codepage support

Support for encoding 8-bit text into Unicode strings exists in ScreenAdapter but is missing in DummyScreenAdapter.

- Moved and rewrote the CP437-related code in src/browser/screen.js to src/lib.js.
- Added support for codepages other than CP437.
- Made the active codepage configurable in the V86 constructor in options.screen.encoding.

src/lib.js
- Added public function get_charmap(encoding)
  Returns the charmap of type Array<number> for given encoding string.
  Supported encodings: "cp437", "cp858" and "ascii", more can be added easily from
  https://github.com/chschnell/v86-i18n/blob/main/codepage-tables/codepage_tables.js
- Added public function to_unicode(text_8bit, charmap)
  Returns the Unicode string representation of given 8-bit text and charmap.
  Supported types for text_8bit are Array<number>, Uint8Array and number.

src/browser/dummy_screen.js
- Added options argument to DummyScreenAdapter constructor (same as for ScreenAdapter).
- Uses options.encoding argument for user-defined encoding, defaults to "cp437".
- DummyScreenAdapter.get_text_row() now uses to_unicode() for string encoding.

src/browser/screen.js
- Uses options.encoding argument for user-defined encoding, defaults to "cp437".
- DummyScreenAdapter.get_text_row() now uses to_unicode() for string encoding.
- Removed CP437 table, no longer needed.

src/browser/starter.js
- Added options argument to DummyScreenAdapter constructor call.
- Removed unneeded assignment to settings.screen_options, this gets overwritten
  a couple of dozen lines below.

src/browser/dummy_screen.js

@@ -1,9 +1,11 @@
 import { dbg_assert } from "../log.js";
+import { get_charmap, to_unicode } from "../lib.js";
 
 /**
  * @constructor
+ * @param {Object=} options
  */
-export function DummyScreenAdapter()
+export function DummyScreenAdapter(options)
 {
     var
         graphic_image_data,
@@ -30,7 +32,11 @@ export function DummyScreenAdapter()
         text_mode_width = 0,
 
         // number of rows
-        text_mode_height = 0;
+        text_mode_height = 0,
+
+        // 8-bit-text to Unicode character map
+        /** @type {!Array<number>} */
+        charmap = get_charmap(options?.encoding || "cp437");
 
     this.put_char = function(row, col, chr, blinking, bg_color, fg_color)
     {
@@ -115,7 +121,7 @@ export function DummyScreenAdapter()
     this.get_text_row = function(i)
     {
         const offset = i * text_mode_width;
-        return String.fromCharCode.apply(String, text_mode_data.subarray(offset, offset + text_mode_width));
+        return to_unicode(text_mode_data.subarray(offset, offset + text_mode_width), charmap);
     };
 
     this.set_size_text(80, 25);

src/browser/screen.js

@@ -1,4 +1,5 @@
 import { dbg_assert } from "../log.js";
+import { get_charmap, to_unicode } from "../lib.js";
 
 // Draws entire buffer and visualizes the layers that would be drawn
 export const DEBUG_SCREEN_LAYERS = DEBUG && false;
@@ -7,6 +8,7 @@ export const DEBUG_SCREEN_LAYERS = DEBUG && false;
  * Adapter to use visual screen in browsers (in contrast to node)
  * @constructor
  * @param {Object} options
+ * @param {function()} screen_fill_buffer
  */
 export function ScreenAdapter(options, screen_fill_buffer)
 {
@@ -95,9 +97,9 @@ export function ScreenAdapter(options, screen_fill_buffer)
         cursor_end,
         cursor_enabled,
 
-        // 8-bit Unicode character maps
-        charmap_default = [],
-        charmap = charmap_default,
+        // 8-bit-text to Unicode character map
+        /** @type {!Array<number>} */
+        charmap = get_charmap(options.encoding || "cp437"),
 
         // render loop state
         timer_id = 0,
@@ -317,52 +319,6 @@ export function ScreenAdapter(options, screen_fill_buffer)
 
