Skip to content

Commit

Permalink
Merge pull request #2 from Vyshantha/Judeo-Arabic-Suggestion-Translit…
Browse files Browse the repository at this point in the history 
…erate

Judeo-Arabic-Suggestion-Transliterate
  • Loading branch information
@Vyshantha
Vyshantha authored Oct 10, 2023
2 parents d762675 + 2d57c10 commit 0647a52
Show file tree
Hide file tree
Showing 3 changed files with 74 additions and 20 deletions.
1 change: 1 addition & 0 deletions index.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -51,6 +51,7 @@
<span class="judeoCopyText">Copy Transliteration text - Judeo-Arabic</span>
</div>
</div>
<div class="vocalised" id="vocalised" title="Vocalised Text in Arabic" onclick="vocalised()"> 🔊 </div>
<div class="hebrewFonts">
<img src="SerifHebrew.png" title="Serif Hebrew font type" alt="Serif Hebrew font type" width="20px" height="20px" onclick="loadSerifHebrew()"/>
<img src="Rashi.png" title="Rashi Hebrew font type" alt="Rashi Hebrew font type" width="25px" height="25px" onclick="loadRashi()"/>
Expand Down
28 changes: 28 additions & 0 deletions style.css
View file
Original file line number Diff line number Diff line change
Expand Up @@ -190,6 +190,34 @@ body{
border:2px solid #233d28;
}

.vocalised {
display:inline-block;
color:#867d7d;
border:1px solid #CCC;
background:#867d7d;
cursor:pointer;
max-width: 100px;
padding: 4px;
text-align: center;
margin-right: 5px;
margin-bottom: 5px;
}
.vocalised:active {
color:red;
box-shadow: 0 0 5px -1px rgba(0,0,0,0.6);
}
.nonvocalised {
display:inline-block;
color:#742222;
border:1px solid #CCC;
background:#742222;
cursor:pointer;
max-width: 100px;
padding: 4px;
text-align: center;
margin-right: 5px;
}

.dropbtn {
background-color: #d3cbcb;
color: white;
Expand Down
65 changes: 45 additions & 20 deletions transliterate.js
View file
Original file line number Diff line number Diff line change
Expand Up @@ -48,6 +48,8 @@ function copyContent2() {
navigator.clipboard.writeText(document.getElementById("textarea2").value);
}

var vocalisedText = "";

function transliterate() {
if (document.getElementById("textarea1").value.indexOf("script>") > -1 || document.getElementById("textarea2").value.indexOf("script>") > -1) {
document.getElementById("textarea1").value = "";
Expand Down Expand Up @@ -85,19 +87,16 @@ function transliterate() {

