DOCX word count analyzer

A word is a whitespace-separated token with at least one letter or digit, so standalone punctuation is not counted, and each CJK character counts as one word on its own. Tracked deletions are excluded and insertions included, field results count but field codes do not, and hidden text is counted. Only named heading styles or outline levels start sections; bold look-alike headings do not. The XML has no layout, so there are no page or line counts, and nothing inside embedded objects or images is read. If the file declares a word count in its properties, it is shown next to the counted total instead of being trusted.