You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@oodt.apache.org by "Brian Foster (Created) (JIRA)" <ji...@apache.org> on 2011/11/30 08:41:52 UTC
[jira] [Created] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
QueryTool treats Lucene list of term queries as ANDed instead of ORed
---------------------------------------------------------------------
Key: OODT-355
URL: https://issues.apache.org/jira/browse/OODT-355
Project: OODT
Issue Type: Bug
Components: file manager
Affects Versions: 0.3
Environment: none
Reporter: Brian Foster
Assignee: Brian Foster
Priority: Minor
Fix For: 0.4
For example:
{code}
NominalDate:2011-01-20 Group:test
{code}
is treated as:
{code}
NominalDate=2011-01-20 AND Group=test
{code}
when really it should be:
{code}
NominalDate=2011-01-20 OR Group=test
{code}
>From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
Boolean Operators
Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
"jakarta apache" jakarta
or
"jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
Posted by "Brian Foster (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OODT-355?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13161330#comment-13161330 ]
Brian Foster commented on OODT-355:
-----------------------------------
my current patch for OODT-351 adds @Deprecated to the top of QueryTool class... LuceneQuery action is pretty much the same code as QueryTool, just with a few changes... if we add a new flag to QueryTool then we are in theory saying we are supporting both and we really only want one... i think it's best to just treat it as a legacy method... keep it around... but don't support it... we have documented the know "issue" (which is not really an issue, just a divergence from the lucene query documentation)... basically if anyone wants the version which is now inline with the lucene query documentation then they can use LuceneQuery action instead of QueryTool... cameron, is that what you had in mind with deprecation?... chris, thoughts?
> QueryTool treats Lucene list of term queries as ANDed instead of ORed
> ---------------------------------------------------------------------
>
> Key: OODT-355
> URL: https://issues.apache.org/jira/browse/OODT-355
> Project: OODT
> Issue Type: Bug
> Components: file manager
> Affects Versions: 0.3
> Environment: none
> Reporter: Brian Foster
> Assignee: Brian Foster
> Priority: Minor
> Fix For: 0.4
>
>
> For example:
> {code}
> NominalDate:2011-01-20 Group:test
> {code}
> is treated as:
> {code}
> NominalDate=2011-01-20 AND Group=test
> {code}
> when really it should be:
> {code}
> NominalDate=2011-01-20 OR Group=test
> {code}
> From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
> Boolean Operators
> Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
> The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
> To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
> "jakarta apache" jakarta
> or
> "jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
Posted by "Cameron Goodale (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OODT-355?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13161144#comment-13161144 ]
Cameron Goodale commented on OODT-355:
--------------------------------------
Chris,
I see your point about the 5 years. My concern is this will create a situation where if you Query the FileManager using CAS-CLI you must do it one way, but if you are using the File Manager's included query_tool then you must do it a different way. I just see it as a fracture in functionality.
I would feel better about leaving the query_tool alone if it were deprecated, that way there would be ONE tool to query the FileManager. That way the old school people who have used the tool successfully for 5 years will not be effected, and we avoid confusion with new users moving forward.
