fix(playback): require a meaningful Cyrillic letter share before choosing CP1251

Codex review round 6 on #1471: an isolated accented CP1252 word decodes
under CP1251 to a tiny pure-Cyrillic word ("À table" -> "А table") whose
single vote flipped the file to Cyrillic. The chooser now also requires
Cyrillic letters to carry a meaningful share of all letters — genuinely
Cyrillic dialogue dominates its own letter count even with embedded Latin
names, while isolated accents never do.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

docs/architecture/player-controls-contract.md

@@ -695,9 +695,11 @@ Per-engine implementations:
   BOMs, strict UTF-8, then `chooseLegacySingleByteDecode`, which picks
   between Windows-1251 and Windows-1252 by the plausibility of the 1251
   candidate's decoded words — pure-Cyrillic words vote for 1251, words
-  mixing Cyrillic with ASCII letters vote against, since misread Latin text
-  like "était" decodes to the mixed-script "йtait" that real subtitles never
-  contain), because `Blob.text()`'s silent UTF-8 substitution turns common
+  mixing Cyrillic with ASCII letters vote against (misread Latin text like
+  "était" decodes to the mixed-script "йtait" that real subtitles never
+  contain), and Cyrillic must also carry a meaningful share of all letters
+  so an isolated accented CP1252 word ("À table" → "А table") cannot flip
+  the file), because `Blob.text()`'s silent UTF-8 substitution turns common
   legacy-encoded SRT files into mojibake. `WebVideoExternalSubtitles` parses
   the file (`external-subtitle-cues.util.ts`) and renders it through a native
   `TextTrack` on the video element, so it works under every source kind. The

libs/ui/playback/src/lib/web-video-support/external-subtitle-cues.util.spec.ts

@@ -76,6 +76,21 @@ describe('external-subtitle-cues.util', () => {
             expect(decoded).toContain('À table');
         });
 
+        it('keeps an isolated accented CP1252 word Latin ("À table")', () => {
+            // À=0xC0 decodes under CP1251 to the pure-Cyrillic one-letter
+            // word "А"; without the letter-share guard that single vote
+            // flips the whole file to Cyrillic.
+            const cp1252 = [
+                0xc0, // À
+                ...Array.from(
+                    ' table !\n1\n00:00:01,000 --> 00:00:02,000\n'
+                ).map((c) => c.charCodeAt(0)),
+            ];
+            expect(decodeExternalSubtitleBytes(toBuffer(cp1252))).toContain(
+                'À table'
+            );
+        });
+
         it('decodes mostly-ASCII Windows-1252 bytes (sparse accents)', () => {
             // "resume: cafe" with two accented letters among ASCII.
             const cp1252 = [

libs/ui/playback/src/lib/web-video-support/external-subtitle-cues.util.ts

@@ -82,14 +82,30 @@ export function decodeExternalSubtitleBytes(buffer: ArrayBuffer): string {
  */
 function chooseLegacySingleByteDecode(buffer: ArrayBuffer): string {
     const decoded1251 = new TextDecoder('windows-1251').decode(buffer);
-    let score = 0;
+    let cyrillicLetters = 0;
+    let mixedLetters = 0;
+    let asciiLetters = 0;
     for (const word of decoded1251.split(/[^\p{L}]+/u)) {
-        if (!/[\u0400-\u04ff]/.test(word)) {
+        if (!word) {
             continue;
         }
-        score += /[A-Za-z]/.test(word) ? -word.length : word.length;
+        if (!/[\u0400-\u04ff]/.test(word)) {
+            asciiLetters += word.length;
+        } else if (/[A-Za-z]/.test(word)) {
+            mixedLetters += word.length;
+        } else {
+            cyrillicLetters += word.length;
+        }
     }
-    return score > 0
+    // Two guards: mixed-script words are strong evidence of misread Latin
+    // text, and isolated accented CP1252 words ("\u00c0 table" \u2192 "\u0410 table")
+    // masquerade as tiny pure-Cyrillic words \u2014 so Cyrillic must also carry a
+    // meaningful share of all letters before 1251 wins. Genuinely Cyrillic
+    // dialogue dominates its own letter count even with embedded Latin names.
+    const plausiblyCyrillic =
+        cyrillicLetters > mixedLetters * 2 &&
+        cyrillicLetters * 4 > asciiLetters;
+    return plausiblyCyrillic
         ? decoded1251
         : new TextDecoder('windows-1252').decode(buffer);
 }