I've tried both, and when I don't separately index the contents table, I don't 
get any hits on a fairly unique string that I know is in the contents.plain 
field when I search on titles. The indexing process makes it clear that this 
relation is not being touched -- when I index the contents table, the result is 
this:

collected 1099 docs, 913.5 MB

When I index the titles (including the related field in contents) I only get 
this:

collected 1603 docs, 24.6 MB

contents.plain is the only large part of contents, it's just title_id and 
timestamps besides the plain column. If contents.plain was being accessed as a 
related column (and thus added to the metadata from the titles table), I would 
expect the result from the title indexing process to be in the 1GB 
neighborhood. That would make sense to me, given the amount of data being 
indexed.

What I have done as a fallback for now is added an index to the contents table, 
and rigged up my search results page to display those results as if they were 
to the title. But I'd still like to be searching once across all titles, and 
finding hits whether they were against the metadata (in the titles table) or 
the content (in the contents table). Is there some join syntax I need to use 
here? Remember -- this is Sphinx 2, not 3.

Thanks so much for your excellent support!

Walter

On Mar 21, 2014, at 11:04 PM, Pat Allan wrote:

> * The indexes line you're using will work fine - also, it'll work with the 
> arguments being shifted to method calls:
> 
>  indexes contents.plain, :as => :content
> 
> If there's still issues, do get in touch.

-- 
You received this message because you are subscribed to the Google Groups 
"Thinking Sphinx" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/thinking-sphinx.
For more options, visit https://groups.google.com/d/optout.

Reply via email to