2 ***********************************************************************
\r
3 * Copyright (C) 2005-2007, International Business Machines *
\r
4 * Corporation and others. All Rights Reserved. *
\r
5 ***********************************************************************
\r
9 package com.ibm.icu.dev.tool.charsetdet.sbcs;
\r
13 import java.util.Arrays;
\r
14 import java.util.ArrayList;
\r
15 import java.util.Collections;
\r
16 import java.util.Iterator;
\r
17 import java.util.List;
\r
19 import com.ibm.icu.impl.Utility;
\r
24 * TODO To change the template for this generated type comment go to
\r
25 * Window - Preferences - Java - Code Style - Code Templates
\r
27 public class StatisticsTool implements NGramParser.NGramParserClient, NGramList.NGramKeyMapper
\r
29 /* TODO Make this usage string more sane. */
\r
30 private static final String usageString =
\r
31 "\nUsage: StatisticsTool [OPTIONS] [FILES]\n\n" +
\r
32 "This program will read in a Unicode text file of text in a particular language\n" +
\r
33 "and compute the statistics needed to detected that language and character set.\n " +
\r
35 "-e specify the target encoding\n" +
\r
36 "-h or -? print this usage text.\n" +
\r
37 "-v also generate statistics for visual order.\n" +
\r
38 "-l only generate statistics for logical order (cancel -v)." +
\r
39 "-c run the checker.\n" +
\r
40 "-t run the encoding test.\n" +
\r
41 "example: com.ibm.icu.dev.tool.charset.StatisticsTool -e 8859-1 Spanish.txt";
\r
43 private static final int BUFFER_SIZE = 1024;
\r
45 private char[] buffer;
\r
46 private int bufIndex;
\r
49 private InputFile inputFile;
\r
51 private NGramList ngrams;
\r
53 private static byte[] allBytes = {
\r
54 (byte) 0x00, (byte) 0x01, (byte) 0x02, (byte) 0x03, (byte) 0x04, (byte) 0x05, (byte) 0x06, (byte) 0x07,
\r
55 (byte) 0x08, (byte) 0x09, (byte) 0x0A, (byte) 0x0B, (byte) 0x0C, (byte) 0x0D, (byte) 0x0E, (byte) 0x0F,
\r
56 (byte) 0x10, (byte) 0x11, (byte) 0x12, (byte) 0x13, (byte) 0x14, (byte) 0x15, (byte) 0x16, (byte) 0x17,
\r
57 (byte) 0x18, (byte) 0x19, (byte) 0x1A, (byte) 0x1B, (byte) 0x1C, (byte) 0x1D, (byte) 0x1E, (byte) 0x1F,
\r
58 (byte) 0x20, (byte) 0x21, (byte) 0x22, (byte) 0x23, (byte) 0x24, (byte) 0x25, (byte) 0x26, (byte) 0x27,
\r
59 (byte) 0x28, (byte) 0x29, (byte) 0x2A, (byte) 0x2B, (byte) 0x2C, (byte) 0x2D, (byte) 0x2E, (byte) 0x2F,
\r
60 (byte) 0x30, (byte) 0x31, (byte) 0x32, (byte) 0x33, (byte) 0x34, (byte) 0x35, (byte) 0x36, (byte) 0x37,
\r
61 (byte) 0x38, (byte) 0x39, (byte) 0x3A, (byte) 0x3B, (byte) 0x3C, (byte) 0x3D, (byte) 0x3E, (byte) 0x3F,
\r
62 (byte) 0x40, (byte) 0x41, (byte) 0x42, (byte) 0x43, (byte) 0x44, (byte) 0x45, (byte) 0x46, (byte) 0x47,
\r
63 (byte) 0x48, (byte) 0x49, (byte) 0x4A, (byte) 0x4B, (byte) 0x4C, (byte) 0x4D, (byte) 0x4E, (byte) 0x4F,
\r
64 (byte) 0x50, (byte) 0x51, (byte) 0x52, (byte) 0x53, (byte) 0x54, (byte) 0x55, (byte) 0x56, (byte) 0x57,
\r
65 (byte) 0x58, (byte) 0x59, (byte) 0x5A, (byte) 0x5B, (byte) 0x5C, (byte) 0x5D, (byte) 0x5E, (byte) 0x5F,
\r
