Document translation tutorial

Introduction

Document translation parses text from a document and uses the Machine Translation engine to translate it. The service then creates a new file that contains the translated text in the same format as the original document.

Limits

• Supported languages: For more information, see language codes
• Supported document types: doc, docx, xls, xlsx, csv, ppt, pptx, txt, srt, html, htm, xml, xhtml, json, xliff, tmx, dita, md, markdown, idml, and pdf.
• File size: Up to 50 MB.
• The file URL (FileUrl) must be publicly accessible, use a domain name instead of an IP address, and not contain spaces. Avoid using Chinese characters in the URL.
• The callback URL (CallbackUrl) must be publicly accessible, use a domain name instead of an IP address, and not contain spaces.
• The result file URL (TranslateFileUrl) is valid for one hour. After the document translation task is complete, download the result file within this one-hour period.

Steps

• You can create a document translation task (CreateDocTranslateTask).

• Call the CreateDocTranslateTask API to create a document translation task. Pass the document location in the FileUrl parameter. The FileUrl must be a publicly accessible URL. Host your document on Object Storage Service (OSS). Then, generate a signed URL to allow the Machine Translation service to read the document.

• After the task is created, the API returns a task ID (TaskId). Use the GetDocTranslateTask API and the TaskId to query the status of the translation task.

• When the GetDocTranslateTask API returns a task status of translated, retrieve the URL of the translated document from the TranslateFileUrl field in the response.

Task status codes