Edit report at https://bugs.php.net/bug.php?id=60380&edit=1
ID: 60380
Comment by: ni...@php.net
Reported by: evrinoma at gmail dot com
Summary: preg_split and old version split
Status: Open
Type: Bug
Package: *Programming Data Structures
Operating System: Linux CenOS5 CentOS6
PHP Version: 5.3.8
Block user comment: N
Private report: N
New Comment:
You are dealing with two levels of unescaping here. The first is the usual for
string, the second is additionally done by PCRE.
E.g. if you write "\\\\" the actually resulting string will just be "\\"
(because \\ is unescaped to \). This is then passed to PCRE, which itself again
does unescaping, so the \\ is turned into \. That's why you need four
backslashes to produce one literal backslash in a regular expression.
In your particular case three work too, but only because the following
character does not form an escape sequence.
Previous Comments:
------------------------------------------------------------------------
[2012-08-30 19:24:38] cjbf at cam dot ac dot uk
Three \s are needed to match one \. Could that at least be documented if it's
not wrong?
$text='AA\\nB and B\r';// contains \n and \r
print_r(preg_split('/\\\n|\\\r/', $text));
gives
Array
(
[0] => AA
[1] => B and B
[2] =>
)
------------------------------------------------------------------------
[2012-08-30 19:02:43] cjbf at cam dot ac dot uk
Put simply, preg_split treats "\\" wrongly. It ought to search for an actual
"\" but never matches it.
------------------------------------------------------------------------
[2011-11-25 07:15:00] evrinoma at gmail dot com
Description:
------------
--splite-------
array[0]=0123456789--
array[1]=qwertyuiop--
array[2]=9876543210--
array[3]=poiuytrewq--
array[4]=0123456789--
array[5]=asdfghjkl--
--preg_splite--
array[0]=0123456789\qwertyuiop-- <<<<<<<<< ???
array[1]=9876543210--
array[2]=poiuytrewq--
array[3]=0123456789--
array[4]=asdfghjkl--
array[5]=--
---------------
Test script:
---------------
<HTML>
<HEAD>
<TITLE> BUG </TITLE>
</HEAD>
<BODY>
<?
printf(" --splite------- <BR>\n");
$date="0123456789\\qwertyuiop 9876543210.poiuytrewq:0123456789 asdfghjkl";
$var = split( "[-\\. :]",$date );
printf(" array[0]=%s-- <BR>\n",$var[0]);
printf(" array[1]=%s-- <BR>\n",$var[1]);
printf(" array[2]=%s-- <BR>\n",$var[2]);
printf(" array[3]=%s-- <BR>\n",$var[3]);
printf(" array[4]=%s-- <BR>\n",$var[4]);
printf(" array[5]=%s-- <BR>\n",$var[5]);
printf(" --preg_splite-- <BR>\n");
$var = preg_split( "#[-\\. :]#",$date);
printf(" array[0]=%s-- <BR>\n",$var[0]);
printf(" array[1]=%s-- <BR>\n",$var[1]);
printf(" array[2]=%s-- <BR>\n",$var[2]);
printf(" array[3]=%s-- <BR>\n",$var[3]);
printf(" array[4]=%s-- <BR>\n",$var[4]);
printf(" array[5]=%s-- <BR>\n",$var[5]);
printf(" --------------- <BR>\n");
?>
</BODY>
</HTML>
------------------------------------------------------------------------
--
Edit this bug report at https://bugs.php.net/bug.php?id=60380&edit=1
