<?php
/**
 * Zend Framework (http://framework.zend.com/)
 *
 * @link      http://github.com/zendframework/zf2 for the canonical source repository
 * @copyright Copyright (c) 2005-2015 Zend Technologies USA Inc. (http://www.zend.com)
 * @license   http://framework.zend.com/license/new-bsd New BSD License
 */

namespace Zend\Code\Scanner;

use Zend\Code\Annotation\AnnotationCollection;
use Zend\Code\Annotation\AnnotationManager;
use Zend\Code\NameInformation;

class AnnotationScanner extends AnnotationCollection implements ScannerInterface
{
    /**
     * @var bool
     */
    protected $isScanned = false;

    /**
     * @var string
     */
    protected $docComment = null;

    /**
     * @var NameInformation
     */
    protected $nameInformation = null;

    /**
     * @var AnnotationManager
     */
    protected $annotationManager = null;

    /**
     * @var array
     */
    protected $annotations = array();

    /**
     * @param  AnnotationManager $annotationManager
     * @param  string $docComment
     * @param  NameInformation $nameInformation
     * @return AnnotationScanner
     */
    public function __construct(
        AnnotationManager $annotationManager,
        $docComment,
        NameInformation $nameInformation = null
    ) {
        $this->annotationManager = $annotationManager;
        $this->docComment        = $docComment;
        $this->nameInformation   = $nameInformation;
        $this->scan($this->tokenize());
    }

    /**
     * @param NameInformation $nameInformation
     */
    public function setNameInformation(NameInformation $nameInformation)
    {
        $this->nameInformation = $nameInformation;
    }

    /**
     * @param  array $tokens
     */
    protected function scan(array $tokens)
    {
        $annotations     = array();
        $annotationIndex = -1;
        $contentEnd      = false;

        reset($tokens);

        SCANNER_TOP:
        $token = current($tokens);

        switch ($token[0]) {

            case 'ANNOTATION_CLASS':

                $contentEnd = false;
                $annotationIndex++;
                $class                         = substr($token[1], 1);
                $class                         = $this->nameInformation->resolveName($class);
                $annotations[$annotationIndex] = array($class, null);
                goto SCANNER_CONTINUE;
                // goto no break needed

            case 'ANNOTATION_CONTENT_START':

                $annotations[$annotationIndex][1] = '';
                //fall-through

            case 'ANNOTATION_CONTENT_END':
            case 'ANNOTATION_CONTENT':
            case 'ANNOTATION_WHITESPACE':
            case 'ANNOTATION_NEWLINE':

                if (!$contentEnd && isset($annotations[$annotationIndex]) && is_string($annotations[$annotationIndex][1])) {
                    $annotations[$annotationIndex][1] .= $token[1];
                }

                if ($token[0] === 'ANNOTATION_CONTENT_END') {
                    $contentEnd = true;
                }

                goto SCANNER_CONTINUE;
        }

        SCANNER_CONTINUE:
        if (next($tokens) === false) {
            goto SCANNER_END;
        }
        goto SCANNER_TOP;

        SCANNER_END:

        foreach ($annotations as $annotation) {
            $annotation[]     = '@' . $annotation[0] . $annotation[1];
            $annotationObject = $this->annotationManager->createAnnotation($annotation);
            if ($annotationObject) {
                $this->append($annotationObject);
            }
        }
    }

    /**
     * @return array
     */
    protected function tokenize()
    {
        static $CONTEXT_DOCBLOCK = 0x01;
        static $CONTEXT_ASTERISK = 0x02;
        static $CONTEXT_CLASS = 0x04;
        static $CONTEXT_CONTENT = 0x08;

        $context     = 0x00;
        $stream      = $this->docComment;
        $streamIndex = null;
        $tokens      = array();
        $tokenIndex  = null;
        $currentChar = null;
        $currentWord = null;
        $currentLine = null;

        $annotationParentCount = 0;

