Is this the wrong forum for this question - or does no-one else have a clue
about this either?
At 09:43 AM 1/12/06, Frank Bax wrote:
As I understand the docs for preg_replace(), I can enclose an PCRE
expression in parenthesis and use a backreference in the replace string;
but it's not working!
Data coming from another system contains a lot of data in one text record
which I must parse. Individual elements in the record are separated by
$. I want to "fix" elements that contain
dollar-alpha-space-space-digits-dollar
and replace the second space (between word and number) with a plus
character to end up with
dollar-alpha-space-plus-digits-dollar
An example of input string:
$prop = '$Fencing 11$Lumber 17$Weight: 317 Stones$Energy Resist
2%$';
Notice that Fencing and Lumber have two spaces. I want to end up with:
$prop = '$Fencing +11$Lumber +17$Weight: 317 Stones$Energy Resist
2%$';
Notice that original and final strings are the same length.
preg_replace( ' (\d+\$)', '+$1', $prop ); /* results in
dollar-alpha-space-space-plus, why have digits-dollar have disappeared? */
$Fencing +Lumber +Weight: 317 Stones$Energy Resist 2%$
Although the docs say that $0 should backreference the whole pattern,
instead it seems to match the pattern in parenthesis, but this code :
preg_replace( ' (\d+\$)', '+$0', $prop ); /* results in
dollar-alpha-space-space-plus-digits-dollar */
$Fencing +11$Lumber +17$Weight: 317 Stones$Energy Resist 2%$
Still contains the two blanks!!
In neither of my test cases is the result string the same length as original.
I'm running PHP 4.4.0 on OpenBSD 3.7
--
PHP General Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php