     this.init = function()
     {
-        // map 8-bit DOS codepage 437 character range 0-31 to 16-bit Unicode codepoints
-        const charmap_low = new Uint16Array([
-            0x20,   0x263A, 0x263B, 0x2665, 0x2666, 0x2663, 0x2660, 0x2022,
-            0x25D8, 0x25CB, 0x25D9, 0x2642, 0x2640, 0x266A, 0x266B, 0x263C,
-            0x25BA, 0x25C4, 0x2195, 0x203C, 0xB6,   0xA7,   0x25AC, 0x21A8,
-            0x2191, 0x2193, 0x2192, 0x2190, 0x221F, 0x2194, 0x25B2, 0x25BC
-        ]);
-        // map 8-bit DOS codepage 437 character range 127-255 to 16-bit Unicode codepoints
-        const charmap_high = new Uint16Array([
-            0x2302,
-            0xC7, 0xFC, 0xE9, 0xE2, 0xE4, 0xE0, 0xE5, 0xE7,
-            0xEA, 0xEB, 0xE8, 0xEF, 0xEE, 0xEC, 0xC4, 0xC5,
-            0xC9, 0xE6, 0xC6, 0xF4, 0xF6, 0xF2, 0xFB, 0xF9,
-            0xFF, 0xD6, 0xDC, 0xA2, 0xA3, 0xA5, 0x20A7, 0x192,
-            0xE1, 0xED, 0xF3, 0xFA, 0xF1, 0xD1, 0xAA, 0xBA,
-            0xBF, 0x2310, 0xAC, 0xBD, 0xBC, 0xA1, 0xAB, 0xBB,
-            0x2591, 0x2592, 0x2593, 0x2502, 0x2524, 0x2561, 0x2562, 0x2556,
-            0x2555, 0x2563, 0x2551, 0x2557, 0x255D, 0x255C, 0x255B, 0x2510,
-            0x2514, 0x2534, 0x252C, 0x251C, 0x2500, 0x253C, 0x255E, 0x255F,
-            0x255A, 0x2554, 0x2569, 0x2566, 0x2560, 0x2550, 0x256C, 0x2567,
-            0x2568, 0x2564, 0x2565, 0x2559, 0x2558, 0x2552, 0x2553, 0x256B,
-            0x256A, 0x2518, 0x250C, 0x2588, 0x2584, 0x258C, 0x2590, 0x2580,
-            0x3B1, 0xDF, 0x393, 0x3C0, 0x3A3, 0x3C3, 0xB5, 0x3C4,
-            0x3A6, 0x398, 0x3A9, 0x3B4, 0x221E, 0x3C6, 0x3B5, 0x2229,
-            0x2261, 0xB1, 0x2265, 0x2264, 0x2320, 0x2321, 0xF7,
-            0x2248, 0xB0, 0x2219, 0xB7, 0x221A, 0x207F, 0xB2, 0x25A0, 0xA0
-        ]);
-
-        // initialize 8-bit DOS codepage 437 map charmap[256] (Uint8 -> String[1])
-        for(var i = 0, chr; i < 256; i++)
-        {
-            if(i > 126)
-            {
-                chr = charmap_high[i - 0x7F];
-            }
-            else if(i < 32)
-            {
-                chr = charmap_low[i];
-            }
-            else
-            {
-                chr = i;
-            }
-            charmap_default.push(String.fromCharCode(chr));
-        }
-
         // setup text mode cursor DOM element
         cursor_element.classList.add("cursor");
         cursor_element.style.position = "absolute";
@@ -417,7 +373,7 @@ export function ScreenAdapter(options, screen_fill_buffer)
                     context.fillStyle = number_as_color(bg_color);
                     context.fillRect(x * char_size[0], y * char_size[1], char_size[0], char_size[1]);
                     context.fillStyle = number_as_color(fg_color);
-                    context.fillText(charmap[character], x * char_size[0], y * char_size[1]);
+                    context.fillText(to_unicode(character, charmap), x * char_size[0], y * char_size[1]);
                 }
             }
 
@@ -716,11 +672,6 @@ export function ScreenAdapter(options, screen_fill_buffer)
         update_scale_graphic();
     };
 
-    this.set_charmap = function(text_charmap)
-    {
-        charmap = text_charmap || charmap_default;
-    };
-
     this.set_scale = function(s_x, s_y)
     {
         scale_x = s_x;
@@ -858,8 +809,7 @@ export function ScreenAdapter(options, screen_fill_buffer)
 
         var blinking,
             bg_color,
-            fg_color,
-            text;
+            fg_color;
 
         row_element = text_screen.childNodes[row];
         fragment = document.createElement("div");
@@ -880,7 +830,7 @@ export function ScreenAdapter(options, screen_fill_buffer)
             color_element.style.backgroundColor = number_as_color(bg_color);
             color_element.style.color = number_as_color(fg_color);
 
-            text = "";
+            const text = [];
 
             // put characters of the same color in one element
             while(i < text_mode_width &&
@@ -888,10 +838,7 @@ export function ScreenAdapter(options, screen_fill_buffer)
                 text_mode_data[offset + BG_COLOR_INDEX] === bg_color &&
                 text_mode_data[offset + FG_COLOR_INDEX] === fg_color)
             {
-                var ascii = text_mode_data[offset + CHARACTER_INDEX];
-
-                text += charmap[ascii];
-                dbg_assert(charmap[ascii]);
+                text.push(text_mode_data[offset + CHARACTER_INDEX]);
 
                 i++;
                 offset += TEXT_BUF_COMPONENT_SIZE;
@@ -914,7 +861,7 @@ export function ScreenAdapter(options, screen_fill_buffer)
                 }
             }
 
-            color_element.textContent = text;
+            color_element.textContent = to_unicode(text, charmap);
             fragment.appendChild(color_element);
         }
 
