Translators receive source text in every condition imaginable: copied from PDFs with broken line breaks, riddled with smart quotes and non-breaking spaces, and salted with invisible characters that confuse CAT-tool segmentation. Cleaning the source before it enters the workflow saves hours of fighting the tool later.
This workflow prepares source text for translation: fix line breaks and whitespace, normalize quotes and Unicode, and remove invisible characters so segmentation and QA behave.
