|
1
|
|
|
<?php |
|
2
|
|
|
|
|
3
|
|
|
declare(strict_types=1); |
|
4
|
|
|
|
|
5
|
|
|
namespace PhpMyAdmin\SqlParser\Utils; |
|
6
|
|
|
|
|
7
|
|
|
use PhpMyAdmin\SqlParser\Components\JoinKeyword; |
|
8
|
|
|
use PhpMyAdmin\SqlParser\Lexer; |
|
9
|
|
|
use PhpMyAdmin\SqlParser\Parser; |
|
10
|
|
|
use PhpMyAdmin\SqlParser\Token; |
|
11
|
|
|
use PhpMyAdmin\SqlParser\TokensList; |
|
12
|
|
|
use PhpMyAdmin\SqlParser\TokenType; |
|
13
|
|
|
|
|
14
|
|
|
use function array_merge; |
|
15
|
|
|
use function array_pop; |
|
16
|
|
|
use function end; |
|
17
|
|
|
use function htmlspecialchars; |
|
18
|
|
|
use function in_array; |
|
19
|
|
|
use function mb_strlen; |
|
20
|
|
|
use function str_contains; |
|
21
|
|
|
use function str_repeat; |
|
22
|
|
|
use function str_replace; |
|
23
|
|
|
use function strtoupper; |
|
24
|
|
|
|
|
25
|
|
|
use const ENT_NOQUOTES; |
|
26
|
|
|
use const PHP_SAPI; |
|
27
|
|
|
|
|
28
|
|
|
/** |
|
29
|
|
|
* Utilities that are used for formatting queries. |
|
30
|
|
|
*/ |
|
31
|
|
|
class Formatter |
|
32
|
|
|
{ |
|
33
|
|
|
/** |
|
34
|
|
|
* The formatting options. |
|
35
|
|
|
* |
|
36
|
|
|
* @var array<string, bool|string|array<int, array<string, int|string>>> |
|
37
|
|
|
*/ |
|
38
|
|
|
public array $options; |
|
39
|
|
|
|
|
40
|
|
|
/** |
|
41
|
|
|
* Clauses that are usually short. |
|
42
|
|
|
* |
|
43
|
|
|
* These clauses share the line with the next clause. |
|
44
|
|
|
* |
|
45
|
|
|
* E.g. if INSERT was not here, the formatter would produce: |
|
46
|
|
|
* |
|
47
|
|
|
* INSERT |
|
48
|
|
|
* INTO foo |
|
49
|
|
|
* VALUES(0, 0, 0),(1, 1, 1); |
|
50
|
|
|
* |
|
51
|
|
|
* Instead of: |
|
52
|
|
|
* |
|
53
|
|
|
* INSERT INTO foo |
|
54
|
|
|
* VALUES(0, 0, 0),(1, 1, 1) |
|
55
|
|
|
* |
|
56
|
|
|
* @var array<string, bool> |
|
57
|
|
|
*/ |
|
58
|
|
|
public static array $shortClauses = [ |
|
59
|
|
|
'CREATE' => true, |
|
60
|
|
|
'INSERT' => true, |
|
61
|
|
|
]; |
|
62
|
|
|
|
|
63
|
|
|
/** |
|
64
|
|
|
* Clauses that must be inlined. |
|
65
|
|
|
* |
|
66
|
|
|
* These clauses usually are short and it's nicer to have them inline. |
|
67
|
|
|
* |
|
68
|
|
|
* @var array<string, bool> |
|
69
|
|
|
*/ |
|
70
|
|
|
public static array $inlineClauses = [ |
|
71
|
|
|
'CREATE' => true, |
|
72
|
|
|
'INTO' => true, |
|
73
|
|
|
'LIMIT' => true, |
|
74
|
|
|
'PARTITION BY' => true, |
|
75
|
|
|
'PARTITION' => true, |
|
76
|
|
|
'PROCEDURE' => true, |
|
77
|
|
|
'SUBPARTITION BY' => true, |
|
78
|
|
|
'VALUES' => true, |
|
79
|
|
|
]; |
|
80
|
|
|
|
|
81
|
|
|
private const FORMATTERS = [ |
|
82
|
|
|
'PARTITION BY', |
|
83
|
|
|
'SUBPARTITION BY', |
|
84
|
|
|
]; |
|
85
|
|
|
|
|
86
|
|
|
/** @param array<string, bool|string|array<int, array<string, int|string>>> $options the formatting options */ |
|
87
|
48 |
|
public function __construct(array $options = []) |
|
88
|
|
|
{ |
|
89
|
48 |
|
$this->options = $this->getMergedOptions($options); |
|
90
|
|
|
} |
|
91
|
|
|
|
|
92
|
|
|
/** |
|
93
|
|
|
* The specified formatting options are merged with the default values. |
|
94
|
|
|
* |
|
95
|
|
|
* @param array<string, bool|string|array<int, array<string, int|string>>> $options |
|
96
|
|
|
* |
|
97
|
|
|
* @return array<string, bool|string|array<int, array<string, int|string>>> |
|
98
|
|
|
*/ |
|
99
|
56 |
|
protected function getMergedOptions(array $options): array |
