Closed (fixed)
Project:
Apache Solr Search
Version:
6.x-1.x-dev
Component:
Code
Priority:
Normal
Category:
Bug report
Assigned:
Unassigned
Reporter:
Created:
7 Feb 2009 at 16:46 UTC
Updated:
3 Mar 2009 at 23:20 UTC
Jump to comment: Most recent file
Comments
Comment #1
damien tournoud commentedComment #2
pwolanin commentedI don't think we want this definition as-is. However, it would probably make sense to at least lowercase string fields?
Comment #3
damien tournoud commentedAdded LowerCase and MappingCharFilterFactory to those fields to reduce character forms.
Comment #4
damien tournoud commentedThe same patch, but without mapping on order fields.
Comment #5
pwolanin commentedLet's combine proposed schema changes with those from: http://drupal.org/node/370436
Comment #6
pwolanin commentedI'm not sure about changing the dynamic fields this way - need to look at how this affects faceting on those fields.
Comment #7
pwolanin commentedSo, as I feared, if you specify the lower case factory, the possible facet values are lower-cased in the response. I don't think that's what we want for names or taxonomy terms. For example:
The result in the a doc has:
but the facet values look like:
Comment #8
pwolanin commentedHere's another go at it after much beating around.
Comment #9
pwolanin commentedplay around with wildcard copyField directives in discussion w/ Damien.
Also, is there any reason at this point we need to use, e.g
ssfield*rather thanss_*or*_ss?Comment #10
pwolanin commentedComment #11
pwolanin commentedmissed some changes.
Comment #12
pwolanin commenteddiscussed w/ Jacob - we probably don't want the punctuation stripping out of the box. That's more of a site-specific requirement.
Comment #13
pwolanin commentedre-titleing the issue since we are also shortening the field names
Comment #14
pwolanin commentedtaking out the non-analyzed fields - string fields are not analyzed in any case, so I don't think this really matters.
Comment #15
damien tournoud commentedThe patch looks good at first sight, but I haven't actually tested it.
Comment #16
pwolanin commentedminor comment cleanup. Committing this.
Comment #17
pwolanin commentedone bug in the committed schema
Comment #18
pwolanin commentedcommitted
Comment #19
pwolanin commentedper suggestion by Damien and agreement from Jacob, we should go for consistency with sort_name, sort_title, sort_ss_*
Comment #20
pwolanin commentedpatch - also changes the irnored filed to mult valued (needs testing) since had reports of an error from Solr when trying to post an unknown multi-valued field.
Comment #21
pwolanin commentedComment #22
pwolanin commentedmy testing does show that making the ignored field multi-values is better at preventing errors.
see also: https://issues.apache.org/jira/browse/SOLR-1022
Comment #23
pwolanin commentedoops - need to also change the copyField directives.
Comment #24
pwolanin commentedcommitted to 6.x