-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/59252/#review174876
-----------------------------------------------------------



The comments that I made in the first few occurrences of the potential issues 
apply everywhere else where you use String.intern() or rewrite hashCode().


sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/provider/db/service/model/MPath.java
Line 30 (original), 30 (patched)
<https://reviews.apache.org/r/59252/#comment248100>

    Looks like path can be null, in which case you will get NPE here. When I 
worked on similar code in Hive, I ended up with a special static method String 
internIfNotNull(String s) in some common util class, to avoid writing the same 
line over and over again.



sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/provider/db/service/model/MPath.java
Line 41 (original), 41 (patched)
<https://reviews.apache.org/r/59252/#comment248101>

    I don't know whether 31 is a more sensible hashcode than 0 when path is 
null... maybe it just makes sense to keep this compatible with the old code?



sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/provider/db/service/model/MSentryGMPrivilege.java
Line 82 (original), 82 (patched)
<https://reviews.apache.org/r/59252/#comment248102>

    Probably it's safer to use the proposed internIfNotNull() everywhere to 
avoid any problems.



sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/provider/db/service/model/MSentryGroup.java
Line 86 (original), 82 (patched)
<https://reviews.apache.org/r/59252/#comment248103>

    Similar consideration about 0 vs 31.


- Misha Dmitriev


On May 13, 2017, 12:54 a.m., Alexander Kolbasov wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/59252/
> -----------------------------------------------------------
> 
> (Updated May 13, 2017, 12:54 a.m.)
> 
> 
> Review request for sentry, Brian Towles, Misha Dmitriev, Hao Hao, kalyan 
> kumar kalvagadda, Na Li, and Vamsee Yarlagadda.
> 
> 
> Bugs: SENTRY-1758
>     https://issues.apache.org/jira/browse/SENTRY-1758
> 
> 
> Repository: sentry
> 
> 
> Description
> -------
> 
> SENTRY-1758 Improve Sentry memory usage by interning object names
> 
> 
> Diffs
> -----
> 
>   
> sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/provider/db/service/model/MAuthzPathsMapping.java
>  c710701e06921c5ee85cda5c11021384aeb4053f 
>   
> sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/provider/db/service/model/MPath.java
>  c74384688ca920c79fb2987d225042e135cdfd53 
>   
> sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/provider/db/service/model/MSentryGMPrivilege.java
>  749af2a3e9d0420d3597952b0fcf13578f16a52f 
>   
> sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/provider/db/service/model/MSentryGroup.java
>  7e41c93294c64dd50b5672ca6fa3e003955ed365 
>   
> sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/provider/db/service/model/MSentryPrivilege.java
>  4c3af7992c90ba6ce33ff38ca6c5a3eb3492dd8b 
>   
> sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/provider/db/service/model/MSentryRole.java
>  f9da589d8de78bd4f8cb820bd7f545241000c417 
>   
> sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/provider/db/service/model/MSentryUser.java
>  ff5724999531fe4c1c6ccf8e206ee695749661c4 
>   
> sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/provider/db/service/model/MSentryVersion.java
>  ff8830f11a4f53801babbc0bb5deb386cfcbab42 
> 
> 
> Diff: https://reviews.apache.org/r/59252/diff/1/
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> Alexander Kolbasov
> 
>

Reply via email to