|
100
|
|
|
{ |
|
101
|
56 |
|
$options = array_merge( |
|
102
|
56 |
|
$this->getDefaultOptions(), |
|
103
|
56 |
|
$options, |
|
104
|
56 |
|
); |
|
105
|
|
|
|
|
106
|
56 |
|
if (isset($options['formats'])) { |
|
107
|
8 |
|
$options['formats'] = self::mergeFormats($this->getDefaultFormats(), $options['formats']); |
|
108
|
|
|
} else { |
|
109
|
48 |
|
$options['formats'] = $this->getDefaultFormats(); |
|
110
|
|
|
} |
|
111
|
|
|
|
|
112
|
56 |
|
if ($options['line_ending'] === null) { |
|
113
|
48 |
|
$options['line_ending'] = $options['type'] === 'html' ? '<br/>' : "\n"; |
|
114
|
|
|
} |
|
115
|
|
|
|
|
116
|
56 |
|
if ($options['indentation'] === null) { |
|
117
|
56 |
|
$options['indentation'] = $options['type'] === 'html' ? ' ' : ' '; |
|
118
|
|
|
} |
|
119
|
|
|
|
|
120
|
|
|
// `parts_newline` requires `clause_newline` |
|
121
|
56 |
|
$options['parts_newline'] &= $options['clause_newline']; |
|
122
|
|
|
|
|
123
|
56 |
|
return $options; |
|
124
|
|
|
} |
|
125
|
|
|
|
|
126
|
|
|
/** |
|
127
|
|
|
* The default formatting options. |
|
128
|
|
|
* |
|
129
|
|
|
* @return array<string, bool|string|null> |
|
130
|
|
|
* @psalm-return array{ |
|
131
|
|
|
* type: ('cli'|'text'), |
|
132
|
|
|
* line_ending: null, |
|
133
|
|
|
* indentation: null, |
|
134
|
|
|
* remove_comments: false, |
|
135
|
|
|
* clause_newline: true, |
|
136
|
|
|
* parts_newline: true, |
|
137
|
|
|
* indent_parts: true |
|
138
|
|
|
* } |
|
139
|
|
|
*/ |
|
140
|
48 |
|
protected function getDefaultOptions(): array |
|
141
|
|
|
{ |
|
142
|
48 |
|
return [ |
|
143
|
|
|
/* |
|
144
|
|
|
* The format of the result. |
|
145
|
|
|
* |
|
146
|
|
|
* @var string The type ('text', 'cli' or 'html') |
|
147
|
|
|
*/ |
|
148
|
48 |
|
'type' => PHP_SAPI === 'cli' ? 'cli' : 'text', |
|
149
|
|
|
|
|
150
|
|
|
/* |
|
151
|
|
|
* The line ending used. |
|
152
|
|
|
* By default, for text this is "\n" and for HTML this is "<br/>". |
|
153
|
|
|
* |
|
154
|
|
|
* @var string |
|
155
|
|
|
*/ |
|
156
|
48 |
|
'line_ending' => null, |
|
157
|
|
|
|
|
158
|
|
|
/* |
|
159
|
|
|
* The string used for indentation. |
|
160
|
|
|
* |
|
161
|
|
|
* @var string |
|
162
|
|
|
*/ |
|
163
|
48 |
|
'indentation' => null, |
|
164
|
|
|
|
|
165
|
|
|
/* |
|
166
|
|
|
* Whether comments should be removed or not. |
|
167
|
|
|
* |
|
168
|
|
|
* @var bool |
|
169
|
|
|
*/ |
|
170
|
48 |
|
'remove_comments' => false, |
|
171
|
|
|
|
|
172
|
|
|
/* |
|
173
|
|
|
* Whether each clause should be on a new line. |
|
174
|
|
|
* |
|
175
|
|
|
* @var bool |
|
176
|
|
|
*/ |
|
177
|
48 |
|
'clause_newline' => true, |
|
178
|
|
|
|
|
179
|
|
|
/* |
|
180
|
|
|
* Whether each part should be on a new line. |
|
181
|
|
|
* Parts are delimited by brackets and commas. |
|
182
|
|
|
* |
|
183
|
|
|
* @var bool |
|
184
|
|
|
*/ |
|
185
|
48 |
|
'parts_newline' => true, |
|
186
|
|
|
|
|
187
|
|
|
/* |
|
188
|
|
|
* Whether each part of each clause should be indented. |
|
189
|
|
|
* |
|
190
|
|
|
* @var bool |
|
191
|
|
|
*/ |
|
192
|
48 |
|
'indent_parts' => true, |
|
193
|
48 |
|
]; |
|
194
|
|
|
} |
|
195
|
|
|
|
|
196
|
|
|
/** |
|
197
|
|
|
* The styles used for HTML formatting. |
|
198
|
|
|
* [$type, $flags, $span, $callback]. |
|
199
|
|
|
* |
|
200
|
|
|
* @return array<int, array<string, int|string>> |
|
201
|
