Allow non-ASCII terminal encodings to accept unicode values for the

[mandoc.git] / term.c

diff --git a/term.c b/term.c
index bfedefb0ff31e718b8f5f80d5ce3ddd619853414..2bcdb69dfffe9d84bf840c7788ba15c99d321d97 100644 (file)
--- a/term.c
+++ b/term.c
@@ -1,4 +1,4 @@
-/*     $Id: term.c,v 1.192 2011/05/17 11:55:08 kristaps Exp $ */
+/*     $Id: term.c,v 1.196 2011/05/20 15:44:55 kristaps Exp $ */

 /*
  * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2010, 2011 Ingo Schwarze <schwarze@openbsd.org>
 /*
  * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2010, 2011 Ingo Schwarze <schwarze@openbsd.org>


@@ -36,6 +36,7 @@
 static void             adjbuf(struct termp *p, int);
 static void             bufferc(struct termp *, char);
 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)


@@ -69,18 +70,6 @@ term_end(struct termp *p)
        (*p->end)(p);
 }
 
        (*p->end)(p);
 }
 

-
-struct termp *
-term_alloc(enum termenc enc)
-{
-       struct termp    *p;
-
-       p = mandoc_calloc(1, sizeof(struct termp));
-       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


@@ -415,7 +404,7 @@ term_word(struct termp *p, const char *word)
 {
        const char      *seq, *cp;
        char             c;
 {
        const char      *seq, *cp;
        char             c;

-       int              sz;
+       int              sz, uc;

        size_t           ssz;
        enum mandoc_esc  esc;
 
        size_t           ssz;
        enum mandoc_esc  esc;
 


@@ -450,12 +439,37 @@ term_word(struct termp *p, const char *word)
                if (ESCAPE_ERROR == esc)
                        break;
 
                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_PREDEF):
+                               uc = mchars_res2cp(p->symtab, seq, sz);
+                               if (uc <= 0)
+                                       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):
                switch (esc) {
                case (ESCAPE_UNICODE):

-                       encode(p, "?", 1);
+                       encode1(p, '?');

                        break;
                case (ESCAPE_NUMBERED):
                        break;
                case (ESCAPE_NUMBERED):

-                       if ('\0' != (c = mchars_num2char(seq, sz)))
+                       c = mchars_num2char(seq, sz);
+                       if ('\0' != c)

                                encode(p, &c, 1);
                        break;
                case (ESCAPE_PREDEF):
                                encode(p, &c, 1);
                        break;
                case (ESCAPE_PREDEF):


@@ -476,6 +490,8 @@ term_word(struct termp *p, const char *word)
                case (ESCAPE_FONTITALIC):
                        term_fontrepl(p, TERMFONT_UNDER);
                        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_FONTROMAN):
                        term_fontrepl(p, TERMFONT_NONE);
                        break;


@@ -515,6 +531,33 @@ bufferc(struct termp *p, char c)
        p->buf[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)
 {
 static void
 encode(struct termp *p, const char *word, size_t sz)
 {


@@ -573,6 +616,7 @@ term_strlen(const struct termp *p, const char *cp)
        size_t           sz, rsz, i;
        int              ssz, c;
        const char      *seq, *rhs;
        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' };
 
        /*
        static const char rej[] = { '\\', ASCII_HYPH, ASCII_NBRSP, '\0' };
 
        /*


@@ -591,16 +635,45 @@ term_strlen(const struct termp *p, const char *cp)
                switch (*cp) {
                case ('\\'):
                        cp++;
                switch (*cp) {
                case ('\\'):
                        cp++;

-                       rhs = NULL;
-                       switch (mandoc_escape(&cp, &seq, &ssz)) {
-                       case (ESCAPE_ERROR):
+                       esc = mandoc_escape(&cp, &seq, &ssz);
+                       if (ESCAPE_ERROR == esc)

                                return(sz);
                                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_PREDEF):
+                                       c = mchars_res2cp
+                                               (p->symtab, seq, ssz);
+                                       if (c <= 0)
+                                               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):
                        case (ESCAPE_UNICODE):

-                               c = '?';
-                               /* FALLTHROUGH */
+                               sz += (*p->width)(p, '?');
+                               break;

                        case (ESCAPE_NUMBERED):
                        case (ESCAPE_NUMBERED):

-                               if ('\0' != c)
-                                       c = mchars_num2char(seq, ssz);
+                               c = mchars_num2char(seq, ssz);

                                if ('\0' != c)
                                        sz += (*p->width)(p, c);
                                break;
                                if ('\0' != c)
                                        sz += (*p->width)(p, c);
                                break;