Solr is looking in its index for tokens that start with "xyz". In order to have a match you need to have a token like that in the index. Which brings us to the indexing part.
When you put "xyz" in a wiki page, the tokens that get added to the index are not exactly the words from the page content. It depends very much:
on the field (page content, title, name, object property, etc.)
the page locale
the indexing type (the same field is often indexed in multiple ways)
For English locale a word like "cars" leads to a "car" token, but that may be different for other locales. So when you try to search for cars* you won't get any results most probably because the token that was indexed is car. Wild card search should be used on fields that are indexed as strings (i.e. the field value produces a single token). Using it on fields where the root of each word is extracted leads to unexpected results, like in this case.
This message was sent by Atlassian Jira (v9.3.0#930000-sha1:287aeb6)
If image attachments aren't displayed, see this article.