Thank you Steven,
I'll look into this
On Fri, Apr 20, 2012 at 12:43 AM, Steven A Rowe wrote:
> Hi Vladimir,
>
>> The most uncomfortable in new behaviour to me is that in past I used
>> to search by subdomain like bbb.com: and have displayed results
>> with www.bbb.com:, aaa.bbb.com:
On Thu, Apr 19, 2012 at 4:51 PM, Vladimir Gubarkov wrote:
> So it's now imposible to find this document with query: "site.com".
> I'm having an RSS subscription for that search, and now it's broken.
>
Just to point out, its not impossible, as i suggested before, if you
were happy with the old tok
On Thu, Apr 19, 2012 at 4:51 PM, Vladimir Gubarkov wrote:
> Hmmm... I know this and I reindexed!
> I'll try to explain the problem (fortunately, already solved by using
> LUCENE_30) ones again:
> When indexing with new analyzer the whole lexeme "some.cool.site.com"
> goes to index, not 4 lexems "
Thank you Robert for detailed reply
On Fri, Apr 20, 2012 at 12:37 AM, Robert Muir wrote:
> On Thu, Apr 19, 2012 at 7:26 AM, Vladimir Gubarkov wrote:
>> New analyzer:
>> [aaa.bbb.com, , a, b, c, d'e, f, g, h, i, j, k, l_m, n, o, p, q,
>> r, s, t, u, v, z, y, z]
>> Old analyzer:
>> [aaa, bbb,
Hi Vladimir,
> The most uncomfortable in new behaviour to me is that in past I used
> to search by subdomain like bbb.com: and have displayed results
> with www.bbb.com:, aaa.bbb.com: and so on. Now I have 0
> results.
About domain names, see my response to a similar question today on
On Thu, Apr 19, 2012 at 7:26 AM, Vladimir Gubarkov wrote:
> New analyzer:
> [aaa.bbb.com, , a, b, c, d'e, f, g, h, i, j, k, l_m, n, o, p, q,
> r, s, t, u, v, z, y, z]
> Old analyzer:
> [aaa, bbb, com, , a, b, c, d, e, f, g, h, i, j, k, l, m, n, o, p,
> q, r, s, t, u, v, z, y, z]
>
> Please
On Thu, Apr 19, 2012 at 7:57 PM, Uwe Schindler wrote:
>> My questions are: 1) it this change is by design (not a mistake) and
>> 2) is the only option to achieve old behaviour is to use
>> Version.LUCENE_30 for creating analyzer?
>
> This is why this option is there!
Right and it's great, but thi
> My questions are: 1) it this change is by design (not a mistake) and
> 2) is the only option to achieve old behaviour is to use
> Version.LUCENE_30 for creating analyzer?
This is why this option is there!
-
To unsubscribe, e-m
