autocorrect 2.14.2

// autocorrect: false
use regex::Regex;
use std::{borrow::Cow, collections::HashMap};

const SPCIAL_PUNCTUATIONS: &str = "[.:!]([ ]*)";
const NORMAL_PUNCTUATIONS: &str = "[,?]([ ]*)";

lazy_static! {
    static ref FULLWIDTH_MAPS: HashMap<&'static str, &'static str> = map!(
      "," => "，",
      "." => "。",
      ";" => "；",
      ":" => "：",
      "!" => "！",
      "?" => "？",
    );
    static ref PUNCTUATION_WITH_LEFT_CJK_RE: Regex = regexp!(
        "{}{}{}",
        r"[\p{CJ}\w\d]+",
        NORMAL_PUNCTUATIONS,
        r"[\p{CJ}]+"
    );
    static ref PUNCTUATION_WITH_RIGHT_CJK_RE: Regex =
        regexp!("{}{}", r"[\p{CJ}]+", NORMAL_PUNCTUATIONS);
    static ref PUNCTUATION_WITH_SPEICAL_CJK_RE: Regex =
        regexp!("{}{}{}", r"[\p{CJ}]+", SPCIAL_PUNCTUATIONS, r"[\p{CJ}]+");
    static ref PUNCTUATION_WITH_SPEICAL_LAST_CJK_RE: Regex =
        regexp!("{}{}{}", r"[\p{CJ}]+", SPCIAL_PUNCTUATIONS, r#"["']?$"#);
    static ref PUNCTUATIONS_RE: Regex =
        regexp!("({}|{})", SPCIAL_PUNCTUATIONS, NORMAL_PUNCTUATIONS);
}

// fullwidth correct punctuations near the CJK chars
pub fn format(text: &str) -> Cow<'_, str> {
    let patterns = [
        &*PUNCTUATION_WITH_LEFT_CJK_RE,
        &*PUNCTUATION_WITH_RIGHT_CJK_RE,
        &*PUNCTUATION_WITH_SPEICAL_CJK_RE,
        &*PUNCTUATION_WITH_SPEICAL_LAST_CJK_RE,
    ];

    let mut result = Cow::Borrowed(text);
    for pattern in &patterns {
        if let Cow::Owned(new_text) = pattern.replace_all(&result, |cap: &regex::Captures| {
            fullwidth_replace_part(&cap[0])
        }) {
            result = Cow::Owned(new_text);
        }
    }

    if let Cow::Owned(new_text) = result {
        Cow::Owned(new_text)
    } else {
        Cow::Borrowed(text)
    }
}

fn fullwidth_replace_part(part: &str) -> String {
    PUNCTUATIONS_RE
        .replace_all(part, |cap: &regex::Captures| FULLWIDTH_MAPS[&cap[0].trim()])
        .to_string()
}

#[cfg(test)]
mod tests {
    use super::*;

    #[track_caller]
    fn assert_cases(cases: HashMap<&str, &str>) {
        for (source, exptected) in cases.into_iter() {
            let actual = format(source);
            assert_eq!(exptected, actual);
        }
    }

    #[test]
    fn test_fullwidth() {
        let cases = map!(
          "你好,这是一个句子." => "你好，这是一个句子。",
          // For Programming String with quote: "Hello你好." => "Hello 你好。"
          "\"请求参数错误.\"" => "\"请求参数错误。\"",
          "'请求参数错误.'" => "'请求参数错误。'",
          "!开头不处理." => "!开头不处理。",
          "刚刚买了一部 iPhone,好开心!" => "刚刚买了一部 iPhone，好开心！",
          "蚂蚁集团上市后有多大的上涨空间?" =>  "蚂蚁集团上市后有多大的上涨空间？",
          "我们需要一位熟悉 JavaScript、HTML5,至少理解一种框架 (如 Backbone.js、AngularJS、React 等) 的前端开发者." => "我们需要一位熟悉 JavaScript、HTML5，至少理解一种框架 (如 Backbone.js、AngularJS、React 等) 的前端开发者。",
          "蚂蚁疾奔:蚂蚁集团两地上市~全速推进!" =>                                                        "蚂蚁疾奔：蚂蚁集团两地上市~全速推进！",
          "蚂蚁集团是阿里巴巴 (BABA.N) 旗下金融科技子公司" =>                                                "蚂蚁集团是阿里巴巴 (BABA.N) 旗下金融科技子公司",
          "Dollar 的演示 $阿里巴巴.US$ 股票标签" =>                                                   "Dollar 的演示 $阿里巴巴.US$ 股票标签",
          "确保&quot;&gt;HTML Entity&lt;&quot;的字符&#34;不会被处理&#34; Ruby&amp;Go" => "确保&quot;&gt;HTML Entity&lt;&quot;的字符&#34;不会被处理&#34; Ruby&amp;Go",
          // No ; convert, because it's danger.
          "中文;中文" => "中文;中文"
        );

        assert_cases(cases);
    }

    #[test]
    fn test_fullwidth_with_cjk() {
        let cases = map!(
            "你好,这是一个句子." => "你好，这是一个句子。",
            "你好,這是一個句子." => "你好，這是一個句子。",
            "でもっと多くのことができるようになります.そんな新機能の数々をさっそく体験してみましょう." => "でもっと多くのことができるようになります。そんな新機能の数々をさっそく体験してみましょう。",
            "근면, 검소, 협동은 우리 겨레의 미덕이다." => "근면, 검소, 협동은 우리 겨레의 미덕이다."
        );

        assert_cases(cases);
    }
}