/*
TODO CORRECTIONS
1. Suggestions : ta-marbuta (-h & -at) and for ג - ج & غ
2. Shadda for double letters לל = لّ
3. No diacritics in Hebrew as shadda, harakat, maddah do not exist in Hebrew Unicode
TODO CORRECTIONS : Shadda for double letters לל = لّ
Problems with UNICODE > scope for improvement & new L2 for Unicode
Arabic diakritics used in Hebrew - typing problem ta-marbuta, harakat, maddah, shadda-*
"David sprach sich dafür aus die hebräische Komponente des Judäo-arabischen NICHT zu übersetzen, sondern einfach nur zu transliterieren und so stehen lassen.""
a. Hebrew woerter > direkt transkibiert oder uebersetzen : أدوناي : adonay = י״י / יהוה / ה / יי / ד
b. Hebrew abkurzungen ? ר׳ = Rabbi ; ש״צ = https://www.halachipedia.com/index.php?title=Shaliach_Tzibur ב״ר = Ben Rabbi = Sohn von Rabbi ; נ״ע = נשמתו עדן abkurzung
c. BL in hebrew then should be translated or transliterated ?
d. determine if anything needs to be fixed : https://unicode.org/L2/L2003/03299-hebrew-issues.pdf
Problems with UNICODE > scope for improvement & new L2 for Unicode
Arabic diakritics used in Hebrew - find FONT ? Judeo-arabisch > typing problem ta-marbuta, harakat, maddah, shadda-*
VALIDATION TEST ground truth await Arabic transliteration
I Judeo-arabisch :
Expand Down Expand Up @@ -159,29 +158,29 @@ function transliterate() {
console.log("nordafrikanischen - Texten אוול für أول")
resultAr = (!judeoToArabic[textJud[u+5]] || (judeoToArabic[textJud[u+5]] && typesOfWordDemarkers.indexOf(judeoToArabic[textJud[u+5]]) == -1) || (judeoToArabic[textJud[u+5]] && nonjoining.indexOf(judeoToArabic[textJud[u+5]]) > -1)) ? resultAr + "أول" : resultAr + "أولـ";
u = u + 3;
} else if (typesOfWordDemarkers.indexOf(textJud[u+3]) > -1 && textJud[u+2] == "ל" && textJud[u+1] == "ו" && textJud[u] == "א" && typesOfWordDemarkers.indexOf(textJud[u-1]) > -1) { //
} else if (typesOfWordDemarkers.indexOf(textJud[u+3]) > -1 && textJud[u+2] == "ל" && textJud[u+1] == "ו" && textJud[u] == "א" && typesOfWordDemarkers.indexOf(textJud[u-1]) > -1) {
console.log("nordafrikanischen - Texten für أول : vergleiche jedoch אול bei Saadiya Gaon")
resultAr = resultAr + "أول";
u = u + 2;
} else if (textJud[u+2] == " " && textJud[u+1] == "ל" && textJud[u] == "א" && typesOfWordDemarkers.indexOf(textJud[u-1]) > -1) { // irakischen - der Artikel oft als אל als seperates Wort z.B. אל רחמאן anstelle von الرحمن hier schreibt das judäoarabische in רחמאן auch ein Alef, wo im Arabischen gar kein Alif ist. Teilweise erscheint אל auch als Ligatur ﭏ.
} else if (textJud[u+2] == " " && textJud[u+1] == "ל" && textJud[u] == "א" && typesOfWordDemarkers.indexOf(textJud[u-1]) > -1) {
console.log("irakischen - Texten der Artikel אל : seperates Wort")
resultAr = (!judeoToArabic[textJud[u+2]] || (judeoToArabic[textJud[u+2]] && typesOfWordDemarkers.indexOf(judeoToArabic[textJud[u+2]]) == -1) || (judeoToArabic[textJud[u+2]] && nonjoining.indexOf(judeoToArabic[textJud[u+2]]) > -1)) ? resultAr + "ال" : resultAr + "ال ";
resultAr = (!judeoToArabic[textJud[u+2]] || (judeoToArabic[textJud[u+2]] && typesOfWordDemarkers.indexOf(judeoToArabic[textJud[u+2]]) == -1) || (judeoToArabic[textJud[u+2]] && nonjoining.indexOf(judeoToArabic[textJud[u+2]]) > -1)) ? resultAr + "ال" : resultAr + "ال "; // irakischen - der Artikel oft als אל als seperates Wort z.B. אל רחמאן anstelle von الرحمن hier schreibt das judäoarabische in רחמאן auch ein Alef, wo im Arabischen gar kein Alif ist. Teilweise erscheint אל auch als Ligatur ﭏ.
u = u + 2;
// TODO ligature ?
} else if (textJud[u] == "ל" && typesOfWordDemarkers.indexOf(textJud[u-1]) > -1) { // nordafrikanischen - den Artikel jedoch auch oft nur als ל gesehen z.B. ליהוד "die Juden" für اليهود in der nordafrikanischen Zeitung בית ישראל
} else if (textJud[u] == "ל" && typesOfWordDemarkers.indexOf(textJud[u-1]) > -1) {
console.log("nordafrikanischen - Texten der Artikel nur als ל geschrieben")
resultAr = resultAr + "ال ";
resultAr = resultAr + "ال "; // nordafrikanischen - den Artikel jedoch auch oft nur als ל gesehen z.B. ליהוד "die Juden" für اليهود in der nordafrikanischen Zeitung בית ישראל
} else if (textJud[u] && textJud[u+1] && judeoToArabic[textJud[u] + textJud[u+1]]) {
console.log("Double letter consonant") // include joined Article ال (al / el / il)
resultAr = resultAr + judeoToArabic[textJud[u] + textJud[u+1]];
u = u + 1;
} else if (typesOfWordDemarkers.indexOf(textJud[u+1]) > -1 && textJud[u] && textJud[u] == "ה" && judeoToArabic[textJud[u]]) { // TODO per word basis suggestion
} else if (typesOfWordDemarkers.indexOf(textJud[u+1]) > -1 && textJud[u] && textJud[u] == "ה" && judeoToArabic[textJud[u]]) {
console.log("Single letter tāʾ marbūṭa or character - ה case with suggestions")
resultAr = (nonjoining.indexOf(judeoToArabic[textJud[u-1]]) > -1) ? resultAr + "ة " + resultAr + "ه " : resultAr + "ـة " + resultAr + "ـه ";
resultAr = (nonjoining.indexOf(judeoToArabic[textJud[u-1]]) > -1) ? resultAr + "ة " : resultAr + "ـة ";
u = u + 1;
} else if (typesOfWordDemarkers.indexOf(textJud[u+1]) > -1 && textJud[u] && textJud[u] == "ת") { // TODO per word basis suggestion
} else if (typesOfWordDemarkers.indexOf(textJud[u+1]) > -1 && textJud[u] && textJud[u] == "ת") {
console.log("Single letter tāʾ marbūṭa or character - ת case with suggestions")
resultAr = (nonjoining.indexOf(judeoToArabic[textJud[u-1]]) > -1) ? resultAr + "ة " + resultAr + "ه " : resultAr + "ـة " + resultAr + "ـه ";
resultAr = (nonjoining.indexOf(judeoToArabic[textJud[u-1]]) > -1) ? resultAr + "ة " : resultAr + "ـة ";
} else if (textJud[u] == "ّ" && textJud[u-1] == "י") {
console.log("Yod followed by Shadda then alef to be included")
resultAr = resultAr + "ّ" + "ا";
Expand All @@ -201,9 +200,6 @@ function transliterate() {
} else if (typesOfWordDemarkers.indexOf(textJud[u+1]) > -1 && textJud[u] && judeoToArabic[textJud[u]] && textJud[u-1] == " " && ((textJud[u-2] == "ק" && textJud[u-3] == "ר" && textJud[u-4] == "פ") || (textJud[u-2] == "ל" && textJud[u-3] == "צ" && textJud[u-4] == "פ"))) {
console.log("1 digit number representation with letter consonant after capital or paragraph פרק or פצל")
resultAr = resultAr + indoToArabicNumerals[numeralsIndo[textJud[u]]];
} else if ((textJud[u+1] == "\u0307" && textJud[u] == "ג") || textJud[u] == "ג̇" || textJud[u] == "גׄ") { // TODO per word basis suggestion
console.log("גׄ suggestions")
resultAr = resultAr + "غ" + resultAr + "ج";
} else if (textJud[u] == "ס" && textJud[u-1] == "נ" && textJud[u-2] == "ו") {
console.log("nicht emphatische Konsonanten und halb emphatisch selten Texte")
resultAr = resultAr + "ص"; // in manchen Fällen werden auch emphatische Konsonanten nicht emphatisch geschrieben, oder nicht emphatische Konsonanten emphatisch z.B. ונס für ونص "und halb", dies ist jedoch eher eine seltene Ausnahme die mir bisher nur in marokkanischen Texten begegnet ist.
Expand All @@ -221,9 +217,21 @@ function transliterate() {
} else if (unprocessed[i].startsWith("ا")) {
console.log("Multi-word suggestion when beginning with ا ");
processed = processed + unprocessed[i] + ' ' + unprocessed[i].replace("ا","أ") + ' ';
} else */ if (unprocessed[i].endsWith("ي")) {
} else */

