Total Complexity | 106 |
Total Lines | 566 |
Duplicated Lines | 0 % |
Changes | 0 |
Complex classes like ExpressionLexer 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 ExpressionLexer, and based on these observations, apply Extract Interface, too.
1 | <?php |
||
30 | class ExpressionLexer |
||
31 | { |
||
32 | /** |
||
33 | * Suffix for single literals. |
||
34 | * |
||
35 | * @var char |
||
36 | */ |
||
37 | const SINGLE_SUFFIX_LOWER = 'f'; |
||
38 | |||
39 | /** |
||
40 | * Suffix for single literals. |
||
41 | * |
||
42 | * @var char |
||
43 | */ |
||
44 | const SINGLE_SUFFIX_UPPER = 'F'; |
||
45 | |||
46 | /** |
||
47 | * Text being parsed. |
||
48 | * |
||
49 | * @var string |
||
50 | */ |
||
51 | private $text; |
||
52 | |||
53 | /** |
||
54 | * Length of text being parsed. |
||
55 | * |
||
56 | * @var int |
||
57 | */ |
||
58 | private $textLen; |
||
59 | |||
60 | /** |
||
61 | * Position on text being parsed. |
||
62 | * |
||
63 | * @var int |
||
64 | */ |
||
65 | private $textPos; |
||
66 | |||
67 | /** |
||
68 | * Character being processed. |
||
69 | * |
||
70 | * @var string |
||
71 | */ |
||
72 | private $ch; |
||
73 | |||
74 | /** |
||
75 | * ExpressionToken being processed. |
||
76 | * |
||
77 | * @var ExpressionToken |
||
78 | */ |
||
79 | private $token; |
||
80 | |||
81 | /** |
||
82 | * Initialize a new instance of ExpressionLexer. |
||
83 | * |
||
84 | * @param string $expression Expression to parse |
||
85 | * @throws ODataException |
||
86 | */ |
||
87 | public function __construct(string $expression) |
||
94 | } |
||
95 | |||
96 | /** |
||
97 | * Set the text position. |
||
98 | * |
||
99 | * @param int $pos Value to position |
||
100 | */ |
||
101 | private function setTextPos(int $pos): void |
||
102 | { |
||
103 | $this->textPos = $pos; |
||
104 | $nextChar = $this->textPos < $this->textLen ? $this->text[$this->textPos] : '\0'; |
||
105 | assert(2 >= strlen($nextChar)); |
||
106 | $this->ch = $nextChar; |
||
107 | } |
||
108 | |||
109 | /** |
||
110 | * Reads the next token, skipping whitespace as necessary. |
||
111 | * @throws ODataException |
||
112 | */ |
||
113 | public function nextToken(): void |
||
114 | { |
||
115 | while (Char::isWhiteSpace($this->ch)) { |
||
116 | $this->nextChar(); |
||
117 | } |
||
118 | |||
119 | $t = null; |
||
120 | $tokenPos = $this->textPos; |
||
121 | switch ($this->ch) { |
||
122 | case '(': |
||
123 | $this->nextChar(); |
||
124 | $t = ExpressionTokenId::OPENPARAM(); |
||
125 | break; |
||
126 | case ')': |
||
127 | $this->nextChar(); |
||
128 | $t = ExpressionTokenId::CLOSEPARAM(); |
||
129 | break; |
||
130 | case ',': |
||
131 | $this->nextChar(); |
||
132 | $t = ExpressionTokenId::COMMA(); |
||
133 | break; |
||
134 | case '-': |
||
135 | $hasNext = $this->textPos + 1 < $this->textLen; |
||
136 | if ($hasNext && Char::isDigit($this->text[$this->textPos + 1])) { |
||
137 | $this->nextChar(); |
||
138 | $t = $this->parseFromDigit(); |
||
139 | if (self::isNumeric($t)) { |
||
140 | break; |
||
141 | } |
||
142 | } elseif ($hasNext && $this->text[$tokenPos + 1] == 'I') { |
||
143 | $this->nextChar(); |
||
144 | $this->parseIdentifier(); |
