Translate an XML file or product feed with every tag and attribute kept
Upload an XML file, such as a product feed, a content export or a configuration with user-facing text, and get the same file back with the text inside the elements translated. Element names, IDs, links, comments and the XML declaration stay exactly as they were; readable attributes such as alt, title and label are translated too. A file of up to 300 values is free in full; bigger files are $5, paid once.
- Text inside elements and readable attributes: title, alt, label, placeholder, caption, tooltip, description, summary, text, message, hint, subtitle, heading
- Element names, other attributes, comments and processing instructions never change
- script, style, code and pre are skipped, and so is anything marked translate="no"
- CDATA sections stay CDATA; text with & or < is escaped correctly when written back
- Up to 50,000 values for $5; Unlimited up to 200,000 values
How to translate an XML file
-
1
Upload the XML file
Drop the .xml file, choose the language, and list extra element names to skip if your feed has its own codes.
-
2
Check the first values free
The first text values appear translated next to the original, with their element path, such as catalog/product/name.
-
3
Unlock and download
Unlock the whole file for $5 and download it with the language code in its name.
How the file is read
The file is not parsed and printed again. The tool finds the exact position of each text node and each readable attribute value and replaces only those characters, so indentation, attribute order, quote style and comments stay byte for byte the same.
Elements whose names are identifiers, such as sku, id, url, code or status, keep their text. Their alt and title attributes, if they have any, are still translated: an img element keeps its src but gets a translated alt. In mixed content, such as a paragraph with a bold word inside, each piece of text is translated separately so the inner tags stay where they are.
Product feeds and special formats
Product feeds usually mix text (titles, descriptions) with values a marketplace matches on (GTINs, prices, availability, links). Only readable text is translated, and values that look like codes, such as in_stock, are left alone. If a feed field holds text you want to keep, list its element name as a key to skip.
Android strings.xml files (with a resources root) and XLIFF files are recognized and handled by their own formats, so plural forms and target elements are written the way those formats expect.
How the tool works, what it costs and how your files are handled is on the Translate JSON, XML and YAML page; prices for every tool are on the pricing page.
Last updated
Questions and answers
Will translating change my XML structure?
No. Only the characters of text nodes and readable attribute values are replaced. Element names, other attributes, comments and indentation stay as they were.
Are attributes translated?
Readable ones are: title, alt, label, placeholder, caption, tooltip, description, summary, text, message, hint, subtitle and heading. Attributes such as id, href, src and class are never touched.
What about CDATA?
Text inside CDATA is translated and written back inside the same CDATA section.
How much does it cost?
A file of up to 300 values is free in full. Bigger files are $5, paid once, for up to 50,000 values; Unlimited at $19 a week covers up to 200,000.