fix(xtream): pair brackets and strip the new tag forms when matching

Greptile: the bracket prefix chose its opening and closing delimiter independently, so a malformed "[EN)" was read as a language tag. Codex: recognizing a prefix is only half the job — normalizeTitleKeys has to strip the same tag, or the tagged copy never matches the bare one and multi-source cannot offer the film at all. Its leading-tag rule now shares the pipe-lookalike set and, on the pipe branch only, takes the same Latin+Cyrillic any-case alphabet with no required trailing space. Dash and colon keep their uppercase-Latin spaced form: those are ordinary punctuation, and loosening them would amputate "ОНО: Часть 2" the way a case-insensitive rule amputates "It: Chapter Two". Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-10-16 22:26:16 -08:00 · 2026-08-12 00:58:46 +02:00
1 parent c5349e728a
commit 2ddd38facc
@@ -100,12 +100,8 @@ describe('provider tag stripping', () => {
    it('strips long and compound leading tags', () => {
        expect(normalizeTitle('EXYU| Fallout')).toBe('fallout');
        expect(normalizeTitle('MULTI| Breaking Bad')).toBe('breaking bad');
-        expect(normalizeTitle('4K-DE - The Pitt (2025) (US)')).toBe(
-            'the pitt'
-        );
-        expect(normalizeTitle('AR-SUBS - Fallout (2024) (US)')).toBe(
-            'fallout'
-        );
+        expect(normalizeTitle('4K-DE - The Pitt (2025) (US)')).toBe('the pitt');
+        expect(normalizeTitle('AR-SUBS - Fallout (2024) (US)')).toBe('fallout');
        expect(normalizeTitle('4K-OSN+ - The Last of Us (2023)')).toBe(
            'the last of us'
        );
@@ -117,6 +113,26 @@ describe('provider tag stripping', () => {
        );
    });

+    it('strips Cyrillic and lowercase tags before a pipe', () => {
+        // A pipe is a strong tag signal, so its branch takes the wider
+        // alphabet and needs no trailing space: without this the tagged copy
+        // and the bare one never match, and multi-source cannot offer the
+        // film at all.
+        expect(normalizeTitle('РУС | Дюна')).toBe('дюна');
+        expect(normalizeTitle('укр│ Дюна')).toBe('дюна');
+        expect(normalizeTitle('ru| Dune')).toBe('dune');
+        expect(normalizeTitle('EN|Dune')).toBe('dune');
+        expect(normalizeTitle('|РУС| Дюна')).toBe('дюна');
+    });
+
+    it('keeps that widening away from dash and colon', () => {
+        // Where the separator is ordinary punctuation, an uppercase-Latin
+        // restriction is the only thing standing between a tag and a real
+        // title — "ОНО: Часть 2" is the Cyrillic "It: Chapter Two".
+        expect(normalizeTitle('ОНО: Часть 2')).toBe('оно часть 2');
+        expect(normalizeTitle('ОНО - Часть 2')).toBe('оно часть 2');
+    });
+
    it('keeps bare 4-5 char words before a spaced dash (real titles)', () => {
        expect(normalizeTitle('DUNE - Part Two')).toBe('dune part two');
        expect(normalizeTitle('ALIEN - Covenant')).toBe('alien covenant');
@@ -170,40 +186,89 @@ describe('provider tag stripping', () => {
    });

    const pittCorpus = [
-        'The Pitt (2025)_sub', 'The Pitt (2025)-it', 'The Pitt (2025)',
-        'The Pitt (Hindi)', 'The Pitt (2025) 4K', 'The Pitt (2025) DE',
-        'The Pitt (2025) ES', 'The Pitt (2025) FR', 'The Pitt (2025)_eng',
-        'The Pitt [MULTI-SUB]', 'The Pitt (2025) (4K DV)', 'GR - The Pitt',
-        '4K-DE - The Pitt (2025) (US)', '4K-TR - The Pitt (2025) (US)',
-        'AR-SUBS - The Pitt (2025) (US)', 'DE - The Pitt (2025) (US)',
-        'ALB| The Pitt', 'EXYU| The Pitt', '|ALB| The Pitt', '|DE| The Pitt',
+        'The Pitt (2025)_sub',
+        'The Pitt (2025)-it',
+        'The Pitt (2025)',
+        'The Pitt (Hindi)',
+        'The Pitt (2025) 4K',
+        'The Pitt (2025) DE',
+        'The Pitt (2025) ES',
+        'The Pitt (2025) FR',
+        'The Pitt (2025)_eng',
+        'The Pitt [MULTI-SUB]',
+        'The Pitt (2025) (4K DV)',
+        'GR - The Pitt',
+        '4K-DE - The Pitt (2025) (US)',
+        '4K-TR - The Pitt (2025) (US)',
+        'AR-SUBS - The Pitt (2025) (US)',
+        'DE - The Pitt (2025) (US)',
+        'ALB| The Pitt',
+        'EXYU| The Pitt',
+        '|ALB| The Pitt',
+        '|DE| The Pitt',
    ];