if (unprocessed[i].indexOf("غ") > -1) {
console.log("Suggestions - גׄ case")
processed = processed + unprocessed[i] + ' ' + unprocessed[i].replace("غ", "ج") + ' ';
}

if (unprocessed[i].endsWith("ي")) {
console.log("Word end to be processed alef maksura")
processed = processed + unprocessed[i].replace("ي","ى") + ' ';
} else if (unprocessed[i].endsWith("ة") || unprocessed[i].endsWith("ـة")) {
console.log("Suggestions tāʾ marbūṭa - ה & ת cases")
processed = processed + unprocessed[i] + ' ' + unprocessed[i].replace("ة", "ه").replace("ـة","ـه") + ' ';
} else if (unprocessed[i].indexOf("الما") > -1) {
processed = processed + "الماء" + ' ';
} else {
console.log("Un processed word");
processed = processed + unprocessed[i] + ' ';
Expand All @@ -238,6 +246,23 @@ function transliterate() {
}
}

function vocalised() {
if (localStorage.getItem("vocalised") == 'NO' || localStorage.getItem("vocalised") == null || localStorage.getItem("vocalised") == undefined) {
localStorage.setItem("vocalised","YES");
document.getElementById("vocalised").classList.add('vocalised');
document.getElementById("vocalised").classList.remove('nonvocalised');
document.getElementById("vocalised").title = "Vocalised Text in Arabic";
document.getElementById("textarea2").value = vocalisedText;
} else if (localStorage.getItem("vocalised") == 'YES') {
localStorage.setItem("vocalised","NO");
document.getElementById("vocalised").classList.add('nonvocalised');
document.getElementById("vocalised").classList.remove('vocalised');
document.getElementById("vocalised").title = "Non-vocalised Text in Arabic";
vocalisedText = document.getElementById("textarea2").value;
document.getElementById("textarea2").value = document.getElementById("textarea2").value.replaceAll("\uFE70","").replaceAll("\uFE71","").replaceAll("\uFE72","").replaceAll("\uFE74","").replaceAll("\u08F0","").replaceAll("\u08F1","").replaceAll("\u08F2","").replaceAll("\u064C","").replaceAll("\u064D","").replaceAll("\u064B","").replaceAll("\u064E","").replaceAll("\u0618","").replaceAll("\uFE76","").replaceAll("\uFE77","").replaceAll("\u064F","").replaceAll("\u0619","").replaceAll("\uFE78","").replaceAll("\uFE79","").replaceAll("\u0650","").replaceAll("\uFE7A","").replaceAll("\uFE7B","").replaceAll("\u061A","").replaceAll("\uFE7E","").replaceAll("\u0652","").replaceAll("\uFC5E","").replaceAll("\uFC60","").replaceAll("\uFC61","").replaceAll("\uFC62","").replaceAll("\uFC63","").replaceAll("\uFCF2","").replaceAll("\uFCF3","").replaceAll("\uFCF4","").replaceAll("\uFC5F","").replaceAll("\u0651","").replaceAll("\uFE7D","").replaceAll("\uFE7C","").replaceAll("\u0670","");
}
}

function swap(json) {
var ret = {};
for (var key in json) {
Expand Down

0 comments on commit 0647a52

Please sign in to comment.