lochmueller / sourceopt

Classes/Service/CleanHtmlService.php

Indentation +370 added lines, -370 removed lines

@@ -17,374 +17,374 @@
  */
 class CleanHtmlService implements SingletonInterface
 {
-    /**
-     * Enable Debug comment in footer.
-     */
-    protected bool $debugComment = false;
-
-    /**
-     * Format Type.
-     */
-    protected int $formatType = 0;
-
-    /**
-     * Tab character.
-     */
-    protected string $tab = "\t";
-
-    /**
-     * Newline character.
-     */
-    protected string $newline = "\n";
-
-    /**
-     * Configured extra header comment.
-     */
-    protected string $headerComment = '';
-
-    /**
-     * Empty space char.
-     */
-    protected string $emptySpaceChar = ' ';
-
-    /**
-     * Set variables based on given config.
-     */
-    public function setVariables(array $config): void
-    {
-        if (isset($config['headerComment']) && !empty($config['headerComment'])) {
-            $this->headerComment = $config['headerComment'];
-        }
-
-        if (isset($config['formatHtml']) && is_numeric($config['formatHtml'])) {
-            $this->formatType = (int) $config['formatHtml'];
-        }
-
-        if (isset($config['formatHtml.']['tabSize']) && is_numeric($config['formatHtml.']['tabSize'])) {
-            $this->tab = str_pad('', (int) $config['formatHtml.']['tabSize'], ' ');
-        }
-
-        if (isset($config['formatHtml.']['debugComment'])) {
-            $this->debugComment = (bool) $config['formatHtml.']['debugComment'];
-        }
-
-        if (isset($config['dropEmptySpaceChar']) && (bool) $config['dropEmptySpaceChar']) {
-            $this->emptySpaceChar = '';
-        }
-    }
-
-    /**
-     * Clean given HTML with formatter.
-     */
-    public function clean(string $html, array $config = []): string
-    {
-        if (!empty($config)) {
-            $this->setVariables($config);
-        }
-
-        // convert line-breaks to UNIX
-        $this->convNlOs($html);
-
-        $manipulations = [];
-
-        if (isset($config['removeGenerator']) && (bool) $config['removeGenerator']) {
-            $manipulations['removeGenerator'] = GeneralUtility::makeInstance(RemoveGenerator::class);
-        }
-
-        if (isset($config['removeComments']) && (bool) $config['removeComments']) {
-            $manipulations['removeComments'] = GeneralUtility::makeInstance(RemoveComments::class);
-        }
-
-        if (!empty($this->headerComment)) {
-            $this->includeHeaderComment($html);
-        }
-
-        foreach ($manipulations as $key => $manipulation) {
-            /** @var ManipulationInterface $manipulation */
-            $configuration = isset($config[$key . '.']) && \is_array($config[$key . '.']) ? $config[$key . '.'] : [];
-            $html = $manipulation->manipulate($html, $configuration);
-        }
-
-        // cleanup HTML5 self-closing elements
-        if (!isset($GLOBALS['TSFE']->config['config']['doctype'])
-            || 'x' !== substr($GLOBALS['TSFE']->config['config']['doctype'], 0, 1)) {
-            $html = preg_replace(
-                '/<((?:area|base|br|col|embed|hr|img|input|link|meta|param|source|track|wbr)\s[^>]+?)\s*\\\?\/>/',
-                '<$1>',
-                $html
-            );
-        }
-
-        if ($this->formatType > 0) {
-            $html = $this->formatHtml($html);
-        }
-
-        // remove white space after line ending
-        $this->rTrimLines($html);
-
-        // recover line-breaks
-        if (Environment::isWindows()) {
-            $html = str_replace($this->newline, "\r\n", $html);
-        }
-
-        return (string) $html;
-    }
-
-    /**
-     * Formats the (X)HTML code:
-     *  - taps according to the hirarchy of the tags
-     *  - removes empty spaces between tags
-     *  - removes linebreaks within tags (spares where necessary: pre, textarea, comments, ..)
-     *  choose from five options:
-     *    0 => off
-     *    1 => no line break at all  (code in one line)
-     *    2 => minimalistic line breaks (structure defining box-elements)
-     *    3 => aesthetic line breaks (important box-elements)
-     *    4 => logic line breaks (all box-elements)
-     *    5 => max line breaks (all elements).
