We extract the text with its coordinates, translate whole paragraphs rather than word by word, then write them back in place. Images, tables and logos stay untouched.