first commit

vendor/symfony/css-selector/Parser/Tokenizer/Tokenizer.php

@@ -0,0 +1,73 @@
+<?php
+
+/*
+ * This file is part of the Symfony package.
+ *
+ * (c) Fabien Potencier <fabien@symfony.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+namespace Symfony\Component\CssSelector\Parser\Tokenizer;
+
+use Symfony\Component\CssSelector\Parser\Handler;
+use Symfony\Component\CssSelector\Parser\Reader;
+use Symfony\Component\CssSelector\Parser\Token;
+use Symfony\Component\CssSelector\Parser\TokenStream;
+
+/**
+ * CSS selector tokenizer.
+ *
+ * This component is a port of the Python cssselect library,
+ * which is copyright Ian Bicking, @see https://github.com/SimonSapin/cssselect.
+ *
+ * @author Jean-François Simon <jeanfrancois.simon@sensiolabs.com>
+ *
+ * @internal
+ */
+class Tokenizer
+{
+    /**
+     * @var Handler\HandlerInterface[]
+     */
+    private array $handlers;
+
+    public function __construct()
+    {
+        $patterns = new TokenizerPatterns();
+        $escaping = new TokenizerEscaping($patterns);
+
+        $this->handlers = [
+            new Handler\WhitespaceHandler(),
+            new Handler\IdentifierHandler($patterns, $escaping),
+            new Handler\HashHandler($patterns, $escaping),
+            new Handler\StringHandler($patterns, $escaping),
+            new Handler\NumberHandler($patterns),
+            new Handler\CommentHandler(),
+        ];
+    }
+
+    /**
+     * Tokenize selector source code.
+     */
+    public function tokenize(Reader $reader): TokenStream
+    {
+        $stream = new TokenStream();
+
+        while (!$reader->isEOF()) {
+            foreach ($this->handlers as $handler) {
+                if ($handler->handle($reader, $stream)) {
+                    continue 2;
+                }
+            }
+
+            $stream->push(new Token(Token::TYPE_DELIMITER, $reader->getSubstring(1), $reader->getPosition()));
+            $reader->moveForward(1);
+        }
+
+        return $stream
+            ->push(new Token(Token::TYPE_FILE_END, null, $reader->getPosition()))
+            ->freeze();
+    }
+}

vendor/symfony/css-selector/Parser/Tokenizer/TokenizerEscaping.php

@@ -0,0 +1,65 @@
+<?php
+
+/*
+ * This file is part of the Symfony package.
+ *
+ * (c) Fabien Potencier <fabien@symfony.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+namespace Symfony\Component\CssSelector\Parser\Tokenizer;
+
+/**
+ * CSS selector tokenizer escaping applier.
+ *
+ * This component is a port of the Python cssselect library,
+ * which is copyright Ian Bicking, @see https://github.com/SimonSapin/cssselect.
+ *
+ * @author Jean-François Simon <jeanfrancois.simon@sensiolabs.com>
+ *
+ * @internal
+ */
+class TokenizerEscaping
+{
+    private TokenizerPatterns $patterns;
+
+    public function __construct(TokenizerPatterns $patterns)
+    {
+        $this->patterns = $patterns;
+    }
+
+    public function escapeUnicode(string $value): string
+    {
+        $value = $this->replaceUnicodeSequences($value);
+
+        return preg_replace($this->patterns->getSimpleEscapePattern(), '$1', $value);
+    }
+
+    public function escapeUnicodeAndNewLine(string $value): string
+    {
+        $value = preg_replace($this->patterns->getNewLineEscapePattern(), '', $value);
+
+        return $this->escapeUnicode($value);
+    }
+
+    private function replaceUnicodeSequences(string $value): string
+    {
+        return preg_replace_callback($this->patterns->getUnicodeEscapePattern(), function ($match) {
+            $c = hexdec($match[1]);
+
+            if (0x80 > $c %= 0x200000) {
+                return \chr($c);
+            }
+            if (0x800 > $c) {
+                return \chr(0xC0 | $c >> 6).\chr(0x80 | $c & 0x3F);
+            }
+            if (0x10000 > $c) {
+                return \chr(0xE0 | $c >> 12).\chr(0x80 | $c >> 6 & 0x3F).\chr(0x80 | $c & 0x3F);
+            }
+
+            return '';
+        }, $value);
+    }
+}

