On 11/17/2015 7:07 AM, Mugeesh Husain wrote:
> when i fire query fq=date:[NOW/DAY-7DAYS TO NOW/DAY+1DAY], it is giving
> below error
>
> "fq":"initial_release_date:[NOW/DAY-7DAYS TO NOW/DAY 1DAY]",
> "rows":"32"}},
> "error":{
> "msg":"org.apache.solr.search.SyntaxError: Cannot parse
> 'initial_release_date:[NOW/DAY-7DAYS TO NOW/DAY 1DAY]': Encountered \"
> <RANGE_GOOP> \"1DAY \"\" at line 1, column 47.\nWas expecting one of:\n
> \"]\" ...\n \"}\" ...\n ",
> "code":400}}
How are you sending the queries to Solr? I'm betting that you are
constructing a URL manually in your own code and sending it to Solr
directly via HTTP, rather than using a Solr library like SolrJ,
Solarium, Sunspot, etc.
https://wiki.apache.org/solr/IntegratingSolr
What's happening here is that the query is not URL escaped, which means
the URL contains an actual plus sign for the "NOW/DAY+1DAY" part of your
query. A plus sign in a URL is interpreted by the webserver as a space,
in accordance with the standards that govern HTTP. This is not a bug.
A Solr library would have taken care of all the URL escaping for
characters that require it, and the user code is typically a lot easier
to write than URL construction code.
Whatever programming language you are using to construct your queries
very likely has a function for doing URL escaping on the parameters for
your URL. You would want to be careful to only run it on the values in
your parameters, not the entire URL, or it would escape everything and
the URL would not work.
The URL escaped version of a plus sign is %2B if you want to quickly fix
this before you look into URL escaping functions or a Solr library.
Thanks,
Shawn
