1 | <?php |
||
2 | /** |
||
3 | * @link https://www.yiiframework.com/ |
||
4 | * @copyright Copyright (c) 2008 Yii Software LLC |
||
5 | * @license https://www.yiiframework.com/license/ |
||
6 | */ |
||
7 | |||
8 | namespace yii\db; |
||
9 | |||
10 | use yii\base\Component; |
||
11 | use yii\base\InvalidArgumentException; |
||
12 | |||
13 | /** |
||
14 | * SqlTokenizer splits an SQL query into individual SQL tokens. |
||
15 | * |
||
16 | * It can be used to obtain an addition information from an SQL code. |
||
17 | * |
||
18 | * Usage example: |
||
19 | * |
||
20 | * ```php |
||
21 | * $tokenizer = new SqlTokenizer("SELECT * FROM user WHERE id = 1"); |
||
22 | * $root = $tokeinzer->tokenize(); |
||
23 | * $sqlTokens = $root->getChildren(); |
||
24 | * ``` |
||
25 | * |
||
26 | * Tokens are instances of [[SqlToken]]. |
||
27 | * |
||
28 | * @author Sergey Makinen <[email protected]> |
||
29 | * @since 2.0.13 |
||
30 | */ |
||
31 | abstract class SqlTokenizer extends Component |
||
32 | { |
||
33 | /** |
||
34 | * @var string SQL code. |
||
35 | */ |
||
36 | public $sql; |
||
37 | |||
38 | /** |
||
39 | * @var int SQL code string length. |
||
40 | */ |
||
41 | protected $length; |
||
42 | /** |
||
43 | * @var int SQL code string current offset. |
||
44 | */ |
||
45 | protected $offset; |
||
46 | |||
47 | /** |
||
48 | * @var \SplStack stack of active tokens. |
||
49 | */ |
||
50 | private $_tokenStack; |
||
51 | /** |
||
52 | * @var SqlToken active token. It's usually a top of the token stack. |
||
53 | */ |
||
54 | private $_currentToken; |
||
55 | /** |
||
56 | * @var string[] cached substrings. |
||
57 | */ |
||
58 | private $_substrings; |
||
59 | /** |
||
60 | * @var string current buffer value. |
||
61 | */ |
||
62 | private $_buffer = ''; |
||
63 | /** |
||
64 | * @var SqlToken resulting token of a last [[tokenize()]] call. |
||
65 | */ |
||
66 | private $_token; |
||
67 | |||
68 | |||
69 | /** |
||
70 | * Constructor. |
||
71 | * @param string $sql SQL code to be tokenized. |
||
72 | * @param array $config name-value pairs that will be used to initialize the object properties |
||
73 | */ |
||
74 | public function __construct($sql, $config = []) |
||
75 | { |
||
76 | $this->sql = $sql; |
||
77 | parent::__construct($config); |
||
78 | } |
||
79 | |||
80 | /** |
||
81 | * Tokenizes and returns a code type token. |
||
82 | * @return SqlToken code type token. |
||
83 | */ |
||
84 | public function tokenize() |
||
85 | { |
||
86 | $this->length = mb_strlen($this->sql, 'UTF-8'); |
||
87 | $this->offset = 0; |
||
88 | $this->_substrings = []; |
||
89 | $this->_buffer = ''; |
||
90 | $this->_token = new SqlToken([ |
||
91 | 'type' => SqlToken::TYPE_CODE, |
||
92 | 'content' => $this->sql, |
||
93 | ]); |
||
94 | $this->_tokenStack = new \SplStack(); |
||
95 | $this->_tokenStack->push($this->_token); |
||
96 | $this->_token[] = new SqlToken(['type' => SqlToken::TYPE_STATEMENT]); |
||
97 | $this->_tokenStack->push($this->_token[0]); |
||
98 | $this->_currentToken = $this->_tokenStack->top(); |
||
99 | while (!$this->isEof()) { |
||
100 | if ($this->isWhitespace($length) || $this->isComment($length)) { |
||
101 | $this->addTokenFromBuffer(); |
||
102 | $this->advance($length); |
||
103 | continue; |
||
104 | } |
||
105 | |||
106 | if ($this->tokenizeOperator($length) || $this->tokenizeDelimitedString($length)) { |
||
107 | $this->advance($length); |
||
108 | continue; |
||
109 | } |
||
110 | |||
111 | $this->_buffer .= $this->substring(1); |
||
112 | $this->advance(1); |
||
113 | } |
||
114 | $this->addTokenFromBuffer(); |
||
115 | if ($this->_token->getHasChildren() && !$this->_token[-1]->getHasChildren()) { |
||
0 ignored issues
–
show
|
|||
116 | unset($this->_token[-1]); |
||
117 | } |
||
118 | |||
119 | return $this->_token; |
||
120 | } |
||
121 | |||
122 | /** |
||
123 | * Returns whether there's a whitespace at the current offset. |
||
124 | * If this methos returns `true`, it has to set the `$length` parameter to the length of the matched string. |
