[
https://issues.apache.org/jira/browse/THRIFT-105?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12680123#action_12680123
]
Alexander Shigin commented on THRIFT-105:
-----------------------------------------
Esteve, I've review your patch. I'm biassed, but I prefer my patch :).
I don't get the idea of thrift_limits. In the original version of fast binary
protocol the boundaries are checked by length of thrift_spec. I've choose the
same way. And your patch has unnecessary old check with length of thrift_spec.
The second issue with your patch is unnatural order of fields in thrift_spec.
The David's idea is briliant, but I prefer to duplicate default values in
thrift_spec and __init__ to increase readability. Still the THRIFT-105 +
THRIFT-339 patchs smaller than your patch.
> make a thrift_spec for a structures with negative tags
> ------------------------------------------------------
>
> Key: THRIFT-105
> URL: https://issues.apache.org/jira/browse/THRIFT-105
> Project: Thrift
> Issue Type: Improvement
> Components: Compiler (Python), Library (Python), Test Suite
> Reporter: Alexander Shigin
> Assignee: Alexander Shigin
> Attachments: thrift-python-offset-v2.patch,
> thrift-python-offset-v3.patch, thrift-python-offset.patch
>
>
> The patch introduces thrift_offset variable for a python thrift structures.
> The change allows to use fastbinary module for structures with negative tags.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
