CompletionContext::tokenizeString() - Code Metrics - Inspection of "Merge pull request #82 from stecman/string-argumen..." - stecman/symfony-console-completion - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( bd07a2...bc095f )

by Stephen

created 2019-04-26 12:14 UTC

CompletionContext::tokenizeString() B

↳ Parent: CompletionContext

Complexity

Conditions	6
Paths	6

Size

Total Lines

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
dl	0
loc	64
rs	8.1632
c	0
b	0
f	0
cc	6
nc	6
nop	1

How to fix Long Method

<?php


namespace Stecman\Component\Symfony\Console\BashCompletion;

/**
 * Command line context for completion
 *
 * Represents the current state of the command line that is being completed
 */
class CompletionContext
{
    /**
     * The current contents of the command line as a single string
     *
     * Bash equivalent: COMP_LINE
     *
     * @var string
     */
    protected $commandLine;

    /**
     * The index of the user's cursor relative to the start of the command line.
     *
     * If the current cursor position is at the end of the current command,
     * the value of this variable is equal to the length of $this->commandLine
     *
     * Bash equivalent: COMP_POINT
     *
     * @var int
     */
    protected $charIndex = 0;

    /**
     * An array of the individual words in the current command line.
     *
     * This is not set until $this->splitCommand() is called, when it is populated by
     * $commandLine exploded by $wordBreaks
     *
     * Bash equivalent: COMP_WORDS
     *
     * @var string[]|null
     */
    protected $words = null;

    /**
     * Words from the currently command-line before quotes and escaping is processed
     *
     * This is indexed the same as $this->words, but in their raw input terms are in their input form, including
     * quotes and escaping.
     *
     * @var string[]|null
     */
    protected $rawWords = null;

    /**
     * The index in $this->words containing the word at the current cursor position.
     *
     * This is not set until $this->splitCommand() is called.
     *
     * Bash equivalent: COMP_CWORD
     *
     * @var int|null
     */
    protected $wordIndex = null;

    /**
     * Characters that $this->commandLine should be split on to get a list of individual words
     *
     * Bash equivalent: COMP_WORDBREAKS
     *
     * @var string
     */
    protected $wordBreaks = "= \t\n";

    /**
     * Set the whole contents of the command line as a string
     *
     * @param string $commandLine
     */
    public function setCommandLine($commandLine)
    {
        $this->commandLine = $commandLine;
        $this->reset();
    }

    /**
     * Return the current command line verbatim as a string
     *
     * @return string
     */
    public function getCommandLine()
    {
        return $this->commandLine;
    }

    /**
     * Return the word from the command line that the cursor is currently in
     *
     * Most of the time this will be a partial word. If the cursor has a space before it,
     * this will return an empty string, indicating a new word.
     *
     * @return string
     */
    public function getCurrentWord()
    {
        if (isset($this->words[$this->wordIndex])) {
            return $this->words[$this->wordIndex];
        }

        return '';
    }

    /**
     * Return the unprocessed string for the word under the cursor
     *
     * This preserves any quotes and escaping that are present in the input command line.
     *
     * @return string
     */
    public function getRawCurrentWord()
    {
        if (isset($this->rawWords[$this->wordIndex])) {
            return $this->rawWords[$this->wordIndex];
        }

        return '';
    }

    /**
     * Return a word by index from the command line
     *
     * @see $words, $wordBreaks
     * @param int $index
     * @return string
     */
    public function getWordAtIndex($index)
    {
        if (isset($this->words[$index])) {
            return $this->words[$index];
        }

        return '';
    }

    /**
     * Get the contents of the command line, exploded into words based on the configured word break characters
     *
     * @see $wordBreaks, setWordBreaks
     * @return array
     */
    public function getWords()
    {
        if ($this->words === null) {
            $this->splitCommand();
        }

        return $this->words;
    }

    /**
     * Get the unprocessed/literal words from the command line
     *
     * This is indexed the same as getWords(), but preserves any quoting and escaping from the command line
     *
     * @return string[]
     */
    public function getRawWords()
    {
        if ($this->rawWords === null) {
            $this->splitCommand();
        }

        return $this->rawWords;
    }

    /**
     * Get the index of the word the cursor is currently in
     *
     * @see getWords, getCurrentWord
     * @return int
     */
    public function getWordIndex()
    {
        if ($this->wordIndex === null) {
            $this->splitCommand();
        }

        return $this->wordIndex;
    }

    /**
     * Get the character index of the user's cursor on the command line
     *
     * This is in the context of the full command line string, so includes word break characters.
     * Note that some shells can only provide an approximation for character index. Under ZSH for
     * example, this will always be the character at the start of the current word.
     *
     * @return int
     */
    public function getCharIndex()
    {
        return $this->charIndex;
    }