||
125 | * @param int $length length of the matched string. |
||
126 | * @return bool whether there's a whitespace at the current offset. |
||
127 | */ |
||
128 | abstract protected function isWhitespace(&$length); |
||
129 | |||
130 | /** |
||
131 | * Returns whether there's a commentary at the current offset. |
||
132 | * If this methos returns `true`, it has to set the `$length` parameter to the length of the matched string. |
||
133 | * @param int $length length of the matched string. |
||
134 | * @return bool whether there's a commentary at the current offset. |
||
135 | */ |
||
136 | abstract protected function isComment(&$length); |
||
137 | |||
138 | /** |
||
139 | * Returns whether there's an operator at the current offset. |
||
140 | * If this methos returns `true`, it has to set the `$length` parameter to the length of the matched string. |
||
141 | * It may also set `$content` to a string that will be used as a token content. |
||
142 | * @param int $length length of the matched string. |
||
143 | * @param string $content optional content instead of the matched string. |
||
144 | * @return bool whether there's an operator at the current offset. |
||
145 | */ |
||
146 | abstract protected function isOperator(&$length, &$content); |
||
147 | |||
148 | /** |
||
149 | * Returns whether there's an identifier at the current offset. |
||
150 | * If this methos returns `true`, it has to set the `$length` parameter to the length of the matched string. |
||
151 | * It may also set `$content` to a string that will be used as a token content. |
||
152 | * @param int $length length of the matched string. |
||
153 | * @param string $content optional content instead of the matched string. |
||
154 | * @return bool whether there's an identifier at the current offset. |
||
155 | */ |
||
156 | abstract protected function isIdentifier(&$length, &$content); |
||
157 | |||
158 | /** |
||
159 | * Returns whether there's a string literal at the current offset. |
||
160 | * If this methos returns `true`, it has to set the `$length` parameter to the length of the matched string. |
||
161 | * It may also set `$content` to a string that will be used as a token content. |
||
162 | * @param int $length length of the matched string. |
||
163 | * @param string $content optional content instead of the matched string. |
||
164 | * @return bool whether there's a string literal at the current offset. |
||
165 | */ |
||
166 | abstract protected function isStringLiteral(&$length, &$content); |
||
167 | |||
168 | /** |
||
169 | * Returns whether the given string is a keyword. |
||
170 | * The method may set `$content` to a string that will be used as a token content. |
||
171 | * @param string $string string to be matched. |
||
172 | * @param string $content optional content instead of the matched string. |
||
173 | * @return bool whether the given string is a keyword. |
||
174 | */ |
||
175 | abstract protected function isKeyword($string, &$content); |
||
176 | |||
177 | /** |
||
178 | * Returns whether the longest common prefix equals to the SQL code of the same length at the current offset. |
||
179 | * @param string[] $with strings to be tested. |
||
180 | * The method **will** modify this parameter to speed up lookups. |
||
181 | * @param bool $caseSensitive whether to perform a case sensitive comparison. |
||
182 | * @param int|null $length length of the matched string. |
||
183 | * @param string|null $content matched string. |
||
184 | * @return bool whether a match is found. |
||
185 | */ |
||
186 | protected function startsWithAnyLongest(array &$with, $caseSensitive, &$length = null, &$content = null) |
||
187 | { |
||
188 | if (empty($with)) { |
||
189 | return false; |
||
190 | } |
||
191 | |||
192 | if (!is_array(reset($with))) { |
||
193 | usort($with, function ($string1, $string2) { |
||
194 | return mb_strlen($string2, 'UTF-8') - mb_strlen($string1, 'UTF-8'); |
||
195 | }); |
||