> QueryTool treats Lucene list of term queries as ANDed instead of ORed
> ---------------------------------------------------------------------
>
> Key: OODT-355
> URL: https://issues.apache.org/jira/browse/OODT-355
> Project: OODT
> Issue Type: Bug
> Components: file manager
> Affects Versions: 0.3
> Environment: none
> Reporter: Brian Foster
> Assignee: Brian Foster
> Priority: Minor
> Fix For: 0.4
>
>
> For example:
> {code}
> NominalDate:2011-01-20 Group:test
> {code}
> is treated as:
> {code}
> NominalDate=2011-01-20 AND Group=test
> {code}
> when really it should be:
> {code}
> NominalDate=2011-01-20 OR Group=test
> {code}
> From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
> Boolean Operators
> Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
> The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
> To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
> "jakarta apache" jakarta
> or
> "jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
Posted by "Brian Foster (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OODT-355?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13161046#comment-13161046 ]
Brian Foster commented on OODT-355:
-----------------------------------
hey cameron,
i'm all for changing it, however my main concern is that this QueryTool was used in many scripts for deleting products for OCO, and SMAP inherited theses scripts (probably other projects too)... changing the AND behavior to OR behavior is extremely dangerous for delete scripts since more will be deleted than expected since queries will return more results then previously expected... after LuceneQuery action is added this will become the new "QueryTool" which should be used... QueryTool will continue to exist, but should no longer be used by new projects... existing projects should be encouraged to make the switch to LuceneQuery (or SqlQuery if they are using the sql functionality of it)... shouldn't be too hard to convenience them since LuceneQuery will support OR operator, where QueryTool does not (and SqlQuery has many more options as well)... anyway you and chris hash it out and let me know what you want me to do here
> QueryTool treats Lucene list of term queries as ANDed instead of ORed
> ---------------------------------------------------------------------
>
> Key: OODT-355
> URL: https://issues.apache.org/jira/browse/OODT-355
> Project: OODT
> Issue Type: Bug
> Components: file manager
> Affects Versions: 0.3
> Environment: none
> Reporter: Brian Foster
> Assignee: Brian Foster
> Priority: Minor
> Fix For: 0.4
>
>
> For example:
> {code}
> NominalDate:2011-01-20 Group:test
> {code}
> is treated as:
> {code}
> NominalDate=2011-01-20 AND Group=test
> {code}
> when really it should be:
> {code}
> NominalDate=2011-01-20 OR Group=test
> {code}
> From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
> Boolean Operators
> Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
> The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
> To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
> "jakarta apache" jakarta
> or
> "jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
Posted by "Brian Foster (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OODT-355?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13160267#comment-13160267 ]
Brian Foster commented on OODT-355:
-----------------------------------
In order to support backwards compatibility and not mess up current scripts which may currently wrap QueryTool, maybe QueryTool should be left as is, and i will make this change only in the new LuceneQuery action... what you guys think?
> QueryTool treats Lucene list of term queries as ANDed instead of ORed
> ---------------------------------------------------------------------
>
> Key: OODT-355
> URL: https://issues.apache.org/jira/browse/OODT-355
> Project: OODT
> Issue Type: Bug
> Components: file manager
> Affects Versions: 0.3
> Environment: none
> Reporter: Brian Foster
> Assignee: Brian Foster
> Priority: Minor
> Fix For: 0.4
>
>
> For example:
> {code}
> NominalDate:2011-01-20 Group:test
> {code}
> is treated as:
> {code}
> NominalDate=2011-01-20 AND Group=test
> {code}
> when really it should be:
> {code}
> NominalDate=2011-01-20 OR Group=test
> {code}
> From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
> Boolean Operators
> Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
> The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
> To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
> "jakarta apache" jakarta
> or
> "jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
Posted by "Cameron Goodale (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OODT-355?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13160885#comment-13160885 ]
Cameron Goodale commented on OODT-355:
--------------------------------------
Guys,
My take is that you are talking about a helper tool for the File Manager (query_tool) which has a bug in the way the lucene queries are being executed. I say fix the bug in query_tool and forget about 'backward compatibility' since we aren't changing some critical interface that will break core functionality.
So my vote is to fix the issue in BOTH places, QueryTool AND LuceneQuery action.
> QueryTool treats Lucene list of term queries as ANDed instead of ORed
> ---------------------------------------------------------------------
>
> Key: OODT-355
> URL: https://issues.apache.org/jira/browse/OODT-355
> Project: OODT
> Issue Type: Bug
> Components: file manager
> Affects Versions: 0.3
> Environment: none
> Reporter: Brian Foster
> Assignee: Brian Foster
> Priority: Minor
> Fix For: 0.4
>
>
> For example:
> {code}
> NominalDate:2011-01-20 Group:test
> {code}
> is treated as:
> {code}
> NominalDate=2011-01-20 AND Group=test
> {code}
> when really it should be:
> {code}
> NominalDate=2011-01-20 OR Group=test
> {code}
> From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
> Boolean Operators
> Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
> The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
> To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
> "jakarta apache" jakarta
> or
> "jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
Posted by "Chris A. Mattmann (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OODT-355?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13161399#comment-13161399 ]
Chris A. Mattmann commented on OODT-355:
----------------------------------------
Nah, you're fine Cam. +1 to Brian's last suggestion and his latest patch. Let's move it forward. I think having CLI and QueryTool with slightly diff functionality will allow us to see which one users prefer. Over time the one they don't care about will become less and less used and we can slowly phase them out.