    const falloutCorpus = [
-        'Fallout', 'DE - Fallout (2024)', 'Fallout (2024) - 4K',
-        'Fallout (2024) FR-EN', 'Fallout (2024) Multi', 'Fallout (2024)_fr',
-        'Fallout_esp', 'Fallout (4K)', '4K-AMZ - Fallout (2024) (US)',
-        'AL - Fallout (2024)', 'AMZ - Fallout (2024) (US)',
-        'AR-DE - Fallout (US)', 'LA - Fallout', 'EN| Fallout - 4K',
-        'MULTI| Fallout - 4K', 'Fallout ( مدبلج )', 'Fallout (Telugu)',
-        '|EN| Fallout - 4K', '|MULTI| Fallout', '|TR| Fallout',
+        'Fallout',
+        'DE - Fallout (2024)',
+        'Fallout (2024) - 4K',
+        'Fallout (2024) FR-EN',
+        'Fallout (2024) Multi',
+        'Fallout (2024)_fr',
+        'Fallout_esp',
+        'Fallout (4K)',
+        '4K-AMZ - Fallout (2024) (US)',
+        'AL - Fallout (2024)',
+        'AMZ - Fallout (2024) (US)',
+        'AR-DE - Fallout (US)',
+        'LA - Fallout',
+        'EN| Fallout - 4K',
+        'MULTI| Fallout - 4K',
+        'Fallout ( مدبلج )',
+        'Fallout (Telugu)',
+        '|EN| Fallout - 4K',
+        '|MULTI| Fallout',
+        '|TR| Fallout',
    ];

    const lastOfUsCorpus = [
-        'The Last of Us', 'The Last Of Us', 'The Last of Us (2023) 4K',
-        'The Last of Us (2023) AF', 'The Last of Us_tr',
-        'The Last of Us--esp', 'The Last of Us-DE', 'The Last of Us-esp',
-        'The Last of Us [L]', 'The Last of Us ( HD )',
-        '4K-OSN+ - The Last of Us (2023)', 'IS - The Last of Us (2023) (US)',
-        'RU - The Last of Us', 'ALB| The Last of Us',
+        'The Last of Us',
+        'The Last Of Us',
+        'The Last of Us (2023) 4K',
+        'The Last of Us (2023) AF',
+        'The Last of Us_tr',
+        'The Last of Us--esp',
+        'The Last of Us-DE',
+        'The Last of Us-esp',
+        'The Last of Us [L]',
+        'The Last of Us ( HD )',
+        '4K-OSN+ - The Last of Us (2023)',
+        'IS - The Last of Us (2023) (US)',
+        'RU - The Last of Us',
+        'ALB| The Last of Us',
    ];

    const breakingBadCorpus = [
-        'Breaking Bad', 'Breaking Bad (2008)_fr', 'Breaking Bad (US)_msub',
-        'Breaking Bad_it', 'Breaking Bad-DE', 'Breaking Bad-eng',
-        'Breaking Bad ( عائلي )', 'Breaking Bad (Pure)',
-        'Breaking Bad - Multi', 'Breaking Bad ES', 'AR-DE - Breaking Bad',
-        'EN| Breaking Bad SUB', 'MULTI| Breaking Bad', 'AR| Breaking Bad',
+        'Breaking Bad',
+        'Breaking Bad (2008)_fr',
+        'Breaking Bad (US)_msub',
+        'Breaking Bad_it',
+        'Breaking Bad-DE',
+        'Breaking Bad-eng',
+        'Breaking Bad ( عائلي )',
+        'Breaking Bad (Pure)',
+        'Breaking Bad - Multi',
+        'Breaking Bad ES',
+        'AR-DE - Breaking Bad',
+        'EN| Breaking Bad SUB',
+        'MULTI| Breaking Bad',
+        'AR| Breaking Bad',
+        // Pipe lookalikes and a lowercase tag: identical on screen to the
+        // forms above, so they have to reach the same key.
+        'EN │ Breaking Bad',
+        'DE ¦ Breaking Bad',
+        'FR｜Breaking Bad',
+        'en| Breaking Bad',
    ];

    it.each([
@@ -30,11 +30,38 @@ const QUALITY_TAGS = new Set([
 ]);

