HtmlConverter::convertChildren() - Code Metrics - Inspection of "Add support for tables (#203)" - thephpleague/html-to-markdown - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( 0868ae...a5d47a )

by Colin

created 2021-03-28 21:51 UTC

HtmlConverter::convertChildren() B

↳ Parent: HtmlConverter

Complexity

Conditions	6
Paths	5

Size

Total Lines

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	12
CRAP Score	6

Importance

Changes

Metric	Value
dl	0
loc	29
c	0
b	0
f	0
ccs	12
cts	12
cp	1
rs	8.8337
cc	6
nc	5
nop	1
crap	6

<?php

namespace League\HTMLToMarkdown;

use League\HTMLToMarkdown\Converter\PreConverterInterface;

/**
 * Class HtmlConverter
 *
 * A helper class to convert HTML to Markdown.
 *
 * @author Colin O'Dell <[email protected]>
 * @author Nick Cernis <[email protected]>
 *
 * @link https://github.com/thephpleague/html-to-markdown/ Latest version on GitHub.
 *
 * @license http://www.opensource.org/licenses/mit-license.php MIT
 */
class HtmlConverter implements HtmlConverterInterface
{
    /**
     * @var Environment
     */
    protected $environment;

    /**
     * Constructor
     *
     * @param Environment|array $options Environment object or configuration options
     */
    public function __construct($options = array())
    {
        if ($options instanceof Environment) {
            $this->environment = $options;
        } elseif (is_array($options)) {
            $defaults = array(
                'header_style' => 'setext', // Set to 'atx' to output H1 and H2 headers as # Header1 and ## Header2
                'suppress_errors' => true, // Set to false to show warnings when loading malformed HTML
                'strip_tags' => false, // Set to true to strip tags that don't have markdown equivalents. N.B. Strips tags, not their content. Useful to clean MS Word HTML output.
                'bold_style' => '**', // DEPRECATED: Set to '__' if you prefer the underlined style
                'italic_style' => '*', // DEPRECATED: Set to '_' if you prefer the underlined style
                'remove_nodes' => '', // space-separated list of dom nodes that should be removed. example: 'meta style script'
                'hard_break' => false, // Set to true to turn <br> into `\n` instead of `  \n`
                'list_item_style' => '-', // Set the default character for each <li> in a <ul>. Can be '-', '*', or '+'
                'preserve_comments' => false, // Set to true to preserve comments, or set to an array of strings to preserve specific comments
                'use_autolinks' => true, // Set to true to use simple link syntax if possible. Will always use []() if set to false
                'table_pipe_escape' => '\|', // Replacement string for pipe characters inside markdown table cells
                'table_caption_side' => 'top', // Set to 'top' or 'bottom' to show <caption> content before or after table, null to suppress
            );

            $this->environment = Environment::createDefaultEnvironment($defaults);

            $this->environment->getConfig()->merge($options);
        }
    }

    /**
     * @return Environment
     */
    public function getEnvironment()
    {
        return $this->environment;
    }

    /**
     * @return Configuration
     */
    public function getConfig()
    {
        return $this->environment->getConfig();
    }

    /**
     * Convert
     *
     * @see HtmlConverter::convert
     *
     * @param string $html
     *
     * @return string The Markdown version of the html
     */
    public function __invoke($html)
    {
        return $this->convert($html);
    }

    /**
     * Convert
     *
     * Loads HTML and passes to getMarkdown()
     *
     * @param string $html
     *
     * @throws \InvalidArgumentException
     *
     * @return string The Markdown version of the html
     */
    public function convert($html)
    {
        if (trim($html) === '') {
            return '';
        }

        $document = $this->createDOMDocument($html);

        // Work on the entire DOM tree (including head and body)
        if (!($root = $document->getElementsByTagName('html')->item(0))) {
            throw new \InvalidArgumentException('Invalid HTML was provided');
        }

        $rootElement = new Element($root);
        $this->convertChildren($rootElement);

        // Store the now-modified DOMDocument as a string
        $markdown = $document->saveHTML();

        return $this->sanitize($markdown);
    }

    /**
     * @param string $html
     *
     * @return \DOMDocument
     */
    private function createDOMDocument($html)
    {
        $document = new \DOMDocument();

        if ($this->getConfig()->getOption('suppress_errors')) {
            // Suppress conversion errors (from http://bit.ly/pCCRSX)
            libxml_use_internal_errors(true);
        }

