[
https://issues.apache.org/jira/browse/SOLR-2876?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13144197#comment-13144197
]
Erick Erickson commented on SOLR-2876:
--------------------------------------
Dear God! You mean the intent here is:
return (pf.filter==null && pf.postFilter==null) ? qr.getDocSet() : null;
??? gotta confess I just saw the pattern and jumped to conclusions, hadn't
examined the code at all. But you've just made the job tougher... Wait, I know!
I'll just write something in awk to convert *everything* in the source tree.
The ternary operator really annoys me for this very reason... rant, rant,
rant.....
> Precedence operator in conditionals with ternary operator needs to be
> examined.
> -------------------------------------------------------------------------------
>
> Key: SOLR-2876
> URL: https://issues.apache.org/jira/browse/SOLR-2876
> Project: Solr
> Issue Type: Bug
> Affects Versions: 3.5, 4.0
> Environment: all
> Reporter: Erick Erickson
> Assignee: Erick Erickson
> Labels: operator, precedence, ternary
>
> This is an offshoot of 2829 where the root of the bug was that precedence in
> the ternary operator along with && without appropriate parentheses was a
> problem.
> && this.parser == null ? other.parser == null : this.parser.getClass() ==
> other.parser.getClass() (from ShortFieldSource.java).
> So that got me curious whether this pattern was repeated. A quick grep with
> the following REs produced one hit I wasn't related to 2829 with && and more
> with || (3x code base). I'll try to get to it over the weekend. Please don't
> grab it just yet, I'm fixing this partially for 2829, but if anyone wants to
> try the grep and see if I'm hallucinating, I'd appreciate it. I'd *really*
> appreciate any tests for things people see...
> Some of the returns are false hits, but not others. See
> SolrIndexSearcher.getDocListAndSetNC()
> the last line is: return pf.filter==null && pf.postFilter==null ?
> qr.getDocSet() : null;
> REs (using them in IntelliJ)
> \|\|[\sa-z\.0-9A-Z]+==.*\?
> &&[\sa-z\.0-9A-Z]+==.*\?
> I got some hits with the above and didn't pursue it any further, but if
> anyone wants to suggest more comprehensive REs, please attach. I'm trying for
> "&& or || followed by anything without an open parentheses followed by ==
> followed by anything followed by ?" I'd rather get a manageable number of
> false positives than miss things.
--
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
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