    /**
     * Set the cursor position as a character index relative to the start of the command line
     *
     * @param int $index
     */
    public function setCharIndex($index)
    {
        $this->charIndex = $index;
        $this->reset();
    }

    /**
     * Set characters to use as split points when breaking the command line into words
     *
     * This defaults to a sane value based on BASH's word break characters and shouldn't
     * need to be changed unless your completions contain the default word break characters.
     *
     * @deprecated This is becoming an internal setting that doesn't make sense to expose publicly.
     *
     * @see wordBreaks
     * @param string $charList - a single string containing all of the characters to break words on
     */
    public function setWordBreaks($charList)
    {
        // Drop quotes from break characters - strings are handled separately to word breaks now
        $this->wordBreaks = str_replace(array('"', '\''), '', $charList);;
        $this->reset();
    }

    /**
     * Split the command line into words using the configured word break characters
     *
     * @return string[]
     */
    protected function splitCommand()
    {
        $tokens = $this->tokenizeString($this->commandLine);

        foreach ($tokens as $token) {
            if ($token['type'] != 'break') {
                $this->words[] = $this->getTokenValue($token);
                $this->rawWords[] = $token['value'];
            }

            // Determine which word index the cursor is inside once we reach it's offset
            if ($this->wordIndex === null && $this->charIndex <= $token['offsetEnd']) {
                $this->wordIndex = count($this->words) - 1;

                if ($token['type'] == 'break') {
                    // Cursor is in the break-space after a word
                    // Push an empty word at the cursor to allow completion of new terms at the cursor, ignoring words ahead
                    $this->wordIndex++;
                    $this->words[] = '';
                    $this->rawWords[] = '';
                    continue;
                }

                if ($this->charIndex < $token['offsetEnd']) {
                    // Cursor is inside the current word - truncate the word at the cursor to complete on
                    // This emulates BASH completion's behaviour with COMP_CWORD

                    // Create a copy of the token with its value truncated
                    $truncatedToken = $token;
                    $relativeOffset = $this->charIndex - $token['offset'];
                    $truncatedToken['value'] = substr($token['value'], 0, $relativeOffset);

                    // Replace the current word with the truncated value
                    $this->words[$this->wordIndex] = $this->getTokenValue($truncatedToken);
                    $this->rawWords[$this->wordIndex] = $truncatedToken['value'];
                }
            }
        }

        // Cursor position is past the end of the command line string - consider it a new word
        if ($this->wordIndex === null) {
            $this->wordIndex = count($this->words);
            $this->words[] = '';
            $this->rawWords[] = '';
        }
    }

    /**
     * Return a token's value with escaping and quotes removed
     *
     * @see self::tokenizeString()
     * @param array $token
     * @return string
     */
    protected function getTokenValue($token)
    {
        $value = $token['value'];

        // Remove outer quote characters (or first quote if unclosed)
        if ($token['type'] == 'quoted') {
            $value = preg_replace('/^(?:[\'"])(.*?)(?:[\'"])?$/', '$1', $value);
        }

        // Remove escape characters
        $value = preg_replace('/\\\\(.)/', '$1', $value);

        return $value;
    }

    /**
     * Break a string into words, quoted strings and non-words (breaks)
     *
     * Returns an array of unmodified segments of $string with offset and type information.
     *
     * @param string $string
     * @return array as [ [type => string, value => string, offset => int], ... ]
     */
    protected function tokenizeString($string)
    {
        // Map capture groups to returned token type
        $typeMap = array(
            'double_quote_string' => 'quoted',
            'single_quote_string' => 'quoted',
            'word' => 'word',
            'break' => 'break',
        );

        // Escape every word break character including whitespace
        // preg_quote won't work here as it doesn't understand the ignore whitespace flag ("x")
        $breaks = preg_replace('/(.)/', '\\\$1', $this->wordBreaks);

