Re: dblook usage message bug?

[myrnap <my...@Golux.Com>]

Ron Reuben wrote:

> Using Derby version 10.0.2.1 - (106978), when I execute
> /    java org.apache.derby.tools.dblook/
> instead of getting the usage message, I get
> /    CSLOOK_Usage
> /Looks like a bug.  Has anyone seen or reported this?  If not, I'll go 
> ahead and file it in Jira.
>
Ron, I tried it and see the same thing. I have not seen anyone log a bug.
Myrna

Re: [PATCH] dblook and dblook_test cleanup/enhancement

[Andrew McIntyre <fu...@nonintuitive.com>]

On Dec 6, 2004, at 11:47 AM, Army wrote:

> Army wrote:
>
>> I have a cleanup/enhancement patch for dblook_test that, among other 
>> things, will add a test case to verify the messages so that this kind 
>> of problem is caught by the harness in the future.  I hope to post 
>> the patch later today.
>
> Attached (and in-lined below) is a patch for dblook and dblook_test 
> that does the following:

<snip>

Committed revision 111185.

Unless there is any objection, I will merge this to the 10.0 branch 
tomorrow.

andrew

[PATCH] dblook and dblook_test cleanup/enhancement

[Army <ar...@golux.com>]

Army wrote:

> I have a cleanup/enhancement patch for dblook_test that, among other things, will add a test case to verify the 
> messages so that this kind of problem is caught by the harness in the future.  I hope to post the patch later today.
> 

Attached (and in-lined below) is a patch for dblook and dblook_test that does the following:

1) Adds new scenarios to dblook_test.java to test the dblook messages (per the "dblook usage message bug?" thread on 
derby-users list).

2) Changes dblook.java to use "DBJARS" as the directory for storing jar files, instead of "CSJARS".

3) Changes dblook_test so that, instead of passing an explicit database name to dblook, it lets dblook use the database 
name that it (dblook) parses out of the URL.  This allows us to verify that the URL parsing is returning the correct 
database name (which we don't currently test).

4) Cleans up error handling in both dblook.java and dblook_test.java, to allow better debugging of output in cases where 
dblook_test fails.

As part of this patch, two new files have been created:

java/testing/org/apache/derbyTesting/functionTests/tests/tools/dblook_test_sed.properties
java/testing/org/apache/derbyTesting/functionTests/tests/tools/dblook_makeDB_2.sql

I went in and changed the "svn:eol-style" property to "native" on these 2 files, and I noticed that it appears to be 
included as part of the patch--but of course, this should be confirmed by whoever submits...

As always, feel free to comment...
Army

Re: dblook usage message bug?

[Army <ar...@golux.com>]

myrnap wrote:
> myrnap wrote:
> 
> Looks like the message had changed to DBLOOK_Usage with revision 76095, 
> but the dblook tool only got updated with revision 109527.
> So, this has already been fixed...just not in the release candidate.
> 

Yes, this has been fixed with 109527.  I have a cleanup/enhancement patch for
dblook_test that, among other things, will add a test case to verify the messages
so that this kind of problem is caught by the harness in the future.  I
hope to post the patch later today.

Army

Re: dblook usage message bug?

[myrnap <my...@Golux.Com>]

myrnap wrote:

> Ron Reuben wrote:
>
>> Using Derby version 10.0.2.1 - (106978), when I execute
>> /    java org.apache.derby.tools.dblook/
>> instead of getting the usage message, I get
>> /    CSLOOK_Usage
>> /Looks like a bug.  Has anyone seen or reported this?  If not, I'll 
>> go ahead and file it in Jira.
>>
> Ron, I tried it and see the same thing. I have not seen anyone log a bug.
> Myrna
>
Looks like the message had changed to DBLOOK_Usage with revision 76095, 
but the dblook tool only got updated with revision 109527.
So, this has already been fixed...just not in the release candidate.

Myrna