Chris@0: <?php
Chris@0: /*
Chris@0:  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
Chris@0:  * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
Chris@0:  * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
Chris@0:  * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
Chris@0:  * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
Chris@0:  * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
Chris@0:  * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
Chris@0:  * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
Chris@0:  * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
Chris@0:  * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
Chris@0:  * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
Chris@0:  *
Chris@0:  * This software consists of voluntary contributions made by many individuals
Chris@0:  * and is licensed under the MIT license. For more information, see
Chris@0:  * <http://www.doctrine-project.org>.
Chris@0:  */
Chris@0: 
Chris@0: namespace Doctrine\Common\Annotations;
Chris@0: 
Chris@0: /**
Chris@0:  * Parses a file for namespaces/use/class declarations.
Chris@0:  *
Chris@0:  * @author Fabien Potencier <fabien@symfony.com>
Chris@0:  * @author Christian Kaps <christian.kaps@mohiva.com>
Chris@0:  */
Chris@0: class TokenParser
Chris@0: {
Chris@0:     /**
Chris@0:      * The token list.
Chris@0:      *
Chris@0:      * @var array
Chris@0:      */
Chris@0:     private $tokens;
Chris@0: 
Chris@0:     /**
Chris@0:      * The number of tokens.
Chris@0:      *
Chris@0:      * @var int
Chris@0:      */
Chris@0:     private $numTokens;
Chris@0: 
Chris@0:     /**
Chris@0:      * The current array pointer.
Chris@0:      *
Chris@0:      * @var int
Chris@0:      */
Chris@0:     private $pointer = 0;
Chris@0: 
Chris@0:     /**
Chris@0:      * @param string $contents
Chris@0:      */
Chris@0:     public function __construct($contents)
Chris@0:     {
Chris@0:         $this->tokens = token_get_all($contents);
Chris@0: 
Chris@0:         // The PHP parser sets internal compiler globals for certain things. Annoyingly, the last docblock comment it
Chris@0:         // saw gets stored in doc_comment. When it comes to compile the next thing to be include()d this stored
Chris@0:         // doc_comment becomes owned by the first thing the compiler sees in the file that it considers might have a
Chris@0:         // docblock. If the first thing in the file is a class without a doc block this would cause calls to
Chris@0:         // getDocBlock() on said class to return our long lost doc_comment. Argh.
Chris@0:         // To workaround, cause the parser to parse an empty docblock. Sure getDocBlock() will return this, but at least
Chris@0:         // it's harmless to us.
Chris@0:         token_get_all("<?php\n/**\n *\n */");
Chris@0: 
Chris@0:         $this->numTokens = count($this->tokens);
Chris@0:     }
Chris@0: 
Chris@0:     /**
Chris@0:      * Gets the next non whitespace and non comment token.
Chris@0:      *
Chris@0:      * @param boolean $docCommentIsComment If TRUE then a doc comment is considered a comment and skipped.
Chris@0:      *                                     If FALSE then only whitespace and normal comments are skipped.
Chris@0:      *
Chris@0:      * @return array|null The token if exists, null otherwise.
Chris@0:      */
Chris@0:     public function next($docCommentIsComment = TRUE)
Chris@0:     {
Chris@0:         for ($i = $this->pointer; $i < $this->numTokens; $i++) {
Chris@0:             $this->pointer++;
Chris@0:             if ($this->tokens[$i][0] === T_WHITESPACE ||
Chris@0:                 $this->tokens[$i][0] === T_COMMENT ||
Chris@0:                 ($docCommentIsComment && $this->tokens[$i][0] === T_DOC_COMMENT)) {
Chris@0: 
Chris@0:                 continue;
Chris@0:             }
Chris@0: 
Chris@0:             return $this->tokens[$i];
Chris@0:         }
Chris@0: 
Chris@0:         return null;
Chris@0:     }
Chris@0: 
Chris@0:     /**
Chris@0:      * Parses a single use statement.
Chris@0:      *
Chris@0:      * @return array A list with all found class names for a use statement.
