From ab232ceb32b51ac8553692cf8a2b1f86fa975f7d Mon Sep 17 00:00:00 2001
From: Alex Auvolat <alex@adnab.me>
Date: Tue, 28 Nov 2023 16:31:28 +0100
Subject: add furigana override file to fix edge cases

---
 src/datafiles.rs | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)

(limited to 'src/datafiles.rs')

diff --git a/src/datafiles.rs b/src/datafiles.rs
index 0e526ef..3065fbf 100644
--- a/src/datafiles.rs
+++ b/src/datafiles.rs
@@ -176,6 +176,26 @@ pub fn read_examples(all_kanji: &Charset) -> Result<Vec<Example>> {
     Ok(ret)
 }
 
+pub fn read_furigana_overrides() -> Result<HashMap<String, String>> {
+    let file = fs::File::open("data/furigana_overrides")?;
+
+    let mut ret = HashMap::new();
+    let re = regex::Regex::new(r#"\|\|\w+\]\]"#)?;
+
+    for line in io::BufReader::new(file).lines() {
+        let line = line?;
+        let line = line.trim();
+        if !line.is_empty() {
+            let clean = re.replace_all(line, "").replace("[[", "");
+            if clean != line {
+                ret.insert(clean, line.to_string());
+            }
+        }
+    }
+
+    Ok(ret)
+}
+
 #[derive(Clone, Debug, Serialize, Deserialize, PartialEq)]
 pub struct JlptVocab {
     pub level: String,
-- 
cgit v1.2.3