Greetings!
Mikhail Khludnev, in his post to the thread "Google didn't help on this
one!", has pointed out one bug in Solr-5.3.0, and I was able to uncover
another one (which I wrote about in the same thread). Therefore, and
thankfully, I've been able to get past my configuration issues.
So now I've been able to try spell-checking on my local configuration
for the first time. My query string was "Anothr text containig
missspelled wordz." Of these 5 words, the only correct one is "text";
the others are not spelled correctly. Yet my query results gave me only
two suggestions ("test" and "Test") and they were for the one word that
*is* spelled correctly! This is the polar opposite of what I expected.
I understand why, though. Because I am using
solr.IndexBasedSpellchecker, so my data's index used as the dictionary.
This is not entirely a bad thing, because we use a lot of technical
terms and industry accepted spellings (like "gage" instead of "gauge").
But for the most part, I want to use a Webster-like dictionary against
which to check my spelling. Does this mean I need to find an English
dictionary file on the web and add Solr's FileBasedSpellChecker?? Or
does Solr already have what I need and it's a matter of me learning how
to configure that properly?? (If so, how?)
Mark
- Good ol' Websters Mark Fenbers
-
