Ben, just to clarify: you're searching for 'similar' paragraphs to the  
one in question? If so, then what Mickaël said is correct: Sphinx  
isn't going to be much help.

Although... if you could trim down the paragraph text to keywords -  
remove the more common words (like 'the'), then use the remainder in a  
search something like:

   Paragraph.search "ninja | pirate | deadly", :match_mode => :extended

I *think* this will give priority to matches that have more than one  
of those words, but I'm not sure. Something to think about, anyway.

Cheers

-- 
Pat

On 26/03/2009, at 11:47 PM, Ben Scofield wrote:

>
> That's what I figured (and was afraid of). Thanks for verifying it!
>
> Ben
>
> On Mar 25, 9:22 pm, Mickaël <[email protected]> wrote:
>> Okey, I understand.
>>
>> We can imagine a database like it : Texts <=> Paragraphs
>>
>> Text 1 has_many Paragraphs
>>
>> And we ask for "In the Text 1, paragraph #12, i want the number of
>> 'turtle'"
>> And the result will be 4.
>>
>> We could retrieves the Paragraphs content with find_by_id(12)
>> And parse the result to check if the word is present one or more time
>> in the String.
>>
>> I'm not sure Thinking Sphinx is design for the purpose which want.
>>
>> Mickaël.
>>
>> On 26 mar, 02:01, Ben Scofield <[email protected]> wrote:
>>
>>
>>
>>> Thanks, Mickaël - but I'm afraid this isn't quite what I need. My
>>> model has a field with long text passages, and faceting on that  
>>> field
>>> doesn't help. I'm looking for the individual terms out of that field
>>> that would return a given instance.
>>
>>> Maybe an example would help. Say I've got a Paragraph model with a
>>> content field, and I load my database up with the contents of a  
>>> book,
>>> split into individual paragraphs. What I want is to pick out a
>>> paragraph (#32, for instance), and see which terms from the
>>> paragraph's content would return it — so "the" wouldn't be very
>>> useful, but "ninja" might appear in only five of the paragraphs, and
>>> I'd want to see those five.
>>
>>> Does that make more sense?
>>
>>> Thanks!
>>> Ben
>>
>>> On Mar 24, 6:23 pm, Mickaël <[email protected]> wrote:
>>
>>>> Hi Ben,
>>
>>>> You want to sort the result into categories like : "Results are :
>>>> [words] [number_words] in Model docs" ?
>>
>>>> It's like faceting isn't it ?
>>
>>>> @facet_model = Model.facet("query")
>>>> And after, in @facet_model you will see each indexes of you model
>>>> sorted by attribute.
>>
>>>> I don't know if this method is right for you.
>>
>>>> Mickaël.
>>
>>>> On Mar 24, 5:45 pm, Ben Scofield <[email protected]> wrote:
>>
>>>>> Hi all,
>>
>>>>> Does anyone know of a way to ask Sphinx for the terms that a given
>>>>> document in a set scores most highly on? I.e., I've got 1000  
>>>>> docs, and
>>>>> I want to see the terms that would return #472 above the others...
>>
>>>>> Thanks!
>>>>> Ben Scofield
> >


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"Thinking Sphinx" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/thinking-sphinx?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to