Onward!
> QueryTool treats Lucene list of term queries as ANDed instead of ORed
> ---------------------------------------------------------------------
>
> Key: OODT-355
> URL: https://issues.apache.org/jira/browse/OODT-355
> Project: OODT
> Issue Type: Bug
> Components: file manager
> Affects Versions: 0.3
> Environment: none
> Reporter: Brian Foster
> Assignee: Brian Foster
> Priority: Minor
> Fix For: 0.4
>
>
> For example:
> {code}
> NominalDate:2011-01-20 Group:test
> {code}
> is treated as:
> {code}
> NominalDate=2011-01-20 AND Group=test
> {code}
> when really it should be:
> {code}
> NominalDate=2011-01-20 OR Group=test
> {code}
> From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
> Boolean Operators
> Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
> The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
> To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
> "jakarta apache" jakarta
> or
> "jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
Posted by "Chris A. Mattmann (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OODT-355?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13161197#comment-13161197 ]
Chris A. Mattmann commented on OODT-355:
----------------------------------------
Hey Cam:
{quote}
I see your point about the 5 years. My concern is this will create a situation where if you Query the FileManager using CAS-CLI you must do it one way, but if you are using the File Manager's included query_tool then you must do it a different way. I just see it as a fracture in functionality.
{quote}
Well, QueryTool existed first :-) And, there's nothing that says what Brian is doing in terms of adding QueryTool to the FM CLI isn't something completely different or new. Just depends on how you think of it. So, I wouldn't call it a fracture but maybe that's just me.
{quote}
I would feel better about leaving the query_tool alone if it were deprecated, that way there would be ONE tool to query the FileManager. That way the old school people who have used the tool successfully for 5 years will not be effected, and we avoid confusion with new users moving forward.
{quote}
Well this is a good question. Perhaps another way to solve this would be to make a configuration option in queryTool that says by default interpret spaces to AND but allow override to interpret to OR. That way, with a switch query tool can behave like the CLI. Does that make sense? Brian, thoughts?
> QueryTool treats Lucene list of term queries as ANDed instead of ORed
> ---------------------------------------------------------------------
>
> Key: OODT-355
> URL: https://issues.apache.org/jira/browse/OODT-355
> Project: OODT
> Issue Type: Bug
> Components: file manager
> Affects Versions: 0.3
> Environment: none
> Reporter: Brian Foster
> Assignee: Brian Foster
> Priority: Minor
> Fix For: 0.4
>
>
> For example:
> {code}
> NominalDate:2011-01-20 Group:test
> {code}
> is treated as:
> {code}
> NominalDate=2011-01-20 AND Group=test
> {code}
> when really it should be:
> {code}
> NominalDate=2011-01-20 OR Group=test
> {code}
> From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
> Boolean Operators
> Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
> The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
> To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
> "jakarta apache" jakarta
> or
> "jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
Posted by "Brian Foster (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OODT-355?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13159889#comment-13159889 ]
Brian Foster commented on OODT-355:
-----------------------------------
NOTE: this change would change query results for existing Lucene query which uses this syntax... while integrating CAS-CLI with CAS-Filemgr i created an action, LuceneQuery, which ports QueryTool's Lucene query to an action... in this process i was improved a few things with it, like make it support the OR operator... in fact QueryTool's current implementation will treat a specified Lucene OR as an AND: "NominalDate=2011-01-20 AND Group=test" and "NominalDate=2011-01-20 OR Group=test" will currently give the same query results
> QueryTool treats Lucene list of term queries as ANDed instead of ORed
> ---------------------------------------------------------------------
>
> Key: OODT-355
> URL: https://issues.apache.org/jira/browse/OODT-355
