more

src/lib/diacritics-helpers.ts

@@ -274,7 +274,7 @@ export const phonemeTable: Record<Phoneme, PhonemeInfo> = {
     const willIgnore = ["?", " ", "`", ".", "…", ","];
     
     const result: Phoneme[] = [];
-    const f = removeAccents(fIn);
+    const f = removeAccents(fIn).replace(/ă/g, "a");
     let index = 0;
     while (index < f.length) {
         const isLastTwoLetters = (index === f.length - 2 || f[index + 2] === " ");
@@ -336,6 +336,7 @@ export enum PhonemeStatus {
     ShortAForAlefBeforeFathatan,
     NOnFathatan,
     HamzaOnWow,
+    ArabicDefiniteArticleUl,
 }
 
 export function stateInfo({ state, i, phonemes, phoneme }: {
@@ -348,11 +349,11 @@ export function stateInfo({ state, i, phonemes, phoneme }: {
     const prevPLetter = last(state.pOut);
     const currentPLetter = state.pIn[0];
     const nextPLetter = state.pIn[1];
-    const isBeginningOfWord = state.pOut === "" || prevPLetter === " ";
-    const isEndOfWord = isOutOfWord(nextPLetter);
-    const phonemeInfo = phonemeTable[phoneme];
     const nextPhoneme = phonemes[i+1];
     const previousPhoneme = i > 0 && phonemes[i-1];
+    const isBeginningOfWord = (state.pOut === "" || prevPLetter === " ") || (previousPhoneme === "-Ul-" && prevPLetter === "ل");
+    const isEndOfWord = isOutOfWord(nextPLetter);
+    const phonemeInfo = phonemeTable[phoneme];
     const previousPhonemeInfo = (!isBeginningOfWord && i > 0) && phonemeTable[phonemes[i-1]];
     // const nextPhoneme = (phonemes.length > (i + 1)) && phonemes[i+1];
     // const nextPhonemeInfo = nextPhoneme ? phonemeTable[nextPhoneme] : undefined;
@@ -391,6 +392,9 @@ export function stateInfo({ state, i, phonemes, phoneme }: {
         if (isBeginningOfWord && phoneme === "u" && prevPLetter === " " && lastNonWhitespace(state.pOut) === "د") {
             return PhonemeStatus.EndOfDuParticle
         }
+        if (isBeginningOfWord && phoneme === "-Ul-" && currentPLetter === "ا" && nextPLetter === "ل") {
+            return PhonemeStatus.ArabicDefiniteArticleUl;
+        }
         if (phoneme === "a" && previousPhoneme === "U" && currentPLetter === "و") {
             return PhonemeStatus.HamzaOnWow;
         }

src/lib/diacritics.test.ts

@@ -586,6 +586,13 @@ const diacriticsSections: {
                 },
                 out: null,
             },
+            {
+                in: {
+                    p: "سختسری",
+                    f: "sakht sărey",
+                },
+                out: "سَخْتْسَری",
+            },
         ],
     },
     {
@@ -805,6 +812,20 @@ const diacriticsSections: {
                 },
                 out: "تَوَقُّع",
             },
+            {
+                in: {
+                    p: "راجع کېدل",
+                    f: "raaji kedul",
+                },
+                out: "راجِع کېد" + zwarakey + "ل",
+            },
+            {
+                in: {
+                    p: "ربیع",
+                    f: "rabee'",
+                },
+                out: "رَبِیع",
+            },
         ],
     },
     {
@@ -1054,6 +1075,44 @@ const diacriticsSections: {
             },
         ],
     },
+    {
+        describe: "With Arabic definate article -Ul- ال",
+        tests: [
+            {
+                in: {
+                    p: "حق الاجاره",
+                    f: "haq-Ul-ijaara",
+                },
+                out: "حَق اُلاِجارَه",
+            },
+            {
+                in: {
+                    p: "دار العلوم",
+                    f: "daar-Ul-Ulóom",
+                },
+                out: "دار اُلعُلُوم",
+            },
+        ],
+    },
+    // {
+    //     describe: "double consonants on end of words",
+    //     tests: [
+    //         {
+    //             in: {
+    //                 p: "حق",
+    //                 f: "haqq",
+    //             },
+    //             out: "حَقّ",
+    //         },
+    //         {
+    //             in: {
+    //                 p: "حق پر",
+    //                 f: "haqq par",
+    //             },
+    //             out: "حَقّ پَر",
+    //         },
+    //     ],
+    // },
 ];
 
 diacriticsSections.forEach((section) => {

src/lib/diacritics.ts

@@ -195,6 +195,12 @@ function processPhoneme(
                 addP(hamzaAbove),
                 addP(diacritic),
             )(state)
+        : (phs === PhonemeStatus.ArabicDefiniteArticleUl) ?
+            pipe(
+                advanceP,
+                addP(pesh),
+                advanceP,
+            )(state)
         : state;
 
 

src/lib/phonetics-to-diacritics.test.ts

@@ -970,7 +970,6 @@ const toTest: Array<{
         },
         out: "حَقّ پَر",
     },
-    // TODO: Allow ' in there
     {
         in: {
             p: "راجع کېدل",