1 /* 2 * Copyright (c) 2013, Oracle and/or its affiliates. All rights reserved. 3 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. 4 * 5 * This code is free software; you can redistribute it and/or modify it 6 * under the terms of the GNU General Public License version 2 only, as 7 * published by the Free Software Foundation. 8 * 9 * This code is distributed in the hope that it will be useful, but WITHOUT 10 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or 11 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License 12 * version 2 for more details (a copy is included in the LICENSE file that 13 * accompanied this code). 14 * 15 * You should have received a copy of the GNU General Public License version 16 * 2 along with this work; if not, write to the Free Software Foundation, 17 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. 18 * 19 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA 20 * or visit www.oracle.com if you need additional information or have any 21 * questions. 22 */ 23 24 25 package tidystats; 26 27 import java.io.IOException; 28 import java.nio.charset.Charset; 29 import java.nio.file.FileSystem; 30 import java.nio.file.FileSystems; 31 import java.nio.file.Files; 32 import java.nio.file.Path; 33 import java.util.ArrayList; 34 import java.util.Comparator; 35 import java.util.HashMap; 36 import java.util.List; 37 import java.util.Map; 38 import java.util.Set; 39 import java.util.TreeMap; 40 import java.util.TreeSet; 41 import java.util.regex.Matcher; 42 import java.util.regex.Pattern; 43 44 /** 45 * Generate statistics from the files generated by tidy.sh. 46 * 47 * <p>The tidy.sh script is used to run tidy on all the HTML files 48 * in a directory, creating files in a new directory, and for each 49 * HTML file, it writes the console output from tidy into a file 50 * beside the fixed up file, with an additional .tidy extension. 51 * 52 * <p>This program will scan a directory for *.tidy files and 53 * analyze the messages reported by tidy, in order to generate a 54 * report with statistics on the various messages that were 55 * reported by tidy. 56 * 57 * <p>Typical usage: 58 * <pre> 59 * $ bash /path/to/tidy.sh /path/to/htmldir 60 * $ javac -d /path/to/classes /path/to/Main.java 61 * $ java -cp /path/to/classes tidystats.Main /path/to/htmldir.tidy 62 * </pre> 63 * 64 * <p>Internally, the program works by matching lines in the *.tidy 65 * files against a series of regular expressions that are used to 66 * categorize the messages. The set of regular expressions was 67 * empirically determined by running the program on the output from 68 * running tidy.sh on all the generated JDK documentation. It is 69 * possible that tidy may generate more/different messages on other 70 * doc sets, in which case, the set of regexes in the program should 71 * be updated. 72 */ 73 public class Main { main(String... args)74 public static void main(String... args) throws IOException { 75 new Main().run(args); 76 } 77 run(String... args)78 void run(String... args) throws IOException { 79 FileSystem fs = FileSystems.getDefault(); 80 List<Path> paths = new ArrayList<>(); 81 82 int i; 83 for (i = 0; i < args.length; i++) { 84 String arg = args[i]; 85 if (arg.startsWith("-")) 86 throw new IllegalArgumentException(arg); 87 else 88 break; 89 } 90 91 for ( ; i < args.length; i++) { 92 Path p = fs.getPath(args[i]); 93 paths.add(p); 94 } 95 96 for (Path p: paths) { 97 scan(p); 98 } 99 100 print("%6d files read", files); 101 print("%6d files had no errors or warnings", ok); 102 print("%6d files reported \"Not all warnings/errors were shown.\"", overflow); 103 print("%6d errors found", errs); 104 print("%6d warnings found", warns); 105 print("%6d recommendations to use CSS", css); 106 print(""); 107 108 Map<Integer, Set<String>> sortedCounts = new TreeMap<>( 109 new Comparator<Integer>() { 110 @Override 111 public int compare(Integer o1, Integer o2) { 112 return o2.compareTo(o1); 113 } 114 }); 115 116 for (Map.Entry<Pattern, Integer> e: counts.entrySet()) { 117 Pattern p = e.getKey(); 118 Integer n = e.getValue(); 119 Set<String> set = sortedCounts.get(n); 120 if (set == null) 121 sortedCounts.put(n, (set = new TreeSet<>())); 122 set.add(p.toString()); 123 } 124 125 for (Map.Entry<Integer, Set<String>> e: sortedCounts.entrySet()) { 126 for (String p: e.getValue()) { 127 if (p.startsWith(".