Blog HTML Translator

An Apify Actor that scrapes blog pages and translates all visible text content to multiple target languages using AI-powered localization, while preserving the original HTML structure.

🌟 Features

• Multi-language support: Translate to 30+ languages simultaneously
• HTML structure preservation: Keeps all HTML tags, attributes, and formatting intact
• Smart text extraction: Only translates visible user-facing text (ignores scripts, styles, hidden elements)
• Translation dictionary: Optionally outputs a JSON mapping of original texts to translations
• Auto language detection: Automatically detect source language or specify manually
• Concurrent processing: Fast translation with parallel API calls
• Detailed logging: Track progress and debug issues easily

📥 Input

Field	Type	Required	Description
blog_urls	Array	Yes	URLs of blog pages to translate
source_language	String	No	Source language code or "auto" for detection
target_languages	Array	Yes	Target language codes (e.g., ["es", "fr", "de"])
lingo_api_key	String	Yes	Your Lingo.dev API key
include_translation_dictionary	Boolean	No	Include original→translated text mapping
max_concurrent_pages	Integer	No	Max pages to process concurrently (1-10)
request_timeout_secs	Integer	No	Timeout per page request (10-120 seconds)
proxy_configuration	Object	No	Optional proxy settings

Example Input

{
    "blog_urls": [
        { "url": "https://blog.example.com/my-article" },
        { "url": "https://blog.example.com/another-post" }
    ],
    "source_language": "auto",
    "target_languages": ["es", "fr", "de", "it"],
    "lingo_api_key": "your-api-key-here",
    "include_translation_dictionary": true,
    "max_concurrent_pages": 3
}

📤 Output

Dataset

Each translation produces a dataset item with:

{
    "url": "https://blog.example.com/my-article",
    "page_title": "My Amazing Article",
    "source_language": "en",
    "target_language": "es",
    "target_language_name": "Spanish",
    "html_key": "html-my-article-a1b2c3d4_es",
    "texts_extracted": 45,
    "texts_translated": 45,
    "word_count": 1250,
    "processing_time_ms": 3420,
    "status": "success",
    "timestamp": "2024-01-15T10:30:00Z",
    "translation_dictionary": {
        "Hello World": "Hola Mundo",
        "Read more": "Leer más"
    }
}

Key-Value Store

Translated HTML files are stored in the key-value store with keys like:

• html-{url_slug}_{language_code} - Individual translated HTML files
• SUMMARY - JSON summary of all translations

🌍 Supported Languages

Code	Language	Code	Language
en	English	pl	Polish
es	Spanish	ru	Russian
fr	French	uk	Ukrainian
de	German	ja	Japanese
it	Italian	zh	Chinese
pt	Portuguese	ko	Korean
nl	Dutch	ar	Arabic
tr	Turkish	sv	Swedish
da	Danish	no	Norwegian
fi	Finnish	cs	Czech
hu	Hungarian	ro	Romanian
el	Greek	th	Thai
vi	Vietnamese	id	Indonesian
ms	Malay	hi	Hindi
bn	Bengali	ta	Tamil
te	Telugu	he	Hebrew
fa	Persian

🔧 How It Works

1. Scraping Phase: Uses Playwright to render JavaScript-heavy pages and extract full HTML
2. Text Extraction: Parses HTML and extracts only visible text nodes (excluding scripts, styles, hidden elements)
3. Translation: Sends text batches to AI localization engine for translation
4. HTML Reconstruction: Replaces original texts with translations while preserving all HTML structure
5. Output: Saves translated HTMLs to key-value store and metadata to dataset

⚡ Performance Tips

• Use max_concurrent_pages: 1 for rate-limited APIs
• Set appropriate request_timeout_secs for slow-loading pages
• Enable proxy if scraping sites with geo-restrictions
• Use auto language detection only when source language varies

🔑 Getting Your API Key

1. Visit lingo.dev
2. Create an account
3. Navigate to Projects → API Key
4. Copy your API key

📝 Use Cases

• Content localization: Translate blog posts for international audiences
• Website migration: Prepare content for multi-language CMS
• SEO expansion: Create localized versions for different markets
• Content analysis: Extract translation dictionaries for terminology management

⚠️ Limitations

• Large pages with many text nodes may take longer to process
• Some dynamically loaded content may not be captured
• Translation quality depends on the AI localization service

🐛 Troubleshooting

"Translation failed" errors: Check your API key is valid and has sufficient credits

Missing translations: Some text may be in elements not captured (try adjusting page load time)

Empty results: Ensure the URLs are accessible and contain visible text content