Re: [Groff] accented characters in MAILTO requests

[Gaius Mulley]

Hi Werner,

> the following doesn't work as expected:
> 
>  .MAILTO address@hidden "J\(:orgen H\(:agg"
>
> I get the following warning message from troff:
>
>  a special character is illegal within \X
>
> Can you fix this?

yes :-) - here is the patch. troff wants to encode such
characters by the tcommand 'tC:o` and of course the device
specific 'special' character tcommand is encoded as:

x X etc

and we cannot nest tcommands -
so I've had to alter troff to encode any named character
as \(name\) within the x X command. This functionality
is turned on from within the devhtml/DESC file which
now contains:

use_charnames_in_special

So, hopefully, these changes will not break any other devices.
I've also added a 'pass_filenames' line to the html DESC file
as this will be used to pass the current filename from troff
to the device driver - at a later date - to fix the image file
name problem. I've altered grohtml to decode the named characters
within the x X tcommand. Are these alterations acceptable?
Hopefully it will allow the homepage to be constructed now :-)

cheers Gaius

diff -aruN groff-cvs/include/font.h groff-html/include/font.h
--- groff-cvs/include/font.h    Sat Jan 15 11:35:20 2000
+++ groff-html/include/font.h   Fri Jan 21 01:38:39 2000
@@ -73,6 +73,8 @@
   static int spare2;
   static int sizescale;
   static int tcommand;
+  static int pass_filenames;
+  static int use_charnames_in_special;
 
   static const char **font_name_table;
   static const char **style_table;
diff -aruN groff-cvs/troff/charinfo.h groff-html/troff/charinfo.h
--- groff-cvs/troff/charinfo.h  Fri May 21 05:50:38 1999
+++ groff-html/troff/charinfo.h Thu Jan 20 09:12:21 2000
@@ -74,6 +74,7 @@
   void set_number(int);
   int get_number();
   int numbered();
+  symbol *get_symbol();
 };
 
 charinfo *get_charinfo(symbol);
@@ -162,4 +163,9 @@
     not_found = 1;
     return 1;
   }
+}
+
+inline symbol *charinfo::get_symbol()
+{
+  return( &nm );
 }
diff -aruN groff-cvs/troff/input.cc groff-html/troff/input.cc
--- groff-cvs/troff/input.cc    Sat Jan 15 11:35:25 2000
+++ groff-html/troff/input.cc   Thu Jan 20 10:00:56 2000
@@ -3927,6 +3927,38 @@
   return new non_interpreted_node(mac);
 }
 
+static void encode_char (macro *mac, char c)
+{
+  if (c == '\0') {
+    if (tok.special() && (font::use_charnames_in_special)) {
+      charinfo *ci=tok.get_char(1);
+      const char *s=ci->get_symbol()->contents();
+
+      if (s[0] != (char)0) {
+       mac->append('\\');
+       mac->append('(');
+       int i=0;
+       while (s[i] != (char)0) {
+         mac->append(s[i]);
+         i++;
+       }
+       mac->append('\\');
+       mac->append(')');
+      }
+    } else {
+      error("%1 is illegal within \\X", tok.description());
+    }
+  } else {
+    if ((font::use_charnames_in_special) && (c == '\\')) {
+      /*
+       * add escape escape sequence
+       */
+      mac->append(c);
+    }
+    mac->append(c);
+  }
+}
+
 node *do_special()
 {
   token start;
@@ -3956,10 +3988,7 @@
       c = '\b';
     else
       c = tok.ch();
-    if (c == '\0')
-      error("%1 is illegal within \\X", tok.description());
-    else 
-      mac.append(c);
+    encode_char(&mac, c);
   }
   return new special_node(mac);
 }
diff -aruN groff-cvs/troff/token.h groff-html/troff/token.h
--- groff-cvs/troff/token.h     Fri May 21 05:50:39 1999
+++ groff-html/troff/token.h    Wed Jan 19 20:50:47 2000
@@ -68,6 +68,7 @@
   int nspaces();               // 1 if space, 2 if double space, 0 otherwise
   int space();                 // is it a space or double space?
   int white_space();           // is the current token space or tab?
+  int special();                // is the current token a special character?
   int newline();               // is the current token a newline?
   int tab();                   // is the current token a tab?
   int leader();
@@ -122,6 +123,11 @@
 inline int token::space()
 { 
   return type == TOKEN_SPACE;
+}
+
+inline int token::special()
+{ 
+  return type == TOKEN_SPECIAL;
 }
 
