Github user LucVL commented on the pull request:
https://github.com/apache/lucene-solr/pull/210#issuecomment-158335822
Thereâs good documentation on combining git pull-requests with jira
issues on the apache wiki (thereâs a bot for that)
Github user elyograg commented on the pull request:
https://github.com/apache/lucene-solr/pull/210#issuecomment-158147256
I've created the issue in Jira. Let's continue the discussion there. When
the patch is ready, I will commit it with the keywords that will close this PR.
Github user elyograg commented on the pull request:
https://github.com/apache/lucene-solr/pull/210#issuecomment-158099893
I think you've got the default values backwards, but other than that it
looks good to me. For 5.x, we need the default to be false so existing setups
do not
Github user elyograg commented on the pull request:
https://github.com/apache/lucene-solr/pull/210#issuecomment-158115946
Now that I've gotten a closer look, I see that you went with
splitKeywordTokens rather than skipKeywordTokens, so your defaults were correct.
---
If your project
Github user elyograg commented on the pull request:
https://github.com/apache/lucene-solr/pull/210#issuecomment-158117696
What name would you like to have in the CHANGES.txt giving you credit for
the change?
I believe the only thing that's missing from the patch now is
Github user smartprix commented on the pull request:
https://github.com/apache/lucene-solr/pull/210#issuecomment-158072439
The behavior should is now configurable. I have updated the pull request to
reflect that. A new attribute "splitKeywordTokens" which is false by default
for
GitHub user smartprix opened a pull request:
https://github.com/apache/lucene-solr/pull/210
WordDelimiterFilter - Don't split words marked as keyword
Currently WordDelimiterFilter also splits keywords into tokens. eg. if
128GB is maked as a keyword using KeywordMarkerFilter,
Github user elyograg commented on the pull request:
https://github.com/apache/lucene-solr/pull/210#issuecomment-157833770
I like this idea. I think I'd go one step further -- make this behavior
configurable with an attribute named something like skipKeywordTokens. It
should default