// wikiCode.java 
// -------------------------------------
// part of YACY
//
// (C) 2005, 2006 by Alexander Schier 
//                   Marc Nause, Franz Brausze
//
//
// last change: $LastChangedDate: $ by $LastChangedBy: $
//
// This program is free software; you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation; either version 2 of the License, or
// (at your option) any later version.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
// GNU General Public License for more details.
//
// You should have received a copy of the GNU General Public License
// along with this program; if not, write to the Free Software
// Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA

// Contains contributions from Alexander Schier [AS]
// Franz Brausze [FB] and Marc Nause [MN]

package de.anomic.data.wiki;

import java.io.BufferedReader;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.HashMap;

import de.anomic.htmlFilter.htmlFilterCharacterCoding;
import de.anomic.server.serverCore;

/** This class provides methods to handle texts that have been posted in the yacyWiki or other
  * parts of YaCy that use this class, like the blog or the profile.
  */
public class wikiCode extends abstractWikiParser implements wikiParser {
    
    /* Table properties */
    private static final String[] tps = { "rowspan", "colspan", "vspace", "hspace", "cellspacing", "cellpadding", "border" };
    private static final HashMap/* <String,String[]> */<String, String[]> ps = new HashMap<String, String[]>();
    static {
        Arrays.sort(tps);
        String[] array;
        Arrays.sort(array = new String[] { "void", "above", "below", "hsides", "lhs", "rhs", "vsides", "box", "border" });
        ps.put("frame", array);
        Arrays.sort(array = new String[] { "none", "groups", "rows", "cols", "all" });
        ps.put("rules", array);
        Arrays.sort(array = new String[] { "top", "middle", "bottom", "baseline" });
        ps.put("valign", array);
        Arrays.sort(array = new String[] { "left", "right", "center" });
        ps.put("align", array);
    }
    
    private String numListLevel="";
    private String ListLevel="";
    private String defListLevel="";
    private boolean cellprocessing=false;       //needed for prevention of double-execution of replaceHTML
    private boolean defList = false;            //needed for definition lists
    private boolean escape = false;             //needed for escape
    private boolean escaped = false;            //needed for <pre> not getting in the way
    private boolean newrowstart=false;          //needed for the first row not to be empty
    private boolean nolist = false;             //needed for handling of [= and <pre> in lists
    private boolean preformatted = false;       //needed for preformatted text
    private boolean preformattedSpan = false;   //needed for <pre> and </pre> spanning over several lines
    private boolean replacedHTML = false;       //indicates if method replaceHTML has been used with line already
    private boolean table = false;              //needed for tables, because they reach over several lines
    private int preindented = 0;                //needed for indented <pre>s
    private int headlines = 0;                  //number of headlines in page
    private final ArrayList<String> dirElements = new ArrayList<String>();    //list of headlines used to create diectory of page

    /** Constructor of the class wikiCode */
    public wikiCode(String address){
        super(address);
    }
    
    protected String transform(
            final BufferedReader reader,
            final int length) throws IOException {
        final StringBuilder out = new StringBuilder(length);
        String line;
        while ((line = reader.readLine()) != null)
            out.append(transformLine(line)).append(serverCore.CRLF_STRING);
        return out.insert(0, directory()).toString();
    }

    /** This method processes tables in the wiki code.
      * @param a string that might contain parts of a table
      * @return a string with wiki code of parts of table replaced by HTML code for table
      */
      //[FB], changes by [MN]
    private String processTable(String result){
        //some variables that make it easier to change codes for the table
        String line = "";
        final String tableStart = "&#123;&#124;";                 // {|
        final String newLine = "&#124;-";                         // |-
        final String cellDivider = "&#124;&#124;";                // ||
        final String tableEnd = "&#124;&#125;";                   // |}
        final String attribDivider = "&#124;";                    // |
        final int lenTableStart = tableStart.length();
        final int lenCellDivider = cellDivider.length();
        final int lenTableEnd = tableEnd.length();
        final int lenAttribDivider = attribDivider.length();