66 (byte) 0x60, (byte) 0x61, (byte) 0x62, (byte) 0x63, (byte) 0x64, (byte) 0x65, (byte) 0x66, (byte) 0x67,
\r
67 (byte) 0x68, (byte) 0x69, (byte) 0x6A, (byte) 0x6B, (byte) 0x6C, (byte) 0x6D, (byte) 0x6E, (byte) 0x6F,
\r
68 (byte) 0x70, (byte) 0x71, (byte) 0x72, (byte) 0x73, (byte) 0x74, (byte) 0x75, (byte) 0x76, (byte) 0x77,
\r
69 (byte) 0x78, (byte) 0x79, (byte) 0x7A, (byte) 0x7B, (byte) 0x7C, (byte) 0x7D, (byte) 0x7E, (byte) 0x7F,
\r
70 (byte) 0x80, (byte) 0x81, (byte) 0x82, (byte) 0x83, (byte) 0x84, (byte) 0x85, (byte) 0x86, (byte) 0x87,
\r
71 (byte) 0x88, (byte) 0x89, (byte) 0x8A, (byte) 0x8B, (byte) 0x8C, (byte) 0x8D, (byte) 0x8E, (byte) 0x8F,
\r
72 (byte) 0x90, (byte) 0x91, (byte) 0x92, (byte) 0x93, (byte) 0x94, (byte) 0x95, (byte) 0x96, (byte) 0x97,
\r
73 (byte) 0x98, (byte) 0x99, (byte) 0x9A, (byte) 0x9B, (byte) 0x9C, (byte) 0x9D, (byte) 0x9E, (byte) 0x9F,
\r
74 (byte) 0xA0, (byte) 0xA1, (byte) 0xA2, (byte) 0xA3, (byte) 0xA4, (byte) 0xA5, (byte) 0xA6, (byte) 0xA7,
\r
75 (byte) 0xA8, (byte) 0xA9, (byte) 0xAA, (byte) 0xAB, (byte) 0xAC, (byte) 0xAD, (byte) 0xAE, (byte) 0xAF,
\r
76 (byte) 0xB0, (byte) 0xB1, (byte) 0xB2, (byte) 0xB3, (byte) 0xB4, (byte) 0xB5, (byte) 0xB6, (byte) 0xB7,
\r
77 (byte) 0xB8, (byte) 0xB9, (byte) 0xBA, (byte) 0xBB, (byte) 0xBC, (byte) 0xBD, (byte) 0xBE, (byte) 0xBF,
\r
78 (byte) 0xC0, (byte) 0xC1, (byte) 0xC2, (byte) 0xC3, (byte) 0xC4, (byte) 0xC5, (byte) 0xC6, (byte) 0xC7,
\r
79 (byte) 0xC8, (byte) 0xC9, (byte) 0xCA, (byte) 0xCB, (byte) 0xCC, (byte) 0xCD, (byte) 0xCE, (byte) 0xCF,
\r
80 (byte) 0xD0, (byte) 0xD1, (byte) 0xD2, (byte) 0xD3, (byte) 0xD4, (byte) 0xD5, (byte) 0xD6, (byte) 0xD7,
\r
81 (byte) 0xD8, (byte) 0xD9, (byte) 0xDA, (byte) 0xDB, (byte) 0xDC, (byte) 0xDD, (byte) 0xDE, (byte) 0xDF,
\r
82 (byte) 0xE0, (byte) 0xE1, (byte) 0xE2, (byte) 0xE3, (byte) 0xE4, (byte) 0xE5, (byte) 0xE6, (byte) 0xE7,
\r
83 (byte) 0xE8, (byte) 0xE9, (byte) 0xEA, (byte) 0xEB, (byte) 0xEC, (byte) 0xED, (byte) 0xEE, (byte) 0xEF,
\r
84 (byte) 0xF0, (byte) 0xF1, (byte) 0xF2, (byte) 0xF3, (byte) 0xF4, (byte) 0xF5, (byte) 0xF6, (byte) 0xF7,
\r
85 (byte) 0xF8, (byte) 0xF9, (byte) 0xFA, (byte) 0xFB, (byte) 0xFC, (byte) 0xFD, (byte) 0xFE, (byte) 0xFF
\r
91 public StatisticsTool()
\r
93 buffer = new char[BUFFER_SIZE];
\r
100 private static void usage()
\r
102 System.out.println(usageString);
\r
105 // private static void exceptionError(Exception e)
\r
107 // System.err.println("ioError: " + e.toString());
\r
110 private int nextBuffer(InputFile inputFileArg)
\r
114 return inputFileArg.read(buffer);
\r
117 public char nextChar()
\r
119 if (bufIndex >= bufMax) {
\r
120 bufMax = nextBuffer(inputFile);
\r
127 return buffer[bufIndex++];
\r
130 public void handleNGram(String key)
\r
135 public Object mapKey(String key)
\r
140 private NGramList dumpNGrams()
\r
142 String filename = inputFile.getPath();
\r
143 int extension = filename.lastIndexOf(".");
\r
144 String outputFileName = filename.substring(0, extension) + ".raw" + filename.substring(extension);
\r
145 PrintStream output;
\r
146 double cumulative = 0;
\r
149 output = new PrintStream(
\r