        $pattern = <<<"REGEX"
            /(?:
                (?P<double_quote_string>
                    "(\\\\.|[^\"\\\\])*(?:"|$)
                ) |
                (?P<single_quote_string>
                    '(\\\\.|[^'\\\\])*(?:'|$)
                ) |
                (?P<word>
                    (?:\\\\.|[^$breaks])+
                ) |
                (?P<break>
                     [$breaks]+
                )
            )/x
REGEX;

        $tokens = array();

        if (!preg_match_all($pattern, $string, $matches, PREG_OFFSET_CAPTURE | PREG_SET_ORDER)) {
            return $tokens;
        }

        foreach ($matches as $set) {
            foreach ($set as $groupName => $match) {

                // Ignore integer indices preg_match outputs (duplicates of named groups)
                if (is_integer($groupName)) {
                    continue;
                }

                // Skip if the offset indicates this group didn't match
                if ($match[1] === -1) {

                    continue;
                }

                $tokens[] = array(
                    'type' => $typeMap[$groupName],
                    'value' => $match[0],
                    'offset' => $match[1],
                    'offsetEnd' => $match[1] + strlen($match[0])
                );

                // Move to the next set (only one group should match per set)
                continue;
            }
        }

        return $tokens;
    }

    /**
     * Reset the computed words so that $this->splitWords is forced to run again
     */
    protected function reset()
    {
        $this->words = null;
        $this->wordIndex = null;
    }
}


