deva2laos.cs

using System;
using System.Collections;
using System.IO;
using System.Text;
using System.Text.RegularExpressions;

namespace VRI.CSCD.Conversion
{
    class Deva2Laos
    {
        static void Main(string[] args)
        {
            try
            {
                if (args.Length < 2 || args.Length > 3)
                {
                    PrintUsage();
                    return;
                }

                // validate args
                FileInfo fi = new FileInfo(args[0]);
                if (fi.Exists == false)
                {
                    Console.WriteLine("Input file does not exist");
                    return;
                }

                DirectoryInfo di = new DirectoryInfo(args[1]);
                if (di.Exists == false)
                {
                    Console.Write("Output directory does not exist. Would you like to create it?[y/n] ");
                    if (Console.ReadLine().ToLower().Substring(0, 1) == "y")
                    {
                        di.Create();
                    }
                    else
                    {
                        return;
                    }
                }

                Deva2Laos d2 = new Deva2Laos();
                d2.InputFilePath = args[0];
                d2.OutputFilePath = di.FullName + "\\" + fi.Name;
                d2.Convert();
            }
            catch (Exception ex)
            {
                Console.WriteLine("Exception: " + ex);
            }
        }

        static void PrintUsage()
        {
            Console.WriteLine("Transliterates Unicode Devanagari to Unicode Lao script (Pali)");
            Console.WriteLine("syntax:");
            Console.WriteLine("deva2laos input [output]");
        }
        // end static methods


        private Hashtable dev2Lao;

        public Deva2Laos()
        {
            dev2Lao = new Hashtable();

            // Niggahita (Lao Niggahita)
            dev2Lao['\x0902'] = '\x0ECD'; 

            // independent vowels (Using Or O + Vowel sign)
            dev2Lao['\x0905'] = '\x0EAD'; // a (Or O)
            dev2Lao['\x0906'] = "\x0EAD\x0EB2"; // aa
            dev2Lao['\x0907'] = "\x0EAD\x0EB4"; // i
            dev2Lao['\x0908'] = "\x0EAD\x0EB5"; // ii
            dev2Lao['\x0909'] = "\x0EAD\x0EB8"; // u
            dev2Lao['\x090A'] = "\x0EAD\x0EB9"; // uu
            dev2Lao['\x090F'] = "\x0EC0\x0EAD"; // e (Sala E + Or O)
            dev2Lao['\x0913'] = "\x0EC2\x0EAD"; // o (Sala O + Or O)

            // velar stops
            dev2Lao['\x0915'] = '\x0E81'; // ka (Ko Kay)
            dev2Lao['\x0916'] = '\x0E82'; // kha (Kho Khay)
            dev2Lao['\x0917'] = '\x0E84'; // ga (Kho Khouay)
            dev2Lao['\x0918'] = '\x0E84'; // gha (Kho Khouay - Lao maps Gh to Kh often in modern types)
            dev2Lao['\x0919'] = '\x0E87'; // nga (Ngo Ngu)
            
            // palatal stops
            dev2Lao['\x091A'] = '\x0E88'; // ca (Jo Jua)
            dev2Lao['\x091B'] = '\x0E29'; // cha (So Seua - used for Pali Cha often, or sometimes So Xang)
            dev2Lao['\x091C'] = '\x0E8A'; // ja (Xo Xang)
            dev2Lao['\x091D'] = '\x0E8A'; // jha (Xo Xang - Lao maps Jh to X)
            dev2Lao['\x091E'] = '\x0E8D'; // ña (Nyo Nyung)

            // retroflex stops (Using standard Lao approximate glyphs for Pali)
            dev2Lao['\x091F'] = '\x0E8F'; // t underdot a (To Patak - rare but in Unicode)
            dev2Lao['\x0920'] = '\x0E90'; // t underdot ha (Tho Than - rare)
            dev2Lao['\x0921'] = '\x0E91'; // d underdot a (Do Montho - rare)
            dev2Lao['\x0922'] = '\x0E92'; // d underdot ha (Tho Phuthao - rare)
            dev2Lao['\x0923'] = '\x0E93'; // n underdot a (No Nen)

            // dental stops
            dev2Lao['\x0924'] = '\x0E95'; // ta (To Ta)
            dev2Lao['\x0925'] = '\x0E96'; // tha (Tho Thung)
            dev2Lao['\x0926'] = '\x0E97'; // da (Tho Thung - Pali D)
            dev2Lao['\x0927'] = '\x0E98'; // dha (Tho Thong)
            dev2Lao['\x0928'] = '\x0E99'; // na (No Nok)

            // labial stops
            dev2Lao['\x092A'] = '\x0E9B'; // pa (Po Pa)
            dev2Lao['\x092B'] = '\x0E9C'; // pha (Pho Pheung)
            dev2Lao['\x092C'] = '\x0E9E'; // ba (Pho Phu - Pali B)
            dev2Lao['\x092D'] = '\x0E9E'; // bha (Pho Phu - Pali Bh)
            dev2Lao['\x092E'] = '\x0EA1'; // ma (Mo Ma)

