用新行替换 \n

Question

我有一个由文本组成的字符串，可以包含多个\和n。

在 php 中，使用 preg_replace，我只想在需要的地方\n替换[NewLine]。说替换\n我的意思是替换字符反斜杠-n（\和n）。

我浏览了类似的问题，但没有找到一个来处理\\n不应该用新行替换的情况。

例如，字符串hello\nworld应该是hello[NewLine]world，

并且字符串hello\\nworld应该是hello\nworld.

意义 ：

<?php

$source1 = <<<'EOD'

\n

EOD;

$source2 = <<<'EOD'

\\n

EOD;

$source3 = <<<'EOD'

\\\n

EOD;

//$source4 with 4 backslaskes , 5 with 5 etc.

//preg_replace `$source1` so it ends up as `[NewLine]`

//preg_replace `$source2` to `\\n`

//preg_replace `$source3` to `\\[NewLine]`

//etc

//....

//even number of consecutive \ followed by n means no [NewLine]

//odd number of consecutive \ followed by n means new [NewLine]

我有一个错误的替换代码

$a = preg_replace('{([^{\\\}])\\\n}', "$1\n", $string); 

//work as expected when $string is "first-line\\nsecond-line";

//doesn't work as expected when $string is "first-line\\\nsecond-line";

任何人都知道在正则表达式中需要做什么才能使它正常工作？

Answer 1

您可以使用回溯控制动词&nbsp;跳过部分字符串(*SKIP)：$result&nbsp;=&nbsp;preg_replace('~\\\\.(*SKIP)(?<=n)~',&nbsp;"\n",&nbsp;$source);演示(*SKIP)当它之后的子模式失败时，跳过它之前的子模式已经到达的字符串的位置。(?<=n)是一个回顾断言，意思是前面有 n。当它失败时，正则表达式引擎跳转到转义字符位置（由 定义的位置(*SKIP)）之后进行新的尝试，而不是测试下一个位置（反斜杠之后的位置）。您也可以使用以下方法执行相同的操作strtr：$result&nbsp;=&nbsp;strtr($source,&nbsp;['\\\\'&nbsp;=>&nbsp;'\\\\',&nbsp;'\\n'&nbsp;=>&nbsp;"\n"]);（由于为两个连续的反斜杠定义了替换，因此第二个反斜杠不能用于最终的后续n.）

Answer 2

您需要匹配成对的反斜杠\\或成对的\n. 通过指定回调，您可以检查是否需要替换它，如果不需要，只需逐字返回匹配的字符：<?php$inputs = [&nbsp; &nbsp; "\n is [NewLine]",&nbsp; &nbsp; "\\n is \n",&nbsp; &nbsp; "\\\n is \[NewLine]",&nbsp; &nbsp; "\n and \\n and \\\n and \\\\n"];$nl = function($matches) {&nbsp; &nbsp; return $matches[0] == "\n" ? "[NewLine]" : $matches[0];};print_r(preg_replace_callback("/\\\\|\\n/", $nl, $inputs));通过将\\first 放在正则表达式交替中，它将首先被消耗，并且只有当反斜杠的数量为奇数时，才会\n匹配。（确保转义正则表达式模式中的反斜杠！）最后一个示例输入的输出：[NewLine] and \n and \[NewLine] and \\n如果您的输入包含文字反斜杠字符和n(0x5c 0x6e) 而不是换行符 (0x0a)——那么您的输入是真的\\n而不是\n——您需要添加一个额外的转义级别。正则表达式变为/\\\\\\\\|\\\\n/回调必须检查"\\n"

Answer 3

找到了一个更简单的解决方案。就像 knittl 所说的那样，对匹配\是关键，而 regxp 并不是真正需要的。strtr($input, ['\\\\' => '\\','\\n' => "[NL]"]);在https://3v4l.org/5tebZ测试$um = [&nbsp; &nbsp; &nbsp; &nbsp; '\\\\' => '\\',&nbsp; &nbsp; &nbsp; &nbsp; '\\n' => "[NL]",&nbsp; &nbsp; ];$inputs = [];for ($i=1;$i<=14;$i++) {&nbsp; &nbsp;$inputs[] = ''.str_repeat("\\", $i)."n";}foreach ($inputs as $input) {&nbsp; &nbsp; $results[] = strtr($input, $um);}foreach ($inputs as $key=>$value){&nbsp; &nbsp; echo str_pad($value, 15, ' ',STR_PAD_LEFT)." =======> ".$results[$key]."\n";}输出&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;\n =======> [NL]&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; \\n =======> \n&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;\\\n =======> \[NL]&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; \\\\n =======> \\n&nbsp; &nbsp; &nbsp; &nbsp; &nbsp;\\\\\n =======> \\[NL]&nbsp; &nbsp; &nbsp; &nbsp; \\\\\\n =======> \\\n&nbsp; &nbsp; &nbsp; &nbsp;\\\\\\\n =======> \\\[NL]&nbsp; &nbsp; &nbsp; \\\\\\\\n =======> \\\\n&nbsp; &nbsp; &nbsp;\\\\\\\\\n =======> \\\\[NL]&nbsp; &nbsp; \\\\\\\\\\n =======> \\\\\n&nbsp; &nbsp;\\\\\\\\\\\n =======> \\\\\[NL]&nbsp; \\\\\\\\\\\\n =======> \\\\\\n&nbsp;\\\\\\\\\\\\\n =======> \\\\\\[NL]\\\\\\\\\\\\\\n =======> \\\\\\\n