-     */
-    protected function formatHtml(string $html): string
-    {
-        // Save original formated pre, textarea, comments, styles and scripts & replace them with markers
-        preg_match_all(
-            '/(?s)((<!--.*?-->)|(<[ \n\r]*pre[^>]*>.*?<[ \n\r]*\/pre[^>]*>)|(<[ \n\r]*textarea[^>]*>.*?<[ \n\r]*\/textarea[^>]*>)|(<[ \n\r]*style[^>]*>.*?<[ \n\r]*\/style[^>]*>)|(<[ \n\r]*script[^>]*>.*?<[ \n\r]*\/script[^>]*>))/im',
-            $html,
-            $matches
-        );
-        $noFormat = $matches[0]; // do not format these block elements
-        for ($i = 0; $i < \count($noFormat); ++$i) {
-            $html = str_replace($noFormat[$i], "\n<!-- ELEMENT {$i} -->", $html);
-        }
-
-        // define box elements for formatting
-        $trueBoxElements = 'address|blockquote|center|dir|div|dl|fieldset|form|h1|h2|h3|h4|h5|h6|hr|isindex|menu|noframes|noscript|ol|p|pre|table|ul|article|aside|details|figcaption|figure|footer|header|hgroup|menu|nav|section';
-        $functionalBoxElements = 'dd|dt|frameset|li|tbody|td|tfoot|th|thead|tr|colgroup';
-        $usableBoxElements = 'applet|button|del|iframe|ins|map|object|script';
-        $imagineBoxElements = 'html|body|head|meta|title|link|script|base|!--';
-        $allBoxLikeElements = '(?>' . $trueBoxElements . '|' . $functionalBoxElements . '|' . $usableBoxElements . '|' . $imagineBoxElements . ')';
-        $esteticBoxLikeElements = '(?>html|head|body|meta name|title|div|table|h1|h2|h3|h4|h5|h6|p|form|pre|center|!--)';
-        $structureBoxLikeElements = '(?>html|head|body|div|!--)';
-
-        // split html into it's elements
-        $htmlArrayTemp = preg_split(
-            '/(<(?:[^<>]+(?:"[^"]*"|\'[^\']*\')?)+>)/',
-            $html,
-            -1,
-            \PREG_SPLIT_DELIM_CAPTURE | \PREG_SPLIT_NO_EMPTY
-        );
-
-        if (false === $htmlArrayTemp) {
-            // Restore saved comments, styles and scripts
-            for ($i = 0; $i < \count($noFormat); ++$i) {
-                $html = str_replace("<!-- ELEMENT {$i} -->", $noFormat[$i], $html);
-            }
-
-            return $html;
-        }
-        // remove empty lines
-        $htmlArray = [''];
-        $index = 1;
-        for ($x = 0; $x < \count($htmlArrayTemp); ++$x) {
-            $text = trim($htmlArrayTemp[$x]);
-            $htmlArray[$index] = '' !== $text ? $htmlArrayTemp[$x] : $this->emptySpaceChar;
-            ++$index;
-        }
-
-        // rebuild html
-        $html = '';
-        $tabs = 0;
-        for ($x = 0; $x < \count($htmlArray); ++$x) {
-            $htmlArrayBefore = $htmlArray[$x - 1] ?? '';
-            $htmlArrayCurrent = $htmlArray[$x] ?? '';
-
-            // check if the element should stand in a new line
-            $newline = false;
-            if ('<?xml' == substr($htmlArrayBefore, 0, 5)) {
-                $newline = true;
-            } elseif (2 == $this->formatType && ( // minimalistic line break
-                // this element has a line break before itself
-                preg_match(
-                    '/<' . $structureBoxLikeElements . '(.*)>/Usi',
-                    $htmlArrayCurrent
-                ) || preg_match(
-                    '/<' . $structureBoxLikeElements . '(.*) \/>/Usi',
-                    $htmlArrayCurrent
-                ) // one element before is a element that has a line break after
-                || preg_match(
-                    '/<\/' . $structureBoxLikeElements . '(.*)>/Usi',
-                    $htmlArrayBefore
-                ) || '<!--' == substr(
-                    $htmlArrayBefore,
-                    0,
-                    4
-                ) || preg_match('/<' . $structureBoxLikeElements . '(.*) \/>/Usi', $htmlArrayBefore))
-            ) {
-                $newline = true;
-            } elseif (3 == $this->formatType && ( // aestetic line break
-                // this element has a line break before itself
-                preg_match(
-                    '/<' . $esteticBoxLikeElements . '(.*)>/Usi',
-                    $htmlArrayCurrent
-                ) || preg_match(
-                    '/<' . $esteticBoxLikeElements . '(.*) \/>/Usi',
-                    $htmlArrayCurrent
-                ) // one element before is a element that has a line break after