        if ((result.startsWith(tableStart)) && (!table)) {
            table=true;
            newrowstart=true;
            line="<table";
            if (result.trim().length()>lenTableStart) {
                line+=parseTableProperties(result.substring(lenTableStart).trim()).toString();
            }
            line+=">";
            result=line;
        }
        else if (result.startsWith(newLine) && (table)) {          // new row
            if (!newrowstart) {
                line+="\t</tr>\n";
            } else {
                newrowstart=false;
            }
            line=line+"\t<tr>";
            result=line;
        }
        else if ((result.startsWith(cellDivider)) && (table)) {
            line+="\t\t<td";
            final int cellEnd=(result.indexOf(cellDivider,lenCellDivider)>0)?(result.indexOf(cellDivider,lenCellDivider)):(result.length());
            int propEnd = result.indexOf(attribDivider,lenCellDivider);
            final int occImage = result.indexOf("[[Image:",lenCellDivider);
            final int occEscape = result.indexOf("[=",lenCellDivider);
            //If resultOf("[[Image:") is less than propEnd, that means that there is no
            //property for this cell, only an image. Without this, YaCy could get confused
            //by a | in [[Image:picture.png|alt-text]] or [[Image:picture.png|alt-text]]
            //Same for [= (part of [= =])
            if((propEnd > lenCellDivider)
            &&((occImage > propEnd)||( occImage < 0))
            &&((occEscape> propEnd)||( occEscape < 0))
            ){
                propEnd = result.indexOf(attribDivider,lenCellDivider)+lenAttribDivider;
            }
            else {
                propEnd = cellEnd;
            }
            // both point at same place => new line
            if (propEnd==cellEnd) {
                propEnd=lenCellDivider;
            }
            else {
                line+=parseTableProperties(result.substring(lenCellDivider,propEnd-lenAttribDivider).trim()).toString();
            }
            // quick&dirty fix [MN]
            if(propEnd > cellEnd){
                propEnd = lenCellDivider;
            }
            table=false; cellprocessing=true;
            line+=">"+processTable(result.substring(propEnd,cellEnd).trim())+"</td>";
            table=true; cellprocessing=false;
            if (cellEnd<result.length()) {
                line+="\n"+processTable(result.substring(cellEnd));
            }
            result=line;
        }
        else if (result.startsWith(tableEnd) && (table)) {          // Table end
            table=false;
            line+="\t</tr>\n</table>"+result.substring(lenTableEnd);
            result=line;
        }
        return result;
    }
    
    // contributed by [MN], changes by [FB]
    /** This method takes possible table properties and tests if they are valid.
      * Valid in this case means if they are a property for the table, tr or td
      * tag as stated in the HTML Pocket Reference by Jennifer Niederst (1st edition)
      * The method is important to avoid XSS attacks on the wiki via table properties.
      * @param properties A string that may contain several table properties and/or junk.
      * @return A string that only contains table properties.
      */
    private static StringBuilder parseTableProperties(final String properties) {
        final String[] values = properties.replaceAll("&quot;", "").split("[= ]");     //splitting the string at = and blanks
        final StringBuilder sb = new StringBuilder(properties.length());
        String key, value;
        String[] posVals;
        final int numberofvalues = values.length;
        for (int i=0; i<numberofvalues; i++) {
            key = values[i].trim();
            if (key.equals("nowrap")) {
                addPair("nowrap", "nowrap", sb);
            } else if (i + 1 < numberofvalues) {
                value = values[++i].trim();
                if (
                        (key.equals("summary")) ||
                        (key.equals("bgcolor") && value.matches("#{0,1}[0-9a-fA-F]{1,6}|[a-zA-Z]{3,}")) ||
                        ((key.equals("width") || key.equals("height")) && value.matches("\\d+%{0,1}")) ||
                        ((posVals = ps.get(key)) != null && Arrays.binarySearch(posVals, value) >= 0) ||
                        (Arrays.binarySearch(tps, key) >= 0 && value.matches("\\d+"))
                ) {
                    addPair(key, value, sb);
                }
            }
        }
        return sb;
    }
    
    private static StringBuilder addPair(final String key, final String value, final StringBuilder sb) {
        return sb.append(" ").append(key).append("=\"").append(value).append("\"");
    }