            // liquids, fricatives, etc.
            dev2Lao['\x092F'] = '\x0EA2'; // ya (Yo Ya)
            dev2Lao['\x0930'] = '\x0EA3'; // ra (Ro Rot - Lo Loot in modern, but Ro for Pali)
            dev2Lao['\x0932'] = '\x0EA5'; // la (Lo Ling)
            dev2Lao['\x0935'] = '\x0EA7'; // va (Vo Vi)
            dev2Lao['\x0938'] = '\x0EAA'; // sa (So Seua)
            dev2Lao['\x0939'] = '\x0EAB'; // ha (Ho Han)
            dev2Lao['\x0933'] = '\x0EAC'; // l underdot a (Lo Lo)

            // dependent vowel signs
            dev2Lao['\x093E'] = '\x0EB2'; // aa
            dev2Lao['\x093F'] = '\x0EB4'; // i
            dev2Lao['\x0940'] = '\x0EB5'; // ii
            dev2Lao['\x0941'] = '\x0EB8'; // u
            dev2Lao['\x0942'] = '\x0EB9'; // uu
            dev2Lao['\x0947'] = '\x0EC0'; // e (Sala E)
            dev2Lao['\x094B'] = '\x0EC2'; // o (Sala O)

            // Virama (Lao uses Pali Virama - often represented by 0EBA)
            dev2Lao['\x094D'] = '\x0EBA'; 

            // numerals (Lao Digits)
            dev2Lao['\x0966'] = '\x0ED0';
            dev2Lao['\x0967'] = '\x0ED1';
            dev2Lao['\x0968'] = '\x0ED2';
            dev2Lao['\x0969'] = '\x0ED3';
            dev2Lao['\x096A'] = '\x0ED4';
            dev2Lao['\x096B'] = '\x0ED5';
            dev2Lao['\x096C'] = '\x0ED6';
            dev2Lao['\x096D'] = '\x0ED7';
            dev2Lao['\x096E'] = '\x0ED8';
            dev2Lao['\x096F'] = '\x0ED9';

            // other
            dev2Lao['\x0970'] = '.'; // Dev. abbreviation sign
            dev2Lao['\x200C'] = ""; // ZWNJ (remove)
            dev2Lao['\x200D'] = ""; // ZWJ (remove)
        }

        public string InputFilePath
        {
            get { return inputFilePath; }
            set { inputFilePath = value; }
        }
        private string inputFilePath;

        public string OutputFilePath
        {
            get { return outputFilePath; }
            set { outputFilePath = value; }
        }
        private string outputFilePath;


        public void Convert()
        {
            StreamReader sr = new StreamReader(InputFilePath);
            string devStr = sr.ReadToEnd();
            sr.Close();

            // change name of stylesheet for Lao
            devStr = devStr.Replace("tipitaka-deva.xsl", "tipitaka-lao.xsl");

            string str = Convert(devStr);

            str = ConvertDandas(str);
            str = CleanupPunctuation(str);

            StreamWriter sw = new StreamWriter(OutputFilePath, false, Encoding.Unicode);
            sw.Write(str);
            sw.Flush();
            sw.Close();
        }

        public string Convert(string devStr)
        {
            // first remove all the ZWJs
            devStr = devStr.Replace("\x200D", "");

            // pre-processing step for Lao: put the E vowel before its consonants
            // Regex for Lao is same logic as Thai (Left-side vowels)
            devStr = Regex.Replace(devStr, "([\x0915-\x0939])\x0947", "\x0EC0$1");

            // pre-processing step for Lao: put the O vowel before its consonants
            devStr = Regex.Replace(devStr, "([\x0915-\x0939])\x094B", "\x0EC2$1");

            StringBuilder sb = new StringBuilder();
            foreach (char c in devStr.ToCharArray())
            {
                if (dev2Lao.ContainsKey(c))
                    sb.Append(dev2Lao[c]);
                else
                    sb.Append(c);
            }

            return sb.ToString();
        }

        public string ConvertDandas(string str)
        {
            // in gathas, single dandas convert to semicolon, double to period
            str = Regex.Replace(str, "<p rend=\"gatha[a-z0-9]*\">.+?</p>",
                new MatchEvaluator(this.ConvertGathaDandas));

            // remove double dandas around namo tassa
            str = Regex.Replace(str, "<p rend=\"centre\">.+?</p>",
                new MatchEvaluator(this.ConvertNamoTassaDandas));

            // convert all others to Lao Ellipsis (Paiyannoi equivalent: \x0EAF)
            str = str.Replace("\x0964", "\x0EAF");
            str = str.Replace("\x0965", "\x0EAF");
            return str;
        }

        public string ConvertGathaDandas(Match m)
        {
            string str = m.Value;
            str = str.Replace("\x0964", ";");
            str = str.Replace("\x0965", "\x0EAF");
            return str;
        }

        public string ConvertNamoTassaDandas(Match m)
        {
            string str = m.Value;
            return str.Replace("\x0965", "\x0EAF");
        }

        public string CleanupPunctuation(string str)
        {
            str = str.Replace(" ,", ",");
            str = str.Replace(" ?", "?");
            str = str.Replace(" !", "!");
            str = str.Replace(" ;", ";");
            str = str.Replace(" .", ".");
            return str;
        }
    }
}