From 420cbc1f71653fdcff334eba24a23ebc84456845 Mon Sep 17 00:00:00 2001
From: hanan-nana <hanan.dev04@gmail.com>
Date: Fri, 24 Oct 2025 23:23:09 +0900
Subject: [PATCH] =?UTF-8?q?=ED=95=9C=EA=B8=80=20=EC=98=81=EC=96=B4=20?=
 =?UTF-8?q?=EB=AC=B8=EC=9E=A5=20=ED=8C=90=EB=B3=84?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 libs/braillify/src/sentence_analyzer.rs | 80 +++++++++++++++++++++++++
 1 file changed, 80 insertions(+)
 create mode 100644 libs/braillify/src/sentence_analyzer.rs

diff --git a/libs/braillify/src/sentence_analyzer.rs b/libs/braillify/src/sentence_analyzer.rs
new file mode 100644
index 0000000..1f45f39
--- /dev/null
+++ b/libs/braillify/src/sentence_analyzer.rs
@@ -0,0 +1,80 @@
+use crate::utils;
+
+/// 문장의 언어 주체를 판단하는 분석기
+pub struct SentenceAnalyzer;
+
+impl SentenceAnalyzer {
+    /// 문장을 공백으로 단어 분리
+    pub fn parse_sentence(text: &str) -> Vec<String> {
+        text.split_whitespace()
+            .map(|s| s.to_string())
+            .collect()
+    }
+    
+    /// 단어가 혼합 단어인지 확인 (영어+한글+특수기호)
+    pub fn is_mixed_word(word: &str) -> bool {
+        let has_english = word.chars().any(|c| c.is_ascii_alphabetic());
+        let has_korean = word.chars().any(|c| utils::is_korean_char(c));
+        
+        // 영어와 한글이 모두 있으면 혼합 단어
+        has_english && has_korean
+    }
+    
+    /// 문장에 혼합 단어가 있는지 확인
+    pub fn has_mixed_words(text: &str) -> bool {
+        let words = Self::parse_sentence(text);
+        words.iter().any(|word| Self::is_mixed_word(word))
+    }
+    
+    /// 문장의 언어 주체 판단
+    pub fn determine_language_dominance(text: &str) -> LanguageDominance {
+        // 1. 혼합 단어가 있으면 한글 문장
+        if Self::has_mixed_words(text) {
+            return LanguageDominance::Korean;
+        }
+        
+        // 2. 혼합 단어가 없으면 비율 기반 판단
+        let words = Self::parse_sentence(text);
+        let total_words = words.len();
+        
+        if total_words == 0 {
+            return LanguageDominance::English;
+        }
+        
+        let korean_words = words.iter()
+            .filter(|word| word.chars().any(|c| utils::is_korean_char(c)))
+            .count();
+        
+        let english_words = words.iter()
+            .filter(|word| word.chars().all(|c| c.is_ascii_alphabetic()))
+            .count();
+        
+        let korean_ratio = korean_words as f64 / total_words as f64;
+        let english_ratio = english_words as f64 / total_words as f64;
+        
+        if korean_ratio > english_ratio {
+            LanguageDominance::Korean
+        } else if english_ratio > korean_ratio {
+            LanguageDominance::English
+        } else {
+            // 비율이 같으면 첫 단어 기준
+            if let Some(first_word) = words.first() {
+                if first_word.chars().any(|c| utils::is_korean_char(c)) {
+                    LanguageDominance::Korean
+                } else {
+                    LanguageDominance::English
+                }
+            } else {
+                LanguageDominance::English
+            }
+        }
+    }
+}
+
+/// 문장의 언어 주체
+#[derive(Debug, Clone, Copy, PartialEq)]
+pub enum LanguageDominance {
+    Korean,   // 한글이 주체
+    English,  // 영어가 주체
+}
+