 inline int token::nspaces()
diff -aruN groff-cvs/libgroff/font.cc groff-html/libgroff/font.cc
--- groff-cvs/libgroff/font.cc  Sat Jan 15 11:35:21 2000
+++ groff-html/libgroff/font.cc Thu Jan 20 10:00:26 2000
@@ -767,6 +767,12 @@
     else if (strcmp("tcommand", p) == 0) {
       tcommand = 1;
     }
+    else if (strcmp("pass_filenames", p) == 0) {
+      pass_filenames = 1;
+    }
+    else if (strcmp("use_charnames_in_special", p) == 0) {
+      use_charnames_in_special = 1;
+    }
     else if (strcmp("family", p) == 0) {
       p = strtok(0, WS);
       if (!p) {
diff -aruN groff-cvs/libgroff/fontfile.cc groff-html/libgroff/fontfile.cc
--- groff-cvs/libgroff/fontfile.cc      Fri May 21 05:50:38 1999
+++ groff-html/libgroff/fontfile.cc     Thu Jan 20 10:00:29 2000
@@ -43,6 +43,8 @@
 int font::spare2 = 0;
 int font::sizescale = 1;
 int font::tcommand = 0;
+int font::pass_filenames = 0;
+int font::use_charnames_in_special = 0;
 const char **font::font_name_table = 0;
 int *font::sizes = 0;
 const char *font::family = 0;
diff -aruN groff-cvs/grohtml/html.cc groff-html/grohtml/html.cc
--- groff-cvs/grohtml/html.cc   Sat Jan 15 11:30:17 2000
+++ groff-html/grohtml/html.cc  Thu Jan 20 10:47:29 2000
@@ -1585,12 +1585,16 @@
 {
   int  index;
 
-  index = f->name_to_index(name);
-  if (index == 0) {
-    error("character `%s' not found", name);
-    return( NULL );
+  if (f == 0) {
+    return( 0 );
   } else {
-    return( (char *)f->get_special_device_encoding(index) );
+    index = f->name_to_index(name);
+    if (index == 0) {
+      error("character `%s' not found", name);
+      return( NULL );
+    } else {
+      return( (char *)f->get_special_device_encoding(index) );
+    }
   }
 }
 
@@ -1598,9 +1602,11 @@
  *  str_translate_to_html - converts a string, str, into html text. It places
  *                          the output input buffer, buf. It truncates string, 
str, if
  *                          there is not enough space in buf.
+ *                          It looks up the html character encoding of single 
characters
+ *                          if, and_single, is TRUE. Characters such as < > & 
etc.
  */
 
-void str_translate_to_html (font *f, char *buf, int buflen, char *str, int len)
+void str_translate_to_html (font *f, char *buf, int buflen, char *str, int 
len, int and_single)
 {
   int         l;
   char       *translation;
@@ -1651,17 +1657,27 @@
        }
       }
     } else {
-      char name[2];
+      if (and_single) {
+       char name[2];
 
