table_function push() for NULL value doesn't work
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Drizzle |
Fix Released
|
High
|
Stewart Smith | ||
Cherry |
Fix Released
|
High
|
Stewart Smith | ||
Trunk |
Fix Released
|
High
|
Stewart Smith |
Bug Description
push() for a table_function column value of NULL doesn't work, instead you get the value of the last row.
embedded_
+++ /home/stewart/
@@ -76,7 +76,7 @@
open_files ULINT 300
read_io_threads ULINT 4
write_io_threads ULINT 4
-pre_rollback_hook CALLBACK NULL
+pre_rollback_hook CALLBACK 4
print_verbose_log BOOL true
rollback_
stats_sample_pages ULINT 8
@@ -86,7 +86,7 @@
version TEXT 1.0.6.6750
SELECT NAME,VALUE FROM DATA_DICTIONARY
NAME VALUE
-data_file_path NULL
+data_file_path true
data_home_dir ./
SELECT NAME,VALUE FROM DATA_DICTIONARY
NAME VALUE
(the failure is the current behaviour, the result file is what it *should* be).
This probably explains the timestamp default problem we saw the other day.
Related branches
- Brian Aker: Pending requested
- Drizzle Developers: Pending requested
-
Diff: 1271 lines (+351/-297) (has conflicts)23 files modifieddrizzled/plugin/table_function.cc (+20/-9)
drizzled/plugin/table_function.h (+2/-1)
plugin/collation_dictionary/tests/r/data_dictionary.result (+2/-2)
plugin/schema_dictionary/columns.cc (+7/-2)
plugin/schema_dictionary/tests/r/binary_default.result (+11/-0)
plugin/schema_dictionary/tests/r/data_dictionary.result (+1/-1)
plugin/schema_dictionary/tests/r/default_null.result (+5/-0)
plugin/schema_dictionary/tests/r/no_default.result (+7/-0)
plugin/schema_dictionary/tests/r/show.result (+33/-33)
plugin/schema_dictionary/tests/t/binary_default.test (+5/-0)
plugin/schema_dictionary/tests/t/default_null.test (+3/-0)
plugin/schema_dictionary/tests/t/no_default.test (+6/-0)
tests/r/alter_table.result (+2/-2)
tests/r/comment_column.result (+52/-52)
tests/r/comment_index.result (+103/-103)
tests/r/comment_table.result (+24/-24)
tests/r/create.result (+1/-1)
tests/r/data_dictionary_like_info.result (+16/-16)
tests/r/select.result (+12/-12)
tests/r/show_check.result (+3/-3)
tests/r/type_ranges.result (+27/-27)
tests/r/type_timestamp.result (+7/-7)
tests/t/data_dictionary_like_info.test (+2/-2)
Changed in drizzle: | |
assignee: | Brian Aker (brianaker) → Stewart Smith (stewart-flamingspork) |
milestone: | none → 2010-03-29 |
Hi!
Do you have the declaration you made for this? (aka in code for the
column)
Cheers,
-Brian
On Mar 14, 2010, at 10:53 PM, Stewart Smith wrote:
> ** Branch linked: lp:~stewart-flamingspork/drizzle/embedded-innodb- table-function table-function /bugs.launchpad .net/bugs/ 538930 innodb. config_ table_function [ fail ] --- /home/stewart/ embedded- innodb/ tests/. ./plugin/ embedded_ innodb/ tests/r/ table_function. result2010- 03-15 08:31:25.384813577 +0300 drizzle/ embedded- innodb/ tests/. ./plugin/ innodb/ tests/r/ config_ table_function. reject 2010-03-15 .INNODB_ CONFIGURATION WHERE .INNODB_ CONFIGURATION WHERE
> configuration-
>
> ** Branch unlinked: lp:~stewart-flamingspork/drizzle/embedded-innodb-
> configuration-
>
> --
> table_function push() for NULL value doesn't work
> https:/
> You received this bug notification because you are a member of
> Drizzle-
> developers, which is subscribed to Drizzle.
>
> Status in A Lightweight SQL Database for Cloud and Web: Confirmed
>
> Bug description:
> push() for a table_function column value of NULL doesn't work,
> instead you get the value of the last row.
>
>
> embedded_
> drizzle/
> config_
> +++ /home/stewart/
> embedded_
> 08:31:27.304812653 +0300
> @@ -76,7 +76,7 @@
> open_files ULINT 300
> read_io_threads ULINT 4
> write_io_threads ULINT 4
> -pre_rollback_hook CALLBACK NULL
> +pre_rollback_hook CALLBACK 4
> print_verbose_log BOOL true
> rollback_on_timeout BOOL true
> stats_sample_pages ULINT 8
> @@ -86,7 +86,7 @@
> version TEXT 1.0.6.6750
> SELECT NAME,VALUE FROM DATA_DICTIONARY
> NAME IN ("data_file_path", "data_home_dir");
> NAME VALUE
> -data_file_path NULL
> +data_file_path true
> data_home_dir ./
> SELECT NAME,VALUE FROM DATA_DICTIONARY
> NAME = "file_per_table";
> NAME VALUE
>
>
> (the failure is the current behaviour, the result file is what it
> *should* be).
>
> This probably explains the timestamp default problem we saw the
> other day.
>
>