Complete multi-language localization for news indexes with dynamic content support

[Copilot]

Fix: All 7 PR Review Comments Addressed

🎯 Summary

Fixed all issues identified in PR review #3791772340 with full test validation. All 673 tests passing.

---

✅ Issues Fixed

1. Breadcrumb Localization in Articles (Comment #2799591380)

Problem: BreadcrumbList in article-template.js hardcoded to English ("Home", "News") regardless of article language

Solution: Added comprehensive breadcrumb translations for all 14 languages

Implementation:

const BREADCRUMB_TRANSLATIONS = {
  en: { home: 'Home', news: 'News' },
  sv: { home: 'Hem', news: 'Nyheter' },
  de: { home: 'Startseite', news: 'Nachrichten' },
  fr: { home: 'Accueil', news: 'Actualités' },
  ar: { home: 'الرئيسية', news: 'أخبار' },
  he: { home: 'בית', news: 'חדשות' },
  ja: { home: 'ホーム', news: 'ニュース' },
  // ... all 14 languages
};

function getBreadcrumbName(lang, type) {
  return BREADCRUMB_TRANSLATIONS[lang]?.[type] || BREADCRUMB_TRANSLATIONS.en[type];
}

Template Updated:

"name": "${getBreadcrumbName(lang, 'home')}"  // Instead of "Home"
"name": "${getBreadcrumbName(lang, 'news')}"  // Instead of "News"

Impact:

• Articles now match their language in structured data
• Better SEO consistency across all 14 languages
• Same breadcrumb translations as index pages

---

2-4. JSON-LD Extraction Regex (Comments #2799591413, #2799591427, #2799591441)

Problem: Regex (\{[\s\S]*?\}) stopped at first closing brace, breaking with nested objects

Solution: Changed to ([\s\S]*?) to capture full script content (3 occurrences)

Before:

const jsonLDMatch = html.match(/<script type="application\/ld\+json">\s*(\{[\s\S]*?\})\s*<\/script>/);
// Stops at first } - breaks with nested author.affiliation, publisher.logo, etc.

After:

const jsonLDMatch = html.match(/<script type="application\/ld\+json">\s*([\s\S]*?)\s*<\/script>/);
// Captures full script content including all nested objects

Fixed in:

• tests/seo-structured-data.test.js line 55 (enhanced NewsArticle properties test)
• tests/seo-structured-data.test.js line 70 (author information test)
• tests/seo-structured-data.test.js line 81 (publisher information test)

Impact: Tests now properly parse complete NewsArticle schema with nested structures

---

5. Language Parsing Test (Comment #2799591461)

Problem: Test expected exact pattern languagesArg.split('=')[1].split(',') but actual code splits in two steps

Solution: Updated test to check for actual parsing approach

Before:

expect(content).toContain("languagesArg.split('=')[1].split(',')");

After:

expect(content).toContain("languagesArg.split('=')[1]");
expect(content).toContain("languagesInput.split(',')");

Impact: Test now validates actual two-step parsing logic

---

6-7. Default Language Tests (Comments #2799591485, #2799591503)

Problem: Tests expected array literal ['en', 'sv'] but actual code uses string 'en,sv' with split

Solution: Updated 2 tests to check for string pattern instead of array literal

Fixed Tests:

1. "should default to EN and SV"
2. "should maintain default EN+SV behavior"

Before:

expect(content).toMatch(/\['en',\s*'sv'\]/);

After:

expect(content).toContain("'en,sv'");

Impact: Tests accurately reflect string-based default with runtime splitting

---

📊 Test Results

All Tests Passing ✅

Test Files  22 passed (22)
Tests       673 passed (673)
Duration    10.39s

Breakdown:

• ✅ SEO & Structured Data: 49 tests
• ✅ Multi-Language: 52 tests
• ✅ Agentic Workflow: 31 tests
• ✅ All other tests: 541 tests

Pass Rate: 100% (673/673)

---

🔧 Technical Changes

Files Modified (3)

1. scripts/article-template.js

   • Added BREADCRUMB_TRANSLATIONS constant (14 languages)
   • Added getBreadcrumbName() helper function
   • Updated BreadcrumbList template to use localized names
   • +31 lines

