Subject: | Legacy fulltext index does not work with utf8 string values |
Dear Mark,
I'm not sure is it a bug or maybe some kind of my fault. I have a bunch of nodes with text properties payload, say
{code}
(n:MyNode {name_ua: 'some utf8 Cyrillic text in Ukrainian'})
{code}
I want to do fulltext search on name_ua, so:
- I created the fulltext name_ua_index,
- run script which performs query
{code}
MATCH (n:MyNode) RETURN n
{code}
- script loops fetching node array refs $node_aref one by one and does
{code}
unless( defined($node_aref->[0]) ) {
last MYNODE;
};
$name_ua_index->add_entry( $node_aref->[0],
'name_ua' => $node_aref->[0]->get_property( 'name_ua' ) ) or
die "Failed to add " . $node_aref->[0]->get_property( 'name_ua' ) . " into index, stopped";
{code}
Ok so far, but given I am 100% sure I *do* have an indexed string (n:MyNode {name_ua: 'Вітаю зі святом'}), I can not match it with START n = node:name_ua_index("name_ua:віта*") whatever I try.
What I am doing wrong?
Thanks in advance!
WBR,
Andrii