1			<?php
2
3
4			namespace Stecman\Component\Symfony\Console\BashCompletion;
5
6			/**
7			* Command line context for completion
8			*
9			* Represents the current state of the command line that is being completed
10			*/
11			class CompletionContext
12			{
13			/**
14			* The current contents of the command line as a single string
15			*
16			* Bash equivalent: COMP_LINE
17			*
18			* @var string
19			*/
20			protected $commandLine;
21
22			/**
23			* The index of the user's cursor relative to the start of the command line.
24			*
25			* If the current cursor position is at the end of the current command,
26			* the value of this variable is equal to the length of $this->commandLine
27			*
28			* Bash equivalent: COMP_POINT
29			*
30			* @var int
31			*/
32			protected $charIndex = 0;
33
34			/**
35			* An array of the individual words in the current command line.
36			*
37			* This is not set until $this->splitCommand() is called, when it is populated by
38			* $commandLine exploded by $wordBreaks
39			*
40			* Bash equivalent: COMP_WORDS
41			*
42			* @var string[]\|null
43			*/
44			protected $words = null;
45
46			/**
47			* Words from the currently command-line before quotes and escaping is processed
48			*
49			* This is indexed the same as $this->words, but in their raw input terms are in their input form, including
50			* quotes and escaping.
51			*
52			* @var string[]\|null
53			*/
54			protected $rawWords = null;
55
56			/**
57			* The index in $this->words containing the word at the current cursor position.
58			*
59			* This is not set until $this->splitCommand() is called.
60			*
61			* Bash equivalent: COMP_CWORD
62			*
63			* @var int\|null
64			*/
65			protected $wordIndex = null;
66
67			/**
68			* Characters that $this->commandLine should be split on to get a list of individual words
69			*
70			* Bash equivalent: COMP_WORDBREAKS
71			*
72			* @var string
73			*/
74			protected $wordBreaks = "= \t\n";
75
76			/**
77			* Set the whole contents of the command line as a string
78			*
79			* @param string $commandLine
80			*/
81			public function setCommandLine($commandLine)
82			{
83			$this->commandLine = $commandLine;
84			$this->reset();
85			}
86
87			/**
88			* Return the current command line verbatim as a string
89			*
90			* @return string
91			*/
92			public function getCommandLine()
93			{
94			return $this->commandLine;
95			}
96
97			/**
98			* Return the word from the command line that the cursor is currently in
99			*
100			* Most of the time this will be a partial word. If the cursor has a space before it,
101			* this will return an empty string, indicating a new word.
102			*
103			* @return string
104			*/
105			public function getCurrentWord()
106			{
107			if (isset($this->words[$this->wordIndex])) {
108			return $this->words[$this->wordIndex];
109			}
110
111			return '';
112			}
113
114			/**
115			* Return the unprocessed string for the word under the cursor
116			*
117			* This preserves any quotes and escaping that are present in the input command line.
118			*
119			* @return string
120			*/
121			public function getRawCurrentWord()
122			{
123			if (isset($this->rawWords[$this->wordIndex])) {
124			return $this->rawWords[$this->wordIndex];
125			}
126
127			return '';
128			}
129
130			/**
131			* Return a word by index from the command line
132			*
133			* @see $words, $wordBreaks
134			* @param int $index
135			* @return string
136			*/
137			public function getWordAtIndex($index)
138			{
139			if (isset($this->words[$index])) {
140			return $this->words[$index];
141			}
142
143			return '';
144			}
145
146			/**
147			* Get the contents of the command line, exploded into words based on the configured word break characters
148			*
149			* @see $wordBreaks, setWordBreaks
150			* @return array
151			*/
152			public function getWords()
153			{
154			if ($this->words === null) {
155			$this->splitCommand();
156			}
157
158			return $this->words;
159			}
160
161			/**
162			* Get the unprocessed/literal words from the command line
163			*
164			* This is indexed the same as getWords(), but preserves any quoting and escaping from the command line
165			*
166			* @return string[]
167			*/
168			public function getRawWords()
169			{
170			if ($this->rawWords === null) {
171			$this->splitCommand();
172			}
173
174			return $this->rawWords;
175			}
176
177			/**
178			* Get the index of the word the cursor is currently in
179			*
180			* @see getWords, getCurrentWord
181			* @return int
182			*/
183			public function getWordIndex()
184			{
185			if ($this->wordIndex === null) {
186			$this->splitCommand();
187			}
188
189			return $this->wordIndex;
190			}
191
192			/**
193			* Get the character index of the user's cursor on the command line
194			*
195			* This is in the context of the full command line string, so includes word break characters.
196			* Note that some shells can only provide an approximation for character index. Under ZSH for
197			* example, this will always be the character at the start of the current word.
198			*
199			* @return int
200			*/
201			public function getCharIndex()
202			{
203			return $this->charIndex;
204			}
205
206			/**
207			* Set the cursor position as a character index relative to the start of the command line
208			*
209			* @param int $index
210			*/
211			public function setCharIndex($index)
212			{
213			$this->charIndex = $index;
214			$this->reset();
215			}
216
217			/**
218			* Set characters to use as split points when breaking the command line into words
219			*
220			* This defaults to a sane value based on BASH's word break characters and shouldn't
221			* need to be changed unless your completions contain the default word break characters.
222			*
