Produce a software that reduces any french .doc file to its 1,200 less current words

I have word files of 3 to 15 pages on various topics, written in french. I want each of them reduced to only the 1,200 less current of their words. This way little words like "le", "la"... ("the") will disappear. The coder has to find the database where the frequency of french words is to be documented. The number of results on a google research could be one ("sport" has 2.1 billion results, "?conomie" 48 million). ## Deliverables - Attached a sample of a source .doc file. - Titles (level1, level2 or level3) are to be left as they are. Only the body text content is to be filtered. - The requested software allows to load a .doc file, then it proceeds to the reduction to 1,200 words, and it produces a result .doc file.