use human_regex::{exactly, one_or_more, or, punctuation, whitespace, word_boundary};
use stop_words::{get, LANGUAGE};

fn main() {
    #[cfg(all(any(feature = "nltk", feature = "iso"), not(feature = "constructed")))]
    {
        // Read in a file
        let document = std::fs::read_to_string("examples/foreword.txt").expect("Cannot read file");

        // Print the contents
        println!("Original text:\n{}", document);

        // Get the stopwords
        let words = get(LANGUAGE::English);

        // Remove punctuation and lowercase the text to make parsing easier
        let lowercase_doc = document.to_ascii_lowercase();
        let regex_for_punctuation = one_or_more(punctuation());
        let text_without_punctuation = regex_for_punctuation
            .to_regex()
            .replace_all(&lowercase_doc, "");

        // Make a regex to match stopwords with trailing spaces and punctuation
        let regex_for_stop_words =
            word_boundary() + exactly(1, or(&words)) + word_boundary() + one_or_more(whitespace());

        // Remove stop words
        let clean_text = regex_for_stop_words
            .to_regex()
            .replace_all(&text_without_punctuation, "");
        println!("\nClean text:\n{}", clean_text);
    }
}