vendor/symfony/css-selector/Parser/Tokenizer/TokenizerPatterns.php

@@ -0,0 +1,89 @@
+<?php
+
+/*
+ * This file is part of the Symfony package.
+ *
+ * (c) Fabien Potencier <fabien@symfony.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+namespace Symfony\Component\CssSelector\Parser\Tokenizer;
+
+/**
+ * CSS selector tokenizer patterns builder.
+ *
+ * This component is a port of the Python cssselect library,
+ * which is copyright Ian Bicking, @see https://github.com/SimonSapin/cssselect.
+ *
+ * @author Jean-François Simon <jeanfrancois.simon@sensiolabs.com>
+ *
+ * @internal
+ */
+class TokenizerPatterns
+{
+    private string $unicodeEscapePattern;
+    private string $simpleEscapePattern;
+    private string $newLineEscapePattern;
+    private string $escapePattern;
+    private string $stringEscapePattern;
+    private string $nonAsciiPattern;
+    private string $nmCharPattern;
+    private string $nmStartPattern;
+    private string $identifierPattern;
+    private string $hashPattern;
+    private string $numberPattern;
+    private string $quotedStringPattern;
+
+    public function __construct()
+    {
+        $this->unicodeEscapePattern = '\\\\([0-9a-f]{1,6})(?:\r\n|[ \n\r\t\f])?';
+        $this->simpleEscapePattern = '\\\\(.)';
+        $this->newLineEscapePattern = '\\\\(?:\n|\r\n|\r|\f)';
+        $this->escapePattern = $this->unicodeEscapePattern.'|\\\\[^\n\r\f0-9a-f]';
+        $this->stringEscapePattern = $this->newLineEscapePattern.'|'.$this->escapePattern;
+        $this->nonAsciiPattern = '[^\x00-\x7F]';
+        $this->nmCharPattern = '[_a-z0-9-]|'.$this->escapePattern.'|'.$this->nonAsciiPattern;
+        $this->nmStartPattern = '[_a-z]|'.$this->escapePattern.'|'.$this->nonAsciiPattern;
+        $this->identifierPattern = '-?(?:'.$this->nmStartPattern.')(?:'.$this->nmCharPattern.')*';
+        $this->hashPattern = '#((?:'.$this->nmCharPattern.')+)';
+        $this->numberPattern = '[+-]?(?:[0-9]*\.[0-9]+|[0-9]+)';
+        $this->quotedStringPattern = '([^\n\r\f\\\\%s]|'.$this->stringEscapePattern.')*';
+    }
+
+    public function getNewLineEscapePattern(): string
+    {
+        return '~'.$this->newLineEscapePattern.'~';
+    }
+
+    public function getSimpleEscapePattern(): string
+    {
+        return '~'.$this->simpleEscapePattern.'~';
+    }
+
+    public function getUnicodeEscapePattern(): string
+    {
+        return '~'.$this->unicodeEscapePattern.'~i';
+    }
+
+    public function getIdentifierPattern(): string
+    {
+        return '~^'.$this->identifierPattern.'~i';
+    }
+
+    public function getHashPattern(): string
+    {
+        return '~^'.$this->hashPattern.'~i';
+    }
+
+    public function getNumberPattern(): string
+    {
+        return '~^'.$this->numberPattern.'~';
+    }
+
+    public function getQuotedStringPattern(string $quote): string
+    {
+        return '~^'.sprintf($this->quotedStringPattern, $quote).'~i';
+    }
+}