-                || preg_match('/<\/' . $esteticBoxLikeElements . '(.*)>/Usi', $htmlArrayBefore) || '<!--' == substr(
-                    $htmlArrayBefore,
-                    0,
-                    4
-                ) || preg_match('/<' . $esteticBoxLikeElements . '(.*) \/>/Usi', $htmlArrayBefore))
-            ) {
-                $newline = true;
-            } elseif ($this->formatType >= 4 && ( // logical line break
-                // this element has a line break before itself
-                preg_match(
-                    '/<' . $allBoxLikeElements . '(.*)>/Usi',
-                    $htmlArrayCurrent
-                ) || preg_match(
-                    '/<' . $allBoxLikeElements . '(.*) \/>/Usi',
-                    $htmlArrayCurrent
-                ) // one element before is a element that has a line break after
-                || preg_match('/<\/' . $allBoxLikeElements . '(.*)>/Usi', $htmlArrayBefore) || '<!--' == substr(
-                    $htmlArrayBefore,
-                    0,
-                    4
-                ) || preg_match('/<' . $allBoxLikeElements . '(.*) \/>/Usi', $htmlArrayBefore))
-            ) {
-                $newline = true;
-            }
-
-            // count down a tab
-            if ('</' == substr($htmlArrayCurrent, 0, 2)) {
-                --$tabs;
-            }
-
-            // add tabs and line breaks in front of the current tag
-            if ($newline) {
-                $html .= $this->newline;
-                for ($y = 0; $y < $tabs; ++$y) {
-                    $html .= $this->tab;
-                }
-            }
-
-            // remove white spaces and line breaks and add current tag to the html-string
-            if ('<![CDATA[' == substr($htmlArrayCurrent, 0, 9) // remove multiple white space in CDATA / XML
-                || '<?xml' == substr($htmlArrayCurrent, 0, 5)
-            ) {
-                $html .= $this->killWhiteSpace($htmlArrayCurrent);
-            } else { // remove all line breaks
-                $html .= $this->killLineBreaks($htmlArrayCurrent);
-            }
-
-            // count up a tab
-            if ('<' == substr($htmlArrayCurrent, 0, 1) && '/' != substr($htmlArrayCurrent, 1, 1)) {
-                if (' ' !== substr($htmlArrayCurrent, 1, 1)
-                    && 'img' !== substr($htmlArrayCurrent, 1, 3)
-                    && 'source' !== substr($htmlArrayCurrent, 1, 6)
-                    && 'br' !== substr($htmlArrayCurrent, 1, 2)
-                    && 'hr' !== substr($htmlArrayCurrent, 1, 2)
-                    && 'input' !== substr($htmlArrayCurrent, 1, 5)
-                    && 'link' !== substr($htmlArrayCurrent, 1, 4)
-                    && 'meta' !== substr($htmlArrayCurrent, 1, 4)
-                    && 'col ' !== substr($htmlArrayCurrent, 1, 4)
-                    && 'frame' !== substr($htmlArrayCurrent, 1, 5)
-                    && 'isindex' !== substr($htmlArrayCurrent, 1, 7)
-                    && 'param' !== substr($htmlArrayCurrent, 1, 5)
-                    && 'area' !== substr($htmlArrayCurrent, 1, 4)
-                    && 'base' !== substr($htmlArrayCurrent, 1, 4)
-                    && '<!' !== substr($htmlArrayCurrent, 0, 2)
-                    && '<?xml' !== substr($htmlArrayCurrent, 0, 5)
-                ) {
-                    ++$tabs;
-                }
-            }
-        }
-
-        // Remove empty lines
-        if ($this->formatType > 1) {
-            $this->removeEmptyLines($html);
-        }
-
-        // Restore saved comments, styles and scripts
-        for ($i = 0; $i < \count($noFormat); ++$i) {
-            $html = str_replace("<!-- ELEMENT {$i} -->", $noFormat[$i], $html);
-        }
-
-        // include debug comment at the end
-        if (0 != $tabs && true === $this->debugComment) {
-            $html .= "<!-- {$tabs} open elements found -->";
-        }
-
-        return $html;
-    }
-
-    /**
-     * Remove ALL line breaks and multiple white space.
-     */
-    protected function killLineBreaks(string $html): string
-    {
-        $html = str_replace($this->newline, '', $html);
-
-        return preg_replace('/\s\s+/u', ' ', $html);
-        // ? return preg_replace('/\n|\s+(\s)/u', '$1', $html);
-    }
-
-    /**
-     * Remove multiple white space, keeps line breaks.