223			* @deprecated This is becoming an internal setting that doesn't make sense to expose publicly.
224			*
225			* @see wordBreaks
226			* @param string $charList - a single string containing all of the characters to break words on
227			*/
228			public function setWordBreaks($charList)
229			{
230			// Drop quotes from break characters - strings are handled separately to word breaks now
231			$this->wordBreaks = str_replace(array('"', '\''), '', $charList);;
232			$this->reset();
233			}
234
235			/**
236			* Split the command line into words using the configured word break characters
237			*
238			* @return string[]
239			*/
240			protected function splitCommand()
241			{
242			$tokens = $this->tokenizeString($this->commandLine);
243
244			foreach ($tokens as $token) {
245			if ($token['type'] != 'break') {
246			$this->words[] = $this->getTokenValue($token);
247			$this->rawWords[] = $token['value'];
248			}
249
250			// Determine which word index the cursor is inside once we reach it's offset
251			if ($this->wordIndex === null && $this->charIndex <= $token['offsetEnd']) {
252			$this->wordIndex = count($this->words) - 1;
253
254			if ($token['type'] == 'break') {
255			// Cursor is in the break-space after a word
256			// Push an empty word at the cursor to allow completion of new terms at the cursor, ignoring words ahead
257			$this->wordIndex++;
258			$this->words[] = '';
259			$this->rawWords[] = '';
260			continue;
261			}
262
263			if ($this->charIndex < $token['offsetEnd']) {
264			// Cursor is inside the current word - truncate the word at the cursor to complete on
265			// This emulates BASH completion's behaviour with COMP_CWORD
266
267			// Create a copy of the token with its value truncated
268			$truncatedToken = $token;
269			$relativeOffset = $this->charIndex - $token['offset'];
270			$truncatedToken['value'] = substr($token['value'], 0, $relativeOffset);
271
272			// Replace the current word with the truncated value
273			$this->words[$this->wordIndex] = $this->getTokenValue($truncatedToken);
274			$this->rawWords[$this->wordIndex] = $truncatedToken['value'];
275			}
276			}
277			}
278
279			// Cursor position is past the end of the command line string - consider it a new word
280			if ($this->wordIndex === null) {
281			$this->wordIndex = count($this->words);
282			$this->words[] = '';
283			$this->rawWords[] = '';
284			}
285			}
286
287			/**
288			* Return a token's value with escaping and quotes removed
289			*
290			* @see self::tokenizeString()
291			* @param array $token
292			* @return string
293			*/
294			protected function getTokenValue($token)
295			{
296			$value = $token['value'];
297
298			// Remove outer quote characters (or first quote if unclosed)
299			if ($token['type'] == 'quoted') {
300			$value = preg_replace('/^(?:[\'"])(.*?)(?:[\'"])?$/', '$1', $value);
301			}
302
303			// Remove escape characters
304			$value = preg_replace('/\\\\(.)/', '$1', $value);
305
306			return $value;
307			}
308
309			/**
310			* Break a string into words, quoted strings and non-words (breaks)
311			*
312			* Returns an array of unmodified segments of $string with offset and type information.
313			*
314			* @param string $string
315			* @return array as [ [type => string, value => string, offset => int], ... ]
316			*/
317			protected function tokenizeString($string)
318			{
319			// Map capture groups to returned token type
320			$typeMap = array(
321			'double_quote_string' => 'quoted',
322			'single_quote_string' => 'quoted',
323			'word' => 'word',
324			'break' => 'break',
325			);
326
327			// Escape every word break character including whitespace
328			// preg_quote won't work here as it doesn't understand the ignore whitespace flag ("x")
329			$breaks = preg_replace('/(.)/', '\\\$1', $this->wordBreaks);
330
331			$pattern = <<<"REGEX"
332			/(?:
333			(?P<double_quote_string>
334			"(\\\\.\|[^\"\\\\])*(?:"\|$)
335			) \|
336			(?P<single_quote_string>
337			'(\\\\.\|[^'\\\\])*(?:'\|$)
338			) \|
339			(?P<word>
340			(?:\\\\.\|[^$breaks])+
341			) \|
342			(?P<break>
343			[$breaks]+
344			)
345			)/x
346			REGEX;
347
348			$tokens = array();
349
350			if (!preg_match_all($pattern, $string, $matches, PREG_OFFSET_CAPTURE \| PREG_SET_ORDER)) {
351			return $tokens;
352			}
353
354			foreach ($matches as $set) {
355			foreach ($set as $groupName => $match) {
356
357			// Ignore integer indices preg_match outputs (duplicates of named groups)
358			if (is_integer($groupName)) {
359			continue;
360			}
361
362			// Skip if the offset indicates this group didn't match
363			if ($match[1] === -1) {
			0 ignored issues – show Unused Code Bug introduced 2018-02-12 13:13 UTC by Report Bug Copy Issue Report The strict comparison `===` seems to always evaluate to `false` as the types of `$match[1]` (`string`) and `-1` (`integer`) can never be identical. Maybe you want to use a loose comparison `==` instead? Loading history...
364			continue;
365			}
366
367			$tokens[] = array(
368			'type' => $typeMap[$groupName],
369			'value' => $match[0],
370			'offset' => $match[1],
371			'offsetEnd' => $match[1] + strlen($match[0])
372			);
373
374			// Move to the next set (only one group should match per set)
375			continue;
376			}
377			}
378
379			return $tokens;
380			}
381
382			/**
383			* Reset the computed words so that $this->splitWords is forced to run again
384			*/
385			protected function reset()
386			{
387			$this->words = null;
388			$this->wordIndex = null;
389			}
390			}
391

stecman / symfony-console-completion

Push — master ( bd07a2...bc095f )

CompletionContext::tokenizeString() B

Complexity

Size

Duplication

Importance

How to fix Long Method

Long Method

Duplication Side-by-Side

Filter issues like