Question

I have a task to accomplish, and I’ve been trying with regex for hours to no avail. A little help and some teaching would be appreciated. The .ini file format is:

ES=Spanish
EN=English
TOKEN=Token
FILES=Files

The goal of my routine is to

1. Get a key that needs to be changed,
2. Figure out what the current value of that key is,
3. add some comments about the old value and who changed it,
4. change the value.

So, for example, if I want to change the value belonging to EN from English to Eigo, I would end up with:

ES=Spanish
#Changed by Jane Doe on <date>
#Old value: EN=English
EN=Eigo
TOKEN=Token
FILES=Files

My code:

        $content = "EN=English\n"
            . "ES=Spanish\n"
            . "TOKEN=Token\n"
            . "FILES=Files\n";

            $key = 'EN';
            $newValue = 'Eigo";
            $editor = 'Jane Doe';

             //get the old value
    $matches = array();
    preg_match('~'.$key.'\s?=[^\n$]+~iu',$content,$matches);

    $commentLines = '# Change by ' . $editor . ' on ' . date("m/d/Y g:i a") . "\n";
    $commentLines .= '# Old value: ' . $matches[0] . "\n";
    $newValueLine = $key.'='.$newValue. "\n";
    $newEntry = $commentLines . $newValueLine;
    $content = preg_replace('~'.$key.'\s?=[^\n$]+~iu',$newEntry,$content);

This was working fine, until I realized that if someone changed the key for a short string, like EN, then my regex matches the EN in TOKEN, and also changes that, but messes up the whole file:

TOK#Changed by ....

So, a few questions:

1. Because each KEY should be unique, should I even be using regex for this? Is there a faster/cleaner/better method I should be using?

2. Why when I add a ^ to the regex on the line with preg_replace doesn’t this match the beginning of the line and get rid of my TOKEN matching problem?

   preg_replace('~^'.$key.'\s?=[^\n$]+~iu',$newEntry,$content)
   

Answer 1

The information you give is a bit vague. Is this being stored in a file? If so, it seems like what you really want is version control software like subversion. It will keep track of what the state of the code/ini files were like at any point in time, who made a change, and if people want to put in a message about what they were doing, it will handle that as well.

Your regex seems a bit overcomplicated. And for multiline search and replace, you need to use the m modifier.

Try:

$key = 'EN';
preg_replace('~^'.$key.'=.*$~m', $newEntry, $content);

Tested at http://www.spaweditor.com/scripts/regex/index.php with:

Regex: 
    /^EN=.*$/m

Data:
    EN=English
    ES=Spanish
    TOKEN=Token
    FILES=Files

Replace: 
    FOO=bar

Function:
    preg_replace

Result:
    FOO=bar
    ES=Spanish
    TOKEN=Token
    FILES=Files