package dokumenty.dekompozycja;

import java.io.BufferedReader;
import java.io.File;
import java.io.FileReader;
import java.io.FileWriter;
import java.util.ArrayList;
import java.util.Collections;
import java.util.HashMap;
import java.util.LinkedList;
import java.util.List;
import java.util.Scanner;

import dokumenty.Element;

import pl.wroc.pwr.file.ImageFolderScanner;
import pl.wroc.pwr.image.IRGBImage;
import pl.wroc.pwr.image.io.GenericLoader;
import pl.wroc.pwr.image.io.GenericSaver;
import pl.wroc.pwr.image.model.RGBImage;

public class DekompozycjaDokumentuPWR
{
	public String OCRExec = "tesseract";
	public String TempFolder = "";
	public String ElementExtension = ".result";
	
	public static void main(String[] args)
	{
		System.out.println("XDD (PWR variant) build: 5.10.2012");
		if (args.length != 5)
		{
			System.out.println("Illegal execution arguments!");
			System.out.println("Should be:   java -Xmx1024m -Xms512m -jar XDD.jar [output file] [region extension] [temp folder] [source folder] [file prefix]");
			System.out.println("For example: java -Xmx1024m -Xms512m -jar XDD.jar C:\\NEKST\\Temp\\b001.f001.PWR.xml .expected C:\\NEKST\\Temp\\ C:\\NEKST\\ b001.f001_");
		}
		else
		{	
			DekompozycjaDokumentuPWR XDD = new DekompozycjaDokumentuPWR();
			XDD.TempFolder = args[2];
			XDD.OCRExec = "tesseract " + XDD.TempFolder + "temp.png " + XDD.TempFolder + "temp -l pol";
			XDD.ElementExtension = args[1];
				
			LinkedList<String> FileList = new LinkedList<String>();
			File InFile = new File(args[3]);
			if (InFile.isFile())
			{
				System.out.println("A folder is required for this operation!");
			}
			else if (InFile.isDirectory())
			{
				System.out.print("  [!] ");
				for(File CurrentFile : new ImageFolderScanner().apply(InFile))
					if (CurrentFile.isFile() && CurrentFile.getName().startsWith(args[4]))
						FileList.add(CurrentFile.getAbsolutePath());
			}
			
			Collections.sort(FileList);
			String[] Files = new String[FileList.size()];
			FileList.toArray(Files);
			
			XDD.Decompose(Files, args[0]);
		}
	}
	
	private List<Element> HeaderElements;
	private List<Element> TitleElements;
	private List<Element> ContentElements;
	private List<Element> FooterElements;
	private List<Element> OtherElements;
	public void Decompose(String[] Input, String Output)
	{
		try
		{
			System.out.println("  [?] Files found: " + Input.length);
			LinkedList<String[]> Chunks = new LinkedList<String[]>();
			LinkedList<String[]> TrailingChunks = new LinkedList<String[]>();
			//Build Chunk List
			for(int i = 0; i < Input.length; i++)
			{
				String File = Input[i];
				HeaderElements = new ArrayList<Element>();
				TitleElements = new ArrayList<Element>();
				ContentElements = new ArrayList<Element>();
				FooterElements = new ArrayList<Element>();
				OtherElements = new ArrayList<Element>();
				
				IRGBImage Source = null;
				
				System.out.println("  [!] Currently decomposing: " + File);
				LinkedList<String[]> CurrentChunks = new LinkedList<String[]>();
				
				//Load Files
				System.out.print("    [•] Reading Image... ");
				Source = new GenericLoader(File).apply();
				System.out.println("Ok");
				System.out.print("    [•] Reading Elements... ");
			    BufferedReader In = new BufferedReader(new FileReader(File + ElementExtension));
			    String Line;
			    while ((Line = In.readLine()) != null)
			    {
			       	Element New = Element.ParsujLinie(Line);
			       	if (HeaderClasses.contains(New.pobierzTyp().toLowerCase())) HeaderElements.add(New);
			       	else if (TitleClasses.contains(New.pobierzTyp().toLowerCase())) TitleElements.add(New);
			       	else if (ContentClasses.contains(New.pobierzTyp().toLowerCase())) ContentElements.add(New);
			       	else if (FooterClasses.contains(New.pobierzTyp().toLowerCase())) FooterElements.add(New);
			       	else OtherElements.add(New);
			    }
			    In.close();
			    System.out.println("Ok!");
			    System.out.println("    [!] Elements read...");
				System.out.println("    [?] Header Elements:  " + HeaderElements.size());
				System.out.println("    [?] Title Elements:   " + TitleElements.size());
				System.out.println("    [?] Content Elements: " + ContentElements.size());
				System.out.println("    [?] Footer Elements:  " + FooterElements.size());
				System.out.println("    [?] Other Elements:   " + OtherElements.size());
				
				//Order Chuncks
				TitleElements = SzeregowanieParagrafow.Sort(TitleElements, OtherElements, "header", "line");
				for (Element TitleElement : TitleElements)
				{
					CurrentChunks.add(DecomposeGenericTextObject(TitleElement, Source));
					CurrentChunks.getLast()[0] = "page." + i + "." + CurrentChunks.getLast()[0];
				}
				ContentElements = SzeregowanieParagrafow.Sort(ContentElements, OtherElements, "header", "line");
				for (Element ContentElement : ContentElements)
				{
					CurrentChunks.add(DecomposeGenericTextObject(ContentElement, Source));
					CurrentChunks.getLast()[0] = "page." + i + "." + CurrentChunks.getLast()[0];
				}
				for (Element Caption : OtherElements)
					if (Caption.pobierzTyp().toLowerCase().equals("caption"))
					{
						TrailingChunks.add(DecomposeGenericTextObject(Caption, Source));
						TrailingChunks.getLast()[0] = "page." + i + "." + TrailingChunks.getLast()[0];
					}
					