    /** This method processes ordered lists.
      */
    private String orderedList(String result){
        if(!nolist){    //lists only get processed if not forbidden (see code for [= and <pre>). [MN]
            int p0 = 0;
            int p1 = 0;
            //# sorted Lists contributed by [AS]
            //## Sublist
            if(result.startsWith(numListLevel + "#")){ //more #
                p0 = result.indexOf(numListLevel);
                p1 = result.length();
                result = "<ol>" + serverCore.CRLF_STRING +
                    "<li>" +
                    result.substring(numListLevel.length() + 1, p1) +
                    "</li>";
                numListLevel += "#";
            }else if(numListLevel.length() > 0 && result.startsWith(numListLevel)){ //equal number of #
                p0 = result.indexOf(numListLevel);
                p1 = result.length();
                result = "<li>" +
                    result.substring(numListLevel.length(), p1) +
                    "</li>";
            }else if(numListLevel.length() > 0){ //less #
                int i = numListLevel.length();
                String tmp = "";

                while(! result.startsWith(numListLevel.substring(0,i)) ){
                    tmp += "</ol>";
                    i--;
                }
                numListLevel = numListLevel.substring(0,i);
                p0 = numListLevel.length();
                p1 = result.length();

                if(numListLevel.length() > 0){
                    result = tmp +
                         "<li>" +
                        result.substring(p0, p1) +
                        "</li>";
                }else{
                    result = tmp + result.substring(p0, p1);
                }
            }
            // end contrib [AS]
        }
        return result;
    }

    /** This method processes unordered lists.
      */
    //contributed by [AS] put into it's own method by [MN]
    private String unorderedList(String result){
        if(!nolist){    //lists only get processed if not forbidden (see code for [= and <pre>). [MN]
            int p0 = 0;
            int p1 = 0;
            //contributed by [AS]
            if(result.startsWith(ListLevel + "*")){ //more stars
               p0 = result.indexOf(ListLevel);
                p1 = result.length();
                result = "<ul>" + serverCore.CRLF_STRING +
                    "<li>" +
                result.substring(ListLevel.length() + 1, p1) +
                    "</li>";
                ListLevel += "*";
            }else if(ListLevel.length() > 0 && result.startsWith(ListLevel)){ //equal number of stars
                p0 = result.indexOf(ListLevel);
                p1 = result.length();
                result = "<li>" +
                    result.substring(ListLevel.length(), p1) +
                    "</li>";
            }else if(ListLevel.length() > 0){ //less stars
                int i = ListLevel.length();
                String tmp = "";

                while(! result.startsWith(ListLevel.substring(0,i))){
                    tmp += "</ul>";
                    i--;
                }
                ListLevel = ListLevel.substring(0,i);
                p0 = ListLevel.length();
                p1 = result.length();

                if(ListLevel.length() > 0){
                    result = tmp +
                        "<li>" +
                        result.substring(p0, p1) +
                        "</li>";
                }else{
                    result = tmp + result.substring(p0, p1);
                }
            }
            //end contrib [AS]
        }
        return result;
    }