-     */
-    protected function killWhiteSpace(string $html): string
-    {
-        $temp = explode($this->newline, $html);
-        for ($i = 0; $i < \count($temp); ++$i) {
-            if (!trim($temp[$i])) {
-                unset($temp[$i]);
-                continue;
-            }
-
-            $temp[$i] = trim($temp[$i]);
-            $temp[$i] = preg_replace('/\s\s+/', ' ', $temp[$i]);
-        }
-
-        return implode($this->newline, $temp);
-    }
-
-    /**
-     * Remove white space at the end of lines, keeps other white space and line breaks.
-     */
-    protected function rTrimLines(string &$html): void
-    {
-        $html = preg_replace('/\s+$/m', '', $html);
-    }
-
-    /**
-     * Convert newlines according to the current OS.
-     */
-    protected function convNlOs(string &$html): void
-    {
-        $html = preg_replace("(\r\n|\r)", $this->newline, $html);
-    }
-
-    /**
-     * Remove empty lines.
-     */
-    protected function removeEmptyLines(string &$html): void
-    {
-        $temp = explode($this->newline, $html);
-        $result = [];
-        for ($i = 0; $i < \count($temp); ++$i) {
-            if ('' == trim($temp[$i])) {
-                continue;
-            }
-            $result[] = $temp[$i];
-        }
-        $html = implode($this->newline, $result);
-    }
-
-    /**
-     * Include configured header comment in HTML content block.
-     */
-    public function includeHeaderComment(string &$html): void
-    {
-        $html = preg_replace('/^(-->)$/m', "\n\t" . $this->headerComment . "\n$1", $html);
-    }
+	/**
+	 * Enable Debug comment in footer.
+	 */
+	protected bool $debugComment = false;
+
+	/**
+	 * Format Type.
+	 */
+	protected int $formatType = 0;
+
+	/**
+	 * Tab character.
+	 */
+	protected string $tab = "\t";
+
+	/**
+	 * Newline character.
+	 */
+	protected string $newline = "\n";
+
+	/**
+	 * Configured extra header comment.
+	 */
+	protected string $headerComment = '';
+
+	/**
+	 * Empty space char.
+	 */
+	protected string $emptySpaceChar = ' ';
+
+	/**
+	 * Set variables based on given config.
+	 */
+	public function setVariables(array $config): void
+	{
+		if (isset($config['headerComment']) && !empty($config['headerComment'])) {
+			$this->headerComment = $config['headerComment'];
+		}
+
+		if (isset($config['formatHtml']) && is_numeric($config['formatHtml'])) {
+			$this->formatType = (int) $config['formatHtml'];
+		}
+
+		if (isset($config['formatHtml.']['tabSize']) && is_numeric($config['formatHtml.']['tabSize'])) {
+			$this->tab = str_pad('', (int) $config['formatHtml.']['tabSize'], ' ');
+		}
+
+		if (isset($config['formatHtml.']['debugComment'])) {
+			$this->debugComment = (bool) $config['formatHtml.']['debugComment'];
+		}
+
+		if (isset($config['dropEmptySpaceChar']) && (bool) $config['dropEmptySpaceChar']) {
+			$this->emptySpaceChar = '';
+		}
+	}
+
+	/**
+	 * Clean given HTML with formatter.
+	 */
+	public function clean(string $html, array $config = []): string
+	{
+		if (!empty($config)) {
+			$this->setVariables($config);
+		}
+
+		// convert line-breaks to UNIX
+		$this->convNlOs($html);
+
+		$manipulations = [];
+
+		if (isset($config['removeGenerator']) && (bool) $config['removeGenerator']) {
+			$manipulations['removeGenerator'] = GeneralUtility::makeInstance(RemoveGenerator::class);
+		}
+
+		if (isset($config['removeComments']) && (bool) $config['removeComments']) {
+			$manipulations['removeComments'] = GeneralUtility::makeInstance(RemoveComments::class);
+		}
+
+		if (!empty($this->headerComment)) {
+			$this->includeHeaderComment($html);
+		}
+
+		foreach ($manipulations as $key => $manipulation) {
+			/** @var ManipulationInterface $manipulation */
+			$configuration = isset($config[$key . '.']) && \is_array($config[$key . '.']) ? $config[$key . '.'] : [];
+			$html = $manipulation->manipulate($html, $configuration);
+		}
+
+		// cleanup HTML5 self-closing elements
+		if (!isset($GLOBALS['TSFE']->config['config']['doctype'])
+			|| 'x' !== substr($GLOBALS['TSFE']->config['config']['doctype'], 0, 1)) {
+			$html = preg_replace(
+				'/<((?:area|base|br|col|embed|hr|img|input|link|meta|param|source|track|wbr)\s[^>]+?)\s*\\\?\/>/',
+				'<$1>',
+				$html
+			);
+		}
+
+		if ($this->formatType > 0) {
+			$html = $this->formatHtml($html);
+		}
+
+		// remove white space after line ending
+		$this->rTrimLines($html);
+
+		// recover line-breaks
+		if (Environment::isWindows()) {
+			$html = str_replace($this->newline, "\r\n", $html);
+		}
+
+		return (string) $html;
+	}
+
+	/**
+	 * Formats the (X)HTML code:
+	 *  - taps according to the hirarchy of the tags
+	 *  - removes empty spaces between tags
+	 *  - removes linebreaks within tags (spares where necessary: pre, textarea, comments, ..)