||
145 | $currentIdentifier = substr($this->text, $tokenPos + 1, $this->textPos - $tokenPos - 1); |
||
146 | |||
147 | if (self::isInfinityLiteralDouble($currentIdentifier)) { |
||
148 | $t = ExpressionTokenId::DOUBLE_LITERAL(); |
||
149 | break; |
||
150 | } elseif (self::isInfinityLiteralSingle($currentIdentifier)) { |
||
151 | $t = ExpressionTokenId::SINGLE_LITERAL(); |
||
152 | break; |
||
153 | } |
||
154 | |||
155 | // If it looked like '-INF' but wasn't we'll rewind and fall through to a simple '-' token. |
||
156 | } |
||
157 | $this->setTextPos($tokenPos); |
||
158 | $this->nextChar(); |
||
159 | $t = ExpressionTokenId::MINUS(); |
||
160 | break; |
||
161 | case '=': |
||
162 | $this->nextChar(); |
||
163 | $t = ExpressionTokenId::EQUAL(); |
||
164 | break; |
||
165 | case '/': |
||
166 | $this->nextChar(); |
||
167 | $t = ExpressionTokenId::SLASH(); |
||
168 | break; |
||
169 | case '?': |
||
170 | $this->nextChar(); |
||
171 | $t = ExpressionTokenId::QUESTION(); |
||
172 | break; |
||
173 | case '.': |
||
174 | $this->nextChar(); |
||
175 | $t = ExpressionTokenId::DOT(); |
||
176 | break; |
||
177 | case '\'': |
||
178 | $quote = $this->ch; |
||
179 | do { |
||
180 | $this->nextChar(); |
||
181 | while ($this->textPos < $this->textLen && $this->ch != $quote) { |
||
182 | $this->nextChar(); |
||
183 | } |
||
184 | |||
185 | if ($this->textPos == $this->textLen) { |
||
186 | $this->parseError( |
||
187 | Messages::expressionLexerUnterminatedStringLiteral( |
||
188 | $this->textPos, |
||
189 | $this->text |
||
190 | ) |
||
191 | ); |
||
192 | } |
||
193 | |||
194 | $this->nextChar(); |
||
195 | } while ($this->ch == $quote); |
||
196 | $t = ExpressionTokenId::STRING_LITERAL(); |
||
197 | break; |
||
198 | case '*': |
||
199 | $this->nextChar(); |
||
200 | $t = ExpressionTokenId::STAR(); |
||
201 | break; |
||
202 | default: |
||
203 | if (Char::isLetter($this->ch) || $this->ch == '_') { |
||
204 | $this->parseIdentifier(); |
||
205 | $t = ExpressionTokenId::IDENTIFIER(); |
||
206 | break; |
||
207 | } |
||
208 | |||
209 | if (Char::isDigit($this->ch)) { |
||
210 | $t = $this->parseFromDigit(); |
||
211 | break; |
||
212 | } |
||
213 | |||
214 | if ($this->textPos == $this->textLen) { |
||
215 | $t = ExpressionTokenId::END(); |
||
216 | break; |
||
217 | } |
||
218 | |||
219 | $this->parseError( |
||
220 | Messages::expressionLexerInvalidCharacter( |
||
221 | $this->ch, |
||
222 | $this->textPos |
||
223 | ) |
||
224 | ); |
||
225 | } |
||
226 | |||
227 | $this->token->setId($t); |
||
228 | $this->token->Text = substr($this->text, $tokenPos, $this->textPos - $tokenPos); |
||
229 | $this->token->Position = $tokenPos; |
||
230 | |||
231 | // Handle type-prefixed literals such as binary, datetime or guid. |
||
232 | $this->handleTypePrefixedLiterals(); |
||
233 | |||
234 | // Handle keywords. |
||
235 | if ($this->token->getId() == ExpressionTokenId::IDENTIFIER()) { |
||
236 | if (self::isInfinityOrNaNDouble($this->token->Text)) { |
||
237 | $this->token->setId(ExpressionTokenId::DOUBLE_LITERAL()); |
||
238 | } elseif (self::isInfinityOrNanSingle($this->token->Text)) { |
||