    /** This method processes definition lists.
      */
    //contributed by [MN] based on unordered list code by [AS]
    private String definitionList(String result){
        if(!nolist){    //lists only get processed if not forbidden (see code for [= and <pre>). [MN]
            int p0 = 0;
            int p1 = 0;
            if(result.startsWith(defListLevel + ";")){ //more semicolons
                String dt = ""; 
                String dd = "";
                p0 = result.indexOf(defListLevel);
                p1 = result.length();
                final String resultCopy = result.substring(defListLevel.length() + 1, p1);
                if((p0 = resultCopy.indexOf(":")) > 0){
                    dt = resultCopy.substring(0,p0);
                    dd = resultCopy.substring(p0+1);
                    result = "<dl>" + "<dt>" + dt + "</dt>" + "<dd>" + dd;
                    defList = true;
                }
                defListLevel += ";";
            }else if(defListLevel.length() > 0 && result.startsWith(defListLevel)){ //equal number of semicolons
                String dt = ""; 
                String dd = "";
                p0 = result.indexOf(defListLevel);
                p1 = result.length();
                final String resultCopy = result.substring(defListLevel.length(), p1);
                if((p0 = resultCopy.indexOf(":")) > 0){
                    dt = resultCopy.substring(0,p0);
                    dd = resultCopy.substring(p0+1);
                    result = "<dt>" + dt + "</dt>" + "<dd>" + dd;
                    defList = true;
                }
            }else if(defListLevel.length() > 0){ //less semicolons
                String dt = ""; 
                String dd = "";
                int i = defListLevel.length();
                String tmp = "";
                while(! result.startsWith(defListLevel.substring(0,i)) ){
                    tmp += "</dd></dl>";
                    i--;
                }
                defListLevel = defListLevel.substring(0,i);
                p0 = defListLevel.length();
                p1 = result.length();
                if(defListLevel.length() > 0){
                    final String resultCopy = result.substring(p0, p1);
                    if((p0 = resultCopy.indexOf(":")) > 0){
                        dt = resultCopy.substring(0,p0);
                        dd = resultCopy.substring(p0+1);
                        result = tmp + "<dt>" + dt + "</dt>" + "<dd>" + dd;
                        defList = true;
                    }
                }else{
                    result = tmp + result.substring(p0, p1);
                }
            }
        }
        return result; 
    }

    /** This method processes links and images.
      */
    //contributed by [AS] except where stated otherwise
    private String linksAndImages(String result) {

        // create links
        String kl, kv, alt, align;
        int p;
        int p0 = 0;
        int p1 = 0;
        // internal links and images
        while ((p0 = result.indexOf("[[")) >= 0) {
            p1 = result.indexOf("]]", p0 + 2);
            if (p1 <= p0) break;
            kl = result.substring(p0 + 2, p1);

            // this is the part of the code that's responsible for images
            // contributed by [MN]
            if (kl.startsWith("Image:")) {
                alt = "";
                align = "";
                kv = "";
                kl = kl.substring(6);

                // are there any arguments for the image?
                if ((p = kl.indexOf("&#124;")) > 0) {
                    kv = kl.substring(p + 6);
                    kl = kl.substring(0, p);
                    // if there are 2 arguments, write them into ALIGN and ALT
                    if ((p = kv.indexOf("&#124;")) > 0) {
                        align = kv.substring(0, p);
                        //checking validity of value for align. Only non browser specific
                        //values get supported. Not supported: absmiddle, baseline, texttop
                        if ((align.equals("bottom"))||
                            (align.equals("center"))||
                            (align.equals("left"))||
                            (align.equals("middle"))||
                            (align.equals("right"))||
                            (align.equals("top")))
                        {
                            align = " align=\"" + align + "\"";
                        }
                        else align = "";
                        alt = " alt=\"" + kv.substring(p + 6) + "\"";
                    }
                    // if there is just one, put it into ALT
                    else
                        alt = " alt=\"" + kv + "\"";
                }

                // replace incomplete URLs and make them point to http://peerip:port/...
                // with this feature you can access an image in DATA/HTDOCS/share/yacy.gif
                // using the wikicode [[Image:share/yacy.gif]]
                // or an image DATA/HTDOCS/grafics/kaskelix.jpg with [[Image:grafics/kaskelix.jpg]]
                // you are free to use other sub-paths of DATA/HTDOCS
                if (kl.indexOf("://")<1) {
                    kl = "http://" + super.address + "/" + kl;
                }

                result = result.substring(0, p0) + "<img src=\"" + kl + "\"" + align + alt + ">" + result.substring(p1 + 2);
            }
            // end contrib [MN]

            // if it's no image, it might be an internal link
            else {
                if ((p = kl.indexOf("&#124;")) > 0) {
                    kv = kl.substring(p + 6);
                    kl = kl.substring(0, p);
                } else {
                    kv = kl;
                }
                result = result.substring(0, p0) + "<a class=\"known\" href=\"Wiki.html?page=" + kl + "\">" + kv + "</a>" + result.substring(p1 + 2); // oob exception in append() !
            }
        }