@@ -971,16 +918,14 @@ export function ScreenAdapter(options, screen_fill_buffer)
 
     this.get_text_row = function(y)
     {
-        let result = "";
-
-        for(let x = 0; x < text_mode_width; x++)
+        const start = y * text_mode_width * TEXT_BUF_COMPONENT_SIZE + CHARACTER_INDEX;
+        const end = start + text_mode_width * TEXT_BUF_COMPONENT_SIZE;
+        const row = [];
+        for(let i = start; i < end; i += TEXT_BUF_COMPONENT_SIZE)
         {
-            const index = (y * text_mode_width + x) * TEXT_BUF_COMPONENT_SIZE;
-            const character = text_mode_data[index + CHARACTER_INDEX];
-            result += charmap[character];
+            row.push(text_mode_data[i]);
         }
-
-        return result;
+        return to_unicode(row, charmap);
     };
 
     this.init();

src/browser/starter.js

@@ -235,7 +235,6 @@ V86.prototype.continue_init = async function(emulator, options)
     settings.cpuid_level = options.cpuid_level;
     settings.virtio_balloon = options.virtio_balloon;
     settings.virtio_console = options.virtio_console;
-    settings.screen_options = options.screen_options;
 
     const relay_url = options.network_relay_url || options.net_device && options.net_device.relay_url;
     if(relay_url)
@@ -285,7 +284,7 @@ V86.prototype.continue_init = async function(emulator, options)
     }
     else
     {
-        this.screen_adapter = new DummyScreenAdapter();
+        this.screen_adapter = new DummyScreenAdapter(screen_options);
     }
     settings.screen = this.screen_adapter;
     settings.screen_options = screen_options;

src/lib.js

@@ -728,3 +728,60 @@ export function read_sized_string_from_mem(mem, offset, len)
     len >>>= 0;
     return String.fromCharCode(...new Uint8Array(mem.buffer, offset, len));
 }
+
+/**
+ * Unicode mappings of supported 8-bit code pages.
+ *
+ * Supported encodings:
+ * - "cp437": DOS Latin US (default)
+ * - "cp858": ISO 8859-1 (the lower 128 characters are identical to "cp437")
+ * - "ascii": same as "cp437" with lower 32 and upper 128 characters mapped to "."
+ */
+const CP437 = "\u0000☺☻♥♦♣♠•◘○◙♂♀♪♫☼►◄↕‼¶§▬↨↑↓→←∟↔▲▼ !\"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~⌂ÇüéâäàåçêëèïîìÄÅÉæÆôöòûùÿÖÜ¢£¥₧ƒáíóúñѪº¿⌐¬½¼¡«»░▒▓│┤╡╢╖╕╣║╗╝╜╛┐└┴┬├─┼╞╟╚╔╩╦╠═╬╧╨╤╥╙╘╒╓╫╪┘┌█▄▌▐▀αßΓπΣσµτΦΘΩδ∞φε∩≡±≥≤⌠⌡÷≈°∙·√ⁿ²■ ";
+const CP858 = "ÇüéâäàåçêëèïîìÄÅÉæÆôöòûùÿÖÜø£Ø׃áíóúñѪº¿®¬½¼¡«»░▒▓│┤ÁÂÀ©╣║╗╝¢¥┐└┴┬├─┼ãÃ╚╔╩╦╠═╬¤ðÐÊËÈ€ÍÎÏ┘┌█▄¦Ì▀ÓßÔÒõÕµþÞÚÛÙýݯ´­±‗¾¶§÷¸°¨·¹³²■ ";
+
+/**
+ * @type {Object<string, Array<number>>}
+ */
+const CHARMAPS =
+{
+    cp437: CP437.split("").map(ch_str => ch_str.charCodeAt(0)),
+    cp858: CP858.split("").map(ch_str => ch_str.charCodeAt(0))
+};
+
+CHARMAPS.cp858 = CHARMAPS.cp437.slice(0, 128) + CHARMAPS.cp858;
+CHARMAPS.ascii = CHARMAPS.cp437.map((c, i) => i > 31 && i < 128 ? c : 0x2E);
+
+/**
+ * Return charmap for given case-insensitve encoding id.
+ *
+ * @param {!string} encoding
+ * @return {!Array<number>}
+ */
+export function get_charmap(encoding)
+{
+    return CHARMAPS[encoding.toLowerCase()] || CHARMAPS.cp437;
+}
+
+/**
+ * Decode 8-bit encoded text into its Unicode string.
+ *
+ * @param {!Array<number>|!Uint8Array|number} text_8bit
+ * @param {!Array<number>} charmap
+ * @return {!string}
+ */
+export function to_unicode(text_8bit, charmap)
+{
+    if(Array.isArray(text_8bit))
+    {
+        return String.fromCharCode(...text_8bit.map(ch_byte => charmap[ch_byte]));
+    }
+    else if(text_8bit instanceof Uint8Array)
+    {
+        return String.fromCharCode(...new Uint16Array(text_8bit).map(ch_byte => charmap[ch_byte]));
+    }
+    else
+    {
+        return String.fromCharCode(charmap[text_8bit]);
+    }
+}