LFFATE /
ccg
| Total Complexity | 47 |
| Total Lines | 364 |
| Duplicated Lines | 0 % |
| Changes | 1 | ||
| Bugs | 0 | Features | 0 |
Complex classes like Diff often do a lot of different things. To break such a class down, we need to identify a cohesive component within that class. A common approach to find such a component is to look for fields/methods that share the same prefixes, or suffixes.
Once you have determined the fields that belong together, you can apply the Extract Class refactoring. If the component makes sense as a sub-class, Extract Subclass is also a candidate, and is often faster.
While breaking up the class, it is a good idea to analyze how other classes use Diff, and based on these observations, apply Extract Interface, too.
| 1 | <?php |
||
| 21 | class Diff{ |
||
| 22 | |||
| 23 | // define the constants |
||
| 24 | const UNMODIFIED = 0; |
||
| 25 | const DELETED = 1; |
||
| 26 | const INSERTED = 2; |
||
| 27 | |||
| 28 | /* Returns the diff for two strings. The return value is an array, each of |
||
| 29 | * whose values is an array containing two values: a line (or character, if |
||
| 30 | * $compareCharacters is true), and one of the constants DIFF::UNMODIFIED (the |
||
| 31 | * line or character is in both strings), DIFF::DELETED (the line or character |
||
| 32 | * is only in the first string), and DIFF::INSERTED (the line or character is |
||
| 33 | * only in the second string). The parameters are: |
||
| 34 | * |
||
| 35 | * $string1 - the first string |
||
| 36 | * $string2 - the second string |
||
| 37 | * $compareCharacters - true to compare characters, and false to compare |
||
| 38 | * lines; this optional parameter defaults to false |
||
| 39 | */ |
||
| 40 | public static function compare( |
||
| 41 | $string1, $string2, $compareCharacters = false){ |
||
| 42 | |||
| 43 | // initialise the sequences and comparison start and end positions |
||
| 44 | $start = 0; |
||
| 45 | if ($compareCharacters){ |
||
| 46 | $sequence1 = $string1; |
||
| 47 | $sequence2 = $string2; |
||
| 48 | $end1 = strlen($string1) - 1; |
||
| 49 | $end2 = strlen($string2) - 1; |
||
| 50 | }else{ |
||
| 51 | $sequence1 = preg_split('/\R/', $string1); |
||
| 52 | $sequence2 = preg_split('/\R/', $string2); |
||
| 53 | $end1 = count($sequence1) - 1; |
||
|
Bug
introduced
by
|
|||
| 54 | $end2 = count($sequence2) - 1; |
||
| 55 | } |
||
| 56 | |||
| 57 | // skip any common prefix |
||
| 58 | while ($start <= $end1 && $start <= $end2 |
||
| 59 | && $sequence1[$start] == $sequence2[$start]){ |
||
| 60 | $start ++; |
||
| 61 | } |
||
| 62 | |||
| 63 | // skip any common suffix |
||
| 64 | while ($end1 >= $start && $end2 >= $start |
||
| 65 | && $sequence1[$end1] == $sequence2[$end2]){ |
||
| 66 | $end1 --; |
||
| 67 | $end2 --; |
||
| 68 | } |
||
| 69 | |||
| 70 | // compute the table of longest common subsequence lengths |
||
| 71 | $table = self::computeTable($sequence1, $sequence2, $start, $end1, $end2); |
||
| 72 | |||
| 73 | // generate the partial diff |
||
| 74 | $partialDiff = |
||
| 75 | self::generatePartialDiff($table, $sequence1, $sequence2, $start); |
||
| 76 | |||
| 77 | // generate the full diff |
||
| 78 | $diff = array(); |
||
| 79 | for ($index = 0; $index < $start; $index ++){ |
||
| 80 | $diff[] = array($sequence1[$index], self::UNMODIFIED); |
||
| 81 | } |
||
| 82 | while (count($partialDiff) > 0) $diff[] = array_pop($partialDiff); |
||
| 83 | for ($index = $end1 + 1; |
||
| 84 | $index < ($compareCharacters ? strlen($sequence1) : count($sequence1)); |
||
| 85 | $index ++){ |
||
| 86 | $diff[] = array($sequence1[$index], self::UNMODIFIED); |
||
| 87 | } |
||
| 88 | |||
| 89 | // return the diff |
||
| 90 | return $diff; |
||
| 91 | |||
| 92 | } |
||
| 93 | |||
| 94 | /* Returns the diff for two files. The parameters are: |
||
| 95 | * |
||
| 96 | * $file1 - the path to the first file |
||
| 97 | * $file2 - the path to the second file |
||
| 98 | * $compareCharacters - true to compare characters, and false to compare |
||
| 99 | * lines; this optional parameter defaults to false |
||
| 100 | */ |
||
| 101 | public static function compareFiles( |