+	 *  choose from five options:
+	 *    0 => off
+	 *    1 => no line break at all  (code in one line)
+	 *    2 => minimalistic line breaks (structure defining box-elements)
+	 *    3 => aesthetic line breaks (important box-elements)
+	 *    4 => logic line breaks (all box-elements)
+	 *    5 => max line breaks (all elements).
+	 */
+	protected function formatHtml(string $html): string
+	{
+		// Save original formated pre, textarea, comments, styles and scripts & replace them with markers
+		preg_match_all(
+			'/(?s)((<!--.*?-->)|(<[ \n\r]*pre[^>]*>.*?<[ \n\r]*\/pre[^>]*>)|(<[ \n\r]*textarea[^>]*>.*?<[ \n\r]*\/textarea[^>]*>)|(<[ \n\r]*style[^>]*>.*?<[ \n\r]*\/style[^>]*>)|(<[ \n\r]*script[^>]*>.*?<[ \n\r]*\/script[^>]*>))/im',
+			$html,
+			$matches
+		);
+		$noFormat = $matches[0]; // do not format these block elements
+		for ($i = 0; $i < \count($noFormat); ++$i) {
+			$html = str_replace($noFormat[$i], "\n<!-- ELEMENT {$i} -->", $html);
+		}
+
+		// define box elements for formatting
+		$trueBoxElements = 'address|blockquote|center|dir|div|dl|fieldset|form|h1|h2|h3|h4|h5|h6|hr|isindex|menu|noframes|noscript|ol|p|pre|table|ul|article|aside|details|figcaption|figure|footer|header|hgroup|menu|nav|section';
+		$functionalBoxElements = 'dd|dt|frameset|li|tbody|td|tfoot|th|thead|tr|colgroup';
+		$usableBoxElements = 'applet|button|del|iframe|ins|map|object|script';
+		$imagineBoxElements = 'html|body|head|meta|title|link|script|base|!--';
+		$allBoxLikeElements = '(?>' . $trueBoxElements . '|' . $functionalBoxElements . '|' . $usableBoxElements . '|' . $imagineBoxElements . ')';
+		$esteticBoxLikeElements = '(?>html|head|body|meta name|title|div|table|h1|h2|h3|h4|h5|h6|p|form|pre|center|!--)';
+		$structureBoxLikeElements = '(?>html|head|body|div|!--)';
+
+		// split html into it's elements
+		$htmlArrayTemp = preg_split(
+			'/(<(?:[^<>]+(?:"[^"]*"|\'[^\']*\')?)+>)/',
+			$html,
+			-1,
+			\PREG_SPLIT_DELIM_CAPTURE | \PREG_SPLIT_NO_EMPTY
+		);
+
+		if (false === $htmlArrayTemp) {
+			// Restore saved comments, styles and scripts
+			for ($i = 0; $i < \count($noFormat); ++$i) {
+				$html = str_replace("<!-- ELEMENT {$i} -->", $noFormat[$i], $html);
+			}
+
+			return $html;
+		}
+		// remove empty lines
+		$htmlArray = [''];
+		$index = 1;
+		for ($x = 0; $x < \count($htmlArrayTemp); ++$x) {
+			$text = trim($htmlArrayTemp[$x]);
+			$htmlArray[$index] = '' !== $text ? $htmlArrayTemp[$x] : $this->emptySpaceChar;
+			++$index;
+		}
+
+		// rebuild html
+		$html = '';
+		$tabs = 0;
+		for ($x = 0; $x < \count($htmlArray); ++$x) {
+			$htmlArrayBefore = $htmlArray[$x - 1] ?? '';
+			$htmlArrayCurrent = $htmlArray[$x] ?? '';
+
+			// check if the element should stand in a new line
+			$newline = false;
+			if ('<?xml' == substr($htmlArrayBefore, 0, 5)) {
+				$newline = true;
+			} elseif (2 == $this->formatType && ( // minimalistic line break
+				// this element has a line break before itself
+				preg_match(
+					'/<' . $structureBoxLikeElements . '(.*)>/Usi',
+					$htmlArrayCurrent
+				) || preg_match(
+					'/<' . $structureBoxLikeElements . '(.*) \/>/Usi',
+					$htmlArrayCurrent
+				) // one element before is a element that has a line break after
+				|| preg_match(
+					'/<\/' . $structureBoxLikeElements . '(.*)>/Usi',
+					$htmlArrayBefore
+				) || '<!--' == substr(
+					$htmlArrayBefore,
+					0,
+					4
+				) || preg_match('/<' . $structureBoxLikeElements . '(.*) \/>/Usi', $htmlArrayBefore))
+			) {