196 | $map = []; |
||
197 | foreach ($with as $string) { |
||
198 | $map[mb_strlen($string, 'UTF-8')][$caseSensitive ? $string : mb_strtoupper($string, 'UTF-8')] = true; |
||
199 | } |
||
200 | $with = $map; |
||
201 | } |
||
202 | foreach ($with as $testLength => $testValues) { |
||
203 | $content = $this->substring($testLength, $caseSensitive); |
||
204 | if (isset($testValues[$content])) { |
||
205 | $length = $testLength; |
||
206 | return true; |
||
207 | } |
||
208 | } |
||
209 | |||
210 | return false; |
||
211 | } |
||
212 | |||
213 | /** |
||
214 | * Returns a string of the given length starting with the specified offset. |
||
215 | * @param int $length string length to be returned. |
||
216 | * @param bool $caseSensitive if it's `false`, the string will be uppercased. |
||
217 | * @param int|null $offset SQL code offset, defaults to current if `null` is passed. |
||
218 | * @return string result string, it may be empty if there's nothing to return. |
||
219 | */ |
||
220 | protected function substring($length, $caseSensitive = true, $offset = null) |
||
221 | { |
||
222 | if ($offset === null) { |
||
223 | $offset = $this->offset; |
||
224 | } |
||
225 | if ($offset + $length > $this->length) { |
||
226 | return ''; |
||
227 | } |
||
228 | |||
229 | $cacheKey = $offset . ',' . $length; |
||
230 | if (!isset($this->_substrings[$cacheKey . ',1'])) { |
||
231 | $this->_substrings[$cacheKey . ',1'] = mb_substr($this->sql, $offset, $length, 'UTF-8'); |
||
232 | } |
||
233 | if (!$caseSensitive && !isset($this->_substrings[$cacheKey . ',0'])) { |
||
234 | $this->_substrings[$cacheKey . ',0'] = mb_strtoupper($this->_substrings[$cacheKey . ',1'], 'UTF-8'); |
||
235 | } |
||
236 | |||
237 | return $this->_substrings[$cacheKey . ',' . (int) $caseSensitive]; |
||
238 | } |
||
239 | |||
240 | /** |
||
241 | * Returns an index after the given string in the SQL code starting with the specified offset. |
||
242 | * @param string $string string to be found. |
||
243 | * @param int|null $offset SQL code offset, defaults to current if `null` is passed. |
||
244 | * @return int index after the given string or end of string index. |
||
245 | */ |
||
246 | protected function indexAfter($string, $offset = null) |
||
247 | { |
||
248 | if ($offset === null) { |
||
249 | $offset = $this->offset; |
||
250 | } |
||
251 | if ($offset + mb_strlen($string, 'UTF-8') > $this->length) { |
||
252 | return $this->length; |
||
253 | } |
||
254 | |||
255 | $afterIndexOf = mb_strpos($this->sql, $string, $offset, 'UTF-8'); |
||
256 | if ($afterIndexOf === false) { |
||
257 | $afterIndexOf = $this->length; |
||
258 | } else { |
||
259 | $afterIndexOf += mb_strlen($string, 'UTF-8'); |
||
260 | } |
||
261 | |||
262 | return $afterIndexOf; |
||
263 | } |
||
264 | |||
265 | /** |
||
266 | * Determines whether there is a delimited string at the current offset and adds it to the token children. |
||
267 | * @param int $length |
||
268 | * @return bool |
||
269 | */ |
||
270 | private function tokenizeDelimitedString(&$length) |
||
271 | { |
||
272 | $isIdentifier = $this->isIdentifier($length, $content); |
||
273 | $isStringLiteral = !$isIdentifier && $this->isStringLiteral($length, $content); |
||
274 | if (!$isIdentifier && !$isStringLiteral) { |
||
275 | return false; |
||
276 | } |
||
277 | |||
278 | $this->addTokenFromBuffer(); |
||
279 | $this->_currentToken[] = new SqlToken([ |
||
280 | 'type' => $isIdentifier ? SqlToken::TYPE_IDENTIFIER : SqlToken::TYPE_STRING_LITERAL, |
||
281 | 'content' => is_string($content) ? $content : $this->substring($length), |
||
282 | 'startOffset' => $this->offset, |
||
283 | 'endOffset' => $this->offset + $length, |
||
284 | ]); |
||
285 | return true; |
||
286 | } |
||
287 | |||
288 | /** |
||
289 | * Determines whether there is an operator at the current offset and adds it to the token children. |
||
290 | * @param int $length |
||
291 | * @return bool |
||
292 | */ |
||
293 | private function tokenizeOperator(&$length) |
