webmozart /
glob
Duplicate code is one of the most pungent code smells. A rule that is often used is to re-structure code once it is duplicated in three or more places.
Common duplication problems, and corresponding solutions are:
| 1 | <?php |
||
| 75 | final class Glob |
||
| 76 | { |
||
| 77 | /** |
||
| 78 | * Flag: Enable escaping of special characters with leading backslashes. |
||
| 79 | */ |
||
| 80 | const ESCAPE = 1; |
||
| 81 | |||
| 82 | /** |
||
| 83 | * Globs the file system paths matching the glob. |
||
| 84 | * |
||
| 85 | * The glob may contain the wildcard "*". This wildcard matches any number |
||
| 86 | * of characters, *including* directory separators. |
||
| 87 | * |
||
| 88 | * ```php |
||
| 89 | * foreach (Glob::glob('/project/**.twig') as $path) { |
||
| 90 | * // do something... |
||
| 91 | * } |
||
| 92 | * ``` |
||
| 93 | * |
||
| 94 | * @param string $glob The canonical glob. The glob should contain forward |
||
| 95 | * slashes as directory separators only. It must not |
||
| 96 | * contain any "." or ".." segments. Use the |
||
| 97 | * "webmozart/path-util" utility to canonicalize globs |
||
| 98 | * prior to calling this method. |
||
| 99 | * @param int $flags A bitwise combination of the flag constants in this |
||
| 100 | * class. |
||
| 101 | * |
||
| 102 | * @return string[] The matching paths. The keys of the array are |
||
| 103 | * incrementing integers. |
||
| 104 | */ |
||
| 105 | 4 | public static function glob($glob, $flags = 0) |
|
| 113 | |||
| 114 | /** |
||
| 115 | * Matches a path against a glob. |
||
| 116 | * |
||
| 117 | * ```php |
||
| 118 | * if (Glob::match('/project/views/index.html.twig', '/project/**.twig')) { |
||
| 119 | * // path matches |
||
| 120 | * } |
||
| 121 | * ``` |
||
| 122 | * |
||
| 123 | * @param string $path The path to match. |
||
| 124 | * @param string $glob The canonical glob. The glob should contain forward |
||
| 125 | * slashes as directory separators only. It must not |
||
| 126 | * contain any "." or ".." segments. Use the |
||
| 127 | * "webmozart/path-util" utility to canonicalize globs |
||
| 128 | * prior to calling this method. |
||
| 129 | * @param int $flags A bitwise combination of the flag constants in |
||
| 130 | * this class. |
||
| 131 | * |
||
| 132 | * @return bool Returns `true` if the path is matched by the glob. |
||
| 133 | */ |
||
| 134 | 12 | public static function match($path, $glob, $flags = 0) |
|
| 150 | |||
| 151 | /** |
||
| 152 | * Filters an array for paths matching a glob. |
||
| 153 | * |
||
| 154 | * The filtered array is returned. This array preserves the keys of the |
||
| 155 | * passed array. |
||
| 156 | * |
||
| 157 | * ```php |
||
| 158 | * $filteredPaths = Glob::filter($paths, '/project/**.twig'); |
||
| 159 | * ``` |
||
| 160 | * |
||
| 161 | * @param string[] $paths A list of paths. |
||
| 162 | * @param string $glob The canonical glob. The glob should contain |
||
| 163 | * forward slashes as directory separators only. It |
||
| 164 | * must not contain any "." or ".." segments. Use the |
||
| 165 | * "webmozart/path-util" utility to canonicalize |
||
| 166 | * globs prior to calling this method. |
||
| 167 | * @param int $flags A bitwise combination of the flag constants in |
||
| 168 | * this class. |
||
| 169 | * |
||
| 170 | * @return string[] The paths matching the glob indexed by their original |
||
| 171 | * keys. |
||
| 172 | */ |
||
| 173 | 5 | public static function filter(array $paths, $glob, $flags = 0) |
|
| 190 | |||
| 191 | /** |
||
| 192 | * Returns the base path of a glob. |
||
| 193 | * |
||
| 194 | * This method returns the most specific directory that contains all files |
||
| 195 | * matched by the glob. If this directory does not exist on the file system, |
||