2. tests/seo-structured-data.test.js

   • Fixed JSON-LD extraction regex (3 occurrences)
   • Changed from greedy {...} pattern to full script capture
   • +3 changes

3. tests/multi-language.test.js

   • Updated language parsing test (1 change)
   • Updated default language tests (2 changes)
   • +3 changes

Total: 37 lines changed across 3 files

---

🌍 Breadcrumb Translations

Language	Home	News
English	Home	News
Swedish	Hem	Nyheter
Danish	Hjem	Nyheder
Norwegian	Hjem	Nyheter
Finnish	Etusivu	Uutiset
German	Startseite	Nachrichten
French	Accueil	Actualités
Spanish	Inicio	Noticias
Dutch	Home	Nieuws
Arabic	الرئيسية	أخبار
Hebrew	בית	חדשות
Japanese	ホーム	ニュース
Korean	홈	뉴스
Chinese	主页	新闻

Total: 14 languages × 2 breadcrumbs = 28 translations added

---

✅ Verification

Test Execution

✓ All 673 tests passing
✓ Zero failures
✓ Zero flaky tests
✓ Fast execution (10.39s)

Code Quality

✓ Breadcrumb helper function with fallback
✓ JSON-LD extraction handles nested objects
✓ Tests accurately reflect implementation
✓ No breaking changes

SEO Impact

✓ Structured data matches page language
✓ Better search engine understanding
✓ Consistent breadcrumb translations
✓ Proper localization across all pages

---

🎯 Impact Summary

Fix	Code	Tests	SEO	i18n
Breadcrumb Localization	✅	✅	✅	✅
JSON-LD Regex (3×)	-	✅	✅	-
Language Parsing Test	-	✅	-	-
Default Language Tests (2×)	-	✅	-	-

Overall: 1 code improvement, 6 test improvements, SEO enhanced, full i18n

---

🚀 Production Ready

All quality gates passing:

✅ All 673 tests passing (100%)
✅ Breadcrumbs localized (14 languages)
✅ JSON-LD extraction robust (nested objects)
✅ Tests match implementation (accurate validation)
✅ Zero breaking changes
✅ Backward compatible
✅ SEO improved (language-specific structured data)

Status: Ready for merge and deployment

---

Fixes: 7/7 (100%)
Tests: 673/673 passing (100%)
Languages: 14/14 breadcrumbs
Date: 2026-02-12

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

[github-actions]

🔍 Lighthouse Performance Audit

Category	Score	Status
Performance	85/100	🟡
Accessibility	95/100	🟢
Best Practices	90/100	🟢
SEO	95/100	🟢

📥 Download full Lighthouse report

Budget Compliance: Performance budgets enforced via budget.json

[Copilot]

Pull request overview

Copilot reviewed 37 out of 42 changed files in this pull request and generated 8 comments.

[Copilot]

These committed metadata files indicate a failed generation run (generated: 0, errors: 1, articles: []). If these files are meant to reflect the last successful generation, this will be misleading and may affect any automation that reads them. Consider re-running generation to produce a successful last-generation.json (or avoid committing transient error-state metadata).

[Copilot]

Fixed in 2509b23. Added news/metadata/*.json to .gitignore and removed both metadata files from git tracking. These are runtime state files that will be generated fresh on each run, preventing misleading error states in the repository.

[Copilot]

generation-result.json now records errors: 1 with generated: 0 and an empty articles list. If this JSON is consumed by workflows/CI to determine success, committing an error-state result can cause incorrect downstream behavior. Prefer updating this file only on successful runs or generating it during CI rather than versioning it.

[Copilot]

Fixed in 2509b23. Metadata files removed from tracking and added to .gitignore. They're now generated during CI rather than being versioned.

[Copilot]

The JSON-LD extraction regex stops at the first closing brace, which will produce invalid/partial JSON for the NewsArticle block (it contains nested objects like author/publisher/image). This will make the test flaky/broken depending on content. Consider extracting the full <script type="application/ld+json"> block (e.g., match until </script> and JSON.parse the whole payload) or parsing all JSON-LD scripts and selecting the one with "@type":"NewsArticle".