||
| 109 | |||
| 110 | } |
||
| 111 | |||
| 112 | /* Returns the table of longest common subsequence lengths for the specified |
||
| 113 | * sequences. The parameters are: |
||
| 114 | * |
||
| 115 | * $sequence1 - the first sequence |
||
| 116 | * $sequence2 - the second sequence |
||
| 117 | * $start - the starting index |
||
| 118 | * $end1 - the ending index for the first sequence |
||
| 119 | * $end2 - the ending index for the second sequence |
||
| 120 | */ |
||
| 121 | private static function computeTable( |
||
| 154 | |||
| 155 | } |
||
| 156 | |||
| 157 | /* Returns the partial diff for the specificed sequences, in reverse order. |
||
| 158 | * The parameters are: |
||
| 159 | * |
||
| 160 | * $table - the table returned by the computeTable function |
||
| 161 | * $sequence1 - the first sequence |
||
| 162 | * $sequence2 - the second sequence |
||
| 163 | * $start - the starting index |
||
| 164 | */ |
||
| 165 | private static function generatePartialDiff( |
||
| 166 | $table, $sequence1, $sequence2, $start){ |
||
| 167 | |||
| 168 | // initialise the diff |
||
| 169 | $diff = array(); |
||
| 170 | |||
| 171 | // initialise the indices |
||
| 172 | $index1 = count($table) - 1; |
||
| 173 | $index2 = count($table[0]) - 1; |
||
| 174 | |||
| 175 | // loop until there are no items remaining in either sequence |
||
| 176 | while ($index1 > 0 || $index2 > 0){ |
||
| 177 | |||
| 178 | // check what has happened to the items at these indices |
||
| 179 | if ($index1 > 0 && $index2 > 0 |
||
| 180 | && $sequence1[$index1 + $start - 1] |
||
| 181 | == $sequence2[$index2 + $start - 1]){ |
||
| 182 | |||
| 183 | // update the diff and the indices |
||
| 184 | $diff[] = array($sequence1[$index1 + $start - 1], self::UNMODIFIED); |
||
| 185 | $index1 --; |
||
| 186 | $index2 --; |
||
| 187 | |||
| 188 | }elseif ($index2 > 0 |
||
| 189 | && $table[$index1][$index2] == $table[$index1][$index2 - 1]){ |
||
| 190 | |||
| 191 | // update the diff and the indices |
||
| 192 | $diff[] = array($sequence2[$index2 + $start - 1], self::INSERTED); |
||
| 193 | $index2 --; |
||
| 194 | |||
| 195 | }else{ |
||
| 196 | |||
| 197 | // update the diff and the indices |
||
| 198 | $diff[] = array($sequence1[$index1 + $start - 1], self::DELETED); |
||
| 199 | $index1 --; |
||
| 200 | |||
| 201 | } |
||
| 202 | |||
| 203 | } |
||
| 204 | |||
| 205 | // return the diff |
||
| 206 | return $diff; |
||
| 207 | |||
| 208 | } |
||
| 209 | |||
| 210 | /* Returns a diff as a string, where unmodified lines are prefixed by ' ', |
||
| 211 | * deletions are prefixed by '- ', and insertions are prefixed by '+ '. The |
||
| 212 | * parameters are: |
||
| 213 | * |
||
| 214 | * $diff - the diff array |
||
| 215 | * $separator - the separator between lines; this optional parameter defaults |
||
| 216 | * to "\n" |
||
| 217 | */ |
||
| 218 | public static function toString($diff, $separator = "\n"){ |
||
| 219 | |||
| 220 | // initialise the string |
||
| 221 | $string = ''; |
||
| 222 | |||
| 223 | // loop over the lines in the diff |
||
| 224 | foreach ($diff as $line){ |
||
| 225 | |||
| 226 | // extend the string with the line |
||
| 227 | switch ($line[1]){ |
||
| 228 | case self::UNMODIFIED : $string .= ' ' . $line[0];break; |
||
| 229 | case self::DELETED : $string .= '- ' . $line[0];break; |
||
| 230 | case self::INSERTED : $string .= '+ ' . $line[0];break; |
||
| 231 | } |
||
| 232 | |||
| 233 | // extend the string with the separator |
||
| 234 | $string .= $separator; |
||
| 235 | |||
| 236 | } |
||
| 237 | |||
| 238 | // return the string |
||
| 239 | return $string; |
||
| 240 | |||
| 241 | } |
||
| 242 | |||
| 243 | /* Returns a diff as an HTML string, where unmodified lines are contained |
||
| 244 | * within 'span' elements, deletions are contained within 'del' elements, and |
||
| 245 | * insertions are contained within 'ins' elements. The parameters are: |
||
| 246 | * |
||
| 247 | * $diff - the diff array |
||
| 248 | * $separator - the separator between lines; this optional parameter defaults |
||
| 249 | * to '<br>' |
||
| 250 | */ |
||