> Project: OODT
> Issue Type: Bug
> Components: file manager
> Affects Versions: 0.3
> Environment: none
> Reporter: Brian Foster
> Assignee: Brian Foster
> Priority: Minor
> Fix For: 0.4
>
>
> For example:
> {code}
> NominalDate:2011-01-20 Group:test
> {code}
> is treated as:
> {code}
> NominalDate=2011-01-20 AND Group=test
> {code}
> when really it should be:
> {code}
> NominalDate=2011-01-20 OR Group=test
> {code}
> From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
> Boolean Operators
> Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
> The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
> To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
> "jakarta apache" jakarta
> or
> "jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
Posted by "Cameron Goodale (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OODT-355?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13161337#comment-13161337 ]
Cameron Goodale commented on OODT-355:
--------------------------------------
Brian,
I am not sure what the deprecation policy or practice is for OODT. I have always been under the impression that CLI would be the New and Improved way to interact with File Manager and Workflow Manager and other components in the future. If instead CLI's purpose is to offer another way to interact then I agree with Chris, we should leave things as they are. Meaning. if CLI is an optional component that people can use if they feel like it, then I have no problems with different functionality. My thought was that one was meant to replace the other (hence deprecate the one being replaced, which in this case is QueryTool).
Chances are I have jumped into the CLI fray without really understanding CLI's end game. My apologies if that is the case.
> QueryTool treats Lucene list of term queries as ANDed instead of ORed
> ---------------------------------------------------------------------
>
> Key: OODT-355
> URL: https://issues.apache.org/jira/browse/OODT-355
> Project: OODT
> Issue Type: Bug
> Components: file manager
> Affects Versions: 0.3
> Environment: none
> Reporter: Brian Foster
> Assignee: Brian Foster
> Priority: Minor
> Fix For: 0.4
>
>
> For example:
> {code}
> NominalDate:2011-01-20 Group:test
> {code}
> is treated as:
> {code}
> NominalDate=2011-01-20 AND Group=test
> {code}
> when really it should be:
> {code}
> NominalDate=2011-01-20 OR Group=test
> {code}
> From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
> Boolean Operators
> Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
> The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
> To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
> "jakarta apache" jakarta
> or
> "jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
Posted by "Chris A. Mattmann (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OODT-355?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13160272#comment-13160272 ]
Chris A. Mattmann commented on OODT-355:
----------------------------------------
+1, I think that makes perfect sense.
> QueryTool treats Lucene list of term queries as ANDed instead of ORed
> ---------------------------------------------------------------------
>
> Key: OODT-355
> URL: https://issues.apache.org/jira/browse/OODT-355
> Project: OODT
> Issue Type: Bug
> Components: file manager
> Affects Versions: 0.3
> Environment: none
> Reporter: Brian Foster
> Assignee: Brian Foster
> Priority: Minor
> Fix For: 0.4
>
>
> For example:
> {code}
> NominalDate:2011-01-20 Group:test
> {code}
> is treated as:
> {code}
> NominalDate=2011-01-20 AND Group=test
> {code}
> when really it should be:
> {code}
> NominalDate=2011-01-20 OR Group=test
> {code}
> From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
> Boolean Operators
> Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
> The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
> To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
> "jakarta apache" jakarta
> or
> "jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
Posted by "Chris A. Mattmann (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OODT-355?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13161094#comment-13161094 ]
Chris A. Mattmann commented on OODT-355:
----------------------------------------
+1 to getting them to migrate to the new Action, but leaving query_tool as is.
Cam, thoughts?