239 | $this->token->setId(ExpressionTokenId::SINGLE_LITERAL()); |
||
240 | } elseif ($this->token->Text == ODataConstants::KEYWORD_TRUE |
||
241 | || $this->token->Text == ODataConstants::KEYWORD_FALSE |
||
242 | ) { |
||
243 | $this->token->setId(ExpressionTokenId::BOOLEAN_LITERAL()); |
||
244 | } elseif ($this->token->Text == ODataConstants::KEYWORD_NULL) { |
||
245 | $this->token->setId(ExpressionTokenId::NULL_LITERAL()); |
||
246 | } |
||
247 | } |
||
248 | } |
||
249 | |||
250 | /** |
||
251 | * Advance to next character. |
||
252 | */ |
||
253 | private function nextChar(): void |
||
254 | { |
||
255 | if ($this->textPos < $this->textLen) { |
||
256 | ++$this->textPos; |
||
257 | } |
||
258 | |||
259 | $nextChar = $this->textPos < $this->textLen ? $this->text[$this->textPos] : '\0'; |
||
260 | assert(2 >= strlen($nextChar)); |
||
261 | $this->ch = $nextChar; |
||
262 | } |
||
263 | |||
264 | /** |
||
265 | * Parses a token that starts with a digit. |
||
266 | * |
||
267 | * @throws ODataException |
||
268 | * @return ExpressionTokenId The kind of token recognized |
||
269 | */ |
||
270 | private function parseFromDigit(): ExpressionTokenId |
||
324 | } |
||
325 | |||
326 | /** |
||
327 | * Validate current character is a digit. |
||
328 | * @throws ODataException |
||
329 | */ |
||
330 | private function validateDigit(): void |
||
331 | { |
||
332 | if (!Char::isDigit($this->ch)) { |
||
333 | $this->parseError(Messages::expressionLexerDigitExpected($this->textPos)); |
||
334 | } |
||
335 | } |
||
336 | |||
337 | /** |
||
338 | * Throws parser error. |
||
339 | * |
||
340 | * @param string $message The error message |
||
341 | * |
||
342 | * @throws ODataException |
||
343 | */ |
||
344 | private function parseError(string $message): void |
||
345 | { |
||
346 | throw ODataException::createSyntaxError($message); |
||
347 | } |
||
348 | |||
349 | /** |
||
350 | * Whether the specified token identifier is a numeric literal. |
||
351 | * |
||
352 | * @param ExpressionTokenId $id Token identifier to check |
||
353 | * |
||
354 | * @return bool true if it's a numeric literal; false otherwise |
||
355 | */ |
||
356 | public static function isNumeric(ExpressionTokenId $id): bool |
||
364 | } |
||
365 | |||
366 | /** |
||
367 | * Parses an identifier by advancing the current character. |
||
368 | */ |
||
369 | private function parseIdentifier(): void |
||
370 | { |
||
371 | do { |
||
372 | $this->nextChar(); |
||
373 | } while (Char::isLetterOrDigit($this->ch) || $this->ch == '_'); |
||
374 | } |
||
375 | |||
376 | /** |
||
377 | * Check if the parameter ($text) is INF. |
||
378 | * |
||
379 | * @param string $text Text to look in |
||
380 | * |
||
381 | * @return bool true if match found, false otherwise |
||
382 | */ |
||
383 | private static function isInfinityLiteralDouble(string $text): bool |
||
386 | } |
||
387 | |||
388 | /** |
||
389 | * Checks whether parameter ($text) EQUALS to 'INFf' or 'INFF' at position. |
||
390 | * |
||
391 | * @param string $text Text to look in |
||
392 | * |
||
393 | * @return bool true if the substring is equal using an ordinal comparison; false otherwise |
||
394 | */ |
||