+				$newline = true;
+			} elseif (3 == $this->formatType && ( // aestetic line break
+				// this element has a line break before itself
+				preg_match(
+					'/<' . $esteticBoxLikeElements . '(.*)>/Usi',
+					$htmlArrayCurrent
+				) || preg_match(
+					'/<' . $esteticBoxLikeElements . '(.*) \/>/Usi',
+					$htmlArrayCurrent
+				) // one element before is a element that has a line break after
+				|| preg_match('/<\/' . $esteticBoxLikeElements . '(.*)>/Usi', $htmlArrayBefore) || '<!--' == substr(
+					$htmlArrayBefore,
+					0,
+					4
+				) || preg_match('/<' . $esteticBoxLikeElements . '(.*) \/>/Usi', $htmlArrayBefore))
+			) {
+				$newline = true;
+			} elseif ($this->formatType >= 4 && ( // logical line break
+				// this element has a line break before itself
+				preg_match(
+					'/<' . $allBoxLikeElements . '(.*)>/Usi',
+					$htmlArrayCurrent
+				) || preg_match(
+					'/<' . $allBoxLikeElements . '(.*) \/>/Usi',
+					$htmlArrayCurrent
+				) // one element before is a element that has a line break after
+				|| preg_match('/<\/' . $allBoxLikeElements . '(.*)>/Usi', $htmlArrayBefore) || '<!--' == substr(
+					$htmlArrayBefore,
+					0,
+					4
+				) || preg_match('/<' . $allBoxLikeElements . '(.*) \/>/Usi', $htmlArrayBefore))
+			) {
+				$newline = true;
+			}
+
+			// count down a tab
+			if ('</' == substr($htmlArrayCurrent, 0, 2)) {
+				--$tabs;
+			}
+
+			// add tabs and line breaks in front of the current tag
+			if ($newline) {
+				$html .= $this->newline;
+				for ($y = 0; $y < $tabs; ++$y) {
+					$html .= $this->tab;
+				}
+			}
+
+			// remove white spaces and line breaks and add current tag to the html-string
+			if ('<![CDATA[' == substr($htmlArrayCurrent, 0, 9) // remove multiple white space in CDATA / XML
+				|| '<?xml' == substr($htmlArrayCurrent, 0, 5)
+			) {
+				$html .= $this->killWhiteSpace($htmlArrayCurrent);
+			} else { // remove all line breaks
+				$html .= $this->killLineBreaks($htmlArrayCurrent);
+			}
+
+			// count up a tab
+			if ('<' == substr($htmlArrayCurrent, 0, 1) && '/' != substr($htmlArrayCurrent, 1, 1)) {
+				if (' ' !== substr($htmlArrayCurrent, 1, 1)
+					&& 'img' !== substr($htmlArrayCurrent, 1, 3)
+					&& 'source' !== substr($htmlArrayCurrent, 1, 6)
+					&& 'br' !== substr($htmlArrayCurrent, 1, 2)
+					&& 'hr' !== substr($htmlArrayCurrent, 1, 2)
+					&& 'input' !== substr($htmlArrayCurrent, 1, 5)
+					&& 'link' !== substr($htmlArrayCurrent, 1, 4)
+					&& 'meta' !== substr($htmlArrayCurrent, 1, 4)
+					&& 'col ' !== substr($htmlArrayCurrent, 1, 4)
+					&& 'frame' !== substr($htmlArrayCurrent, 1, 5)
+					&& 'isindex' !== substr($htmlArrayCurrent, 1, 7)
+					&& 'param' !== substr($htmlArrayCurrent, 1, 5)
+					&& 'area' !== substr($htmlArrayCurrent, 1, 4)
+					&& 'base' !== substr($htmlArrayCurrent, 1, 4)
+					&& '<!' !== substr($htmlArrayCurrent, 0, 2)
+					&& '<?xml' !== substr($htmlArrayCurrent, 0, 5)
+				) {
+					++$tabs;
+				}
+			}
+		}
+
+		// Remove empty lines
+		if ($this->formatType > 1) {
+			$this->removeEmptyLines($html);
+		}
+
+		// Restore saved comments, styles and scripts
+		for ($i = 0; $i < \count($noFormat); ++$i) {
+			$html = str_replace("<!-- ELEMENT {$i} -->", $noFormat[$i], $html);
+		}
+
+		// include debug comment at the end
+		if (0 != $tabs && true === $this->debugComment) {
+			$html .= "<!-- {$tabs} open elements found -->";
+		}
+
+		return $html;
+	}
+
+	/**
+	 * Remove ALL line breaks and multiple white space.