        // Hack to load utf-8 HTML (from http://bit.ly/pVDyCt)
        $document->loadHTML('<?xml encoding="UTF-8">' . $html);
        $document->encoding = 'UTF-8';

        if ($this->getConfig()->getOption('suppress_errors')) {
            libxml_clear_errors();
        }

        return $document;
    }

    /**
     * Convert Children
     *
     * Recursive function to drill into the DOM and convert each node into Markdown from the inside out.
     *
     * Finds children of each node and convert those to #text nodes containing their Markdown equivalent,
     * starting with the innermost element and working up to the outermost element.
     *
     * @param ElementInterface $element
     */
    private function convertChildren(ElementInterface $element)
    {
        // Don't convert HTML code inside <code> and <pre> blocks to Markdown - that should stay as HTML
        // except if the current node is a code tag, which needs to be converted by the CodeConverter.
        if ($element->isDescendantOf(array('pre', 'code')) && $element->getTagName() !== 'code') {
            return;
        }

        // Give converter a chance to inspect/modify the DOM before children are converted
        $converter = $this->environment->getConverterByTag($element->getTagName());
        if ($converter instanceof PreConverterInterface) {
            $converter->preConvert($element);
        }

        // If the node has children, convert those to Markdown first
        if ($element->hasChildren()) {
            foreach ($element->getChildren() as $child) {
                $this->convertChildren($child);
            }
        }

        // Now that child nodes have been converted, convert the original node
        $markdown = $this->convertToMarkdown($element);

        // Create a DOM text node containing the Markdown equivalent of the original node

        // Replace the old $node e.g. '<h3>Title</h3>' with the new $markdown_node e.g. '### Title'
        $element->setFinalMarkdown($markdown);
<?php

function getDate($date)
{
    if ($date !== null) {
        return new DateTime($date);
    }

    return false;
}
    }

    /**
     * Convert to Markdown
     *
     * Converts an individual node into a #text node containing a string of its Markdown equivalent.
     *
     * Example: An <h3> node with text content of 'Title' becomes a text node with content of '### Title'
     *
     * @param ElementInterface $element
     *
     * @return string The converted HTML as Markdown
     */
    protected function convertToMarkdown(ElementInterface $element)
    {
        $tag = $element->getTagName();

        // Strip nodes named in remove_nodes
        $tags_to_remove = explode(' ', $this->getConfig()->getOption('remove_nodes'));
        if (in_array($tag, $tags_to_remove)) {
            return false;
        }

        $converter = $this->environment->getConverterByTag($tag);

        return $converter->convert($element);
    }

    /**
     * @param string $markdown
     *
     * @return string
     */
    protected function sanitize($markdown)
    {
        $markdown = html_entity_decode($markdown, ENT_QUOTES, 'UTF-8');
        $markdown = preg_replace('/<!DOCTYPE [^>]+>/', '', $markdown); // Strip doctype declaration
        $markdown = trim($markdown); // Remove blank spaces at the beggining of the html

        /*
         * Removing unwanted tags. Tags should be added to the array in the order they are expected.
         * XML, html and body opening tags should be in that order. Same case with closing tags
         */
        $unwanted = array('<?xml encoding="UTF-8">', '<html>', '</html>', '<body>', '</body>', '<head>', '</head>', '&#xD;');

        foreach ($unwanted as $tag) {
            if (strpos($tag, '/') === false) {
                // Opening tags
                if (strpos($markdown, $tag) === 0) {
                    $markdown = substr($markdown, strlen($tag));
                }
            } else {
                // Closing tags
                if (strpos($markdown, $tag) === strlen($markdown) - strlen($tag)) {
                    $markdown = substr($markdown, 0, -strlen($tag));
                }
            }
        }

        return trim($markdown, "\n\r\0\x0B");
    }

    /**
     * Pass a series of key-value pairs in an array; these will be passed
     * through the config and set.
     * The advantage of this is that it can allow for static use (IE in Laravel).
     * An example being:
     *
     * HtmlConverter::setOptions(['strip_tags' => true])->convert('<h1>test</h1>');
     */
    public function setOptions(array $options)
    {
        $config = $this->getConfig();

        foreach ($options as $key => $option) {
            $config->setOption($key, $option);
        }

