PHP - 在字符串中搜索关键字并提高提取关键字的质量和准确性

Question

我有一段PHP代码如下:

$Keywords = array(
            ', JOE.' => '1',
            ', JOE' => '2',
            'JOE' => '3',
            'JOE.' => '4',
            '/JOE' => '5',
            '/JOE/' => '6',
            'JOE/.' => '7',
            ',JOE.' => '8'
    );
$Text = "JOE is JOE is JOE is JOE is JOE is JOE is JOE. Hello , JOE. Hey ,JOE. Come on , JOE. Dude,JOE/. Shut up ,JOE. What is the meaning of /JOE/? Of course, JOE";

extract_keyword ($Keywords, $Text);

function extract_keyword ($Keywords, $Text){
    mb_internal_encoding('UTF-8');

    uksort($Keywords, function ($a, $b) {
        $as = mb_strlen($a);
        $bs = mb_strlen($b);

        if ($as > $bs) {
            return -1;
        }
        else if ($bs > $as) {
            return 1;
        }
        return 0;

    });

    $Keywords_ci = array();

    foreach ($Keywords as $k => $v) {
        $Keywords_ci[$k] = $v;
    }

    $re = '/\b(?:' . join('|', array_map(function($keyword) {
        return preg_quote($keyword, '/');
    }, array_keys($Keywords))) . ')\b/i';

    $KeywordArrayKey = array();
    $KeywordArrayValue = array();
    $NewArray = array();
    preg_match_all($re, $Text, $matches);
    foreach ($matches[0] as $keyword) {
        $KeywordArrayKey[] = $keyword;
        $KeywordArrayValue[] = $Keywords_ci[$keyword];
        if(!empty($keyword) && !empty($Keywords_ci[$keyword])) {
        $NewArray[] = array($keyword => $Keywords_ci[$keyword]); 
        }
    } 
    print_r($NewArray) ."<br><br>";
}

代码回应如下:

Array ( 
[0] => Array ( [JOE] => 3 ) 
[1] => Array ( [JOE] => 3 ) 
[2] => Array ( [JOE] => 3 ) 
[3] => Array ( [JOE] => 3 ) 
[4] => Array ( [JOE] => 3 ) 
[5] => Array ( [JOE] => 3 ) 
[6] => Array ( [JOE] => 3 ) 
[7] => Array ( [JOE] => 3 ) 
[8] => Array ( [JOE] => 3 ) 
[9] => Array ( [JOE] => 3 ) 
[10] => Array ( [JOE] => 3 ) 
[11] => Array ( [JOE] => 3 ) 
[12] => Array ( [JOE] => 3 ) 
[13] => Array ( [, JOE] => 2 ) )

可以看到，问题是代码不够准确，提取不到$keywords有关键字的地方，比如', JOE .' => '1' 或 '乔/.' => '7'。事实上，我的目标是将 '/JOE' => '5' 与 '/JOE/' => '6' 或 'JOE .' => '4' 等等。能否请您看一下代码，让我知道如何提高提取关键字的质量/准确性？谢谢你的帮助。

注意 1:print_r($Keywords_ci); 打印 Array ( [, JOE.] => 1 [JOE/.] => 7 [,JOE. ] => 8 [, JOE] => 2 [/JOE/] => 6 [JOE.] => 4 [/JOE] => 5 [JOE] => 3 )，但是我是寻找是回显可用关键字的所有实例，例如 '/JOE/' => '6' 或 ',JOE.' => $Text 中的“8”。

注意 2:下面是 print_r($NewArray) 的预期打印:

Array ( 
[0] => Array ( [JOE] => 3 ) 
[1] => Array ( [JOE] => 3 ) 
[2] => Array ( [JOE] => 3 ) 
[3] => Array ( [JOE] => 3 ) 
[4] => Array ( [JOE] => 3 ) 
[5] => Array ( [JOE] => 3 ) 
[6] => Array ( [JOE.] => 4 ) 
[7] => Array ( [, JOE.] => 1 ) 
[8] => Array ( [,JOE.] => 8 ) 
[9] => Array ( [, JOE.] => 1 ) 
[10] => Array ( [JOE/.] => 7 ) 
[11] => Array ( [,JOE.] => 8 ) 
[12] => Array ( [/JOE/] => 6 ) 
[13] => Array ( [, JOE] => 2 ) )

Answer 1

将关键字从最长到最短排序后，您就会知道将在该字符串的任何可能子集之前检查字符串(/JOE/在/JOE 之前)。因此，您可以使用 str_replace 删除实际匹配项，从而在搜索/JOE 时不匹配/JOE/(假设您之前搜索过/JOE/)。使用 str_replace 的 count 参数获取匹配项目的数量

<?php
$Keywords = array(
            ', JOE.' => '1',
            ', JOE' => '2',
            'JOE' => '3',
            'JOE.' => '4',
            '/JOE' => '5',
            '/JOE/' => '6',
            'JOE/.' => '7',
            ',JOE.' => '8'
    );
$Text = "JOE is JOE. Hello , JOE. Hey ,JOE. Come on , JOE. Dude,JOE/. Shut up ,JOE. What is the meaning of /JOE/? Of course, JOE";

uksort($Keywords, function ($a, $b) {
        $as = mb_strlen($a);
        $bs = mb_strlen($b);

        if ($as > $bs) {
            return -1;
        }
        else if ($bs > $as) {
            return 1;
        }
        return 0;

    });

$copy = $Text;
foreach ($Keywords as $keyword => $value) {
   $copy = str_replace($keyword, '', $copy, $count);
   if ($count > 0) {
       $result[$keyword] = $value;
   }
}

print_r($result);

您可以使用 $count 变量实际计算字符串出现的次数。