| 196 | * it's not necessary to execute the glob algorithm. |
||
| 197 | * |
||
| 198 | * More specifically, the "base path" is the longest path trailed by a "/" |
||
| 199 | * on the left of the first wildcard "*". If the glob does not contain |
||
| 200 | * wildcards, the directory name of the glob is returned. |
||
| 201 | * |
||
| 202 | * ```php |
||
| 203 | * Glob::getBasePath('/css/*.css'); |
||
| 204 | * // => /css |
||
| 205 | * |
||
| 206 | * Glob::getBasePath('/css/style.css'); |
||
| 207 | * // => /css |
||
| 208 | * |
||
| 209 | * Glob::getBasePath('/css/st*.css'); |
||
| 210 | * // => /css |
||
| 211 | * |
||
| 212 | * Glob::getBasePath('/*.css'); |
||
| 213 | * // => / |
||
| 214 | * ``` |
||
| 215 | * |
||
| 216 | * @param string $glob The canonical glob. The glob should contain forward |
||
| 217 | * slashes as directory separators only. It must not |
||
| 218 | * contain any "." or ".." segments. Use the |
||
| 219 | * "webmozart/path-util" utility to canonicalize globs |
||
| 220 | * prior to calling this method. |
||
| 221 | * @param int $flags A bitwise combination of the flag constants in this |
||
| 222 | * class. |
||
| 223 | * |
||
| 224 | * @return string The base path of the glob. |
||
| 225 | */ |
||
| 226 | 47 | public static function getBasePath($glob, $flags = 0) |
|
| 250 | |||
| 251 | /** |
||
| 252 | * Converts a glob to a regular expression. |
||
| 253 | * |
||
| 254 | * Use this method if you need to match many paths against a glob: |
||
| 255 | * |
||
| 256 | * ```php |
||
| 257 | * $staticPrefix = Glob::getStaticPrefix('/project/**.twig'); |
||
| 258 | * $regEx = Glob::toRegEx('/project/**.twig'); |
||
| 259 | * |
||
| 260 | * if (0 !== strpos($path, $staticPrefix)) { |
||
| 261 | * // no match |
||
| 262 | * } |
||
| 263 | * |
||
| 264 | * if (!preg_match($regEx, $path)) { |
||
| 265 | * // no match |
||
| 266 | * } |
||
| 267 | * ``` |
||
| 268 | * |
||
| 269 | * You should always test whether a path contains the static prefix of the |
||
| 270 | * glob returned by {@link getStaticPrefix()} to reduce the number of calls |
||
| 271 | * to the expensive {@link preg_match()}. |
||
| 272 | * |
||
| 273 | * @param string $glob The canonical glob. The glob should contain forward |
||
| 274 | * slashes as directory separators only. It must not |
||
| 275 | * contain any "." or ".." segments. Use the |
||
| 276 | * "webmozart/path-util" utility to canonicalize globs |
||
| 277 | * prior to calling this method. |
||
| 278 | * @param int $flags A bitwise combination of the flag constants in this |
||
| 279 | * class. |
||
| 280 | * |
||
| 281 | * @return string The regular expression for matching the glob. |
||
| 282 | */ |
||
| 283 | 77 | public static function toRegEx($glob, $flags = 0) |
|
| 284 | { |
||
| 285 | 77 | View Code Duplication | if (!Path::isAbsolute($glob) && false === strpos($glob, '://')) { |
| 286 | 1 | throw new InvalidArgumentException(sprintf( |
|
| 287 | 1 | 'The glob "%s" is not absolute and not a URI.', |
|
| 288 | $glob |
||
| 289 | )); |
||
| 290 | } |
||
| 291 | |||
| 292 | // From the PHP manual: To specify a literal single quote, escape it |
||
| 293 | // with a backslash (\). To specify a literal backslash, double it (\\). |
||
| 294 | // All other instances of backslash will be treated as a literal backslash. |
||
| 295 | |||
| 296 | // This method does the following replacements: |
||
| 297 | |||
| 298 | // Normal wildcards: "*" => "[^/]*" (regex match any except separator) |
||
| 299 | // Double wildcards: "**" => ".*" (regex match any) |
||
| 300 | // Sets: "{ab,cd}" => "(ab|cd)" (regex group) |
||
| 301 | |||
| 302 | // with flag Glob::ESCAPE: |
||
| 303 | // Escaped wildcards: "\*" => "\*" (regex star) |
||
| 304 | // Escaped backslashes: "\\" => "\\" (regex backslash) |