        return $this;
    }
}


1		<?php
2
3		namespace League\HTMLToMarkdown;
4
5		use League\HTMLToMarkdown\Converter\PreConverterInterface;
6
7		/**
8		* Class HtmlConverter
9		*
10		* A helper class to convert HTML to Markdown.
11		*
12		* @author Colin O'Dell <[email protected]>
13		* @author Nick Cernis <[email protected]>
14		*
15		* @link https://github.com/thephpleague/html-to-markdown/ Latest version on GitHub.
16		*
17		* @license http://www.opensource.org/licenses/mit-license.php MIT
18		*/
19		class HtmlConverter implements HtmlConverterInterface
20		{
21		/**
22		* @var Environment
23		*/
24		protected $environment;
25
26		/**
27		* Constructor
28		*
29		* @param Environment\|array $options Environment object or configuration options
30		*/
31	93	public function __construct($options = array())
32		{
33	93	if ($options instanceof Environment) {
34	3	$this->environment = $options;
35	90	} elseif (is_array($options)) {
36		$defaults = array(
37	90	'header_style' => 'setext', // Set to 'atx' to output H1 and H2 headers as # Header1 and ## Header2
38		'suppress_errors' => true, // Set to false to show warnings when loading malformed HTML
39		'strip_tags' => false, // Set to true to strip tags that don't have markdown equivalents. N.B. Strips tags, not their content. Useful to clean MS Word HTML output.
40		'bold_style' => '**', // DEPRECATED: Set to '__' if you prefer the underlined style
41		'italic_style' => '*', // DEPRECATED: Set to '_' if you prefer the underlined style
42		'remove_nodes' => '', // space-separated list of dom nodes that should be removed. example: 'meta style script'
43		'hard_break' => false, // Set to true to turn <br> into `\n` instead of ` \n`
44		'list_item_style' => '-', // Set the default character for each <li> in a <ul>. Can be '-', '*', or '+'
45		'preserve_comments' => false, // Set to true to preserve comments, or set to an array of strings to preserve specific comments
46		'use_autolinks' => true, // Set to true to use simple link syntax if possible. Will always use []() if set to false
47		'table_pipe_escape' => '\\|', // Replacement string for pipe characters inside markdown table cells
48		'table_caption_side' => 'top', // Set to 'top' or 'bottom' to show <caption> content before or after table, null to suppress
49		);
50
51	90	$this->environment = Environment::createDefaultEnvironment($defaults);
52
53	90	$this->environment->getConfig()->merge($options);
54		}
55	93	}
56
57		/**
58		* @return Environment
59		*/
60	3	public function getEnvironment()
61		{
62	3	return $this->environment;
63		}
64
65		/**
66		* @return Configuration
67		*/
68	90	public function getConfig()
69		{
70	90	return $this->environment->getConfig();
71		}
72
73		/**
74		* Convert
75		*
76		* @see HtmlConverter::convert
77		*
78		* @param string $html
79		*
80		* @return string The Markdown version of the html
81		*/
82	3	public function __invoke($html)
83		{
84	3	return $this->convert($html);
85		}
86
87		/**
88		* Convert
89		*
90		* Loads HTML and passes to getMarkdown()
91		*
92		* @param string $html
93		*
94		* @throws \InvalidArgumentException
95		*
96		* @return string The Markdown version of the html
97		*/
98	93	public function convert($html)
99		{
100	93	if (trim($html) === '') {
101	3	return '';
102		}
103
104	90	$document = $this->createDOMDocument($html);
105
106		// Work on the entire DOM tree (including head and body)
107	90	if (!($root = $document->getElementsByTagName('html')->item(0))) {
108		throw new \InvalidArgumentException('Invalid HTML was provided');
109		}
110
111	90	$rootElement = new Element($root);
112	90	$this->convertChildren($rootElement);
113
114		// Store the now-modified DOMDocument as a string
115	90	$markdown = $document->saveHTML();
116
117	90	return $this->sanitize($markdown);
118		}
119
120		/**
121		* @param string $html
122		*
123		* @return \DOMDocument
124		*/
125	90	private function createDOMDocument($html)
126		{
127	90	$document = new \DOMDocument();
128
129	90	if ($this->getConfig()->getOption('suppress_errors')) {
130		// Suppress conversion errors (from http://bit.ly/pCCRSX)
131	87	libxml_use_internal_errors(true);
132		}
133
134		// Hack to load utf-8 HTML (from http://bit.ly/pVDyCt)
135	90	$document->loadHTML('<?xml encoding="UTF-8">' . $html);
136	90	$document->encoding = 'UTF-8';
137
138	90	if ($this->getConfig()->getOption('suppress_errors')) {
139	87	libxml_clear_errors();
140		}
141
142	90	return $document;
143		}
144
145		/**
146		* Convert Children
147		*
148		* Recursive function to drill into the DOM and convert each node into Markdown from the inside out.
149		*
150		* Finds children of each node and convert those to #text nodes containing their Markdown equivalent,
151		* starting with the innermost element and working up to the outermost element.
152		*
153		* @param ElementInterface $element
154		*/
155	90	private function convertChildren(ElementInterface $element)
156		{
