preg_match_all包含周围数据

I have a large amount of log data that I need to get some relevant information out of. The syntax is:

WORD1 some text KEYWORD some text WORD2 WORD1 some text KEYWORD some text WORD2 WORD1 some text KEYWORD some text WORD2 WORD1 some text KEYWORD some text WORD2

Would it be possible to use regex to get a certain block out of it - so that when I queried some keyword it would return WORD1 some text THIS_KEYWORD some text WORD2

PS: There could be multiple instances of one keyword, it should return all of the blocks (preg_match_all, PREG_SET_ORDER?).

As per you comment there will always be starting and finishing word for each block.

You can do the following:

  • Split the input string into blocks (array). An end of the block is identified by the whitespace which is followed by WORD2 and preceded by WORD1.
  • Search the array we got above for the keyword.

.

$arr = preg_split('/(?<=WORD2)\s+(?=WORD1)/',$log);
$keyword = 'KEYWORD1';
$result = preg_grep("/\b$keyword\b/",$arr);

Ideone Link

use UNGREEDY modifier

$found = preg_match_all("/WORD1.*KEYWORD.*WORD2/U",$x,$match);
if($found) print_r($match[0]);