bug #48771 [CssSelector] Fix escape patterns (fancyweb)

fabpot · fabpot · commit e7ec8a6373e1 · 2022-12-23T16:21:21.000+01:00
This PR was merged into the 5.4 branch. Discussion ---------- [CssSelector] Fix escape patterns | Q | A | ------------- | --- | Branch? | 5.4 | Bug fix? | yes | New feature? | no | Deprecations? | no | Tickets | #46682 and #44516 | License | MIT | Doc PR | - I rechecked the original Python source code we "borrowed" (https://github.com/scrapy/cssselect/blob/master/cssselect/parser.py) and fixed some issues. Commits ------- 7de8f99 [CssSelector] Fix escape patterns
diff --git a/src/Symfony/Component/CssSelector/Parser/Tokenizer/TokenizerPatterns.php b/src/Symfony/Component/CssSelector/Parser/Tokenizer/TokenizerPatterns.php
@@ -49,22 +49,22 @@ public function __construct()
         $this->identifierPattern = '-?(?:'.$this->nmStartPattern.')(?:'.$this->nmCharPattern.')*';
         $this->hashPattern = '#((?:'.$this->nmCharPattern.')+)';
         $this->numberPattern = '[+-]?(?:[0-9]*\.[0-9]+|[0-9]+)';
-        $this->quotedStringPattern = '([^\n\r\f%s]|'.$this->stringEscapePattern.')*';
+        $this->quotedStringPattern = '([^\n\r\f\\\\%s]|'.$this->stringEscapePattern.')*';
     }
 
     public function getNewLineEscapePattern(): string
     {
-        return '~^'.$this->newLineEscapePattern.'~';
+        return '~'.$this->newLineEscapePattern.'~';
     }
 
     public function getSimpleEscapePattern(): string
     {
-        return '~^'.$this->simpleEscapePattern.'~';
+        return '~'.$this->simpleEscapePattern.'~';
     }
 
     public function getUnicodeEscapePattern(): string
     {
-        return '~^'.$this->unicodeEscapePattern.'~i';
+        return '~'.$this->unicodeEscapePattern.'~i';
     }
 
     public function getIdentifierPattern(): string
diff --git a/src/Symfony/Component/CssSelector/Tests/Parser/ParserTest.php b/src/Symfony/Component/CssSelector/Tests/Parser/ParserTest.php
@@ -138,6 +138,16 @@ public function getParserTestData()
             ['div:not(div.foo)', ['Negation[Element[div]:not(Class[Element[div].foo])]']],
             ['td ~ th', ['CombinedSelector[Element[td] ~ Element[th]]']],
             ['.foo[data-bar][data-baz=0]', ["Attribute[Attribute[Class[Element[*].foo][data-bar]][data-baz = '0']]"]],
+            ['div#foo\.bar', ['Hash[Element[div]#foo.bar]']],
+            ['div.w-1\/3', ['Class[Element[div].w-1/3]']],
+            ['#test\:colon', ['Hash[Element[*]#test:colon]']],
+            [".a\xc1b", ["Class[Element[*].a\xc1b]"]],
+            // unicode escape: \22 == "
+            ['*[aval="\'\22\'"]', ['Attribute[Element[*][aval = \'\'"\'\']]']],
+            ['*[aval="\'\22 2\'"]', ['Attribute[Element[*][aval = \'\'"2\'\']]']],
+            // unicode escape: \20 ==  (space)
+            ['*[aval="\'\20  \'"]', ['Attribute[Element[*][aval = \'\'  \'\']]']],
+            ["*[aval=\"'\\20\r\n '\"]", ['Attribute[Element[*][aval = \'\'  \'\']]']],
         ];
     }
 

Original file line number	Diff line number	Diff line change
`@@ -49,22 +49,22 @@ public function __construct()`
`49`	`49`	`$this->identifierPattern = '-?(?:'.$this->nmStartPattern.')(?:'.$this->nmCharPattern.')*';`
`50`	`50`	`$this->hashPattern = '#((?:'.$this->nmCharPattern.')+)';`
`51`	`51`	`$this->numberPattern = '[+-]?(?:[0-9]*\.[0-9]+\|[0-9]+)';`
`52`		`- $this->quotedStringPattern = '([^\n\r\f%s]\|'.$this->stringEscapePattern.')*';`
	`52`	`+ $this->quotedStringPattern = '([^\n\r\f\\\\%s]\|'.$this->stringEscapePattern.')*';`
`53`	`53`	`}`
`54`	`54`
`55`	`55`	`public function getNewLineEscapePattern(): string`
`56`	`56`	`{`
`57`		`- return '~^'.$this->newLineEscapePattern.'~';`
	`57`	`+ return '~'.$this->newLineEscapePattern.'~';`
`58`	`58`	`}`
`59`	`59`
`60`	`60`	`public function getSimpleEscapePattern(): string`
`61`	`61`	`{`
`62`		`- return '~^'.$this->simpleEscapePattern.'~';`
	`62`	`+ return '~'.$this->simpleEscapePattern.'~';`
`63`	`63`	`}`
`64`	`64`
`65`	`65`	`public function getUnicodeEscapePattern(): string`
`66`	`66`	`{`
`67`		`- return '~^'.$this->unicodeEscapePattern.'~i';`
	`67`	`+ return '~'.$this->unicodeEscapePattern.'~i';`
`68`	`68`	`}`
`69`	`69`
`70`	`70`	`public function getIdentifierPattern(): string`