|
|
* @psalm-return list<array{type: int, flags: int, html: string, cli: string, function: string}> |
|
202
|
|
|
*/ |
|
203
|
48 |
|
protected function getDefaultFormats(): array |
|
204
|
|
|
{ |
|
205
|
48 |
|
return [ |
|
206
|
48 |
|
[ |
|
207
|
48 |
|
'type' => TokenType::Keyword->value, |
|
208
|
48 |
|
'flags' => Token::FLAG_KEYWORD_RESERVED, |
|
209
|
48 |
|
'html' => 'class="sql-reserved"', |
|
210
|
48 |
|
'cli' => "\x1b[35m", |
|
211
|
48 |
|
'function' => 'strtoupper', |
|
212
|
48 |
|
], |
|
213
|
48 |
|
[ |
|
214
|
48 |
|
'type' => TokenType::Keyword->value, |
|
215
|
48 |
|
'flags' => 0, |
|
216
|
48 |
|
'html' => 'class="sql-keyword"', |
|
217
|
48 |
|
'cli' => "\x1b[95m", |
|
218
|
48 |
|
'function' => 'strtoupper', |
|
219
|
48 |
|
], |
|
220
|
48 |
|
[ |
|
221
|
48 |
|
'type' => TokenType::Comment->value, |
|
222
|
48 |
|
'flags' => 0, |
|
223
|
48 |
|
'html' => 'class="sql-comment"', |
|
224
|
48 |
|
'cli' => "\x1b[37m", |
|
225
|
48 |
|
'function' => '', |
|
226
|
48 |
|
], |
|
227
|
48 |
|
[ |
|
228
|
48 |
|
'type' => TokenType::Bool->value, |
|
229
|
48 |
|
'flags' => 0, |
|
230
|
48 |
|
'html' => 'class="sql-atom"', |
|
231
|
48 |
|
'cli' => "\x1b[36m", |
|
232
|
48 |
|
'function' => 'strtoupper', |
|
233
|
48 |
|
], |
|
234
|
48 |
|
[ |
|
235
|
48 |
|
'type' => TokenType::Number->value, |
|
236
|
48 |
|
'flags' => 0, |
|
237
|
48 |
|
'html' => 'class="sql-number"', |
|
238
|
48 |
|
'cli' => "\x1b[92m", |
|
239
|
48 |
|
'function' => 'strtolower', |
|
240
|
48 |
|
], |
|
241
|
48 |
|
[ |
|
242
|
48 |
|
'type' => TokenType::String->value, |
|
243
|
48 |
|
'flags' => 0, |
|
244
|
48 |
|
'html' => 'class="sql-string"', |
|
245
|
48 |
|
'cli' => "\x1b[91m", |
|
246
|
48 |
|
'function' => '', |
|
247
|
48 |
|
], |
|
248
|
48 |
|
[ |
|
249
|
48 |
|
'type' => TokenType::Symbol->value, |
|
250
|
48 |
|
'flags' => Token::FLAG_SYMBOL_PARAMETER, |
|
251
|
48 |
|
'html' => 'class="sql-parameter"', |
|
252
|
48 |
|
'cli' => "\x1b[31m", |
|
253
|
48 |
|
'function' => '', |
|
254
|
48 |
|
], |
|
255
|
48 |
|
[ |
|
256
|
48 |
|
'type' => TokenType::Symbol->value, |
|
257
|
48 |
|
'flags' => 0, |
|
258
|
48 |
|
'html' => 'class="sql-variable"', |
|
259
|
48 |
|
'cli' => "\x1b[36m", |
|
260
|
48 |
|
'function' => '', |
|
261
|
48 |
|
], |
|
262
|
48 |
|
]; |
|
263
|
|
|
} |
|
264
|
|
|
|
|
265
|
|
|
/** |
|
266
|
|
|
* @param array<int, array<string, int|string>> $formats |
|
267
|
|
|
* @param array<int, array<string, int|string>> $newFormats |
|
268
|
|
|
* |
|
269
|
|
|
* @return array<int, array<string, int|string>> |
|
270
|
|
|
*/ |
|
271
|
8 |
|
private static function mergeFormats(array $formats, array $newFormats): array |
|
272
|
|
|
{ |
|
273
|
8 |
|
$added = []; |
|
274
|
8 |
|
$integers = [ |
|
275
|
8 |
|
'flags', |
|
276
|
8 |
|
'type', |
|
277
|
8 |
|
]; |
|
278
|
8 |
|
$strings = [ |
|
279
|
8 |
|
'html', |
|
280
|
8 |
|
'cli', |
|
281
|
8 |
|
'function', |
|
282
|
8 |
|
]; |
|
283
|
|
|
|
|
284
|
|
|
/* Sanitize the array so that we do not have to care later */ |
|
285
|
8 |
|
foreach ($newFormats as $j => $new) { |
|
286
|
8 |
|
foreach ($integers as $name) { |
|
287
|
8 |
|
if (isset($new[$name])) { |
|
288
|
6 |
|
continue; |
|
289
|
|
|
} |
|
290
|
|
|
|
|
291
|
6 |
|
$newFormats[$j][$name] = 0; |
|
292
|
|
|
} |
|
293
|
|
|
|
|
294
|
8 |
|
foreach ($strings as $name) { |
