|
| 1 | +<?php |
| 2 | + |
| 3 | +declare(strict_types=1); |
| 4 | + |
| 5 | +namespace TypeLang\PhpDoc\Tests\Parser\Comment; |
| 6 | + |
| 7 | +use PHPUnit\Framework\Attributes\DataProvider; |
| 8 | +use PHPUnit\Framework\Attributes\Test; |
| 9 | +use TypeLang\PhpDoc\Parser\Comment\CommentParserInterface; |
| 10 | +use TypeLang\PhpDoc\Parser\Comment\RegexCommentParser; |
| 11 | +use TypeLang\PhpDoc\Parser\Comment\Segment; |
| 12 | +use TypeLang\PhpDoc\Tests\TestCase; |
| 13 | + |
| 14 | +final class RegexCommentParserTest extends TestCase |
| 15 | +{ |
| 16 | + /** |
| 17 | + * All implementations of {@see CommentParserInterface} to be verified. |
| 18 | + * |
| 19 | + * @return iterable<string, array{CommentParserInterface}> |
| 20 | + */ |
| 21 | + public static function provideParsers(): iterable |
| 22 | + { |
| 23 | + foreach (self::createParsers() as $name => $parser) { |
| 24 | + yield $name => [$parser]; |
| 25 | + } |
| 26 | + } |
| 27 | + |
| 28 | + /** |
| 29 | + * Cross product of every parser with every "unwrapped" (not a `/** */` |
| 30 | + * comment) input. |
| 31 | + * |
| 32 | + * @return iterable<string, array{CommentParserInterface, string}> |
| 33 | + */ |
| 34 | + public static function provideUnwrappedInputs(): iterable |
| 35 | + { |
| 36 | + $inputs = [ |
| 37 | + 'plain text' => 'Just some text', |
| 38 | + 'empty string' => '', |
| 39 | + 'whitespace only' => ' ', |
| 40 | + 'opening not at start' => 'abc /** x */', |
| 41 | + ]; |
| 42 | + |
| 43 | + foreach (self::createParsers() as $pName => $parser) { |
| 44 | + foreach ($inputs as $iName => $input) { |
| 45 | + yield $pName . ': ' . $iName => [$parser, $input]; |
| 46 | + } |
| 47 | + } |
| 48 | + } |
| 49 | + |
| 50 | + /** |
| 51 | + * Cross product of every parser with every wrapped multi-line comment that |
| 52 | + * uses `\n` line endings (so byte offsets are unambiguous). |
| 53 | + * |
| 54 | + * @return iterable<string, array{CommentParserInterface, string}> |
| 55 | + */ |
| 56 | + public static function provideWrappedInputs(): iterable |
| 57 | + { |
| 58 | + $inputs = [ |
| 59 | + 'multiline description' => self::comment( |
| 60 | + '/**', |
| 61 | + ' * Example line 1', |
| 62 | + ' *', |
| 63 | + ' * @tag1 type Description of tag1', |
| 64 | + ' */', |
| 65 | + ), |
| 66 | + 'single line' => '/** Foo bar */', |
| 67 | + 'leading whitespace before opening' => " \n" . self::comment( |
| 68 | + '/**', |
| 69 | + ' * Hi', |
| 70 | + ' */', |
| 71 | + ), |
| 72 | + 'multiline tag' => self::comment( |
| 73 | + '/**', |
| 74 | + ' * @param int $a first', |
| 75 | + ' * second line', |
| 76 | + ' * @return void', |
| 77 | + ' */', |
| 78 | + ), |
| 79 | + 'body without star prefix' => self::comment( |
| 80 | + '/**', |
| 81 | + 'Plain', |
| 82 | + '*/', |
| 83 | + ), |
| 84 | + ]; |
| 85 | + |
| 86 | + foreach (self::createParsers() as $pName => $parser) { |
| 87 | + foreach ($inputs as $iName => $input) { |
| 88 | + yield $pName . ': ' . $iName => [$parser, $input]; |
| 89 | + } |
| 90 | + } |
| 91 | + } |
| 92 | + |
| 93 | + #[Test] |
| 94 | + #[DataProvider('provideParsers')] |
| 95 | + public function parseReturnsIterable(CommentParserInterface $parser): void |
| 96 | + { |
| 97 | + $this->assertIsIterable($parser->parse('/** example */')); |
| 98 | + } |
| 99 | + |
| 100 | + #[Test] |
| 101 | + #[DataProvider('provideWrappedInputs')] |
| 102 | + public function parseYieldsOnlySegmentInstances(CommentParserInterface $parser, string $input): void |
| 103 | + { |
| 104 | + $this->assertContainsOnlyInstancesOf(Segment::class, self::segments($parser->parse($input))); |
| 105 | + } |
| 106 | + |
| 107 | + #[Test] |
| 108 | + #[DataProvider('provideUnwrappedInputs')] |