150 new FileOutputStream(outputFileName), true, "UTF8");
\r
151 } catch (IOException e) {
\r
152 System.out.println("? Could not open " + outputFileName + " for writing.");
\r
156 System.out.println(inputFile.getFilename() + ": " + ngrams.getUniqueNGrams() + "/" + ngrams.getTotalNGrams());
\r
158 ArrayList array = new ArrayList(ngrams.values());
\r
160 Collections.sort(array);
\r
162 NGramList stats = new NGramList(inputFile);
\r
164 int totalNGrams = ngrams.getTotalNGrams();
\r
166 for (Iterator it = array.iterator(); it.hasNext(); count += 1) {
\r
167 NGramList.NGram ngram = (NGramList.NGram) it.next();
\r
168 String value = ngram.getValue();
\r
169 int refCount = ngram.getRefCount();
\r
170 double ratio = (double) refCount / totalNGrams * 100.0;
\r
172 cumulative += ratio;
\r
174 // TODO check should be count < max && cumulative < maxPercent
\r
179 output.println(value + "\t" + refCount + "\t" + ratio + "%\t" + cumulative + "%");
\r
187 private void writeStatistics(ArrayList keyList, boolean visual)
\r
189 String filename = inputFile.getPath();
\r
190 int extension = filename.lastIndexOf(".");
\r
191 String outputFileName = filename.substring(0, extension) + "-" + inputFile.getEncoding() +
\r
192 (visual? "-visual.dat" : ".dat");
\r
193 PrintStream output;
\r
196 output = new PrintStream(
\r
197 new FileOutputStream(outputFileName), true, "ASCII");
\r
198 } catch (IOException e) {
\r
199 System.out.println("? Could not open " + outputFileName + " for writing.");
\r
205 output.print(" private static int[] ngrams = {");
\r
207 for (Iterator it = keyList.iterator(); it.hasNext(); i += 1) {
\r
208 Integer ngram = (Integer) it.next();
\r
211 output.print("\n ");
\r
214 output.print("0x" + Utility.hex(ngram.intValue(), 6) + ", ");
\r
217 output.println("\n };\n");
\r
220 * Generate the byte map
\r
222 char[] unicodes = inputFile.decode(allBytes);
\r
224 for (int b = 0; b < 256; b += 1) {
\r
225 char unicode = unicodes[b];
\r
226 int charClass = NGramParser.getCharClass(unicode);
\r
228 switch (charClass) {
\r
229 case NGramParser.C_LETTER:
\r
230 unicodes[b] = Character.toLowerCase(unicode);
\r
233 case NGramParser.C_PUNCT:
\r
237 case NGramParser.C_IGNORE:
\r
239 unicodes[b] = '\0';
\r
243 byte[] byteMap = inputFile.encode(unicodes);
\r
245 output.print(" private static byte[] byteMap = {");
\r
247 for (int b = 0; b < 256; b += 1) {
\r
249 output.print("\n ");
\r
252 output.print("(byte) 0x" + Utility.hex(byteMap[b] & 0xFF, 2) + ", ");
\r
255 output.println("\n };");
\r
258 public NGramList collectStatistics(InputFile file)
\r
260 if (!file.open()) {
\r
266 NGramParser parser = new NGramParser(this);
\r
268 ngrams = new NGramList(this);
\r
273 NGramList stats = dumpNGrams();
\r
274 ArrayList statKeys = new ArrayList(stats.keys());
\r
276 Collections.sort(statKeys);
\r
277 writeStatistics(statKeys, false);
\r
279 if (inputFile.getVisualOrder()) {
\r
280 ArrayList reversed = new ArrayList(statKeys.size());
\r
282 for (Iterator it = statKeys.iterator(); it.hasNext();) {
\r
283 Integer key = (Integer) it.next();