> QueryTool treats Lucene list of term queries as ANDed instead of ORed
> ---------------------------------------------------------------------
>
> Key: OODT-355
> URL: https://issues.apache.org/jira/browse/OODT-355
> Project: OODT
> Issue Type: Bug
> Components: file manager
> Affects Versions: 0.3
> Environment: none
> Reporter: Brian Foster
> Assignee: Brian Foster
> Priority: Minor
> Fix For: 0.4
>
>
> For example:
> {code}
> NominalDate:2011-01-20 Group:test
> {code}
> is treated as:
> {code}
> NominalDate=2011-01-20 AND Group=test
> {code}
> when really it should be:
> {code}
> NominalDate=2011-01-20 OR Group=test
> {code}
> From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
> Boolean Operators
> Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
> The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
> To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
> "jakarta apache" jakarta
> or
> "jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
Posted by "Chris A. Mattmann (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OODT-355?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13160950#comment-13160950 ]
Chris A. Mattmann commented on OODT-355:
----------------------------------------
Hi Cam:
My read on it is that folks have been using query_tool since 2006 and Brian Foster is the first person in 5 years to submit some code to update this. 5 years is a long time of people writing programs on top of this tool and expecting it to behave (even incorrectly) in a particular way. Changing that way might be the right thing to do, but I would say that this tool *is* a critical interface at least for folks just getting a quick look into what's there in an FM catalog out of the box.
> QueryTool treats Lucene list of term queries as ANDed instead of ORed
> ---------------------------------------------------------------------
>
> Key: OODT-355
> URL: https://issues.apache.org/jira/browse/OODT-355
> Project: OODT
> Issue Type: Bug
> Components: file manager
> Affects Versions: 0.3
> Environment: none
> Reporter: Brian Foster
> Assignee: Brian Foster
> Priority: Minor
> Fix For: 0.4
>
>
> For example:
> {code}
> NominalDate:2011-01-20 Group:test
> {code}
> is treated as:
> {code}
> NominalDate=2011-01-20 AND Group=test
> {code}
> when really it should be:
> {code}
> NominalDate=2011-01-20 OR Group=test
> {code}
> From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
> Boolean Operators
> Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
> The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
> To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
> "jakarta apache" jakarta
> or
> "jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Resolved] (OODT-355) QueryTool treats Lucene list of term
queries as ANDed instead of ORed
Posted by "Brian Foster (Resolved) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/OODT-355?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Brian Foster resolved OODT-355.
-------------------------------
Resolution: Fixed
- QueryTool has been marked as @Deprecated in OODT-351 and LuceneQuery action now supports OR querys... here are some example queries which LuceneQuery supports:
{code}
ProductId:TestProductId ProductName:TestProductName
ProductId:TestProductId NominalDate:[20020101 TO 20030101]
ProductId:TestProductId NominalDate:{20020101 TO 20030101}
ProductId:TestProductId AND ProductName:TestProductName
ProductId:TestProductId OR ProductName:TestProductName
(ProductId:TestProductId OR ProductName:TestProductName) AND NominalDate:20110120
{code}
> QueryTool treats Lucene list of term queries as ANDed instead of ORed
> ---------------------------------------------------------------------
>
> Key: OODT-355
> URL: https://issues.apache.org/jira/browse/OODT-355
> Project: OODT
> Issue Type: Bug
> Components: file manager
> Affects Versions: 0.3
> Environment: none
> Reporter: Brian Foster
> Assignee: Brian Foster
> Priority: Minor
> Fix For: 0.4
>
>
> For example:
> {code}
> NominalDate:2011-01-20 Group:test
> {code}
> is treated as:
> {code}
> NominalDate=2011-01-20 AND Group=test
> {code}
> when really it should be:
> {code}
> NominalDate=2011-01-20 OR Group=test
> {code}
> From http://lucene.apache.org/java/2_9_1/queryparsersyntax.html:
> Boolean Operators
> Boolean operators allow terms to be combined through logic operators. Lucene supports AND, "+", OR, NOT and "-" as Boolean operators(Note: Boolean operators must be ALL CAPS).
> The OR operator is the default conjunction operator. This means that if there is no Boolean operator between two terms, the OR operator is used. The OR operator links two terms and finds a matching document if either of the terms exist in a document. This is equivalent to a union using sets. The symbol || can be used in place of the word OR.
> To search for documents that contain either "jakarta apache" or just "jakarta" use the query:
> "jakarta apache" jakarta
> or
> "jakarta apache" OR jakarta
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira