Gathering word statistics

Full-text search can handle a lot of data. To give end users more insights into their texts, PostgreSQL offers the pg_stat function, which returns a list of words:

SELECT * FROM ts_stat('SELECT to_tsvector(''english'', comment) FROM pg_available_extensions') ORDER BY 2 DESC LIMIT 3; 
   word   | ndoc | nentry  
----------+------+-------- 
 function |   10 |     10 
 data     |   10 |     10 
 type     |    7 |      7 
(3 rows) 

The word column contains the stemmed word, ndoc tells us about the number of documents a certain word occurs. nentry indicates how often a word was found all together