Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
0.00% |
0 / 102 |
|
0.00% |
0 / 9 |
CRAP | |
0.00% |
0 / 1 |
| PgnGameParser | |
0.00% |
0 / 102 |
|
0.00% |
0 / 9 |
1190 | |
0.00% |
0 / 1 |
| __construct | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getParsedData | |
0.00% |
0 / 5 |
|
0.00% |
0 / 1 |
2 | |||
| getMetadata | |
0.00% |
0 / 14 |
|
0.00% |
0 / 1 |
42 | |||
| getMetadataKeyAndValue | |
0.00% |
0 / 9 |
|
0.00% |
0 / 1 |
2 | |||
| mergeAdjacentComments | |
0.00% |
0 / 25 |
|
0.00% |
0 / 1 |
110 | |||
| getMoves | |
0.00% |
0 / 33 |
|
0.00% |
0 / 1 |
110 | |||
| getMovesAndComments | |
0.00% |
0 / 4 |
|
0.00% |
0 / 1 |
6 | |||
| getMovesAndVariationFromString | |
0.00% |
0 / 6 |
|
0.00% |
0 / 1 |
2 | |||
| getMoveString | |
0.00% |
0 / 5 |
|
0.00% |
0 / 1 |
6 | |||
| 1 | <?php |
| 2 | /** |
| 3 | * This file is a part of ChessBrowser. |
| 4 | * |
| 5 | * ChessBrowser is free software: you can redistribute it and/or modify |
| 6 | * it under the terms of the GNU General Public License as published by |
| 7 | * the Free Software Foundation, either version 3 of the License, or |
| 8 | * (at your option) any later version. |
| 9 | * |
| 10 | * This program is distributed in the hope that it will be useful, |
| 11 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 12 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 13 | * GNU General Public License for more details. |
| 14 | * |
| 15 | * You should have received a copy of the GNU General Public License |
| 16 | * along with this program. If not, see <https://www.gnu.org/licenses/>. |
| 17 | * |
| 18 | * This file is a part of PgnParser |
| 19 | * |
| 20 | * PgnParser is free software: you can redistribute it and/or modify |
| 21 | * it under the terms of the GNU Lesser General Public License as published by |
| 22 | * the Free Software Foundation, either version 3 of the License, or |
| 23 | * (at your option) any later version. |
| 24 | * |
| 25 | * This program is distributed in the hope that it will be useful, |
| 26 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 27 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 28 | * GNU Lesser General Public License for more details. |
| 29 | * |
| 30 | * You should have received a copy of the GNU Lesser General Public License |
| 31 | * along with this program. If not, see <https://www.gnu.org/licenses/>. |
| 32 | * |
| 33 | * @file PgnGameParser |
| 34 | * @ingroup ChessBrowser |
| 35 | * @author Alf Magne Kalleland |
| 36 | */ |
| 37 | |
| 38 | namespace MediaWiki\Extension\ChessBrowser\PgnParser; |
| 39 | |
| 40 | class PgnGameParser { |
| 41 | /** @var string */ |
| 42 | private $pgnGame; |
| 43 | /** @var string */ |
| 44 | private $defaultFen = 'rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1'; |
| 45 | /** @var string[] */ |
| 46 | private $specialMetadata = [ |
| 47 | 'event', |
| 48 | 'site', |
| 49 | 'white', |
| 50 | 'black', |
| 51 | 'result', |
| 52 | 'plycount', |
| 53 | 'eco', |
| 54 | 'fen', |
| 55 | 'timecontrol', |
| 56 | 'round', |
| 57 | 'date', |
| 58 | 'annotator', |
| 59 | 'termination' |
| 60 | ]; |
| 61 | |
| 62 | /** |
| 63 | * Set the parser's pgn |
| 64 | * |
| 65 | * @param string $pgnGame |
| 66 | */ |
| 67 | public function __construct( $pgnGame ) { |
| 68 | $this->pgnGame = trim( $pgnGame ); |
| 69 | } |
| 70 | |
| 71 | /** |
| 72 | * Get the parsed data |
| 73 | * |
| 74 | * @return array |
| 75 | */ |