 /**
- * Wrapped tag at the very start of a provider title: "|DE| ARD",
- * "|MULTI| Fallout". The lookahead requires a letter in the tag so a
- * numeric fragment can never be treated as one.
+ * The pipe and the display lookalikes providers use interchangeably with it
+ * (`¦`, `│`, fullwidth `｜`, …). They are visually identical to `|` in a
+ * catalog, so a rule that reads only U+007C leaves the same tag stripped in
+ * one playlist and welded to the title in another — and the two copies then
+ * never match as the same film.
+ *
+ * Exported because `vod-source-language.util.ts` reads the same separator to
+ * decide a row's language: one set, so the "is this a tag" answer cannot
+ * differ between matching and display.
 */
-const WRAPPED_TAG_PREFIX = /^\s*\|(?=[0-9+]*[A-Z])[A-Z0-9+]{2,5}\|\s*/;
+export const PROVIDER_PIPE_CLASS = '[|¦│┃❘∣⏐⎪︱︳丨｜]';
+
+/**
+ * The alphabet a pipe-delimited tag may use: Latin or Cyrillic, either case
+ * — the same set the language-prefix reader accepts, so the two files agree
+ * on what a tag looks like. Every segment must still contain a letter, so a
+ * numeric fragment is never read as one.
+ *
+ * Wider than the uppercase-Latin `SEG` below because it is only ever used
+ * where a pipe is the separator. Nothing but a tag precedes a pipe: it is
+ * not valid in a Windows filename and does not occur in real titles.
+ */
+const PIPE_LETTER = 'A-Za-zА-Яа-яЁё';
+const PIPE_SEG = `(?=[0-9+]*[${PIPE_LETTER}])[${PIPE_LETTER}0-9+]`;
+
+/**
+ * Wrapped tag at the very start of a provider title: "|DE| ARD",
+ * "|MULTI| Fallout", "|РУС| Дюна".
+ */
+const WRAPPED_TAG_PREFIX = new RegExp(
+    `^\\s*${PROVIDER_PIPE_CLASS}${PIPE_SEG}{2,5}${PROVIDER_PIPE_CLASS}\\s*`
+);

 /**
 * Leading channel/language prefix like "EN - ", "DE| ", "FR: ", including
@@ -50,13 +77,24 @@ const WRAPPED_TAG_PREFIX = /^\s*\|(?=[0-9+]*[A-Z])[A-Z0-9+]{2,5}\|\s*/;
 *   - pipe ("EXYU| "): compound OR 2–5 chars — a pipe is a strong tag signal
 *   - colon ("EN: "): 2–3 chars — longer acronyms are franchise titles
 *     ("NCIS: LA")
+ *
+ * The pipe branch takes the wider `PIPE_SEG` alphabet (Cyrillic as well as
+ * Latin, either case) and does not require a space after the separator.
+ * "РУС | Фильм", "ru| Movie" and "EN|Movie" otherwise keep their tag welded
+ * to the title, so the tagged copy never matches the bare one and
+ * multi-source cannot offer the film at all.
+ *
+ * Neither widening is extended to dash and colon: those are ordinary title
+ * punctuation, and the uppercase-Latin restriction plus the required space
+ * are what keep "ОНО: Часть 2" and "Spider-Man" intact — the Cyrillic and
+ * hyphenated cases of the "It: Chapter Two" hazard.
 */
 const SEG = '(?=[0-9+]*[A-Z])[A-Z0-9+]';
 const COMPOUND_TAG = `${SEG}{2,5}(?:-${SEG}{2,6}){1,2}`;
 const LANGUAGE_PREFIX = new RegExp(
    '^(?:' +
        `(?:${COMPOUND_TAG}|${SEG}{2,3})\\s*-\\s+` +
-        `|(?:${COMPOUND_TAG}|${SEG}{2,5})\\s*\\|\\s+` +
+        `|(?:${COMPOUND_TAG}|${PIPE_SEG}{2,5})\\s*${PROVIDER_PIPE_CLASS}\\s*` +
        `|${SEG}{2,3}\\s*:\\s+` +
        ')'
 );
@@ -69,10 +107,48 @@ const LANGUAGE_PREFIX = new RegExp(
 * suffixes ("NCIS: LA"). US/USA/UK/LA are deliberately absent.
 */
 const TRAILING_TAG_VOCABULARY = new Set([
-    'AF', 'AL', 'ALB', 'AR', 'BY', 'DE', 'DUB', 'EN', 'ENG', 'ES', 'ESP',
-    'EXYU', 'FR', 'FRA', 'GE', 'GR', 'HU', 'IN', 'IR', 'IS', 'IT', 'ITA',
-    'KA', 'KU', 'LAT', 'ML', 'MSUB', 'MULTI', 'NL', 'PL', 'PT', 'RO', 'RU',
-    'SC', 'SE', 'SUB', 'SUBS', 'SW', 'TA', 'TL', 'TR', 'TUR',
+    'AF',
+    'AL',
+    'ALB',
+    'AR',
+    'BY',
+    'DE',
+    'DUB',
+    'EN',
+    'ENG',
+    'ES',
+    'ESP',
+    'EXYU',
+    'FR',
+    'FRA',
+    'GE',
+    'GR',
+    'HU',
+    'IN',
+    'IR',
+    'IS',
+    'IT',
+    'ITA',
+    'KA',
+    'KU',
+    'LAT',
+    'ML',
+    'MSUB',
+    'MULTI',
+    'NL',
+    'PL',
+    'PT',
+    'RO',
+    'RU',
+    'SC',
+    'SE',
+    'SUB',
+    'SUBS',
+    'SW',
+    'TA',
+    'TL',
+    'TR',
+    'TUR',
 ]);

