Add Hebrew voiced reading support and update related tests

app/workbook.py

@@ -15,6 +15,7 @@ whole word is not in LEX. Punctuation stays on its word.
 from __future__ import annotations
 
 import itertools
+import re
 
 from .grammar import MAX_VARIANTS, PRONOUNS, WORDS
 
@@ -205,6 +206,15 @@ def _plain(he: str) -> str:
     return he.translate(POINTS)
 
 
+# words pointed for the voice wherever they appear: unpointed דוד is read "David"
+MISREAD = {"דוד": "דּוֹד", "הדוד": "הַדּוֹד", "דודי": "דּוֹדִי", "ודודי": "וְדוֹדִי", "דודנו": "דּוֹדֵנוּ"}
+_MISREAD = re.compile(rf"(?<![\u05D0-\u05EA])({'|'.join(MISREAD)})(?![\u05D0-\u05EA])")
+
+
+def _voiced(he: str) -> str:
+    return _MISREAD.sub(lambda m: MISREAD[m[1]], he)
+
+
 HE_PRONOUNS = "אֲנִי אַתָּה אַתְּ הוּא הִיא אֲנַחְנוּ אַתֶּם הֵם".split()   # in PRONOUNS order
 AR_PRONOUNS = [alts[0][0] for _, _, _, alts in PRONOUNS]
 # lesson 1's verbs (בניין ראשון, פַעַל): Hebrew past forms and LEX keys, one per pronoun in PRONOUNS order, then the future
@@ -474,8 +484,9 @@ def workbook_phrases() -> list[dict]:
                 words = [_word(t) for t in sentence.split()]
                 phrase = dict(cat=f"l{n}-{tid}", he=he, tr=" ".join(t for _, t, _ in words),
                               variants=_variants(words), read=kind == "read")
-                if he_say:
-                    phrase["he_say"] = he_say[0]
+                say = _voiced(he_say[0] if he_say else he)
+                if say != he:
+                    phrase["he_say"] = say
                 if in_order:
                     phrase["seq"] = seq
                 out.append(phrase)

static/index.html

@@ -226,7 +226,7 @@ function variation(kind){
 /* Hebrew prompt read aloud on every new sentence (option): edge-tts in Edge, gTTS in other browsers. */
 const heEngine=/\bEdg\//.test(navigator.userAgent)?"edge":"gtts";
 const heUrls={};   // phrase id -> object URL
-let heAudio=null, hePending=null;
+let heAudio=null, hePending=null, opened=false;   // opened: the sentence the app opens on is not read aloud
 async function promptUrl(id){
   if(!heUrls[id]){
     try{
@@ -240,7 +240,7 @@ async function promptUrl(id){
 }
 async function readPrompt(){
   if(heAudio){ heAudio.pause(); heAudio=null; }
-  if(driving||!$("readHe").checked||!cur||cur.read) return;   // driving mode reads the prompt itself, in sequence; a read phrase's Hebrew is its answer
+  if(!opened||driving||!$("readHe").checked||!cur||cur.read) return;   // driving mode reads the prompt itself, in sequence; a read phrase's Hebrew is its answer
   const id=cur.id;
   if(!await promptUrl(id)) return;
   if(!cur||cur.id!==id) return;   // moved on while it was loading
@@ -515,7 +515,7 @@ fetch("/api/phrases").then(r=>r.json()).then(d=>{
   for(const v of d.voices||[]) $("voiceId").add(new Option(v.name,v.id));
   const saved=store.get("ihki.voice",""); if((d.voices||[]).some(v=>v.id===saved)) $("voiceId").value=saved;
   if(!CATS.some(c=>c.id===cat)) cat=CATS[0].id;
-  buildCats(); startCat();
+  buildCats(); startCat(); opened=true;
 }).catch(()=>{ $("prompt").textContent="השרת לא זמין."; });
 </script>
 <script>

tests/test_scoring.py

@@ -393,6 +393,9 @@ def test_workbook_lesson_2():
     assert you["tr"] == "בִּנְתֵ-כּ" and you["he_say"] == "הבת שֶׁלָּךְ" and you["variants"] == [("بنتك", "b i n t i k")]
     assert by["l2-poss", "הכלב שלכם"]["variants"] == [("كلبكم", "k a l b k o m"), ("كلبكم", "k a l b k u")]
     assert ("جدّي", "s iː d i") in by["l2-poss", "הסבא שלי"]["variants"]   # the other word for סבא is accepted
+    # דוד is pointed for the prompt voice, which otherwise reads it "David"; דודה is read correctly as it is
+    assert by["l2-family", "דוד (מצד האב)"]["he_say"] == "דּוֹד (מצד האב)" and "he_say" not in by["l2-family", "דודה (מצד האב)"]
+    assert by["l2-say", "איפה כלבכם? הוא אצל דודי (מצד אמא)."]["he_say"] == "איפה כלבכם? הוא אצל דּוֹדִי (מצד אמא)."
     p = next(p for p in PHRASES if p["cat"] == "l2-talk" and p["seq"] == 2)
     assert p["tr"] == "וּעַלֵיכֹּם אֵ(ל)סַّלַאםֻ" and p["variants"][0][1] == "u ʕ a l eː k u m | i s s a l aː m u"
     p = next(p for p in PHRASES if p["cat"] == "l2-say" and p["seq"] == 22)