| 76 | public function getParsedData() { |
| 77 | $gameData = $this->getMetadata(); |
| 78 | $moveReferences = $this->getMoves(); |
| 79 | $gameData[ChessJson::MOVE_MOVES] = $moveReferences[0]; |
| 80 | $gameData[ChessJson::MOVE_COMMENT] = $moveReferences; |
| 81 | return $gameData; |
| 82 | } |
| 83 | |
| 84 | /** |
| 85 | * Get the metadata |
| 86 | * |
| 87 | * @return array |
| 88 | */ |
| 89 | private function getMetadata() { |
| 90 | $ret = [ |
| 91 | ChessJson::GAME_METADATA => [] |
| 92 | ]; |
| 93 | // TODO set lastmoves property by reading last 3-4 moves in moves array |
| 94 | $lines = explode( "\n", $this->pgnGame ); |
| 95 | foreach ( $lines as $line ) { |
| 96 | $line = trim( $line ); |
| 97 | if ( substr( $line, 0, 1 ) === '[' && substr( $line, strlen( $line ) - 1, 1 ) === ']' ) { |
| 98 | $metadata = $this->getMetadataKeyAndValue( $line ); |
| 99 | if ( in_array( $metadata['key'], $this->specialMetadata ) ) { |
| 100 | $ret[$metadata['key']] = $metadata['value']; |
| 101 | } else { |
| 102 | $ret[ChessJson::GAME_METADATA][$metadata['key']] = $metadata['value']; |
| 103 | } |
| 104 | } |
| 105 | } |
| 106 | if ( !isset( $ret[ChessJson::FEN] ) ) { |
| 107 | $ret[ChessJson::FEN] = $this->defaultFen; |
| 108 | } |
| 109 | |
| 110 | return $ret; |
| 111 | } |
| 112 | |
| 113 | /** |
| 114 | * Get the metadata key and value from a string |
| 115 | * |
| 116 | * @param string $metadataString |
| 117 | * @return array |
| 118 | */ |
| 119 | private function getMetadataKeyAndValue( $metadataString ) { |
| 120 | $metadataString = preg_replace( "/[\[\]]/s", "", $metadataString ); |
| 121 | $metadataString = str_replace( '"', '', $metadataString ); |
| 122 | $tokens = explode( " ", $metadataString ); |
| 123 | |
| 124 | $key = $tokens[0]; |
| 125 | $value = implode( " ", array_slice( $tokens, 1 ) ); |
| 126 | return [ |
| 127 | 'key' => strtolower( $key ), |
| 128 | 'value' => $value |
| 129 | ]; |
| 130 | } |
| 131 | |
| 132 | /** |
| 133 | * Determine how many elements are part of the comment |
| 134 | * |
| 135 | * getMovesAndComments() Takes the PGN move string and splits it on the |
| 136 | * special characters `{`, `}`, `;`, and `\n`. These characters *sometimes* |
| 137 | * delimit a comment. From the PGN Standard section 5 |
| 138 | * |
| 139 | * > Brace comments do not nest; a left brace character appearing in a brace |
| 140 | * > comment loses its special meaning and is ignored. A semicolon appearing |
| 141 | * > inside of a brace comment loses its special meaning and is ignored. |
| 142 | * > Braces appearing inside of a semicolon comments lose their special meaning |
| 143 | * > and are ignored. |
| 144 | * |
| 145 | * The result is that a single comment might span multiple elements of |
| 146 | * $moveStringParts if it contains characters that lost their special meaning. This |
| 147 | * function implements a context-sensitive sub-parser to determine how far to move |
| 148 | * the main buffer when it runs into a comment start character. |
| 149 | * |
| 150 | * This function is called whenever getMoves() encounters a comment start character, |
| 151 | * so `{` or `;` and receives the whole $moveStringParts array and the main buffer |
| 152 | * position ($bufferPos) to start sub-parsing the comment. The element at that index |
| 153 | * will be `{` or `;` and the first for loop iteration sets the proper context flag. |