|
295
|
8 |
|
if (isset($new[$name])) { |
|
296
|
6 |
|
continue; |
|
297
|
|
|
} |
|
298
|
|
|
|
|
299
|
8 |
|
$newFormats[$j][$name] = ''; |
|
300
|
|
|
} |
|
301
|
|
|
} |
|
302
|
|
|
|
|
303
|
|
|
/* Process changes to existing formats */ |
|
304
|
8 |
|
foreach ($formats as $i => $original) { |
|
305
|
8 |
|
foreach ($newFormats as $j => $new) { |
|
306
|
8 |
|
if ($new['type'] !== $original['type'] || $original['flags'] !== $new['flags']) { |
|
307
|
6 |
|
continue; |
|
308
|
|
|
} |
|
309
|
|
|
|
|
310
|
6 |
|
$formats[$i] = $new; |
|
311
|
6 |
|
$added[] = $j; |
|
312
|
|
|
} |
|
313
|
|
|
} |
|
314
|
|
|
|
|
315
|
|
|
/* Add not already handled formats */ |
|
316
|
8 |
|
foreach ($newFormats as $j => $new) { |
|
317
|
8 |
|
if (in_array($j, $added)) { |
|
318
|
6 |
|
continue; |
|
319
|
|
|
} |
|
320
|
|
|
|
|
321
|
2 |
|
$formats[] = $new; |
|
322
|
|
|
} |
|
323
|
|
|
|
|
324
|
8 |
|
return $formats; |
|
325
|
|
|
} |
|
326
|
|
|
|
|
327
|
|
|
/** |
|
328
|
|
|
* Formats the given list of tokens. |
|
329
|
|
|
* |
|
330
|
|
|
* @param TokensList $list the list of tokens |
|
331
|
|
|
*/ |
|
332
|
48 |
|
public function formatList(TokensList $list): string |
|
333
|
|
|
{ |
|
334
|
|
|
/** |
|
335
|
|
|
* The query to be returned. |
|
336
|
|
|
*/ |
|
337
|
48 |
|
$ret = ''; |
|
338
|
|
|
|
|
339
|
|
|
/** |
|
340
|
|
|
* The indentation level. |
|
341
|
|
|
*/ |
|
342
|
48 |
|
$indent = 0; |
|
343
|
|
|
|
|
344
|
|
|
/** |
|
345
|
|
|
* Whether the line ended. |
|
346
|
|
|
*/ |
|
347
|
48 |
|
$lineEnded = false; |
|
348
|
|
|
|
|
349
|
|
|
/** |
|
350
|
|
|
* Whether current group is short (no linebreaks). |
|
351
|
|
|
*/ |
|
352
|
48 |
|
$shortGroup = false; |
|
353
|
|
|
|
|
354
|
|
|
/** |
|
355
|
|
|
* The name of the last clause. |
|
356
|
|
|
*/ |
|
357
|
48 |
|
$lastClause = ''; |
|
358
|
|
|
|
|
359
|
|
|
/** |
|
360
|
|
|
* A stack that keeps track of the indentation level every time a new |
|
361
|
|
|
* block is found. |
|
362
|
|
|
*/ |
|
363
|
48 |
|
$blocksIndentation = []; |
|
364
|
|
|
|
|
365
|
|
|
/** |
|
366
|
|
|
* A stack that keeps track of the line endings every time a new block |
|
367
|
|
|
* is found. |
|
368
|
|
|
*/ |
|
369
|
48 |
|
$blocksLineEndings = []; |
|
370
|
|
|
|
|
371
|
|
|
/** |
|
372
|
|
|
* Whether clause's options were formatted. |
|
373
|
|
|
*/ |
|
374
|
48 |
|
$formattedOptions = false; |
|
375
|
|
|
|
|
376
|
|
|
/** |
|
377
|
|
|
* Previously parsed token. |
|
378
|
|
|
*/ |
|
379
|
48 |
|
$prev = null; |
|
380
|
|
|
|
|
381
|
|
|
// In order to be able to format the queries correctly, the next token |
|
382
|
|
|
// must be taken into consideration. The loop below uses two pointers, |
|
383
|
|
|
// `$prev` and `$curr` which store two consecutive tokens. |
|
384
|
|
|
// Actually, at every iteration the previous token is being used. |
|
385
|
48 |
|
for ($list->idx = 0; $list->idx < $list->count; ++$list->idx) { |
|
386
|
|
|
/** |
|
387
|
|
|
* Token parsed at this moment. |
|
388
|
|
|
*/ |
|
389
|
48 |
|
$curr = $list->tokens[$list->idx]; |
|
390
|
48 |
|
if ($list->idx + 1 < $list->count) { |
|
391
|
46 |
|
$next = $list->tokens[$list->idx + 1]; |
|
392
|
|
|
} else { |
|
393
|
48 |
|
$next = null; |
|
394
|
|
|
} |
|
395
|
|
|
|
|
396
|
48 |
|
if ($curr->type === TokenType::Whitespace) { |
|
397
|
|
|