        // external links
        while ((p0 = result.indexOf("[")) >= 0) {
            p1 = result.indexOf("]", p0 + 1);
            if (p1 <= p0) break;
            kl = result.substring(p0 + 1, p1);
            if ((p = kl.indexOf(" ")) > 0) {
                kv = kl.substring(p + 1);
                kl = kl.substring(0, p);
            }
            // No text for the link? -> <a href="http://www.url.com/">http://www.url.com/</a>
            else {
                kv = kl;
            }
            // replace incomplete URLs and make them point to http://peerip:port/...
            // with this feature you can access a file at DATA/HTDOCS/share/page.html
            // using the wikicode [share/page.html]
            // or a file DATA/HTDOCS/www/page.html with [www/page.html]
            // you are free to use other sub-paths of DATA/HTDOCS
            if (kl.indexOf("://") < 1) {
                kl = "http://" + super.address + "/" + kl;
            }
        result = result.substring(0, p0) + "<a class=\"extern\" href=\"" + kl + "\">" + kv + "</a>" + result.substring(p1 + 1);
        }
        return result;
    }

    /** This method handles the preformatted tags <pre> </pre> */
    //contributed by [MN]
    private String preformattedTag(String result){
        int p0 = 0;
        int p1 = 0;
        //implementation very similar to escape code (see above)
        //both <pre> and </pre> in the same line
        if(((p0=result.indexOf("&lt;pre&gt;"))>=0)&&((p1=result.indexOf("&lt;/pre&gt;"))>0)&&(!(escaped))){
            if(p0<p1){
                String preformattedText = "<pre style=\"border:dotted;border-width:thin\">"+result.substring(p0+11,p1)+"</pre>";
                preformattedText = preformattedText.replaceAll("!pre!", "!pre!!");
                result = transformLine(result.substring(0,p0).replaceAll("!pre!", "!pre!!")+"!pre!txt!"+result.substring(p1+12).replaceAll("!pre!", "!pre!!"));
                result = result.replaceAll("!pre!txt!", preformattedText);
                result = result.replaceAll("!pre!!", "!pre!");
            }
            //handles cases like <pre><pre> </pre></pre> <pre> </pre> that would cause an exception otherwise
            else{
                preformatted = true;
                final String temp1 = transformLine(result.substring(0,p0-1).replaceAll("!tmp!","!tmp!!")+"!tmp!txt!");
                nolist = true;
                final String temp2 = transformLine(result.substring(p0));
                nolist = false;
                result = temp1.replaceAll("!tmp!txt!",temp2);
                result = result.replaceAll("!tmp!!", "!tmp!");
                preformatted = false;
            }
        }

        //start <pre>
        else if(((p0 = result.indexOf("&lt;pre&gt;"))>=0)&&(!preformattedSpan)&&(!escaped)){
            preformatted = true;    //prevent surplus line breaks
            String bq ="";  //gets filled with <blockquote>s as needed
            String preformattedText = "<pre style=\"border:dotted;border-width:thin\">"+result.substring(p0+11);
            preformattedText = preformattedText.replaceAll("!pre!", "!pre!!");
            //taking care of indented lines
            while(result.substring(preindented,p0).startsWith(":")){
                preindented++;
                bq = bq + "<blockquote>";
            }
            result = transformLine(result.substring(preindented,p0).replaceAll("!pre!", "!pre!!")+"!pre!txt!");
            result = bq + result.replaceAll("!pre!txt!", preformattedText);
            result = result.replaceAll("!pre!!", "!pre!");
            preformattedSpan = true;
        }