| 251 | public static function toHTML($diff, $separator = '<br>'){ |
||
| 252 | |||
| 253 | // initialise the HTML |
||
| 254 | $html = ''; |
||
| 255 | |||
| 256 | // loop over the lines in the diff |
||
| 257 | foreach ($diff as $line){ |
||
| 258 | |||
| 259 | // extend the HTML with the line |
||
| 260 | switch ($line[1]){ |
||
| 261 | case self::UNMODIFIED : $element = 'span'; break; |
||
| 262 | case self::DELETED : $element = 'del'; break; |
||
| 263 | case self::INSERTED : $element = 'ins'; break; |
||
| 264 | } |
||
| 265 | $html .= |
||
| 266 | '<' . $element . '>' |
||
| 267 | . htmlspecialchars($line[0]) |
||
| 268 | . '</' . $element . '>'; |
||
| 269 | |||
| 270 | // extend the HTML with the separator |
||
| 271 | $html .= $separator; |
||
| 272 | |||
| 273 | } |
||
| 274 | |||
| 275 | // return the HTML |
||
| 276 | return $html; |
||
| 277 | |||
| 278 | } |
||
| 279 | |||
| 280 | /* Returns a diff as an HTML table. The parameters are: |
||
| 281 | * |
||
| 282 | * $diff - the diff array |
||
| 283 | * $indentation - indentation to add to every line of the generated HTML; this |
||
| 284 | * optional parameter defaults to '' |
||
| 285 | * $separator - the separator between lines; this optional parameter |
||
| 286 | * defaults to '<br>' |
||
| 287 | */ |
||
| 288 | public static function toTable($diff, $indentation = '', $separator = '<br>'){ |
||
| 289 | |||
| 290 | // initialise the HTML |
||
| 291 | $html = $indentation . "<table class=\"diff\">\n"; |
||
| 292 | |||
| 293 | // loop over the lines in the diff |
||
| 294 | $index = 0; |
||
| 295 | while ($index < count($diff)){ |
||
| 296 | |||
| 297 | // determine the line type |
||
| 298 | switch ($diff[$index][1]){ |
||
| 299 | |||
| 300 | // display the content on the left and right |
||
| 301 | case self::UNMODIFIED: |
||
| 302 | $leftCell = |
||
| 303 | self::getCellContent( |
||
| 304 | $diff, $indentation, $separator, $index, self::UNMODIFIED); |
||
| 305 | $rightCell = $leftCell; |
||
| 306 | break; |
||
| 307 | |||
| 308 | // display the deleted on the left and inserted content on the right |
||
| 309 | case self::DELETED: |
||
| 310 | $leftCell = |
||
| 311 | self::getCellContent( |
||
| 312 | $diff, $indentation, $separator, $index, self::DELETED); |
||
| 313 | $rightCell = |
||
| 314 | self::getCellContent( |
||
| 315 | $diff, $indentation, $separator, $index, self::INSERTED); |
||
| 316 | break; |
||
| 317 | |||
| 318 | // display the inserted content on the right |
||
| 319 | case self::INSERTED: |
||
| 320 | $leftCell = ''; |
||
| 321 | $rightCell = |
||
| 322 | self::getCellContent( |
||
| 323 | $diff, $indentation, $separator, $index, self::INSERTED); |
||
| 324 | break; |
||
| 325 | |||
| 326 | } |
||
| 327 | |||
| 328 | // extend the HTML with the new row |
||
| 329 | $html .= |
||
| 330 | $indentation |
||
| 331 | . " <tr>\n" |
||
| 332 | . $indentation |
||
| 333 | . ' <td class="diff' |
||
| 334 | . ($leftCell == $rightCell |
||
| 335 | ? 'Unmodified' |
||
| 336 | : ($leftCell == '' ? 'Blank' : 'Deleted')) |
||
| 337 | . '">' |
||
| 338 | . $leftCell |
||
| 339 | . "</td>\n" |
||
| 340 | . $indentation |
||
| 341 | . ' <td class="diff' |
||
| 342 | . ($leftCell == $rightCell |
||
| 343 | ? 'Unmodified' |
||
| 344 | : ($rightCell == '' ? 'Blank' : 'Inserted')) |
||
| 345 | . '">' |
||
| 346 | . $rightCell |
||
| 347 | . "</td>\n" |
||
| 348 | . $indentation |
||
| 349 | . " </tr>\n"; |
||
| 350 | |||
| 351 | } |
||
| 352 | |||
| 353 | // return the HTML |
||
| 354 | return $html . $indentation . "</table>\n"; |
||
| 355 | |||
| 356 | } |
||
| 357 | |||
| 358 | /* Returns the content of the cell, for use in the toTable function. The |
||
| 359 | * parameters are: |
||
| 360 | * |
||
| 361 | * $diff - the diff array |
||
| 362 | * $indentation - indentation to add to every line of the generated HTML |
||
| 363 | * $separator - the separator between lines |
||
| 364 | * $index - the current index, passes by reference |
||
| 365 | * $type - the type of line |
||
| 366 | */ |
||
| 367 | private static function getCellContent( |
||
| 385 | |||
| 386 | } |
||
| 387 | |||
| 388 | } |
||
| 389 |
Loading history...