| 154 | * These flags are used to determine which characters lose special meaning. When |
| 155 | * the appropriate comment end character for the context is hit, the function returns |
| 156 | * an integer ($idx) specifying how far forward to move the main buffer. |
| 157 | * |
| 158 | * See T363230 |
| 159 | * |
| 160 | * @param array $moveStringParts Output of getMovesAndComments() |
| 161 | * @param int $bufferPos Index of where in $moveStringParts the comment starts |
| 162 | * @return int |
| 163 | */ |
| 164 | private function mergeAdjacentComments( array $moveStringParts, int $bufferPos ): int { |
| 165 | // Context flags |
| 166 | $inBraceComment = false; |
| 167 | $inEOLComment = false; |
| 168 | |
| 169 | $endIdx = count( $moveStringParts ) - $bufferPos; |
| 170 | for ( $idx = 0; $idx < $endIdx; $idx++ ) { |
| 171 | $move = $moveStringParts[ $bufferPos + $idx ]; |
| 172 | |
| 173 | /** |
| 174 | * The element following a comment start character or a '}' without |
| 175 | * its special meaning will always be a comment string, so we can |
| 176 | * save some time and potential parsing bugs by incrementing $idx |
| 177 | * past them and skipping the iteration. |
| 178 | * |
| 179 | * See getMovesAndComments for more info on $moveStringParts |
| 180 | */ |
| 181 | switch ( $move ) { |
| 182 | case '{': |
| 183 | // Set context flag if not in EOL comment context |
| 184 | // '{' in a brace comment context loses its special meaning |
| 185 | if ( !$inEOLComment ) { |
| 186 | $inBraceComment = true; |
| 187 | } |
| 188 | // Skip past following comment string |
| 189 | $idx++; |
| 190 | break; |
| 191 | case ';': |
| 192 | // Set context flag if not in brace comment context |
| 193 | // ';' in brace and EOL comments loses special meaning |
| 194 | if ( !$inBraceComment ) { |
| 195 | $inEOLComment = true; |
| 196 | } |
| 197 | // Skip past following comment string |
| 198 | $idx++; |
| 199 | break; |
| 200 | case '}': |
| 201 | // The first '}' ALWAYS has special meaning under the PGN |
| 202 | // standard and ends a brace comment. |
| 203 | if ( $inBraceComment ) { |
| 204 | // Don't include the '}' itself in the buffer increment |
| 205 | return $idx - 1; |
| 206 | } |
| 207 | // '}' in EOL comment loses special meaning |
| 208 | // Skip past following comment string |
| 209 | $idx++; |
| 210 | break; |
| 211 | case "\n": |
| 212 | // \n ALWAYS ends an EOL comment. |
| 213 | if ( $inEOLComment ) { |
| 214 | // Include the newline in the buffer increment, it gets |
| 215 | // removed later as whitespace |
| 216 | return $idx - 0; |
| 217 | } |
| 218 | break; |
| 219 | } |
| 220 | } |
| 221 | // Reached EOF so return a buffer increment past EOF |
| 222 | return $idx + 1; |
| 223 | } |
| 224 | |
| 225 | /** |
| 226 | * Process tokens in the move string |
| 227 | * |
| 228 | * @return array |
| 229 | */ |
| 230 | private function getMoves() { |
| 231 | $moveBuilder = new MoveBuilder(); |
| 232 | |
| 233 | $moveStringParts = $this->getMovesAndComments(); |
| 234 | $lenMSP = count( $moveStringParts ); |
| 235 | for ( $bufferPos = 0; $bufferPos < $lenMSP; $bufferPos++ ) { |
| 236 | $move = trim( $moveStringParts[$bufferPos] ); |
| 237 | |
| 238 | switch ( $move ) { |
| 239 | case '{': |
| 240 | case ';': |
| 241 | $commentBufferIncrement = $this->mergeAdjacentComments( |
| 242 | $moveStringParts, |
| 243 | $bufferPos |
| 244 | ); |
| 245 | $commentSlice = array_slice( |
| 246 | $moveStringParts, |
| 247 | $bufferPos + 1, |