        //end </pre>
        else if(((p0 = result.indexOf("&lt;/pre&gt;"))>=0)&&(preformattedSpan)&&(!escaped)){
            preformattedSpan = false;
            String bq = ""; //gets filled with </blockquote>s as needed
            String preformattedText = result.substring(0,p0)+"</pre>";
            preformattedText = preformattedText.replaceAll("!pre!", "!pre!!");
                //taking care of indented lines
            while (preindented > 0){
                bq = bq + "</blockquote>";
                preindented--;
            }
            result = transformLine("!pre!txt!"+result.substring(p0+12).replaceAll("!pre!", "!pre!!"));
            result = result.replaceAll("!pre!txt!", preformattedText) + bq;
            result = result.replaceAll("!pre!!", "!pre!");
            preformatted = false;
        }
        //Getting rid of surplus </pre>
        else if (((p0 = result.indexOf("&lt;/pre&gt;"))>=0)&&(!preformattedSpan)&&(!escaped)){
            while((p0 = result.indexOf("&lt;/pre&gt;"))>=0){
                result = result.substring(0,p0)+result.substring(p0+12);
            }
            result = transformLine(result);
        }
        return result;
    }

    /** This method creates a directory for a wiki page.
      * @return directory of the wiki
      */
    //method contributed by [MN]
    private String directory(){
        String directory = "";
        String element;
        int s = 0;
        int level = 1;
        int level1 = 0;
        int level2 = 0;
        int level3 = 0;
        int doubles = 0;
        String anchorext = "";
        if((s=dirElements.size())>2){
            for(int i=0;i<s;i++){
                element = dirElements.get(i);
                //counting double headlines
                doubles = 0;
                for(int j=0;j<i;j++){
                    if(dirElements.get(j).substring(1).replaceAll(" ","_").replaceAll("[^a-zA-Z0-9_]","").equals(element.substring(1).replaceAll(" ","_").replaceAll("[^a-zA-Z0-9_]",""))){
                        doubles++;
                    }
                }
                //if there are doubles, create anchorextension
                if(doubles>0){
                    anchorext = "_" + (doubles+1);
                }

                if (element.startsWith("3")){
                    if(level<3){
                        level = 3;
                        level3 = 0;
                    }
                    level3++;
                    final String temp = element.substring(1);
                    element=level1+"."+level2+"."+level3+" "+temp;
                    directory = directory + "&nbsp;&nbsp;&nbsp;&nbsp;<a href=\"#"+temp.replaceAll(" ","_").replaceAll("[^a-zA-Z0-9_]","")+anchorext+"\" class=\"WikiTOC\">"+element+"</a><br />\n";
                }
                else if (element.startsWith("2")){
                    if(level==1){
                        level2 = 0;
                        level = 2;
                    }
                    if(level==3){
                        level = 2;
                    }
                    level2++;
                    final String temp = element.substring(1);
                    element=level1+"."+level2+" "+temp;
                    directory = directory + "&nbsp;&nbsp;<a href=\"#"+temp.replaceAll(" ","_").replaceAll("[^a-zA-Z0-9_]","")+anchorext+"\" class=\"WikiTOC\">"+element+"</a><br />\n";
                }
                else if (element.startsWith("1")) {
                    if (level>1) {
                        level = 1;
                        level2 = 0;
                        level3 = 0;
                    }
                    level1++;
                    final String temp = element.substring(1);
                    element=level1+". "+temp;
                    directory = directory + "<a href=\"#"+temp.replaceAll(" ","_").replaceAll("[^a-zA-Z0-9_]","")+anchorext+"\" class=\"WikiTOC\">"+element+"</a><br />\n";
                }
                anchorext="";
            }
            directory = "<table><tr><td><div class=\"WikiTOCBox\">\n" + directory + "</div></td></tr></table>\n";
        }
        // [MN]
        if(!dirElements.isEmpty()){
            dirElements.clear();
            headlines = 0;
        }
        return directory;
    }