// Keep linebreaks before and after comments |
|
398
|
|
|
if ( |
|
399
|
46 |
|
str_contains($curr->token, "\n") && ( |
|
400
|
46 |
|
($prev !== null && $prev->type === TokenType::Comment) || |
|
401
|
46 |
|
($next !== null && $next->type === TokenType::Comment) |
|
402
|
|
|
) |
|
403
|
|
|
) { |
|
404
|
2 |
|
$lineEnded = true; |
|
405
|
|
|
} |
|
406
|
|
|
|
|
407
|
|
|
// Whitespaces are skipped because the formatter adds its own. |
|
408
|
46 |
|
continue; |
|
409
|
|
|
} |
|
410
|
|
|
|
|
411
|
48 |
|
if ($curr->type === TokenType::Comment && $this->options['remove_comments']) { |
|
412
|
|
|
// Skip Comments if option `remove_comments` is enabled |
|
413
|
2 |
|
continue; |
|
414
|
|
|
} |
|
415
|
|
|
|
|
416
|
|
|
// Checking if pointers were initialized. |
|
417
|
48 |
|
if ($prev !== null) { |
|
418
|
|
|
// Checking if a new clause started. |
|
419
|
46 |
|
if (static::isClause($prev) !== false) { |
|
|
|
|
|
|
420
|
46 |
|
$lastClause = $prev->value; |
|
421
|
46 |
|
$formattedOptions = false; |
|
422
|
|
|
} |
|
423
|
|
|
|
|
424
|
|
|
// The options of a clause should stay on the same line and everything that follows. |
|
425
|
|
|
if ( |
|
426
|
46 |
|
$this->options['parts_newline'] |
|
427
|
46 |
|
&& ! $formattedOptions |
|
428
|
46 |
|
&& empty(self::$inlineClauses[$lastClause]) |
|
429
|
|
|
&& ( |
|
430
|
46 |
|
$curr->type !== TokenType::Keyword |
|
431
|
46 |
|
|| ($curr->flags & Token::FLAG_KEYWORD_FUNCTION) |
|
432
|
|
|
) |
|
433
|
|
|
) { |
|
434
|
42 |
|
$formattedOptions = true; |
|
435
|
42 |
|
$lineEnded = true; |
|
436
|
42 |
|
++$indent; |
|
437
|
|
|
} |
|
438
|
|
|
|
|
439
|
|
|
// Checking if this clause ended. |
|
440
|
46 |
|
$isClause = static::isClause($curr); |
|
441
|
|
|
|
|
442
|
46 |
|
if ($isClause !== false) { |
|
443
|
|
|
if ( |
|
444
|
20 |
|
($isClause === 2 || $this->options['clause_newline']) |
|
445
|
20 |
|
&& empty(self::$shortClauses[$lastClause]) |
|
446
|
|
|
) { |
|
447
|
20 |
|
$lineEnded = true; |
|
448
|
20 |
|
if ($this->options['parts_newline'] && $indent > 0) { |
|
449
|
18 |
|
--$indent; |
|
450
|
|
|
} |
|
451
|
|
|
} |
|
452
|
|
|
} |
|
453
|
|
|
|
|
454
|
|
|
// Inline JOINs |
|
455
|
|
|
if ( |
|
456
|
46 |
|
($prev->type === TokenType::Keyword && isset(JoinKeyword::JOINS[$prev->value])) |
|
457
|
46 |
|
|| (in_array($curr->value, ['ON', 'USING'], true) |
|
458
|
46 |
|
&& isset(JoinKeyword::JOINS[$list->tokens[$list->idx - 2]->value])) |
|
459
|
46 |
|
|| isset($list->tokens[$list->idx - 4], JoinKeyword::JOINS[$list->tokens[$list->idx - 4]->value]) |
|
460
|
46 |
|
|| isset($list->tokens[$list->idx - 6], JoinKeyword::JOINS[$list->tokens[$list->idx - 6]->value]) |
|
461
|
|
|
) { |
|
462
|
2 |
|
$lineEnded = false; |
|
463
|
|
|
} |
|
464
|
|
|
|
|
465
|
|
|
// Indenting BEGIN ... END blocks. |
|
466
|
46 |
|
if ($prev->type === TokenType::Keyword && $prev->keyword === 'BEGIN') { |
|
467
|
2 |
|
$lineEnded = true; |
|
468
|
2 |
|
$blocksIndentation[] = $indent; |
|
469
|
2 |
|
++$indent; |
|
470
|
46 |
|
} elseif ($curr->type === TokenType::Keyword && $curr->keyword === 'END') { |
|
471
|
2 |
|
$lineEnded = true; |
|
472
|
2 |
|
$indent = array_pop($blocksIndentation); |
|
473
|
|
|
} |
|
474
|
|
|
|
|
475
|
|
|
// Formatting fragments delimited by comma. |
|
476
|
46 |
|
if ($prev->type === TokenType::Operator && $prev->value === ',') { |