| 248 | $commentBufferIncrement |
| 249 | ); |
| 250 | $comment = implode( '', $commentSlice ); |
| 251 | if ( $bufferPos == 0 ) { |
| 252 | $moveBuilder->addCommentBeforeFirstMove( $comment ); |
| 253 | } else { |
| 254 | $moveBuilder->addComment( $comment ); |
| 255 | } |
| 256 | $bufferPos += $commentBufferIncrement; |
| 257 | break; |
| 258 | default: |
| 259 | $moves = $this->getMovesAndVariationFromString( $move ); |
| 260 | foreach ( $moves as $move ) { |
| 261 | switch ( $move ) { |
| 262 | case '(': |
| 263 | $moveBuilder->startVariation(); |
| 264 | break; |
| 265 | case ')': |
| 266 | $moveBuilder->endVariation(); |
| 267 | break; |
| 268 | default: |
| 269 | $moveBuilder->addMoves( $move ); |
| 270 | } |
| 271 | } |
| 272 | break; |
| 273 | } |
| 274 | } |
| 275 | |
| 276 | return $moveBuilder->getMoves(); |
| 277 | } |
| 278 | |
| 279 | /** |
| 280 | * Split the move string based on comment indicators |
| 281 | * |
| 282 | * $moveSectionParts is an array split by the PGN special comment characters `{`, `}`, `;`, and `\n`. These |
| 283 | * splitting characters are also included in the array to aid parsing later. The structure for a string like: |
| 284 | * |
| 285 | * ``` |
| 286 | * $inputString = "e4 e5 { King's pawn opening } Nf3 ; Interesting! {Not really}\nNc6 ; A common response"; |
| 287 | * ``` |
| 288 | * |
| 289 | * Would result in the array: |
| 290 | * ``` |
| 291 | * $moveSectionParts = [ |
| 292 | * "e4 e5", |
| 293 | * "{" |
| 294 | * "King's pawn opening", |
| 295 | * "}", |
| 296 | * "Nf3", |
| 297 | * ";", |
| 298 | * "Interesting!", |
| 299 | * "{", |
| 300 | * "Not really", |
| 301 | * "}", |
| 302 | * "", |
| 303 | * "\n", |
| 304 | * "Nc6", |
| 305 | * ";", |
| 306 | * "A common response" |
| 307 | * ]; |
| 308 | * ``` |
| 309 | * |
| 310 | * Notice that even though the split characters `}` and `\n` are adjacent, the split results in an empty string |
| 311 | * being inserted between them. |
| 312 | * |
| 313 | * @return array |
| 314 | */ |
| 315 | private function getMovesAndComments() { |
| 316 | $moveSectionParts = preg_split( "/({|}|;|\n)/s", $this->getMoveString(), 0, PREG_SPLIT_DELIM_CAPTURE ); |
| 317 | if ( !$moveSectionParts[0] ) { |
| 318 | $moveSectionParts = array_slice( $moveSectionParts, 1 ); |
| 319 | } |
| 320 | return $moveSectionParts; |
| 321 | } |
| 322 | |
| 323 | /** |
| 324 | * Get the moves and variations from a string |
| 325 | * |
| 326 | * TODO make static |
| 327 | * |
| 328 | * @param string $string |
| 329 | * @return array |
| 330 | */ |
| 331 | private function getMovesAndVariationFromString( $string ) { |
| 332 | $string = " " . $string; |
| 333 | |
| 334 | $string = preg_replace( "/\d+?\./s", "", $string ); |
| 335 | $string = str_replace( " ..", "", $string ); |
| 336 | $string = str_replace( " ", " ", $string ); |
| 337 | $string = trim( $string ); |
| 338 | |
| 339 | return preg_split( "/(\(|\))/s", $string, 0, PREG_SPLIT_DELIM_CAPTURE ); |
| 340 | } |
| 341 | |
| 342 | /** |
| 343 | * Get a move string |
| 344 | * |
| 345 | * @return string |
| 346 | */ |
| 347 | private function getMoveString() { |
| 348 | $tokens = preg_split( "/\]\n\n/s", $this->pgnGame ); |
| 349 | if ( !isset( $tokens[1] ) ) { |
| 350 | return ""; |
| 351 | } |
| 352 | $gameData = $tokens[1]; |
| 353 | // \n is meaningful so don't trim them |
| 354 | return trim( $gameData, " \r\t\v\x00" ); |
| 355 | } |
| 356 | } |