+	 */
+	protected function killLineBreaks(string $html): string
+	{
+		$html = str_replace($this->newline, '', $html);
+
+		return preg_replace('/\s\s+/u', ' ', $html);
+		// ? return preg_replace('/\n|\s+(\s)/u', '$1', $html);
+	}
+
+	/**
+	 * Remove multiple white space, keeps line breaks.
+	 */
+	protected function killWhiteSpace(string $html): string
+	{
+		$temp = explode($this->newline, $html);
+		for ($i = 0; $i < \count($temp); ++$i) {
+			if (!trim($temp[$i])) {
+				unset($temp[$i]);
+				continue;
+			}
+
+			$temp[$i] = trim($temp[$i]);
+			$temp[$i] = preg_replace('/\s\s+/', ' ', $temp[$i]);
+		}
+
+		return implode($this->newline, $temp);
+	}
+
+	/**
+	 * Remove white space at the end of lines, keeps other white space and line breaks.
+	 */
+	protected function rTrimLines(string &$html): void
+	{
+		$html = preg_replace('/\s+$/m', '', $html);
+	}
+
+	/**
+	 * Convert newlines according to the current OS.
+	 */
+	protected function convNlOs(string &$html): void
+	{
+		$html = preg_replace("(\r\n|\r)", $this->newline, $html);
+	}
+
+	/**
+	 * Remove empty lines.
+	 */
+	protected function removeEmptyLines(string &$html): void
+	{
+		$temp = explode($this->newline, $html);
+		$result = [];
+		for ($i = 0; $i < \count($temp); ++$i) {
+			if ('' == trim($temp[$i])) {
+				continue;
+			}
+			$result[] = $temp[$i];
+		}
+		$html = implode($this->newline, $result);
+	}
+
+	/**
+	 * Include configured header comment in HTML content block.
+	 */
+	public function includeHeaderComment(string &$html): void
+	{
+		$html = preg_replace('/^(-->)$/m', "\n\t" . $this->headerComment . "\n$1", $html);
+	}
 }

Spacing +16 added lines, -16 removed lines

@@ -101,7 +101,7 @@ 
 
         foreach ($manipulations as $key => $manipulation) {
             /** @var ManipulationInterface $manipulation */
-            $configuration = isset($config[$key . '.']) && \is_array($config[$key . '.']) ? $config[$key . '.'] : [];
+            $configuration = isset($config[$key.'.']) && \is_array($config[$key.'.']) ? $config[$key.'.'] : [];
             $html = $manipulation->manipulate($html, $configuration);
         }
 
@@ -161,7 +161,7 @@ 
         $functionalBoxElements = 'dd|dt|frameset|li|tbody|td|tfoot|th|thead|tr|colgroup';
         $usableBoxElements = 'applet|button|del|iframe|ins|map|object|script';
         $imagineBoxElements = 'html|body|head|meta|title|link|script|base|!--';
-        $allBoxLikeElements = '(?>' . $trueBoxElements . '|' . $functionalBoxElements . '|' . $usableBoxElements . '|' . $imagineBoxElements . ')';
+        $allBoxLikeElements = '(?>'.$trueBoxElements.'|'.$functionalBoxElements.'|'.$usableBoxElements.'|'.$imagineBoxElements.')';
         $esteticBoxLikeElements = '(?>html|head|body|meta name|title|div|table|h1|h2|h3|h4|h5|h6|p|form|pre|center|!--)';
         $structureBoxLikeElements = '(?>html|head|body|div|!--)';
 
@@ -170,7 +170,7 @@ 
             '/(<(?:[^<>]+(?:"[^"]*"|\'[^\']*\')?)+>)/',
             $html,
             -1,
-            \PREG_SPLIT_DELIM_CAPTURE | \PREG_SPLIT_NO_EMPTY
+            \PREG_SPLIT_DELIM_CAPTURE|\PREG_SPLIT_NO_EMPTY
         );
 
         if (false === $htmlArrayTemp) {
@@ -204,52 +204,52 @@ 
             } elseif (2 == $this->formatType && ( // minimalistic line break
                 // this element has a line break before itself
                 preg_match(
-                    '/<' . $structureBoxLikeElements . '(.*)>/Usi',
+                    '/<'.$structureBoxLikeElements.'(.*)>/Usi',
                     $htmlArrayCurrent