Chris@0:      */
Chris@0:     public function parseUseStatement()
Chris@0:     {
Chris@12: 
Chris@12:         $groupRoot = '';
Chris@0:         $class = '';
Chris@0:         $alias = '';
Chris@0:         $statements = array();
Chris@0:         $explicitAlias = false;
Chris@0:         while (($token = $this->next())) {
Chris@0:             $isNameToken = $token[0] === T_STRING || $token[0] === T_NS_SEPARATOR;
Chris@0:             if (!$explicitAlias && $isNameToken) {
Chris@0:                 $class .= $token[1];
Chris@0:                 $alias = $token[1];
Chris@0:             } else if ($explicitAlias && $isNameToken) {
Chris@0:                 $alias .= $token[1];
Chris@0:             } else if ($token[0] === T_AS) {
Chris@0:                 $explicitAlias = true;
Chris@0:                 $alias = '';
Chris@0:             } else if ($token === ',') {
Chris@12:                 $statements[strtolower($alias)] = $groupRoot . $class;
Chris@0:                 $class = '';
Chris@0:                 $alias = '';
Chris@0:                 $explicitAlias = false;
Chris@0:             } else if ($token === ';') {
Chris@12:                 $statements[strtolower($alias)] = $groupRoot . $class;
Chris@0:                 break;
Chris@12:             } else if ($token === '{' ) {
Chris@12:                 $groupRoot = $class;
Chris@12:                 $class = '';
Chris@12:             } else if ($token === '}' ) {
Chris@12:                 continue;
Chris@0:             } else {
Chris@0:                 break;
Chris@0:             }
Chris@0:         }
Chris@0: 
Chris@0:         return $statements;
Chris@0:     }
Chris@0: 
Chris@0:     /**
Chris@0:      * Gets all use statements.
Chris@0:      *
Chris@0:      * @param string $namespaceName The namespace name of the reflected class.
Chris@0:      *
Chris@0:      * @return array A list with all found use statements.
Chris@0:      */
Chris@0:     public function parseUseStatements($namespaceName)
Chris@0:     {
Chris@0:         $statements = array();
Chris@0:         while (($token = $this->next())) {
Chris@0:             if ($token[0] === T_USE) {
Chris@0:                 $statements = array_merge($statements, $this->parseUseStatement());
Chris@0:                 continue;
Chris@0:             }
Chris@0:             if ($token[0] !== T_NAMESPACE || $this->parseNamespace() != $namespaceName) {
Chris@0:                 continue;
Chris@0:             }
Chris@0: 
Chris@0:             // Get fresh array for new namespace. This is to prevent the parser to collect the use statements
Chris@0:             // for a previous namespace with the same name. This is the case if a namespace is defined twice
Chris@0:             // or if a namespace with the same name is commented out.
Chris@0:             $statements = array();
Chris@0:         }
Chris@0: 
Chris@0:         return $statements;
Chris@0:     }
Chris@0: 
Chris@0:     /**
Chris@0:      * Gets the namespace.
Chris@0:      *
Chris@0:      * @return string The found namespace.
Chris@0:      */
Chris@0:     public function parseNamespace()
Chris@0:     {
Chris@0:         $name = '';
Chris@0:         while (($token = $this->next()) && ($token[0] === T_STRING || $token[0] === T_NS_SEPARATOR)) {
Chris@0:             $name .= $token[1];
Chris@0:         }
Chris@0: 
Chris@0:         return $name;
Chris@0:     }
Chris@0: 
Chris@0:     /**
Chris@0:      * Gets the class name.
Chris@0:      *
Chris@0:      * @return string The found class name.
Chris@0:      */
Chris@0:     public function parseClass()
Chris@0:     {
Chris@0:         // Namespaces and class names are tokenized the same: T_STRINGs
Chris@0:         // separated by T_NS_SEPARATOR so we can use one function to provide
Chris@0:         // both.
Chris@0:         return $this->parseNamespace();
Chris@0:     }
Chris@0: }