|
477
|
|
|
// Fragments delimited by a comma are broken into multiple |
|
478
|
|
|
// pieces only if the clause is not inlined or this fragment |
|
479
|
|
|
// is between brackets that are on new line. |
|
480
|
|
|
if ( |
|
481
|
8 |
|
end($blocksLineEndings) === true |
|
482
|
|
|
|| ( |
|
483
|
8 |
|
empty(self::$inlineClauses[$lastClause]) |
|
484
|
8 |
|
&& ! $shortGroup |
|
485
|
8 |
|
&& $this->options['parts_newline'] |
|
486
|
|
|
) |
|
487
|
|
|
) { |
|
488
|
6 |
|
$lineEnded = true; |
|
489
|
|
|
} |
|
490
|
|
|
} |
|
491
|
|
|
|
|
492
|
|
|
// Handling brackets. |
|
493
|
|
|
// Brackets are indented only if the length of the fragment between |
|
494
|
|
|
// them is longer than 30 characters. |
|
495
|
46 |
|
if ($prev->type === TokenType::Operator && $prev->value === '(') { |
|
496
|
12 |
|
$blocksIndentation[] = $indent; |
|
497
|
12 |
|
$shortGroup = true; |
|
498
|
12 |
|
if (static::getGroupLength($list) > 30) { |
|
499
|
2 |
|
++$indent; |
|
500
|
2 |
|
$lineEnded = true; |
|
501
|
2 |
|
$shortGroup = false; |
|
502
|
|
|
} |
|
503
|
|
|
|
|
504
|
12 |
|
$blocksLineEndings[] = $lineEnded; |
|
505
|
46 |
|
} elseif ($curr->type === TokenType::Operator && $curr->value === ')') { |
|
506
|
10 |
|
$indent = array_pop($blocksIndentation); |
|
507
|
10 |
|
$lineEnded |= array_pop($blocksLineEndings); |
|
508
|
10 |
|
$shortGroup = false; |
|
509
|
|
|
} |
|
510
|
|
|
|
|
511
|
|
|
// Adding the token. |
|
512
|
46 |
|
$ret .= $this->toString($prev); |
|
|
|
|
|
|
513
|
|
|
|
|
514
|
|
|
// Finishing the line. |
|
515
|
46 |
|
if ($lineEnded) { |
|
516
|
44 |
|
$ret .= $this->options['line_ending'] . str_repeat($this->options['indentation'], (int) $indent); |
|
517
|
44 |
|
$lineEnded = false; |
|
518
|
|
|
} elseif ( |
|
519
|
46 |
|
$prev->keyword === 'DELIMITER' |
|
520
|
46 |
|
|| ! ( |
|
521
|
46 |
|
($prev->type === TokenType::Operator && ($prev->value === '.' || $prev->value === '(')) |
|
522
|
46 |
|
// No space after . ( |
|
523
|
46 |
|
|| ($curr->type === TokenType::Operator |
|
524
|
46 |
|
&& ($curr->value === '.' || $curr->value === ',' |
|
525
|
46 |
|
|| $curr->value === '(' || $curr->value === ')')) |
|
526
|
46 |
|
// No space before . , ( ) |
|
527
|
46 |
|
|| $curr->type === TokenType::Delimiter && mb_strlen((string) $curr->value, 'UTF-8') < 2 |
|
528
|
46 |
|
) |
|
529
|
|
|
) { |
|
530
|
|
|
// If the line ended, there is no point in adding whitespaces. |
|
531
|
|
|
// Also, some tokens do not have spaces before or after them. |
|
532
|
|
|
// A space after delimiters that are longer than 2 characters. |
|
533
|
26 |
|
$ret .= ' '; |
|
534
|
|
|
} |
|
535
|
|
|
} |
|
536
|
|
|
|
|
537
|
|
|
// Iteration finished, consider current token as previous. |
|
538
|
48 |
|
$prev = $curr; |
|
539
|
|
|
} |
|
540
|
|
|
|
|
541
|
48 |
|
if ($this->options['type'] === 'cli') { |
|
542
|
40 |
|
return $ret . "\x1b[0m"; |
|
543
|
|
|
} |
|
544
|
|
|
|
|
545
|
42 |
|
return $ret; |
|
546
|
|
|
} |
|
547
|
|
|
|
|
548
|
38 |
|
public function escapeConsole(string $string): string |
|
549
|
|
|
{ |
|
550
|
38 |
|
return str_replace( |
|
551
|
38 |
|
[ |
|
552
|
38 |
|
"\x00", |
|
553
|
38 |
|
"\x01", |
|
554
|
38 |
|
"\x02", |
|
555
|
38 |
|
"\x03", |
|
556
|
38 |
|
"\x04", |
|
557
|
38 |
|
"\x05", |
|
558
|
38 |
|
"\x06", |
|
559
|
38 |
|
"\x07", |