\r
284 int k = key.intValue();
\r
288 r = (r << 8) | (k & 0xFF);
\r
292 reversed.add(new Integer(r));
\r
295 Collections.sort(reversed);
\r
296 writeStatistics(reversed, true);
\r
302 public static void main(String[] args)
\r
304 List list = Arrays.asList(args);
\r
305 InputFile[] input_files = new InputFile[args.length];
\r
306 int file_count = 0;
\r
307 String encoding = null;
\r
308 boolean run_checker = false;
\r
309 boolean encoding_test = false;
\r
310 boolean visual_order = false;
\r
312 for (Iterator it = list.iterator(); it.hasNext(); /*anything?*/) {
\r
313 String arg = (String) it.next();
\r
315 if (arg.equals("-v")) {
\r
316 visual_order = true;
\r
317 } else if (arg.equals("-l")) {
\r
318 visual_order = false;
\r
319 } else if (arg.equals("-c")) {
\r
320 run_checker = true;
\r
321 } else if (arg.equals("-t")) {
\r
322 encoding_test = true;
\r
323 } else if (arg.equals("-e")) {
\r
324 if (it.hasNext()) {
\r
325 encoding = (String) it.next();
\r
327 System.err.println("Error: missing encoding.");
\r
329 } else if (arg.startsWith("-")) {
\r
330 if (! (arg.equals("-h") || arg.equals("-?"))) {
\r
331 System.err.println("Error: unknown option " + arg);
\r
336 input_files[file_count++] = new InputFile(arg, encoding, visual_order);
\r
340 if(file_count == 0){
\r
341 System.err.println("Error: there are no files to process.");
\r
345 StatisticsTool tool = new StatisticsTool();
\r
346 Checker[] checkers = new Checker[file_count];
\r
348 for(int i = 0; i < file_count; i += 1) {
\r
349 InputFile file = input_files[i];
\r
351 checkers[i] = new Checker(tool.collectStatistics(file), file);
\r
354 System.out.println();
\r
360 for(int c = 0; c < file_count; c += 1) {
\r
361 Checker checker = checkers[c];
\r
363 for(int f = 0; f < file_count; f += 1) {
\r
364 checker.check(input_files[f]);
\r
373 if (encoding_test) {
\r
374 char[] buffer = new char[128];
\r
376 System.out.println("Detection test");
\r
378 for (int f = 0; f < file_count; f += 1) {
\r
379 InputFile file = input_files[f];
\r
380 int[] histogram = new int[file_count];
\r
381 int charCount, misses = 0;
\r
383 System.out.println(file.getFilename() + "(" + file.getEncoding() + "):");
\r
386 for (int c = 0; c < file_count; c += 1) {
\r
387 checkers[c].setMapper(file);
\r
391 // for each checker
\r
392 // call checkBuffer, save score
\r
393 // find highest score, update histogram for that checker
\r
394 // show checker histogram
\r
396 while ((charCount = file.read(buffer)) > 0) {
\r
397 int[] scores = new int[file_count];
\r
398 int bestFit = -1, maxScore = 0;
\r
400 for (int c = 0; c < file_count; c += 1) {
\r
401 scores[c] = checkers[c].checkBuffer(buffer, charCount);
\r
404 for (int c = 0; c < file_count; c += 1) {
\r
405 int score = scores[c];
\r
407 if (score > maxScore) {
\r
413 if (bestFit >= 0) {
\r
414 histogram[bestFit] += 1;
\r
420 for (int c = 0; c < file_count; c += 1) {
\r
421 System.out.println(" " + checkers[c].getLanguage() + ": " + histogram[c]);
\r
425 System.out.println(" NONE: " + misses);
\r
428 System.out.println();
\r