        $MACRO_STREAM_ADVANCE_CHAR = function ($positionsForward = 1) use (&$stream, &$streamIndex, &$currentChar, &$currentWord, &$currentLine) {
            $positionsForward = ($positionsForward > 0) ? $positionsForward : 1;
            $streamIndex      = ($streamIndex === null) ? 0 : $streamIndex + $positionsForward;
            if (!isset($stream[$streamIndex])) {
                $currentChar = false;

                return false;
            }
            $currentChar = $stream[$streamIndex];
            $matches     = array();
            $currentLine = (preg_match('#(.*?)(?:\n|\r\n?)#', $stream, $matches, null, $streamIndex) === 1) ? $matches[1] : substr($stream, $streamIndex);
            if ($currentChar === ' ') {
                $currentWord = (preg_match('#( +)#', $currentLine, $matches) === 1) ? $matches[1] : $currentLine;
            } else {
                $currentWord = (($matches = strpos($currentLine, ' ')) !== false) ? substr($currentLine, 0, $matches) : $currentLine;
            }

            return $currentChar;
        };
        $MACRO_STREAM_ADVANCE_WORD = function () use (&$currentWord, &$MACRO_STREAM_ADVANCE_CHAR) {
            return $MACRO_STREAM_ADVANCE_CHAR(strlen($currentWord));
        };
        $MACRO_STREAM_ADVANCE_LINE = function () use (&$currentLine, &$MACRO_STREAM_ADVANCE_CHAR) {
            return $MACRO_STREAM_ADVANCE_CHAR(strlen($currentLine));
        };
        $MACRO_TOKEN_ADVANCE       = function () use (&$tokenIndex, &$tokens) {
            $tokenIndex          = ($tokenIndex === null) ? 0 : $tokenIndex + 1;
            $tokens[$tokenIndex] = array('ANNOTATION_UNKNOWN', '');
        };
        $MACRO_TOKEN_SET_TYPE      = function ($type) use (&$tokenIndex, &$tokens) {
            $tokens[$tokenIndex][0] = $type;
        };
        $MACRO_TOKEN_APPEND_CHAR   = function () use (&$currentChar, &$tokens, &$tokenIndex) {
            $tokens[$tokenIndex][1] .= $currentChar;
        };
        $MACRO_TOKEN_APPEND_WORD   = function () use (&$currentWord, &$tokens, &$tokenIndex) {
            $tokens[$tokenIndex][1] .= $currentWord;
        };
        $MACRO_TOKEN_APPEND_LINE   = function () use (&$currentLine, &$tokens, &$tokenIndex) {
            $tokens[$tokenIndex][1] .= $currentLine;
        };
        $MACRO_HAS_CONTEXT         = function ($which) use (&$context) {
            return (($context & $which) === $which);
        };

        $MACRO_STREAM_ADVANCE_CHAR();
        $MACRO_TOKEN_ADVANCE();

        TOKENIZER_TOP:

        if ($context === 0x00 && $currentChar === '/' && $currentWord === '/**') {
            $MACRO_TOKEN_SET_TYPE('ANNOTATION_COMMENTSTART');
            $MACRO_TOKEN_APPEND_WORD();
            $MACRO_TOKEN_ADVANCE();
            $context |= $CONTEXT_DOCBLOCK;
            $context |= $CONTEXT_ASTERISK;
            if ($MACRO_STREAM_ADVANCE_WORD() === false) {
                goto TOKENIZER_END;
            }
            goto TOKENIZER_TOP;
        }

        if ($MACRO_HAS_CONTEXT($CONTEXT_CLASS)) {
            if (in_array($currentChar, array(' ', '(', "\n", "\r"))) {
                $context &= ~$CONTEXT_CLASS;
                $MACRO_TOKEN_ADVANCE();
            } else {
                $MACRO_TOKEN_APPEND_CHAR();
                if ($MACRO_STREAM_ADVANCE_CHAR() === false) {
                    goto TOKENIZER_END;
                }
                goto TOKENIZER_TOP;
            }
        }

        // Since we don't know what line endings are used in the file, we check for all scenarios. If we find a
        // cariage return (\r), we check the next character for a line feed (\n). If so we consume it and act as
        // if the cariage return was a line feed.
        $lineEnded = $currentChar === "\n";
        if ($currentChar === "\r") {
            $lineEnded = true;

            $nextChar = $MACRO_STREAM_ADVANCE_CHAR();
            if ($nextChar !== "\n") {
                $streamIndex--;
            }

