就我所知，这里所有的解只有在起点固定的情况下才有效。 允许你转动这个: 悲伤之神，神圣之神，痛苦之神，痛苦之神。Ut enim ad minim veniam。 到这个: 神圣的，神圣的… 如果想要截断一组特定关键字周围的单词，该怎么办?

截断一组特定关键字周围的文本。

我们的目标是能够转换这个:

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna liqua. Ut enim ad minim veniam.

到这个:

...consectetur adipisicing elit, sed do eiusmod tempor...

这是在显示搜索结果、摘要等时非常常见的情况。为了实现这一点，我们可以结合使用以下两种方法:

    /**
     * Return the index of the $haystack matching $needle,
     * or NULL if there is no match.
     *
     * This function is case-insensitive  
     * 
     * @param string $needle
     * @param array $haystack
     * @return false|int
     */
    function regexFindInArray(string $needle, array $haystack): ?int
    {
        for ($i = 0; $i < count($haystack); $i++) {
            if (preg_match('/' . preg_quote($needle) . '/i', $haystack[$i]) === 1) {
                return $i;
            }
        }
        return null;
    }

    /**
     * If the keyword is not present, it returns the maximum number of full 
     * words that the max number of characters provided by $maxLength allow,
     * starting from the left.
     *
     * If the keyword is present, it adds words to both sides of the keyword
     * keeping a balanace between the length of the suffix and the prefix.
     *
     * @param string $text
     * @param string $keyword
     * @param int $maxLength
     * @param string $ellipsis
     * @return string
     */
    function truncateWordSurroundingsByLength(string $text, string $keyword, 
            int $maxLength, string $ellipsis): string
    {
        if (strlen($text) < $maxLength) {
            return $text;
        }

        $pattern = '/' . '^(.*?)\s' .
                   '([^\s]*' . preg_quote($keyword) . '[^\s]*)' .
                   '\s(.*)$' . '/i';
        preg_match($pattern, $text, $matches);

        // break everything into words except the matching keywords, 
        // which can contain spaces
        if (count($matches) == 4) {
            $words = preg_split("/\s+/", $matches[1], -1, PREG_SPLIT_NO_EMPTY);
            $words[] = $matches[2];
            $words = array_merge($words, 
                              preg_split("/\s+/", $matches[3], -1, PREG_SPLIT_NO_EMPTY));
        } else {
            $words = preg_split("/\s+/", $text, -1, PREG_SPLIT_NO_EMPTY);
        }

        // find the index of the matching word
        $firstMatchingWordIndex = regexFindInArray($keyword, $words) ?? 0;

        $length = false;
        $prefixLength = $suffixLength = 0;
        $prefixIndex = $firstMatchingWordIndex - 1;
        $suffixIndex = $firstMatchingWordIndex + 1;

        // Initialize the text with the matching word
        $text = $words[$firstMatchingWordIndex];

        while (($prefixIndex >= 0 or $suffixIndex <= count($words))
                and strlen($text) < $maxLength and strlen($text) !== $length) {
            $length = strlen($text);
            if (isset($words[$prefixIndex])
                and (strlen($text) + strlen($words[$prefixIndex]) <= $maxLength)
                and ($prefixLength <= $suffixLength 
                     or strlen($text) + strlen($words[$suffixIndex]) <= $maxLength)) {
                $prefixLength += strlen($words[$prefixIndex]);
                $text = $words[$prefixIndex] . ' ' . $text;
                $prefixIndex--;
            }
            if (isset($words[$suffixIndex])
                and (strlen($text) + strlen($words[$suffixIndex]) <= $maxLength)
                and ($suffixLength <= $prefixLength 
                     or strlen($text) + strlen($words[$prefixIndex]) <= $maxLength)) {
                $suffixLength += strlen($words[$suffixIndex]);
                $text = $text . ' ' . $words[$suffixIndex];
                $suffixIndex++;
            }
        }

        if ($prefixIndex > 0) {
            $text = $ellipsis . ' ' . $text;
        }
        if ($suffixIndex < count($words)) {
            $text = $text . ' ' . $ellipsis;
        }

        return $text;
    }

现在你可以做:

$text = 'Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do' .
        'iusmod tempor incididunt ut labore et dolore magna liqua. Ut enim' .
        'ad minim veniam.';

$text = truncateWordSurroundingsByLength($text, 'elit', 25, '...');

var_dump($text); // string(32) "... adipisicing elit, sed do ..."

运行代码。

我将使用preg_match函数来做到这一点，因为您想要的是一个相当简单的表达式。

$matches = array();
$result = preg_match("/^(.{1,199})[\s]/i", $text, $matches);

表达式的意思是“匹配从长度1-200开始以空格结尾的任何子字符串”。结果在$result中，匹配在$matches中。这就解决了你最初的问题，即以任意空格结束的问题。如果你想让它以换行符结束，将正则表达式更改为:

$result = preg_match("/^(.{1,199})[\n]/i", $text, $matches);

找到这个问题的完美解决方案是多么棘手，这令人惊讶。我还没有在本页上找到至少在某些情况下不会失败的答案(特别是如果字符串包含换行符或制表符，或者如果单词break不是空格，或者如果字符串有UTF-8多字节字符)。

这里有一个简单的解决方案，适用于所有情况。这里也有类似的答案，但是如果希望它处理多行输入，“s”修饰符很重要，而“u”修饰符使它正确地计算UTF-8多字节字符。

function wholeWordTruncate($s, $characterCount) 
{
    if (preg_match("/^.{1,$characterCount}\b/su", $s, $match)) return $match[0];
    return $s;
}

一个可能的边缘情况是…如果字符串的第一个$characterCount字符中没有任何空格，它将返回整个字符串。如果你喜欢在$characterCount处强制中断，即使它不是一个单词边界，你可以使用这个:

function wholeWordTruncate($s, $characterCount) 
{
    if (preg_match("/^.{1,$characterCount}\b/su", $s, $match)) return $match[0];
    return mb_substr($return, 0, $characterCount);
}

最后一个选项，如果你想让它添加省略号如果它截断字符串…

function wholeWordTruncate($s, $characterCount, $addEllipsis = ' …') 
{
    $return = $s;
    if (preg_match("/^.{1,$characterCount}\b/su", $s, $match)) 
        $return = $match[0];
    else
        $return = mb_substr($return, 0, $characterCount);
    if (strlen($s) > strlen($return)) $return .= $addEllipsis;
    return $return;
}

我发现这很有效:

function abbreviate_string_to_whole_word($string, $max_length, $buffer) {
    if (strlen($string) > $max_length) {
        $string_cropped = substr($string, 0, $max_length - $buffer);
        $last_space = strrpos($string_cropped, " ");
        if ($last_space > 0) {
            $string_cropped = substr($string_cropped, 0, $last_space);
        }
        $abbreviated_string = $string_cropped . " ...";
    }
    else {
        $abbreviated_string = $string;
    }
    return $abbreviated_string;
}

缓冲区允许您调整返回字符串的长度。

下面是基于@Cd-MaN方法的函数。

function shorten($string, $width) {
  if(strlen($string) > $width) {
    $string = wordwrap($string, $width);
    $string = substr($string, 0, strpos($string, "\n"));
  }

  return $string;
}

你可以试试这个

substr( $str, 0, strpos($str, ' ', 200) );