This is a simplistic demonstration of how you can calculate the ratio of dictionary words to all words in a METS Alto OCR XML file.
A dump of Wiktionary is used as source for the dictionary.
The latest dump of the English Wiktionary is used because its available and somewhat sizable: ~2 million words.
$ make dictionary.dbDownloading the dump and creating the dictionary database will take a bit of time.
Afterwarts the script alto_words.py can be used to compute the ratio of dictionary words.
$ make install
$ source ./.venv/bin/activate
$ python alto_words.py example.xml