				//Combine Current Chunk List With Global Chunk List
				for(String[] Chunk : CurrentChunks)
				{
					if (Chunks.size() != 0 && Chunk == CurrentChunks.getFirst())
					{
						String LastChunk = Chunks.getLast()[1].trim();
						if (LastChunk.endsWith(".") || LastChunk.endsWith("?") || LastChunk.endsWith("!"))
						{
							Chunks.add(Chunk);
						}
						else
						{
							String[] OldChunk = Chunks.getLast();
							OldChunk[0] += ":" + Chunk[0];
							OldChunk[1] += " " + Chunk[1];
						}
					}
					else Chunks.add(Chunk);
				}
			}
			//Save Chunk List To File
			//Do Header
			FileWriter Out = new FileWriter(Output);
			Out.write("<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n");
			Out.write("<!DOCTYPE cesAna  SYSTEM 'xcesAnaIPI.dtd'>\n");
			Out.write("<cesAna type=\"pre_morph\" version=\"WROC-1.0\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n");
			Out.write("<chunkList>\n");
			for (String[] Chunk : Chunks)
			{
				Out.write("<chunk type=\"p\" id=\"" + Chunk[0] + "\">" + Chunk[1] + "</chunk>\n");
			}
			for (String[] Chunk : TrailingChunks)
			{
				Out.write("<chunk type=\"caption\" id=\"" + Chunk[0] + "\">" + Chunk[1] + "</chunk>\n");
			}
			//Do Footer
			Out.write("</chunkList>\n");
			Out.write("</cesAna>");
			Out.close();
			System.out.println("Decomposition finished...");
			System.out.println();
		}
		catch (Exception Ex)
		{
	    	Ex.printStackTrace();
	    	return;
	    }
		
	}
	
	public final List<String> HeaderClasses = new ArrayList<String>();
	public final List<String> TitleClasses = new ArrayList<String>();
	public final List<String> ContentClasses = new ArrayList<String>();
	public final List<String> FooterClasses = new ArrayList<String>();
	public DekompozycjaDokumentuPWR()
	{
		HeaderClasses.add("pageheader");
		TitleClasses.add("title");
		TitleClasses.add("author");
		TitleClasses.add("abstract");
		ContentClasses.add("paragraph");
		ContentClasses.add("header");
		ContentClasses.add("enumeration");
		FooterClasses.add("pagefooter");
	}
	
	private String[] DecomposeGenericTextObject(Element E, IRGBImage Source)
	{
		String[] Return = new String[2];
		Return[0] = BuildID(E);
		Return[1] = GenericTesseractInvocation(E, Source);
		return Return;
	}
	private String GenericTesseractInvocation(Element E, IRGBImage Source)
	{
		System.out.print("Invoking TESSERACT... ");
		String Return = "";
		
		int MinX = E.pobierzPudelko()[0] - 10;
		if (MinX < 0) MinX = 0;
		int MinY = E.pobierzPudelko()[1] - 10;
		if (MinY < 0) MinY = 0;
		int MaxX = E.pobierzPudelko()[2] + 10;
		if (MaxX >= Source.getXSize()) MaxX = Source.getXSize() - 1;
		int MaxY = E.pobierzPudelko()[3] + 10;
		if (MaxY >= Source.getYSize()) MaxY = Source.getYSize() - 1;
		
		IRGBImage Temp = new RGBImage((MaxX - MinX) + 1, (MaxY - MinY) + 1);
		
		for (int C = 0; C < 3; C++)
			for (int x = 0; x < Temp.getXSize(); x++)
				for (int y = 0; y < Temp.getYSize(); y++)
					Temp.getChannel(C).setValue(x, y, Source.getChannel(C).getValue(x + MinX, y + MinY));
		new GenericSaver(Temp, new File(TempFolder + "temp.png")).apply();
		
		try
		{
			Process OCR = Runtime.getRuntime().exec(OCRExec);
			OCR.waitFor();
			Scanner ReadFile = new Scanner(new FileReader(TempFolder + "temp.txt"));
			while (ReadFile.hasNextLine())
			{
				String CurrentLine = ReadFile.nextLine().trim();
				if (CurrentLine.endsWith("-")) CurrentLine = CurrentLine.substring(0, CurrentLine.length() - 1);
				else CurrentLine += " ";
				Return += CurrentLine;
			}
			System.out.println("Ok!");
		}
		catch (Exception Ex)
		{
			System.out.println("TESSERACT OR IO ERROR!");
			Return = "TESSERACT OR IO ERROR!";
		}
		
		Return.replaceAll("<", "&lt;");
		Return.replaceAll(">", "&gt;");
		Return.replaceAll("&", "&amp;");
		Return.replaceAll("\"", "&quot;");
		Return.replaceAll("'", "&apos;");
		return Return.trim();
	}
	public String BuildID(Element E)
	{
		return E.pobierzTyp().toLowerCase() + "." + E.pobierzPudelko()[0] + "." + E.pobierzPudelko()[1];
	}
}