HtmlConverter::convertChildren() - Code Metrics - Inspection of "Added stripping of placeholder links (#196)" - thephpleague/html-to-markdown - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( 84e485...c5777c )

by Colin

created 2021-03-28 21:59 UTC

HtmlConverter::convertChildren() B

↳ Parent: HtmlConverter

Complexity

Conditions	6
Paths	5

Size

Total Lines

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	15
CRAP Score	6

Importance

Changes

Metric	Value
dl	0
loc	29
ccs	15
cts	15
cp	1
rs	8.8337
c	0
b	0
f	0
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.
                'strip_placeholder_links' => false, // Set to true to remove <a> that doesn't have href.
                '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 $options Environment object or configuration options
30		*/
31	99	public function __construct($options = array())
32		{
33	99	if ($options instanceof Environment) {
34	3	$this->environment = $options;
35	98	} elseif (is_array($options)) {
36		$defaults = array(
37	96	'header_style' => 'setext', // Set to 'atx' to output H1 and H2 headers as # Header1 and ## Header2
38	64	'suppress_errors' => true, // Set to false to show warnings when loading malformed HTML
39	64	'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	64	'strip_placeholder_links' => false, // Set to true to remove <a> that doesn't have href.
41	64	'bold_style' => '**', // DEPRECATED: Set to '__' if you prefer the underlined style
42	64	'italic_style' => '*', // DEPRECATED: Set to '_' if you prefer the underlined style
43	64	'remove_nodes' => '', // space-separated list of dom nodes that should be removed. example: 'meta style script'
44	64	'hard_break' => false, // Set to true to turn <br> into `\n` instead of ` \n`
45	64	'list_item_style' => '-', // Set the default character for each <li> in a <ul>. Can be '-', '*', or '+'
46	64	'preserve_comments' => false, // Set to true to preserve comments, or set to an array of strings to preserve specific comments
47	64	'use_autolinks' => true, // Set to true to use simple link syntax if possible. Will always use []() if set to false
48	64	'table_pipe_escape' => '\\|', // Replacement string for pipe characters inside markdown table cells
49	64	'table_caption_side' => 'top', // Set to 'top' or 'bottom' to show <caption> content before or after table, null to suppress
50	64	);
51
52	96	$this->environment = Environment::createDefaultEnvironment($defaults);
53
54	96	$this->environment->getConfig()->merge($options);
55	64	}
56	99	}
57
58		/**
59		* @return Environment
60		*/
61	3	public function getEnvironment()
62		{
63	3	return $this->environment;
64		}
65
66		/**
67		* @return Configuration
68		*/
69	96	public function getConfig()
70		{
71	96	return $this->environment->getConfig();
72		}
73
74		/**
75		* Convert
76		*
77		* @see HtmlConverter::convert
78		*
79		* @param string $html
80		*
81		* @return string The Markdown version of the html
82		*/
83	3	public function __invoke($html)
84		{
85	3	return $this->convert($html);
86		}
87
88		/**
89		* Convert
90		*
91		* Loads HTML and passes to getMarkdown()
92		*
93		* @param string $html
94		*
95		* @throws \InvalidArgumentException
96		*
97		* @return string The Markdown version of the html
98		*/
99	99	public function convert($html)
100		{
101	99	if (trim($html) === '') {
102	3	return '';
103		}
104
105	96	$document = $this->createDOMDocument($html);
106
107		// Work on the entire DOM tree (including head and body)
108	96	if (!($root = $document->getElementsByTagName('html')->item(0))) {
109		throw new \InvalidArgumentException('Invalid HTML was provided');
110		}
111
112	96	$rootElement = new Element($root);
113	96	$this->convertChildren($rootElement);
114
115		// Store the now-modified DOMDocument as a string
116	96	$markdown = $document->saveHTML();
117
118	96	return $this->sanitize($markdown);
119		}
120
121		/**
122		* @param string $html
123		*
124		* @return \DOMDocument
125		*/
126	96	private function createDOMDocument($html)
127		{
128	96	$document = new \DOMDocument();
129
130	96	if ($this->getConfig()->getOption('suppress_errors')) {
131		// Suppress conversion errors (from http://bit.ly/pCCRSX)
132	93	libxml_use_internal_errors(true);
133	62	}
134
135		// Hack to load utf-8 HTML (from http://bit.ly/pVDyCt)
136	96	$document->loadHTML('<?xml encoding="UTF-8">' . $html);
137	96	$document->encoding = 'UTF-8';
138
139	96	if ($this->getConfig()->getOption('suppress_errors')) {
140	93	libxml_clear_errors();
141	62	}
142
143	96	return $document;
144		}
145
146		/**
147		* Convert Children
148		*
149		* Recursive function to drill into the DOM and convert each node into Markdown from the inside out.
150		*