[Copilot]

Fixed in 2509b23. Changed regex from (\{[\s\S]*?\}) to ([\s\S]*?) to capture full script content. Now properly extracts NewsArticle schema with nested author/publisher/image objects. All 49 tests passing.

[Copilot]

Meta descriptions generated by the template are truncated with substring(0, 160), which can legitimately be exactly 160 characters. This assertion requires < 160, so it will fail on boundary values; use <= 160 (or reduce the template truncation) to avoid an off-by-one test failure.

[Copilot]

Fixed in 2509b23. Changed test to <= 160 to allow the boundary value. Template truncates to exactly 160 characters which is perfectly valid.

[Copilot]

This test hard-codes article filenames to only allow '-en.html'/'-sv.html'. Since the workflows/scripts now support generating additional languages (via the new languages input / --languages flag), this will fail as soon as other language articles are present in news/. Consider allowing the full supported set of language suffixes (or deriving allowed codes from configuration) so the test reflects the multi-language capability.

[Copilot]

Fixed in 2509b23. Updated regex to support all 14 languages: -(en|sv|da|no|fi|de|fr|es|nl|ar|he|ja|ko|zh)\.html$. Test now validates articles in any supported language.

[Copilot]

The script treats the --languages value as a comma-separated list only. Passing presets like "all" (documented in MULTI_LANGUAGE_STATUS.md / MULTI_LANGUAGE_IMPLEMENTATION.md) will currently produce an output file with '-all.html' and lang="all" rather than expanding to 14 language codes. Add preset expansion (all/nordic/eu-core) and validate/normalize language codes before generation.

[Copilot]

Fixed in 2509b23. Added preset expansion logic in generate-news-enhanced.js:

• --languages=all → all 14 languages
• --languages=nordic → en,sv,da,no,fi
• --languages=eu-core → en,sv,de,fr,es,nl

No more invalid -all.html files created.

[Copilot]

BreadcrumbList structured data uses hardcoded English names ("Home", "News") for all languages. For localized news indexes this makes the structured data inconsistent with the page language and undermines the PR’s localization goals. Consider using language-specific breadcrumb labels (e.g., from LANGUAGES config) for these "name" fields.

[Copilot]

Fixed in 2509b23. Added localized breadcrumb translations for all 14 languages (EN: Home/News, SV: Hem/Nyheter, DE: Startseite/Nachrichten, AR: الرئيسية/أخبار, etc.). BreadcrumbList structured data now uses lang.breadcrumbs.home and lang.breadcrumbs.news for proper localization.

[Copilot]

The plural articleCount message is incorrect (e.g., "8 all types"), which reads like a filter label rather than a count of articles. If this is intended to display an article count, it should use an articles noun (e.g., "8 articles") and optionally include the current filter context separately.

Suggested change

	articleCount: (n) => n === 1 ? '1 article' : `${n} all types`
	articleCount: (n) => n === 1 ? '1 article' : `${n} articles`

[Copilot]

Fixed in 2509b23. Changed from incorrect '${n} all types' to proper '${n} articles'. Applied correct plural forms across all 14 languages (EN: articles, SV: artiklar, DE: Artikel, FR: articles, AR: مقالات, JA: 件の記事, etc.).

[pethers]

@copilot apply changes based on the comments in this thread

[Copilot]

Just as a heads up, I was blocked by some firewall rules while working on your feedback. Expand below for details.

Warning

Firewall rules blocked me from connecting to one or more addresses (expand for details)

I tried to connect to the following addresses, but was blocked by firewall rules:

• cdn.cypress.io
  • Triggering command: /opt/hostedtoolcache/node/24.13.0/x64/bin/node node dist/index.js --exec install (dns block)

If you need me to access, download, or install something from one of these locations, you can either:

• Configure Actions setup steps to set up my environment, which run before the firewall is enabled
• Add the appropriate URLs or hosts to the custom allowlist in this repository's Copilot coding agent settings (admins only)

[github-actions]

🔍 Lighthouse Performance Audit

Category	Score	Status
Performance	85/100	🟡
Accessibility	95/100	🟢
Best Practices	90/100	🟢
SEO	95/100	🟢

📥 Download full Lighthouse report

Budget Compliance: Performance budgets enforced via budget.json

[Copilot]

Pull request overview

Copilot reviewed 37 out of 43 changed files in this pull request and generated 7 comments.