395 | private static function isInfinityLiteralSingle(string $text): bool |
||
396 | { |
||
397 | return strlen($text) == 4 |
||
398 | && ($text[3] == self::SINGLE_SUFFIX_LOWER |
||
399 | || $text[3] == self::SINGLE_SUFFIX_UPPER) |
||
400 | && strncmp($text, ODataConstants::XML_INFINITY_LITERAL, 3) == 0; |
||
401 | } |
||
402 | |||
403 | /** |
||
404 | * Handles the literals that are prefixed by types. |
||
405 | * This method modified the token field as necessary. |
||
406 | * |
||
407 | * |
||
408 | * @throws ODataException |
||
409 | */ |
||
410 | private function handleTypePrefixedLiterals(): void |
||
411 | { |
||
412 | $id = $this->token->getId(); |
||
413 | if ($id != ExpressionTokenId::IDENTIFIER()) { |
||
414 | return; |
||
415 | } |
||
416 | |||
417 | $quoteFollows = $this->ch == '\''; |
||
418 | if (!$quoteFollows) { |
||
419 | return; |
||
420 | } |
||
421 | |||
422 | $tokenText = $this->token->Text; |
||
423 | |||
424 | if (strcasecmp('datetime', $tokenText) == 0) { |
||
425 | $id = ExpressionTokenId::DATETIME_LITERAL(); |
||
426 | } elseif (strcasecmp('guid', $tokenText) == 0) { |
||
427 | $id = ExpressionTokenId::GUID_LITERAL(); |
||
428 | } elseif (strcasecmp('binary', $tokenText) == 0 |
||
429 | || strcasecmp('X', $tokenText) == 0 |
||
430 | || strcasecmp('x', $tokenText) == 0 |
||
431 | ) { |
||
432 | $id = ExpressionTokenId::BINARY_LITERAL(); |
||
433 | } else { |
||
434 | return; |
||
435 | } |
||
436 | |||
437 | $tokenPos = $this->token->Position; |
||
438 | do { |
||
439 | $this->nextChar(); |
||
440 | } while ($this->ch != '\0' && $this->ch != '\''); |
||
441 | |||
442 | if ($this->ch == '\0') { |
||
443 | $this->parseError( |
||
444 | Messages::expressionLexerUnterminatedStringLiteral( |
||
445 | $this->textPos, |
||
446 | $this->text |
||
447 | ) |
||
448 | ); |
||
449 | } |
||
450 | |||
451 | $this->nextChar(); |
||
452 | $this->token->setId($id); |
||
453 | $this->token->Text |
||
454 | = substr($this->text, $tokenPos, $this->textPos - $tokenPos); |
||
455 | } |
||
456 | |||
457 | /** |
||
458 | * Check if the parameter ($tokenText) is INF or NaN. |
||
459 | * |
||
460 | * @param string $tokenText Text to look in |
||
461 | * |
||
462 | * @return bool true if match found, false otherwise |
||
463 | */ |
||
464 | private static function isInfinityOrNaNDouble(string $tokenText): bool |
||
475 | } |
||
476 | |||
477 | /** |
||
478 | * Checks if the parameter ($tokenText) is INFf/INFF or NaNf/NaNF. |
||
479 | * |
||
480 | * @param string $tokenText Input token |
||
481 | * |
||
482 | * @return bool true if match found, false otherwise |
||
483 | */ |
||
484 | private static function isInfinityOrNanSingle(string $tokenText): bool |
||
485 | { |
||
486 | if (strlen($tokenText) == 4) { |
||
487 | if ($tokenText[0] == 'I') { |
||
488 | return self::isInfinityLiteralSingle($tokenText); |
||
489 | } elseif ($tokenText[0] == 'N') { |
||
490 | return ($tokenText[3] == self::SINGLE_SUFFIX_LOWER |
||
491 | || $tokenText[3] == self::SINGLE_SUFFIX_UPPER) |
||
492 | && strncmp($tokenText, ODataConstants::XML_NAN_LITERAL, 3) == 0; |
||
493 | } |
||
494 | } |
||
495 | |||
496 | return false; |
||