|
560
|
38 |
|
"\x08", |
|
561
|
38 |
|
"\x09", |
|
562
|
38 |
|
"\x0A", |
|
563
|
38 |
|
"\x0B", |
|
564
|
38 |
|
"\x0C", |
|
565
|
38 |
|
"\x0D", |
|
566
|
38 |
|
"\x0E", |
|
567
|
38 |
|
"\x0F", |
|
568
|
38 |
|
"\x10", |
|
569
|
38 |
|
"\x11", |
|
570
|
38 |
|
"\x12", |
|
571
|
38 |
|
"\x13", |
|
572
|
38 |
|
"\x14", |
|
573
|
38 |
|
"\x15", |
|
574
|
38 |
|
"\x16", |
|
575
|
38 |
|
"\x17", |
|
576
|
38 |
|
"\x18", |
|
577
|
38 |
|
"\x19", |
|
578
|
38 |
|
"\x1A", |
|
579
|
38 |
|
"\x1B", |
|
580
|
38 |
|
"\x1C", |
|
581
|
38 |
|
"\x1D", |
|
582
|
38 |
|
"\x1E", |
|
583
|
38 |
|
"\x1F", |
|
584
|
38 |
|
], |
|
585
|
38 |
|
[ |
|
586
|
38 |
|
'\x00', |
|
587
|
38 |
|
'\x01', |
|
588
|
38 |
|
'\x02', |
|
589
|
38 |
|
'\x03', |
|
590
|
38 |
|
'\x04', |
|
591
|
38 |
|
'\x05', |
|
592
|
38 |
|
'\x06', |
|
593
|
38 |
|
'\x07', |
|
594
|
38 |
|
'\x08', |
|
595
|
38 |
|
'\x09', |
|
596
|
38 |
|
'\x0A', |
|
597
|
38 |
|
'\x0B', |
|
598
|
38 |
|
'\x0C', |
|
599
|
38 |
|
'\x0D', |
|
600
|
38 |
|
'\x0E', |
|
601
|
38 |
|
'\x0F', |
|
602
|
38 |
|
'\x10', |
|
603
|
38 |
|
'\x11', |
|
604
|
38 |
|
'\x12', |
|
605
|
38 |
|
'\x13', |
|
606
|
38 |
|
'\x14', |
|
607
|
38 |
|
'\x15', |
|
608
|
38 |
|
'\x16', |
|
609
|
38 |
|
'\x17', |
|
610
|
38 |
|
'\x18', |
|
611
|
38 |
|
'\x19', |
|
612
|
38 |
|
'\x1A', |
|
613
|
38 |
|
'\x1B', |
|
614
|
38 |
|
'\x1C', |
|
615
|
38 |
|
'\x1D', |
|
616
|
38 |
|
'\x1E', |
|
617
|
38 |
|
'\x1F', |
|
618
|
38 |
|
], |
|
619
|
38 |
|
$string, |
|
620
|
38 |
|
); |
|
621
|
|
|
} |
|
622
|
|
|
|
|
623
|
|
|
/** |
|
624
|
|
|
* Tries to print the query and returns the result. |
|
625
|
|
|
* |
|
626
|
|
|
* @param Token $token the token to be printed |
|
627
|
|
|
*/ |
|
628
|
46 |
|
public function toString(Token $token): string |
|
629
|
|
|
{ |
|
630
|
46 |
|
$text = $token->token; |
|
631
|
46 |
|
static $prev; |
|
632
|
|
|
|
|
633
|
46 |
|
foreach ($this->options['formats'] as $format) { |
|
634
|
|
|
if ( |
|
635
|
46 |
|
$token->type->value !== $format['type'] || ! (($token->flags & $format['flags']) === $format['flags']) |
|
636
|
|
|
) { |
|
637
|
46 |
|
continue; |
|
638
|
|
|
} |
|
639
|
|
|
|
|
640
|
|
|
// Running transformation function. |
|
641
|
46 |
|
if (! empty($format['function'])) { |
|
642
|
46 |
|
$func = $format['function']; |
|
643
|
46 |
|
$text = $func($text); |
|
644
|
|
|
} |
|
645
|
|
|
|
|
646
|
|
|
// Formatting HTML. |
|
647
|
46 |
|
if ($this->options['type'] === 'html') { |
|
648
|
36 |
|
return '<span ' . $format['html'] . '>' . htmlspecialchars($text, ENT_NOQUOTES) . '</span>'; |
|
649
|
|
|
} |
|
650
|
|
|
|
|
651
|
42 |
|
if ($this->options['type'] === 'cli') { |
|
652
|
38 |
|
if ($prev !== $format['cli']) { |
|
653
|
38 |
|
$prev = $format['cli']; |
|
654
|
|
|
|
|
655
|
38 |
|
return $format['cli'] . $this->escapeConsole($text); |
|
656
|
|
|
} |
|
657
|
|
|
|
|
658
|
10 |
|
return $this->escapeConsole($text); |
|
659
|
|
|
} |
|
660
|
|
|
|
|
661
|
36 |
|
break; |
|
662
|
|
|
} |
|
663
|
|
|
|
|
664
|
36 |
|
if ($this->options['type'] === 'cli') { |
|
665
|
28 |
|
if ($prev !== "\x1b[39m") { |
|
666
|
28 |
|
$prev = "\x1b[39m"; |
|
667
|
|
|
|
|
668
|
28 |
|
return "\x1b[39m" . $this->escapeConsole($text); |
|
669
|
|
|
} |
|
670
|
|
|
|
|
671
|
16 |
|
return $this->escapeConsole($text); |
|
672
|
|
|
} |
|
673
|
|
|
|
|
674
|
