bug #52443 [Yaml] Fix uid binary parsing (mRoca)

fabpot · fabpot · commit a432b12bf4a7 · 2023-11-05T20:14:19.000+01:00
This PR was merged into the 5.4 branch. Discussion ---------- [Yaml] Fix uid binary parsing | Q | A | ------------- | --- | Branch? | 5.4 | Bug fix? | yes | New feature? | no | Deprecations? | no | License | MIT Currently, on SF 5.4 & 6.4, a YAML file containing `{ uid: !!binary Ju0Yh+uqSXOagJZFTlUt8g== }` cannot be parsed because of a `preg_match` error: `Malformed UTF-8 characters, possibly incorrectly encoded`. This is because `preg_match` should not be run over a binary string, but this specific binary string matches the following tests : `\is_string($value) && '' !== $value && '&' === $value[0]` (the string starts with `0x26` , which is the UTF-8 `&` char). Commits ------- ce3e282 [Yaml] Fix uid binary parsing
diff --git a/src/Symfony/Component/Yaml/Inline.php b/src/Symfony/Component/Yaml/Inline.php
@@ -531,7 +531,7 @@ private static function parseMapping(string $mapping, int $flags, int &$i = 0, a
                         if ('<<' === $key) {
                             $output += $value;
                         } elseif ($allowOverwrite || !isset($output[$key])) {
-                            if (!$isValueQuoted && \is_string($value) && '' !== $value && '&' === $value[0] && Parser::preg_match(Parser::REFERENCE_PATTERN, $value, $matches)) {
+                            if (!$isValueQuoted && \is_string($value) && '' !== $value && '&' === $value[0] && !self::isBinaryString($value) && Parser::preg_match(Parser::REFERENCE_PATTERN, $value, $matches)) {
                                 $references[$matches['ref']] = $matches['value'];
                                 $value = $matches['value'];
                             }
diff --git a/src/Symfony/Component/Yaml/Tests/InlineTest.php b/src/Symfony/Component/Yaml/Tests/InlineTest.php
@@ -368,6 +368,9 @@ public static function getTestsForParse()
 
             ['[foo, bar: { foo: bar }]', ['foo', '1' => ['bar' => ['foo' => 'bar']]]],
             ['[foo, \'@foo.baz\', { \'%foo%\': \'foo is %foo%\', bar: \'%foo%\' }, true, \'@service_container\']', ['foo', '@foo.baz', ['%foo%' => 'foo is %foo%', 'bar' => '%foo%'], true, '@service_container']],
+
+            // Binary string not utf8-compliant but starting with and utf8-equivalent "&" character
+            ['{ uid: !!binary Ju0Yh+uqSXOagJZFTlUt8g== }', ['uid' => hex2bin('26ed1887ebaa49739a8096454e552df2')]],
         ];
     }
 

Original file line number	Diff line number	Diff line change
`@@ -531,7 +531,7 @@ private static function parseMapping(string $mapping, int $flags, int &$i = 0, a`
`531`	`531`	`if ('<<' === $key) {`
`532`	`532`	`$output += $value;`
`533`	`533`	`} elseif ($allowOverwrite \|\| !isset($output[$key])) {`
`534`		`- if (!$isValueQuoted && \is_string($value) && '' !== $value && '&' === $value[0] && Parser::preg_match(Parser::REFERENCE_PATTERN, $value, $matches)) {`
	`534`	`+ if (!$isValueQuoted && \is_string($value) && '' !== $value && '&' === $value[0] && !self::isBinaryString($value) && Parser::preg_match(Parser::REFERENCE_PATTERN, $value, $matches)) {`
`535`	`535`	`$references[$matches['ref']] = $matches['value'];`
`536`	`536`	`$value = $matches['value'];`
`537`	`537`	`}`
Original file line number	Diff line number	Diff line change
`@@ -368,6 +368,9 @@ public static function getTestsForParse()`
`368`	`368`
`369`	`369`	`['[foo, bar: { foo: bar }]', ['foo', '1' => ['bar' => ['foo' => 'bar']]]],`
`370`	`370`	`['[foo, \'@foo.baz\', { \'%foo%\': \'foo is %foo%\', bar: \'%foo%\' }, true, \'@service_container\']', ['foo', '@foo.baz', ['%foo%' => 'foo is %foo%', 'bar' => '%foo%'], true, '@service_container']],`
	`371`	`+`
	`372`	`+ // Binary string not utf8-compliant but starting with and utf8-equivalent "&" character`
	`373`	`+ ['{ uid: !!binary Ju0Yh+uqSXOagJZFTlUt8g== }', ['uid' => hex2bin('26ed1887ebaa49739a8096454e552df2')]],`
`371`	`374`	`];`
`372`	`375`	`}`
`373`	`376`