WordDelimiterFilter stemEnglishPossessive option does not work for Tokens ending with a digit [LUCENE-5203]

For Token ending with "{digit}'s" the WordDelimiterFilter ignores the stemEnglishPossessive configuration

e.g. in the text "SD500's poor underwater performance" the "'s" will not be stemmed from SD500 even if stemEnglishPossessive="1".

The reason is that "WordDelimiterIterator#endsWithPossessive(int pos)" checks for

isAlpha(charType(text[pos - 3]))

while it should also allow for isDigits(int type).

---
Migrated from [LUCENE-5203](https://issues.apache.org/jira/browse/LUCENE-5203) by Rupert Westenthaler, 1 vote

Start at WordDelimiterIterator#endsWithPossessive(int pos) and inspect how the character type before "'s" is checked. Confirm that tokens ending in a digit followed by "'s" are handled when stemEnglishPossessive is enabled, and verify the existing WordDelimiterFilter behavior for the reported SD500's example.

Written by the indexing model from the issue text.