36 |
|
if ($this->options['type'] === 'html') { |
|
675
|
28 |
|
return htmlspecialchars($text, ENT_NOQUOTES); |
|
676
|
|
|
} |
|
677
|
|
|
|
|
678
|
36 |
|
return $text; |
|
679
|
|
|
} |
|
680
|
|
|
|
|
681
|
|
|
/** |
|
682
|
|
|
* Formats a query. |
|
683
|
|
|
* |
|
684
|
|
|
* @param string $query The query to be formatted |
|
685
|
|
|
* @param array<string, bool|string|array<int, array<string, int|string>>> $options the formatting options |
|
686
|
|
|
* |
|
687
|
|
|
* @return string the formatted string |
|
688
|
|
|
*/ |
|
689
|
48 |
|
public static function format(string $query, array $options = []): string |
|
690
|
|
|
{ |
|
691
|
48 |
|
$lexer = new Lexer($query); |
|
692
|
48 |
|
$formatter = new self($options); |
|
693
|
|
|
|
|
694
|
48 |
|
return $formatter->formatList($lexer->list); |
|
695
|
|
|
} |
|
696
|
|
|
|
|
697
|
|
|
/** |
|
698
|
|
|
* Computes the length of a group. |
|
699
|
|
|
* |
|
700
|
|
|
* A group is delimited by a pair of brackets. |
|
701
|
|
|
* |
|
702
|
|
|
* @param TokensList $list the list of tokens |
|
703
|
|
|
*/ |
|
704
|
12 |
|
public static function getGroupLength(TokensList $list): int |
|
705
|
|
|
{ |
|
706
|
|
|
/** |
|
707
|
|
|
* The number of opening brackets found. |
|
708
|
|
|
* This counter starts at one because by the time this function called, |
|
709
|
|
|
* the list already advanced one position and the opening bracket was |
|
710
|
|
|
* already parsed. |
|
711
|
|
|
*/ |
|
712
|
12 |
|
$count = 1; |
|
713
|
|
|
|
|
714
|
|
|
/** |
|
715
|
|
|
* The length of this group. |
|
716
|
|
|
*/ |
|
717
|
12 |
|
$length = 0; |
|
718
|
|
|
|
|
719
|
12 |
|
for ($idx = $list->idx; $idx < $list->count; ++$idx) { |
|
720
|
|
|
// Counting the brackets. |
|
721
|
12 |
|
if ($list->tokens[$idx]->type === TokenType::Operator) { |
|
722
|
12 |
|
if ($list->tokens[$idx]->value === '(') { |
|
723
|
2 |
|
++$count; |
|
724
|
12 |
|
} elseif ($list->tokens[$idx]->value === ')') { |
|
725
|
12 |
|
--$count; |
|
726
|
12 |
|
if ($count === 0) { |
|
727
|
12 |
|
break; |
|
728
|
|
|
} |
|
729
|
|
|
} |
|
730
|
|
|
} |
|
731
|
|
|
|
|
732
|
|
|
// Keeping track of this group's length. |
|
733
|
10 |
|
$length += mb_strlen((string) $list->tokens[$idx]->value, 'UTF-8'); |
|
734
|
|
|
} |
|
735
|
|
|
|
|
736
|
12 |
|
return $length; |
|
737
|
|
|
} |
|
738
|
|
|
|
|
739
|
|
|
/** |
|
740
|
|
|
* Checks if a token is a statement or a clause inside a statement. |
|
741
|
|
|
* |
|
742
|
|
|
* @param Token $token the token to be checked |
|
743
|
|
|
* |
|
744
|
|
|
* @psalm-return 1|2|false |
|
745
|
|
|
*/ |
|
746
|
46 |
|
public static function isClause(Token $token): int|false |
|
747
|
|
|
{ |
|
748
|
|
|
if ( |
|
749
|
46 |
|
($token->type === TokenType::Keyword && isset(Parser::STATEMENT_PARSERS[$token->keyword])) |
|
750
|
46 |
|
|| ($token->type === TokenType::None && strtoupper($token->token) === 'DELIMITER') |
|
751
|
|
|
) { |
|
752
|
44 |
|
return 2; |
|
753
|
|
|
} |
|
754
|
|
|
|
|
755
|
|
|
if ( |
|
756
|
46 |
|
$token->type === TokenType::Keyword && ( |
|
757
|
46 |
|
in_array($token->keyword, self::FORMATTERS, true) || isset(Parser::KEYWORD_PARSERS[$token->keyword]) |
|
758
|
|
|
) |
|
759
|
|
|
) { |
|
760
|
20 |
|
return 1; |
|
761
|
|
|
} |
|
762
|
|
|
|
|
763
|
46 |
|
return false; |
|
764
|
|
|
} |
|
765
|
|
|
} |
|
766
|
|
|
|