From a359c9da4ed354f0b0061be88a2376fb34d6348f Mon Sep 17 00:00:00 2001
From: Alex Auvolat <alex@adnab.me>
Date: Fri, 21 Jul 2023 22:17:50 +0200
Subject: filter dictionnary entries to limit to those with correct reading

---
 src/main.rs | 37 +++++++++++++++++++++++--------------
 1 file changed, 23 insertions(+), 14 deletions(-)

diff --git a/src/main.rs b/src/main.rs
index e081454..8a74d82 100644
--- a/src/main.rs
+++ b/src/main.rs
@@ -356,19 +356,20 @@ fn format_batch_aux<'a>(dict_idx: &DictIndex<'a>, count: usize, i: usize, batch:
         writeln!(f, r#"<details><summary>Explanation</summary>"#)?;
         let mut expl_batch = Vec::new();
         let mut expl_all = Vec::new();
-        for w in ex.expl.split(|c| c == ' ' || c == '~') {
-            let w = expl_clean_word(w);
-            let wchars = Charset::new(w);
+        for word in ex.expl.split(|c| c == ' ' || c == '~') {
+            let (keb, reb) = expl_clean_word(word);
+            let wchars = Charset::new(keb);
             if !wchars.intersects(&ex.chars) {
                 continue;
             }
-            if let Some(ents) = dict_idx.get(w) {
+            if let Some(ents) = dict_idx.get(keb) {
                 for ent in ents.iter() {
-                    let s = dict_str(w, ent);
-                    if wchars.intersects(&batch.chars) {
-                        expl_batch.push(s);
-                    } else {
-                        expl_all.push(s);
+                    if let Some(s) = dict_str(keb, reb, ent) {
+                        if wchars.intersects(&batch.chars) {
+                            expl_batch.push(s);
+                        } else {
+                            expl_all.push(s);
+                        }
                     }
                 }
             }
@@ -392,21 +393,29 @@ fn format_batch_aux<'a>(dict_idx: &DictIndex<'a>, count: usize, i: usize, batch:
     Ok(())
 }
 
-fn expl_clean_word(w: &str) -> &str {
+fn expl_clean_word(w: &str) -> (&str, Option<&str>) {
     let mut ret = w;
     for delim in ['(', '{', '['] {
         if let Some((s, _)) = ret.split_once(delim) {
             ret = s;
         }
     }
-    ret
+    let p = w.split_once('(')
+        .and_then(|(_, r)| r.split_once(')'))
+        .map(|(p, _)| p);
+    (ret, p)
 }
 
-fn dict_str<'a>(w: &str, ent: &roxmltree::Node<'a, 'a>) -> String {
+fn dict_str<'a>(qkeb: &str, qreb: Option<&str>, ent: &roxmltree::Node<'a, 'a>) -> Option<String> {
     let r_ele = ent.children().find(|x| x.has_tag_name("r_ele")).unwrap();
     let reb = r_ele.children().find(|x| x.has_tag_name("reb")).unwrap();
+    let reb = reb.text().unwrap().trim();
 
-    let mut ret = format!("{} [{}]", w, reb.text().unwrap().trim());
+    if qreb.map(|x| x != reb).unwrap_or(false) {
+        return None;
+    }
+
+    let mut ret = format!("{} [{}]", qkeb, reb);
 
     for sense in ent.children().filter(|x| x.has_tag_name("sense")) {
         if let Some(s) = sense.children().find(|x| x.has_tag_name("gloss")) {
@@ -417,7 +426,7 @@ fn dict_str<'a>(w: &str, ent: &roxmltree::Node<'a, 'a>) -> String {
     if ret.chars().rev().next() == Some(';') {
         ret.pop();
     }
-    ret
+    Some(ret)
 }
 
 fn format_index(batches: &[Batch], kanji_levels: &[(String, String)]) -> Result<()> {
-- 
cgit v1.2.3