                 ) || preg_match(
-                    '/<' . $structureBoxLikeElements . '(.*) \/>/Usi',
+                    '/<'.$structureBoxLikeElements.'(.*) \/>/Usi',
                     $htmlArrayCurrent
                 ) // one element before is a element that has a line break after
                 || preg_match(
-                    '/<\/' . $structureBoxLikeElements . '(.*)>/Usi',
+                    '/<\/'.$structureBoxLikeElements.'(.*)>/Usi',
                     $htmlArrayBefore
                 ) || '<!--' == substr(
                     $htmlArrayBefore,
                     0,
                     4
-                ) || preg_match('/<' . $structureBoxLikeElements . '(.*) \/>/Usi', $htmlArrayBefore))
+                ) || preg_match('/<'.$structureBoxLikeElements.'(.*) \/>/Usi', $htmlArrayBefore))
             ) {
                 $newline = true;
             } elseif (3 == $this->formatType && ( // aestetic line break
                 // this element has a line break before itself
                 preg_match(
-                    '/<' . $esteticBoxLikeElements . '(.*)>/Usi',
+                    '/<'.$esteticBoxLikeElements.'(.*)>/Usi',
                     $htmlArrayCurrent
                 ) || preg_match(
-                    '/<' . $esteticBoxLikeElements . '(.*) \/>/Usi',
+                    '/<'.$esteticBoxLikeElements.'(.*) \/>/Usi',
                     $htmlArrayCurrent
                 ) // one element before is a element that has a line break after
-                || preg_match('/<\/' . $esteticBoxLikeElements . '(.*)>/Usi', $htmlArrayBefore) || '<!--' == substr(
+                || preg_match('/<\/'.$esteticBoxLikeElements.'(.*)>/Usi', $htmlArrayBefore) || '<!--' == substr(
                     $htmlArrayBefore,
                     0,
                     4
-                ) || preg_match('/<' . $esteticBoxLikeElements . '(.*) \/>/Usi', $htmlArrayBefore))
+                ) || preg_match('/<'.$esteticBoxLikeElements.'(.*) \/>/Usi', $htmlArrayBefore))
             ) {
                 $newline = true;
             } elseif ($this->formatType >= 4 && ( // logical line break
                 // this element has a line break before itself
                 preg_match(
-                    '/<' . $allBoxLikeElements . '(.*)>/Usi',
+                    '/<'.$allBoxLikeElements.'(.*)>/Usi',
                     $htmlArrayCurrent
                 ) || preg_match(
-                    '/<' . $allBoxLikeElements . '(.*) \/>/Usi',
+                    '/<'.$allBoxLikeElements.'(.*) \/>/Usi',
                     $htmlArrayCurrent
                 ) // one element before is a element that has a line break after
-                || preg_match('/<\/' . $allBoxLikeElements . '(.*)>/Usi', $htmlArrayBefore) || '<!--' == substr(
+                || preg_match('/<\/'.$allBoxLikeElements.'(.*)>/Usi', $htmlArrayBefore) || '<!--' == substr(
                     $htmlArrayBefore,
                     0,
                     4
-                ) || preg_match('/<' . $allBoxLikeElements . '(.*) \/>/Usi', $htmlArrayBefore))
+                ) || preg_match('/<'.$allBoxLikeElements.'(.*) \/>/Usi', $htmlArrayBefore))
             ) {
                 $newline = true;
             }
@@ -385,6 +385,6 @@ 
      */
     public function includeHeaderComment(string &$html): void
     {
-        $html = preg_replace('/^(-->)$/m', "\n\t" . $this->headerComment . "\n$1", $html);
+        $html = preg_replace('/^(-->)$/m', "\n\t".$this->headerComment."\n$1", $html);
     }
 }