||
294 | { |
||
295 | if (!$this->isOperator($length, $content)) { |
||
296 | return false; |
||
297 | } |
||
298 | |||
299 | $this->addTokenFromBuffer(); |
||
300 | switch ($this->substring($length)) { |
||
301 | case '(': |
||
302 | $this->_currentToken[] = new SqlToken([ |
||
303 | 'type' => SqlToken::TYPE_OPERATOR, |
||
304 | 'content' => is_string($content) ? $content : $this->substring($length), |
||
305 | 'startOffset' => $this->offset, |
||
306 | 'endOffset' => $this->offset + $length, |
||
307 | ]); |
||
308 | $this->_currentToken[] = new SqlToken(['type' => SqlToken::TYPE_PARENTHESIS]); |
||
309 | $this->_tokenStack->push($this->_currentToken[-1]); |
||
310 | $this->_currentToken = $this->_tokenStack->top(); |
||
311 | break; |
||
312 | case ')': |
||
313 | $this->_tokenStack->pop(); |
||
314 | $this->_currentToken = $this->_tokenStack->top(); |
||
315 | $this->_currentToken[] = new SqlToken([ |
||
316 | 'type' => SqlToken::TYPE_OPERATOR, |
||
317 | 'content' => ')', |
||
318 | 'startOffset' => $this->offset, |
||
319 | 'endOffset' => $this->offset + $length, |
||
320 | ]); |
||
321 | break; |
||
322 | case ';': |
||
323 | if (!$this->_currentToken->getHasChildren()) { |
||
324 | break; |
||
325 | } |
||
326 | |||
327 | $this->_currentToken[] = new SqlToken([ |
||
328 | 'type' => SqlToken::TYPE_OPERATOR, |
||
329 | 'content' => is_string($content) ? $content : $this->substring($length), |
||
330 | 'startOffset' => $this->offset, |
||
331 | 'endOffset' => $this->offset + $length, |
||
332 | ]); |
||
333 | $this->_tokenStack->pop(); |
||
334 | $this->_currentToken = $this->_tokenStack->top(); |
||
335 | $this->_currentToken[] = new SqlToken(['type' => SqlToken::TYPE_STATEMENT]); |
||
336 | $this->_tokenStack->push($this->_currentToken[-1]); |
||
337 | $this->_currentToken = $this->_tokenStack->top(); |
||
338 | break; |
||
339 | default: |
||
340 | $this->_currentToken[] = new SqlToken([ |
||
341 | 'type' => SqlToken::TYPE_OPERATOR, |
||
342 | 'content' => is_string($content) ? $content : $this->substring($length), |
||
343 | 'startOffset' => $this->offset, |
||
344 | 'endOffset' => $this->offset + $length, |
||
345 | ]); |
||
346 | break; |
||
347 | } |
||
348 | |||
349 | return true; |
||
350 | } |
||
351 | |||
352 | /** |
||
353 | * Determines a type of text in the buffer, tokenizes it and adds it to the token children. |
||
354 | */ |
||
355 | private function addTokenFromBuffer() |
||
356 | { |
||
357 | if ($this->_buffer === '') { |
||
358 | return; |
||
359 | } |
||
360 | |||
361 | $isKeyword = $this->isKeyword($this->_buffer, $content); |
||
362 | $this->_currentToken[] = new SqlToken([ |
||
363 | 'type' => $isKeyword ? SqlToken::TYPE_KEYWORD : SqlToken::TYPE_TOKEN, |
||
364 | 'content' => is_string($content) ? $content : $this->_buffer, |
||
365 | 'startOffset' => $this->offset - mb_strlen($this->_buffer, 'UTF-8'), |
||
366 | 'endOffset' => $this->offset, |
||
367 | ]); |
||
368 | $this->_buffer = ''; |
||
369 | } |
||
370 | |||
371 | /** |
||
372 | * Adds the specified length to the current offset. |
||
373 | * @param int $length |
||
374 | * @throws InvalidArgumentException |
||
375 | */ |
||
376 | private function advance($length) |
||
377 | { |
||
378 | if ($length <= 0) { |
||
379 | throw new InvalidArgumentException('Length must be greater than 0.'); |
||
380 | } |
||
381 | |||
382 | $this->offset += $length; |
||
383 | $this->_substrings = []; |
||
384 | } |
||
385 | |||
386 | /** |
||
387 | * Returns whether the SQL code is completely traversed. |
||
388 | * @return bool |
||
389 | */ |
||
390 | private function isEof() |
||
391 | { |
||
392 | return $this->offset >= $this->length; |
||
393 | } |
||
394 | } |
||
395 |
This check looks for calls to methods that do not seem to exist on a given type. It looks for the method on the type itself as well as in inherited classes or implemented interfaces.
This is most likely a typographical error or the method has been renamed.