            $currentChar = "\n";
        }

        if ($lineEnded) {
            $MACRO_TOKEN_SET_TYPE('ANNOTATION_NEWLINE');
            $MACRO_TOKEN_APPEND_CHAR();
            $MACRO_TOKEN_ADVANCE();
            $context &= ~$CONTEXT_ASTERISK;
            $context &= ~$CONTEXT_CLASS;
            if ($MACRO_STREAM_ADVANCE_CHAR() === false) {
                goto TOKENIZER_END;
            }
            goto TOKENIZER_TOP;
        }

        if ($currentChar === ' ') {
            $MACRO_TOKEN_SET_TYPE(($MACRO_HAS_CONTEXT($CONTEXT_ASTERISK)) ? 'ANNOTATION_WHITESPACE' : 'ANNOTATION_WHITESPACE_INDENT');
            $MACRO_TOKEN_APPEND_WORD();
            $MACRO_TOKEN_ADVANCE();
            if ($MACRO_STREAM_ADVANCE_WORD() === false) {
                goto TOKENIZER_END;
            }
            goto TOKENIZER_TOP;
        }

        if ($MACRO_HAS_CONTEXT($CONTEXT_CONTENT) && $MACRO_HAS_CONTEXT($CONTEXT_ASTERISK)) {
            $MACRO_TOKEN_SET_TYPE('ANNOTATION_CONTENT');
            $annotationParentCount += substr_count($currentWord, '(');
            $annotationParentCount -= substr_count($currentWord, ')');

            if ($annotationParentCount === 0) {
                $context &= ~$CONTEXT_CONTENT;
                $MACRO_TOKEN_SET_TYPE('ANNOTATION_CONTENT_END');
            }
            $MACRO_TOKEN_APPEND_WORD();
            $MACRO_TOKEN_ADVANCE();
            if ($MACRO_STREAM_ADVANCE_WORD() === false) {
                goto TOKENIZER_END;
            }
            goto TOKENIZER_TOP;
        }

        if ($currentChar === '(' && $tokens[$tokenIndex - 1][0] === 'ANNOTATION_CLASS') {
            $context |= $CONTEXT_CONTENT;
            $annotationParentCount = 1;
            $MACRO_TOKEN_SET_TYPE('ANNOTATION_CONTENT_START');
            $MACRO_TOKEN_APPEND_CHAR();
            $MACRO_TOKEN_ADVANCE();
            if ($MACRO_STREAM_ADVANCE_CHAR() === false) {
                goto TOKENIZER_END;
            }
            goto TOKENIZER_TOP;
        }

        if ($MACRO_HAS_CONTEXT($CONTEXT_DOCBLOCK) && $currentWord === '*/') {
            $MACRO_TOKEN_SET_TYPE('ANNOTATION_COMMENTEND');
            $MACRO_TOKEN_APPEND_WORD();
            $MACRO_TOKEN_ADVANCE();
            $context &= ~$CONTEXT_DOCBLOCK;
            if ($MACRO_STREAM_ADVANCE_WORD() === false) {
                goto TOKENIZER_END;
            }
            goto TOKENIZER_TOP;
        }

        if ($currentChar === '*') {
            if ($MACRO_HAS_CONTEXT($CONTEXT_DOCBLOCK) && ($MACRO_HAS_CONTEXT($CONTEXT_ASTERISK))) {
                $MACRO_TOKEN_SET_TYPE('ANNOTATION_IGNORE');
            } else {
                $MACRO_TOKEN_SET_TYPE('ANNOTATION_ASTERISK');
                $context |= $CONTEXT_ASTERISK;
            }
            $MACRO_TOKEN_APPEND_CHAR();
            $MACRO_TOKEN_ADVANCE();
            if ($MACRO_STREAM_ADVANCE_CHAR() === false) {
                goto TOKENIZER_END;
            }
            goto TOKENIZER_TOP;
        }

        if ($currentChar === '@') {
            $MACRO_TOKEN_SET_TYPE('ANNOTATION_CLASS');
            $context |= $CONTEXT_CLASS;
            $MACRO_TOKEN_APPEND_CHAR();
            if ($MACRO_STREAM_ADVANCE_CHAR() === false) {
                goto TOKENIZER_END;
            }
            goto TOKENIZER_TOP;
        }

        TOKENIZER_CONTINUE:

        if ($context && $CONTEXT_CONTENT) {
            $MACRO_TOKEN_APPEND_CHAR();
            if ($MACRO_STREAM_ADVANCE_CHAR() === false) {
                goto TOKENIZER_END;
            }
        } else {
            $MACRO_TOKEN_SET_TYPE('ANNOTATION_IGNORE');
            $MACRO_TOKEN_APPEND_LINE();
            $MACRO_TOKEN_ADVANCE();
            if ($MACRO_STREAM_ADVANCE_LINE() === false) {
                goto TOKENIZER_END;
            }
        }
        goto TOKENIZER_TOP;

        TOKENIZER_END:

        array_pop($tokens);

        return $tokens;
    }
}