It was late 2007, I was trying to develop the system that could search queries and displays the result from huge Nepali archives. There were two problems:
1. How to get huge amount of Nepali text ?
2. How to make them searchable and search with high relevancy ?
I developed some state machines (a simple, equivalent regular expression) that can able accumulate news archives from the web dumped of many news sources. A sample machine for Himal News Paper is here.
Fig. State machine that accumulates news of Himal Khabar.
The java source representation of the state machine is:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 | /* * To change this template, choose Tools | Templates * and open the template in the editor. */ /** * * @author Santa */ public class Himal { private boolean isNepali(char ch) { if (ch >= 0x900 && ch <= 0x97f) { return (true); } else if (ch == ',' || ch == '?' || ch=='!') { return (true); } else if (ch == ' ') { return (true); } else if (ch == '\n') { return (true); } else { return (false); } } private StringBuffer filterText(StringBuffer s){ StringBuffer result = new StringBuffer(); char ch; int state = 0; for(int i=0;i<s.length();i++){ ch = s.charAt(i); if(ch==',' || ch=='?' || ch=='เค'){ if(state != 0){ result.append(ch); result.append(" "); } state = 0; } else if(ch==' '){ if(state != 0) result.append(ch); state = 0; } else{ state = 1; result.append(ch); } } return(result); } private StringBuffer dfaHimal(StringBuffer buffer){ StringBuffer result = new StringBuffer(); String tmp = ""; int state = 1; char ch; for(int i=0;i<buffer.length();i++){ ch = buffer.charAt(i); switch(state){ case 1: if(ch=='=') state = 2; else state = 1; break; case 2: if(ch=='"') state = 3; else state = 1; break; case 3: if(ch=='"'){ if(tmp.equals("headline")||tmp.equals("articletext")||tmp.equals("intro")) state = 4; else state = 1; tmp = ""; } else{ tmp += ch; state = 3; } break; case 4: if(isNepali(ch)) result.append(ch); if(ch=='<') state = 5; else state = 4; break; case 5: if(ch=='>'){ if(tmp.equals("br")||tmp.equals("/p")||tmp.equals("/P")||tmp.equals("BR")||tmp.equals("/TABLE")){ result.append("\r\n44 "); state = 4; } else if(tmp.equals("/div")){ result.append("\r\n"); state = 1; } else state = 4; } else if(ch=='='){ state = 6; } else{ tmp += ch; state = 5; } break; case 6: if(ch=='"'){ tmp = ""; state = 7; } else state = 4; break; case 7: if(ch=='"'){ if(tmp.equals("headlines")) return(result); else state = 5; } else{ tmp += ch; state = 7; } } } return(result); } public StringBuffer getNewsText(String file){ StringBuffer result = new StringBuffer(); StringBuffer buffer = new StringBuffer(); FileReadWrite obj = new FileReadWrite(); buffer = obj.readFile(file); result = dfaHimal(buffer); return(result); } } |
It was the beginning of the searching system development.

No comments:
Post a Comment