[ 
https://issues.apache.org/jira/browse/OPENJPA-3075?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18143800#comment-18143800
 ] 

ASF subversion and git services commented on OPENJPA-3075:
----------------------------------------------------------

Commit 04b97759f8e033bcd558cb71bf3cb90b6df894ab in openjpa's branch 
refs/heads/master from Richard Zowalla
[ https://gitbox.apache.org/repos/asf?p=openjpa.git;h=04b97759f ]

[OPENJPA-3076] Decide per dictionary whether reserved words are delimited, not 
JVM-wide (#304)

* [OPENJPA-3076] Add a test utility that records the names the mapping of a set 
of entities gets for a dictionary

* [OPENJPA-3076] Decide per dictionary whether a reserved word is delimited, 
rather than with the identifier rule all dictionaries of the JVM share

* [OPENJPA-3076] Test that persistence units for different databases in one JVM 
get the names they get alone, and stop resetting the shared identifier rule 
after the tests

* [OPENJPA-3076] Describe in the migration considerations and the release notes 
that the delimiting of reserved words is decided per dictionary

* [OPENJPA-3076] Do not delimit a reserved word in a name that is written 
without delimiters, such as the name the database metadata is asked for

* [OPENJPA-3076] Record the names the database metadata is asked for in the 
name dump

* [OPENJPA-3076] Name the sequences, columns and constraints of a schema file 
for the dictionary, as its schemas and tables

* [OPENJPA-3076] Remember the names a dictionary that delimits reserved words 
gave, and take the base name of a qualified name without copying it, as the 
names written into SQL are normalized again

* [OPENJPA-3076] Test how a dictionary normalizes qualified names, constants 
and column definitions

* [OPENJPA-3076] Name the configured schema, the default schema of the 
persistence unit and the given table of a mapping for the dictionary

* [OPENJPA-3076] Test that persistence units for different databases mapped by 
different threads at the same time get the names they get alone

* [OPENJPA-3076] Do not normalize the placeholder table name of a unique 
constraint, and describe how a naming utility without a dictionary normalizes

* [OPENJPA-3076] Use the renamed compatibility option UpdatedReservedWords

* [OPENJPA-3076] Expect the Derby names of OPENJPA-3075 and test every name 
dump setup next to all others

Since OPENJPA-3075 the Derby dictionary reserves WINDOW and renames a default
table name Derby rejects, so Window becomes WINDOW0 on Derby whichever
dictionary is configured first. The per dictionary tests now use RANK and
Partition, which MySQL reserves and Derby does not, for a name MySQL delimits
and Derby leaves alone, and expect WINDOW0 on Derby. A new test maps every
setup of the name dump after the dictionaries of all setups were configured,
in either order, and compares the names with those the setup gets alone.

---------

Co-authored-by: Maxim Solodovnik <[email protected]>

>  DerbyDictionary does not know WINDOW and other words Derby reserves
> --------------------------------------------------------------------
>
>                 Key: OPENJPA-3075
>                 URL: https://issues.apache.org/jira/browse/OPENJPA-3075
>             Project: OpenJPA
>          Issue Type: Bug
>            Reporter: Richard Zowalla
>            Assignee: Richard Zowalla
>            Priority: Major
>             Fix For: 4.2.0
>
>
> DerbyDictionary's reserved words are missing words that Derby rejects as 
> unquoted identifiers, such as WINDOW. An entity named Window therefore 
> produces CREATE TABLE Window, which fails on Derby. Add the missing words, 
> taken from Derby itself.



--
This message was sent by Atlassian Jira
(v10.3.26#10030026)

Reply via email to