| 109 | + public function unwrappedInputBecomesSingleSegment(CommentParserInterface $parser, string $input): void |
| 110 | + { |
| 111 | + $segments = self::segments($parser->parse($input)); |
| 112 | + |
| 113 | + $this->assertCount(1, $segments); |
| 114 | + $this->assertSame($input, $segments[0]->text); |
| 115 | + $this->assertSame(0, $segments[0]->offset); |
| 116 | + } |
| 117 | + |
| 118 | + #[Test] |
| 119 | + #[DataProvider('provideWrappedInputs')] |
| 120 | + public function segmentOffsetPointsToItsTextInSource(CommentParserInterface $parser, string $input): void |
| 121 | + { |
| 122 | + foreach (self::segments($parser->parse($input)) as $segment) { |
| 123 | + $this->assertSame( |
| 124 | + \substr($input, $segment->offset, \strlen($segment->text)), |
| 125 | + $segment->text, |
| 126 | + 'A segment text must be a verbatim slice of the source at its offset', |
| 127 | + ); |
| 128 | + } |
| 129 | + } |
| 130 | + |
| 131 | + #[Test] |
| 132 | + #[DataProvider('provideWrappedInputs')] |
| 133 | + public function segmentsAreReturnedInSourceOrder(CommentParserInterface $parser, string $input): void |
| 134 | + { |
| 135 | + $previous = -1; |
| 136 | + |
| 137 | + foreach (self::segments($parser->parse($input)) as $segment) { |
| 138 | + $this->assertGreaterThan($previous, $segment->offset); |
| 139 | + $previous = $segment->offset; |
| 140 | + } |
| 141 | + } |
| 142 | + |
| 143 | + #[Test] |
| 144 | + #[DataProvider('provideParsers')] |
| 145 | + public function wrappedCommentYieldsSignificantLinesInOrder(CommentParserInterface $parser): void |
| 146 | + { |
| 147 | + $input = self::comment( |
| 148 | + '/**', |
| 149 | + ' * Example line 1', |
| 150 | + ' *', |
| 151 | + ' * @tag1 type Description of tag1', |
| 152 | + ' */', |
| 153 | + ); |
| 154 | + |
| 155 | + $this->assertSame( |
| 156 | + ['Example line 1', '@tag1 type Description of tag1'], |
| 157 | + self::trimmedTexts($parser->parse($input)), |
| 158 | + ); |
| 159 | + } |
| 160 | + |
| 161 | + #[Test] |
| 162 | + #[DataProvider('provideParsers')] |
| 163 | + public function blankCommentLinesAreSkipped(CommentParserInterface $parser): void |
| 164 | + { |
| 165 | + $input = self::comment( |
| 166 | + '/**', |
| 167 | + ' * first', |
| 168 | + ' *', |
| 169 | + ' * ', |
| 170 | + ' * second', |
| 171 | + ' */', |
| 172 | + ); |
| 173 | + |
| 174 | + $this->assertSame(['first', 'second'], self::trimmedTexts($parser->parse($input))); |
| 175 | + } |
| 176 | + |
| 177 | + #[Test] |
| 178 | + #[DataProvider('provideParsers')] |
| 179 | + public function singleLineCommentYieldsSingleSegment(CommentParserInterface $parser): void |
| 180 | + { |
| 181 | + $this->assertSame(['Foo bar'], self::trimmedTexts($parser->parse('/** Foo bar */'))); |
| 182 | + } |
| 183 | + |
| 184 | + #[Test] |
| 185 | + #[DataProvider('provideParsers')] |
| 186 | + public function leadingWhitespaceBeforeOpeningIsTreatedAsComment(CommentParserInterface $parser): void |
| 187 | + { |
| 188 | + $input = " \n" . self::comment( |
| 189 | + '/**', |
| 190 | + ' * Hi', |
| 191 | + ' */', |
| 192 | + ); |
| 193 | + |
| 194 | + $this->assertSame(['Hi'], self::trimmedTexts($parser->parse($input))); |
| 195 | + } |
| 196 | + |
| 197 | + #[Test] |
| 198 | + #[DataProvider('provideParsers')] |
| 199 | + public function tagContinuationLinesBecomeSeparateSegments(CommentParserInterface $parser): void |
| 200 | + { |
| 201 | + $input = self::comment( |
| 202 | + '/**', |
| 203 | + ' * @param int $a first', |
| 204 | + ' * second line', |
| 205 | + ' * @return void', |
| 206 | + ' */', |
| 207 | + ); |
| 208 | + |
| 209 | + $this->assertSame( |
| 210 | + ['@param int $a first', 'second line', '@return void'], |
| 211 | + self::trimmedTexts($parser->parse($input)), |
| 212 | + ); |
| 213 | + } |
