feat: switch text to UTF-8; add Unicode key events, dead keys & Czech layout; fix numpad navigation; add NumLock support

programs/src/wiki/main.cpp

@@ -10,6 +10,7 @@
 
 #include <montauk/syscall.h>
 #include <montauk/string.h>
+#include <montauk/utf8.h>
 #include <http/http.hpp>
 
 extern "C" {
@@ -177,12 +178,23 @@ static int extract_json_string(const char* buf, int len, const char* key,
                             else if (h >= 'A' && h <= 'F') val |= h - 'A' + 10;
                         }
                         p += 4;
-                        if (val < 128) out[j++] = (char)val;
-                        else if (val == 0x2013 || val == 0x2014) out[j++] = '-';
-                        else if (val == 0x2018 || val == 0x2019) out[j++] = '\'';
-                        else if (val == 0x201C || val == 0x201D) out[j++] = '"';
-                        else if (val == 0x2026) { out[j++] = '.'; out[j++] = '.'; out[j++] = '.'; }
-                        else out[j++] = '?';
+                        // The terminal renders UTF-8; join surrogate pairs
+                        if (val >= 0xD800 && val <= 0xDBFF && p + 6 < len &&
+                            buf[p + 1] == '\\' && buf[p + 2] == 'u') {
+                            unsigned lo = 0;
+                            for (int k = 3; k <= 6; k++) {
+                                char h = buf[p + k];
+                                lo <<= 4;
+                                if (h >= '0' && h <= '9') lo |= h - '0';
+                                else if (h >= 'a' && h <= 'f') lo |= h - 'a' + 10;
+                                else if (h >= 'A' && h <= 'F') lo |= h - 'A' + 10;
+                            }
+                            if (lo >= 0xDC00 && lo <= 0xDFFF) {
+                                val = 0x10000 + ((val - 0xD800) << 10) + (lo - 0xDC00);
+                                p += 6;
+                            }
+                        }
+                        j += montauk::utf8::encode(val, out + j);
                     }
                     break;
                 }
@@ -219,9 +231,12 @@ static int wrap_paragraph(const char* text, int textLen, int cols,
         const char* lastSpace = nullptr;
         int col = 0;
 
+        // One terminal column per UTF-8 character
         while (p < end && col < cols) {
             if (*p == ' ') lastSpace = p;
-            p++;
+            int n;
+            montauk::utf8::decode(p, (int)(end - p), &n);
+            p += n;
             col++;
         }
 
@@ -376,8 +391,7 @@ static void render_pager(WikiLine* lines, int totalLines, int scroll,
         int maxW = cols;
         if (ln.type == LINE_SECTION && ln.level > 2)
             maxW -= (ln.level - 2) * 2;
-        int printLen = ln.len;
-        if (printLen > maxW) printLen = maxW;
+        int printLen = montauk::utf8::offset_of(ln.text, ln.len, maxW < 0 ? 0 : maxW);
         for (int c = 0; c < printLen; c++)
             sb_putc(ln.text[c]);
 
@@ -463,7 +477,7 @@ static void run_pager(WikiLine* lines, int totalLines, const char* title,
         case 'g': scroll = 0; break;
         case 'G': scroll = maxScroll; break;
         default:
-            switch (ev.scancode) {
+            switch (montauk::nav_scancode(ev)) {
             case 0x48: if (scroll > 0) scroll--; break;           // Up
             case 0x50: if (scroll < maxScroll) scroll++; break;   // Down
             case 0x49: scroll -= contentRows; if (scroll < 0) scroll = 0; break; // PgUp