Unexpected behavior of preg_replace() with regular expression containing \h on à

Asked Jun 12 '19 at 20:34

Active Jun 12 '19 at 20:34

Viewed 25 times

I am seeing the following unexpected behavior in PHPs preg_replace():

Running

echo("'".preg_replace('#\h\h+#', ' ', 'à')."'\n");
echo("'".preg_replace('#\h\h+#', ' ', 'à ')."'\n");
echo("'".preg_replace('#\h\h+#', ' ', 'á')."'\n");
echo("'".preg_replace('#\h\h+#', ' ', 'á ')."'\n");
echo("'".preg_replace('#\h+#', ' ', 'à ')."'\n");
echo("'".preg_replace('#\h#', ' ', 'à ')."'\n");

yields the output

'à'
'Ã '
'á'
'á '
'Ã '
'Ã  '

Why is the 'à' in some cases replaced by 'Ã '?

Shouldn't my regular expression only replace certain occurrence of horizontal white space characters with a single space and leave all other characters alone?

asked Jun 12 '19 at 20:34

cgogolin

Different encoding in editor/IDE and server/web? – pavel Jun 12 '19 at 20:36
1

Use it with `u` modifier, e.g. `'#\h+#u'`. – Wiktor Stribiżew Jun 12 '19 at 20:50
That does the trick! I somehow implicitly assumed that at least newer versions of php should by default work with unicode, but that is apparently not the case. Thanks! – cgogolin Jun 15 '19 at 10:06

Unexpected behavior of preg_replace() with regular expression containing \h on à

0 Answers0