url-mode: handle wide chars and grapheme clusters when auto-detecting URLs

* Skip spacer cells. This fixes an issue where characters following a
  double-width character weren't detect properly.

* Unpack grapheme clusters (i.e. cells with multiple codepoints), and
  iterate all their codepoints.

Closes #1465

CHANGELOG.md

@@ -76,9 +76,12 @@
   ([#1436][1436]).
 * One frame being rendered at the wrong scale after being hidden by
   another opaque, maximized window ([#1464][1464]).
+* Double-width characters, and grapheme clusters breaking URL
+  auto-detection ([#1465][1465]).
 
 [1436]: https://codeberg.org/dnkl/foot/issues/1436
 [1464]: https://codeberg.org/dnkl/foot/issues/1464
+[1465]: https://codeberg.org/dnkl/foot/issues/1465
 
 
 ### Security

url-mode.c

@@ -326,7 +326,25 @@ auto_detected(const struct terminal *term, enum url_action action,
 
         for (int c = 0; c < term->cols; c++) {
             const struct cell *cell = &row->cells[c];
-            char32_t wc = cell->wc;
+
+            if (cell->wc >= CELL_SPACER)
+                continue;
+
+            const char32_t *wcs = NULL;
+            size_t wc_count = 0;
+
+            if (cell->wc >= CELL_COMB_CHARS_LO && cell->wc <= CELL_COMB_CHARS_HI) {
+                struct composed *composed =
+                    composed_lookup(term->composed, cell->wc - CELL_COMB_CHARS_LO);
+                wcs = composed->chars;
+                wc_count = composed->count;
+            } else {
+                wcs = &cell->wc;
+                wc_count = 1;
+            }
+
+            for (size_t w_idx = 0; w_idx < wc_count; w_idx++) {
+                char32_t wc = wcs[w_idx];
 
                 switch (state) {
                 case STATE_PROTOCOL:
@@ -348,9 +366,11 @@ auto_detected(const struct terminal *term, enum url_action action,
                     if (proto_char_count < prot_len)
                       continue;
 
-                    const char32_t *proto = &proto_chars[max_prot_len - prot_len];
+                    const char32_t *proto =
+                        &proto_chars[max_prot_len - prot_len];
 
-                    if (c32ncasecmp(conf->url.protocols[i], proto, prot_len) == 0) {
+                    if (c32ncasecmp(conf->url.protocols[i], proto, prot_len) ==
+                        0) {
                       state = STATE_URL;
                       start = proto_start[max_prot_len - prot_len];
 
@@ -364,12 +384,9 @@ auto_detected(const struct terminal *term, enum url_action action,
                   break;
 
                 case STATE_URL: {
-                const char32_t *match = bsearch(
+                  const char32_t *match =
-                    &wc,
+                      bsearch(&wc, uri_characters, uri_characters_count,
-                    uri_characters,
+                              sizeof(uri_characters[0]), &c32cmp_single);
-                    uri_characters_count,
-                    sizeof(uri_characters[0]),
-                    &c32cmp_single);
 
                   bool emit_url = false;
 
@@ -449,8 +466,15 @@ auto_detected(const struct terminal *term, enum url_action action,
                     bool done = false;
                     do {
                       switch (url[len - 1]) {
-                        case U'.': case U',': case U':': case U';': case U'?':
+                      case U'.':
-                        case U'!': case U'"': case U'\'': case U'%':
+                      case U',':
+                      case U':':
+                      case U';':
+                      case U'?':
+                      case U'!':
+                      case U'"':
+                      case U'\'':
+                      case U'%':
                         len--;
                         end.col--;
                         if (end.col < 0) {
@@ -474,10 +498,10 @@ auto_detected(const struct terminal *term, enum url_action action,
                     if (url_utf8 != NULL) {
                       tll_push_back(
                           *urls,
-                            ((struct url){
+                          ((struct url){.id = (uint64_t)rand() << 32 | rand(),
-                                .id = (uint64_t)rand() << 32 | rand(),
                                         .url = url_utf8,
-                                .range = {
+                                        .range =
+                                            {
                                                 .start = start,
                                                 .end = end,
                                             },
@@ -494,6 +518,7 @@ auto_detected(const struct terminal *term, enum url_action action,
                 }
             }
         }
+    }
 }
 
 static void