Consider something other than WhitespaceTokenizer. In this case
the tokenizer would split on the period and it’d work. I don’t know
whether that would fit the rest of your problem space or not though.
But to answer your original question, no there’s no a-priori reason you
can’t have WordDelimiter(
Here are links to images for the Analysis tab.
https://pasteboard.co/JfFTYu6.png
https://pasteboard.co/JfFUYXf.png
On Wed, Jul 1, 2020 at 3:03 PM gnandre wrote:
> I am doing that already but it does not help.
>
> Here is the complete analyzer chain.
>
> "100"> "solr.WhitespaceTokenizerFacto
I am doing that already but it does not help.
Here is the complete analyzer chain.
[image: image.png]
[image: image.png]
On Wed, Jul 1, 2020 at 12:29 PM Erick Erickson
wrote:
> Why not just specify preserveOriginal and follow by a lowerCaseFilter and
> use one wordDelimit
Why not just specify preserveOriginal and follow by a lowerCaseFilter and
use one wordDelimiterFilterFactory?
Best,
Erick
> On Jul 1, 2020, at 11:05 AM, gnandre wrote:
>
> Hi,
>
> To satisfy one use-case, I need to apply WordDelimiterFilter with
> splitOnCaseChange
> with 0 once and then with
Hi,
To satisfy one use-case, I need to apply WordDelimiterFilter with
splitOnCaseChange
with 0 once and then with 1 again. Are there some downsides to this
approach?
Use-case is to be able to match results when indexed content is my.camelCase
and search query is camelcase.