    /** Replaces two occurences of a substring in a string by a pair of strings if 
      * that substring occurs twice in the string. This method is not greedy! You'll
      * have to run it in a loop if you want to replace all occurences of the substring.
      * This method provides special treatment for headlines.
      * @param input the string that something is to be replaced in
      * @param pat substring to be replaced
      * @param repl1 string substring gets replaced by on uneven occurences
      * @param repl2 string substring gets replaced by on even occurences
      */
      //[MN]
    private String pairReplace(String input, final String pat, final String repl1, final String repl2){
        String direlem = "";    //string to keep headlines until they get added to List dirElements
        int p0 = 0;
        int p1 = 0;
        final int l = pat.length();
        //replace pattern if a pair of the pattern can be found in the line
        if( ((p0 = input.indexOf(pat)) >= 0) && ((p1 = input.indexOf(pat, p0 + l)) >= 0) ) {
            //extra treatment for headlines
            if((pat.equals("===="))||(pat.equals("==="))||(pat.equals("=="))){
                //add anchor and create headline
                direlem = input.substring(p0 + l, p1);
                //counting double headlines
                int doubles = 0;
                for(int i=0;i<headlines;i++){
                    if(dirElements.size() > i && dirElements.get(i).substring(1).equals(direlem)){
                        doubles++;
                    }
                }
                String anchor = direlem.replaceAll(" ","_").replaceAll("[^a-zA-Z0-9_]",""); //replace blanks with underscores and delete everything thats not a regular character, a number or _
                //if there are doubles, add underscore and number of doubles plus one
                if(doubles>0){
                    anchor = anchor + "_" + (doubles+1);
                }
                input = input.substring(0, p0) + "<a name=\""+anchor+"\"></a>" + repl1 +
                direlem + repl2 + input.substring(p1 + l);
                //add headlines to list of headlines (so TOC can be created)
                if(pat.equals("===="))     dirElements.add("3"+direlem);
                else if(pat.equals("===")) dirElements.add("2"+direlem);
                else if(pat.equals("=="))  dirElements.add("1"+direlem);
                headlines++;
            }
            else{
                input = input.substring(0, p0) +  repl1 +
                (/*direlem =*/ input.substring(p0 + l, p1)) + repl2 +
                input.substring(p1 + l);
            }
        }
        //recursion if a pair of the pattern can still be found in the line
        if( ((p0 = input.indexOf(pat)) >= 0) && (input.indexOf(pat,p0 + l) >= 0) ){
            input = pairReplace(input, pat, repl1, repl2);
        }
        return input;
    }

    /** Replaces wiki tags with HTML tags.
      * @param result a line of text
      * @param switchboard
      * @return the line of text with HTML tags instead of wiki tags
      */
    public String transformLine(String result) {
        //If HTML has not bee replaced yet (can happen if method gets called in recursion), replace now!
        if (!replacedHTML || preformattedSpan){
            result = htmlFilterCharacterCoding.unicode2html(result, true);
            replacedHTML = true;
        }

        //check if line contains preformatted symbols or if we are in a preformatted sequence already.
        if ((result.indexOf("&lt;pre&gt;")>=0)||(result.indexOf("&lt;/pre&gt;")>=0)||(preformattedSpan)){
            result = preformattedTag(result);
        } else {

            //tables first -> wiki-tags in cells can be treated after that
            result = processTable(result);

            // format lines
            if (result.startsWith(" ")) result = "<tt>" + result.substring(1) + "</tt>";
            if (result.startsWith("----")) result = "<hr />";

            // citings contributed by [MN]
            if(result.startsWith(":")){
                String head = "";
                String tail = "";
                while(result.startsWith(":")){
                    head = head + "<blockquote>";
                    tail = tail + "</blockquote>";
                    result = result.substring(1);
                }
                result = head + result + tail;
            }
            // end contrib [MN]	

            // format headers
            result = pairReplace(result,"====","<h4>","</h4>");
            result = pairReplace(result,"===","<h3>","</h3>");
            result = pairReplace(result,"==","<h2>","</h2>");

            result = pairReplace(result,"'''''","<b><i>","</i></b>");
            result = pairReplace(result,"'''","<b>","</b>");
            result = pairReplace(result,"''","<i>","</i>");

            result = unorderedList(result);
            result = orderedList(result);
            result = definitionList(result);

            result = linksAndImages(result);

        }

        if (!preformatted) replacedHTML = false;
        if ((result.endsWith("</li>"))||(defList)||(escape)||(preformatted)||(table)||(cellprocessing)) return result;
        return result + "<br />";
    }
}