151		* Finds children of each node and convert those to #text nodes containing their Markdown equivalent,
152		* starting with the innermost element and working up to the outermost element.
153		*
154		* @param ElementInterface $element
155		*/
156	96	private function convertChildren(ElementInterface $element)
157		{
158		// Don't convert HTML code inside <code> and <pre> blocks to Markdown - that should stay as HTML
159		// except if the current node is a code tag, which needs to be converted by the CodeConverter.
160	96	if ($element->isDescendantOf(array('pre', 'code')) && $element->getTagName() !== 'code') {
161	21	return;
162		}
163
164		// Give converter a chance to inspect/modify the DOM before children are converted
165	96	$converter = $this->environment->getConverterByTag($element->getTagName());
166	96	if ($converter instanceof PreConverterInterface) {
167	3	$converter->preConvert($element);
168	2	}
169
170		// If the node has children, convert those to Markdown first
171	96	if ($element->hasChildren()) {
172	96	foreach ($element->getChildren() as $child) {
173	96	$this->convertChildren($child);
174	64	}
175	64	}
176
177		// Now that child nodes have been converted, convert the original node
178	96	$markdown = $this->convertToMarkdown($element);
179
180		// Create a DOM text node containing the Markdown equivalent of the original node
181
182		// Replace the old $node e.g. '<h3>Title</h3>' with the new $markdown_node e.g. '### Title'
183	96	$element->setFinalMarkdown($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($element)` on line `178` can also be of type `false`; however, `League\HTMLToMarkdown\El...ace::setFinalMarkdown()` does only seem to accept `string`, did you maybe forget to handle an error condition? This check looks for type mismatches where the missing type is `false`. This is usually indicative of an error condtion. Consider the follow example <?php function getDate($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...
184	96	}
185
186		/**
187		* Convert to Markdown
188		*
189		* Converts an individual node into a #text node containing a string of its Markdown equivalent.
190		*
191		* Example: An <h3> node with text content of 'Title' becomes a text node with content of '### Title'
192		*
193		* @param ElementInterface $element
194		*
195		* @return string The converted HTML as Markdown
196		*/
197	96	protected function convertToMarkdown(ElementInterface $element)
198		{
199	96	$tag = $element->getTagName();
200
201		// Strip nodes named in remove_nodes
202	96	$tags_to_remove = explode(' ', $this->getConfig()->getOption('remove_nodes'));
203	96	if (in_array($tag, $tags_to_remove)) {
204	3	return false;
205		}
206
207	96	$converter = $this->environment->getConverterByTag($tag);
208
209	96	return $converter->convert($element);
210		}
211
212		/**
213		* @param string $markdown
214		*
215		* @return string
216		*/
217	96	protected function sanitize($markdown)
218		{
219	96	$markdown = html_entity_decode($markdown, ENT_QUOTES, 'UTF-8');
220	96	$markdown = preg_replace('/<!DOCTYPE [^>]+>/', '', $markdown); // Strip doctype declaration
221	96	$markdown = trim($markdown); // Remove blank spaces at the beggining of the html
222
223		/*
224		* Removing unwanted tags. Tags should be added to the array in the order they are expected.
225		* XML, html and body opening tags should be in that order. Same case with closing tags
226		*/
227	96	$unwanted = array('<?xml encoding="UTF-8">', '<html>', '</html>', '<body>', '</body>', '<head>', '</head>', ' ');
228
229	96	foreach ($unwanted as $tag) {
230	96	if (strpos($tag, '/') === false) {
231		// Opening tags
232	96	if (strpos($markdown, $tag) === 0) {
233	96	$markdown = substr($markdown, strlen($tag));
234	64	}
235	64	} else {
236		// Closing tags
237	96	if (strpos($markdown, $tag) === strlen($markdown) - strlen($tag)) {
238	90	$markdown = substr($markdown, 0, -strlen($tag));
239	58	}
240		}
241	64	}
242
243	96	return trim($markdown, "\n\r\0\x0B");
244		}
245
246		/**
247		* Pass a series of key-value pairs in an array; these will be passed
248		* through the config and set.
249		* The advantage of this is that it can allow for static use (IE in Laravel).
250		* An example being:
251		*
252		* HtmlConverter::setOptions(['strip_tags' => true])->convert('<h1>test</h1>');
253		*/
254		public function setOptions(array $options)
255		{
256		$config = $this->getConfig();
257
258		foreach ($options as $key => $option) {
259		$config->setOption($key, $option);
260		}
261
262		return $this;
263		}
264		}
265

thephpleague / html-to-markdown

Push — master ( 84e485...c5777c )

HtmlConverter::convertChildren() B

Complexity

Size

Duplication

Code Coverage

Importance

Duplication Side-by-Side

Filter issues like