||
| 305 | |||
| 306 | // Other characters are escaped as usual for regular expressions. |
||
| 307 | |||
| 308 | // Quote regex characters |
||
| 309 | 76 | $quoted = preg_quote($glob, '~'); |
|
| 310 | |||
| 311 | 76 | if ($flags & self::ESCAPE) { |
|
| 312 | 36 | $regEx = self::toRegExEscaped($quoted); |
|
| 313 | } else { |
||
| 314 | 41 | $regEx = self::toRegExNonEscaped($quoted); |
|
| 315 | } |
||
| 316 | |||
| 317 | 76 | return '~^'.$regEx.'$~'; |
|
| 318 | } |
||
| 319 | |||
| 320 | /** |
||
| 321 | * Returns the static prefix of a glob. |
||
| 322 | * |
||
| 323 | * The "static prefix" is the part of the glob up to the first wildcard "*". |
||
| 324 | * If the glob does not contain wildcards, the full glob is returned. |
||
| 325 | * |
||
| 326 | * @param string $glob The canonical glob. The glob should contain forward |
||
| 327 | * slashes as directory separators only. It must not |
||
| 328 | * contain any "." or ".." segments. Use the |
||
| 329 | * "webmozart/path-util" utility to canonicalize globs |
||
| 330 | * prior to calling this method. |
||
| 331 | * @param int $flags A bitwise combination of the flag constants in this |
||
| 332 | * class. |
||
| 333 | * |
||
| 334 | * @return string The static prefix of the glob. |
||
| 335 | */ |
||
| 336 | 96 | public static function getStaticPrefix($glob, $flags = 0) |
|
| 337 | { |
||
| 338 | 96 | View Code Duplication | if (!Path::isAbsolute($glob) && false === strpos($glob, '://')) { |
| 339 | 5 | throw new InvalidArgumentException(sprintf( |
|
| 340 | 5 | 'The glob "%s" is not absolute and not a URI.', |
|
| 341 | $glob |
||
| 342 | )); |
||
| 343 | } |
||
| 344 | |||
| 345 | 91 | $prefix = $glob; |
|
| 346 | |||
| 347 | 91 | if ($flags & self::ESCAPE) { |
|
| 348 | // Read backslashes together with the next (the escaped) character |
||
| 349 | // up to the first non-escaped star/brace |
||
| 350 | 32 | if (preg_match('~^('.Symbol::BACKSLASH.'.|[^'.Symbol::BACKSLASH.Symbol::STAR.Symbol::L_BRACE.Symbol::QUESTION_MARK.Symbol::L_BRACKET.'])*~', $glob, $matches)) { |
|
| 351 | 32 | $prefix = $matches[0]; |
|
| 352 | } |
||
| 353 | |||
| 354 | // Replace escaped characters by their unescaped equivalents |
||
| 355 | 32 | $prefix = str_replace( |
|
| 356 | 32 | array('\\\\', '\\*', '\\{', '\\}', '\\?', '\\[', '\\]', '\\^'), |
|
| 357 | 32 | array('\\', '*', '{', '}', '?', '[', ']', '^'), |
|
| 358 | $prefix |
||
| 359 | ); |
||
| 360 | } else { |
||
| 361 | 61 | $pos1 = strpos($glob, '*'); |
|
| 362 | 61 | $pos2 = strpos($glob, '{'); |
|
| 363 | 61 | $pos3 = strpos($glob, '?'); |
|
| 364 | 61 | $pos4 = strpos($glob, '['); |
|
| 365 | |||
| 366 | $positions = array_filter(array($pos1, $pos2, $pos3, $pos4), function ($v) { |
||
| 367 | 61 | return false !== $v; |
|
| 368 | 61 | }); |
|
| 369 | |||
| 370 | 61 | if (!empty($positions)) { |
|
| 371 | 56 | $prefix = substr($glob, 0, min($positions)); |
|
| 372 | } |
||
| 373 | } |
||
| 374 | |||
| 375 | 91 | return $prefix; |
|
| 376 | } |
||
| 377 | |||
| 378 | /** |
||
| 379 | * Returns whether the glob contains a dynamic part. |
||
| 380 | * |
||
| 381 | * The glob contains a dynamic part if it contains an unescaped "*" or |
||
| 382 | * "{" character. |
||
| 383 | * |
||
| 384 | * @param string $glob The glob to test. |
||
| 385 | * |
||
| 386 | * @return bool Returns `true` if the glob contains a dynamic part and |
||
| 387 | * `false` otherwise. |
||
| 388 | */ |
||
| 389 | 34 | public static function isDynamic($glob) |
|
| 393 | |||
| 394 | private function __construct() |
||
| 397 | |||
| 398 | 41 | private static function toRegExNonEscaped($quoted) |
|
| 399 | { |
||
| 433 | |||
| 434 | 36 | private static function toRegExEscaped($quoted) |
|
| 513 | } |
||
| 514 |
Loading history...