Node/Typescript app which demonstrates text search.
There are two input files included in the repo; one with a list of approximately 5,400 first names and one with the entire text of Oliver Twist. The app counts the occurences of each name in the text of the story and writes a file with results sorted from most to least frequent.
The text search is handled by the streamsearch npm package which implements the Boyer–Moore–Horspool algorithm. It takes around 20 seconds to run.
Note that the matching is case sensitive so a search for "Oliver" will return 830 results and a search for "OLIVER" will return 51. I considered lower casing the whole text to return both combined but that resulted in a large number of false potisitves for 2 letter names.
-
clone the repo
$ git clone [email protected]:sauntimo/name-counter.git name-counter -
initialise
$ cd name-counter && npm i -g
$ namecounter
Similar functionality is also provided at via an API and a simple frontend at name-counter-api.herokuapp.com, with code at github.com/sauntimo/name-counter-api.