[Copilot]

BreadcrumbList structured data is hardcoded to English ("Home" / "News") regardless of the article language. Since the site supports 14 languages and index pages already localize breadcrumb names, this produces language-mismatched structured data for non-English articles. Consider adding a small breadcrumb translation map (or sharing the existing one from generate-news-indexes.js) and use it here based on lang/locale.

[Copilot]

Fixed in ae355ff. Added BREADCRUMB_TRANSLATIONS for all 14 languages and getBreadcrumbName() helper. BreadcrumbList now uses localized names (EN: Home/News, SV: Hem/Nyheter, DE: Startseite/Nachrichten, AR: الرئيسية/أخبار, etc.).

[Copilot]

The JSON-LD extraction regex in this test still uses a {...}-bounded non-greedy capture ((\{[\s\S]*?\})), which will truncate as soon as it hits the first closing brace. Since the NewsArticle JSON-LD now contains nested objects (author.affiliation, publisher.logo, image, etc.), this will cause JSON.parse to fail or parse incomplete data. Reuse the same full-script capture regex used in the earlier test (or otherwise target the first ld+json block) so nested objects are handled correctly.

[Copilot]

Fixed in ae355ff. Changed regex from (\{[\s\S]*?\}) to ([\s\S]*?) to capture full script content including nested objects. Test now properly parses NewsArticle with nested author/publisher/image.

[Copilot]

Same issue as above: this test still uses the {...}-bounded JSON-LD regex, which will stop at the first } and break parsing now that NewsArticle JSON-LD contains nested objects. Update this extraction to capture the full <script type="application/ld+json"> contents before JSON.parse.

[Copilot]

Fixed in ae355ff. Updated JSON-LD extraction regex to capture full script content, handling nested objects correctly.

[Copilot]

Same JSON-LD parsing problem here: the regex captures only up to the first closing brace, which is incompatible with the nested NewsArticle JSON-LD structure. Switch to a full-script capture (like /...>\s*([\s\S]*?)\s*<\/script>/) so JSON.parse reliably sees the entire JSON object.

[Copilot]

Fixed in ae355ff. Changed to full-script capture pattern ([\s\S]*?) so JSON.parse reliably sees the entire JSON object.

[Copilot]

This assertion expects the script to contain the exact pattern languagesArg.split('=')[1].split(','), but the current implementation splits in two steps (languagesArg.split('=')[1] into languagesInput, then languagesInput.split(',')). As written, this test will fail even though the behavior is correct. Update the expectation to match the current parsing approach (e.g., assert both substrings exist, or assert the languagesInput variable is split).

[Copilot]

Fixed in ae355ff. Updated test to check for actual two-step parsing: languagesArg.split('=')[1] and languagesInput.split(',') instead of expecting the combined pattern.

[Copilot]

These tests assert that generate-news-enhanced.js contains a default languages array literal ['en', 'sv'], but the script now defaults via a string ('en,sv') and then splits it. This will cause the test to fail despite correct default behavior. Adjust the assertion to check for the actual default (e.g., : 'en,sv' and subsequent split) rather than an array literal.

[Copilot]

Fixed in ae355ff. Updated test to check for string 'en,sv' instead of array literal ['en', 'sv'] to match actual default implementation.

[Copilot]

The Backward Compatibility test repeats the same incorrect assumption that the script defaults via an ['en','sv'] array literal. The current implementation defaults to the string 'en,sv' and splits it, so this assertion will fail. Update it to validate the actual default parsing logic instead of searching for the old literal pattern.

[Copilot]

Fixed in ae355ff. Updated backward compatibility test to validate string-based default 'en,sv' with runtime splitting instead of array literal.

[pethers]

@copilot apply changes based on the comments in this thread

[github-actions]

🔍 Lighthouse Performance Audit

Category	Score	Status
Performance	85/100	🟡
Accessibility	95/100	🟢
Best Practices	90/100	🟢
SEO	95/100	🟢

📥 Download full Lighthouse report

Budget Compliance: Performance budgets enforced via budget.json

[Copilot]

Pull request overview

Copilot reviewed 37 out of 43 changed files in this pull request and generated no new comments.