157		// Don't convert HTML code inside <code> and <pre> blocks to Markdown - that should stay as HTML
158		// except if the current node is a code tag, which needs to be converted by the CodeConverter.
159	90	if ($element->isDescendantOf(array('pre', 'code')) && $element->getTagName() !== 'code') {
160	21	return;
161		}
162
163		// Give converter a chance to inspect/modify the DOM before children are converted
164	90	$converter = $this->environment->getConverterByTag($element->getTagName());
165	90	if ($converter instanceof PreConverterInterface) {
166	3	$converter->preConvert($element);
167		}
168
169		// If the node has children, convert those to Markdown first
170	90	if ($element->hasChildren()) {
171	90	foreach ($element->getChildren() as $child) {
172	90	$this->convertChildren($child);
173		}
174		}
175
176		// Now that child nodes have been converted, convert the original node
177	90	$markdown = $this->convertToMarkdown($element);
178
179		// Create a DOM text node containing the Markdown equivalent of the original node
180
181		// Replace the old $node e.g. '<h3>Title</h3>' with the new $markdown_node e.g. '### Title'
182	90	$element->setFinalMarkdown($markdown);
		0 ignored issues – show Security Bug introduced 2016-02-10 20:51 UTC by Report Bug Copy Issue Report It seems like `$markdown` defined by `$this->convertToMarkdown($element)` on line `177` can also be of type `false`; however, `League\HTMLToMarkdown\El...ace::setFinalMarkdown()` does only seem to accept `string`, did you maybe forget to handle an error condition? This check looks for type mismatches where the missing type is `false`. This is usually indicative of an error condtion. Consider the follow example <?php function getDate($date) { if ($date !== null) { return new DateTime($date); } return false; } This function either returns a new `DateTime` object or false, if there was an error. This is a typical pattern in PHP programming to show that an error has occurred without raising an exception. The calling code should check for this returned `false` before passing on the value to another function or method that may not be able to handle a `false`. Loading history...
183	90	}
184
185		/**
186		* Convert to Markdown
187		*
188		* Converts an individual node into a #text node containing a string of its Markdown equivalent.
189		*
190		* Example: An <h3> node with text content of 'Title' becomes a text node with content of '### Title'
191		*
192		* @param ElementInterface $element
193		*
194		* @return string The converted HTML as Markdown
195		*/
196	90	protected function convertToMarkdown(ElementInterface $element)
197		{
198	90	$tag = $element->getTagName();
199
200		// Strip nodes named in remove_nodes
201	90	$tags_to_remove = explode(' ', $this->getConfig()->getOption('remove_nodes'));
202	90	if (in_array($tag, $tags_to_remove)) {
203	3	return false;
204		}
205
206	90	$converter = $this->environment->getConverterByTag($tag);
207
208	90	return $converter->convert($element);
209		}
210
211		/**
212		* @param string $markdown
213		*
214		* @return string
215		*/
216	90	protected function sanitize($markdown)
217		{
218	90	$markdown = html_entity_decode($markdown, ENT_QUOTES, 'UTF-8');
219	90	$markdown = preg_replace('/<!DOCTYPE [^>]+>/', '', $markdown); // Strip doctype declaration
220	90	$markdown = trim($markdown); // Remove blank spaces at the beggining of the html
221
222		/*
223		* Removing unwanted tags. Tags should be added to the array in the order they are expected.
224		* XML, html and body opening tags should be in that order. Same case with closing tags
225		*/
226	90	$unwanted = array('<?xml encoding="UTF-8">', '<html>', '</html>', '<body>', '</body>', '<head>', '</head>', ' ');
227
228	90	foreach ($unwanted as $tag) {
229	90	if (strpos($tag, '/') === false) {
230		// Opening tags
231	90	if (strpos($markdown, $tag) === 0) {
232	90	$markdown = substr($markdown, strlen($tag));
233		}
234		} else {
235		// Closing tags
236	90	if (strpos($markdown, $tag) === strlen($markdown) - strlen($tag)) {
237	81	$markdown = substr($markdown, 0, -strlen($tag));
238		}
239		}
240		}
241
242	90	return trim($markdown, "\n\r\0\x0B");
243		}
244
245		/**
246		* Pass a series of key-value pairs in an array; these will be passed
247		* through the config and set.
248		* The advantage of this is that it can allow for static use (IE in Laravel).
249		* An example being:
250		*
251		* HtmlConverter::setOptions(['strip_tags' => true])->convert('<h1>test</h1>');
252		*/
253		public function setOptions(array $options)
254		{
255		$config = $this->getConfig();
256
257		foreach ($options as $key => $option) {
258		$config->setOption($key, $option);
259		}
260
261		return $this;
262		}
263		}
264

thephpleague / html-to-markdown

Push — master ( 0868ae...a5d47a )

HtmlConverter::convertChildren() B

Complexity

Size

Duplication

Code Coverage

Importance

Duplication Side-by-Side

Filter issues like