-      name[0] = str[i];
-      name[1] = (char)0;
-      translation = get_html_translation(f, name);
-      if (translation) {
-       l = strlen(translation);
-       t = max(0, min(l, buflen-b));
-       strncpy(&buf[b], translation, t);
-       b += t;
+       name[0] = str[i];
+       name[1] = (char)0;
+       translation = get_html_translation(f, name);
+       if (translation) {
+         l = strlen(translation);
+         t = max(0, min(l, buflen-b));
+         strncpy(&buf[b], translation, t);
+         b += t;
+       } else {
+         if (b<buflen) {
+           buf[b] = str[i];
+           b++;
+         }
+       }
       } else {
+       /*
+        *  do not attempt to encode single characters
+        */
        if (b<buflen) {
          buf[b] = str[i];
          b++;
@@ -1721,7 +1737,7 @@
                   ((start_region == -1) || (t->maxv < start_region))) {
          start_title_vpos     = t->minv;
          end_title_hpos       = t->minh;
-         str_translate_to_html(t->text_style.f, buf, MAX_STRING_LENGTH, 
t->text_string, t->text_length);
+         str_translate_to_html(t->text_style.f, buf, MAX_STRING_LENGTH, 
t->text_string, t->text_length, TRUE);
          strcpy((char *)title.text, buf);
          height               = t->text_style.point_size*r/72;
          found_title_start    = TRUE;
@@ -1736,7 +1752,7 @@
            start_title_vpos = min(t->minv, start_title_vpos);
            end_title_hpos   = max(t->maxh, end_title_hpos);
            strcat(title.text, " ");
-           str_translate_to_html(t->text_style.f, buf, MAX_STRING_LENGTH, 
t->text_string, t->text_length);
+           str_translate_to_html(t->text_style.f, buf, MAX_STRING_LENGTH, 
t->text_string, t->text_length, TRUE);
            strcat(title.text, buf);
            page_contents->words.sub_move_right();
            removed_from_head = ((!page_contents->words.is_empty()) &&
@@ -2819,7 +2835,7 @@
 {
   char buf[MAX_STRING_LENGTH];
 
-  str_translate_to_html(f, buf, MAX_STRING_LENGTH, str, len);
+  str_translate_to_html(f, buf, MAX_STRING_LENGTH, str, len, TRUE);
   strncpy(str, buf, max(len, strlen(buf)+1));
 }
 
@@ -2868,7 +2884,7 @@
   do {
     l = g;
     current_vpos = g->minv;
-    str_translate_to_html(g->text_style.f, buf, MAX_STRING_LENGTH, 
g->text_string, g->text_length);
+    str_translate_to_html(g->text_style.f, buf, MAX_STRING_LENGTH, 
g->text_string, g->text_length, TRUE);
     strcat(header.header_buffer, (char *)buf);
     page_contents->words.move_right();
     g = page_contents->words.get_data();
@@ -5814,7 +5830,7 @@
   char buf[MAX_STRING_LENGTH];
 
   str_translate_to_html(g->text_style.f, buf, MAX_STRING_LENGTH,
-                       g->text_string, g->text_length);
+                       g->text_string, g->text_length, TRUE);
   html.put_string(buf);
 }
 
@@ -6369,8 +6385,17 @@
       }
     } else if (strncmp(s, "html:", 5) == 0) {
       int r=font::res;   // resolution of the device
+      char buf[MAX_STRING_LENGTH];
+      font *f=sbuf_style.f;
 
-      page_contents->add_html_command(&sbuf_style, &s[5], strlen(s)-5,
+      if (f == NULL) {
+       int found=FALSE;
+
+       f = font::load_font("TR", &found);
+      }
+      str_translate_to_html(f, buf, MAX_STRING_LENGTH,
+                           &s[5], strlen(s)-5, FALSE);
+      page_contents->add_html_command(&sbuf_style, buf, strlen(buf),
 
       // need to pass rest of string through to html output during flush
 
diff -aruN groff-cvs/devhtml/DESC groff-html/devhtml/DESC
--- groff-cvs/devhtml/DESC      Sun Sep 12 08:29:59 1999
+++ groff-html/devhtml/DESC     Thu Jan 20 09:53:33 2000
@@ -8,3 +8,5 @@
 vert 1
 unitwidth 10
 postpro grohtml
+use_charnames_in_special
+pass_filenames