*")) p = p.substring(2); 128 print("%6d: %s", e.getKey(), p); 129 } 130 } 131 } 132 scan(Path p)133 void scan(Path p) throws IOException { 134 if (Files.isDirectory(p)) { 135 for (Path c: Files.newDirectoryStream(p)) { 136 scan(c); 137 } 138 } else if (isTidyFile(p)) { 139 scan(Files.readAllLines(p, Charset.defaultCharset())); 140 } 141 } 142 isTidyFile(Path p)143 boolean isTidyFile(Path p) { 144 return Files.isRegularFile(p) && p.getFileName().toString().endsWith(".tidy"); 145 } 146 scan(List<String> lines)147 void scan(List<String> lines) { 148 Matcher m; 149 files++; 150 for (String line: lines) { 151 if (okPattern.matcher(line).matches()) { 152 ok++; 153 } else if ((m = countPattern.matcher(line)).matches()) { 154 warns += Integer.valueOf(m.group(1)); 155 errs += Integer.valueOf(m.group(2)); 156 if (m.group(3) != null) 157 overflow++; 158 } else if ((m = guardPattern.matcher(line)).matches()) { 159 boolean found = false; 160 for (Pattern p: patterns) { 161 if ((m = p.matcher(line)).matches()) { 162 found = true; 163 count(p); 164 break; 165 } 166 } 167 if (!found) 168 System.err.println("Unrecognized line: " + line); 169 } else if (cssPattern.matcher(line).matches()) { 170 css++; 171 } 172 } 173 } 174 175 Map<Pattern, Integer> counts = new HashMap<>(); count(Pattern p)176 void count(Pattern p) { 177 Integer i = counts.get(p); 178 counts.put(p, (i == null) ? 1 : i + 1); 179 } 180 print(String format, Object... args)181 void print(String format, Object... args) { 182 System.out.println(String.format(format, args)); 183 } 184 185 Pattern okPattern = Pattern.compile("No warnings or errors were found."); 186 Pattern countPattern = Pattern.compile("([0-9]+) warnings, ([0-9]+) errors were found!.*?(Not all warnings/errors were shown.)?"); 187 Pattern cssPattern = Pattern.compile("You are recommended to use CSS.*"); 188 Pattern guardPattern = Pattern.compile("line [0-9]+ column [0-9]+ - (Error|Warning):.*"); 189 190 Pattern[] patterns = { 191 Pattern.compile(".*Error: <.*> is not recognized!"), 192 Pattern.compile(".*Error: missing quote mark for attribute value"), 193 Pattern.compile(".*Warning: <.*> anchor \".*\" already defined"), 194 Pattern.compile(".*Warning: <.*> attribute \".*\" has invalid value \".*\""), 195 Pattern.compile(".*Warning: <.*> attribute \".*\" lacks value"), 196 Pattern.compile(".*Warning: <.*> attribute \".*\" lacks value"), 197 Pattern.compile(".*Warning: <.*> attribute with missing trailing quote mark"), 198 Pattern.compile(".*Warning: <.*> dropping value \".*\" for repeated attribute \".*\""), 199 Pattern.compile(".*Warning: <.*> inserting \".*\" attribute"), 200 Pattern.compile(".*Warning: <.*> is probably intended as </.*>"), 201 Pattern.compile(".*Warning: <.*> isn't allowed in <.*> elements"), 202 Pattern.compile(".*Warning: <.*> lacks \".*\" attribute"), 203 Pattern.compile(".*Warning: <.*> missing '>' for end of tag"), 204 Pattern.compile(".*Warning: <.*> proprietary attribute \".*\""), 205 Pattern.compile(".*Warning: <.*> unexpected or duplicate quote mark"), 206 Pattern.compile(".*Warning: <a> cannot copy name attribute to id"), 207 Pattern.compile(".*Warning: <a> escaping malformed URI reference"), 208 Pattern.compile(".*Warning: <blockquote> proprietary attribute \"pre\""), 209 Pattern.compile(".*Warning: discarding unexpected <.*>"), 210 Pattern.compile(".*Warning: discarding unexpected </.*>"), 211 Pattern.compile(".*Warning: entity \".*\" doesn't end in ';'"), 212 Pattern.compile(".*Warning: inserting implicit <.*>"), 213 Pattern.compile(".*Warning: inserting missing 'title' element"), 214 Pattern.compile(".*Warning: missing <!DOCTYPE> declaration"), 215 Pattern.compile(".*Warning: missing <.*>"), 216 Pattern.compile(".*Warning: missing </.*> before <.*>"), 217 Pattern.compile(".*Warning: nested emphasis <.*>"), 218 Pattern.compile(".*Warning: plain text isn't allowed in <.*> elements"), 219 Pattern.compile(".*Warning: replacing <p> by <br>"), 220 Pattern.compile(".*Warning: replacing invalid numeric character reference .*"), 221 Pattern.compile(".*Warning: replacing unexpected .* by </.*>"), 222 Pattern.compile(".*Warning: trimming empty <.*>"), 223 Pattern.compile(".*Warning: unescaped & or unknown entity \".*\""), 224 Pattern.compile(".*Warning: unescaped & which should be written as &"), 225 Pattern.compile(".*Warning: using <br> in place of <p>") 226 }; 227 228 int files; 229 int ok; 230 int warns; 231 int errs; 232 int css; 233 int overflow; 234 } 235 236