 /**
@@ -30,6 +30,12 @@ describe('titleLanguagePrefix', () => {
        expect(titleLanguagePrefix('Movie [EN]')).toBeNull();
    });

+    it('requires the brackets to pair', () => {
+        // Openers and closers matched independently would accept these.
+        expect(titleLanguagePrefix('[EN) Movie')).toBeNull();
+        expect(titleLanguagePrefix('(DE] Film')).toBeNull();
+    });
+
    it('reads an uppercase tag before a spaced dash, and only that form', () => {
        expect(titleLanguagePrefix('EN - Movie')).toBe('EN');
        expect(titleLanguagePrefix('РУС - Фильм')).toBe('РУС');
@@ -1,3 +1,5 @@
+import { PROVIDER_PIPE_CLASS } from './title-normalization.util';
+
 /**
 * Language prefixes for VOD multi-source rows.
 *
@@ -26,11 +28,11 @@
 */

 /**
- * The pipe and its display lookalikes. Panels routinely decorate with
- * `¦`, `│` or fullwidth bars that are visually identical to `|` but fail a
- * literal match — the user sees "EN | Movie" either way.
+ * The pipe and its display lookalikes, shared with title normalization so
+ * that what counts as a tag separator here is exactly what counts as one
+ * when two copies are matched as the same film.
 */
-const PIPE = '[|¦│┃❘∣⏐⎪︱︳丨｜]';
+const PIPE = PROVIDER_PIPE_CLASS;

 /**
 * A candidate language token: 2–4 letters of ONE script, or the `MULTI`
@@ -43,8 +45,16 @@ const TOKEN = '(?:[A-Za-z]{2,4}|[А-Яа-яЁё]{2,4}|[Mm][Uu][Ll][Tt][Ii])';
 /** `EN| Movie`, `ru │ Фильм`, `MULTI ¦ Movie` — any case before a pipe. */
 const PIPE_FORM = new RegExp(`^\\s*(${TOKEN})\\s*${PIPE}`);

-/** `[EN] Movie`, `(RU) Фильм` — a bracketed tag at the very start. */
-const BRACKET_FORM = new RegExp(`^\\s*[[(]\\s*(${TOKEN})\\s*[\\])]`);
+/**
+ * `[EN] Movie`, `(RU) Фильм` — a bracketed tag at the very start.
+ *
+ * The two bracket styles are separate alternatives rather than one class of
+ * openers and one of closers: the latter pairs them independently, so a
+ * malformed `[EN)` would be read as a tag.
+ */
+const BRACKET_FORM = new RegExp(
+    `^\\s*(?:\\[\\s*(${TOKEN})\\s*\\]|\\(\\s*(${TOKEN})\\s*\\))`
+);

 /**
 * `EN - Movie` — dash-separated, and deliberately stricter than the pipe
@@ -73,15 +83,24 @@ export function titleLanguagePrefix(
 ): string | null {
    const title = rawTitle ?? '';

-    const pipe = PIPE_FORM.exec(title);
+    const pipe = capturedTag(PIPE_FORM.exec(title));
    if (pipe) {
-        return pipe[1].toUpperCase();
+        return pipe.toUpperCase();
    }

-    const gated = BRACKET_FORM.exec(title) ?? DASH_FORM.exec(title);
-    return gated && isKnownLanguageTag(gated[1])
-        ? gated[1].toUpperCase()
-        : null;
+    const gated =
+        capturedTag(BRACKET_FORM.exec(title)) ??
+        capturedTag(DASH_FORM.exec(title));
+    return gated && isKnownLanguageTag(gated) ? gated.toUpperCase() : null;
+}
+
+/**
+ * The tag out of whichever alternative matched. Read positionally rather
+ * than as group 1, because a form with several alternatives (brackets) has
+ * one group per alternative and only one of them is filled.
+ */
+function capturedTag(match: RegExpExecArray | null): string | null {
+    return match?.slice(1).find((group) => group !== undefined) ?? null;
 }

 /**