497 | } |
||
498 | |||
499 | /** |
||
500 | * To get the expression token being processed. |
||
501 | * |
||
502 | * @return ExpressionToken |
||
503 | */ |
||
504 | public function getCurrentToken(): ExpressionToken |
||
505 | { |
||
506 | return $this->token; |
||
507 | } |
||
508 | |||
509 | /** |
||
510 | * To set the token being processed. |
||
511 | * |
||
512 | * @param ExpressionToken $token The expression token to set as current |
||
513 | */ |
||
514 | public function setCurrentToken(ExpressionToken $token): void |
||
515 | { |
||
516 | $this->token = $token; |
||
517 | } |
||
518 | |||
519 | /** |
||
520 | * To get the text being parsed. |
||
521 | * |
||
522 | * @return string |
||
523 | */ |
||
524 | public function getExpressionText(): string |
||
525 | { |
||
526 | return $this->text; |
||
527 | } |
||
528 | |||
529 | /** |
||
530 | * Position of the current token in the text being parsed. |
||
531 | * |
||
532 | * @return int |
||
533 | */ |
||
534 | public function getPosition(): int |
||
535 | { |
||
536 | return $this->token->Position; |
||
537 | } |
||
538 | |||
539 | /** |
||
540 | * Returns the next token without advancing the lexer to next token. |
||
541 | * |
||
542 | * @throws ODataException |
||
543 | * @return ExpressionToken |
||
544 | */ |
||
545 | public function peekNextToken(): ExpressionToken |
||
546 | { |
||
547 | $savedTextPos = $this->textPos; |
||
548 | assert(2 >= strlen($this->ch)); |
||
549 | $savedChar = $this->ch; |
||
550 | $savedToken = clone $this->token; |
||
551 | $this->nextToken(); |
||
552 | $result = clone $this->token; |
||
553 | $this->textPos = $savedTextPos; |
||
554 | $this->ch = $savedChar; |
||
555 | $this->token->setId($savedToken->getId()); |
||
556 | $this->token->Position = $savedToken->Position; |
||
557 | $this->token->Text = $savedToken->Text; |
||
558 | |||
559 | return $result; |
||
560 | } |
||
561 | |||
562 | /** |
||
563 | * Starting from an identifier, reads alternate sequence of dots and identifiers |
||
564 | * and returns the text for it. |
||
565 | * |
||
566 | * @throws ODataException |
||
567 | * @return string The dotted identifier starting at the current identifier |
||
568 | */ |
||
569 | public function readDottedIdentifier(): string |
||
570 | { |
||
571 | $this->validateToken(ExpressionTokenId::IDENTIFIER()); |
||
572 | $identifier = $this->token->Text; |
||
573 | $this->nextToken(); |
||
574 | while ($this->token->getId() == ExpressionTokenId::DOT()) { |
||
575 | $this->nextToken(); |
||
576 | $this->validateToken(ExpressionTokenId::IDENTIFIER()); |
||
577 | $identifier = $identifier . '.' . $this->token->Text; |
||
578 | $this->nextToken(); |
||
579 | } |
||
580 | |||
581 | return $identifier; |
||
582 | } |
||
583 | |||
584 | /** |
||
585 | * Validates the current token is of the specified kind. |
||
586 | * |
||
587 | * @param ExpressionTokenId $tokenId Expected token kind |
||
588 | * |
||
589 | * @throws ODataException if current token is not of the |
||
590 | * specified kind |
||
591 | */ |
||
592 | public function validateToken(ExpressionTokenId $tokenId): void |
||
596 | } |
||
597 | } |
||
598 | } |
||
599 |