        enum termfont     f;
 
 static void
 encode(struct termp *p, const char *word, size_t sz)
 {
        enum termfont     f;

-       int               i;
+       int               i, len;
+
+       /* LINTED */
+       len = sz;

 
        /*
         * Encode and buffer a string of characters.  If the current
 
        /*
         * Encode and buffer a string of characters.  If the current


@@ -597,35 +553,34 @@ encode(struct termp *p, const char *word, size_t sz)
         */
 
        if (TERMFONT_NONE == (f = term_fonttop(p))) {
         */
 
        if (TERMFONT_NONE == (f = term_fonttop(p))) {

-               if (p->col + sz >= p->maxcols) 
-                       adjbuf(p, p->col + sz);
-               memcpy(&p->buf[(int)p->col], word, sz);
-               p->col += sz;
+               if (p->col + len >= p->maxcols) 
+                       adjbuf(p, p->col + len);
+               for (i = 0; i < len; i++)
+                       p->buf[p->col++] = word[i];

                return;
        }
 
        /* Pre-buffer, assuming worst-case. */
 
                return;
        }
 
        /* Pre-buffer, assuming worst-case. */
 

-       if (p->col + 1 + (sz * 3) >= p->maxcols)
-               adjbuf(p, p->col + 1 + (sz * 3));
+       if (p->col + 1 + (len * 3) >= p->maxcols)
+               adjbuf(p, p->col + 1 + (len * 3));

 
 

-       for (i = 0; i < (int)sz; i++) {
-               if ( ! isgraph((u_char)word[i])) {
-                       p->buf[(int)p->col++] = word[i];
+       for (i = 0; i < len; i++) {
+               if ( ! isgraph((unsigned char)word[i])) {
+                       p->buf[p->col++] = word[i];

                        continue;
                }
 
                if (TERMFONT_UNDER == f)
                        continue;
                }
 
                if (TERMFONT_UNDER == f)

-                       p->buf[(int)p->col++] = '_';
+                       p->buf[p->col++] = '_';

                else
                else

-                       p->buf[(int)p->col++] = word[i];
+                       p->buf[p->col++] = word[i];

 
 

-               p->buf[(int)p->col++] = 8;
-               p->buf[(int)p->col++] = word[i];
+               p->buf[p->col++] = 8;
+               p->buf[p->col++] = word[i];

        }
 }
 
        }
 }
 

-

 size_t
 term_len(const struct termp *p, size_t sz)
 {
 size_t
 term_len(const struct termp *p, size_t sz)
 {


@@ -637,53 +592,85 @@ term_len(const struct termp *p, size_t sz)
 size_t
 term_strlen(const struct termp *p, const char *cp)
 {
 size_t
 term_strlen(const struct termp *p, const char *cp)
 {

-       size_t           sz, ssz, rsz, i;
-       enum roffdeco    d;
+       size_t           sz, rsz, i;
+       int              ssz, c;

        const char      *seq, *rhs;
        const char      *seq, *rhs;

+       static const char rej[] = { '\\', ASCII_HYPH, ASCII_NBRSP, '\0' };

 
 

-       for (sz = 0; '\0' != *cp; )
-               /*
-                * Account for escaped sequences within string length
-                * calculations.  This follows the logic in term_word()
-                * as we must calculate the width of produced strings.
-                */
-               if ('\\' == *cp) {
-                       seq = ++cp;
-                       cp += a2roffdeco(&d, &seq, &ssz);
+       /*
+        * 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++);

 
 

-                       switch (d) {
-                       case (DECO_RESERVED):
-                               rhs = chars_res2str
+               c = 0;
+               switch (*cp) {
+               case ('\\'):
+                       cp++;
+                       rhs = NULL;
+                       switch (mandoc_escape(&cp, &seq, &ssz)) {
+                       case (ESCAPE_ERROR):
+                               return(sz);
+                       case (ESCAPE_UNICODE):
+                               if (TERMENC_ASCII != p->enc) {
+                                       sz += (*p->width)(p, '?');
+                                       break;
+                               }
+                               c = mchars_num2uc(seq + 1, ssz - 1);
+                               if ('\0' != c)
+                                       sz += (*p->width)(p, c);
+                               break;
+                       case (ESCAPE_NUMBERED):
+                               c = mchars_num2char(seq, ssz);
+                               if ('\0' != c)
+                                       sz += (*p->width)(p, c);
+                               break;
+                       case (ESCAPE_PREDEF):
+                               rhs = mchars_res2str

                                        (p->symtab, seq, ssz, &rsz);
                                break;
                                        (p->symtab, seq, ssz, &rsz);
                                break;

-                       case (DECO_SPECIAL):
-                               /* FALLTHROUGH */
-                       case (DECO_SSPECIAL):
-                               rhs = chars_spec2str
+                       case (ESCAPE_SPECIAL):
+                               rhs = mchars_spec2str

                                        (p->symtab, seq, ssz, &rsz);
 
                                        (p->symtab, seq, ssz, &rsz);
 

-                               /* Allow for one-char escapes. */
-                               if (DECO_SSPECIAL != d || rhs)
+                               if (ssz != 1 || rhs)

                                        break;
 
                                rhs = seq;
                                rsz = ssz;
                                break;
                        default:
                                        break;
 
                                rhs = seq;
                                rsz = ssz;
                                break;
                        default:

-                               rhs = NULL;

                                break;
                        }
 
                                break;
                        }
 

-                       if (rhs)
-                               for (i = 0; i < rsz; i++)
-                                       sz += (*p->width)(p, *rhs++);
-               } else
-                       sz += (*p->width)(p, *cp++);
+                       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);
 }
 
 
        return(sz);
 }
 

-

 /* ARGSUSED */
 size_t
 term_vspan(const struct termp *p, const struct roffsu *su)
 /* ARGSUSED */
 size_t
 term_vspan(const struct termp *p, const struct roffsu *su)


@@ -720,7 +707,6 @@ term_vspan(const struct termp *p, const struct roffsu *su)
                        r);
 }
 
                        r);
 }
 

-

 size_t
 term_hspan(const struct termp *p, const struct roffsu *su)
 {
 size_t
 term_hspan(const struct termp *p, const struct roffsu *su)
 {