| 214 | + |
| 215 | + /** |
| 216 | + * Cross product of every parser with line-ending fixtures. Each fixture |
| 217 | + * declares the EXACT segments (verbatim text including the trailing line |
| 218 | + * terminator, plus its byte offset) the parser is expected to produce. |
| 219 | + * |
| 220 | + * @return iterable<string, array{CommentParserInterface, string, list<array{string, int<0, max>}>}> |
| 221 | + */ |
| 222 | + public static function provideLineEndingCases(): iterable |
| 223 | + { |
| 224 | + $cases = [ |
| 225 | + 'LF line endings' => [ |
| 226 | + "/**\n * a\n * b\n */", |
| 227 | + [["a\n", 7], ["b\n", 12]], |
| 228 | + ], |
| 229 | + 'CRLF line endings' => [ |
| 230 | + "/**\r\n * a\r\n * b\r\n */", |
| 231 | + [["a\r\n", 8], ["b\r\n", 14]], |
| 232 | + ], |
| 233 | + 'CRLF tag line' => [ |
| 234 | + "/**\r\n * @param int \$x d\r\n */", |
| 235 | + [["@param int \$x d\r\n", 8]], |
| 236 | + ], |
| 237 | + 'mixed LF and CRLF' => [ |
| 238 | + "/**\r\n * a\n * b\r\n */", |
| 239 | + [["a\n", 8], ["b\r\n", 13]], |
| 240 | + ], |
| 241 | + ]; |
| 242 | + |
| 243 | + foreach (self::createParsers() as $pName => $parser) { |
| 244 | + foreach ($cases as $cName => [$input, $expected]) { |
| 245 | + yield $pName . ': ' . $cName => [$parser, $input, $expected]; |
| 246 | + } |
| 247 | + } |
| 248 | + } |
| 249 | + |
| 250 | + /** |
| 251 | + * The text of every segment retains its own trailing line terminator |
| 252 | + * verbatim (`\n` or `\r\n`), and offsets are byte offsets, so a `\r\n` |
| 253 | + * line shifts subsequent offsets by the extra carriage return byte. |
| 254 | + * |
| 255 | + * @param list<array{string, int<0, max>}> $expected |
| 256 | + */ |
| 257 | + #[Test] |
| 258 | + #[DataProvider('provideLineEndingCases')] |
| 259 | + public function preservesLineEndingsVerbatim(CommentParserInterface $parser, string $input, array $expected): void |
| 260 | + { |
| 261 | + $segments = self::segments($parser->parse($input)); |
| 262 | + |
| 263 | + $this->assertSame(\array_column($expected, 0), \array_map( |
| 264 | + static fn(Segment $segment): string => $segment->text, |
| 265 | + $segments, |
| 266 | + )); |
| 267 | + |
| 268 | + $this->assertSame(\array_column($expected, 1), \array_map( |
| 269 | + static fn(Segment $segment): int => $segment->offset, |
| 270 | + $segments, |
| 271 | + )); |
| 272 | + } |
| 273 | + |
| 274 | + /** |
| 275 | + * @return iterable<string, CommentParserInterface> |
| 276 | + */ |
| 277 | + private static function createParsers(): iterable |
| 278 | + { |
| 279 | + yield 'RegexCommentParser' => new RegexCommentParser(); |
| 280 | + } |
| 281 | + |
| 282 | + /** |
| 283 | + * Builds a comment source from the given lines using `\n` separators. |
| 284 | + */ |
| 285 | + private static function comment(string ...$lines): string |
| 286 | + { |
| 287 | + return \implode("\n", $lines); |
| 288 | + } |
| 289 | + |
| 290 | + /** |
| 291 | + * Normalizes the {@see CommentParserInterface::parse()} result (which may be |
| 292 | + * any iterable) into a positionally indexed list of segments. |
| 293 | + * |
| 294 | + * @return list<Segment> |
| 295 | + */ |
| 296 | + private static function segments(iterable $result): array |
| 297 | + { |
| 298 | + return \is_array($result) |
| 299 | + ? \array_values($result) |
| 300 | + : \iterator_to_array($result, false); |
| 301 | + } |
| 302 | + |
| 303 | + /** |
| 304 | + * @return list<string> |
| 305 | + */ |
| 306 | + private static function trimmedTexts(iterable $result): array |
| 307 | + { |
| 308 | + return \array_map( |
| 309 | + static fn(Segment $segment): string => \trim($segment->text), |
| 310 | + self::segments($result), |
| 311 | + ); |
| 312 | + } |
| 313 | +} |
0 commit comments